model : update label for LFM2-24B-A2B (#19848)

* model : Update label for LFM2-24B-A2B

```
❯ build/bin/llama-bench -m /data/playground/checkpoints/LFM2-24B-A2B-Preview-Q4_0.gguf,/data/playground/checkpoints/LFM2-8B-A1B-Q4_0.gguf -p 1 -n 0
| model                          |       size |     params | backend    | threads |            test |                  t/s |
| ------------------------------ | ---------: | ---------: | ---------- | ------: | --------------: | -------------------: |
| lfm2moe 24B.A2B Q4_0           |  12.54 GiB |    23.84 B | CPU        |      10 |             pp1 |         30.35 ± 2.49 |
| lfm2moe 8B.A1B Q4_0            |   4.41 GiB |     8.34 B | CPU        |      10 |             pp1 |         49.24 ± 1.93 |
```

* Remove extra line

.devops/rocm.Dockerfile

@@ -1,8 +1,8 @@
 ARG UBUNTU_VERSION=24.04
 
 # This needs to generally match the container host's environment.
-ARG ROCM_VERSION=7.0
-ARG AMDGPU_VERSION=7.0
+ARG ROCM_VERSION=7.2
+ARG AMDGPU_VERSION=7.2
 
 # Target the ROCm build image
 ARG BASE_ROCM_DEV_CONTAINER=rocm/dev-ubuntu-${UBUNTU_VERSION}:${ROCM_VERSION}-complete
@@ -11,13 +11,12 @@ ARG BASE_ROCM_DEV_CONTAINER=rocm/dev-ubuntu-${UBUNTU_VERSION}:${ROCM_VERSION}-co
 FROM ${BASE_ROCM_DEV_CONTAINER} AS build
 
 # Unless otherwise specified, we make a fat build.
-# List from https://github.com/ggml-org/llama.cpp/pull/1087#issuecomment-1682807878
 # This is mostly tied to rocBLAS supported archs.
-# gfx803, gfx900, gfx906, gfx1032, gfx1101, gfx1102,not officialy supported
-# check https://rocm.docs.amd.com/projects/install-on-linux/en/docs-6.4.1/reference/system-requirements.html
+# check https://rocm.docs.amd.com/projects/install-on-linux/en/docs-7.2.0/reference/system-requirements.html
+# check https://rocm.docs.amd.com/projects/radeon-ryzen/en/latest/docs/compatibility/compatibilityrad/native_linux/native_linux_compatibility.html
+# check https://rocm.docs.amd.com/projects/radeon-ryzen/en/latest/docs/compatibility/compatibilityryz/native_linux/native_linux_compatibility.html
 
-ARG ROCM_DOCKER_ARCH='gfx803;gfx900;gfx906;gfx908;gfx90a;gfx942;gfx1010;gfx1030;gfx1032;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201;gfx1151'
-#ARG ROCM_DOCKER_ARCH='gfx1151'
+ARG ROCM_DOCKER_ARCH='gfx908;gfx90a;gfx942;gfx1030;gfx1100;gfx1101;gfx1151;gfx1150;gfx1200;gfx1201'
 
 # Set ROCm architectures
 ENV AMDGPU_TARGETS=${ROCM_DOCKER_ARCH}

.github/workflows/release.yml

@@ -516,6 +516,102 @@ jobs:
           path: llama-bin-win-sycl-x64.zip
           name: llama-bin-win-sycl-x64.zip
 
+  ubuntu-22-rocm:
+    runs-on: ubuntu-22.04
+
+    strategy:
+      matrix:
+        include:
+          - ROCM_VERSION: "7.2"
+            gpu_targets: "gfx908;gfx90a;gfx942;gfx1030;gfx1100;gfx1101;gfx1151;gfx1150;gfx1200;gfx1201"
+            build: 'x64'
+
+    steps:
+      - name: Clone
+        id: checkout
+        uses: actions/checkout@v6
+        with:
+          fetch-depth: 0
+
+      - name: ccache
+        uses: ggml-org/ccache-action@v1.2.16
+        with:
+          key: ubuntu-rocm-cmake-${{ matrix.ROCM_VERSION }}-${{ matrix.build }}
+          evict-old-files: 1d
+
+      - name: Dependencies
+        id: depends
+        run: |
+          sudo apt install -y build-essential git cmake wget
+
+      - name: Setup Legacy ROCm
+        if: matrix.ROCM_VERSION == '7.2'
+        id: legacy_env
+        run: |
+          sudo mkdir --parents --mode=0755 /etc/apt/keyrings
+          wget https://repo.radeon.com/rocm/rocm.gpg.key -O - | \
+            gpg --dearmor | sudo tee /etc/apt/keyrings/rocm.gpg > /dev/null
+
+          sudo tee /etc/apt/sources.list.d/rocm.list << EOF
+          deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/${{ matrix.ROCM_VERSION }} jammy main
+          EOF
+
+          sudo tee /etc/apt/preferences.d/rocm-pin-600 << EOF
+          Package: *
+          Pin: release o=repo.radeon.com
+          Pin-Priority: 600
+          EOF
+
+          sudo apt update
+          sudo apt-get install -y libssl-dev rocm-hip-sdk
+
+      - name: Setup TheRock
+        if: matrix.ROCM_VERSION != '7.2'
+        id: therock_env
+        run: |
+          wget https://repo.amd.com/rocm/tarball/therock-dist-linux-gfx1151-${{ matrix.ROCM_VERSION }}.tar.gz
+          mkdir install
+          tar -xf *.tar.gz -C install
+          export ROCM_PATH=$(pwd)/install
+          echo ROCM_PATH=$ROCM_PATH >> $GITHUB_ENV
+          echo PATH=$PATH:$ROCM_PATH/bin >> $GITHUB_ENV
+          echo LD_LIBRARY_PATH=$ROCM_PATH/lib:$ROCM_PATH/llvm/lib:$ROCM_PATH/lib/rocprofiler-systems >> $GITHUB_ENV
+
+      - name: Build with native CMake HIP support
+        id: cmake_build
+        run: |
+          cmake -B build -S . \
+            -DCMAKE_HIP_COMPILER="$(hipconfig -l)/clang" \
+            -DCMAKE_HIP_FLAGS="-mllvm --amdgpu-unroll-threshold-local=600" \
+            -DCMAKE_BUILD_TYPE=Release \
+            -DGGML_BACKEND_DL=ON \
+            -DGGML_NATIVE=OFF \
+            -DCMAKE_INSTALL_RPATH='$ORIGIN' \
+            -DCMAKE_BUILD_WITH_INSTALL_RPATH=ON \
+            -DGGML_CPU_ALL_VARIANTS=ON \
+            -DGPU_TARGETS="${{ matrix.gpu_targets }}" \
+            -DGGML_HIP=ON \
+            -DHIP_PLATFORM=amd \
+            -DGGML_HIP_ROCWMMA_FATTN=ON \
+            ${{ env.CMAKE_ARGS }}
+          cmake --build build --config Release -j $(nproc)
+
+      - name: Determine tag name
+        id: tag
+        uses: ./.github/actions/get-tag-name
+
+      - name: Pack artifacts
+        id: pack_artifacts
+        run: |
+          cp LICENSE ./build/bin/
+          tar -czvf llama-${{ steps.tag.outputs.name }}-bin-ubuntu-rocm-${{ matrix.ROCM_VERSION }}-${{ matrix.build }}.tar.gz --transform "s,./,llama-${{ steps.tag.outputs.name }}/," -C ./build/bin .
+
+      - name: Upload artifacts
+        uses: actions/upload-artifact@v6
+        with:
+          path: llama-${{ steps.tag.outputs.name }}-bin-ubuntu-rocm-${{ matrix.ROCM_VERSION }}-${{ matrix.build }}.tar.gz
+          name: llama-bin-ubuntu-rocm-${{ matrix.ROCM_VERSION }}-${{ matrix.build }}.tar.gz
+
   windows-hip:
     runs-on: windows-2022
 
@@ -784,6 +880,7 @@ jobs:
       - windows-cuda
       - windows-sycl
       - windows-hip
+      - ubuntu-22-rocm
       - ubuntu-22-cpu
       - ubuntu-22-vulkan
       - macOS-arm64
@@ -868,6 +965,7 @@ jobs:
             **Linux:**
             - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/${{ steps.tag.outputs.name }}/llama-${{ steps.tag.outputs.name }}-bin-ubuntu-x64.tar.gz)
             - [Ubuntu x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/${{ steps.tag.outputs.name }}/llama-${{ steps.tag.outputs.name }}-bin-ubuntu-vulkan-x64.tar.gz)
+            - [Ubuntu x64 (ROCm 7.2)](https://github.com/ggml-org/llama.cpp/releases/download/${{ steps.tag.outputs.name }}/llama-${{ steps.tag.outputs.name }}-bin-ubuntu-rocm-7.2-x64.tar.gz)
             - [Ubuntu s390x (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/${{ steps.tag.outputs.name }}/llama-${{ steps.tag.outputs.name }}-bin-ubuntu-s390x.tar.gz)
 
             **Windows:**

CMakeLists.txt

@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 3.14) # for add_link_options and implicit target directories.
+cmake_minimum_required(VERSION 3.14...3.28) # for add_link_options and implicit target directories.
 project("llama.cpp" C CXX)
 include(CheckIncludeFileCXX)
 

common/chat-parser-xml-toolcall.cpp

@@ -803,7 +803,7 @@ inline void parse_msg_with_xml_tool_calls(common_chat_msg_parser & builder, cons
         }
 
         // remove potential partial suffix
-        if (builder.pos() == builder.input().size()) {
+        if (builder.pos() == builder.input().size() && builder.is_partial()) {
             if (unclosed_reasoning_content.empty()) {
                 rstrip(content);
                 trim_potential_partial_word(content);

common/chat-parser.cpp

@@ -893,23 +893,6 @@ static void common_chat_parse_minimax_m2(common_chat_msg_parser & builder) {
     builder.consume_reasoning_with_xml_tool_calls(form, "<think>", "</think>");
 }
 
-static void common_chat_parse_qwen3_coder_xml(common_chat_msg_parser & builder) {
-    static const xml_tool_call_format form = ([]() {
-        xml_tool_call_format form {};
-        form.scope_start = "<tool_call>";
-        form.tool_start  = "<function=";
-        form.tool_sep    = ">";
-        form.key_start   = "<parameter=";
-        form.key_val_sep = ">";
-        form.val_end     = "</parameter>";
-        form.tool_end    = "</function>";
-        form.scope_end   = "</tool_call>";
-        form.trim_raw_argval = true;
-        return form;
-    })();
-    builder.consume_reasoning_with_xml_tool_calls(form);
-}
-
 static void common_chat_parse_kimi_k2(common_chat_msg_parser & builder) {
     static const xml_tool_call_format form = ([]() {
         xml_tool_call_format form {};
@@ -1590,9 +1573,6 @@ static void common_chat_parse(common_chat_msg_parser & builder) {
         case COMMON_CHAT_FORMAT_KIMI_K2:
             common_chat_parse_kimi_k2(builder);
             break;
-        case COMMON_CHAT_FORMAT_QWEN3_CODER_XML:
-            common_chat_parse_qwen3_coder_xml(builder);
-            break;
         case COMMON_CHAT_FORMAT_APRIEL_1_5:
             common_chat_parse_apriel_1_5(builder);
             break;

common/chat.cpp

@@ -736,7 +736,6 @@ const char * common_chat_format_name(common_chat_format format) {
         case COMMON_CHAT_FORMAT_MINIMAX_M2: return "MiniMax-M2";
         case COMMON_CHAT_FORMAT_GLM_4_5: return "GLM 4.5";
         case COMMON_CHAT_FORMAT_KIMI_K2: return "Kimi K2";
-        case COMMON_CHAT_FORMAT_QWEN3_CODER_XML: return "Qwen3 Coder";
         case COMMON_CHAT_FORMAT_APRIEL_1_5: return "Apriel 1.5";
         case COMMON_CHAT_FORMAT_XIAOMI_MIMO: return "Xiaomi MiMo";
         case COMMON_CHAT_FORMAT_SOLAR_OPEN: return "Solar Open";
@@ -1522,14 +1521,17 @@ static common_chat_params common_chat_params_init_nemotron_v2(const common_chat_
     return data;
 }
 
-static common_chat_params common_chat_params_init_nemotron_v3(const common_chat_template & tmpl, const struct templates_params & inputs) {
+static common_chat_params common_chat_params_init_qwen3_coder(const common_chat_template & tmpl, const struct templates_params & inputs) {
     common_chat_params data;
 
     data.prompt = apply(tmpl, inputs);
     data.format = COMMON_CHAT_FORMAT_PEG_CONSTRUCTED;
 
+    // Nemotron Nano 3 and Step-3.5-Flash use the Qwen3 Coder tool calling with thinking
+    bool supports_reasoning = (tmpl.source().find("<think>") != std::string::npos);
+
     // Handle thinking tags appropriately based on inputs.enable_thinking
-    if (string_ends_with(data.prompt, "<think>\n")) {
+    if (supports_reasoning && string_ends_with(data.prompt, "<think>\n")) {
         if (!inputs.enable_thinking) {
             data.prompt += "</think>";
         } else {
@@ -1538,19 +1540,21 @@ static common_chat_params common_chat_params_init_nemotron_v3(const common_chat_
     }
 
     data.preserved_tokens = {
-        "<think>",
-        "</think>",
         "<tool_call>",
         "</tool_call>",
     };
 
+    if (supports_reasoning) {
+        data.preserved_tokens.insert(data.preserved_tokens.end(), {"<think>", "</think>"});
+    }
+
     auto has_tools = inputs.tools.is_array() && !inputs.tools.empty();
     auto extract_reasoning = inputs.reasoning_format != COMMON_REASONING_FORMAT_NONE;
     auto include_grammar = true;
 
     auto parser = build_chat_peg_constructed_parser([&](auto & p) {
         auto reasoning = p.eps();
-        if (inputs.enable_thinking && extract_reasoning) {
+        if (supports_reasoning && inputs.enable_thinking && extract_reasoning) {
             auto reasoning_content = p.reasoning(p.until("</think>")) + ("</think>" | p.end());
             if (data.thinking_forced_open) {
                 reasoning = reasoning_content;
@@ -1888,38 +1892,6 @@ static common_chat_params common_chat_params_init_minimax_m2(const common_chat_t
     return data;
 }
 
-static common_chat_params common_chat_params_init_qwen3_coder_xml(const common_chat_template & tmpl, const struct templates_params & params) {
-    common_chat_params data;
-    data.grammar_lazy = params.tools.is_array() && !params.tools.empty() && params.tool_choice != COMMON_CHAT_TOOL_CHOICE_REQUIRED;
-
-    data.prompt = apply(tmpl, params);
-    data.format = COMMON_CHAT_FORMAT_QWEN3_CODER_XML;
-
-    data.preserved_tokens = {
-        "<tool_call>",
-        "</tool_call>",
-        "<function=",
-        "</function>",
-        "<parameter=",
-        "</parameter>",
-    };
-
-    // build grammar for tool call
-    static const xml_tool_call_format form {
-        /* form.scope_start = */ "<tool_call>\n",
-        /* form.tool_start  = */ "<function=",
-        /* form.tool_sep    = */ ">\n",
-        /* form.key_start   = */ "<parameter=",
-        /* form.key_val_sep = */ ">\n",
-        /* form.val_end     = */ "\n</parameter>\n",
-        /* form.tool_end    = */ "</function>\n",
-        /* form.scope_end   = */ "</tool_call>",
-    };
-    build_grammar_xml_tool_call(data, params.tools, form);
-
-    return data;
-}
-
 static common_chat_params common_chat_params_init_kimi_k2(const common_chat_template & tmpl, const struct templates_params & params) {
     common_chat_params data;
     data.grammar_lazy = params.tools.is_array() && !params.tools.empty() && params.tool_choice != COMMON_CHAT_TOOL_CHOICE_REQUIRED;
@@ -2043,6 +2015,7 @@ static common_chat_params common_chat_params_init_gpt_oss(const common_chat_temp
         if (has_reasoning_content && has_tool_calls) {
             auto adjusted_message = msg;
             adjusted_message["thinking"] = msg.at("reasoning_content");
+            adjusted_message.erase("content");
             adjusted_messages.push_back(adjusted_message);
         } else {
             adjusted_messages.push_back(msg);
@@ -3140,19 +3113,13 @@ static common_chat_params common_chat_templates_apply_jinja(
     }
 
     // Qwen3-Coder XML format detection (must come before Hermes 2 Pro)
-    // Detect via explicit XML markers unique to Qwen3-Coder to avoid false positives in other templates.
-    // Require presence of <tool_call>, <function=...>, and <parameter=...> blocks.
+    // Detect via XML markers: <tool_call>, <function=...>, and <parameter=...> blocks.
+    // Also matches Step-3.5-Flash and Nemotron 3 Nano which use the same output format.
     if (src.find("<tool_call>") != std::string::npos &&
-        src.find("<function>") != std::string::npos &&
         src.find("<function=") != std::string::npos &&
-        src.find("<parameters>") != std::string::npos &&
         src.find("<parameter=") != std::string::npos) {
         workaround::func_args_not_string(params.messages);
-        // Nemotron 3 Nano 30B A3B
-        if (src.find("<think>") != std::string::npos) {
-            return common_chat_params_init_nemotron_v3(tmpl, params);
-        }
-        return common_chat_params_init_qwen3_coder_xml(tmpl, params);
+        return common_chat_params_init_qwen3_coder(tmpl, params);
     }
 
     // Xiaomi MiMo format detection (must come before Hermes 2 Pro)

common/chat.h

@@ -128,7 +128,6 @@ enum common_chat_format {
     COMMON_CHAT_FORMAT_GLM_4_5,
     COMMON_CHAT_FORMAT_MINIMAX_M2,
     COMMON_CHAT_FORMAT_KIMI_K2,
-    COMMON_CHAT_FORMAT_QWEN3_CODER_XML,
     COMMON_CHAT_FORMAT_APRIEL_1_5,
     COMMON_CHAT_FORMAT_XIAOMI_MIMO,
     COMMON_CHAT_FORMAT_SOLAR_OPEN,

common/common.cpp

@@ -1760,3 +1760,65 @@ float lr_opt::get_lr(float epoch) const {
     LOG_INF("epoch %.2g lr=%.2g\n", epoch, r);
     return r;
 }
+
+bool common_replay_last_token(struct llama_context * ctx, llama_token last_token, int32_t pos) {
+    llama_batch batch = llama_batch_get_one(&last_token, 1);
+    batch.pos = &pos;
+    if (llama_decode(ctx, batch)) {
+        LOG_ERR("%s: failed to replay last token\n", __func__);
+        return false;
+    }
+    return true;
+}
+
+bool common_prompt_batch_decode(
+              struct llama_context * ctx,
+    const std::vector<llama_token> & tokens,
+                               int & n_past,
+                               int   n_batch,
+                  std::string_view   state_path,
+                              bool   save_state) {
+    const int n_eval = tokens.size();
+    if (n_eval == 0) {
+        return true;
+    }
+
+    if (save_state && n_eval > 1) {
+        const int n_tokens_before_last = n_eval - 1;
+
+        GGML_ASSERT(n_eval <= n_batch);
+
+        // Decode all but the last token so we can save the memory state before decoding the last token.
+        // This is done so we can restore the session state later and replay the last token.
+        // Memory implementations in recurrent/hybrid models don't support removing tokens from their
+        // memory, so we can't just remove the last token from the memory and replay the last token which
+        // is the reason for this logic.
+        if (llama_decode(ctx, llama_batch_get_one(const_cast<llama_token*>(tokens.data()), n_tokens_before_last))) {
+            LOG_ERR("%s : failed to eval\n", __func__);
+            return false;
+        }
+        n_past += n_tokens_before_last;
+
+        llama_state_save_file(ctx, state_path.data(), tokens.data(), n_tokens_before_last);
+        LOG_INF("saved session before last token to %s, n_tokens = %d\n", state_path.data(), n_tokens_before_last);
+
+        llama_token last_token = tokens.back();
+        llama_batch batch = llama_batch_get_one(&last_token, 1);
+        int32_t pos = n_past;
+        batch.pos = &pos;
+
+        if (llama_decode(ctx, batch)) {
+            LOG_ERR("%s : failed to eval last token\n", __func__);
+            return false;
+        }
+        n_past++;
+    } else {
+        if (llama_decode(ctx, llama_batch_get_one(const_cast<llama_token*>(tokens.data()), n_eval))) {
+            LOG_ERR("%s : failed to eval\n", __func__);
+            return false;
+        }
+        n_past += n_eval;
+    }
+
+    return true;
+}

common/common.h

@@ -804,6 +804,23 @@ void common_batch_add(
     const std::vector<llama_seq_id> & seq_ids,
                                bool   logits);
 
+// decodes a single batch of tokens for a prompt and manages session tokens
+//
+// Note: We save state before the last token so that we can replay it to ensure
+// compatibility with all memory types. Recurrent/hybrid models cannot remove
+// tokens from memory, so this approach works across all model architectures.
+bool common_prompt_batch_decode(
+              struct llama_context * ctx,
+    const std::vector<llama_token> & embd,
+                               int & n_past,
+                               int   n_batch,
+                  std::string_view   state_path,
+                              bool   save_state);
+
+// replays the last token after loading state to regenerate logits
+// used after loading session state to ensure the sampling context has valid logits
+bool common_replay_last_token(struct llama_context * ctx, llama_token last_token, int32_t pos);
+
 //
 // Vocab utils
 //

common/jinja/runtime.cpp

@@ -85,7 +85,7 @@ value identifier::execute_impl(context & ctx) {
     auto builtins = global_builtins();
     if (!it->is_undefined()) {
         if (ctx.is_get_stats) {
-            it->stats.used = true;
+            value_t::stats_t::mark_used(it);
         }
         JJ_DEBUG("Identifier '%s' found, type = %s", val.c_str(), it->type().c_str());
         return it;
@@ -277,7 +277,7 @@ value binary_expression::execute_impl(context & ctx) {
 static value try_builtin_func(context & ctx, const std::string & name, value & input, bool undef_on_missing = false) {
     JJ_DEBUG("Trying built-in function '%s' for type %s", name.c_str(), input->type().c_str());
     if (ctx.is_get_stats) {
-        input->stats.used = true;
+        value_t::stats_t::mark_used(input);
         input->stats.ops.insert(name);
     }
     auto builtins = input->get_builtins();
@@ -448,7 +448,7 @@ value for_statement::execute_impl(context & ctx) {
 
     // mark the variable being iterated as used for stats
     if (ctx.is_get_stats) {
-        iterable_val->stats.used = true;
+        value_t::stats_t::mark_used(iterable_val);
         iterable_val->stats.ops.insert("array_access");
     }
 
@@ -470,7 +470,7 @@ value for_statement::execute_impl(context & ctx) {
             items.push_back(std::move(tuple));
         }
         if (ctx.is_get_stats) {
-            iterable_val->stats.used = true;
+            value_t::stats_t::mark_used(iterable_val);
             iterable_val->stats.ops.insert("object_access");
         }
     } else {
@@ -480,7 +480,7 @@ value for_statement::execute_impl(context & ctx) {
             items.push_back(item);
         }
         if (ctx.is_get_stats) {
-            iterable_val->stats.used = true;
+            value_t::stats_t::mark_used(iterable_val);
             iterable_val->stats.ops.insert("array_access");
         }
     }
@@ -817,8 +817,9 @@ value member_expression::execute_impl(context & ctx) {
     }
 
     if (ctx.is_get_stats && val && object && property) {
-        val->stats.used = true;
-        object->stats.used = true;
+        value_t::stats_t::mark_used(val);
+        value_t::stats_t::mark_used(object);
+        value_t::stats_t::mark_used(property);
         if (is_val<value_int>(property)) {
             object->stats.ops.insert("array_access");
         } else if (is_val<value_string>(property)) {

common/jinja/value.cpp

@@ -161,6 +161,11 @@ static value tojson(const func_args & args) {
     value val_separators = args.get_kwarg_or_pos("separators",   3);
     value val_sort       = args.get_kwarg_or_pos("sort_keys",    4);
     int indent = -1;
+    if (args.ctx.is_get_stats) {
+        // mark as used (recursively) for stats
+        auto val_input = args.get_pos(0);
+        value_t::stats_t::mark_used(const_cast<value&>(val_input), true);
+    }
     if (is_val<value_int>(val_indent)) {
         indent = static_cast<int>(val_indent->as_int());
     }
@@ -891,6 +896,11 @@ const func_builtins & value_array_t::get_builtins() const {
         }},
         {"string", [](const func_args & args) -> value {
             args.ensure_vals<value_array>();
+            if (args.ctx.is_get_stats) {
+                // mark as used (recursively) for stats
+                auto val_input = args.get_pos(0);
+                value_t::stats_t::mark_used(const_cast<value&>(val_input), true);
+            }
             return mk_val<value_string>(args.get_pos(0)->as_string());
         }},
         {"tojson", tojson},
@@ -1046,6 +1056,11 @@ const func_builtins & value_object_t::get_builtins() const {
         {"tojson", tojson},
         {"string", [](const func_args & args) -> value {
             args.ensure_vals<value_object>();
+            if (args.ctx.is_get_stats) {
+                // mark as used (recursively) for stats
+                auto val_input = args.get_pos(0);
+                value_t::stats_t::mark_used(const_cast<value&>(val_input), true);
+            }
             return mk_val<value_string>(args.get_pos(0)->as_string());
         }},
         {"length", [](const func_args & args) -> value {
@@ -1358,4 +1373,21 @@ std::string value_to_string_repr(const value & val) {
     }
 }
 
+// stats utility
+void value_t::stats_t::mark_used(value & val, bool deep) {
+    val->stats.used = true;
+    if (deep) {
+        if (is_val<value_array>(val)) {
+            for (auto & item : val->val_arr) {
+                mark_used(item, deep);
+            }
+        } else if (is_val<value_object>(val)) {
+            for (auto & pair : val->val_obj) {
+                mark_used(pair.first, deep);
+                mark_used(pair.second, deep);
+            }
+        }
+    }
+}
+
 } // namespace jinja

common/jinja/value.h

@@ -118,6 +118,8 @@ struct value_t {
         bool used = false;
         // ops can be builtin calls or operators: "array_access", "object_access"
         std::set<std::string> ops;
+        // utility to recursively mark value and its children as used
+        static void mark_used(value & val, bool deep = false);
     } stats;
 
     value_t() = default;

convert_hf_to_gguf.py

@@ -1274,6 +1274,9 @@ class TextModel(ModelBase):
         if chkhsh == "b4b8ca1f9769494fbd956ebc4c249de6131fb277a4a3345a7a92c7dd7a55808d":
             # ref: https://huggingface.co/jdopensource/JoyAI-LLM-Flash
             res = "joyai-llm"
+        if chkhsh == "e4d54df1ebc1f2b91acd986c5b51aa50837d5faf7c7398e73c1f9e9ee5d19869":
+            # ref: https://huggingface.co/kakaocorp/kanana-2-30b-a3b-instruct-2601
+            res = "kanana2"
 
         if res is None:
             logger.warning("\n")
@@ -3733,6 +3736,13 @@ class Ernie4_5Model(TextModel):
     def set_vocab(self):
         self._set_vocab_sentencepiece()
 
+        tokenizer_config_file = self.dir_model / 'tokenizer_config.json'
+        if tokenizer_config_file.is_file():
+            with open(tokenizer_config_file, "r", encoding="utf-8") as f:
+                tokenizer_config_json = json.load(f)
+                if "add_prefix_space" in tokenizer_config_json:
+                    self.gguf_writer.add_add_space_prefix(tokenizer_config_json["add_prefix_space"])
+
     def set_gguf_parameters(self):
         super().set_gguf_parameters()
 
@@ -3742,6 +3752,10 @@ class Ernie4_5Model(TextModel):
         if (head_dim := self.hparams.get("head_dim")) is None:
             head_dim = self.hparams["hidden_size"] // num_heads
 
+        if "mlp_AR" in name or "vision_model" in name:
+            # skip vision model and projector tensors
+            return
+
         if "ernie." in name:
             name = name.replace("ernie.", "model.")
         # split the qkv weights
@@ -3851,6 +3865,48 @@ class Ernie4_5MoeModel(Ernie4_5Model):
                 raise ValueError(f"Unprocessed experts: {experts}")
 
 
+@ModelBase.register("PaddleOCRVLForConditionalGeneration")
+class PaddleOCRModel(Ernie4_5Model):
+    model_arch = gguf.MODEL_ARCH.PADDLEOCR
+
+
+@ModelBase.register("PaddleOCRVisionModel")
+class PaddleOCRVisionModel(MmprojModel):
+    # PaddleOCR-VL uses a modified version of Siglip
+    min_pixels: int = 0
+    max_pixels: int = 0
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        assert self.hparams_vision is not None
+        self.min_pixels = self.preprocessor_config["min_pixels"]
+        self.max_pixels = self.preprocessor_config["max_pixels"]
+        self.hparams_vision["image_size"] = int(math.sqrt(self.max_pixels))
+
+    def set_gguf_parameters(self):
+        super().set_gguf_parameters()
+        assert self.hparams_vision is not None
+        hparams = self.hparams_vision
+        self.gguf_writer.add_clip_projector_type(gguf.VisionProjectorType.PADDLEOCR)
+        self.gguf_writer.add_vision_max_pixels(self.max_pixels)
+        self.gguf_writer.add_vision_min_pixels(self.min_pixels)
+        self.gguf_writer.add_vision_use_gelu(True)
+        self.gguf_writer.add_vision_attention_layernorm_eps(hparams.get("rms_norm_eps", 1e-6))
+
+    def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
+        name = name.replace("visual.", "model.")
+
+        if "vision_model" in name or "mlp_AR" in name:
+            if "packing_position_embedding" in name:
+                return # unused
+            elif "vision_model.head" in name:
+                # we don't yet support image embeddings for this model
+                return
+            else:
+                yield from super().modify_tensors(data_torch, name, bid)
+        return # skip other tensors
+
+
 @ModelBase.register(
     "Qwen2VLModel",
     "Qwen2VLForConditionalGeneration",

convert_hf_to_gguf_update.py

@@ -152,6 +152,7 @@ models = [
     {"name": "exaone-moe",       "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/LGAI-EXAONE/K-EXAONE-236B-A23B", },
     {"name": "qwen35",           "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/Qwen/Qwen3.5-9B-Instruct", },
     {"name": "joyai-llm",        "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/jdopensource/JoyAI-LLM-Flash", },
+    {"name": "kanana2",          "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/kakaocorp/kanana-2-30b-a3b-instruct-2601", },
 ]
 
 # some models are known to be broken upstream, so we will skip them as exceptions

docs/ops.md

@@ -31,7 +31,7 @@ Legend:
 |                          CONV_3D | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
 |                CONV_TRANSPOSE_1D | ❌ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ✅ | ❌ | ❌ | ❌ |
 |                CONV_TRANSPOSE_2D | ❌ | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ |
-|                              COS | ❌ | ✅ | ✅ | ✅ | 🟡 | ❌ | ✅ | 🟡 | ❌ | ❌ | ❌ |
+|                              COS | ❌ | ✅ | ✅ | ✅ | 🟡 | ❌ | ✅ | 🟡 | ✅ | ❌ | ❌ |
 |                      COUNT_EQUAL | ❌ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ✅ | ❌ | ❌ | ❌ |
 |                              CPY | ❌ | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | ❌ | ❌ |
 |               CROSS_ENTROPY_LOSS | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
@@ -96,13 +96,13 @@ Legend:
 |                          SIGMOID | ❌ | ✅ | ✅ | 🟡 | 🟡 | 🟡 | ✅ | 🟡 | ✅ | ❌ | ❌ |
 |                             SILU | ❌ | ✅ | ✅ | 🟡 | 🟡 | 🟡 | ✅ | 🟡 | ✅ | ❌ | ❌ |
 |                        SILU_BACK | ❌ | ❌ | ✅ | ✅ | ❌ | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ |
-|                              SIN | ❌ | ✅ | ✅ | ✅ | 🟡 | ❌ | ✅ | 🟡 | ❌ | ❌ | ❌ |
+|                              SIN | ❌ | ✅ | ✅ | ✅ | 🟡 | ❌ | ✅ | 🟡 | ✅ | ❌ | ❌ |
 |                         SOFTPLUS | ❌ | ❌ | ✅ | 🟡 | 🟡 | ❌ | ✅ | 🟡 | ✅ | ❌ | ❌ |
 |                         SOFT_MAX | ❌ | 🟡 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ❌ |
 |                    SOFT_MAX_BACK | ❌ | ❌ | 🟡 | 🟡 | ❌ | ❌ | 🟡 | ✅ | ❌ | ❌ | ❌ |
 |                        SOLVE_TRI | ❌ | ❌ | ✅ | 🟡 | ❌ | ❌ | ❌ | 🟡 | ❌ | ❌ | ❌ |
-|                              SQR | ❌ | ✅ | ✅ | ✅ | 🟡 | ✅ | ✅ | 🟡 | ❌ | ❌ | ❌ |
-|                             SQRT | ❌ | ✅ | ✅ | ✅ | 🟡 | ✅ | ✅ | 🟡 | ❌ | ❌ | ❌ |
+|                              SQR | ❌ | ✅ | ✅ | ✅ | 🟡 | ✅ | ✅ | 🟡 | ✅ | ❌ | ❌ |
+|                             SQRT | ❌ | ✅ | ✅ | ✅ | 🟡 | ✅ | ✅ | 🟡 | ✅ | ❌ | ❌ |
 |                         SSM_CONV | ❌ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
 |                         SSM_SCAN | ❌ | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ | 🟡 | ❌ | ❌ | ❌ |
 |                             STEP | ❌ | ✅ | ✅ | 🟡 | 🟡 | ❌ | ✅ | 🟡 | ✅ | ❌ | ❌ |

docs/ops/WebGPU.csv

@@ -8760,22 +8760,14 @@
 "WebGPU: WebGPU","ADD_ID","type_a=f32,type_b=f32,n_embd=129,n_experts=8,n_experts_used=4,n_token=1","support","0","no","WebGPU"
 "WebGPU: WebGPU","ADD_ID","type_a=f32,type_b=f32,n_embd=129,n_experts=8,n_experts_used=4,n_token=32","support","0","no","WebGPU"
 "WebGPU: WebGPU","ADD_ID","type_a=f32,type_b=f32,n_embd=129,n_experts=8,n_experts_used=4,n_token=129","support","0","no","WebGPU"
-"WebGPU: WebGPU","SQR","type=f16,ne=[10,5,4,3]","support","0","no","WebGPU"
-"WebGPU: WebGPU","SQRT","type=f16,ne=[10,3,3,2]","support","0","no","WebGPU"
 "WebGPU: WebGPU","LOG","type=f16,ne=[10,5,4,3]","support","1","yes","WebGPU"
-"WebGPU: WebGPU","SIN","type=f16,ne=[10,2,2,2]","support","0","no","WebGPU"
-"WebGPU: WebGPU","COS","type=f16,ne=[10,2,2,2]","support","0","no","WebGPU"
 "WebGPU: WebGPU","CLAMP","type=f16,ne=[10,5,4,3],min=-0.500000,max=0.500000","support","1","yes","WebGPU"
 "WebGPU: WebGPU","LEAKY_RELU","type=f16,ne_a=[10,5,4,3],negative_slope=0.100000","support","0","no","WebGPU"
 "WebGPU: WebGPU","FLOOR","type=f16,ne=[10,2,2,2]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","CEIL","type=f16,ne=[10,2,2,2]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","ROUND","type=f16,ne=[10,2,2,2]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","TRUNC","type=f16,ne=[10,2,2,2]","support","1","yes","WebGPU"
-"WebGPU: WebGPU","SQR","type=f16,ne=[7,1,5,3]","support","0","no","WebGPU"
-"WebGPU: WebGPU","SQRT","type=f16,ne=[7,1,5,3]","support","0","no","WebGPU"
 "WebGPU: WebGPU","LOG","type=f16,ne=[7,1,5,3]","support","1","yes","WebGPU"
-"WebGPU: WebGPU","SIN","type=f16,ne=[7,1,5,3]","support","0","no","WebGPU"
-"WebGPU: WebGPU","COS","type=f16,ne=[7,1,5,3]","support","0","no","WebGPU"
 "WebGPU: WebGPU","CLAMP","type=f16,ne=[7,1,5,3],min=-0.500000,max=0.500000","support","1","yes","WebGPU"
 "WebGPU: WebGPU","LEAKY_RELU","type=f16,ne_a=[7,1,5,3],negative_slope=0.100000","support","0","no","WebGPU"
 "WebGPU: WebGPU","FLOOR","type=f16,ne=[7,1,5,3]","support","1","yes","WebGPU"
@@ -8786,22 +8778,14 @@
 "WebGPU: WebGPU","ROUND","type=f16,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","TRUNC","type=f16,ne=[7,1,5,3]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","TRUNC","type=f16,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
-"WebGPU: WebGPU","SQR","type=f32,ne=[10,5,4,3]","support","0","no","WebGPU"
-"WebGPU: WebGPU","SQRT","type=f32,ne=[10,3,3,2]","support","0","no","WebGPU"
 "WebGPU: WebGPU","LOG","type=f32,ne=[10,5,4,3]","support","1","yes","WebGPU"
-"WebGPU: WebGPU","SIN","type=f32,ne=[10,2,2,2]","support","0","no","WebGPU"
-"WebGPU: WebGPU","COS","type=f32,ne=[10,2,2,2]","support","0","no","WebGPU"
 "WebGPU: WebGPU","CLAMP","type=f32,ne=[10,5,4,3],min=-0.500000,max=0.500000","support","1","yes","WebGPU"
 "WebGPU: WebGPU","LEAKY_RELU","type=f32,ne_a=[10,5,4,3],negative_slope=0.100000","support","0","no","WebGPU"
 "WebGPU: WebGPU","FLOOR","type=f32,ne=[10,2,2,2]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","CEIL","type=f32,ne=[10,2,2,2]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","ROUND","type=f32,ne=[10,2,2,2]","support","1","yes","WebGPU"
 "WebGPU: WebGPU","TRUNC","type=f32,ne=[10,2,2,2]","support","1","yes","WebGPU"
-"WebGPU: WebGPU","SQR","type=f32,ne=[7,1,5,3]","support","0","no","WebGPU"
-"WebGPU: WebGPU","SQRT","type=f32,ne=[7,1,5,3]","support","0","no","WebGPU"
 "WebGPU: WebGPU","LOG","type=f32,ne=[7,1,5,3]","support","1","yes","WebGPU"
-"WebGPU: WebGPU","SIN","type=f32,ne=[7,1,5,3]","support","0","no","WebGPU"
-"WebGPU: WebGPU","COS","type=f32,ne=[7,1,5,3]","support","0","no","WebGPU"
 "WebGPU: WebGPU","CLAMP","type=f32,ne=[7,1,5,3],min=-0.500000,max=0.500000","support","1","yes","WebGPU"
 "WebGPU: WebGPU","LEAKY_RELU","type=f32,ne_a=[7,1,5,3],negative_slope=0.100000","support","0","no","WebGPU"
 "WebGPU: WebGPU","FLOOR","type=f32,ne=[7,1,5,3]","support","1","yes","WebGPU"
@@ -18901,3 +18885,27 @@
 "WebGPU: WebGPU","CROSS_ENTROPY_LOSS_BACK","type=f32,ne=[30000,1,1,1]","support","0","no","WebGPU"
 "WebGPU: WebGPU","OPT_STEP_ADAMW","type=f32,ne=[10,5,4,3]","support","0","no","WebGPU"
 "WebGPU: WebGPU","OPT_STEP_SGD","type=f32,ne=[10,5,4,3]","support","0","no","WebGPU"
+"WebGPU: WebGPU","SQR","type=f16,ne=[10,5,4,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQRT","type=f16,ne=[10,3,3,2]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SIN","type=f16,ne=[10,2,2,2]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","COS","type=f16,ne=[10,2,2,2]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQR","type=f16,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQR","type=f16,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQRT","type=f16,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQRT","type=f16,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SIN","type=f16,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SIN","type=f16,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","COS","type=f16,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","COS","type=f16,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQR","type=f32,ne=[10,5,4,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQRT","type=f32,ne=[10,3,3,2]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SIN","type=f32,ne=[10,2,2,2]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","COS","type=f32,ne=[10,2,2,2]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQR","type=f32,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQR","type=f32,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQRT","type=f32,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SQRT","type=f32,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SIN","type=f32,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","SIN","type=f32,ne=[1024,1024,1,1]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","COS","type=f32,ne=[7,1,5,3]","support","1","yes","WebGPU"
+"WebGPU: WebGPU","COS","type=f32,ne=[1024,1024,1,1]","support","1","yes","WebGPU"

examples/model-conversion/Makefile

@@ -77,7 +77,10 @@ causal-verify-embeddings: causal-run-original-embeddings causal-run-converted-em
 	@./scripts/causal/compare-embeddings-logits.sh
 
 causal-inspect-original-model:
-	@./scripts/utils/inspect-org-model.py
+	@./scripts/utils/inspect-org-model.py --list-all -s
+
+causal-list-original-model-tensors:
+	@./scripts/utils/inspect-org-model.py --list-all-short -s
 
 causal-inspect-converted-model:
 	@./scripts/utils/inspect-converted-model.sh
@@ -153,7 +156,7 @@ embedding-verify-logits-st: embedding-run-original-model-st embedding-run-conver
 
 embedding-inspect-original-model:
 	$(call validate_embedding_model_path,embedding-inspect-original-model)
-	@EMBEDDING_MODEL_PATH="$(EMBEDDING_MODEL_PATH)" ./scripts/utils/inspect-org-model.py -m ${EMBEDDING_MODEL_PATH}
+	@EMBEDDING_MODEL_PATH="$(EMBEDDING_MODEL_PATH)" ./scripts/utils/inspect-org-model.py -m ${EMBEDDING_MODEL_PATH} --list-all -s
 
 embedding-inspect-converted-model:
 	@CONVERTED_EMBEDDING_MODEL="$(CONVERTED_EMBEDDING_MODEL)" ./scripts/utils/inspect-converted-model.sh ${CONVERTED_EMBEDDING_MODEL}

examples/model-conversion/scripts/utils/inspect-org-model.py

@@ -1,67 +1,290 @@
 #!/usr/bin/env python3
 
 import argparse
-import os
 import json
+import os
+import re
+import struct
+import sys
+from pathlib import Path
+from typing import Optional
 from safetensors import safe_open
-from collections import defaultdict
 
-parser = argparse.ArgumentParser(description='Process model with specified path')
-parser.add_argument('--model-path', '-m', help='Path to the model')
-args = parser.parse_args()
 
-model_path = os.environ.get('MODEL_PATH', args.model_path)
-if model_path is None:
-    parser.error("Model path must be specified either via --model-path argument or MODEL_PATH environment variable")
+MODEL_SAFETENSORS_FILE = "model.safetensors"
+MODEL_SAFETENSORS_INDEX = "model.safetensors.index.json"
 
-# Check if there's an index file (multi-file model)
-index_path = os.path.join(model_path, "model.safetensors.index.json")
-single_file_path = os.path.join(model_path, "model.safetensors")
+DTYPE_SIZES = {
+    "F64": 8, "I64": 8, "U64": 8,
+    "F32": 4, "I32": 4, "U32": 4,
+    "F16": 2, "BF16": 2, "I16": 2, "U16": 2,
+    "I8": 1, "U8": 1, "BOOL": 1,
+    "F8_E4M3": 1, "F8_E5M2": 1,
+}
 
-if os.path.exists(index_path):
-    # Multi-file model
-    print("Multi-file model detected")
+SIZE_UNITS = ['B', 'KB', 'MB', 'GB', 'TB']
 
-    with open(index_path, 'r') as f:
-        index_data = json.load(f)
 
-    # Get the weight map (tensor_name -> file_name)
-    weight_map = index_data.get("weight_map", {})
+def get_weight_map(model_path: Path) -> Optional[dict[str, str]]:
+    index_file = model_path / MODEL_SAFETENSORS_INDEX
 
-    # Group tensors by file for efficient processing
-    file_tensors = defaultdict(list)
-    for tensor_name, file_name in weight_map.items():
-        file_tensors[file_name].append(tensor_name)
+    if index_file.exists():
+        with open(index_file, 'r') as f:
+            index = json.load(f)
+            return index.get("weight_map", {})
 
-    print("Tensors in model:")
+    return None
 
-    # Process each shard file
-    for file_name, tensor_names in file_tensors.items():
-        file_path = os.path.join(model_path, file_name)
-        print(f"\n--- From {file_name} ---")
 
-        with safe_open(file_path, framework="pt") as f:
-            for tensor_name in sorted(tensor_names):
-                tensor = f.get_tensor(tensor_name)
-                print(f"- {tensor_name} : shape = {tensor.shape}, dtype = {tensor.dtype}")
+def get_all_tensor_names(model_path: Path) -> list[str]:
+    weight_map = get_weight_map(model_path)
 
-elif os.path.exists(single_file_path):
-    # Single file model (original behavior)
-    print("Single-file model detected")
+    if weight_map is not None:
+        return list(weight_map.keys())
 
-    with safe_open(single_file_path, framework="pt") as f:
-        keys = f.keys()
-        print("Tensors in model:")
-        for key in sorted(keys):
-            tensor = f.get_tensor(key)
-            print(f"- {key} : shape = {tensor.shape}, dtype = {tensor.dtype}")
+    single_file = model_path / MODEL_SAFETENSORS_FILE
+    if single_file.exists():
+        try:
+            with safe_open(single_file, framework="pt", device="cpu") as f:
+                return list(f.keys())
+        except Exception as e:
+            print(f"Error reading {single_file}: {e}")
+            sys.exit(1)
 
-else:
-    print(f"Error: Neither 'model.safetensors.index.json' nor 'model.safetensors' found in {model_path}")
-    print("Available files:")
-    if os.path.exists(model_path):
-        for item in sorted(os.listdir(model_path)):
-            print(f"  {item}")
+    print(f"Error: No safetensors files found in {model_path}")
+    sys.exit(1)
+
+
+def find_tensor_file(model_path: Path, tensor_name: str) -> Optional[str]:
+    weight_map = get_weight_map(model_path)
+
+    if weight_map is not None:
+        return weight_map.get(tensor_name)
+
+    single_file = model_path / MODEL_SAFETENSORS_FILE
+    if single_file.exists():
+        return single_file.name
+
+    return None
+
+
+def read_safetensors_header(file_path: Path) -> dict:
+    with open(file_path, 'rb') as f:
+        header_size = struct.unpack('<Q', f.read(8))[0]
+        return json.loads(f.read(header_size))
+
+
+def get_tensor_size_bytes(tensor_meta: dict) -> int:
+    offsets = tensor_meta.get("data_offsets")
+    if offsets and len(offsets) == 2:
+        return offsets[1] - offsets[0]
+    n_elements = 1
+    for d in tensor_meta.get("shape", []):
+        n_elements *= d
+    return n_elements * DTYPE_SIZES.get(tensor_meta.get("dtype", "F32"), 4)
+
+
+def format_size(size_bytes: int) -> str:
+    val = float(size_bytes)
+    for unit in SIZE_UNITS[:-1]:
+        if val < 1024.0:
+            return f"{val:.2f} {unit}"
+        val /= 1024.0
+    return f"{val:.2f} {SIZE_UNITS[-1]}"
+
+
+def get_all_tensor_metadata(model_path: Path) -> dict[str, dict]:
+    weight_map = get_weight_map(model_path)
+
+    if weight_map is not None:
+        file_to_tensors: dict[str, list[str]] = {}
+        for tensor_name, file_name in weight_map.items():
+            file_to_tensors.setdefault(file_name, []).append(tensor_name)
+
+        all_metadata: dict[str, dict] = {}
+        for file_name, tensor_names in file_to_tensors.items():
+            try:
+                header = read_safetensors_header(model_path / file_name)
+                for tensor_name in tensor_names:
+                    if tensor_name in header:
+                        all_metadata[tensor_name] = header[tensor_name]
+            except Exception as e:
+                print(f"Warning: Could not read header from {file_name}: {e}", file=sys.stderr)
+        return all_metadata
+
+    single_file = model_path / MODEL_SAFETENSORS_FILE
+    if single_file.exists():
+        try:
+            header = read_safetensors_header(single_file)
+            return {k: v for k, v in header.items() if k != "__metadata__"}
+        except Exception as e:
+            print(f"Error reading {single_file}: {e}")
+            sys.exit(1)
+
+    print(f"Error: No safetensors files found in {model_path}")
+    sys.exit(1)
+
+
+def normalize_tensor_name(tensor_name: str) -> str:
+    normalized = re.sub(r'\.\d+\.', '.#.', tensor_name)
+    normalized = re.sub(r'\.\d+$', '.#', normalized)
+    return normalized
+
+
+def list_all_tensors(
+    model_path: Path,
+    short: bool = False,
+    show_sizes: bool = False,
+):
+    tensor_names = get_all_tensor_names(model_path)
+
+    metadata: Optional[dict[str, dict]] = None
+    if show_sizes:
+        metadata = get_all_tensor_metadata(model_path)
+
+    total_bytes = 0
+
+    if short:
+        seen: dict[str, str] = {}
+        for tensor_name in sorted(tensor_names):
+            normalized = normalize_tensor_name(tensor_name)
+            if normalized not in seen:
+                seen[normalized] = tensor_name
+        display_pairs = list(sorted(seen.items()))
+        name_width = max((len(n) for n, _ in display_pairs), default=0)
+        for normalized, first_name in display_pairs:
+            if metadata and first_name in metadata:
+                m = metadata[first_name]
+                size = get_tensor_size_bytes(m)
+                total_bytes += size
+                print(f"{normalized:{name_width}}  {m.get('dtype', '?'):6s}  {str(m.get('shape', '')):30s}  {format_size(size)}")
+            else:
+                print(normalized)
     else:
-        print(f"  Directory {model_path} does not exist")
-    exit(1)
+        name_width = max((len(n) for n in tensor_names), default=0)
+        for tensor_name in sorted(tensor_names):
+            if metadata and tensor_name in metadata:
+                m = metadata[tensor_name]
+                size = get_tensor_size_bytes(m)
+                total_bytes += size
+                print(f"{tensor_name:{name_width}}  {m.get('dtype', '?'):6s}  {str(m.get('shape', '')):30s}  {format_size(size)}")
+            else:
+                print(tensor_name)
+
+    if show_sizes:
+        print(f"\nTotal: {format_size(total_bytes)}")
+
+
+def print_tensor_info(model_path: Path, tensor_name: str, num_values: Optional[int] = None):
+    tensor_file = find_tensor_file(model_path, tensor_name)
+
+    if tensor_file is None:
+        print(f"Error: Could not find tensor '{tensor_name}' in model index")
+        print(f"Model path: {model_path}")
+        sys.exit(1)
+
+    file_path = model_path / tensor_file
+
+    try:
+        header = read_safetensors_header(file_path)
+        tensor_meta = header.get(tensor_name, {})
+        dtype_str = tensor_meta.get("dtype")
+
+        with safe_open(file_path, framework="pt", device="cpu") as f:
+            if tensor_name in f.keys():
+                tensor_slice = f.get_slice(tensor_name)
+                shape = tensor_slice.get_shape()
+                print(f"Tensor: {tensor_name}")
+                print(f"File:   {tensor_file}")
+                print(f"Shape:  {shape}")
+                if dtype_str:
+                    print(f"Dtype:  {dtype_str}")
+                if tensor_meta:
+                    print(f"Size:   {format_size(get_tensor_size_bytes(tensor_meta))}")
+                if num_values is not None:
+                    tensor = f.get_tensor(tensor_name)
+                    if not dtype_str:
+                        print(f"Dtype:  {tensor.dtype}")
+                    flat = tensor.flatten()
+                    n = min(num_values, flat.numel())
+                    print(f"Values: {flat[:n].tolist()}")
+            else:
+                print(f"Error: Tensor '{tensor_name}' not found in {tensor_file}")
+                sys.exit(1)
+
+    except FileNotFoundError:
+        print(f"Error: The file '{file_path}' was not found.")
+        sys.exit(1)
+    except Exception as e:
+        print(f"An error occurred: {e}")
+        sys.exit(1)
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Print tensor information from a safetensors model"
+    )
+    parser.add_argument(
+        "tensor_name",
+        nargs="?",
+        help="Name of the tensor to inspect"
+    )
+    parser.add_argument(
+        "-m", "--model-path",
+        type=Path,
+        help="Path to the model directory (default: MODEL_PATH environment variable)"
+    )
+    parser.add_argument(
+        "-l", "--list-all-short",
+        action="store_true",
+        help="List unique tensor patterns (layer numbers replaced with #)"
+    )
+    parser.add_argument(
+        "-la", "--list-all",
+        action="store_true",
+        help="List all tensor names with actual layer numbers"
+    )
+    parser.add_argument(
+        "-n", "--num-values",
+        nargs="?",
+        const=10,
+        default=None,
+        type=int,
+        metavar="N",
+        help="Print the first N values of the tensor flattened (default: 10 if flag is given without a number)"
+    )
+    parser.add_argument(
+        "-s", "--sizes",
+        action="store_true",
+        help="Show dtype, shape, and size for each tensor when listing"
+    )
+
+    args = parser.parse_args()
+
+    model_path = args.model_path
+    if model_path is None:
+        model_path_str = os.environ.get("MODEL_PATH")
+        if model_path_str is None:
+            print("Error: --model-path not provided and MODEL_PATH environment variable not set")
+            sys.exit(1)
+        model_path = Path(model_path_str)
+
+    if not model_path.exists():
+        print(f"Error: Model path does not exist: {model_path}")
+        sys.exit(1)
+
+    if not model_path.is_dir():
+        print(f"Error: Model path is not a directory: {model_path}")
+        sys.exit(1)
+
+    if args.list_all_short or args.list_all:
+        list_all_tensors(model_path, short=args.list_all_short, show_sizes=args.sizes)
+    else:
+        if args.tensor_name is None:
+            print("Error: tensor_name is required when not using --list-all-short or --list-all")
+            sys.exit(1)
+        print_tensor_info(model_path, args.tensor_name, args.num_values)
+
+
+if __name__ == "__main__":
+    main()

examples/model-conversion/scripts/utils/tensor-info.py

@@ -1,174 +0,0 @@
-#!/usr/bin/env python3
-
-import argparse
-import json
-import os
-import re
-import sys
-from pathlib import Path
-from typing import Optional
-from safetensors import safe_open
-
-
-MODEL_SAFETENSORS_FILE = "model.safetensors"
-MODEL_SAFETENSORS_INDEX = "model.safetensors.index.json"
-
-
-def get_weight_map(model_path: Path) -> Optional[dict[str, str]]:
-    index_file = model_path / MODEL_SAFETENSORS_INDEX
-
-    if index_file.exists():
-        with open(index_file, 'r') as f:
-            index = json.load(f)
-            return index.get("weight_map", {})
-
-    return None
-
-
-def get_all_tensor_names(model_path: Path) -> list[str]:
-    weight_map = get_weight_map(model_path)
-
-    if weight_map is not None:
-        return list(weight_map.keys())
-
-    single_file = model_path / MODEL_SAFETENSORS_FILE
-    if single_file.exists():
-        try:
-            with safe_open(single_file, framework="pt", device="cpu") as f:
-                return list(f.keys())
-        except Exception as e:
-            print(f"Error reading {single_file}: {e}")
-            sys.exit(1)
-
-    print(f"Error: No safetensors files found in {model_path}")
-    sys.exit(1)
-
-
-def find_tensor_file(model_path: Path, tensor_name: str) -> Optional[str]:
-    weight_map = get_weight_map(model_path)
-
-    if weight_map is not None:
-        return weight_map.get(tensor_name)
-
-    single_file = model_path / MODEL_SAFETENSORS_FILE
-    if single_file.exists():
-        return single_file.name
-
-    return None
-
-
-def normalize_tensor_name(tensor_name: str) -> str:
-    normalized = re.sub(r'\.\d+\.', '.#.', tensor_name)
-    normalized = re.sub(r'\.\d+$', '.#', normalized)
-    return normalized
-
-
-def list_all_tensors(model_path: Path, unique: bool = False):
-    tensor_names = get_all_tensor_names(model_path)
-
-    if unique:
-        seen = set()
-        for tensor_name in sorted(tensor_names):
-            normalized = normalize_tensor_name(tensor_name)
-            if normalized not in seen:
-                seen.add(normalized)
-                print(normalized)
-    else:
-        for tensor_name in sorted(tensor_names):
-            print(tensor_name)
-
-
-def print_tensor_info(model_path: Path, tensor_name: str, num_values: Optional[int] = None):
-    tensor_file = find_tensor_file(model_path, tensor_name)
-
-    if tensor_file is None:
-        print(f"Error: Could not find tensor '{tensor_name}' in model index")
-        print(f"Model path: {model_path}")
-        sys.exit(1)
-
-    file_path = model_path / tensor_file
-
-    try:
-        with safe_open(file_path, framework="pt", device="cpu") as f:
-            if tensor_name in f.keys():
-                tensor_slice = f.get_slice(tensor_name)
-                shape = tensor_slice.get_shape()
-                print(f"Tensor: {tensor_name}")
-                print(f"File:   {tensor_file}")
-                print(f"Shape:  {shape}")
-                if num_values is not None:
-                    tensor = f.get_tensor(tensor_name)
-                    print(f"Dtype:  {tensor.dtype}")
-                    flat = tensor.flatten()
-                    n = min(num_values, flat.numel())
-                    print(f"Values: {flat[:n].tolist()}")
-            else:
-                print(f"Error: Tensor '{tensor_name}' not found in {tensor_file}")
-                sys.exit(1)
-
-    except FileNotFoundError:
-        print(f"Error: The file '{file_path}' was not found.")
-        sys.exit(1)
-    except Exception as e:
-        print(f"An error occurred: {e}")
-        sys.exit(1)
-
-
-def main():
-    parser = argparse.ArgumentParser(
-        description="Print tensor information from a safetensors model"
-    )
-    parser.add_argument(
-        "tensor_name",
-        nargs="?",  # optional (if --list is used for example)
-        help="Name of the tensor to inspect"
-    )
-    parser.add_argument(
-        "-m", "--model-path",
-        type=Path,
-        help="Path to the model directory (default: MODEL_PATH environment variable)"
-    )
-    parser.add_argument(
-        "-l", "--list",
-        action="store_true",
-        help="List unique tensor patterns in the model (layer numbers replaced with #)"
-    )
-    parser.add_argument(
-        "-n", "--num-values",
-        nargs="?",
-        const=10,
-        default=None,
-        type=int,
-        metavar="N",
-        help="Print the first N values of the tensor flattened (default: 10 if flag is given without a number)"
-    )
-
-    args = parser.parse_args()
-
-    model_path = args.model_path
-    if model_path is None:
-        model_path_str = os.environ.get("MODEL_PATH")
-        if model_path_str is None:
-            print("Error: --model-path not provided and MODEL_PATH environment variable not set")
-            sys.exit(1)
-        model_path = Path(model_path_str)
-
-    if not model_path.exists():
-        print(f"Error: Model path does not exist: {model_path}")
-        sys.exit(1)
-
-    if not model_path.is_dir():
-        print(f"Error: Model path is not a directory: {model_path}")
-        sys.exit(1)
-
-    if args.list:
-        list_all_tensors(model_path, unique=True)
-    else:
-        if args.tensor_name is None:
-            print("Error: tensor_name is required when not using --list")
-            sys.exit(1)
-        print_tensor_info(model_path, args.tensor_name, args.num_values)
-
-
-if __name__ == "__main__":
-    main()

examples/save-load-state/save-load-state.cpp

@@ -5,12 +5,15 @@
 #include <vector>
 #include <cstdio>
 
+
 int main(int argc, char ** argv) {
     common_params params;
 
     params.prompt = "The quick brown fox";
     params.sampling.seed = 1234;
 
+    const std::string_view state_file = "dump_state.bin";
+
     if (!common_params_parse(argc, argv, params, LLAMA_EXAMPLE_COMMON)) {
         return 1;
     }
@@ -53,35 +56,16 @@ int main(int argc, char ** argv) {
     // tokenize prompt
     auto tokens = common_tokenize(ctx, params.prompt, true);
 
-    // prepare the batch
-    llama_batch batch = llama_batch_init(tokens.size(), 0, 1);
-    for (size_t i = 0; i < tokens.size(); i++) {
-        common_batch_add(batch, tokens[i], i, {0}, false);
+    const bool save_state = true;
+    if (!common_prompt_batch_decode(ctx, tokens, n_past, params.n_batch, state_file, save_state)) {
+        return 1;
     }
-    batch.logits[batch.n_tokens - 1] = true; // generate next token
-
-    // evaluate prompt
-    llama_decode(ctx, batch);
-    n_past += batch.n_tokens;
-
-    // save state (rng, logits, embedding and kv_cache) to file
-    {
-        std::vector<uint8_t> state_mem(llama_state_get_size(ctx));
-        const size_t written = llama_state_get_data(ctx, state_mem.data(), state_mem.size());
-
-        FILE *fp_write = fopen("dump_state.bin", "wb");
-        fwrite(state_mem.data(), 1, written, fp_write);
-        fclose(fp_write);
-
-        fprintf(stderr, "%s : serialized state into %zd out of a maximum of %zd bytes\n", __func__, written, state_mem.size());
-    }
-
-    // save state (last tokens)
-    const auto n_past_saved = n_past;
 
     // first run
     printf("\nfirst run: %s", params.prompt.c_str());
 
+    llama_batch batch = llama_batch_init(1, 0, 1);
+
     for (auto i = 0; i < params.n_predict; i++) {
         auto next_token     = llama_sampler_sample(smpl, ctx, -1);
         auto next_token_str = common_token_to_piece(ctx, next_token);
@@ -111,27 +95,23 @@ int main(int argc, char ** argv) {
 
     printf("\nsecond run: %s", params.prompt.c_str());
 
-    // load state (rng, logits, embedding and kv_cache) from file
-    {
-        std::vector<uint8_t> state_mem;
+    // load state from file
+    std::vector<llama_token> unused_sts(tokens.size()); // unused session tokens.
+    size_t n_token_count_out = 0;
 
-        FILE * fp_read = fopen("dump_state.bin", "rb");
-        fseek(fp_read, 0, SEEK_END);
-        state_mem.resize(ftell(fp_read));
-        fseek(fp_read, 0, SEEK_SET);
-        const size_t read = fread(state_mem.data(), 1, state_mem.size(), fp_read);
-        fclose(fp_read);
-
-        if (read != llama_state_set_data(ctx2, state_mem.data(), state_mem.size())) {
-            fprintf(stderr, "\n%s : failed to read state\n", __func__);
-            return 1;
-        }
-
-        fprintf(stderr, "%s : deserialized state from %zd out of a maximum of %zd bytes\n", __func__, read, state_mem.size());
+    if (!llama_state_load_file(ctx2, state_file.data(), unused_sts.data(), unused_sts.size(), &n_token_count_out)) {
+        fprintf(stderr, "\n%s : failed to load state\n", __func__);
+        return 1;
     }
 
+    fprintf(stderr, "%s : loaded state with %zu tokens\n", __func__, n_token_count_out);
+
     // restore state (last tokens)
-    n_past = n_past_saved;
+    n_past = n_token_count_out;
+    if (!common_replay_last_token(ctx2, tokens.back(), n_past)) {
+        return 1;
+    }
+    ++n_past;
 
     // second run
     for (auto i = 0; i < params.n_predict; i++) {
@@ -160,7 +140,9 @@ int main(int argc, char ** argv) {
     }
 
     // make new context
-    llama_context * ctx3 = llama_init_from_model(model, common_context_params_to_llama(params));
+    auto params_ctx3 = common_context_params_to_llama(params);
+    params_ctx3.n_seq_max = 2;
+    llama_context * ctx3 = llama_init_from_model(model, params_ctx3);
 
     llama_sampler * smpl3 = llama_sampler_chain_init(sparams);
 
@@ -169,26 +151,21 @@ int main(int argc, char ** argv) {
     printf("\nsingle seq run: %s", params.prompt.c_str());
 
     // load state (rng, logits, embedding and kv_cache) from file
-    {
-        std::vector<uint8_t> state_mem;
+    n_token_count_out = 0;
 
-        FILE * fp_read = fopen("dump_state.bin", "rb");
-        fseek(fp_read, 0, SEEK_END);
-        state_mem.resize(ftell(fp_read));
-        fseek(fp_read, 0, SEEK_SET);
-        const size_t read = fread(state_mem.data(), 1, state_mem.size(), fp_read);
-        fclose(fp_read);
-
-        if (read != llama_state_set_data(ctx3, state_mem.data(), state_mem.size())) {
-            fprintf(stderr, "\n%s : failed to read state\n", __func__);
-            return 1;
-        }
-
-        fprintf(stderr, "%s : deserialized state from %zd out of a maximum of %zd bytes\n", __func__, read, state_mem.size());
+    if (!llama_state_load_file(ctx3, state_file.data(), unused_sts.data(), unused_sts.size(), &n_token_count_out)) {
+        fprintf(stderr, "\n%s : failed to load state\n", __func__);
+        return 1;
     }
 
+    fprintf(stderr, "%s : loaded state with %zu tokens\n", __func__, n_token_count_out);
+
     // restore state (last tokens)
-    n_past = n_past_saved;
+    n_past = n_token_count_out;
+    if (!common_replay_last_token(ctx3, tokens.back(), n_past)) {
+        return 1;
+    }
+    ++n_past;
 
     // save seq 0 and load into seq 1
     {

ggml/include/ggml.h

@@ -730,6 +730,10 @@ extern "C" {
     GGML_API size_t  ggml_type_size(enum ggml_type type);             // size in bytes for all elements in a block
     GGML_API size_t  ggml_row_size (enum ggml_type type, int64_t ne); // size in bytes for all elements in a row
 
+    GGML_DEPRECATED(
+    GGML_API double ggml_type_sizef(enum ggml_type type), // ggml_type_size()/ggml_blck_size() as float
+    "use ggml_row_size() instead");
+
     GGML_API const char * ggml_type_name(enum ggml_type type);
     GGML_API const char * ggml_op_name  (enum ggml_op   op);
     GGML_API const char * ggml_op_symbol(enum ggml_op   op);

ggml/src/ggml-cpu/arch-fallback.h

@@ -42,6 +42,7 @@
 #define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
 #define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
 #define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
+#define ggml_gemv_q5_K_8x4_q8_K_generic ggml_gemv_q5_K_8x4_q8_K
 #define ggml_gemv_q5_K_8x8_q8_K_generic ggml_gemv_q5_K_8x8_q8_K
 #define ggml_gemv_q6_K_8x4_q8_K_generic ggml_gemv_q6_K_8x4_q8_K
 #define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
@@ -55,9 +56,10 @@
 #define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
 #define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
 #define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
+#define ggml_gemm_q5_K_8x4_q8_K_generic ggml_gemm_q5_K_8x4_q8_K
 #define ggml_gemm_q5_K_8x8_q8_K_generic ggml_gemm_q5_K_8x8_q8_K
 #define ggml_gemm_q6_K_8x4_q8_K_generic ggml_gemm_q6_K_8x4_q8_K
-#define ggml_gemm_q6_K_8x8_q8_K_generic   ggml_gemm_q6_K_8x8_q8_K
+#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
 #define ggml_gemm_iq4_nl_4x4_q8_0_generic ggml_gemm_iq4_nl_4x4_q8_0
 #define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
 #define ggml_gemm_q8_0_4x4_q8_0_generic ggml_gemm_q8_0_4x4_q8_0
@@ -77,6 +79,7 @@
 #define ggml_gemv_q4_0_4x4_q8_0_generic ggml_gemv_q4_0_4x4_q8_0
 #define ggml_gemv_q4_0_4x8_q8_0_generic ggml_gemv_q4_0_4x8_q8_0
 #define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
+#define ggml_gemv_q5_K_8x4_q8_K_generic ggml_gemv_q5_K_8x4_q8_K
 #define ggml_gemv_q5_K_8x8_q8_K_generic ggml_gemv_q5_K_8x8_q8_K
 #define ggml_gemv_q6_K_8x4_q8_K_generic ggml_gemv_q6_K_8x4_q8_K
 #define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
@@ -86,6 +89,7 @@
 #define ggml_gemm_q4_0_4x4_q8_0_generic ggml_gemm_q4_0_4x4_q8_0
 #define ggml_gemm_q4_0_4x8_q8_0_generic ggml_gemm_q4_0_4x8_q8_0
 #define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
+#define ggml_gemm_q5_K_8x4_q8_K_generic ggml_gemm_q5_K_8x4_q8_K
 #define ggml_gemm_q5_K_8x8_q8_K_generic ggml_gemm_q5_K_8x8_q8_K
 #define ggml_gemm_q6_K_8x4_q8_K_generic ggml_gemm_q6_K_8x4_q8_K
 #define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
@@ -110,6 +114,7 @@
 #define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
 #define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
 #define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
+#define ggml_gemv_q5_K_8x4_q8_K_generic ggml_gemv_q5_K_8x4_q8_K
 #define ggml_gemv_q5_K_8x8_q8_K_generic ggml_gemv_q5_K_8x8_q8_K
 #define ggml_gemv_q6_K_8x4_q8_K_generic ggml_gemv_q6_K_8x4_q8_K
 #define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
@@ -123,6 +128,7 @@
 #define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
 #define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
 #define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
+#define ggml_gemm_q5_K_8x4_q8_K_generic ggml_gemm_q5_K_8x4_q8_K
 #define ggml_gemm_q5_K_8x8_q8_K_generic ggml_gemm_q5_K_8x8_q8_K
 #define ggml_gemm_q6_K_8x4_q8_K_generic ggml_gemm_q6_K_8x4_q8_K
 #define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
@@ -148,6 +154,7 @@
 #define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
 #define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
 #define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
+#define ggml_gemv_q5_K_8x4_q8_K_generic ggml_gemv_q5_K_8x4_q8_K
 #define ggml_gemv_q5_K_8x8_q8_K_generic ggml_gemv_q5_K_8x8_q8_K
 #define ggml_gemv_q6_K_8x4_q8_K_generic ggml_gemv_q6_K_8x4_q8_K
 #define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
@@ -161,6 +168,7 @@
 #define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
 #define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
 #define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
+#define ggml_gemm_q5_K_8x4_q8_K_generic ggml_gemm_q5_K_8x4_q8_K
 #define ggml_gemm_q5_K_8x8_q8_K_generic ggml_gemm_q5_K_8x8_q8_K
 #define ggml_gemm_q6_K_8x4_q8_K_generic ggml_gemm_q6_K_8x4_q8_K
 #define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
@@ -171,15 +179,9 @@
 #elif defined(__riscv)
 // quants.c
 #define quantize_row_q8_K_generic quantize_row_q8_K
-#define ggml_vec_dot_tq1_0_q8_K_generic ggml_vec_dot_tq1_0_q8_K
-#define ggml_vec_dot_tq2_0_q8_K_generic ggml_vec_dot_tq2_0_q8_K
 #define ggml_vec_dot_iq2_xxs_q8_K_generic ggml_vec_dot_iq2_xxs_q8_K
 #define ggml_vec_dot_iq2_xs_q8_K_generic ggml_vec_dot_iq2_xs_q8_K
-#define ggml_vec_dot_iq2_s_q8_K_generic ggml_vec_dot_iq2_s_q8_K
 #define ggml_vec_dot_iq3_xxs_q8_K_generic ggml_vec_dot_iq3_xxs_q8_K
-#define ggml_vec_dot_iq3_s_q8_K_generic ggml_vec_dot_iq3_s_q8_K
-#define ggml_vec_dot_iq1_s_q8_K_generic ggml_vec_dot_iq1_s_q8_K
-#define ggml_vec_dot_iq1_m_q8_K_generic ggml_vec_dot_iq1_m_q8_K
 #define ggml_vec_dot_iq4_nl_q8_0_generic ggml_vec_dot_iq4_nl_q8_0
 #define ggml_vec_dot_iq4_xs_q8_K_generic ggml_vec_dot_iq4_xs_q8_K
 #define ggml_vec_dot_mxfp4_q8_0_generic ggml_vec_dot_mxfp4_q8_0
@@ -193,6 +195,7 @@
 #define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
 #define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
 #define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
+#define ggml_gemv_q5_K_8x4_q8_K_generic ggml_gemv_q5_K_8x4_q8_K
 #define ggml_gemv_q5_K_8x8_q8_K_generic ggml_gemv_q5_K_8x8_q8_K
 #define ggml_gemv_q6_K_8x4_q8_K_generic ggml_gemv_q6_K_8x4_q8_K
 #define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
@@ -205,6 +208,7 @@
 #define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
 #define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
 #define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
+#define ggml_gemm_q5_K_8x4_q8_K_generic ggml_gemm_q5_K_8x4_q8_K
 #define ggml_gemm_q5_K_8x8_q8_K_generic ggml_gemm_q5_K_8x8_q8_K
 #define ggml_gemm_q6_K_8x4_q8_K_generic ggml_gemm_q6_K_8x4_q8_K
 #define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
@@ -236,6 +240,7 @@
 #define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
 #define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
 #define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
+#define ggml_gemv_q5_K_8x4_q8_K_generic ggml_gemv_q5_K_8x4_q8_K
 #define ggml_gemv_q5_K_8x8_q8_K_generic ggml_gemv_q5_K_8x8_q8_K
 #define ggml_gemv_q6_K_8x4_q8_K_generic ggml_gemv_q6_K_8x4_q8_K
 #define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
@@ -249,6 +254,7 @@
 #define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
 #define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
 #define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
+#define ggml_gemm_q5_K_8x4_q8_K_generic ggml_gemm_q5_K_8x4_q8_K
 #define ggml_gemm_q5_K_8x8_q8_K_generic ggml_gemm_q5_K_8x8_q8_K
 #define ggml_gemm_q6_K_8x4_q8_K_generic ggml_gemm_q6_K_8x4_q8_K
 #define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
@@ -282,6 +288,7 @@
 #define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
 #define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
 #define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
+#define ggml_gemv_q5_K_8x4_q8_K_generic ggml_gemv_q5_K_8x4_q8_K
 #define ggml_gemv_q5_K_8x8_q8_K_generic ggml_gemv_q5_K_8x8_q8_K
 #define ggml_gemv_q6_K_8x4_q8_K_generic ggml_gemv_q6_K_8x4_q8_K
 #define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
@@ -295,6 +302,7 @@
 #define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
 #define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
 #define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
+#define ggml_gemm_q5_K_8x4_q8_K_generic ggml_gemm_q5_K_8x4_q8_K
 #define ggml_gemm_q5_K_8x8_q8_K_generic ggml_gemm_q5_K_8x8_q8_K
 #define ggml_gemm_q6_K_8x4_q8_K_generic ggml_gemm_q6_K_8x4_q8_K
 #define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K

ggml/src/ggml-cpu/arch/arm/repack.cpp

@@ -785,6 +785,165 @@ void ggml_gemv_q4_K_8x8_q8_K(int                        n,
     ggml_gemv_q4_K_8x8_q8_K_generic(n, s, bs, vx, vy, nr, nc);
 }
 
+void ggml_gemv_q5_K_8x4_q8_K(int                        n,
+                             float * GGML_RESTRICT      s,
+                             size_t                     bs,
+                             const void * GGML_RESTRICT vx,
+                             const void * GGML_RESTRICT vy,
+                             int                        nr,
+                             int                        nc) {
+    constexpr int qk = QK_K;
+    const int     nb = n / qk;
+
+    constexpr int ncols_interleaved = 8;
+    constexpr int blocklen          = 4;
+
+    assert(n % qk == 0);
+    assert(nc % ncols_interleaved == 0);
+
+    UNUSED(nb);
+    UNUSED(ncols_interleaved);
+    UNUSED(blocklen);
+
+#if defined(__aarch64__) && defined(__ARM_NEON) && defined(__ARM_FEATURE_DOTPROD)
+    constexpr int    col_groups = ncols_interleaved / 4;  // 0123 and 4567
+    const uint8x16_t m4b        = vdupq_n_u8(0x0f);
+    const uint8x16_t mone       = vdupq_n_u8(1);
+    const uint8x16_t mtwo       = vdupq_n_u8(2);
+
+    // 1x8 tile = 2 x 4
+    float32x4_t acc_f32[col_groups];
+
+    const block_q8_K * GGML_RESTRICT q8_ptr = (const block_q8_K *) vy;
+
+    for (int x = 0; x < nc / ncols_interleaved; x++) {
+        const block_q5_Kx8 * GGML_RESTRICT q5_ptr = (const block_q5_Kx8 *) vx + (x * nb);
+
+        for (int i = 0; i < col_groups; i++) {
+            acc_f32[i] = vdupq_n_f32(0);
+        }
+
+        for (int b = 0; b < nb; b++) {
+            float32x4_t q5_d_0        = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].d));      // d0 d1 d2 d3
+            float32x4_t q5_d_1        = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].d + 4));  // d4 d5 d6 d7
+            float32x4_t q8_d          = vdupq_n_f32(q8_ptr[b].d);
+            float32x4_t sb_scale_0123 = vmulq_f32(q5_d_0, q8_d);
+            float32x4_t sb_scale_4567 = vmulq_f32(q5_d_1, q8_d);
+            float32x4_t q5_dmin_0     = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].dmin));      // dmin 0..3
+            float32x4_t q5_dmin_1     = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].dmin + 4));  // dmin 4..7
+            float32x4_t sb_min_0123   = vmulq_f32(q5_dmin_0, q8_d);
+            float32x4_t sb_min_4567   = vmulq_f32(q5_dmin_1, q8_d);
+
+            // interleaved bias_acc: [0]->r0 0123, [1]->r0 4567
+            int32x4_t bias_acc[2] = { vdupq_n_s32(0), vdupq_n_s32(0) };
+            int32x4_t acc_lo[col_groups];
+            int32x4_t acc_hi[col_groups];
+
+            // Each bsum is 16 elements, pairwise add leaves us with the 8 bsums of the entire block
+            const int16x8_t bsums = vpaddq_s16(vld1q_s16(q8_ptr[b].bsums), vld1q_s16(q8_ptr[b].bsums + 8));
+            int16_t         bsums_arr[8];
+            vst1q_s16(bsums_arr, bsums);
+
+            uint8x16_t qh[col_groups][8];
+            for (int c = 0; c < col_groups; c++) {
+                for (int i = 0; i < 8; i++) {
+                    qh[c][i] = vld1q_u8(q5_ptr[b].qh + i * 32 + 16 * c);
+                }
+            }
+
+            for (int sb = 0; sb < QK_K / 64; sb++) {
+                for (int i = 0; i < col_groups; i++) {
+                    acc_lo[i] = vdupq_n_s32(0);
+                    acc_hi[i] = vdupq_n_s32(0);
+                }
+                // Need scales for the low and high nibbles
+                // 2 * 12 = 24 bytes per subblock, 4 sbs -> 4 * 24 = 96 bytes total
+                int16x8_t q5sb_mins[2];
+                int16x8_t q5sb_scales[2];
+                for (int i = 0; i < 2; i++) {
+                    int8_t    aux_q5sb[8];
+                    const int offset = sb * 24 + i * 12;
+                    decode_q_Kx8_6bit_scales(&q5_ptr[b].scales[offset], &q5sb_mins[i], aux_q5sb);
+                    q5sb_scales[i] = vmovl_s8(vld1_s8(aux_q5sb));
+                }
+
+                int8x16_t q8_qs[4];
+                for (int i = 0; i < 4; i++) {
+                    q8_qs[i] = vld1q_s8(q8_ptr[b].qs + sb * 64 + i * 16);
+                }
+
+                for (int c = 0; c < col_groups; c++) {
+                    uint8x16_t q5_cols[8];
+                    uint8x16_t hbit_lo[8];
+                    uint8x16_t hbit_hi[8];
+                    int8x16_t  q5_lo[8];
+                    int8x16_t  q5_hi[8];
+
+                    for (int i = 0; i < 8; i++) {
+                        q5_cols[i] = vld1q_u8(q5_ptr[b].qs + sb * QK_K + i * 32 + 16 * c);
+                        hbit_lo[i] = vandq_u8(qh[c][i], mone);
+                        hbit_hi[i] = vshlq_n_u8(vandq_u8(qh[c][i], mtwo), 3);
+                        qh[c][i]   = vshrq_n_u8(qh[c][i], 2);
+                        q5_lo[i]   = vreinterpretq_s8_u8(vsliq_n_u8(vandq_u8(q5_cols[i], m4b), hbit_lo[i], 4));
+                        q5_hi[i]   = vreinterpretq_s8_u8(vorrq_u8(vshrq_n_u8(q5_cols[i], 4), hbit_hi[i]));
+                    }
+
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[0], q8_qs[0], 0);
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[1], q8_qs[0], 1);
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[2], q8_qs[0], 2);
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[3], q8_qs[0], 3);
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[4], q8_qs[1], 0);
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[5], q8_qs[1], 1);
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[6], q8_qs[1], 2);
+                    acc_lo[c] = vdotq_laneq_s32(acc_lo[c], q5_lo[7], q8_qs[1], 3);
+
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[0], q8_qs[2], 0);
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[1], q8_qs[2], 1);
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[2], q8_qs[2], 2);
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[3], q8_qs[2], 3);
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[4], q8_qs[3], 0);
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[5], q8_qs[3], 1);
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[6], q8_qs[3], 2);
+                    acc_hi[c] = vdotq_laneq_s32(acc_hi[c], q5_hi[7], q8_qs[3], 3);
+                }
+
+                // Scales
+                // row c0123 blk0 and blk1
+                const int16x4_t   sc_0123_lo = vget_low_s16(q5sb_scales[0]);
+                const int16x4_t   sc_0123_hi = vget_low_s16(q5sb_scales[1]);
+                const float32x4_t sumf_0123  = vcvtq_f32_s32(vaddq_s32(vmulq_s32(vmovl_s16(sc_0123_lo), acc_lo[0]),
+                                                                       vmulq_s32(vmovl_s16(sc_0123_hi), acc_hi[0])));
+                acc_f32[0]                   = vfmaq_f32(acc_f32[0], sb_scale_0123, sumf_0123);
+                // row c4567 blk0 and blk1
+                const int16x4_t   sc_4567_lo = vget_high_s16(q5sb_scales[0]);
+                const int16x4_t   sc_4567_hi = vget_high_s16(q5sb_scales[1]);
+                const float32x4_t sumf_4567  = vcvtq_f32_s32(vaddq_s32(vmulq_s32(vmovl_s16(sc_4567_lo), acc_lo[1]),
+                                                                       vmulq_s32(vmovl_s16(sc_4567_hi), acc_hi[1])));
+                acc_f32[1]                   = vfmaq_f32(acc_f32[1], sb_scale_4567, sumf_4567);
+
+                // Bias Correction
+                const int16x4_t bsums_vec_lo = vdup_n_s16(bsums_arr[2 * sb + 0]);
+                const int16x4_t bsums_vec_hi = vdup_n_s16(bsums_arr[2 * sb + 1]);
+
+                bias_acc[0] = vmlal_s16(bias_acc[0], bsums_vec_lo, vget_low_s16(q5sb_mins[0]));
+                bias_acc[0] = vmlal_s16(bias_acc[0], bsums_vec_hi, vget_low_s16(q5sb_mins[1]));
+                bias_acc[1] = vmlal_s16(bias_acc[1], bsums_vec_lo, vget_high_s16(q5sb_mins[0]));
+                bias_acc[1] = vmlal_s16(bias_acc[1], bsums_vec_hi, vget_high_s16(q5sb_mins[1]));
+            }  // for sb
+
+            acc_f32[0] = vmlsq_f32(acc_f32[0], vcvtq_f32_s32(bias_acc[0]), sb_min_0123);
+            acc_f32[1] = vmlsq_f32(acc_f32[1], vcvtq_f32_s32(bias_acc[1]), sb_min_4567);
+        }  // for b
+
+        int base = x * ncols_interleaved;
+        vst1q_f32(s + base, acc_f32[0]);
+        vst1q_f32(s + base + 4, acc_f32[1]);
+    }  // for x
+    return;
+#endif  // defined(__aarch64__) && defined(__ARM_NEON) && defined(__ARM_FEATURE_DOTPROD)
+    ggml_gemv_q5_K_8x4_q8_K_generic(n, s, bs, vx, vy, nr, nc);
+}
+
 void ggml_gemv_q5_K_8x8_q8_K(int                        n,
                              float * GGML_RESTRICT      s,
                              size_t                     bs,
@@ -3205,6 +3364,235 @@ void ggml_gemm_q4_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const vo
     ggml_gemm_q4_K_8x4_q8_K_generic(n, s, bs, vx, vy, nr, nc);
 }
 
+void ggml_gemm_q5_K_8x4_q8_K(int                        n,
+                             float * GGML_RESTRICT      s,
+                             size_t                     bs,
+                             const void * GGML_RESTRICT vx,
+                             const void * GGML_RESTRICT vy,
+                             int                        nr,
+                             int                        nc) {
+    constexpr int qk = QK_K;
+    const int     nb = n / qk;
+
+    constexpr int ncols_interleaved = 8;
+    constexpr int blocklen          = 4;
+
+    assert(n % qk == 0);
+    assert(nr % 4 == 0);
+    assert(nc % ncols_interleaved == 0);
+
+    UNUSED(nb);
+    UNUSED(ncols_interleaved);
+    UNUSED(blocklen);
+
+#if defined(__aarch64__) && defined(__ARM_NEON) && defined(__ARM_FEATURE_DOTPROD)
+    constexpr int    q8_k_blocklen = 4;
+    constexpr int    acc_size      = 2 * 4;  // 2 row pairs, 4 col pairs
+    constexpr int    col_groups    = ncols_interleaved / 4;
+    const uint8x16_t m4b           = vdupq_n_u8(0x0f);
+    const uint8x16_t mone          = vdupq_n_u8(1);
+    const uint8x16_t mtwo          = vdupq_n_u8(2);
+
+    // 8 accumulators: 2 row pairs, 4 col pairs
+    float32x4_t acc_f32[acc_size];
+
+    for (int y = 0; y < nr / q8_k_blocklen; y++) {
+        const block_q8_Kx4 * GGML_RESTRICT q8_ptr = (const block_q8_Kx4 *) vy + (y * nb);
+
+        for (int x = 0; x < nc / ncols_interleaved; x++) {
+            const block_q5_Kx8 * GGML_RESTRICT q5_ptr = (const block_q5_Kx8 *) vx + (x * nb);
+
+            for (int i = 0; i < acc_size; i++) {
+                acc_f32[i] = vdupq_n_f32(0);
+            }
+
+            for (int b = 0; b < nb; b++) {
+                // d5 0 1 2 3, 4 5 6 7
+                float32x4_t q5_d_0123    = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].d));
+                float32x4_t q5_d_4567    = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].d + 4));
+                // d8 0 1 2 3
+                float32x4_t q8_d_0123    = vld1q_f32(q8_ptr[b].d);
+                // mins
+                float32x4_t q5_dmin_0123 = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].dmin));
+                float32x4_t q5_dmin_4567 = vcvt_f32_f16(vld1_f16((const __fp16 *) q5_ptr[b].dmin + 4));
+
+                // Precomputation of scales and mins
+                float32x4_t sbd_scale_0123[q8_k_blocklen];
+                float32x4_t sbd_scale_4567[q8_k_blocklen];
+                float32x4_t sbd_min_0123[q8_k_blocklen];
+                float32x4_t sbd_min_4567[q8_k_blocklen];
+
+                sbd_scale_0123[0] = vmulq_laneq_f32(q5_d_0123, q8_d_0123, 0);
+                sbd_scale_4567[0] = vmulq_laneq_f32(q5_d_4567, q8_d_0123, 0);
+                sbd_min_0123[0]   = vmulq_laneq_f32(q5_dmin_0123, q8_d_0123, 0);
+                sbd_min_4567[0]   = vmulq_laneq_f32(q5_dmin_4567, q8_d_0123, 0);
+
+                sbd_scale_0123[1] = vmulq_laneq_f32(q5_d_0123, q8_d_0123, 1);
+                sbd_scale_4567[1] = vmulq_laneq_f32(q5_d_4567, q8_d_0123, 1);
+                sbd_min_0123[1]   = vmulq_laneq_f32(q5_dmin_0123, q8_d_0123, 1);
+                sbd_min_4567[1]   = vmulq_laneq_f32(q5_dmin_4567, q8_d_0123, 1);
+
+                sbd_scale_0123[2] = vmulq_laneq_f32(q5_d_0123, q8_d_0123, 2);
+                sbd_scale_4567[2] = vmulq_laneq_f32(q5_d_4567, q8_d_0123, 2);
+                sbd_min_0123[2]   = vmulq_laneq_f32(q5_dmin_0123, q8_d_0123, 2);
+                sbd_min_4567[2]   = vmulq_laneq_f32(q5_dmin_4567, q8_d_0123, 2);
+
+                sbd_scale_0123[3] = vmulq_laneq_f32(q5_d_0123, q8_d_0123, 3);
+                sbd_scale_4567[3] = vmulq_laneq_f32(q5_d_4567, q8_d_0123, 3);
+                sbd_min_0123[3]   = vmulq_laneq_f32(q5_dmin_0123, q8_d_0123, 3);
+                sbd_min_4567[3]   = vmulq_laneq_f32(q5_dmin_4567, q8_d_0123, 3);
+
+                // Precomputation of bsums, each vpaddq calcs all the bsums for each row
+                const int16x8_t bsums[q8_k_blocklen] = {
+                    vpaddq_s16(vld1q_s16(q8_ptr[b].bsums + 16 * 0), vld1q_s16(q8_ptr[b].bsums + 16 * 0 + 8)),
+                    vpaddq_s16(vld1q_s16(q8_ptr[b].bsums + 16 * 1), vld1q_s16(q8_ptr[b].bsums + 16 * 1 + 8)),
+                    vpaddq_s16(vld1q_s16(q8_ptr[b].bsums + 16 * 2), vld1q_s16(q8_ptr[b].bsums + 16 * 2 + 8)),
+                    vpaddq_s16(vld1q_s16(q8_ptr[b].bsums + 16 * 3), vld1q_s16(q8_ptr[b].bsums + 16 * 3 + 8)),
+                };
+                int16_t bsums_arr[QK_K / 64][8];
+                for (int q8_row = 0; q8_row < 4; q8_row++) {
+                    vst1q_s16(bsums_arr[q8_row], bsums[q8_row]);
+                }
+
+                // interleaved bias_acc: [0]->r0 0123, [1]->r1 0123, .., [4]->r0 4567, [5]->r1 4567 ..
+                int32x4_t bias_acc[acc_size];
+                for (int i = 0; i < acc_size; i++) {
+                    bias_acc[i] = vdupq_n_s32(0);
+                }
+
+                uint8x16_t qh[col_groups][8];
+                for (int c = 0; c < col_groups; c++) {
+                    for (int i = 0; i < 8; i++) {
+                        qh[c][i] = vld1q_u8(q5_ptr[b].qh + i * 32 + 16 * c);
+                    }
+                }
+
+                for (int sb = 0; sb < QK_K / 64; sb++) {
+                    // Int accumulators for qs vecdot (4 row * 2 col quartets)
+                    int32x4_t acc_lo[acc_size];
+                    int32x4_t acc_hi[acc_size];
+                    for (int i = 0; i < acc_size; i++) {
+                        acc_lo[i] = vdupq_n_s32(0);
+                        acc_hi[i] = vdupq_n_s32(0);
+                    }
+                    // Need scales for the low and high nibbles
+                    // 2 * 12 = 24 bytes per subblock, 4 sbs -> 4 * 24 = 96 bytes total
+                    int16x8_t q5sb_scales[2];
+                    int16x8_t q5sb_mins[2];
+                    for (int i = 0; i < 2; i++) {
+                        int8_t    aux_q5sb[8];
+                        const int offset = sb * 24 + i * 12;
+                        decode_q_Kx8_6bit_scales(&q5_ptr[b].scales[offset], &q5sb_mins[i], aux_q5sb);
+                        q5sb_scales[i] = vmovl_s8(vld1_s8(aux_q5sb));
+                    }
+
+                    constexpr int reads_per_sb = 8;  // 8 * 16 bytes each => 32 qs * 4 rows
+                    for (int k = 0; k < reads_per_sb; k++) {
+                        const int8x16_t q8_blk0 = vld1q_s8(q8_ptr[b].qs + sb * 256 + 16 * k);
+                        const int8x16_t q8_blk1 = vld1q_s8(q8_ptr[b].qs + sb * 256 + 16 * k + 128);
+
+                        // 0..3 & 32..35
+                        const uint8x16_t q5_0123 = vld1q_u8(q5_ptr[b].qs + sb * QK_K + 32 * k);
+                        const uint8x16_t q5_4567 = vld1q_u8(q5_ptr[b].qs + sb * QK_K + 32 * k + 16);
+
+                        // NOTE: This is the only difference with q4_K
+                        const uint8x16_t hbit_lo_0123 = vandq_u8(qh[0][k], mone);
+                        const uint8x16_t hbit_hi_0123 = vshlq_n_u8(vandq_u8(qh[0][k], mtwo), 3);
+                        qh[0][k]                      = vshrq_n_u8(qh[0][k], 2);
+                        const uint8x16_t hbit_lo_4567 = vandq_u8(qh[1][k], mone);
+                        const uint8x16_t hbit_hi_4567 = vshlq_n_u8(vandq_u8(qh[1][k], mtwo), 3);
+                        qh[1][k]                      = vshrq_n_u8(qh[1][k], 2);
+                        // From here, same as q4_K
+
+                        const int8x16_t q5_0123_lo =
+                            vreinterpretq_s8_u8(vsliq_n_u8(vandq_u8(q5_0123, m4b), hbit_lo_0123, 4));
+                        const int8x16_t q5_0123_hi =
+                            vreinterpretq_s8_u8(vorrq_u8(vshrq_n_u8(q5_0123, 4), hbit_hi_0123));
+
+                        acc_lo[0] = vdotq_laneq_s32(acc_lo[0], q5_0123_lo, q8_blk0, 0);  //  0..3  r0 c0123
+                        acc_lo[1] = vdotq_laneq_s32(acc_lo[1], q5_0123_lo, q8_blk0, 1);  //  0..3  r1 c0123
+                        acc_lo[2] = vdotq_laneq_s32(acc_lo[2], q5_0123_lo, q8_blk0, 2);  //  0..3  r2 c0123
+                        acc_lo[3] = vdotq_laneq_s32(acc_lo[3], q5_0123_lo, q8_blk0, 3);  //  0..3  r3 c0123
+
+                        acc_hi[0] = vdotq_laneq_s32(acc_hi[0], q5_0123_hi, q8_blk1, 0);  // 32..35 r0 c0123
+                        acc_hi[1] = vdotq_laneq_s32(acc_hi[1], q5_0123_hi, q8_blk1, 1);  // 32..35 r1 c0123
+                        acc_hi[2] = vdotq_laneq_s32(acc_hi[2], q5_0123_hi, q8_blk1, 2);  // 32..35 r2 c0123
+                        acc_hi[3] = vdotq_laneq_s32(acc_hi[3], q5_0123_hi, q8_blk1, 3);  // 32..35 r3 c0123
+
+                        const int8x16_t q5_4567_lo =
+                            vreinterpretq_s8_u8(vsliq_n_u8(vandq_u8(q5_4567, m4b), hbit_lo_4567, 4));
+                        const int8x16_t q5_4567_hi =
+                            vreinterpretq_s8_u8(vorrq_u8(vshrq_n_u8(q5_4567, 4), hbit_hi_4567));
+
+                        acc_lo[4] = vdotq_laneq_s32(acc_lo[4], q5_4567_lo, q8_blk0, 0);  //  0..3  r0 c4567
+                        acc_lo[5] = vdotq_laneq_s32(acc_lo[5], q5_4567_lo, q8_blk0, 1);  //  0..3  r1 c4567
+                        acc_lo[6] = vdotq_laneq_s32(acc_lo[6], q5_4567_lo, q8_blk0, 2);  //  0..3  r2 c4567
+                        acc_lo[7] = vdotq_laneq_s32(acc_lo[7], q5_4567_lo, q8_blk0, 3);  //  0..3  r3 c4567
+
+                        acc_hi[4] = vdotq_laneq_s32(acc_hi[4], q5_4567_hi, q8_blk1, 0);  // 32..35 r0 c4567
+                        acc_hi[5] = vdotq_laneq_s32(acc_hi[5], q5_4567_hi, q8_blk1, 1);  // 32..35 r1 c4567
+                        acc_hi[6] = vdotq_laneq_s32(acc_hi[6], q5_4567_hi, q8_blk1, 2);  // 32..35 r2 c4567
+                        acc_hi[7] = vdotq_laneq_s32(acc_hi[7], q5_4567_hi, q8_blk1, 3);  // 32..35 r3 c4567
+                    }
+
+                    // Scale and bias application
+                    // acc is stored interleaved to match output layout
+                    const int16x4_t sc_0123_lo = vget_low_s16(q5sb_scales[0]);
+                    const int16x4_t sc_4567_lo = vget_high_s16(q5sb_scales[0]);
+                    const int16x4_t sc_0123_hi = vget_low_s16(q5sb_scales[1]);
+                    const int16x4_t sc_4567_hi = vget_high_s16(q5sb_scales[1]);
+                    for (int row = 0; row < q8_k_blocklen; row++) {
+                        // Bias correction
+                        // row c0123 blk0 and blk1
+                        const float32x4_t sumf_0123 =
+                            vcvtq_f32_s32(vaddq_s32(vmulq_s32(vmovl_s16(sc_0123_lo), acc_lo[row]),
+                                                    vmulq_s32(vmovl_s16(sc_0123_hi), acc_hi[row])));
+                        acc_f32[2 * row] = vfmaq_f32(acc_f32[2 * row], sbd_scale_0123[row], sumf_0123);
+
+                        // row c4567 blk0 and blk1
+                        const float32x4_t sumf_4567 =
+                            vcvtq_f32_s32(vaddq_s32(vmulq_s32(vmovl_s16(sc_4567_lo), acc_lo[row + 4]),
+                                                    vmulq_s32(vmovl_s16(sc_4567_hi), acc_hi[row + 4])));
+                        acc_f32[2 * row + 1] = vfmaq_f32(acc_f32[2 * row + 1], sbd_scale_4567[row], sumf_4567);
+
+                        // Bias
+                        const int16x4_t bsums_vec_lo = vdup_n_s16(bsums_arr[sb][row * 2]);
+                        const int16x4_t bsums_vec_hi = vdup_n_s16(bsums_arr[sb][row * 2 + 1]);
+
+                        // row c0123 blk0 and blk1
+                        bias_acc[2 * row] = vmlal_s16(bias_acc[2 * row], bsums_vec_lo, vget_low_s16(q5sb_mins[0]));
+                        bias_acc[2 * row] = vmlal_s16(bias_acc[2 * row], bsums_vec_hi, vget_low_s16(q5sb_mins[1]));
+
+                        // row c4567 blk0 and blk1
+                        bias_acc[2 * row + 1] =
+                            vmlal_s16(bias_acc[2 * row + 1], bsums_vec_lo, vget_high_s16(q5sb_mins[0]));
+                        bias_acc[2 * row + 1] =
+                            vmlal_s16(bias_acc[2 * row + 1], bsums_vec_hi, vget_high_s16(q5sb_mins[1]));
+                    }
+                }  // for sb
+
+                for (int row = 0; row < q8_k_blocklen; row++) {
+                    acc_f32[2 * row] = vmlsq_f32(acc_f32[2 * row], vcvtq_f32_s32(bias_acc[2 * row]), sbd_min_0123[row]);
+                    acc_f32[2 * row + 1] =
+                        vmlsq_f32(acc_f32[2 * row + 1], vcvtq_f32_s32(bias_acc[2 * row + 1]), sbd_min_4567[row]);
+                }
+            }  // for b
+
+            for (int i = 0; i < q8_k_blocklen; i++) {
+                int row = y * q8_k_blocklen + i;
+                for (int j = 0; j < 2; j++) {
+                    int col    = x * ncols_interleaved + j * 4;
+                    int offset = row * bs + col;
+                    vst1q_f32(s + offset, acc_f32[2 * i + j]);
+                }
+            }
+        }  // for x
+    }  // for y
+    return;
+#endif  // defined(__aarch64__) && defined(__ARM_NEON) && defined(__ARM_FEATURE_DOTPROD)
+    ggml_gemm_q5_K_8x4_q8_K_generic(n, s, bs, vx, vy, nr, nc);
+}
+
 void ggml_gemm_q4_K_8x8_q8_K(int                        n,
                              float * GGML_RESTRICT      s,
                              size_t                     bs,

ggml/src/ggml-cpu/arch/riscv/quants.c

@@ -1954,3 +1954,773 @@ void ggml_vec_dot_q6_K_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const voi
 #endif
 }
 
+static const uint8_t sign_gather_indices_arr[64] = {
+    0,0,0,0,0,0,0,0, 1,1,1,1,1,1,1,1, 2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,
+    4,4,4,4,4,4,4,4, 5,5,5,5,5,5,5,5, 6,6,6,6,6,6,6,6, 7,7,7,7,7,7,7,7
+};
+
+static const uint8_t sign_bit_masks_arr[64] = {
+    1,2,4,8,16,32,64,128, 1,2,4,8,16,32,64,128, 1,2,4,8,16,32,64,128, 1,2,4,8,16,32,64,128,
+    1,2,4,8,16,32,64,128, 1,2,4,8,16,32,64,128, 1,2,4,8,16,32,64,128, 1,2,4,8,16,32,64,128
+};
+
+static void ggml_vec_dot_iq2_s_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+    assert(n % QK_K == 0);
+    UNUSED(nrc); UNUSED(bx); UNUSED(by); UNUSED(bs);
+
+    const block_iq2_s * GGML_RESTRICT x = vx;
+    const block_q8_K  * GGML_RESTRICT y = vy;
+
+    const int nb = n / QK_K;
+    const uint64_t * grid64 = (const uint64_t *)iq2s_grid;
+
+    // --- Pre-load Constants ---
+    uint16_t gather_qh_arr[8] = {0, 0, 0, 0, 1, 1, 1, 1};
+    vuint16mf2_t v_gather_qh = __riscv_vle16_v_u16mf2(gather_qh_arr, 8);
+    uint16_t shift_qh_arr[8] = {11, 9, 7, 5, 11, 9, 7, 5};
+    vuint16mf2_t v_shift_qh = __riscv_vle16_v_u16mf2(shift_qh_arr, 8);
+
+    // Constants for sign extraction
+    vuint8m2_t v_sign_gather_indices = __riscv_vle8_v_u8m2(sign_gather_indices_arr, 64);
+    vuint8m2_t v_sign_masks = __riscv_vle8_v_u8m2(sign_bit_masks_arr, 64);
+
+    float sumf = 0.0f;
+
+    for (int i = 0; i < nb; ++i) {
+        const float combined_scale = GGML_CPU_FP16_TO_FP32(x[i].d) * y[i].d;
+
+        const uint8_t * GGML_RESTRICT qs = x[i].qs;
+        const uint8_t * GGML_RESTRICT qh = x[i].qh;
+        const uint8_t * GGML_RESTRICT scales = x[i].scales;
+        const int8_t  * GGML_RESTRICT q8 = y[i].qs;
+
+        const uint8_t * signs_ptr = qs + 32;
+
+        float sum_block = 0.0f;
+
+        for (int ib = 0; ib < 4; ++ib) {
+            // Combine low + high bits
+            vuint8mf4_t v_qs_u8 = __riscv_vle8_v_u8mf4(qs, 8);
+            qs += 8;
+            uint16_t qh_val;
+            memcpy(&qh_val, qh, 2);
+            qh += 2;
+            vuint8mf8_t v_qh_raw = __riscv_vle8_v_u8mf8((const uint8_t*)&qh_val, 2);
+            vuint16mf4_t v_qh_u16 = __riscv_vwcvtu_x_x_v_u16mf4(v_qh_raw, 2);
+            vuint16mf2_t v_qh_u16_ext = __riscv_vlmul_ext_v_u16mf4_u16mf2(v_qh_u16);
+            vuint16mf2_t v_qh_expanded = __riscv_vrgather_vv_u16mf2(v_qh_u16_ext, v_gather_qh, 8);
+            v_qh_expanded = __riscv_vsll_vv_u16mf2(v_qh_expanded, v_shift_qh, 8);
+
+            // Mask: We want bits 11-12. 0x1800 = 0001 1000 0000 0000
+            v_qh_expanded = __riscv_vand_vx_u16mf2(v_qh_expanded, 0x1800, 8);
+            vuint16mf2_t v_qs_u16 = __riscv_vwcvtu_x_x_v_u16mf2(v_qs_u8, 8);
+
+            // Multiply by 8 to get byte offset, instead of element offset
+            v_qs_u16 = __riscv_vsll_vx_u16mf2(v_qs_u16, 3, 8);
+            vuint16mf2_t v_grid_offsets = __riscv_vor_vv_u16mf2(v_qs_u16, v_qh_expanded, 8);
+
+            // Lookup Grid using Byte Offsets
+            vuint64m2_t v_grid_vals = __riscv_vluxei16_v_u64m2(grid64, v_grid_offsets, 8);
+
+            vuint8m2_t v_grid_u8 = __riscv_vreinterpret_v_u64m2_u8m2(v_grid_vals);
+            vint8m2_t v_grid_i8 = __riscv_vreinterpret_v_u8m2_i8m2(v_grid_u8);
+
+            // Load signs and generate sign mask
+            vuint8mf4_t v_signs_raw = __riscv_vle8_v_u8mf4(signs_ptr, 8);
+            signs_ptr += 8;
+
+            vuint8m2_t v_signs_source = __riscv_vlmul_ext_v_u8mf4_u8m2(v_signs_raw);
+            vuint8m2_t v_signs_bcast = __riscv_vrgather_vv_u8m2(v_signs_source, v_sign_gather_indices, 64);
+
+            vuint8m2_t v_sign_bits = __riscv_vand_vv_u8m2(v_signs_bcast, v_sign_masks, 64);
+            vbool4_t m_negative = __riscv_vmsne_vx_u8m2_b4(v_sign_bits, 0, 64);
+
+            vint8m2_t v_q8 = __riscv_vle8_v_i8m2(q8, 64);
+            q8 += 64;
+
+            vint8m2_t v_q8_signed = __riscv_vrsub_vx_i8m2_mu(m_negative, v_q8, v_q8, 0, 64);
+            vint16m4_t v_dot = __riscv_vwmul_vv_i16m4(v_grid_i8, v_q8_signed, 64);
+
+            vint32m1_t v_zero = __riscv_vmv_v_x_i32m1(0, 1);
+
+            int32_t s0 = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m1_i32m1(
+                __riscv_vget_v_i16m4_i16m1(v_dot, 0), v_zero, 16));
+            int32_t s1 = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m1_i32m1(
+                __riscv_vget_v_i16m4_i16m1(v_dot, 1), v_zero, 16));
+            int32_t s2 = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m1_i32m1(
+                __riscv_vget_v_i16m4_i16m1(v_dot, 2), v_zero, 16));
+            int32_t s3 = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m1_i32m1(
+                __riscv_vget_v_i16m4_i16m1(v_dot, 3), v_zero, 16));
+
+            uint8_t sc0 = scales[0];
+            uint8_t sc1 = scales[1];
+            scales += 2;
+
+            sum_block += s0 * (2 * (sc0 & 0xF) + 1);
+            sum_block += s1 * (2 * (sc0 >> 4)  + 1);
+            sum_block += s2 * (2 * (sc1 & 0xF) + 1);
+            sum_block += s3 * (2 * (sc1 >> 4)  + 1);
+        }
+        sumf += sum_block * combined_scale;
+    }
+    *s = 0.125f * sumf;
+}
+
+static void ggml_vec_dot_iq2_s_q8_K_vl128(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+    assert(n % QK_K == 0);
+    UNUSED(nrc); UNUSED(bx); UNUSED(by); UNUSED(bs);
+
+    const block_iq2_s * GGML_RESTRICT x = vx;
+    const block_q8_K  * GGML_RESTRICT y = vy;
+
+    const int nb = n / QK_K;
+    const uint64_t * grid64 = (const uint64_t *)iq2s_grid;
+
+    // Pre-load Constants
+    vuint8m2_t v_ids = __riscv_vid_v_u8m2(32);
+    vuint8m2_t v_sign_gather_indices = __riscv_vsrl_vx_u8m2(v_ids, 3, 32);
+    vuint8m2_t v_ones = __riscv_vmv_v_x_u8m2(1, 32);
+    vuint8m2_t v_shift_amts = __riscv_vand_vx_u8m2(v_ids, 7, 32);
+    vuint8m2_t v_sign_masks = __riscv_vsll_vv_u8m2(v_ones, v_shift_amts, 32);
+    uint16_t shift_qh_arr[4] = {11, 9, 7, 5};
+    vuint16mf2_t v_shift_qh = __riscv_vle16_v_u16mf2(shift_qh_arr, 4);
+
+    float sumf = 0.0f;
+
+    for (int i = 0; i < nb; ++i) {
+        const float combined_scale = GGML_CPU_FP16_TO_FP32(x[i].d) * y[i].d;
+
+        const uint8_t * GGML_RESTRICT qs = x[i].qs;
+        const uint8_t * GGML_RESTRICT qh = x[i].qh;
+        const uint8_t * GGML_RESTRICT scales = x[i].scales;
+        const int8_t  * GGML_RESTRICT q8 = y[i].qs;
+
+        const uint8_t * signs_ptr = qs + 32;
+        float sum_block = 0.0f;
+
+        for (int ib = 0; ib < 8; ++ib) {
+
+            // Load Low Bits [4 bytes]
+            vuint8mf4_t v_qs_u8 = __riscv_vle8_v_u8mf4(qs, 4);
+            qs += 4;
+
+            // Load 1 byte. It contains bits for 4 mini-blocks.
+            uint8_t qh_val = *qh++;
+
+            // Combine Low + High bits of 10bit indices
+            vuint8mf4_t v_qh_raw = __riscv_vmv_v_x_u8mf4(qh_val, 4);
+            vuint16mf2_t v_qh_u16 = __riscv_vwcvtu_x_x_v_u16mf2(v_qh_raw, 4);
+            vuint16mf2_t v_qh_mf2 = __riscv_vsll_vv_u16mf2(v_qh_u16, v_shift_qh, 4);
+            v_qh_mf2 = __riscv_vand_vx_u16mf2(v_qh_mf2, 0x1800, 4);
+            vuint16mf2_t v_qs_u16_mf2 = __riscv_vwcvtu_x_x_v_u16mf2(v_qs_u8, 4);
+            vuint16mf2_t v_qs_u16 = __riscv_vsll_vx_u16mf2(v_qs_u16_mf2, 3, 4);
+            vuint16mf2_t v_grid_offsets = __riscv_vor_vv_u16mf2(v_qs_u16, v_qh_mf2, 4);
+
+            // Lookup Grid
+            vint8m2_t v_grid_i8 = __riscv_vreinterpret_v_u8m2_i8m2(__riscv_vreinterpret_v_u64m2_u8m2(__riscv_vluxei16_v_u64m2(grid64, v_grid_offsets, 4)));
+
+            vuint8mf4_t v_signs_raw = __riscv_vle8_v_u8mf4(signs_ptr, 4);
+            signs_ptr += 4;
+            vuint8m2_t v_signs_source = __riscv_vlmul_ext_v_u8mf4_u8m2(v_signs_raw);
+            vuint8m2_t v_signs_bcast = __riscv_vrgather_vv_u8m2(v_signs_source, v_sign_gather_indices, 32);
+
+            // generating sign mask
+            vuint8m2_t v_sign_bits = __riscv_vand_vv_u8m2(v_signs_bcast, v_sign_masks, 32);
+            vbool4_t m_negative = __riscv_vmsne_vx_u8m2_b4(v_sign_bits, 0, 32);
+
+            vint8m2_t v_q8 = __riscv_vle8_v_i8m2(q8, 32);
+            q8 += 32;
+
+            // apply signs
+            vint8m2_t v_q8_signed = __riscv_vrsub_vx_i8m2_mu(m_negative,v_q8, v_q8, 0, 32);
+            vint16m4_t v_dot = __riscv_vwmul_vv_i16m4(v_grid_i8, v_q8_signed, 32);
+
+            // Reduction
+            vint32m1_t v_zero = __riscv_vmv_v_x_i32m1(0, 1);
+
+            // Reduce 0-15 (First Half)
+            int32_t s0 = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(
+                __riscv_vget_v_i16m4_i16m2(v_dot, 0), v_zero, 16));
+
+            // Reduce 16-31 (Second Half)
+            int32_t s1 = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(
+                __riscv_vget_v_i16m4_i16m2(v_dot, 1), v_zero, 16));
+
+            // Apply sub Scales
+            uint8_t sc = *scales++;
+
+            sum_block += s0 * (2 * (sc & 0xF) + 1);
+            sum_block += s1 * (2 * (sc >> 4)  + 1);
+        }
+        sumf += sum_block * combined_scale;
+    }
+    *s = 0.125f * sumf;
+}
+
+void ggml_vec_dot_iq2_s_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+#if defined __riscv_v_intrinsic
+    switch (__riscv_vlenb() * 8) {
+        case 128:
+            ggml_vec_dot_iq2_s_q8_K_vl128(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+        case 256:
+            ggml_vec_dot_iq2_s_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+        default:
+            ggml_vec_dot_iq2_s_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+    }
+#else
+    ggml_vec_dot_iq2_s_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+#endif
+}
+
+static void ggml_vec_dot_iq3_s_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+    assert(n % QK_K == 0);
+    UNUSED(nrc);
+    UNUSED(bx);
+    UNUSED(by);
+    UNUSED(bs);
+
+    const block_iq3_s * GGML_RESTRICT x = vx;
+    const block_q8_K  * GGML_RESTRICT y = vy;
+
+    const int nb = n / QK_K;
+
+    const uint64_t * grid64 = (const uint64_t *)iq3s_grid;
+
+    // --- Pre-load Constants ---
+    const uint16_t qh_bit_shifts_arr[16] = {
+        0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
+    };
+    vuint8m2_t v_sign_gather_indices = __riscv_vle8_v_u8m2(sign_gather_indices_arr, 64);
+    vuint8m2_t v_sign_masks = __riscv_vle8_v_u8m2(sign_bit_masks_arr, 64);
+    vuint16m1_t v_qh_shifts = __riscv_vle16_v_u16m1(qh_bit_shifts_arr, 16);
+
+    float sumf = 0.0f;
+
+    for (int i = 0; i < nb; ++i) {
+        const float d = GGML_CPU_FP16_TO_FP32(x[i].d);
+        const float combined_scale = d * y[i].d;
+
+        const uint8_t * GGML_RESTRICT qs = x[i].qs;
+        const uint8_t * GGML_RESTRICT qh = x[i].qh;
+        const uint8_t * GGML_RESTRICT scales = x[i].scales;
+        const uint8_t * GGML_RESTRICT signs = x[i].signs;
+        const int8_t  * GGML_RESTRICT q8 = y[i].qs;
+
+        float sum_block = 0.0f;
+
+        // Loop: Process 64 weights (16 mini-blocks of 4) per iteration
+        for (int ib = 0; ib < 4; ++ib) {
+
+            vuint8mf2_t v_qs_u8 = __riscv_vle8_v_u8mf2(qs, 16);
+            qs += 16;
+
+            uint16_t qh_val;
+            memcpy(&qh_val, qh, 2);
+            qh += 2;
+
+            vuint16m1_t v_qh_val = __riscv_vmv_v_x_u16m1(qh_val, 16);
+            // Extract bits: (qh >> i) & 1
+            v_qh_val = __riscv_vsrl_vv_u16m1(v_qh_val, v_qh_shifts, 16);
+            v_qh_val = __riscv_vand_vx_u16m1(v_qh_val, 1, 16);
+
+            vuint16m1_t v_qs_u16 = __riscv_vwcvtu_x_x_v_u16m1(v_qs_u8, 16);
+            v_qs_u16 = __riscv_vsll_vx_u16m1(v_qs_u16, 2, 16);
+            v_qh_val = __riscv_vsll_vx_u16m1(v_qh_val, 10, 16);
+            vuint16m1_t v_grid_offsets = __riscv_vor_vv_u16m1(v_qs_u16, v_qh_val, 16);
+
+            // Grid value is 4xuint8
+            vuint32m2_t v_grid_packed = __riscv_vluxei16_v_u32m2((const uint32_t *)grid64, v_grid_offsets, 16);
+            vuint8m2_t v_grid_u8 = __riscv_vreinterpret_v_u32m2_u8m2(v_grid_packed);
+            vuint8mf4_t v_signs_raw = __riscv_vle8_v_u8mf4(signs, 8);
+            signs += 8;
+
+            // Generate sign mask
+            vuint8m2_t v_signs_source = __riscv_vlmul_ext_v_u8mf4_u8m2(v_signs_raw);
+            vuint8m2_t v_signs_bcast = __riscv_vrgather_vv_u8m2(v_signs_source, v_sign_gather_indices, 64);
+            vuint8m2_t v_sign_bits = __riscv_vand_vv_u8m2(v_signs_bcast, v_sign_masks, 64);
+            vbool4_t m_negative = __riscv_vmsne_vx_u8m2_b4(v_sign_bits, 0, 64);
+
+            vint8m2_t v_q8 = __riscv_vle8_v_i8m2(q8, 64);
+            q8 += 64;
+
+            // Apply Signs
+            vint8m2_t v_q8_signed = __riscv_vrsub_vx_i8m2_mu(m_negative, v_q8, v_q8, 0, 64);
+            vint16m4_t v_dot = __riscv_vwmulsu_vv_i16m4(v_q8_signed, v_grid_u8, 64);
+
+            // Reduction
+            vint16m2_t v_dot_lo = __riscv_vget_v_i16m4_i16m2(v_dot, 0);
+            vint16m2_t v_dot_hi = __riscv_vget_v_i16m4_i16m2(v_dot, 1);
+            vint32m1_t v_zero = __riscv_vmv_v_x_i32m1(0, 1);
+
+            int32_t s_lo = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(v_dot_lo, v_zero, 32));
+            int32_t s_hi = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(v_dot_hi, v_zero, 32));
+
+            // Apply sub-scales
+            uint8_t sc_byte = *scales++;
+            int sc_lo = (sc_byte & 0xF) * 2 + 1;
+            int sc_hi = (sc_byte >> 4)  * 2 + 1;
+
+            sum_block += s_lo * sc_lo + s_hi * sc_hi;
+        }
+        sumf += sum_block * combined_scale;
+    }
+    *s = 0.125f * sumf;
+}
+
+void ggml_vec_dot_iq3_s_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+#if defined __riscv_v_intrinsic
+    switch (__riscv_vlenb() * 8) {
+        case 256:
+            ggml_vec_dot_iq3_s_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+        default:
+            ggml_vec_dot_iq3_s_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+    }
+#else
+    ggml_vec_dot_iq3_s_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+#endif
+}
+
+static void ggml_vec_dot_tq1_0_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+    assert(nrc == 1);
+    UNUSED(nrc);
+    UNUSED(bx);
+    UNUSED(by);
+    UNUSED(bs);
+
+    const block_tq1_0 * GGML_RESTRICT x = vx;
+    const block_q8_K  * GGML_RESTRICT y = vy;
+
+    const int nb = n / QK_K;
+
+    float sumf = 0.0f;
+    uint8_t pow[16] = {1, 1, 1, 1, 3, 3, 3, 3, 9, 9, 9, 9, 27, 27, 27, 27};
+
+    for (int i = 0; i < nb; i++) {
+        // First loop.
+        vint32m4_t suml1;
+        {
+            const int vl = 32;
+            vuint8m1_t tq = __riscv_vle8_v_u8m1(x[i].qs, vl);
+
+            vuint16m2_t tq0 = __riscv_vsrl_vx_u16m2(__riscv_vwmulu_vx_u16m2(tq, 3, vl), 8, vl);
+            vuint16m2_t tq1 = __riscv_vsrl_vx_u16m2(__riscv_vwmulu_vx_u16m2(__riscv_vmul_vx_u8m1(tq, 3, vl), 3, vl), 8, vl);
+            vuint16m2_t tq2 = __riscv_vsrl_vx_u16m2(__riscv_vwmulu_vx_u16m2(__riscv_vmul_vx_u8m1(tq, 9, vl), 3, vl), 8, vl);
+            vuint16m2_t tq3 = __riscv_vsrl_vx_u16m2(__riscv_vwmulu_vx_u16m2(__riscv_vmul_vx_u8m1(tq, 27, vl), 3, vl), 8, vl);
+            vuint16m2_t tq4 = __riscv_vsrl_vx_u16m2(__riscv_vwmulu_vx_u16m2(__riscv_vmul_vx_u8m1(tq, 81, vl), 3, vl), 8, vl);
+
+            vint16m2_t q80 = __riscv_vwcvt_x_x_v_i16m2(__riscv_vle8_v_i8m1(y[i].qs + 0, vl), vl);
+            vint16m2_t q81 = __riscv_vwcvt_x_x_v_i16m2(__riscv_vle8_v_i8m1(y[i].qs + 32, vl), vl);
+            vint16m2_t q82 = __riscv_vwcvt_x_x_v_i16m2(__riscv_vle8_v_i8m1(y[i].qs + 64, vl), vl);
+            vint16m2_t q83 = __riscv_vwcvt_x_x_v_i16m2(__riscv_vle8_v_i8m1(y[i].qs + 96, vl), vl);
+            vint16m2_t q84 = __riscv_vwcvt_x_x_v_i16m2(__riscv_vle8_v_i8m1(y[i].qs + 128, vl), vl);
+
+            vint16m2_t sum0 = __riscv_vmul_vv_i16m2(__riscv_vreinterpret_v_u16m2_i16m2(__riscv_vsub_vx_u16m2(tq0, 1, vl)), q80, vl);
+            vint16m2_t sum1 = __riscv_vmul_vv_i16m2(__riscv_vreinterpret_v_u16m2_i16m2(__riscv_vsub_vx_u16m2(tq1, 1, vl)), q81, vl);
+            vint16m2_t sum2 = __riscv_vmul_vv_i16m2(__riscv_vreinterpret_v_u16m2_i16m2(__riscv_vsub_vx_u16m2(tq2, 1, vl)), q82, vl);
+            vint16m2_t sum3 = __riscv_vmul_vv_i16m2(__riscv_vreinterpret_v_u16m2_i16m2(__riscv_vsub_vx_u16m2(tq3, 1, vl)), q83, vl);
+            vint16m2_t sum4 = __riscv_vmul_vv_i16m2(__riscv_vreinterpret_v_u16m2_i16m2(__riscv_vsub_vx_u16m2(tq4, 1, vl)), q84, vl);
+
+            vint32m4_t sumi0 = __riscv_vwadd_vv_i32m4(sum0, sum1, vl);
+            vint32m4_t sumi1 = __riscv_vwadd_vv_i32m4(sum2, sum3, vl);
+            suml1 = __riscv_vadd_vv_i32m4(__riscv_vwcvt_x_x_v_i32m4(sum4, vl), __riscv_vadd_vv_i32m4(sumi0, sumi1, vl), vl);
+        }
+
+        // Second loop.
+        vint32m2_t suml2;
+        {
+            const int vl = 16;
+            vuint8mf2_t tq = __riscv_vle8_v_u8mf2(x[i].qs + 32, vl);
+
+            vuint16m1_t tq0 = __riscv_vsrl_vx_u16m1(__riscv_vwmulu_vx_u16m1(tq, 3 * 1, vl), 8, vl);
+            vuint16m1_t tq1 = __riscv_vsrl_vx_u16m1(__riscv_vwmulu_vx_u16m1(__riscv_vmul_vx_u8mf2(tq, 3, vl), 3, vl), 8, vl);
+            vuint16m1_t tq2 = __riscv_vsrl_vx_u16m1(__riscv_vwmulu_vx_u16m1(__riscv_vmul_vx_u8mf2(tq, 9, vl), 3, vl), 8, vl);
+            vuint16m1_t tq3 = __riscv_vsrl_vx_u16m1(__riscv_vwmulu_vx_u16m1(__riscv_vmul_vx_u8mf2(tq, 27, vl), 3, vl), 8, vl);
+            vuint16m1_t tq4 = __riscv_vsrl_vx_u16m1(__riscv_vwmulu_vx_u16m1(__riscv_vmul_vx_u8mf2(tq, 81, vl), 3, vl), 8, vl);
+
+            vint16m1_t q80 = __riscv_vwcvt_x_x_v_i16m1(__riscv_vle8_v_i8mf2(y[i].qs + 160, vl), vl);
+            vint16m1_t q81 = __riscv_vwcvt_x_x_v_i16m1(__riscv_vle8_v_i8mf2(y[i].qs + 176, vl), vl);
+            vint16m1_t q82 = __riscv_vwcvt_x_x_v_i16m1(__riscv_vle8_v_i8mf2(y[i].qs + 192, vl), vl);
+            vint16m1_t q83 = __riscv_vwcvt_x_x_v_i16m1(__riscv_vle8_v_i8mf2(y[i].qs + 208, vl), vl);
+            vint16m1_t q84 = __riscv_vwcvt_x_x_v_i16m1(__riscv_vle8_v_i8mf2(y[i].qs + 224, vl), vl);
+
+            vint16m1_t sum0 = __riscv_vmul_vv_i16m1(__riscv_vreinterpret_v_u16m1_i16m1(__riscv_vsub_vx_u16m1(tq0, 1, vl)), q80, vl);
+            vint16m1_t sum1 = __riscv_vmul_vv_i16m1(__riscv_vreinterpret_v_u16m1_i16m1(__riscv_vsub_vx_u16m1(tq1, 1, vl)), q81, vl);
+            vint16m1_t sum2 = __riscv_vmul_vv_i16m1(__riscv_vreinterpret_v_u16m1_i16m1(__riscv_vsub_vx_u16m1(tq2, 1, vl)), q82, vl);
+            vint16m1_t sum3 = __riscv_vmul_vv_i16m1(__riscv_vreinterpret_v_u16m1_i16m1(__riscv_vsub_vx_u16m1(tq3, 1, vl)), q83, vl);
+            vint16m1_t sum4 = __riscv_vmul_vv_i16m1(__riscv_vreinterpret_v_u16m1_i16m1(__riscv_vsub_vx_u16m1(tq4, 1, vl)), q84, vl);
+
+            vint32m2_t sumi0 = __riscv_vwadd_vv_i32m2(sum0, sum1, vl);
+            vint32m2_t sumi1 = __riscv_vwadd_vv_i32m2(sum2, sum3, vl);
+            suml2 = __riscv_vadd_vv_i32m2(__riscv_vwcvt_x_x_v_i32m2(sum4, vl), __riscv_vadd_vv_i32m2(sumi0, sumi1, vl), vl);
+        }
+
+        // Third loop.
+        vint32m2_t suml3;
+        {
+            const int vl = 16;
+
+            uint32_t qh;
+            memcpy(&qh, &x[i].qh[0], 4);
+            // Prevent fusion with vmv.
+            __asm__ __volatile__("" : "+r"(qh));
+            vuint8mf2_t tq = __riscv_vreinterpret_v_u32mf2_u8mf2(__riscv_vmv_v_x_u32mf2(qh, vl / 4));
+
+            vuint8mf2_t p = __riscv_vle8_v_u8mf2(pow, vl);
+
+            vuint16m1_t tq0 = __riscv_vsrl_vx_u16m1(__riscv_vwmulu_vx_u16m1(__riscv_vmul_vv_u8mf2(tq, p, vl), 3, vl), 8, vl);
+
+            vint16m1_t q80 = __riscv_vwcvt_x_x_v_i16m1(__riscv_vle8_v_i8mf2(y[i].qs + 240, vl), vl);
+
+            vint16m1_t sum0 = __riscv_vmul_vv_i16m1(__riscv_vreinterpret_v_u16m1_i16m1(__riscv_vsub_vx_u16m1(tq0, 1, vl)), q80, vl);
+            suml3 = __riscv_vwcvt_x_x_v_i32m2(sum0, vl);
+        }
+
+        vint32m2_t sumb = __riscv_vadd_vv_i32m2(__riscv_vget_v_i32m4_i32m2(suml1, 0), __riscv_vget_v_i32m4_i32m2(suml1, 1), 16);
+        sumb = __riscv_vadd_vv_i32m2(sumb, suml2, 16);
+        sumb = __riscv_vadd_vv_i32m2(sumb, suml3, 16);
+
+        vint32m1_t sum = __riscv_vredsum_vs_i32m2_i32m1(sumb, __riscv_vmv_v_x_i32m1(0, 1), 16);
+        sumf += __riscv_vmv_x_s_i32m1_i32(sum) * y[i].d * GGML_CPU_FP16_TO_FP32(x[i].d);
+    }
+
+    *s = sumf;
+}
+
+void ggml_vec_dot_tq1_0_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+#if defined __riscv_v_intrinsic
+    switch (__riscv_vlenb() * 8) {
+        case 256:
+            ggml_vec_dot_tq1_0_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+        default:
+            ggml_vec_dot_tq1_0_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+    }
+#else
+    ggml_vec_dot_tq1_0_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+#endif
+}
+
+static void ggml_vec_dot_tq2_0_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+    assert(n % QK_K == 0);
+    assert(nrc == 1);
+    UNUSED(nrc);
+    UNUSED(bx);
+    UNUSED(by);
+    UNUSED(bs);
+
+    const block_tq2_0 * GGML_RESTRICT x = vx;
+    const block_q8_K  * GGML_RESTRICT y = vy;
+
+    const int nb = n / QK_K;
+
+    float sumf = 0.0f;
+    for (int i = 0; i < nb; ++i) {
+        int32_t sumi = 0;
+
+        for (size_t j = 0; j < sizeof(x[0].qs); j += 32) {
+            const int8_t * py0 = &y[i].qs[j * 4 + 0 * 32];
+            const int8_t * py1 = &y[i].qs[j * 4 + 1 * 32];
+            const int8_t * py2 = &y[i].qs[j * 4 + 2 * 32];
+            const int8_t * py3 = &y[i].qs[j * 4 + 3 * 32];
+            const uint8_t* px  = &x[i].qs[j];
+
+            size_t vlmax_16m2 = __riscv_vsetvl_e16m2(32);
+            vint16m2_t vacc16 = __riscv_vmv_v_x_i16m2(0, vlmax_16m2);
+
+            size_t vl = __riscv_vsetvl_e8m1(32);
+
+            vuint8m1_t vx_u8 = __riscv_vle8_v_u8m1(px, vl);
+
+            vint8m1_t vy0 = __riscv_vle8_v_i8m1(py0 , vl);
+            vint8m1_t vy1 = __riscv_vle8_v_i8m1(py1, vl);
+            vint8m1_t vy2 = __riscv_vle8_v_i8m1(py2, vl);
+            vint8m1_t vy3 = __riscv_vle8_v_i8m1(py3, vl);
+
+            // l=0 (bits 1:0)
+            vuint8m1_t t0 = __riscv_vand_vx_u8m1(vx_u8, 0x03, vl);
+            vint8m1_t vq0 = __riscv_vsub_vx_i8m1(__riscv_vreinterpret_v_u8m1_i8m1(t0), 1, vl);
+
+            // l=1 (bits 3:2)
+            vuint8m1_t t1 = __riscv_vand_vx_u8m1(__riscv_vsrl_vx_u8m1(vx_u8, 2, vl), 0x03, vl);
+            vint8m1_t vq1 = __riscv_vsub_vx_i8m1(__riscv_vreinterpret_v_u8m1_i8m1(t1), 1, vl);
+
+            // l=2 (bits 5:4)
+            vuint8m1_t t2 = __riscv_vand_vx_u8m1(__riscv_vsrl_vx_u8m1(vx_u8, 4, vl), 0x03, vl);
+            vint8m1_t vq2 = __riscv_vsub_vx_i8m1(__riscv_vreinterpret_v_u8m1_i8m1(t2), 1, vl);
+
+            // l=3 (bits 7:6)
+            vuint8m1_t t3 = __riscv_vsrl_vx_u8m1(vx_u8, 6, vl); // No final AND needed as vsrl shifts in zeros
+            vint8m1_t vq3 = __riscv_vsub_vx_i8m1(__riscv_vreinterpret_v_u8m1_i8m1(t3), 1, vl);
+
+            // 4. Multiply and accumulate
+            vacc16 = __riscv_vwmacc_vv_i16m2(vacc16, vq0, vy0, vl);
+            vacc16 = __riscv_vwmacc_vv_i16m2(vacc16, vq1, vy1, vl);
+            vacc16 = __riscv_vwmacc_vv_i16m2(vacc16, vq2, vy2, vl);
+            vacc16 = __riscv_vwmacc_vv_i16m2(vacc16, vq3, vy3, vl);
+
+            vlmax_16m2 = __riscv_vsetvl_e16m2(32);
+            vint32m1_t vzero32 = __riscv_vmv_v_x_i32m1(0, 1);
+            vint32m1_t vred32 = __riscv_vwredsum_vs_i16m2_i32m1(vacc16, vzero32, vlmax_16m2);
+
+            sumi += __riscv_vmv_x_s_i32m1_i32(vred32);
+        }
+        const float d = y[i].d * GGML_CPU_FP16_TO_FP32(x[i].d);
+        sumf += (float)sumi * d;
+    }
+
+    *s = sumf;
+}
+
+void ggml_vec_dot_tq2_0_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+#if defined __riscv_v_intrinsic
+    switch (__riscv_vlenb() * 8) {
+        case 256:
+            ggml_vec_dot_tq2_0_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+        default:
+            ggml_vec_dot_tq2_0_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+    }
+#else
+    ggml_vec_dot_tq2_0_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+#endif
+}
+
+static void ggml_vec_dot_iq1_s_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+    assert(n % QK_K == 0);
+    assert(nrc == 1);
+    UNUSED(nrc);
+    UNUSED(bx);
+    UNUSED(by);
+    UNUSED(bs);
+
+    const block_iq1_s * GGML_RESTRICT x = vx;
+    const block_q8_K  * GGML_RESTRICT y = vy;
+
+    const int nb = n / QK_K;
+
+    float sumf = 0;
+    for (int i = 0; i < nb; ++i) {
+        // Load qh once for the entire superblock.
+        vuint16mf2_t qh = __riscv_vle16_v_u16mf2(x[i].qh, 8);
+
+        // Calculate ls.
+        vuint16mf2_t temp = __riscv_vsrl_vx_u16mf2(qh, 12, 8);
+        temp = __riscv_vand_vx_u16mf2(temp, 7, 8);
+        vint32m1_t ls = __riscv_vreinterpret_v_u32m1_i32m1(__riscv_vwmulu_vx_u32m1(temp, 2, 8));
+        ls = __riscv_vadd_vx_i32m1(ls, 1, 8);
+
+        // Calculate delta.
+        vbool32_t mask = __riscv_vmseq_vx_u16mf2_b32(__riscv_vand_vx_u16mf2(qh, 0x8000, 8), 0, 8);
+        vint32m1_t delta_neg = __riscv_vmv_v_x_i32m1(-1, 8);
+        vint32m1_t delta_pos = __riscv_vmv_v_x_i32m1(1, 8);
+        vint32m1_t delta = __riscv_vmerge_vvm_i32m1(delta_neg, delta_pos, mask, 8);
+
+        // Load qs.
+        vuint8m1_t qs = __riscv_vle8_v_u8m1(x[i].qs, 32);
+
+        // Prepare the indices.
+        const uint64_t shift = 0x0009000600030000;
+        vuint16m2_t qh_shift = __riscv_vreinterpret_v_u64m2_u16m2(__riscv_vmv_v_x_u64m2(shift, 8));
+        vuint16m2_t qh_gather_index = __riscv_vreinterpret_v_i16m2_u16m2(
+            __riscv_vdiv_vx_i16m2(__riscv_vreinterpret_v_u16m2_i16m2(__riscv_vid_v_u16m2(32)), 4, 32));
+        vuint16m2_t qh_ext = __riscv_vlmul_ext_v_u16m1_u16m2(__riscv_vlmul_ext_v_u16mf2_u16m1(qh));
+        vuint16m2_t qh_index = __riscv_vrgather_vv_u16m2(qh_ext, qh_gather_index, 32);
+        qh_index = __riscv_vsrl_vv_u16m2(qh_index, qh_shift, 32);
+        qh_index = __riscv_vand_vx_u16m2(qh_index, 7, 32);
+        qh_index = __riscv_vsll_vx_u16m2(qh_index, 8, 32);
+        qh_index = __riscv_vor_vv_u16m2(qh_index, __riscv_vzext_vf2_u16m2(qs, 32), 32);
+        vuint16m2_t index = __riscv_vsll_vx_u16m2(qh_index, 3, 32);
+
+        // Final lsums.
+        int32_t lsums_s[8];
+        vint32m1_t one_scalar = __riscv_vmv_v_x_i32m1(0, 1);
+
+        // Sub-blocks 1-4
+        {
+            vuint16m1_t grid_index0 = __riscv_vget_v_u16m2_u16m1(index, 0);
+            vint8m4_t grid0 = __riscv_vreinterpret_v_i64m4_i8m4(__riscv_vluxei16_v_i64m4((const int64_t*)iq1s_grid, grid_index0, 16));
+            vint8m4_t q80 = __riscv_vle8_v_i8m4(y[i].qs, 128);
+            vint16m8_t lsum0 = __riscv_vwmul_vv_i16m8(grid0, q80, 128);
+            lsums_s[0] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum0, 0), one_scalar, 32));
+            lsums_s[1] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum0, 1), one_scalar, 32));
+            lsums_s[2] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum0, 2), one_scalar, 32));
+            lsums_s[3] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum0, 3), one_scalar, 32));
+        }
+        __asm__ __volatile__("" ::: "memory");
+        // Sub-blocks 5-8
+        {
+            vuint16m1_t grid_index1 = __riscv_vget_v_u16m2_u16m1(index, 1);
+            vint8m4_t grid1 = __riscv_vreinterpret_v_i64m4_i8m4(__riscv_vluxei16_v_i64m4((const int64_t*)iq1s_grid, grid_index1, 16));
+            vint8m4_t q81 = __riscv_vle8_v_i8m4(&y[i].qs[128], 128);
+            vint16m8_t lsum1 = __riscv_vwmul_vv_i16m8(grid1, q81, 128);
+            lsums_s[4] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum1, 0), one_scalar, 32));
+            lsums_s[5] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum1, 1), one_scalar, 32));
+            lsums_s[6] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum1, 2), one_scalar, 32));
+            lsums_s[7] = __riscv_vmv_x_s_i32m1_i32(__riscv_vwredsum_vs_i16m2_i32m1(__riscv_vget_v_i16m8_i16m2(lsum1, 3), one_scalar, 32));
+        }
+        __asm__ __volatile__("" ::: "memory");
+        vint32m1_t lsums = __riscv_vle32_v_i32m1(&lsums_s[0], 8);
+
+        // Calculate the bsums.
+        vint16m1_t bsums_0 = __riscv_vle16_v_i16m1(y[i].bsums, 16);
+        const vuint32m1_t bsums_i32 = __riscv_vreinterpret_v_u16m1_u32m1(__riscv_vreinterpret_v_i16m1_u16m1(bsums_0));
+        const vint16mf2_t bsums_i32_0 = __riscv_vreinterpret_v_u16mf2_i16mf2(__riscv_vnsrl_wx_u16mf2(bsums_i32, 0, 8));
+        const vint16mf2_t bsums_i32_1 = __riscv_vreinterpret_v_u16mf2_i16mf2(__riscv_vnsrl_wx_u16mf2(bsums_i32, 16, 8));
+        const vint32m1_t bsums = __riscv_vwadd_vv_i32m1(bsums_i32_0, bsums_i32_1, 8);
+
+        // Accumulation.
+        vint32m1_t sumi_v = __riscv_vmul_vv_i32m1(ls, lsums, 8);
+        vint32m1_t sumi1_v = __riscv_vmul_vv_i32m1(__riscv_vmul_vv_i32m1(ls, delta, 8), bsums, 8);
+
+        // Update sumf.
+        int sumi = __riscv_vmv_x_s_i32m1_i32(__riscv_vredsum_vs_i32m1_i32m1(sumi_v, __riscv_vmv_v_x_i32m1(0.0f, 1), 8));
+        int sumi1 = __riscv_vmv_x_s_i32m1_i32(__riscv_vredsum_vs_i32m1_i32m1(sumi1_v, __riscv_vmv_v_x_i32m1(0.0f, 1), 8));
+        sumf += GGML_CPU_FP16_TO_FP32(x[i].d) * y[i].d * (sumi + IQ1S_DELTA * sumi1);
+    }
+
+    *s = sumf;
+}
+
+void ggml_vec_dot_iq1_s_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+#if defined __riscv_v_intrinsic
+    switch (__riscv_vlenb() * 8) {
+        case 256:
+            ggml_vec_dot_iq1_s_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+        default:
+            ggml_vec_dot_iq1_s_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+    }
+#else
+    ggml_vec_dot_iq1_s_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+#endif
+}
+
+static void ggml_vec_dot_iq1_m_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+    assert(n % QK_K == 0);
+    assert(nrc == 1);
+    UNUSED(nrc);
+    UNUSED(bx);
+    UNUSED(by);
+    UNUSED(bs);
+
+    const block_iq1_m * GGML_RESTRICT x = vx;
+    const block_q8_K  * GGML_RESTRICT y = vy;
+
+    const int nb = n / QK_K;
+
+    iq1m_scale_t scale;
+    float sumf = 0.0f;
+    for (int i = 0; i < nb; ++i) {
+        const int8_t   * q8 = y[i].qs;
+        const uint8_t  * qs = x[i].qs;
+        const uint8_t  * qh = x[i].qh;
+        const uint16_t * sc = (const uint16_t *)x[i].scales;
+
+        scale.u16 = (sc[0] >> 12) | ((sc[1] >> 8) & 0x00f0) | ((sc[2] >> 4) & 0x0f00) | (sc[3] & 0xf000);
+
+        // Accumulators.
+        vint32m2_t acc1 = __riscv_vmv_v_x_i32m2(0, 16);
+        vint32m2_t acc2 = __riscv_vmv_v_x_i32m2(0, 16);
+
+        // We process 4 sub-blocks together.
+        for (int ib = 0; ib < QK_K/128; ib++) {
+            // Load qh for 4 sub-blocks.
+            const vuint8mf4_t qh_8 = __riscv_vle8_v_u8mf4(qh, 8);
+            const vuint16mf2_t qh_16_lo = __riscv_vzext_vf2_u16mf2(qh_8, 8);
+            const vuint16mf2_t qh_16_hi = __riscv_vsll_vx_u16mf2(qh_16_lo, 8, 8);
+            const vuint16m1_t qhb = __riscv_vzext_vf2_u16m1(
+                __riscv_vreinterpret_v_u16mf2_u8mf2(__riscv_vor_vv_u16mf2(qh_16_lo, qh_16_hi, 8)), 16);
+            qh += 8;
+
+            // Prepare grid indices.
+            const vuint16m1_t qsb = __riscv_vzext_vf2_u16m1(__riscv_vle8_v_u8mf2(&qs[0], 16), 16);
+            const vuint16m1_t shift = __riscv_vreinterpret_v_u32m1_u16m1(__riscv_vmv_v_x_u32m1(0x00040008, 8));
+            vuint16m1_t index = __riscv_vor_vv_u16m1(qsb, __riscv_vand_vx_u16m1(__riscv_vsll_vv_u16m1(qhb, shift, 16), 0x700, 16), 16);
+            index = __riscv_vsll_vx_u16m1(index, 3, 16);
+            qs += 16;
+
+            // Load the grid.
+            const vint8m4_t iq1b = __riscv_vreinterpret_v_i64m4_i8m4(__riscv_vreinterpret_v_u64m4_i64m4(
+                __riscv_vluxei16_v_u64m4(iq1s_grid, index, 16)));
+
+            // Prepare the deltas.
+            const vbool16_t mask = __riscv_vmsgtu_vx_u16m1_b16(
+                __riscv_vand_vv_u16m1(qhb, __riscv_vreinterpret_v_u32m1_u16m1(__riscv_vmv_v_x_u32m1(0x00800008, 8)), 16), 0, 16);
+            const vint64m4_t delta_pos = __riscv_vmv_v_x_i64m4(0x0101010101010101, 16);
+            const vint64m4_t delta_neg = __riscv_vmv_v_x_i64m4(0xffffffffffffffff, 16);
+            const vint8m4_t delta = __riscv_vreinterpret_v_i64m4_i8m4(
+                __riscv_vmerge_vvm_i64m4(delta_pos, delta_neg, mask, 16));
+
+            // Load q8 for sub-blocks.
+            const vint8m4_t q8b = __riscv_vle8_v_i8m4(q8, 128);
+            q8 += 128;
+
+            // Calculate the lsums.
+            const vint16m8_t lsum1 = __riscv_vwmul_vv_i16m8(iq1b, q8b, 128);
+            const vint16m8_t lsum2 = __riscv_vwmul_vv_i16m8(delta, q8b, 128);
+
+            // Prepare the scales.
+            const int16_t ls_0_0 = 2*((sc[0] >> 0) & 0x7) + 1;
+            const int16_t ls_0_1 = 2*((sc[0] >> 3) & 0x7) + 1;
+            const int16_t ls_1_0 = 2*((sc[0] >> 6) & 0x7) + 1;
+            const int16_t ls_1_1 = 2*((sc[0] >> 9) & 0x7) + 1;
+            const int16_t ls_2_0 = 2*((sc[1] >> 0) & 0x7) + 1;
+            const int16_t ls_2_1 = 2*((sc[1] >> 3) & 0x7) + 1;
+            const int16_t ls_3_0 = 2*((sc[1] >> 6) & 0x7) + 1;
+            const int16_t ls_3_1 = 2*((sc[1] >> 9) & 0x7) + 1;
+            sc += 2;
+
+            // Accumulate in acc0 and acc1 for each sub-block.
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_0_0, __riscv_vget_v_i16m8_i16m1(lsum1, 0), 16);
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_0_1, __riscv_vget_v_i16m8_i16m1(lsum1, 1), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_0_0, __riscv_vget_v_i16m8_i16m1(lsum2, 0), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_0_1, __riscv_vget_v_i16m8_i16m1(lsum2, 1), 16);
+            //
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_1_0, __riscv_vget_v_i16m8_i16m1(lsum1, 2), 16);
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_1_1, __riscv_vget_v_i16m8_i16m1(lsum1, 3), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_1_0, __riscv_vget_v_i16m8_i16m1(lsum2, 2), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_1_1, __riscv_vget_v_i16m8_i16m1(lsum2, 3), 16);
+            //
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_2_0, __riscv_vget_v_i16m8_i16m1(lsum1, 4), 16);
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_2_1, __riscv_vget_v_i16m8_i16m1(lsum1, 5), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_2_0, __riscv_vget_v_i16m8_i16m1(lsum2, 4), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_2_1, __riscv_vget_v_i16m8_i16m1(lsum2, 5), 16);
+            //
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_3_0, __riscv_vget_v_i16m8_i16m1(lsum1, 6), 16);
+            acc1 = __riscv_vwmacc_vx_i32m2(acc1, ls_3_1, __riscv_vget_v_i16m8_i16m1(lsum1, 7), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_3_0, __riscv_vget_v_i16m8_i16m1(lsum2, 6), 16);
+            acc2 = __riscv_vwmacc_vx_i32m2(acc2, ls_3_1, __riscv_vget_v_i16m8_i16m1(lsum2, 7), 16);
+        }
+
+        // Reduce and accumulate in `sumf`.
+        vint32m1_t one = __riscv_vmv_v_x_i32m1(0, 1);
+        int sumi1 = __riscv_vmv_x_s_i32m1_i32(__riscv_vredsum_vs_i32m2_i32m1(acc1, one, 16));
+        int sumi2 = __riscv_vmv_x_s_i32m1_i32(__riscv_vredsum_vs_i32m2_i32m1(acc2, one, 16));
+        sumf += y[i].d * GGML_CPU_FP16_TO_FP32(scale.f16) * (sumi1 + IQ1M_DELTA * sumi2);
+    }
+
+    *s = sumf;
+}
+
+void ggml_vec_dot_iq1_m_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
+#if defined __riscv_v_intrinsic
+    switch (__riscv_vlenb() * 8) {
+        case 256:
+            ggml_vec_dot_iq1_m_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+        default:
+            ggml_vec_dot_iq1_m_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+            break;
+    }
+#else
+    ggml_vec_dot_iq1_m_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
+#endif
+}

ggml/src/ggml-cpu/repack.cpp

@@ -450,6 +450,208 @@ static void ggml_gemm_q6_K_NxM_q8_K_generic_impl(int                        n,
     }
 }
 
+template <int M, int N>
+static void ggml_gemv_q5_K_NxM_q8_K_generic_impl(int                        n,
+                                                 float * GGML_RESTRICT      s,
+                                                 size_t                     bs,
+                                                 const void * GGML_RESTRICT vx,
+                                                 const void * GGML_RESTRICT vy,
+                                                 int                        nr,
+                                                 int                        nc) {
+    constexpr int         blocklen          = M;
+    constexpr int         ncols_interleaved = N;
+    const int             qk                = QK_K;
+    const int             nb                = n / qk;
+    static const uint32_t kmask1            = 0x3f3f3f3f;
+    static const uint32_t kmask2            = 0x0f0f0f0f;
+    static const uint32_t kmask3            = 0x03030303;
+
+    assert(n % qk == 0);
+    assert(nc % ncols_interleaved == 0);
+
+    UNUSED(bs);
+    UNUSED(nr);
+
+    float    sumf[ncols_interleaved];
+    float    sum_minf[ncols_interleaved];
+    uint32_t utmp[32];
+    int      sumi1;
+    int      sumi2;
+    int      sumi;
+
+    const block_q8_K * a_ptr = (const block_q8_K *) vy;
+    for (int x = 0; x < nc / ncols_interleaved; x++) {
+        const block_q5_Kx8 * b_ptr = (const block_q5_Kx8 *) vx + (x * nb);
+
+        for (int j = 0; j < ncols_interleaved; j++) {
+            sumf[j]     = 0.0;
+            sum_minf[j] = 0.0;
+        }
+        for (int l = 0; l < nb; l++) {
+            for (int sb = 0; sb < 8; sb++) {
+                memcpy(utmp + sb * 4, b_ptr[l].scales + sb * K_SCALE_SIZE, K_SCALE_SIZE);
+                utmp[sb * 4 + 3]      = ((utmp[sb * 4 + 2] >> 4) & kmask2) | (((utmp[sb * 4 + 1] >> 6) & kmask3) << 4);
+                const uint32_t uaux_0 = utmp[sb * 4 + 1] & kmask1;
+                utmp[sb * 4 + 1]      = (utmp[sb * 4 + 2] & kmask2) | (((utmp[sb * 4 + 0] >> 6) & kmask3) << 4);
+                utmp[sb * 4 + 2]      = uaux_0;
+                utmp[sb * 4 + 0] &= kmask1;
+            }
+            for (int k = 0; k < (qk / (2 * blocklen)); k++) {
+                constexpr int scale_stride = 32;
+                uint8_t *     scales_0     = (uint8_t *) utmp + (k / (32 / blocklen)) * scale_stride;
+                uint8_t *     scales_1     = (uint8_t *) utmp + (k / (32 / blocklen)) * scale_stride + 16;
+
+                const int qh_shift = (k / (32 / blocklen)) * 2;
+                for (int j = 0; j < ncols_interleaved; j++) {
+                    sumi1 = 0;
+                    sumi2 = 0;
+                    sumi  = 0;
+                    for (int i = 0; i < blocklen; ++i) {
+                        const int b_qs_offset = k * ncols_interleaved * blocklen + j * blocklen + i;
+
+                        const int qh_idx      = (k * blocklen + i) % 32;
+                        const int qh_chunk    = qh_idx / blocklen;
+                        const int qh_pos      = qh_idx % blocklen;
+                        const int b_qh_offset = qh_chunk * (blocklen * ncols_interleaved) + j * blocklen + qh_pos;
+
+                        const uint8_t qh_val = b_ptr[l].qh[b_qh_offset];
+                        const uint8_t h0     = (qh_val >> qh_shift) & 1;
+                        const uint8_t h1     = (qh_val >> (qh_shift + 1)) & 1;
+
+                        const int v0 = (int8_t) ((b_ptr[l].qs[b_qs_offset] & 0xF) | (h0 << 4));
+                        const int v1 = (int8_t) ((b_ptr[l].qs[b_qs_offset] >> 4) | (h1 << 4));
+
+                        const int q8_offset = (k / (32 / blocklen)) * 64 + (k % (32 / blocklen)) * blocklen + i;
+
+                        sumi1 = (v0 * a_ptr[l].qs[q8_offset]);
+                        sumi2 = (v1 * a_ptr[l].qs[q8_offset + 32]);
+                        sumi1 = sumi1 * scales_0[j];
+                        sumi2 = sumi2 * scales_1[j];
+                        sumi += sumi1 + sumi2;
+                    }
+                    sumf[j] += sumi * GGML_CPU_FP16_TO_FP32(b_ptr[l].d[j]) * a_ptr[l].d;
+                }
+            }
+            for (int sb = 0; sb < 8; sb++) {
+                uint8_t * mins = (uint8_t *) utmp + 8 + sb * 16;
+                for (int j = 0; j < ncols_interleaved; j++) {
+                    sum_minf[j] += mins[j] * (a_ptr[l].bsums[sb * 2] + a_ptr[l].bsums[sb * 2 + 1]) *
+                                   GGML_CPU_FP16_TO_FP32(b_ptr[l].dmin[j]) * a_ptr[l].d;
+                }
+            }
+        }
+        for (int j = 0; j < ncols_interleaved; j++) {
+            s[x * ncols_interleaved + j] = sumf[j] - sum_minf[j];
+        }
+    }
+}
+
+template <int M, int N>
+static void ggml_gemm_q5_K_NxM_q8_K_generic_impl(int                        n,
+                                                 float * GGML_RESTRICT      s,
+                                                 size_t                     bs,
+                                                 const void * GGML_RESTRICT vx,
+                                                 const void * GGML_RESTRICT vy,
+                                                 int                        nr,
+                                                 int                        nc) {
+    constexpr int         blocklen          = M;
+    constexpr int         ncols_interleaved = N;
+    const int             qk                = QK_K;
+    const int             nb                = n / qk;
+    static const uint32_t kmask1            = 0x3f3f3f3f;
+    static const uint32_t kmask2            = 0x0f0f0f0f;
+    static const uint32_t kmask3            = 0x03030303;
+
+    assert(n % qk == 0);
+    assert(nr % 4 == 0);
+    assert(nc % ncols_interleaved == 0);
+
+    float    sumf[4][ncols_interleaved];
+    float    sum_minf[4][ncols_interleaved];
+    uint32_t utmp[32];
+    int      sumi1;
+    int      sumi2;
+    int      sumi;
+
+    for (int y = 0; y < nr / 4; y++) {
+        const block_q8_Kx4 * a_ptr = (const block_q8_Kx4 *) vy + (y * nb);
+        for (int x = 0; x < nc / ncols_interleaved; x++) {
+            const block_q5_Kx8 * b_ptr = (const block_q5_Kx8 *) vx + (x * nb);
+            for (int m = 0; m < 4; m++) {
+                for (int j = 0; j < ncols_interleaved; j++) {
+                    sumf[m][j]     = 0.0;
+                    sum_minf[m][j] = 0.0;
+                }
+            }
+            for (int l = 0; l < nb; l++) {
+                for (int sb = 0; sb < 8; sb++) {
+                    memcpy(utmp + sb * 4, b_ptr[l].scales + sb * K_SCALE_SIZE, K_SCALE_SIZE);
+                    utmp[sb * 4 + 3] = ((utmp[sb * 4 + 2] >> 4) & kmask2) | (((utmp[sb * 4 + 1] >> 6) & kmask3) << 4);
+                    const uint32_t uaux_0 = utmp[sb * 4 + 1] & kmask1;
+                    utmp[sb * 4 + 1]      = (utmp[sb * 4 + 2] & kmask2) | (((utmp[sb * 4 + 0] >> 6) & kmask3) << 4);
+                    utmp[sb * 4 + 2]      = uaux_0;
+                    utmp[sb * 4 + 0] &= kmask1;
+                }
+                for (int k = 0; k < (qk / (2 * blocklen)); k++) {
+                    constexpr int scale_stride = 32;
+                    uint8_t *     scales_0     = (uint8_t *) utmp + (k / (32 / blocklen)) * scale_stride;
+                    uint8_t *     scales_1     = (uint8_t *) utmp + (k / (32 / blocklen)) * scale_stride + 16;
+
+                    const int qh_shift = (k / (32 / blocklen)) * 2;
+                    for (int m = 0; m < 4; m++) {
+                        for (int j = 0; j < ncols_interleaved; j++) {
+                            sumi1 = 0;
+                            sumi2 = 0;
+                            sumi  = 0;
+                            for (int i = 0; i < blocklen; ++i) {
+                                const int b_qs_offset = k * ncols_interleaved * blocklen + j * blocklen + i;
+
+                                const int qh_idx   = (k * blocklen + i) % 32;
+                                const int qh_chunk = qh_idx / blocklen;
+                                const int qh_pos   = qh_idx % blocklen;
+                                const int b_qh_offset =
+                                    qh_chunk * (blocklen * ncols_interleaved) + j * blocklen + qh_pos;
+
+                                const uint8_t qh_val = b_ptr[l].qh[b_qh_offset];
+                                const uint8_t h0     = (qh_val >> qh_shift) & 1;
+                                const uint8_t h1     = (qh_val >> (qh_shift + 1)) & 1;
+
+                                const int v0 = (int8_t) ((b_ptr[l].qs[b_qs_offset] & 0xF) | (h0 << 4));
+                                const int v1 = (int8_t) ((b_ptr[l].qs[b_qs_offset] >> 4) | (h1 << 4));
+
+                                const int q8_offset = (k / (32 / blocklen)) * 256 +
+                                                      (k % (32 / blocklen)) * 4 * blocklen + m * blocklen + i;
+
+                                sumi1 = (v0 * a_ptr[l].qs[q8_offset]);
+                                sumi2 = (v1 * a_ptr[l].qs[q8_offset + 128]);
+                                sumi1 = sumi1 * scales_0[j];
+                                sumi2 = sumi2 * scales_1[j];
+                                sumi += sumi1 + sumi2;
+                            }
+                            sumf[m][j] += sumi * GGML_CPU_FP16_TO_FP32(b_ptr[l].d[j]) * a_ptr[l].d[m];
+                        }
+                    }
+                }
+                for (int sb = 0; sb < 8; sb++) {
+                    uint8_t * mins = (uint8_t *) utmp + 8 + sb * 16;
+                    for (int m = 0; m < 4; m++) {
+                        const int16_t * bsums = a_ptr[l].bsums + (sb * 8) + (m * 4) - ((sb % 2) * 6);
+                        for (int j = 0; j < ncols_interleaved; j++) {
+                            sum_minf[m][j] += mins[j] * (bsums[0] + bsums[1]) *
+                                              GGML_CPU_FP16_TO_FP32(b_ptr[l].dmin[j]) * a_ptr[l].d[m];
+                        }
+                    }
+                }
+            }
+            for (int m = 0; m < 4; m++) {
+                for (int j = 0; j < ncols_interleaved; j++) {
+                    s[(y * 4 + m) * bs + x * ncols_interleaved + j] = sumf[m][j] - sum_minf[m][j];
+                }
+            }
+        }
+    }
+}
+
 extern "C" {
 
 void ggml_gemv_q4_0_4x4_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
@@ -803,98 +1005,12 @@ void ggml_gemv_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs,
     }
 }
 
-void ggml_gemv_q5_K_8x8_q8_K_generic(int                        n,
-                                     float * GGML_RESTRICT      s,
-                                     size_t                     bs,
-                                     const void * GGML_RESTRICT vx,
-                                     const void * GGML_RESTRICT vy,
-                                     int                        nr,
-                                     int                        nc) {
-    const int             qk                = QK_K;
-    const int             nb                = n / qk;
-    const int             ncols_interleaved = 8;
-    const int             blocklen          = 8;
-    static const uint32_t kmask1            = 0x3f3f3f3f;
-    static const uint32_t kmask2            = 0x0f0f0f0f;
-    static const uint32_t kmask3            = 0x03030303;
+void ggml_gemv_q5_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
+    ggml_gemv_q5_K_NxM_q8_K_generic_impl<4, 8>(n, s, bs, vx, vy, nr, nc);
+}
 
-    assert(n % qk == 0);
-    assert(nc % ncols_interleaved == 0);
-
-    UNUSED(bs);
-    UNUSED(nr);
-
-    float    sumf[8];
-    float    sum_minf[8];
-    uint32_t utmp[32];
-    int      sumi1;
-    int      sumi2;
-    int      sumi;
-
-    const block_q8_K * a_ptr = (const block_q8_K *) vy;
-    for (int x = 0; x < nc / ncols_interleaved; x++) {
-        const block_q5_Kx8 * b_ptr = (const block_q5_Kx8 *) vx + (x * nb);
-
-        for (int j = 0; j < ncols_interleaved; j++) {
-            sumf[j]     = 0.0;
-            sum_minf[j] = 0.0;
-        }
-        for (int l = 0; l < nb; l++) {
-            for (int sb = 0; sb < 8; sb++) {
-                memcpy(utmp + sb * 4, b_ptr[l].scales + sb * 12, 12);
-                utmp[sb * 4 + 3]      = ((utmp[sb * 4 + 2] >> 4) & kmask2) | (((utmp[sb * 4 + 1] >> 6) & kmask3) << 4);
-                const uint32_t uaux_0 = utmp[sb * 4 + 1] & kmask1;
-                utmp[sb * 4 + 1]      = (utmp[sb * 4 + 2] & kmask2) | (((utmp[sb * 4 + 0] >> 6) & kmask3) << 4);
-                utmp[sb * 4 + 2]      = uaux_0;
-                utmp[sb * 4 + 0] &= kmask1;
-            }
-            for (int k = 0; k < (qk / (2 * blocklen)); k++) {
-                uint8_t * scales_0 = (uint8_t *) utmp + (k / 4) * 32;
-                uint8_t * scales_1 = (uint8_t *) utmp + (k / 4) * 32 + 16;
-
-                const int qh_shift = (k / 4) * 2;
-                for (int j = 0; j < ncols_interleaved; j++) {
-                    sumi1 = 0;
-                    sumi2 = 0;
-                    sumi  = 0;
-                    for (int i = 0; i < blocklen; ++i) {
-                        const int b_qs_offset = k * ncols_interleaved * blocklen + j * blocklen + i;
-
-                        const int qh_idx      = (k * 8 + i) % 32;
-                        const int qh_chunk    = qh_idx / 8;
-                        const int qh_pos      = qh_idx % 8;
-                        const int b_qh_offset = qh_chunk * 64 + j * 8 + qh_pos;
-
-                        const uint8_t qh_val = b_ptr[l].qh[b_qh_offset];
-                        const uint8_t h0     = (qh_val >> qh_shift) & 1;
-                        const uint8_t h1     = (qh_val >> (qh_shift + 1)) & 1;
-
-                        const int v0 = (int8_t) ((b_ptr[l].qs[b_qs_offset] & 0xF) | (h0 << 4));
-                        const int v1 = (int8_t) ((b_ptr[l].qs[b_qs_offset] >> 4) | (h1 << 4));
-
-                        const int q8_offset = (k >> 2) * 64 + (k % 4) * blocklen + i;
-
-                        sumi1 = (v0 * a_ptr[l].qs[q8_offset]);
-                        sumi2 = (v1 * a_ptr[l].qs[q8_offset + 32]);
-                        sumi1 = sumi1 * scales_0[j];
-                        sumi2 = sumi2 * scales_1[j];
-                        sumi += sumi1 + sumi2;
-                    }
-                    sumf[j] += sumi * GGML_CPU_FP16_TO_FP32(b_ptr[l].d[j]) * a_ptr[l].d;
-                }
-            }
-            for (int sb = 0; sb < 8; sb++) {
-                uint8_t * mins = (uint8_t *) utmp + 8 + sb * 16;
-                for (int j = 0; j < ncols_interleaved; j++) {
-                    sum_minf[j] += mins[j] * (a_ptr[l].bsums[sb * 2] + a_ptr[l].bsums[sb * 2 + 1]) *
-                                   GGML_CPU_FP16_TO_FP32(b_ptr[l].dmin[j]) * a_ptr[l].d;
-                }
-            }
-        }
-        for (int j = 0; j < ncols_interleaved; j++) {
-            s[x * ncols_interleaved + j] = sumf[j] - sum_minf[j];
-        }
-    }
+void ggml_gemv_q5_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
+    ggml_gemv_q5_K_NxM_q8_K_generic_impl<8, 8>(n, s, bs, vx, vy, nr, nc);
 }
 
 
@@ -1494,107 +1610,12 @@ void ggml_gemm_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs,
     }
 }
 
-void ggml_gemm_q5_K_8x8_q8_K_generic(int                        n,
-                                     float * GGML_RESTRICT      s,
-                                     size_t                     bs,
-                                     const void * GGML_RESTRICT vx,
-                                     const void * GGML_RESTRICT vy,
-                                     int                        nr,
-                                     int                        nc) {
-    const int qk                = QK_K;
-    const int nb                = n / qk;
-    const int ncols_interleaved = 8;
-    const int blocklen          = 8;
+void ggml_gemm_q5_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
+    ggml_gemm_q5_K_NxM_q8_K_generic_impl<4, 8>(n, s, bs, vx, vy, nr, nc);
+}
 
-    constexpr uint32_t kmask1 = 0x3f3f3f3f;
-    constexpr uint32_t kmask2 = 0x0f0f0f0f;
-    constexpr uint32_t kmask3 = 0x03030303;
-
-    assert(n % qk == 0);
-    assert(nr % 4 == 0);
-    assert(nc % ncols_interleaved == 0);
-
-    float    sumf[4][8];
-    float    sum_minf[4][8];
-    uint32_t utmp[32];
-    int      sumi1;
-    int      sumi2;
-    int      sumi;
-
-    for (int y = 0; y < nr / 4; y++) {
-        const block_q8_Kx4 * a_ptr = (const block_q8_Kx4 *) vy + (y * nb);
-        for (int x = 0; x < nc / ncols_interleaved; x++) {
-            const block_q5_Kx8 * b_ptr = (const block_q5_Kx8 *) vx + (x * nb);
-            for (int m = 0; m < 4; m++) {
-                for (int j = 0; j < ncols_interleaved; j++) {
-                    sumf[m][j]     = 0.0;
-                    sum_minf[m][j] = 0.0;
-                }
-            }
-            for (int l = 0; l < nb; l++) {
-                for (int sb = 0; sb < 8; sb++) {
-                    memcpy(utmp + sb * 4, b_ptr[l].scales + sb * 12, 12);
-                    utmp[sb * 4 + 3] = ((utmp[sb * 4 + 2] >> 4) & kmask2) | (((utmp[sb * 4 + 1] >> 6) & kmask3) << 4);
-                    const uint32_t uaux_0 = utmp[sb * 4 + 1] & kmask1;
-                    utmp[sb * 4 + 1]      = (utmp[sb * 4 + 2] & kmask2) | (((utmp[sb * 4 + 0] >> 6) & kmask3) << 4);
-                    utmp[sb * 4 + 2]      = uaux_0;
-                    utmp[sb * 4 + 0] &= kmask1;
-                }
-                for (int k = 0; k < (qk / (2 * blocklen)); k++) {
-                    uint8_t * scales_0 = (uint8_t *) utmp + (k / 4) * 32;
-                    uint8_t * scales_1 = (uint8_t *) utmp + (k / 4) * 32 + 16;
-
-                    const int qh_shift = (k / 4) * 2;
-                    for (int m = 0; m < 4; m++) {
-                        for (int j = 0; j < ncols_interleaved; j++) {
-                            sumi1 = 0;
-                            sumi2 = 0;
-                            sumi  = 0;
-                            for (int i = 0; i < blocklen; ++i) {
-                                const int b_qs_offset = k * ncols_interleaved * blocklen + j * blocklen + i;
-
-                                const int qh_idx      = (k * 8 + i) % 32;
-                                const int qh_chunk    = qh_idx / 8;
-                                const int qh_pos      = qh_idx % 8;
-                                const int b_qh_offset = qh_chunk * 64 + j * 8 + qh_pos;
-
-                                const uint8_t qh_val = b_ptr[l].qh[b_qh_offset];
-                                const uint8_t h0     = (qh_val >> qh_shift) & 1;
-                                const uint8_t h1     = (qh_val >> (qh_shift + 1)) & 1;
-
-                                const int v0 = (int8_t) ((b_ptr[l].qs[b_qs_offset] & 0xF) | (h0 << 4));
-                                const int v1 = (int8_t) ((b_ptr[l].qs[b_qs_offset] >> 4) | (h1 << 4));
-
-                                const int q8_offset = (k >> 2) * 256 + (k % 4) * 4 * blocklen + m * blocklen + i;
-
-                                sumi1 = (v0 * a_ptr[l].qs[q8_offset]);
-                                sumi2 = (v1 * a_ptr[l].qs[q8_offset + 128]);
-                                sumi1 = sumi1 * scales_0[j];
-                                sumi2 = sumi2 * scales_1[j];
-                                sumi += sumi1 + sumi2;
-                            }
-                            sumf[m][j] += sumi * GGML_CPU_FP16_TO_FP32(b_ptr[l].d[j]) * a_ptr[l].d[m];
-                        }
-                    }
-                }
-                for (int sb = 0; sb < 8; sb++) {
-                    uint8_t * mins = (uint8_t *) utmp + 8 + sb * 16;
-                    for (int m = 0; m < 4; m++) {
-                        const int16_t * bsums = a_ptr[l].bsums + (sb * 8) + (m * 4) - ((sb % 2) * 6);
-                        for (int j = 0; j < ncols_interleaved; j++) {
-                            sum_minf[m][j] += mins[j] * (bsums[0] + bsums[1]) *
-                                              GGML_CPU_FP16_TO_FP32(b_ptr[l].dmin[j]) * a_ptr[l].d[m];
-                        }
-                    }
-                }
-            }
-            for (int m = 0; m < 4; m++) {
-                for (int j = 0; j < ncols_interleaved; j++) {
-                    s[(y * 4 + m) * bs + x * ncols_interleaved + j] = sumf[m][j] - sum_minf[m][j];
-                }
-            }
-        }
-    }
+void ggml_gemm_q5_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
+    ggml_gemm_q5_K_NxM_q8_K_generic_impl<8, 8>(n, s, bs, vx, vy, nr, nc);
 }
 
 void ggml_gemm_q6_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
@@ -2029,18 +2050,16 @@ static block_q5_Kx8 make_block_q5_Kx8(block_q5_K * in, unsigned int blck_size_in
 
     const int end = QK_K * 4 / blck_size_interleave;
 
-    // Interleave Q5_K quants by taking 8 bytes at a time
+    // Interleave Q5_K quants by taking blck_size_interleave bytes at a time
     for (int i = 0; i < end; ++i) {
         int src_id     = i % 8;
         int src_offset = (i / 8) * blck_size_interleave;
         int dst_offset = i * blck_size_interleave;
 
-        uint64_t elems;
-        memcpy(&elems, &in[src_id].qs[src_offset], sizeof(uint64_t));
-        memcpy(&out.qs[dst_offset], &elems, sizeof(uint64_t));
+        memcpy(&out.qs[dst_offset], &in[src_id].qs[src_offset], blck_size_interleave);
     }
 
-    // Repeat for low bits 8 bytes at a time as well, since
+    // Repeat for high bits with the same chunk size, since
     // the high bits are interleaved in Q5_K and the index is
     // qh_idx = (qs_idx % 32);
     // qh_val = qh[qh_idx] >> (qs_idx / 32);
@@ -2049,9 +2068,7 @@ static block_q5_Kx8 make_block_q5_Kx8(block_q5_K * in, unsigned int blck_size_in
         int src_offset = (i / 8) * blck_size_interleave;
         int dst_offset = i * blck_size_interleave;
 
-        uint64_t elems;
-        memcpy(&elems, &in[src_id].qh[src_offset], sizeof(uint64_t));
-        memcpy(&out.qh[dst_offset], &elems, sizeof(uint64_t));
+        memcpy(&out.qh[dst_offset], &in[src_id].qh[src_offset], blck_size_interleave);
     }
 
     // The below logic is copied over from Q4_K
@@ -2249,7 +2266,7 @@ static int repack_q5_K_to_q5_K_8_bl(struct ggml_tensor *       t,
                                     const void * GGML_RESTRICT data,
                                     size_t                     data_size) {
     GGML_ASSERT(t->type == GGML_TYPE_Q5_K);
-    GGML_ASSERT(interleave_block == 8);
+    GGML_ASSERT(interleave_block == 4 || interleave_block == 8);
     constexpr int nrows_interleaved = 8;
 
     block_q5_Kx8 *     dst = (block_q5_Kx8 *) t->data;
@@ -2523,6 +2540,10 @@ template <> int repack<block_q2_K, 8, 8>(struct ggml_tensor * t, const void * da
     return repack_q2_K_to_q2_K_8_bl(t, 8, data, data_size);
 }
 
+template <> int repack<block_q5_K, 4, 8>(struct ggml_tensor * t, const void * data, size_t data_size) {
+    return repack_q5_K_to_q5_K_8_bl(t, 4, data, data_size);
+}
+
 template <> int repack<block_q5_K, 8, 8>(struct ggml_tensor * t, const void * data, size_t data_size) {
     return repack_q5_K_to_q5_K_8_bl(t, 8, data, data_size);
 }
@@ -2591,6 +2612,10 @@ template <> void gemv<block_q4_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t
     ggml_gemv_q4_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
 }
 
+template <> void gemv<block_q5_K, 4, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
+    ggml_gemv_q5_K_8x4_q8_K(n, s, bs, vx, vy, nr, nc);
+}
+
 template <> void gemv<block_q5_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
     ggml_gemv_q5_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
 }
@@ -2654,6 +2679,10 @@ template <> void gemm<block_q4_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t
     ggml_gemm_q4_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
 }
 
+template <> void gemm<block_q5_K, 4, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
+    ggml_gemm_q5_K_8x4_q8_K(n, s, bs, vx, vy, nr, nc);
+}
+
 template <> void gemm<block_q5_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
     ggml_gemm_q5_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
 }
@@ -3068,6 +3097,7 @@ static const ggml::cpu::tensor_traits * ggml_repack_get_optimal_repack_type(cons
     static const ggml::cpu::repack::tensor_traits<block_q4_K, 8, 8, GGML_TYPE_Q8_K> q4_K_8x8_q8_K;
 
     // instance for Q5_K
+    static const ggml::cpu::repack::tensor_traits<block_q5_K, 4, 8, GGML_TYPE_Q8_K> q5_K_8x4_q8_K;
     static const ggml::cpu::repack::tensor_traits<block_q5_K, 8, 8, GGML_TYPE_Q8_K> q5_K_8x8_q8_K;
 
     // instance for Q6_K
@@ -3130,6 +3160,11 @@ static const ggml::cpu::tensor_traits * ggml_repack_get_optimal_repack_type(cons
                 return &q5_K_8x8_q8_K;
             }
         }
+        if (ggml_cpu_has_neon() && ggml_cpu_has_dotprod()) {
+            if (cur->ne[1] % 8 == 0) {
+                return &q5_K_8x4_q8_K;
+            }
+        }
     } else if (cur->type == GGML_TYPE_Q6_K) {
         if (ggml_cpu_has_neon() && ggml_cpu_has_matmul_int8()) {
             if (cur->ne[1] % 8 == 0) {

ggml/src/ggml-cpu/repack.h

@@ -111,6 +111,7 @@ void ggml_gemv_q4_0_8x8_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const vo
 void ggml_gemv_q2_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q4_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q4_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
+void ggml_gemv_q5_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q5_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q6_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q6_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
@@ -122,6 +123,7 @@ void ggml_gemm_q4_0_8x8_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const vo
 void ggml_gemm_q2_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q4_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q4_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
+void ggml_gemm_q5_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q5_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q6_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q6_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
@@ -143,6 +145,7 @@ void ggml_gemv_q4_0_8x8_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs,
 void ggml_gemv_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q4_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q4_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
+void ggml_gemv_q5_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q5_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q6_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemv_q6_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
@@ -154,6 +157,7 @@ void ggml_gemm_q4_0_8x8_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs,
 void ggml_gemm_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q4_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q4_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
+void ggml_gemm_q5_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q5_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q6_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
 void ggml_gemm_q6_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);

ggml/src/ggml-cuda/common.cuh

@@ -1149,8 +1149,7 @@ struct ggml_cuda_graph {
     size_t num_nodes = 0;
     std::vector<cudaGraphNode_t> nodes;
     bool disable_due_to_gpu_arch = false;
-    bool disable_due_to_too_many_updates = false;
-    int number_consecutive_updates = 0;
+    bool warmup_complete = false;
     std::vector<ggml_cuda_graph_node_properties> props;
 
     // these are extra tensors (inputs) that participate in the ggml graph but are not nodes
@@ -1159,21 +1158,9 @@ struct ggml_cuda_graph {
     // ref: https://github.com/ggml-org/llama.cpp/pull/19165
     std::vector<ggml_cuda_graph_node_properties> extra;
 
-    void record_update(bool use_graph, bool update_required) {
-        if (use_graph && update_required) {
-            number_consecutive_updates++;
-        } else {
-            number_consecutive_updates = 0;
-        }
-        if (number_consecutive_updates >= 4) {
-            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to too many consecutive updates\n", __func__);
-            disable_due_to_too_many_updates = true;
-        }
-    }
-
     bool is_enabled() const {
         static const bool disable_cuda_graphs_due_to_env = (getenv("GGML_CUDA_DISABLE_GRAPHS") != nullptr);
-        return !(disable_due_to_gpu_arch || disable_cuda_graphs_due_to_env || disable_due_to_too_many_updates);
+        return !(disable_due_to_gpu_arch || disable_cuda_graphs_due_to_env);
     }
 #endif
 };

ggml/src/ggml-cuda/ggml-cuda.cu

@@ -2979,10 +2979,6 @@ static bool ggml_cuda_graph_update_required(ggml_backend_cuda_context * cuda_ctx
     const void * graph_key = ggml_cuda_graph_get_key(cgraph);
     ggml_cuda_graph * graph = cuda_ctx->cuda_graph(graph_key);
 
-    if (graph->instance == nullptr) {
-        res = true;
-    }
-
     // Check if the graph size has changed
     if (graph->props.size() != (size_t)cgraph->n_nodes) {
         res = true;
@@ -3931,14 +3927,35 @@ static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t backend,
 #ifdef USE_CUDA_GRAPH
     graph_key = ggml_cuda_graph_get_key(cgraph);
 
-    use_cuda_graph = ggml_cuda_graph_set_enabled(cuda_ctx, graph_key);
+    ggml_cuda_graph_set_enabled(cuda_ctx, graph_key);
 
     ggml_cuda_graph * graph = cuda_ctx->cuda_graph(graph_key);
     if (graph->is_enabled()) {
-        cuda_graph_update_required = ggml_cuda_graph_update_required(cuda_ctx, cgraph);
-        use_cuda_graph             = ggml_cuda_graph_check_compability(cgraph);
+        const bool graph_compatible = ggml_cuda_graph_check_compability(cgraph);
+        if (graph_compatible) {
+            const bool properties_changed = ggml_cuda_graph_update_required(cuda_ctx, cgraph);
 
-        graph->record_update(use_cuda_graph, cuda_graph_update_required);
+            if (!graph->warmup_complete) {
+                // Warmup: need at least 2 calls with no property change on the 2nd call
+                if (!properties_changed) {
+                    graph->warmup_complete = true;
+                    GGML_LOG_DEBUG("%s: CUDA graph warmup complete\n", __func__);
+                    use_cuda_graph = true;
+                    cuda_graph_update_required = true;
+                }
+                // else: properties changed or first call - execute directly (use_cuda_graph stays false)
+            } else {
+                // Post-warmup: normal CUDA graph operation
+                if (properties_changed) {
+                    // Properties changed - reset warmup, execute directly until stable again
+                    graph->warmup_complete = false;
+                    GGML_LOG_DEBUG("%s: CUDA graph warmup reset\n", __func__);
+                } else {
+                    use_cuda_graph = true;
+                    cuda_graph_update_required = graph->instance == nullptr;
+                }
+            }
+        }
     }
 #endif // USE_CUDA_GRAPH
 

ggml/src/ggml-hexagon/ggml-hexagon.cpp

@@ -1749,23 +1749,6 @@ static inline bool ggml_backend_buffer_is_hexagon_repack(const struct ggml_backe
     return b->buft->iface.alloc_buffer == ggml_backend_hexagon_repack_buffer_type_alloc_buffer;
 }
 
-static bool hex_supported_dims2(const struct ggml_tensor * x, const struct ggml_tensor * y) {
-    if (x->ne[0] != y->ne[0]) {
-        return false;
-    }
-    if (x->ne[1] != y->ne[1]) {
-        return false;
-    }
-    if (x->ne[2] != y->ne[2]) {
-        return false;
-    }
-    if (x->ne[3] != y->ne[3]) {
-        return false;
-    }
-
-    return true;
-}
-
 static bool ggml_hexagon_supported_flash_attn_ext(const struct ggml_hexagon_session * sess, const struct ggml_tensor * op) {
     const struct ggml_tensor * src0 = op->src[0];
     const struct ggml_tensor * src1 = op->src[1];
@@ -1797,43 +1780,6 @@ static bool ggml_hexagon_supported_flash_attn_ext(const struct ggml_hexagon_sess
     return opt_experimental;
 }
 
-static bool hex_supported_src0_type(ggml_type t) {
-    return t == GGML_TYPE_F32;
-}
-
-static bool hex_supported_src1_type(ggml_type t) {
-    return t == GGML_TYPE_F32;
-}
-
-static bool hex_supported_src2_type(ggml_type t) {
-    return t == GGML_TYPE_F32;
-}
-
-static bool hex_supported_src1_type2(ggml_type t) {
-    return t == GGML_TYPE_F16;
-}
-
-static bool hex_supported_src1_type3(ggml_type t) {
-    return t == GGML_TYPE_I32;
-}
-
-static bool hex_supported_dst_type(ggml_type t) {
-    return t == GGML_TYPE_F32;
-}
-
-static bool hex_supported_dims(const struct ggml_tensor * x, const struct ggml_tensor * y) {
-    // TODO: support broadcast for ne[2 and 3]
-    if (x->ne[0] != y->ne[0]) {
-        return false;
-    }
-    if (x->ne[2] != y->ne[2]) {
-        return false;
-    }
-    if (x->ne[3] != y->ne[3]) {
-        return false;
-    }
-    return true;
-}
 
 static bool ggml_hexagon_supported_mul_mat(const struct ggml_hexagon_session * sess, const struct ggml_tensor * dst) {
     const struct ggml_tensor * src0 = dst->src[0];
@@ -1919,19 +1865,19 @@ static bool ggml_hexagon_supported_binary(const struct ggml_hexagon_session * se
     const struct ggml_tensor * src1 = op->src[1];
     const struct ggml_tensor * dst  = op;
 
-    if (!hex_supported_src0_type(src0->type)) {
+    if (src0->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_src1_type(src1->type)) {
+    if (src1->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dst_type(dst->type)) {
+    if (dst->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dims2(src0, dst)) {
+    if (!ggml_are_same_shape(src0, dst)) {
         return false;
     }
-    if (!ggml_can_repeat(src1, src0)) {
+    if (!ggml_can_repeat(src1, src0) || ggml_is_permuted(src1)) {
         return false;
     }
 
@@ -1943,16 +1889,16 @@ static bool ggml_hexagon_supported_add_id(const struct ggml_hexagon_session * se
     const struct ggml_tensor * src1 = op->src[1];
     const struct ggml_tensor * dst  = op;
 
-    if (!hex_supported_src0_type(src0->type)) {
+    if (src0->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_src1_type(src1->type)) {
+    if (src1->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dst_type(dst->type)) {
+    if (dst->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dims2(src0, dst)) {
+    if (!ggml_are_same_shape(src0, dst)) {
         return false;
     }
 
@@ -1968,13 +1914,13 @@ static bool ggml_hexagon_supported_unary(const struct ggml_hexagon_session * ses
     const struct ggml_tensor * src0 = op->src[0];
     const struct ggml_tensor * dst  = op;
 
-    if (!hex_supported_src0_type(src0->type)) {
+    if (src0->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dst_type(dst->type)) {
+    if (dst->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dims2(src0, dst)) {
+    if (!ggml_are_same_shape(src0, dst)) {
         return false;
     }
 
@@ -1990,10 +1936,10 @@ static bool ggml_hexagon_supported_sum_rows(const struct ggml_hexagon_session *
     const struct ggml_tensor * src0 = op->src[0];
     const struct ggml_tensor * dst  = op;
 
-    if (!hex_supported_src0_type(src0->type)) {
+    if (src0->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dst_type(dst->type)) {
+    if (dst->type != GGML_TYPE_F32) {
         return false;
     }
 
@@ -2011,10 +1957,10 @@ static bool ggml_hexagon_supported_activations(const struct ggml_hexagon_session
     const struct ggml_tensor * src1 = op->src[1];
     const struct ggml_tensor * dst  = op;
 
-    if (!hex_supported_src0_type(src0->type)) {
+    if (src0->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dst_type(dst->type)) {
+    if (dst->type != GGML_TYPE_F32) {
         return false;
     }
 
@@ -2023,10 +1969,10 @@ static bool ggml_hexagon_supported_activations(const struct ggml_hexagon_session
     }
 
     if (src1) {
-        if (!hex_supported_src1_type(src1->type)) {
+        if (src1->type != GGML_TYPE_F32) {
             return false;
         }
-        if (!hex_supported_dims2(src0, src1)) {
+        if (!ggml_are_same_shape(src0, src1)) {
             return false;
         }
         if (!ggml_is_contiguous(src1)) {
@@ -2047,15 +1993,15 @@ static bool ggml_hexagon_supported_softmax(const struct ggml_hexagon_session * s
         return false;  // FIXME: add support for sinks
     }
 
-    if (!hex_supported_src0_type(src0->type)) {
+    if (src0->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_dst_type(dst->type)) {
+    if (dst->type != GGML_TYPE_F32) {
         return false;
     }
 
     if (src1) {
-        if (!hex_supported_src1_type(src1->type) && !hex_supported_src1_type2(src1->type)) {
+        if (src1->type != GGML_TYPE_F32 && src1->type != GGML_TYPE_F16) {
             return false;
         }
         if (src0->ne[0] != src1->ne[0]) {
@@ -2162,17 +2108,17 @@ static bool ggml_hexagon_supported_rope(const struct ggml_hexagon_session * sess
     const struct ggml_tensor * src2 = op->src[2];
     const struct ggml_tensor * dst  = op;
 
-    if (!hex_supported_src0_type(src0->type)) {
+    if (src0->type != GGML_TYPE_F32) {
         return false;  // FIXME: add support for GGML_TYPE_F16 for src0
     }
-    if (!hex_supported_dst_type(dst->type)) {
+    if (dst->type != GGML_TYPE_F32) {
         return false;
     }
-    if (!hex_supported_src1_type3(src1->type)) {
+    if (src1->type != GGML_TYPE_I32) {
         return false;
     }
     if (src2) {
-        if (!hex_supported_src2_type(src2->type)) {
+        if (src2->type != GGML_TYPE_F32) {
             return false;
         }
         int n_dims = op_params[1];

ggml/src/ggml-hexagon/htp/act-ops.c

@@ -69,27 +69,45 @@
     const uint32_t nb2 = dst->nb[2];   \
     const uint32_t nb3 = dst->nb[3];
 
-static void glu_swiglu_f32_per_thread(const struct htp_tensor * src0,
-                                       const struct htp_tensor * src1,
-                                       struct htp_tensor *       dst,
-                                       const int32_t *           op_params,
-                                       struct htp_spad *         src0_spad,
-                                       struct htp_spad *         src1_spad,
-                                       struct htp_spad *         dst_spad,
-                                       uint32_t                  nth,
-                                       uint32_t                  ith,
-                                       uint32_t                  src0_nrows_per_thread,
-                                       dma_queue *               dma_queue) {
+struct htp_act_context {
+    struct htp_ops_context *  octx;
+
+    // Precomputed values
+    const uint8_t *           data_src0;
+    const uint8_t *           data_src1;
+    uint8_t *                 data_dst;
+
+    size_t                    src0_row_size;
+    size_t                    src1_row_size;
+    size_t                    dst_row_size;
+
+    size_t                    src0_row_size_aligned;
+    size_t                    src1_row_size_aligned;
+    size_t                    dst_row_size_aligned;
+
+    size_t                    src0_spad_half_size;
+    size_t                    src1_spad_half_size;
+    size_t                    dst_spad_half_size;
+
+    uint32_t                  block;
+    uint32_t                  src0_nrows;
+    uint32_t                  src0_nrows_per_thread;
+    int                       nc;
+};
+
+static void glu_swiglu_f32_per_thread(unsigned int nth, unsigned int ith, void * data) {
+    struct htp_act_context * actx = (struct htp_act_context *) data;
+    const struct htp_tensor * src0 = &actx->octx->src0;
+    const struct htp_tensor * src1 = &actx->octx->src1;
+    const struct htp_tensor * dst  = &actx->octx->dst;
     htp_act_preamble3;
 
-    size_t src0_row_size = nb01;
-    size_t src1_row_size = nb11;
-    size_t dst_row_size  = nb1;
-
-
-
-    const uint32_t src0_nrows = ne01 * ne02 * ne03;  // src0 rows
+    size_t src0_row_size = actx->src0_row_size;
+    size_t src1_row_size = actx->src1_row_size;
+    size_t dst_row_size  = actx->dst_row_size;
 
+    const uint32_t src0_nrows = actx->src0_nrows;
+    const uint32_t src0_nrows_per_thread = actx->src0_nrows_per_thread;
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
 
@@ -101,43 +119,34 @@ static void glu_swiglu_f32_per_thread(const struct htp_tensor * src0,
     uint64_t t1, t2;
     t1 = HAP_perf_get_qtimer_count();
 
-    const uint8_t * restrict data_src0 = (const uint8_t *) src0->data;
-    const uint8_t * restrict data_src1 = (const uint8_t *) src1->data;
-    uint8_t * restrict data_dst        = (uint8_t *) dst->data;
+    const uint8_t * restrict data_src0 = actx->data_src0;
+    const uint8_t * restrict data_src1 = actx->data_src1;
+    uint8_t * restrict data_dst        = actx->data_dst;
 
-    const bool src1_valid = src1->ne[0];
-    const int  nc         = (src1_valid) ? ne00 : ne00 / 2;
-    if (!src1_valid) {
-        const int32_t swapped = op_params[1];
-        data_src1             = data_src0;
-        src1_row_size         = src0_row_size;
+    const int  nc = actx->nc;
 
-        const size_t nc_in_bytes = nc * SIZEOF_FP32;
-        data_src0 += swapped ? nc_in_bytes : 0;
-        data_src1 += swapped ? 0 : nc_in_bytes;
-    }
+    const size_t src0_row_size_aligned = actx->src0_row_size_aligned;
+    const size_t src1_row_size_aligned = actx->src1_row_size_aligned;
+    const size_t dst_row_size_aligned  = actx->dst_row_size_aligned;
 
-    const size_t src0_row_size_aligned = hex_round_up(src0_row_size, VLEN);
-    const size_t src1_row_size_aligned = hex_round_up(src1_row_size, VLEN);
-    const size_t dst_row_size_aligned  = hex_round_up(dst_row_size, VLEN);
+    uint8_t * restrict src0_spad_data = actx->octx->src0_spad.data + (ith * actx->octx->src0_spad.size_per_thread);
+    uint8_t * restrict src1_spad_data = actx->octx->src1_spad.data + (ith * actx->octx->src1_spad.size_per_thread);
+    uint8_t * restrict dst_spad_data  = actx->octx->dst_spad.data + (ith * actx->octx->dst_spad.size_per_thread);
 
-    uint8_t * restrict src0_spad_data = src0_spad->data + (ith * src0_spad->size_per_thread);
-    uint8_t * restrict src1_spad_data = src1_spad->data + (ith * src1_spad->size_per_thread);
-    uint8_t * restrict dst_spad_data  = dst_spad->data + (ith * dst_spad->size_per_thread);
+    size_t src0_spad_half_size = actx->src0_spad_half_size;
+    size_t src1_spad_half_size = actx->src1_spad_half_size;
+    size_t dst_spad_half_size  = actx->dst_spad_half_size;
 
-    // While given src0_spad->size_per_thread, divide it to two ping-pong buffer for src0
-    size_t src0_spad_half_size = src0_spad->size_per_thread / 2;
-    size_t src1_spad_half_size = src1_spad->size_per_thread / 2;
-    size_t dst_spad_half_size  = dst_spad->size_per_thread / 2;
-
-    const int BLOCK = src0_spad_half_size / src0_row_size_aligned;  // How many rows can we process in one block
+    const int BLOCK = actx->block;
     if (BLOCK == 0) {
         FARF(ERROR,
              "swiglu-f32 : current VTCM reservation %zu is too small for even 1 row per thread, needed at least %zu\n",
-             src0_spad->size_per_thread, src0_row_size_aligned);
+             actx->octx->src0_spad.size_per_thread, src0_row_size_aligned);
         return;
     }
 
+    dma_queue * dma_queue = actx->octx->ctx->dma[ith];
+
     // See discussion: https://github.com/ggml-org/llama.cpp/pull/18151#issuecomment-3678235379
     for (uint32_t ir = src0_start_row, spad_idx = 0; ir < src0_end_row && spad_idx < 2; ir += BLOCK, spad_idx++) {
         const uint32_t block_size = MIN(BLOCK, src0_end_row - ir);
@@ -196,27 +205,22 @@ static void glu_swiglu_f32_per_thread(const struct htp_tensor * src0,
          (unsigned) HAP_perf_qtimer_count_to_us(t2 - t1));
 }
 
-static void glu_swiglu_oai_f32_per_thread(const struct htp_tensor * src0,
-                                           const struct htp_tensor * src1,
-                                           struct htp_tensor *       dst,
-                                           const int32_t *           op_params,
-                                           struct htp_spad *         src0_spad,
-                                           struct htp_spad *         src1_spad,
-                                           struct htp_spad *         dst_spad,
-                                           uint32_t                  nth,
-                                           uint32_t                  ith,
-                                           uint32_t                  src0_nrows_per_thread,
-                                           dma_queue *               dma_queue) {
+static void glu_swiglu_oai_f32_per_thread(unsigned int nth, unsigned int ith, void * data) {
+    struct htp_act_context * actx = (struct htp_act_context *) data;
+    const struct htp_tensor * src0 = &actx->octx->src0;
+    const struct htp_tensor * src1 = &actx->octx->src1;
+    const struct htp_tensor * dst  = &actx->octx->dst;
     htp_act_preamble3;
 
     uint64_t t1, t2;
     t1 = HAP_perf_get_qtimer_count();
 
-    size_t src0_row_size = nb01;
-    size_t src1_row_size = nb11;
-    size_t dst_row_size  = nb1;
+    size_t src0_row_size = actx->src0_row_size;
+    size_t src1_row_size = actx->src1_row_size;
+    size_t dst_row_size  = actx->dst_row_size;
 
-    const uint32_t src0_nrows = ne01 * ne02 * ne03;  // src0 rows
+    const uint32_t src0_nrows = actx->src0_nrows;
+    const uint32_t src0_nrows_per_thread = actx->src0_nrows_per_thread;
 
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
@@ -226,45 +230,36 @@ static void glu_swiglu_oai_f32_per_thread(const struct htp_tensor * src0,
         return;
     }
 
-    const uint8_t * restrict data_src0 = (const uint8_t *) src0->data;
-    const uint8_t * restrict data_src1 = (const uint8_t *) src1->data;
-    uint8_t * restrict data_dst        = (uint8_t *) dst->data;
+    const uint8_t * restrict data_src0 = actx->data_src0;
+    const uint8_t * restrict data_src1 = actx->data_src1;
+    uint8_t * restrict data_dst        = actx->data_dst;
 
-    const bool src1_valid = src1->ne[0];
-    const int  nc         = (src1_valid) ? ne00 : ne00 / 2;
-    if (!src1_valid) {
-        const int32_t swapped = op_params[1];
-        data_src1             = data_src0;
-        src1_row_size         = src0_row_size;
+    const int nc = actx->nc;
 
-        const size_t nc_in_bytes = nc * SIZEOF_FP32;
-        data_src0 += swapped ? nc_in_bytes : 0;
-        data_src1 += swapped ? 0 : nc_in_bytes;
-    }
+    const size_t src0_row_size_aligned = actx->src0_row_size_aligned;
+    const size_t src1_row_size_aligned = actx->src1_row_size_aligned;
+    const size_t dst_row_size_aligned  = actx->dst_row_size_aligned;
 
-    const size_t src0_row_size_aligned = hex_round_up(src0_row_size, VLEN);
-    const size_t src1_row_size_aligned = hex_round_up(src1_row_size, VLEN);
-    const size_t dst_row_size_aligned  = hex_round_up(dst_row_size, VLEN);
+    uint8_t * restrict src0_spad_data = actx->octx->src0_spad.data + (ith * actx->octx->src0_spad.size_per_thread);
+    uint8_t * restrict src1_spad_data = actx->octx->src1_spad.data + (ith * actx->octx->src1_spad.size_per_thread);
+    uint8_t * restrict dst_spad_data  = actx->octx->dst_spad.data + (ith * actx->octx->dst_spad.size_per_thread);
 
-    uint8_t * restrict src0_spad_data = src0_spad->data + (ith * src0_spad->size_per_thread);
-    uint8_t * restrict src1_spad_data = src1_spad->data + (ith * src1_spad->size_per_thread);
-    uint8_t * restrict dst_spad_data  = dst_spad->data + (ith * dst_spad->size_per_thread);
+    size_t src0_spad_half_size = actx->src0_spad_half_size;
+    size_t src1_spad_half_size = actx->src1_spad_half_size;
+    size_t dst_spad_half_size  = actx->dst_spad_half_size;
 
-    // While given src0_spad->size_per_thread, divide it to two ping-pong buffer for src0
-    size_t src0_spad_half_size = src0_spad->size_per_thread / 2;
-    size_t src1_spad_half_size = src1_spad->size_per_thread / 2;
-    size_t dst_spad_half_size  = dst_spad->size_per_thread / 2;
-
-    const int BLOCK = src0_spad_half_size / src0_row_size_aligned;  // How many rows can we process in one block
+    const int BLOCK = actx->block;
     if (BLOCK == 0) {
         FARF(ERROR,
              "swiglu-oai-f32 : current VTCM reservation %zu is too small for even 1 row per thread, needed at least "
              "%zu\n",
-             src0_spad->size_per_thread, src0_row_size_aligned);
+             actx->octx->src0_spad.size_per_thread, src0_row_size_aligned);
         return;
     }
-    const float alpha = ((const float *) (op_params))[2];
-    const float limit = ((const float *) (op_params))[3];
+    const float alpha = ((const float *) (actx->octx->op_params))[2];
+    const float limit = ((const float *) (actx->octx->op_params))[3];
+
+    dma_queue * dma_queue = actx->octx->ctx->dma[ith];
 
     // See discussion: https://github.com/ggml-org/llama.cpp/pull/18151#issuecomment-3678235379
     for (uint32_t ir = src0_start_row, spad_idx = 0; ir < src0_end_row && spad_idx < 2; ir += BLOCK, spad_idx++) {
@@ -335,26 +330,22 @@ static void glu_swiglu_oai_f32_per_thread(const struct htp_tensor * src0,
 }
 
 
-static void unary_gelu_f32_per_thread(const struct htp_tensor * src0,
-                                       struct htp_tensor *       dst,
-                                       const int32_t *           op_params,
-                                       struct htp_spad *         src0_spad,
-                                       struct htp_spad *         dst_spad,
-                                       uint32_t                  nth,
-                                       uint32_t                  ith,
-                                       uint32_t                  src0_nrows_per_thread,
-                                       dma_queue *               dma_queue) {
+static void unary_gelu_f32_per_thread(unsigned int nth, unsigned int ith, void * data) {
+    struct htp_act_context * actx = (struct htp_act_context *) data;
+    const struct htp_tensor * src0 = &actx->octx->src0;
+    const struct htp_tensor * dst  = &actx->octx->dst;
     htp_act_preamble2;
 
     uint64_t t1, t2;
     t1 = HAP_perf_get_qtimer_count();
 
-    const size_t src0_row_size = nb01;
-    const size_t dst_row_size  = nb1;
-    const size_t src0_row_size_aligned = hex_round_up(src0_row_size, VLEN);
-    const size_t dst_row_size_aligned  = hex_round_up(dst_row_size, VLEN);
+    const size_t src0_row_size = actx->src0_row_size;
+    const size_t dst_row_size  = actx->dst_row_size;
+    const size_t src0_row_size_aligned = actx->src0_row_size_aligned;
+    const size_t dst_row_size_aligned  = actx->dst_row_size_aligned;
 
-    const uint32_t src0_nrows = ne01 * ne02 * ne03;
+    const uint32_t src0_nrows = actx->src0_nrows;
+    const uint32_t src0_nrows_per_thread = actx->src0_nrows_per_thread;
 
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
@@ -364,25 +355,29 @@ static void unary_gelu_f32_per_thread(const struct htp_tensor * src0,
         return;
     }
 
-    const uint8_t * data_src0 = (const uint8_t *) src0->data;
-    uint8_t * data_dst        = (uint8_t *) dst->data;
+    const uint8_t * data_src0 = actx->data_src0;
+    uint8_t * data_dst        = actx->data_dst;
 
-    uint8_t * src0_spad_data = src0_spad->data + (ith * src0_spad->size_per_thread);
-    uint8_t * dst_spad_data  = dst_spad->data  + (ith * dst_spad->size_per_thread);
+    // nc/ne0 matches.
+    const int ne0_val = actx->nc; // == dst->ne[0]
 
-    // While given src0_spad->size_per_thread, divide it to two ping-pong buffer for src0
-    size_t src0_spad_half_size = src0_spad->size_per_thread / 2;
-    size_t dst_spad_half_size  = dst_spad->size_per_thread  / 2;
+    uint8_t * src0_spad_data = actx->octx->src0_spad.data + (ith * actx->octx->src0_spad.size_per_thread);
+    uint8_t * dst_spad_data  = actx->octx->dst_spad.data  + (ith * actx->octx->dst_spad.size_per_thread);
+
+    size_t src0_spad_half_size = actx->src0_spad_half_size;
+    size_t dst_spad_half_size  = actx->dst_spad_half_size;
 
     // In gelu = x*sigmoid(x*1.702)
-    const int BLOCK = src0_spad_half_size / src0_row_size_aligned; // How many rows can we process in one block
+    const int BLOCK = actx->block;
 
     if (BLOCK == 0) {
         FARF(ERROR, "gelu-f32 : current VTCM reservation %zu is too small for even 1 row per thread, needed at least %zu\n",
-                src0_spad->size_per_thread, src0_row_size_aligned);
+                actx->octx->src0_spad.size_per_thread, src0_row_size_aligned);
         return;
     }
 
+    dma_queue * dma_queue = actx->octx->ctx->dma[ith];
+
     // See discussion: https://github.com/ggml-org/llama.cpp/pull/18151#issuecomment-3678235379
     for (uint32_t ir = src0_start_row, spad_idx = 0; ir < src0_end_row && spad_idx < 2; ir += BLOCK, spad_idx++) {
         const uint32_t block_size = MIN(BLOCK, src0_end_row - ir);
@@ -408,9 +403,9 @@ static void unary_gelu_f32_per_thread(const struct htp_tensor * src0,
             float* dst_spad_ptr        = dst_spad  + ib * (dst_row_size_aligned  / sizeof(float));
 
             // gelu = x * sigmoid(1.702 * x) // current implementation
-            hvx_mul_scalar_f32((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, (float) 1.702, ne0);
-            hvx_sigmoid_f32_aa((uint8_t *) dst_spad_ptr, (const uint8_t *) dst_spad_ptr, ne0);
-            hvx_mul_f32_aaa((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, (const uint8_t *) dst_spad_ptr, ne0);
+            hvx_mul_scalar_f32((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, (float) 1.702, ne0_val);
+            hvx_sigmoid_f32_aa((uint8_t *) dst_spad_ptr, (const uint8_t *) dst_spad_ptr, ne0_val);
+            hvx_mul_f32_aaa((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, (const uint8_t *) dst_spad_ptr, ne0_val);
         }
 
         dma_queue_push_vtcm_to_ddr(dma_queue,
@@ -435,34 +430,23 @@ static void unary_gelu_f32_per_thread(const struct htp_tensor * src0,
          ne03, src0_start_row, src0_end_row, ne0, ne1, ne2, ne3, (unsigned) HAP_perf_qtimer_count_to_us(t2 - t1));
 }
 
-static void unary_gelu_f32(unsigned int n, unsigned int i, void * data) {
-    struct htp_ops_context * octx = (struct htp_ops_context *) data;
-    unary_gelu_f32_per_thread(&octx->src0, &octx->dst, octx->op_params, &octx->src0_spad, &octx->dst_spad, n, i,
-                               octx->src0_nrows_per_thread, octx->ctx->dma[i]);
-}
 
-
-
-static void unary_silu_f32_per_thread(const struct htp_tensor * src0,
-                                       struct htp_tensor *       dst,
-                                       const int32_t *           op_params,
-                                       struct htp_spad *         src0_spad,
-                                       struct htp_spad *         dst_spad,
-                                       uint32_t                  nth,
-                                       uint32_t                  ith,
-                                       uint32_t                  src0_nrows_per_thread,
-                                       dma_queue *               dma_queue) {
+static void unary_silu_f32_per_thread(unsigned int nth, unsigned int ith, void * data) {
+    struct htp_act_context * actx = (struct htp_act_context *) data;
+    const struct htp_tensor * src0 = &actx->octx->src0;
+    const struct htp_tensor * dst  = &actx->octx->dst;
     htp_act_preamble2;
 
     uint64_t t1, t2;
     t1 = HAP_perf_get_qtimer_count();
 
-    const size_t src0_row_size = nb01;
-    const size_t dst_row_size  = nb1;
-    const size_t src0_row_size_aligned = hex_round_up(src0_row_size, VLEN);
-    const size_t dst_row_size_aligned  = hex_round_up(dst_row_size, VLEN);
+    const size_t src0_row_size = actx->src0_row_size;
+    const size_t dst_row_size  = actx->dst_row_size;
+    const size_t src0_row_size_aligned = actx->src0_row_size_aligned;
+    const size_t dst_row_size_aligned  = actx->dst_row_size_aligned;
 
-    const uint32_t src0_nrows = ne01 * ne02 * ne03;
+    const uint32_t src0_nrows = actx->src0_nrows;
+    const uint32_t src0_nrows_per_thread = actx->src0_nrows_per_thread;
 
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
@@ -472,24 +456,27 @@ static void unary_silu_f32_per_thread(const struct htp_tensor * src0,
         return;
     }
 
-    const uint8_t * data_src0 = (const uint8_t *) src0->data;
-    uint8_t * data_dst        = (uint8_t *) dst->data;
+    const uint8_t * data_src0 = actx->data_src0;
+    uint8_t * data_dst        = actx->data_dst;
 
-    uint8_t * src0_spad_data = src0_spad->data + (ith * src0_spad->size_per_thread);
-    uint8_t * dst_spad_data  = dst_spad->data  + (ith * dst_spad->size_per_thread);
+    const int ne0_val = actx->nc; // == dst->ne[0]
 
-    // While given src0_spad->size_per_thread, divide it to two ping-pong buffer for src0
-    size_t src0_spad_half_size = src0_spad->size_per_thread / 2;
-    size_t dst_spad_half_size  = dst_spad->size_per_thread  / 2;
+    uint8_t * src0_spad_data = actx->octx->src0_spad.data + (ith * actx->octx->src0_spad.size_per_thread);
+    uint8_t * dst_spad_data  = actx->octx->dst_spad.data  + (ith * actx->octx->dst_spad.size_per_thread);
 
-    const int BLOCK = src0_spad_half_size / src0_row_size_aligned; // How many rows can we process in one block
+    size_t src0_spad_half_size = actx->src0_spad_half_size;
+    size_t dst_spad_half_size  = actx->dst_spad_half_size;
+
+    const int BLOCK = actx->block;
 
     if (BLOCK == 0) {
         FARF(ERROR, "silu-f32 : current VTCM reservation %zu is too small for even 1 row per thread, needed at least %zu\n",
-                src0_spad->size_per_thread, src0_row_size_aligned);
+                actx->octx->src0_spad.size_per_thread, src0_row_size_aligned);
         return;
     }
 
+    dma_queue * dma_queue = actx->octx->ctx->dma[ith];
+
     // See discussion: https://github.com/ggml-org/llama.cpp/pull/18151#issuecomment-3678235379
     for (uint32_t ir = src0_start_row, spad_idx = 0; ir < src0_end_row && spad_idx < 2; ir += BLOCK, spad_idx++) {
         const uint32_t block_size = MIN(BLOCK, src0_end_row - ir);
@@ -515,8 +502,8 @@ static void unary_silu_f32_per_thread(const struct htp_tensor * src0,
             float* dst_spad_ptr        = dst_spad  + ib * (dst_row_size_aligned  / sizeof(float));
 
             // silu = x * sigmoid(x)
-            hvx_sigmoid_f32_aa((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, ne0);
-            hvx_mul_f32_aaa((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, (const uint8_t *) dst_spad_ptr, ne0);
+            hvx_sigmoid_f32_aa((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, ne0_val);
+            hvx_mul_f32_aaa((uint8_t *) dst_spad_ptr, (const uint8_t *) src0_spad_ptr, (const uint8_t *) dst_spad_ptr, ne0_val);
         }
 
         dma_queue_push_vtcm_to_ddr(dma_queue,
@@ -544,27 +531,22 @@ static void unary_silu_f32_per_thread(const struct htp_tensor * src0,
 static const float GELU_COEF_A     = 0.044715f;
 static const float SQRT_2_OVER_PI  = 0.79788456080286535587989211986876f;
 
-static void glu_geglu_f32_per_thread(const struct htp_tensor * src0,
-                                       const struct htp_tensor * src1,
-                                       struct htp_tensor *       dst,
-                                       const int32_t *           op_params,
-                                       struct htp_spad *         src0_spad,
-                                       struct htp_spad *         src1_spad,
-                                       struct htp_spad *         dst_spad,
-                                       uint32_t                  nth,
-                                       uint32_t                  ith,
-                                       uint32_t                  src0_nrows_per_thread,
-                                       dma_queue *               dma_queue) {
+static void glu_geglu_f32_per_thread(unsigned int nth, unsigned int ith, void * data) {
+    struct htp_act_context * actx = (struct htp_act_context *) data;
+    const struct htp_tensor * src0 = &actx->octx->src0;
+    const struct htp_tensor * src1 = &actx->octx->src1;
+    const struct htp_tensor * dst  = &actx->octx->dst;
     htp_act_preamble3;
 
-    size_t src0_row_size = nb01;
-    size_t src1_row_size = nb11;
-    size_t dst_row_size  = nb1;
+    size_t src0_row_size = actx->src0_row_size;
+    size_t src1_row_size = actx->src1_row_size;
+    size_t dst_row_size  = actx->dst_row_size;
 
     uint64_t t1, t2;
     t1 = HAP_perf_get_qtimer_count();
 
-    const uint32_t src0_nrows = ne01 * ne02 * ne03;  // src0 rows
+    const uint32_t src0_nrows = actx->src0_nrows;
+    const uint32_t src0_nrows_per_thread = actx->src0_nrows_per_thread;
 
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
@@ -574,43 +556,34 @@ static void glu_geglu_f32_per_thread(const struct htp_tensor * src0,
         return;
     }
 
-    const uint8_t * restrict data_src0 = (const uint8_t *) src0->data;
-    const uint8_t * restrict data_src1 = (const uint8_t *) src1->data;
-    uint8_t * restrict data_dst        = (uint8_t *) dst->data;
+    const uint8_t * restrict data_src0 = actx->data_src0;
+    const uint8_t * restrict data_src1 = actx->data_src1;
+    uint8_t * restrict data_dst        = actx->data_dst;
 
-    const bool src1_valid = src1->ne[0];
-    const int  nc         = (src1_valid) ? ne00 : ne00 / 2;
-    if (!src1_valid) {
-        const int32_t swapped = op_params[1];
-        data_src1             = data_src0;
-        src1_row_size         = src0_row_size;
+    const int nc = actx->nc;
 
-        const size_t nc_in_bytes = nc * SIZEOF_FP32;
-        data_src0 += swapped ? nc_in_bytes : 0;
-        data_src1 += swapped ? 0 : nc_in_bytes;
-    }
+    const size_t src0_row_size_aligned = actx->src0_row_size_aligned;
+    const size_t src1_row_size_aligned = actx->src1_row_size_aligned;
+    const size_t dst_row_size_aligned  = actx->dst_row_size_aligned;
 
-    const size_t src0_row_size_aligned = hex_round_up(src0_row_size, VLEN);
-    const size_t src1_row_size_aligned = hex_round_up(src1_row_size, VLEN);
-    const size_t dst_row_size_aligned  = hex_round_up(dst_row_size, VLEN);
+    uint8_t * restrict src0_spad_data = actx->octx->src0_spad.data + (ith * actx->octx->src0_spad.size_per_thread);
+    uint8_t * restrict src1_spad_data = actx->octx->src1_spad.data + (ith * actx->octx->src1_spad.size_per_thread);
+    uint8_t * restrict dst_spad_data  = actx->octx->dst_spad.data + (ith * actx->octx->dst_spad.size_per_thread);
 
-    uint8_t * restrict src0_spad_data = src0_spad->data + (ith * src0_spad->size_per_thread);
-    uint8_t * restrict src1_spad_data = src1_spad->data + (ith * src1_spad->size_per_thread);
-    uint8_t * restrict dst_spad_data  = dst_spad->data + (ith * dst_spad->size_per_thread);
+    size_t src0_spad_half_size = actx->src0_spad_half_size;
+    size_t src1_spad_half_size = actx->src1_spad_half_size;
+    size_t dst_spad_half_size  = actx->dst_spad_half_size;
 
-    // While given src0_spad->size_per_thread, divide it to two ping-pong buffer for src0
-    size_t src0_spad_half_size = src0_spad->size_per_thread / 2;
-    size_t src1_spad_half_size = src1_spad->size_per_thread / 2;
-    size_t dst_spad_half_size  = dst_spad->size_per_thread / 2;
-
-    const int BLOCK = src0_spad_half_size / src0_row_size_aligned;  // How many rows can we process in one block
+    const int BLOCK = actx->block;
     if (BLOCK == 0) {
         FARF(ERROR,
              "geglu-f32 : current VTCM reservation %zu is too small for even 1 row per thread, needed at least %zu\n",
-             src0_spad->size_per_thread, src0_row_size_aligned);
+             actx->octx->src0_spad.size_per_thread, src0_row_size_aligned);
         return;
     }
 
+    dma_queue * dma_queue = actx->octx->ctx->dma[ith];
+
     // See discussion: https://github.com/ggml-org/llama.cpp/pull/18151#issuecomment-3678235379
     for (uint32_t ir = src0_start_row, spad_idx = 0; ir < src0_end_row && spad_idx < 2; ir += BLOCK, spad_idx++) {
         const uint32_t block_size = MIN(BLOCK, src0_end_row - ir);
@@ -678,33 +651,7 @@ static void glu_geglu_f32_per_thread(const struct htp_tensor * src0,
          (unsigned) HAP_perf_qtimer_count_to_us(t2 - t1));
 }
 
-static void unary_silu_f32(unsigned int n, unsigned int i, void * data) {
-    struct htp_ops_context * octx = (struct htp_ops_context *) data;
-    unary_silu_f32_per_thread(&octx->src0, &octx->dst, octx->op_params, &octx->src0_spad, &octx->dst_spad, n, i,
-                               octx->src0_nrows_per_thread, octx->ctx->dma[i]);
-}
-
-static void glu_swiglu_f32(unsigned int n, unsigned int i, void * data) {
-    struct htp_ops_context * octx = (struct htp_ops_context *) data;
-    glu_swiglu_f32_per_thread(&octx->src0, &octx->src1, &octx->dst, octx->op_params, &octx->src0_spad,
-                               &octx->src1_spad, &octx->dst_spad, n, i, octx->src0_nrows_per_thread, octx->ctx->dma[i]);
-}
-
-static void glu_swiglu_oai_f32(unsigned int n, unsigned int i, void * data) {
-    struct htp_ops_context * octx = (struct htp_ops_context *) data;
-    glu_swiglu_oai_f32_per_thread(&octx->src0, &octx->src1, &octx->dst, octx->op_params, &octx->src0_spad,
-                                   &octx->src1_spad, &octx->dst_spad, n, i, octx->src0_nrows_per_thread, octx->ctx->dma[i]);
-}
-
-static void glu_geglu_f32(unsigned int n, unsigned int i, void * data) {
-    struct htp_ops_context * octx = (struct htp_ops_context *) data;
-    glu_geglu_f32_per_thread(&octx->src0, &octx->src1, &octx->dst, octx->op_params, &octx->src0_spad,
-                               &octx->src1_spad, &octx->dst_spad, n, i, octx->src0_nrows_per_thread, octx->ctx->dma[i]);
-}
-
 static int execute_op_activations_f32(struct htp_ops_context * octx) {
-    int err = HTP_STATUS_OK;
-
     const struct htp_tensor * src0 = &octx->src0;
     const struct htp_tensor * src1 = &octx->src1;
     struct htp_tensor *       dst  = &octx->dst;
@@ -719,26 +666,26 @@ static int execute_op_activations_f32(struct htp_ops_context * octx) {
 
     switch (octx->op) {
         case HTP_OP_UNARY_SILU:
-            act_op_func = unary_silu_f32;
+            act_op_func = (worker_callback_t)unary_silu_f32_per_thread;
             op_type     = "silu-f32";
             break;
 
         case HTP_OP_GLU_SWIGLU:
-            act_op_func = glu_swiglu_f32;
+            act_op_func = (worker_callback_t)glu_swiglu_f32_per_thread;
             op_type     = "swiglu-f32";
             break;
 
         case HTP_OP_GLU_SWIGLU_OAI:
-            act_op_func = glu_swiglu_oai_f32;
+            act_op_func = (worker_callback_t)glu_swiglu_oai_f32_per_thread;
             op_type     = "swiglu-oai-f32";
             break;
         case HTP_OP_UNARY_GELU:
-            act_op_func = unary_gelu_f32;
+            act_op_func = (worker_callback_t)unary_gelu_f32_per_thread;
             op_type     = "gelu-f32";
             break;
 
         case HTP_OP_GLU_GEGLU:
-            act_op_func = glu_geglu_f32;
+            act_op_func = (worker_callback_t)glu_geglu_f32_per_thread;
             op_type     = "geglu-f32";
             break;
         default:
@@ -797,13 +744,58 @@ static int execute_op_activations_f32(struct htp_ops_context * octx) {
              octx->src0_spad.size, octx->src1_spad.size, octx->dst_spad.size);
     }
 
-    if (!(octx->flags & HTP_OPFLAGS_SKIP_COMPUTE)) {
-        uint32_t n_jobs = MIN(n_threads, src0_nrows);
-        octx->src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
-        worker_pool_run_func(octx->ctx->worker_pool, act_op_func, octx, n_jobs);
+    if ((octx->flags & HTP_OPFLAGS_SKIP_COMPUTE)) {
+        return HTP_STATUS_OK;
     }
 
-    return err;
+    uint32_t n_jobs = MIN(n_threads, src0_nrows);
+
+    // Prepare context
+    struct htp_act_context actx;
+    actx.octx = octx;
+
+    actx.src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
+
+    actx.src0_row_size = src0_row_size;
+    actx.src1_row_size = src1_row_size;
+    actx.dst_row_size  = dst_row_size;
+
+    actx.src0_row_size_aligned = src0_row_size_aligned;
+    actx.src1_row_size_aligned = src1_row_size_aligned;
+    actx.dst_row_size_aligned  = dst_row_size_aligned;
+
+    actx.src0_spad_half_size = octx->src0_spad.size_per_thread / 2;
+    actx.src1_spad_half_size = octx->src1_spad.size_per_thread / 2;
+    actx.dst_spad_half_size  = octx->dst_spad.size_per_thread / 2;
+
+    actx.block = actx.src0_spad_half_size / actx.src0_row_size_aligned;
+    actx.src0_nrows = src0_nrows;
+
+    actx.nc = dst->ne[0];
+
+    // Pointers and GLU logic
+    const uint8_t * data_src0 = (const uint8_t *) src0->data;
+    const uint8_t * data_src1 = (const uint8_t *) src1->data;
+
+    if (!src1_valid && (octx->op == HTP_OP_GLU_SWIGLU || octx->op == HTP_OP_GLU_SWIGLU_OAI || octx->op == HTP_OP_GLU_GEGLU)) {
+         const int32_t swapped = octx->op_params[1];
+         data_src1 = data_src0;
+         actx.src1_row_size = actx.src0_row_size;
+
+         size_t nc_in_bytes = actx.nc * SIZEOF_FP32;
+         if (swapped) {
+             data_src0 += nc_in_bytes;
+         } else {
+             data_src1 += nc_in_bytes;
+         }
+    }
+
+    actx.data_src0 = data_src0;
+    actx.data_src1 = data_src1;
+    actx.data_dst  = (uint8_t *) dst->data;
+
+    worker_pool_run_func(octx->ctx->worker_pool, act_op_func, &actx, n_jobs);
+    return HTP_STATUS_OK;
 }
 
 int op_activations(struct htp_ops_context * octx) {

ggml/src/ggml-hexagon/htp/get-rows-ops.c

@@ -15,6 +15,13 @@
 #include "htp-ops.h"
 #include "hvx-utils.h"
 
+struct get_rows_context {
+    struct htp_ops_context * octx;
+    uint32_t src1_nrows_per_thread;
+    struct fastdiv_values get_rows_div_ne10;
+    struct fastdiv_values get_rows_div_ne10_ne11;
+};
+
 #define get_rows_preamble \
     const uint32_t ne00 = octx->src0.ne[0]; \
     const uint32_t ne01 = octx->src0.ne[1]; \
@@ -39,20 +46,22 @@
                                             \
     const uint32_t nr = ne10 * ne11 * ne12;
 
-static int get_rows_thread_f32_f32(struct htp_ops_context * octx, const int nth, const int ith) {
+static void get_rows_thread_f32_f32(unsigned int nth, unsigned int ith, void *data) {
+    struct get_rows_context * grctx = (struct get_rows_context *)data;
+    struct htp_ops_context * octx = grctx->octx;
     get_rows_preamble;
 
     // parallelize by src1 elements (which correspond to dst rows)
-    const uint32_t dr  = octx->src1_nrows_per_thread;
+    const uint32_t dr  = grctx->src1_nrows_per_thread;
     const uint32_t ir0 = dr * ith;
     const uint32_t ir1 = (ir0 + dr < nr) ? (ir0 + dr) : nr;
 
     const bool is_i32 = (octx->src1.type == HTP_TYPE_I32);
 
     for (uint32_t i = ir0; i < ir1; ++i) {
-        const uint32_t i12 = fastdiv(i, &octx->get_rows_div_ne10_ne11);
+        const uint32_t i12 = fastdiv(i, &grctx->get_rows_div_ne10_ne11);
         const uint32_t rem = i - i12 * ne11 * ne10;
-        const uint32_t i11 = fastdiv(rem, &octx->get_rows_div_ne10);
+        const uint32_t i11 = fastdiv(rem, &grctx->get_rows_div_ne10);
         const uint32_t i10 = rem - i11 * ne10;
 
         const uintptr_t src1_addr = octx->src1.data + i10*nb10 + i11*nb11 + i12*nb12;
@@ -68,12 +77,6 @@ static int get_rows_thread_f32_f32(struct htp_ops_context * octx, const int nth,
         const uintptr_t dst_ptr  = octx->dst.data  + i10*nb1  + i11*nb2  + i12*nb3;
         hvx_copy_f32_uu((uint8_t *)dst_ptr, (const uint8_t *)src0_ptr, ne00);
     }
-
-    return HTP_STATUS_OK;
-}
-
-static void get_rows_work_f32_f32(unsigned int n, unsigned int i, void *data) {
-    get_rows_thread_f32_f32((struct htp_ops_context *) data, n, i);
 }
 
 int op_get_rows(struct htp_ops_context * octx) {
@@ -95,12 +98,14 @@ int op_get_rows(struct htp_ops_context * octx) {
         return HTP_STATUS_OK;
     }
 
-    octx->get_rows_div_ne10      = init_fastdiv_values(octx->src1.ne[0]);
-    octx->get_rows_div_ne10_ne11 = init_fastdiv_values(octx->src1.ne[0] * octx->src1.ne[1]);
+    struct get_rows_context grctx;
+    grctx.octx = octx;
+    grctx.get_rows_div_ne10      = init_fastdiv_values(octx->src1.ne[0]);
+    grctx.get_rows_div_ne10_ne11 = init_fastdiv_values(octx->src1.ne[0] * octx->src1.ne[1]);
 
     const uint32_t n_jobs = MIN(nr, octx->n_threads);
-    octx->src1_nrows_per_thread = (nr + n_jobs - 1) / n_jobs;
+    grctx.src1_nrows_per_thread = (nr + n_jobs - 1) / n_jobs;
 
-    worker_pool_run_func(octx->ctx->worker_pool, get_rows_work_f32_f32, octx, n_jobs);
+    worker_pool_run_func(octx->ctx->worker_pool, get_rows_thread_f32_f32, &grctx, n_jobs);
     return HTP_STATUS_OK;
 }

ggml/src/ggml-hexagon/htp/hex-dma.h

@@ -102,7 +102,7 @@ static inline bool dma_queue_push(dma_queue * q,
     dmlink(q->tail, desc);
     q->tail = desc;
 
-    // FARF(ERROR, "dma-push: i %u len %u dst %p src %p\n", q->push_idx, len, dst, src);
+    // FARF(ERROR, "dma-push: i %u width %u nrows %d dst %p src %p\n", q->push_idx, width, nrows, dptr.dst, dptr.src);
     q->push_idx = (q->push_idx + 1) & q->idx_mask;
     return true;
 }
@@ -144,11 +144,37 @@ static inline dma_ptr dma_queue_pop(dma_queue * q) {
 
     dptr = q->dptr[q->pop_idx];
 
-    // FARF(ERROR, "dma-pop: i %u dst %p\n", q->pop_idx, dst);
+    // FARF(ERROR, "dma-pop: i %u dst %p src %p\n", q->pop_idx, dptr.dst, dptr.src);
     q->pop_idx = (q->pop_idx + 1) & q->idx_mask;
     return dptr;
 }
 
+static inline dma_ptr dma_queue_pop_nowait(dma_queue * q) {
+    dma_ptr dptr  = { NULL };
+
+    if (q->push_idx == q->pop_idx) {
+        return dptr;
+    }
+
+    dptr = q->dptr[q->pop_idx];
+
+    // FARF(ERROR, "dma-pop-nowait: i %u dst %p src %p\n", q->pop_idx, dptr.dst, dptr.src);
+    q->pop_idx = (q->pop_idx + 1) & q->idx_mask;
+    return dptr;
+}
+
+static inline bool dma_queue_empty(dma_queue * q) {
+    return q->push_idx == q->pop_idx;
+}
+
+static inline uint32_t dma_queue_depth(dma_queue * q) {
+    return (q->push_idx - q->pop_idx) & q->idx_mask;
+}
+
+static inline uint32_t dma_queue_capacity(dma_queue * q) {
+    return q->capacity;
+}
+
 #ifdef __cplusplus
 }  // extern "C"
 #endif

ggml/src/ggml-hexagon/htp/htp-ops.h

@@ -44,32 +44,6 @@ struct htp_ops_context {
     uint32_t src0_nrows_per_thread;
     uint32_t src1_nrows_per_thread;
 
-    struct fastdiv_values src0_div1;  // fastdiv values for ne1
-    struct fastdiv_values src0_div2;  // fastdiv values for ne2
-    struct fastdiv_values src0_div3;  // fastdiv values for ne3
-    struct fastdiv_values src0_div21; // fastdiv values for ne2 * ne1
-
-    struct fastdiv_values src1_div1;  // fastdiv values for ne1
-    struct fastdiv_values src1_div2;  // fastdiv values for ne2
-    struct fastdiv_values src1_div3;  // fastdiv values for ne3
-    struct fastdiv_values src1_div21; // fastdiv values for ne2 * ne1
-
-    struct fastdiv_values src3_div1;  // fastdiv values for ne1
-    struct fastdiv_values src3_div2;  // fastdiv values for ne2
-    struct fastdiv_values src3_div3;  // fastdiv values for ne3
-    struct fastdiv_values src3_div21; // fastdiv values for ne2 * ne1
-
-    struct fastdiv_values broadcast_rk2;
-    struct fastdiv_values broadcast_rk3;
-    struct fastdiv_values broadcast_rv2;
-    struct fastdiv_values broadcast_rv3;
-
-    struct fastdiv_values set_rows_div_ne12; // fastdiv values for ne12
-    struct fastdiv_values set_rows_div_ne11; // fastdiv values for ne11
-
-    struct fastdiv_values get_rows_div_ne10;      // fastdiv values for ne10
-    struct fastdiv_values get_rows_div_ne10_ne11; // fastdiv values for ne10 * ne11
-
     uint32_t flags;
 };
 

ggml/src/ggml-hexagon/htp/matmul-ops.c

@@ -49,62 +49,6 @@ struct htp_matmul_context {
     struct fastdiv_values mm_div_r3;
 };
 
-// vdelta control to replicate first 4x fp32 values across lanes
-static const uint8_t __attribute__((aligned(128))) repl_4x_f32[128] = {
-    0x00, 0x00, 0x00, 0x00, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x10, 0x10, 0x10,
-    0x10, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x20, 0x20,
-    0x20, 0x20, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x10, 0x10, 0x10, 0x10, 0x04,
-    0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x40, 0x40, 0x40, 0x40,
-    0x44, 0x44, 0x44, 0x44, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x10, 0x10, 0x10, 0x10, 0x04, 0x04, 0x04,
-    0x04, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x20, 0x20, 0x20, 0x20, 0x04, 0x04,
-    0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x10, 0x10, 0x10, 0x10,
-};
-
-// vdelta control to replicate and interleave first 8x fp32 values across lanes
-static const uint8_t __attribute__((aligned(128))) repl_interleave_8x_f32[128] = {
-    0x00, 0x00, 0x00, 0x00, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x00, 0x00, 0x00,
-    0x00, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x20, 0x20,
-    0x20, 0x20, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x20, 0x20, 0x20, 0x20, 0x04,
-    0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x40, 0x40, 0x40, 0x40,
-    0x44, 0x44, 0x44, 0x44, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x40, 0x40, 0x40, 0x40, 0x44, 0x44, 0x44,
-    0x44, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x20, 0x20, 0x20, 0x20, 0x04, 0x04,
-    0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x20, 0x20, 0x20, 0x20,
-};
-
-// vdelta control to replicate first fp32 value across all elements
-static const uint8_t __attribute__((aligned(128))) repl_1x_f32[128] = {
-    0x00, 0x00, 0x00, 0x00, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x10, 0x10, 0x10,
-    0x10, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x20, 0x20, 0x20, 0x20, 0x04, 0x04,
-    0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x10, 0x10, 0x10, 0x10, 0x04, 0x04, 0x04, 0x04, 0x08,
-    0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x40, 0x40, 0x40, 0x40, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08,
-    0x04, 0x04, 0x04, 0x04, 0x10, 0x10, 0x10, 0x10, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04,
-    0x04, 0x20, 0x20, 0x20, 0x20, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04, 0x10, 0x10,
-    0x10, 0x10, 0x04, 0x04, 0x04, 0x04, 0x08, 0x08, 0x08, 0x08, 0x04, 0x04, 0x04, 0x04,
-};
-
-// vdelta control to replicate first fp16 value across all elements
-static const uint8_t __attribute__((aligned(128))) repl_1x_f16[128] = {
-    0x00, 0x00, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x10, 0x10, 0x02,
-    0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x20, 0x20, 0x02, 0x02, 0x04, 0x04,
-    0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x10, 0x10, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08,
-    0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x40, 0x40, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02,
-    0x04, 0x04, 0x02, 0x02, 0x10, 0x10, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02,
-    0x02, 0x20, 0x20, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x10, 0x10,
-    0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-};
-
-// vdelta control to replicate first fp16 value across all elements
-static const uint8_t __attribute__((aligned(128))) repl_2x_f16[128] = {
-    0x00, 0x00, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-    0x10, 0x10, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-    0x20, 0x20, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-    0x10, 0x10, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-    0x00, 0x00, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-    0x10, 0x10, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-    0x20, 0x20, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-    0x10, 0x10, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02, 0x08, 0x08, 0x02, 0x02, 0x04, 0x04, 0x02, 0x02,
-};
-
 // vdelta control to expand first 32 e8m0 values into 32 uint32 elements
 static const uint8_t __attribute__((aligned(128))) expand_x32_e8m0[128] = {
     0x00, 0x00, 0x00, 0x00, 0x01, 0x04, 0x00, 0x00, 0x02, 0x00, 0x08, 0x08, 0x01, 0x02, 0x00, 0x04, 0x04, 0x00, 0x00,
@@ -2067,10 +2011,10 @@ static inline void quantize_block_f32_q8x1(float * restrict x, uint8_t * restric
     HVX_Vector vx3_qf = Q6_Vqf32_vsub_VsfVsf(vx[3], zero);  // 32 elements
 
     // Convert to QF32
-    HVX_Vector vmax0_qf = Q6_Vqf32_vsub_VsfVsf(vmax0_sf, zero);
-    HVX_Vector vmax1_qf = Q6_Vqf32_vsub_VsfVsf(vmax1_sf, zero);
-    HVX_Vector vmax2_qf = Q6_Vqf32_vsub_VsfVsf(vmax2_sf, zero);
-    HVX_Vector vmax3_qf = Q6_Vqf32_vsub_VsfVsf(vmax3_sf, zero);
+    HVX_Vector vmax0_qf = Q6_Vqf32_vsub_VsfVsf(vmax0_sf, zero); // replicated over all lanes
+    HVX_Vector vmax1_qf = Q6_Vqf32_vsub_VsfVsf(vmax1_sf, zero); // replicated over all lanes
+    HVX_Vector vmax2_qf = Q6_Vqf32_vsub_VsfVsf(vmax2_sf, zero); // replicated over all lanes
+    HVX_Vector vmax3_qf = Q6_Vqf32_vsub_VsfVsf(vmax3_sf, zero); // replicated over all lanes
 
     // Combine and convert to fp16
     HVX_Vector vmax01_hf = Q6_Vh_vdeal_Vh(Q6_Vhf_equals_Wqf32(Q6_W_vcombine_VV(vmax1_qf, vmax0_qf)));
@@ -2080,11 +2024,6 @@ static inline void quantize_block_f32_q8x1(float * restrict x, uint8_t * restric
     HVX_Vector vx01_hf = Q6_Vh_vdeal_Vh(Q6_Vhf_equals_Wqf32(Q6_W_vcombine_VV(vx1_qf, vx0_qf)));
     HVX_Vector vx23_hf = Q6_Vh_vdeal_Vh(Q6_Vhf_equals_Wqf32(Q6_W_vcombine_VV(vx3_qf, vx2_qf)));
 
-    // Replicate first fp16 scale across all lanes
-    HVX_Vector ctrl = *(const HVX_Vector *) repl_2x_f16;
-    vmax01_hf         = Q6_V_vdelta_VV(vmax01_hf, ctrl);
-    vmax23_hf         = Q6_V_vdelta_VV(vmax23_hf, ctrl);
-
     HVX_Vector vd01_qf16 = Q6_Vqf16_vmpy_VhfVhf(vmax01_hf, Q6_Vh_vsplat_R(0x2008));  // 1.0 / 127.0
     HVX_Vector vd23_qf16 = Q6_Vqf16_vmpy_VhfVhf(vmax23_hf, Q6_Vh_vsplat_R(0x2008));  // 1.0 / 127.0
     HVX_Vector vd01_hf   = Q6_Vhf_equals_Vqf16(vd01_qf16);
@@ -2130,13 +2069,8 @@ static inline void quantize_block_f32_q8x2(float * restrict x, uint8_t * restric
     HVX_Vector vx23_hf = Q6_Vh_vdeal_Vh(Q6_Vhf_equals_Wqf32(Q6_W_vcombine_VV(vx3_qf, vx2_qf)));
 
     // Compute max and scale
-    HVX_Vector vmax01_hf = hvx_vec_reduce_max_f16(hvx_vec_abs_f16(vx01_hf));
-    HVX_Vector vmax23_hf = hvx_vec_reduce_max_f16(hvx_vec_abs_f16(vx23_hf));
-
-    // Replicate first fp16 scale across all lanes
-    HVX_Vector ctrl = *(const HVX_Vector *) repl_1x_f16;
-    vmax01_hf         = Q6_V_vdelta_VV(vmax01_hf, ctrl);
-    vmax23_hf         = Q6_V_vdelta_VV(vmax23_hf, ctrl);
+    HVX_Vector vmax01_hf = hvx_vec_reduce_max_f16(hvx_vec_abs_f16(vx01_hf)); // replicated over all lanes
+    HVX_Vector vmax23_hf = hvx_vec_reduce_max_f16(hvx_vec_abs_f16(vx23_hf)); // replicated over all lanes
 
     HVX_Vector vd01_qf16 = Q6_Vqf16_vmpy_VhfVhf(vmax01_hf, Q6_Vh_vsplat_R(0x2008));  // 1.0 / 127.0
     HVX_Vector vd23_qf16 = Q6_Vqf16_vmpy_VhfVhf(vmax23_hf, Q6_Vh_vsplat_R(0x2008));  // 1.0 / 127.0
@@ -2179,11 +2113,7 @@ static inline void quantize_block_f32_q8x4(float * restrict x, uint8_t * restric
 
     // Compute max and scale
     HVX_Vector vmax_hf = hvx_vec_reduce_max_f16(hvx_vec_abs_f16(vx01_hf));
-    vmax_hf            = hvx_vec_reduce_max2_f16(hvx_vec_abs_f16(vx23_hf), vmax_hf);
-
-    // Replicate first fp16 scale across all lanes
-    HVX_Vector ctrl = *(const HVX_Vector *) repl_1x_f16;
-    vmax_hf         = Q6_V_vdelta_VV(vmax_hf, ctrl);
+    vmax_hf            = hvx_vec_reduce_max2_f16(hvx_vec_abs_f16(vx23_hf), vmax_hf); // replicated over all lanes
 
     HVX_Vector vd_qf16 = Q6_Vqf16_vmpy_VhfVhf(vmax_hf, Q6_Vh_vsplat_R(0x2008));  // 1.0 / 127.0
     HVX_Vector vd_hf   = Q6_Vhf_equals_Vqf16(vd_qf16);

ggml/src/ggml-hexagon/htp/rope-ops.c

@@ -10,6 +10,7 @@
 
 #include "hex-dma.h"
 #include "hvx-utils.h"
+#include "hex-fastdiv.h"
 
 #define GGML_COMMON_DECL_C
 #include "ggml-common.h"
@@ -21,6 +22,9 @@
 #define HTP_ROPE_TYPE_NORMAL 0
 #define HTP_ROPE_TYPE_NEOX   2
 
+#define HTP_ROPE_SPAD_NROWS  16
+#define HTP_ROPE_SPAD_BLOCK  (HTP_ROPE_SPAD_NROWS/2)
+
 #define htp_rope_preamble              \
     const uint32_t ne00 = src0->ne[0]; \
     const uint32_t ne01 = src0->ne[1]; \
@@ -42,7 +46,7 @@
     const uint32_t nb2 = dst->nb[2];   \
     const uint32_t nb3 = dst->nb[3];
 
-struct rope_th_ctx {
+struct htp_rope_context {
     int32_t n_dims;
     int32_t mode;
     int32_t n_ctx_orig;
@@ -57,7 +61,19 @@ struct rope_th_ctx {
     float theta_scale;
     float corr_dims[2];
 
+    uint32_t src0_nrows_per_thread;
+    size_t spad_stride;
+
     struct htp_ops_context * octx;
+
+    size_t src0_row_size;
+    size_t dst_row_size;
+    size_t src0_row_size_aligned;
+    size_t dst_row_size_aligned;
+    size_t theta_cache_offset;
+    uint32_t src0_nrows;
+
+    uint64_t t_start;
 };
 
 static float rope_yarn_ramp(const float low, const float high, const int i0) {
@@ -117,64 +133,23 @@ static void rope_corr_dims(int     n_dims,
     dims[1]     = MIN(n_dims - 1, end);
 }
 
-static void init_rope_ctx(struct rope_th_ctx * rope_ctx, struct htp_ops_context * octx) {
-    memset(rope_ctx, 0, sizeof(struct rope_th_ctx));
+static inline void hvx_rope_neox_f32_aa(float * restrict dst, const float * restrict src0, uint32_t ne, const float * restrict theta_cache) {
+    const HVX_Vector * restrict vsrc   = (const HVX_Vector *) src0;
+    const HVX_Vector * restrict vtheta = (const HVX_Vector *) theta_cache;
+    HVX_Vector       * restrict vdst   = (HVX_Vector *) dst;
 
-    const int32_t * op_params = &octx->op_params[0];
+    uint32_t nvec = (ne / (VLEN_FP32 * 2) * 2); // 2 vecs per loop, step of 2
 
-    rope_ctx->n_dims     = ((const int32_t *) op_params)[1];
-    rope_ctx->mode       = ((const int32_t *) op_params)[2];
-    rope_ctx->n_ctx_orig = ((const int32_t *) op_params)[4];
+    uint32_t he = ne / 2;         // half_dims offset in elements
+    uint32_t hv = he / VLEN_FP32; // half_dims offset in vectors
 
-    memcpy(&rope_ctx->freq_base, (int32_t *) op_params + 5, sizeof(float));
-    memcpy(&rope_ctx->freq_scale, (int32_t *) op_params + 6, sizeof(float));
-    memcpy(&rope_ctx->ext_factor, (int32_t *) op_params + 7, sizeof(float));
-    memcpy(&rope_ctx->attn_factor, (int32_t *) op_params + 8, sizeof(float));
-    memcpy(&rope_ctx->beta_fast, (int32_t *) op_params + 9, sizeof(float));
-    memcpy(&rope_ctx->beta_slow, (int32_t *) op_params + 10, sizeof(float));
-    memcpy(&rope_ctx->sections, (int32_t *) op_params + 11, sizeof(int) * 4);
+    #pragma unroll(2)
+    for (uint32_t i = 0; i < nvec; i += 2) {
+        HVX_Vector v0 = vsrc[i/2+0];
+        HVX_Vector v1 = vsrc[i/2+hv];
 
-    rope_ctx->theta_scale = powf(rope_ctx->freq_base, -2.0f / rope_ctx->n_dims);
-
-    rope_corr_dims(rope_ctx->n_dims, rope_ctx->n_ctx_orig, rope_ctx->freq_base, rope_ctx->beta_fast,
-                   rope_ctx->beta_slow, rope_ctx->corr_dims);
-
-    rope_ctx->octx = octx;
-    FARF(HIGH, "rope-f32 n_dims:%d, ext_factor:%.6f, theta_scale:%.6f, attn_factor:%.6f\n", rope_ctx->n_dims,
-         rope_ctx->ext_factor, rope_ctx->theta_scale, rope_ctx->attn_factor);
-}
-
-static void hvx_calc_rope_neox_f32(const float * restrict src0,
-                                   float * restrict dst,
-                                   const int num_elems,
-                                   const float * restrict theta_cache) {
-    // for (int i = 0; i < num_elems; i += 2) {
-    //const float cos_theta = theta_cache[i + 0];
-    //const float sin_theta = theta_cache[i + 1];
-
-    //const float x0 = src[0];
-    //const float x1 = src[num_elems/2];
-
-    //dst[0] = x0*cos_theta - x1*sin_theta;
-    //dst[num_elems/2] = x0*sin_theta + x1*cos_theta;
-
-    //src += 1;
-    //dst += 1;
-    // }
-
-    const uint8_t * restrict src0_curr  = (const uint8_t *) src0;
-    const uint8_t * restrict theta_curr = (const uint8_t *) theta_cache;
-    uint8_t * restrict dst_curr         = (uint8_t *) dst;
-
-    int step_of_1 = num_elems >> 6;  // 6 because we process two vectors at once
-    int half_size = (sizeof(float) * (num_elems / 2));
-
-    for (int i = 0; i < step_of_1; i++) {
-        HVX_Vector v0 = *(HVX_Vector *) src0_curr;
-        HVX_Vector v1 = *(HVX_Vector *) (src0_curr + half_size);
-
-        HVX_Vector v2 = *(HVX_Vector *) theta_curr;
-        HVX_Vector v3 = *(HVX_Vector *) (theta_curr + VLEN);
+        HVX_Vector v2 = vtheta[i+0];
+        HVX_Vector v3 = vtheta[i+1];
 
         HVX_VectorPair vcos_sin = Q6_W_vdeal_VVR(v3, v2, -4);  // vcos_sin[0] = cos_theta, vcos_sin[1] = sin_theta
 
@@ -186,45 +161,34 @@ static void hvx_calc_rope_neox_f32(const float * restrict src0,
         HVX_Vector v4 = Q6_Vqf32_vsub_Vqf32Vqf32(vx0_c, vx1_s);
         HVX_Vector v5 = Q6_Vqf32_vadd_Vqf32Vqf32(vx0_s, vx1_c);
 
-        *(HVX_Vector *) dst_curr               = Q6_Vsf_equals_Vqf32(v4);
-        *(HVX_Vector *) (dst_curr + half_size) = Q6_Vsf_equals_Vqf32(v5);
+        vdst[i/2+0]  = Q6_Vsf_equals_Vqf32(v4);
+        vdst[i/2+hv] = Q6_Vsf_equals_Vqf32(v5);
+    }
 
-        src0_curr += VLEN;
-        theta_curr += 2 * VLEN;
-        dst_curr += VLEN;
+    for (uint32_t i = nvec * VLEN_FP32; i < ne; i += 2) {
+        const float cos_theta = theta_cache[i+0];
+        const float sin_theta = theta_cache[i+1];
+        float x0 = src0[i/2];
+        float x1 = src0[i/2 + he];
+        dst[i/2]      = x0 * cos_theta - x1 * sin_theta;
+        dst[i/2 + he] = x0 * sin_theta + x1 * cos_theta;
     }
 }
 
-static void hvx_calc_rope_f32(const float * restrict src0,
-                              float * restrict dst,
-                              const int num_elems,
-                              const float * restrict theta_cache) {
-    // for (int i = 0; i < num_elems; i += 2) {
-    //const float cos_theta = theta_cache[i + 0];
-    //const float sin_theta = theta_cache[i + 1];
+static inline void hvx_rope_f32_aa(float * restrict dst, const float * restrict src0, uint32_t ne, const float * restrict theta_cache) {
+    const HVX_Vector * restrict vsrc   = (const HVX_Vector *) src0;
+    const HVX_Vector * restrict vtheta = (const HVX_Vector *) theta_cache;
+    HVX_Vector       * restrict vdst   = (HVX_Vector *) dst;
 
-    //const float x0 = src[0];
-    //const float x1 = src[1];
+    uint32_t nvec = (ne / (VLEN_FP32 * 2)) * 2; // 2 vecs per loop, step of two
 
-    //dst[0] = x0*cos_theta - x1*sin_theta;
-    //dst[1] = x0*sin_theta + x1*cos_theta;
+    #pragma unroll(2)
+    for (uint32_t i = 0; i < nvec; i+=2) {
+        HVX_Vector v0 = vsrc[i+0];
+        HVX_Vector v1 = vsrc[i+1];
 
-    //src += 2;
-    //dst += 2;
-    // }
-
-    const uint8_t * restrict src0_curr  = (const uint8_t *) src0;
-    const uint8_t * restrict theta_curr = (const uint8_t *) theta_cache;
-    uint8_t * restrict dst_curr         = (uint8_t *) dst;
-
-    int step_of_1 = num_elems >> 6;  // 6 because we process two vectors at once
-
-    for (int i = 0; i < step_of_1; i++) {
-        HVX_Vector v0 = *(HVX_Vector *) src0_curr;
-        HVX_Vector v1 = *(HVX_Vector *) (src0_curr + VLEN);
-
-        HVX_Vector v2 = *(HVX_Vector *) theta_curr;
-        HVX_Vector v3 = *(HVX_Vector *) (theta_curr + VLEN);
+        HVX_Vector v2 = vtheta[i+0];
+        HVX_Vector v3 = vtheta[i+1];
 
         HVX_VectorPair vx0_x1   = Q6_W_vdeal_VVR(v1, v0, -4);  // vx0_x1[0] = x0, vx0_x1[1] = x1
         HVX_VectorPair vcos_sin = Q6_W_vdeal_VVR(v3, v2, -4);  // vcos_sin[0] = cos_theta, vcos_sin[1] = sin_theta
@@ -239,116 +203,65 @@ static void hvx_calc_rope_f32(const float * restrict src0,
 
         HVX_VectorPair vstore = Q6_W_vshuff_VVR(Q6_Vsf_equals_Vqf32(v5), Q6_Vsf_equals_Vqf32(v4), -4);
 
-        *(HVX_Vector *) dst_curr          = Q6_V_lo_W(vstore);
-        *(HVX_Vector *) (dst_curr + VLEN) = Q6_V_hi_W(vstore);
+        vdst[i+0] = Q6_V_lo_W(vstore);
+        vdst[i+1] = Q6_V_hi_W(vstore);
+    }
 
-        src0_curr += 2 * VLEN;
-        theta_curr += 2 * VLEN;
-        dst_curr += 2 * VLEN;
+    for (uint32_t i = nvec * VLEN_FP32; i < ne; i += 2) {
+        const float cos_theta = theta_cache[i+0];
+        const float sin_theta = theta_cache[i+1];
+        float x0 = src0[i+0];
+        float x1 = src0[i+1];
+        dst[i+0] = x0 * cos_theta - x1 * sin_theta;
+        dst[i+1] = x0 * sin_theta + x1 * cos_theta;
     }
 }
 
-static void rope_hex_f32(struct rope_th_ctx * rope_ctx,
-                         const uint32_t       ir0,
-                         const uint32_t       ir1,
-                         int                  nth,
-                         int                  ith,
-                         const int            opt_path) {
-    struct htp_ops_context * octx = rope_ctx->octx;
+static void inline rope_basic_f32(struct htp_rope_context * rctx, uint8_t * restrict dst, uint8_t * restrict src,
+                   uint32_t nr, uint32_t ne0, const float * restrict theta_cache) {
+    #pragma unroll(4)
+    for (uint32_t i = 0; i < nr; i++) {
+        float * d = (float *) (dst + i * rctx->dst_row_size_aligned);
+        float * s = (float *) (src + i * rctx->src0_row_size_aligned);
+
+        hvx_rope_f32_aa(d, s, rctx->n_dims, theta_cache);
+
+        // fill the remain channels with data from src tensor
+        if (rctx->n_dims < ne0) {
+            hvx_copy_f32_uu((uint8_t *)(d + rctx->n_dims), (uint8_t *)(s + rctx->n_dims), ne0 - rctx->n_dims);
+        }
+    }
+}
+
+static void inline rope_neox_f32(struct htp_rope_context * rctx, uint8_t * restrict dst, uint8_t * restrict src,
+                   uint32_t nr, uint32_t ne0, const float * restrict theta_cache) {
+    #pragma unroll(4)
+    for (uint32_t i = 0; i < nr; i++) {
+        float * d = (float *) (dst + i * rctx->dst_row_size_aligned);
+        float * s = (float *) (src + i * rctx->src0_row_size_aligned);
+
+        hvx_rope_neox_f32_aa(d, s, rctx->n_dims, theta_cache);
+
+        // fill the remain channels with data from src tensor
+        if (rctx->n_dims < ne0) {
+            hvx_copy_f32_uu((uint8_t *)(d + rctx->n_dims), (uint8_t *)(s + rctx->n_dims), ne0 - rctx->n_dims);
+        }
+    }
+}
+
+static void rope_job_f32(unsigned int nth, unsigned int ith, void * data) {
+    struct htp_rope_context * rctx = (struct htp_rope_context *) data;
+    struct htp_ops_context * octx = rctx->octx;
 
     const struct htp_tensor * src0 = &octx->src0;
     const struct htp_tensor * src1 = &octx->src1;
     const struct htp_tensor * src2 = &octx->src2;
     struct htp_tensor *       dst  = &octx->dst;
 
-    const int32_t mode    = rope_ctx->mode;
-    const bool    is_neox = mode & HTP_ROPE_TYPE_NEOX;
-
     htp_rope_preamble;
 
-    const int32_t * pos = (const int32_t *) src1->data;
-
-    float * wp0 = (float *) (octx->src0_spad.data + (ith * nb01));
-
-    const float * freq_factors = NULL;
-    if (src2 != NULL) {
-        freq_factors = (const float *) src2->data;
-    }
-
-    const uint32_t i1_end       = MIN(ir1, ne1);
-    const int32_t  half_dims    = rope_ctx->n_dims / 2;
-    const size_t   remain_bytes = (ne0 - rope_ctx->n_dims) * sizeof(float);
-    for (uint32_t i3 = 0; i3 < ne3; i3++) {      // batch
-        for (uint32_t i2 = 0; i2 < ne2; i2++) {  // seq-len
-            const int32_t p = pos[i2];
-
-            rope_cache_init(p, rope_ctx->freq_scale, freq_factors, rope_ctx->corr_dims, ne0, rope_ctx->ext_factor,
-                            rope_ctx->attn_factor, wp0, rope_ctx->theta_scale);
-
-            for (uint32_t i1 = ir0; i1 < i1_end; i1++) {  // attn-heads
-                const float * src      = (float *) ((char *) src0->data + i3 * nb03 + i2 * nb02 + i1 * nb01);
-                float *       dst_data = (float *) ((char *) dst->data + i3 * nb3 + i2 * nb2 + i1 * nb1);
-
-                const float * src_loc      = src;
-                float *       dst_data_loc = dst_data;
-
-                if (1 == opt_path) {
-                    if (is_neox) {
-                        hvx_calc_rope_neox_f32(src_loc, dst_data_loc, rope_ctx->n_dims, wp0);
-                    } else {
-                        hvx_calc_rope_f32(src_loc, dst_data_loc, rope_ctx->n_dims, wp0);
-                    }
-
-                    src_loc += rope_ctx->n_dims;
-                    dst_data_loc += rope_ctx->n_dims;
-                } else {
-                    for (uint32_t i0 = 0; i0 < rope_ctx->n_dims; i0 += 2) {
-                        const float cos_theta = wp0[i0 + 0];
-                        const float sin_theta = wp0[i0 + 1];
-
-                        if (is_neox) {
-                            const float x0 = src_loc[0];
-                            const float x1 = src_loc[half_dims];
-
-                            dst_data_loc[0]         = x0 * cos_theta - x1 * sin_theta;
-                            dst_data_loc[half_dims] = x0 * sin_theta + x1 * cos_theta;
-
-                            src_loc += 1;
-                            dst_data_loc += 1;
-                        } else {
-                            const float x0 = src_loc[0];
-                            const float x1 = src_loc[1];
-
-                            dst_data_loc[0] = x0 * cos_theta - x1 * sin_theta;
-                            dst_data_loc[1] = x0 * sin_theta + x1 * cos_theta;
-
-                            src_loc += 2;
-                            dst_data_loc += 2;
-                        }
-                    }
-
-                    src_loc += (is_neox ? half_dims : 0);
-                    dst_data_loc += (is_neox ? half_dims : 0);
-                }
-
-                // TODO: use simd to speed up the remaining elements copy
-                memcpy(dst_data_loc, src_loc, remain_bytes);
-            }
-        }
-    }
-}
-
-static void rope_job_f32_per_thread(struct rope_th_ctx * rope_ctx, int nth, int ith) {
-    struct htp_ops_context * octx = rope_ctx->octx;
-
-    const struct htp_tensor * src0 = &octx->src0;
-    const struct htp_tensor * src1 = &octx->src1;
-    struct htp_tensor *       dst  = &octx->dst;
-
-    htp_rope_preamble;
-
-    const uint32_t src0_nrows            = ne01 * ne02 * ne03;  // src0 rows
-    const uint32_t src0_nrows_per_thread = octx->src0_nrows_per_thread;
+    const uint32_t src0_nrows = rctx->src0_nrows;
+    const uint32_t src0_nrows_per_thread = rctx->src0_nrows_per_thread;
 
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
@@ -358,32 +271,114 @@ static void rope_job_f32_per_thread(struct rope_th_ctx * rope_ctx, int nth, int
         return;
     }
 
-    uint64_t t1, t2;
-    t1 = HAP_perf_get_qtimer_count();
+    uint64_t tt = HAP_perf_get_qtimer_count();
 
-    int is_aligned = 1;
-    int opt_path   = 0;
-    if ((0 == hex_is_aligned((void *) src0->data, VLEN)) || (0 == hex_is_aligned((void *) src1->data, VLEN)) ||
-        (0 == hex_is_aligned((void *) dst->data, VLEN))) {
-        FARF(HIGH, "rope-f32: unaligned addresses in rope op, possibly slower execution\n");
-        is_aligned = 0;
-    }
-    if ((1 == is_aligned) && !(nb01 & (VLEN - 1))) {
-        opt_path = 1;
+    const int32_t mode    = rctx->mode;
+    const bool    is_neox = mode & HTP_ROPE_TYPE_NEOX;
+
+    // VTCM setup
+    uint8_t * src0_spad_base = octx->src0_spad.data + (ith * octx->src0_spad.size_per_thread);
+    float *   theta_cache    = (float *) (src0_spad_base);
+              src0_spad_base = src0_spad_base + rctx->theta_cache_offset;
+    uint8_t * dst_spad_base  = octx->dst_spad.data + (ith * octx->dst_spad.size_per_thread);
+
+    dma_queue * dma_queue = octx->ctx->dma[ith];
+    const int32_t * pos = (const int32_t *) src1->data;
+    const float * freq_factors = src2->data ? (const float *) src2->data : NULL;
+
+    uint32_t ir = 0;
+    uint32_t prev_i2 = (uint32_t) -1;
+
+    for (uint32_t i3 = 0; i3 < ne3; i3++) { // batch
+        for (uint32_t i2 = 0; i2 < ne2; i2++) { // seq-len
+            for (uint32_t i1 = 0; i1 < ne1; ) { // attn-heads
+                if (ir < src0_start_row) { ir++; i1++; continue; }
+                if (ir >= src0_end_row) goto done;
+
+                // Rows in this block
+                const uint32_t nrows = MIN(src0_end_row - ir, ne1 - i1);
+
+                // Depth before prefetch
+                uint32_t dma_depth = dma_queue_depth(dma_queue);
+
+                // FARF(HIGH, "rope-block %u: ir %u n-rows %u dma-depth %u : usec %u", ith, ir, nrows, dma_depth,
+                //             (unsigned) HAP_perf_qtimer_count_to_us(HAP_perf_get_qtimer_count() - rctx->t_start));
+
+                // Prefetch loop
+                for (uint32_t pnr = 0, pr = 0; pr < nrows && pr < HTP_ROPE_SPAD_NROWS; pr += pnr) {
+                    pnr = MIN(nrows - pr, HTP_ROPE_SPAD_BLOCK);
+
+                    uint32_t pi1 = i1 + pr;
+                    uint32_t pir = ir + pr;
+
+                    // Dummy DMA transaction for sequencing (interleaving dst,src,dst,...)
+                    dma_queue_push_vtcm_to_ddr(dma_queue, dma_make_ptr((void *) dst->data, dst_spad_base + pr * rctx->dst_row_size_aligned), 0, 0, 0);
+
+                    const uint8_t * src_addr = (const uint8_t *) src0->data + i3 * nb03 + i2 * nb02 + pi1 * nb01;
+                          uint8_t * src_spad = src0_spad_base + pr * rctx->src0_row_size_aligned;
+                    dma_queue_push_ddr_to_vtcm(dma_queue, dma_make_ptr(src_spad, src_addr),
+                        rctx->src0_row_size_aligned, rctx->src0_row_size, pnr);
+
+                    // FARF(HIGH, "rope-prefetch %u: pr %u i1 %u i2 %u i3 %u src-spad %p src-addr %p pnr %u", ith, pir, pi1, i2, i3, src_spad, src_addr, pnr);
+                }
+
+                // Update theta cache
+                if (i2 != prev_i2) {
+                    prev_i2 = i2;
+
+                    const int32_t p = pos[i2];
+                    rope_cache_init(p, rctx->freq_scale, freq_factors, rctx->corr_dims, ne0, rctx->ext_factor, rctx->attn_factor, theta_cache, rctx->theta_scale);
+
+                    // FARF(HIGH, "rope-theta %u: ir %u i1 %u i2 %u i3 %u cache %p : usec %u", ith, ir, i1, i2, i3, theta_cache,
+                    //         (unsigned) HAP_perf_qtimer_count_to_us(HAP_perf_get_qtimer_count() - rctx->t_start));
+                }
+
+                // Skip DMA transactions from prev block (if any)
+                // No need to wait for these since the DMA is setup for in-order processing
+                for (uint32_t d=0; d < dma_depth; d++) { dma_queue_pop_nowait(dma_queue); }
+
+                // Compute loop
+                for (uint32_t cnr = 0, cr = 0; cr < nrows; cr += cnr, ir += cnr, i1 += cnr) {
+                    // Number of rows to compute
+                    cnr = MIN(nrows - cr, HTP_ROPE_SPAD_BLOCK);
+
+                    uint8_t * dst_spad = (uint8_t *) dma_queue_pop(dma_queue).src;
+                    uint8_t * src_spad = (uint8_t *) dma_queue_pop(dma_queue).dst;
+
+                    // FARF(HIGH, "rope-compute %u: ir %u i1 %u i2 %u i3 %u src-spad %p cnr %u : usec %u", ith, ir, i1, i2, i3, src_spad, cnr,
+                    //         (unsigned) HAP_perf_qtimer_count_to_us(HAP_perf_get_qtimer_count() - rctx->t_start));
+
+                    if (is_neox) {
+                        rope_neox_f32(rctx, dst_spad, src_spad, cnr, ne0, theta_cache);
+                    } else {
+                        rope_basic_f32(rctx, dst_spad, src_spad, cnr, ne0, theta_cache);
+                    }
+
+                    uint8_t * dst_addr = (uint8_t *) dst->data + i3 * nb3 + i2 * nb2 + i1 * nb1;
+                    dma_queue_push_vtcm_to_ddr(dma_queue, dma_make_ptr(dst_addr, dst_spad), rctx->dst_row_size, rctx->dst_row_size_aligned, cnr);
+
+                    // Prefetch more rows (if any)
+                    if ((cr + HTP_ROPE_SPAD_NROWS) < nrows) {
+                        uint32_t pnr = MIN(nrows - (cr + HTP_ROPE_SPAD_NROWS), HTP_ROPE_SPAD_BLOCK);
+                        uint32_t pi1 = i1 + HTP_ROPE_SPAD_NROWS;
+                        uint32_t pir = ir + HTP_ROPE_SPAD_NROWS;
+
+                        const uint8_t * src_addr = (const uint8_t *) src0->data + i3 * nb03 + i2 * nb02 + pi1 * nb01;
+                        dma_queue_push_ddr_to_vtcm(dma_queue, dma_make_ptr(src_spad, src_addr),
+                            rctx->src0_row_size_aligned, rctx->src0_row_size, pnr);
+
+                        // FARF(HIGH, "rope-prefetch %u: pr %u i1 %u i2 %u i3 %u src-spad %p src-addr %p pnr %u", ith, pir, pi1, i2, i3, src_spad, src_addr, pnr);
+                    }
+                }
+            }
+        }
     }
 
-    rope_hex_f32(rope_ctx, src0_start_row, src0_end_row, nth, ith, opt_path);
+done:
+    dma_queue_flush(dma_queue);
+    tt = HAP_perf_get_qtimer_count() - tt;
 
-    t2 = HAP_perf_get_qtimer_count();
-
-    FARF(HIGH, "rope-f32: %d/%d/%d: (%u:%u) usec %u\n", ith, nth, opt_path, src0_start_row, src0_end_row,
-         (unsigned) HAP_perf_qtimer_count_to_us(t2 - t1));
-}
-
-static void rope_job_dispatcher_f32(unsigned int n, unsigned int i, void * data) {
-    struct rope_th_ctx * rope_ctx = (struct rope_th_ctx *) data;
-
-    rope_job_f32_per_thread(rope_ctx, n, i);
+    FARF(HIGH, "rope-f32: %d/%d: (%u:%u) usec %u\n", ith, nth, src0_start_row, src0_end_row, (unsigned) HAP_perf_qtimer_count_to_us(tt));
 }
 
 static int execute_op_rope_f32(struct htp_ops_context * octx) {
@@ -394,17 +389,10 @@ static int execute_op_rope_f32(struct htp_ops_context * octx) {
     const struct htp_tensor * src2 = &octx->src2;
     struct htp_tensor *       dst  = &octx->dst;
 
-    worker_callback_t op_func;
-    const char *      op_type = NULL;
-
-    struct rope_th_ctx rope_ctx;
+    const char * op_type = "rope-f32";
 
     switch (octx->op) {
         case HTP_OP_ROPE:
-            op_func = rope_job_dispatcher_f32;
-            op_type = "rope-f32";
-
-            init_rope_ctx(&rope_ctx, octx);
             break;
 
         default:
@@ -415,49 +403,79 @@ static int execute_op_rope_f32(struct htp_ops_context * octx) {
     const uint32_t n_threads = octx->n_threads;
 
     const size_t src0_row_size = src0->nb[1];
-    const size_t src1_row_size = src0_row_size;
     const size_t dst_row_size  = dst->nb[1];
 
-    // VTCM scratchpads for all tensors
-    // N rows per thread, padded to HVX vector size
-    octx->dst_spad.size  = hex_round_up(dst_row_size, 128) * n_threads;
-    octx->src0_spad.size = hex_round_up(src0_row_size, 128) * n_threads;
-    octx->src1_spad.size = hex_round_up(src1_row_size, 128) * n_threads;
+    // Aligned row sizes for VTCM
+    const size_t src0_row_size_aligned    = hex_round_up(src0_row_size, VLEN);
+    const size_t dst_row_size_aligned     = hex_round_up(dst_row_size, VLEN);
+    const size_t theta_cache_size_aligned = hex_round_up(src0->ne[0] * sizeof(float), 128);
 
-    size_t spad_size = octx->src0_spad.size + octx->src1_spad.size + octx->dst_spad.size;
+    // Calculate spad sizes per thread
+    size_t src0_spad_per_thread = theta_cache_size_aligned + HTP_ROPE_SPAD_NROWS * src0_row_size_aligned;
+    size_t dst_spad_per_thread  = HTP_ROPE_SPAD_NROWS * dst_row_size_aligned;
+    size_t spad_per_thread = src0_spad_per_thread + dst_spad_per_thread;
 
-    if (src2->ne[0]) {
-        FARF(HIGH,
-             "%s: %ux%ux%ux%u (x %ux%ux%ux%u x %ux%ux%ux%u) -> %ux%ux%ux%u : src0-spad-size %u src1-spad-size %u "
-             "dst-spad-size %u\n",
-             op_type, src0->ne[0], src0->ne[1], src0->ne[2], src0->ne[3], src1->ne[0], src1->ne[1], src1->ne[2],
-             src1->ne[3], src2->ne[0], src2->ne[1], src2->ne[2], src2->ne[3], dst->ne[0], dst->ne[1], dst->ne[2],
-             dst->ne[3], octx->src0_spad.size, octx->src1_spad.size, octx->dst_spad.size);
-    } else {
-        FARF(HIGH,
-             "%s: %ux%ux%ux%u (%ux%ux%ux%u) -> %ux%ux%ux%u : src0-spad-size %u src1-spad-size %u dst-spad-size %u\n",
-             op_type, src0->ne[0], src0->ne[1], src0->ne[2], src0->ne[3], src1->ne[0], src1->ne[1], src1->ne[2],
-             src1->ne[3], dst->ne[0], dst->ne[1], dst->ne[2], dst->ne[3], octx->src0_spad.size, octx->src1_spad.size,
-             octx->dst_spad.size);
-    }
-
-    // Make sure the reserved vtcm size is sufficient
-    if (octx->ctx->vtcm_size < spad_size) {
-        FARF(ERROR, "%s : current VTCM reservation %zu is too small, needed %zu\n", op_type, octx->ctx->vtcm_size,
-             spad_size);
+    // Check if we fit in VTCM
+    size_t total_vtcm_needed = spad_per_thread * n_threads;
+    if (octx->ctx->vtcm_size < total_vtcm_needed) {
+        FARF(ERROR, "%s : current VTCM reservation %zu is too small, needed %zu\n", op_type, octx->ctx->vtcm_size, total_vtcm_needed);
         return HTP_STATUS_VTCM_TOO_SMALL;
     }
 
-    octx->src0_spad.data = octx->ctx->vtcm_base;
-    octx->src1_spad.data = octx->src0_spad.data + octx->src0_spad.size;
-    octx->dst_spad.data  = octx->src1_spad.data + octx->src1_spad.size;
+    // Assign sizes
+    octx->src0_spad.size_per_thread = src0_spad_per_thread;
+    octx->dst_spad.size_per_thread  = dst_spad_per_thread;
+    octx->src0_spad.size = n_threads * src0_spad_per_thread;
+    octx->dst_spad.size  = n_threads * dst_spad_per_thread;
+    octx->src1_spad.size = 0;
 
+    // Assign pointers
+    octx->src0_spad.data = octx->ctx->vtcm_base;
+    octx->src1_spad.data = NULL;
+    octx->dst_spad.data  = octx->src0_spad.data + octx->src0_spad.size;
+
+    // Fill context
+    struct htp_rope_context rctx;
+    memset(&rctx, 0, sizeof(struct htp_rope_context));
+
+    rctx.t_start = HAP_perf_get_qtimer_count();
+
+    rctx.octx = octx;
+
+    const int32_t * op_params = &octx->op_params[0];
+    rctx.n_dims     = ((const int32_t *) op_params)[1];
+    rctx.mode       = ((const int32_t *) op_params)[2];
+    rctx.n_ctx_orig = ((const int32_t *) op_params)[4];
+
+    memcpy(&rctx.freq_base,   (int32_t *) op_params + 5,  sizeof(float));
+    memcpy(&rctx.freq_scale,  (int32_t *) op_params + 6,  sizeof(float));
+    memcpy(&rctx.ext_factor,  (int32_t *) op_params + 7,  sizeof(float));
+    memcpy(&rctx.attn_factor, (int32_t *) op_params + 8,  sizeof(float));
+    memcpy(&rctx.beta_fast,   (int32_t *) op_params + 9,  sizeof(float));
+    memcpy(&rctx.beta_slow,   (int32_t *) op_params + 10, sizeof(float));
+    memcpy(&rctx.sections,    (int32_t *) op_params + 11, sizeof(int) * 4);
+
+    rctx.theta_scale = powf(rctx.freq_base, -2.0f / rctx.n_dims);
+
+    rope_corr_dims(rctx.n_dims, rctx.n_ctx_orig, rctx.freq_base, rctx.beta_fast, rctx.beta_slow, rctx.corr_dims);
+
+    rctx.src0_row_size = src0_row_size;
+    rctx.dst_row_size  = dst_row_size;
+    rctx.src0_row_size_aligned = src0_row_size_aligned;
+    rctx.dst_row_size_aligned  = dst_row_size_aligned;
+    rctx.theta_cache_offset    = theta_cache_size_aligned;
+
+    uint32_t ne0 = dst->ne[0];
     uint32_t src0_nrows = src0->ne[1] * src0->ne[2] * src0->ne[3];
+    rctx.src0_nrows = src0_nrows;
+
+    FARF(HIGH, "rope-f32 n-rows %u n-dims %d ne0 %u ext-factor %.6f theta-scale %.6f attn-factor %.6f\n", rctx.src0_nrows, rctx.n_dims, ne0,
+         rctx.ext_factor, rctx.theta_scale, rctx.attn_factor);
 
     if (!(octx->flags & HTP_OPFLAGS_SKIP_COMPUTE)) {
-        uint32_t n_jobs             = MIN(n_threads, src0_nrows);
-        octx->src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
-        worker_pool_run_func(octx->ctx->worker_pool, op_func, &rope_ctx, n_jobs);
+        uint32_t n_jobs = MIN(n_threads, src0_nrows);
+        rctx.src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
+        worker_pool_run_func(octx->ctx->worker_pool, rope_job_f32, &rctx, n_jobs);
     }
 
     return err;

ggml/src/ggml-hexagon/htp/set-rows-ops.c

@@ -43,11 +43,21 @@
                                             \
     const uint32_t nr  = ne01;
 
-static int set_rows_thread_f32_f32(struct htp_ops_context * octx, const int nth, const int ith) {
+struct htp_set_rows_context {
+    struct htp_ops_context * octx;
+    struct fastdiv_values div_ne12;
+    struct fastdiv_values div_ne11;
+    uint32_t src0_nrows_per_thread;
+};
+
+static void set_rows_thread_f32_f32(unsigned int nth, unsigned int ith, void *data) {
+    struct htp_set_rows_context * srctx = (struct htp_set_rows_context *)data;
+    struct htp_ops_context * octx = srctx->octx;
+
     set_rows_preamble;
 
     // parallelize by rows of src0
-    const uint32_t dr  = octx->src0_nrows_per_thread;
+    const uint32_t dr  = srctx->src0_nrows_per_thread;
     const uint32_t ir0 = dr * ith;
     const uint32_t ir1 = (ir0 + dr < nr) ? (ir0 + dr) : nr;
 
@@ -56,8 +66,8 @@ static int set_rows_thread_f32_f32(struct htp_ops_context * octx, const int nth,
     for (uint32_t i03 = 0; i03 < ne03; ++i03) {
         for (uint32_t i02 = 0; i02 < ne02; ++i02) {
             for (uint32_t i = ir0; i < ir1; ++i) {
-                const uint32_t i12 = fastmodulo(i03, ne12, &octx->set_rows_div_ne12);
-                const uint32_t i11 = fastmodulo(i02, ne11, &octx->set_rows_div_ne11);
+                const uint32_t i12 = fastmodulo(i03, ne12, &srctx->div_ne12);
+                const uint32_t i11 = fastmodulo(i02, ne11, &srctx->div_ne11);
                 const uint32_t i10 = i;
 
                 const uintptr_t src1_addr = octx->src1.data + i10*nb10 + i11*nb11 + i12*nb12;
@@ -76,15 +86,16 @@ static int set_rows_thread_f32_f32(struct htp_ops_context * octx, const int nth,
             }
         }
     }
-
-    return HTP_STATUS_OK;
 }
 
-static int set_rows_thread_f16_f32(struct htp_ops_context * octx, const int nth, const int ith) {
+static void set_rows_thread_f16_f32(unsigned int nth, unsigned int ith, void *data) {
+    struct htp_set_rows_context * srctx = (struct htp_set_rows_context *)data;
+    struct htp_ops_context * octx = srctx->octx;
+
     set_rows_preamble;
 
     // parallelize by rows of src0
-    const uint32_t dr  = octx->src0_nrows_per_thread;
+    const uint32_t dr  = srctx->src0_nrows_per_thread;
     const uint32_t ir0 = dr * ith;
     const uint32_t ir1 = (ir0 + dr < nr) ? (ir0 + dr) : nr;
 
@@ -93,8 +104,8 @@ static int set_rows_thread_f16_f32(struct htp_ops_context * octx, const int nth,
     for (uint32_t i03 = 0; i03 < ne03; ++i03) {
         for (uint32_t i02 = 0; i02 < ne02; ++i02) {
             for (uint32_t i = ir0; i < ir1; ++i) {
-                const uint32_t i12 = fastmodulo(i03, ne12, &octx->set_rows_div_ne12);
-                const uint32_t i11 = fastmodulo(i02, ne11, &octx->set_rows_div_ne11);
+                const uint32_t i12 = fastmodulo(i03, ne12, &srctx->div_ne12);
+                const uint32_t i11 = fastmodulo(i02, ne11, &srctx->div_ne11);
                 const uint32_t i10 = i;
 
                 const uintptr_t src1_addr = octx->src1.data + i10*nb10 + i11*nb11 + i12*nb12;
@@ -112,16 +123,6 @@ static int set_rows_thread_f16_f32(struct htp_ops_context * octx, const int nth,
             }
         }
     }
-
-    return HTP_STATUS_OK;
-}
-
-static void set_rows_work_f16_f32(unsigned int n, unsigned int i, void *data) {
-    set_rows_thread_f16_f32((struct htp_ops_context *) data, n, i);
-}
-
-static void set_rows_work_f32_f32(unsigned int n, unsigned int i, void *data) {
-    set_rows_thread_f32_f32((struct htp_ops_context *) data, n, i);
 }
 
 int op_set_rows(struct htp_ops_context * octx) {
@@ -143,18 +144,20 @@ int op_set_rows(struct htp_ops_context * octx) {
         return HTP_STATUS_OK;
     }
 
-    octx->set_rows_div_ne12 = init_fastdiv_values(ne12);
-    octx->set_rows_div_ne11 = init_fastdiv_values(ne11);
+    struct htp_set_rows_context srctx;
+    srctx.octx = octx;
+    srctx.div_ne12 = init_fastdiv_values(ne12);
+    srctx.div_ne11 = init_fastdiv_values(ne11);
 
     const uint32_t n_jobs = MIN(nr, octx->n_threads);
-    octx->src0_nrows_per_thread = (nr + n_jobs - 1) / n_jobs;
+    srctx.src0_nrows_per_thread = (nr + n_jobs - 1) / n_jobs;
 
     switch(octx->dst.type) {
     case HTP_TYPE_F32:
-        worker_pool_run_func(octx->ctx->worker_pool, set_rows_work_f32_f32, octx, n_jobs);
+        worker_pool_run_func(octx->ctx->worker_pool, set_rows_thread_f32_f32, &srctx, n_jobs);
         break;
     case HTP_TYPE_F16:
-        worker_pool_run_func(octx->ctx->worker_pool, set_rows_work_f16_f32, octx, n_jobs);
+        worker_pool_run_func(octx->ctx->worker_pool, set_rows_thread_f16_f32, &srctx, n_jobs);
         break;
     default:
         return HTP_STATUS_NO_SUPPORT;

ggml/src/ggml-hexagon/htp/softmax-ops.c

@@ -10,6 +10,7 @@
 
 #include "hex-dma.h"
 #include "hvx-utils.h"
+#include "hex-fastdiv.h"
 
 #define GGML_COMMON_DECL_C
 #include "ggml-common.h"
@@ -48,7 +49,7 @@
     const uint32_t nb2 = dst->nb[2];                       \
     const uint32_t nb3 = dst->nb[3];
 
-struct softmax_th_ctx {
+struct htp_softmax_context {
     bool     use_f16;
     bool     use_src1;
     uint32_t n_head;
@@ -59,28 +60,48 @@ struct softmax_th_ctx {
     float m0;
     float m1;
 
+    uint32_t src0_nrows_per_thread;
+    struct fastdiv_values fastdiv_ne01;
+    struct fastdiv_values fastdiv_ne02;
+    struct fastdiv_values fastdiv_ne12; // For mask broadcasting
+    struct fastdiv_values fastdiv_ne13; // For mask broadcasting
+    size_t spad_stride;
+
     struct htp_ops_context * octx;
 };
 
-static void init_softmax_ctx(struct softmax_th_ctx * softmax_ctx, struct htp_ops_context * octx) {
+static void init_softmax_ctx(struct htp_softmax_context * smctx, struct htp_ops_context * octx) {
     const struct htp_tensor * src0 = &octx->src0;
     const struct htp_tensor * src1 = &octx->src1;
 
-    memset(softmax_ctx, 0, sizeof(struct softmax_th_ctx));
+    memset(smctx, 0, sizeof(struct htp_softmax_context));
 
-    memcpy(&softmax_ctx->scale, (float *) octx->op_params, sizeof(float));
-    memcpy(&softmax_ctx->max_bias, (float *) octx->op_params + 1, sizeof(float));
+    memcpy(&smctx->scale, (float *) octx->op_params, sizeof(float));
+    memcpy(&smctx->max_bias, (float *) octx->op_params + 1, sizeof(float));
 
-    softmax_ctx->n_head      = src0->ne[2];
-    softmax_ctx->n_head_log2 = 1u << (uint32_t) floor(log2(softmax_ctx->n_head));
+    smctx->n_head      = src0->ne[2];
+    smctx->n_head_log2 = 1u << (uint32_t) floor(log2(smctx->n_head));
 
-    softmax_ctx->m0 = powf(2.0f, -(softmax_ctx->max_bias) / softmax_ctx->n_head_log2);
-    softmax_ctx->m1 = powf(2.0f, -(softmax_ctx->max_bias / 2.0f) / softmax_ctx->n_head_log2);
+    smctx->m0 = powf(2.0f, -(smctx->max_bias) / smctx->n_head_log2);
+    smctx->m1 = powf(2.0f, -(smctx->max_bias / 2.0f) / smctx->n_head_log2);
 
-    softmax_ctx->use_src1 = (src1->ne[0] != 0);
-    softmax_ctx->use_f16  = (src1->ne[0] != 0) && (src1->type == HTP_TYPE_F16);
+    smctx->use_src1 = (src1->ne[0] != 0);
+    smctx->use_f16  = (src1->ne[0] != 0) && (src1->type == HTP_TYPE_F16);
 
-    softmax_ctx->octx = octx;
+    smctx->octx = octx;
+
+    // Initialize fastdiv values
+    const uint32_t ne01 = src0->ne[1];
+    const uint32_t ne02 = src0->ne[2];
+
+    if (ne01 > 0) smctx->fastdiv_ne01 = init_fastdiv_values(ne01);
+    if (ne02 > 0) smctx->fastdiv_ne02 = init_fastdiv_values(ne02);
+
+    const uint32_t ne12 = (src1->ne[0]) ? src1->ne[2] : 1;
+    const uint32_t ne13 = (src1->ne[0]) ? src1->ne[3] : 1;
+
+    if (ne12 > 0) smctx->fastdiv_ne12 = init_fastdiv_values(ne12);
+    if (ne13 > 0) smctx->fastdiv_ne13 = init_fastdiv_values(ne13);
 }
 
 static void hvx_fast_softmax_prep_f32(const uint8_t * restrict src,
@@ -139,8 +160,7 @@ static void hvx_fast_softmax_f32(const uint8_t * restrict src,
         max_vec       = Q6_Vsf_vmax_VsfVsf(max_vec, v1);
     }
 
-    HVX_Vector v = hvx_vec_reduce_max_f32(max_vec);
-    max_vec      = hvx_vec_repl4(v);
+    max_vec = hvx_vec_reduce_max_f32(max_vec); // replicated over all lanes
 
     #pragma unroll(4)
     for (int i = 0; i < step_of_1; i++) {
@@ -154,8 +174,7 @@ static void hvx_fast_softmax_f32(const uint8_t * restrict src,
         v_pad[i] = v3;
     }
 
-    v       = hvx_vec_reduce_sum_f32(Q6_Vsf_equals_Vqf32(sum_vec));
-    sum_vec = hvx_vec_repl4(v);
+    sum_vec = hvx_vec_reduce_sum_f32(Q6_Vsf_equals_Vqf32(sum_vec)); // replicated over all lanes
 
     HVX_VectorPred pos_sum   = Q6_Q_vcmp_gt_VwVw(sum_vec, zero_v);
     HVX_Vector     v4        = hvx_vec_inverse_f32(sum_vec);
@@ -183,83 +202,9 @@ static float hvx_softmax_f32(const uint8_t * restrict src,
     return sum;
 }
 
-static void softmax_htp_f32(int nth, int ith, struct softmax_th_ctx * softmax_ctx, int opt_path) {
-    struct htp_ops_context * octx = softmax_ctx->octx;
-
-    const struct htp_tensor * src0 = &octx->src0;
-    const struct htp_tensor * src1 = &octx->src1;
-    const struct htp_tensor * dst  = &octx->dst;
-
-    htp_softmax_preamble3;
-
-    uint8_t * src0_spad_data = octx->src0_spad.data + (ith * nb01);
-    uint8_t * src1_spad_data = octx->src1_spad.data + (ith * nb01);
-    uint8_t * dst_spad_data  = octx->dst_spad.data + (ith * nb1);
-
-    float * wp0 = (float *) src0_spad_data;
-    float * wp1 = (float *) src1_spad_data;
-    float * wp2 = (float *) dst_spad_data;
-
-    for (uint32_t i03 = 0; i03 < ne03; i03++) {
-        for (uint32_t i02 = 0; i02 < ne02; i02++) {
-            for (uint32_t i01 = ith; i01 < ne01; i01 += nth) {
-                const uint32_t i11 = i01;
-                const uint32_t i12 = i02 % ne12;
-                const uint32_t i13 = i03 % ne13;
-
-                // ALiBi
-                const uint32_t h = i02;  // head
-
-                const float slope = (softmax_ctx->max_bias > 0.0f) ?
-                                        h < softmax_ctx->n_head_log2 ?
-                                        powf(softmax_ctx->m0, h + 1) :
-                                        powf(softmax_ctx->m1, 2 * (h - softmax_ctx->n_head_log2) + 1) :
-                                        1.0f;
-
-                float * sp = (float *) ((char *) octx->src0.data + i01 * nb01 + i02 * nb02 + i03 * nb03);
-                float * dp = (float *) ((char *) octx->dst.data + i01 * nb1 + i02 * nb2 + i03 * nb3);
-
-                // broadcast the mask across rows
-                __fp16 * mp_f16 = (softmax_ctx->use_src1) ?
-                                      (__fp16 *) ((char *) octx->src1.data + i11 * nb11 + i12 * nb12 + i13 * nb13) :
-                                      NULL;
-                float *  mp_f32 = (softmax_ctx->use_src1) ?
-                                      (float *) ((char *) octx->src1.data + i11 * nb11 + i12 * nb12 + i13 * nb13) :
-                                      NULL;
-
-                if ((1 == opt_path) && (mp_f32) && !(softmax_ctx->use_f16)) {
-                    hvx_fast_softmax_prep_f32((const uint8_t *) sp, (uint8_t *) wp0, ne00, softmax_ctx->scale,
-                                              (const uint8_t *) mp_f32, slope);
-                } else {
-                    hvx_scale_f32((uint8_t *) wp0, (const uint8_t *) sp, ne00, softmax_ctx->scale);
-                    if (mp_f32) {
-                        if (softmax_ctx->use_f16) {
-                            for (int i = 0; i < ne00; ++i) {
-                                wp0[i] += slope * (float) mp_f16[i];
-                            }
-                        } else {
-                            for (int i = 0; i < ne00; ++i) {
-                                wp0[i] += slope * mp_f32[i];
-                            }
-                        }
-                    }
-                }
-
-                if (1 == opt_path) {
-                    hvx_fast_softmax_f32((const uint8_t *) wp0, (uint8_t *) dp, (uint8_t *) wp1, ne00);
-                } else {
-                    float max = hvx_reduce_max_f32((const uint8_t *) wp0, ne00);
-                    float sum = hvx_softmax_f32((const uint8_t *) wp0, (uint8_t *) wp2, (uint8_t *) wp1, ne00, max);
-                    sum       = sum > 0.0 ? (1.0 / sum) : 1;
-                    hvx_scale_f32((uint8_t *) dp, (const uint8_t *) wp2, ne00, sum);
-                }
-            }
-        }
-    }
-}
-
-static void softmax_job_f32_per_thread(struct softmax_th_ctx * softmax_ctx, int nth, int ith) {
-    struct htp_ops_context * octx = softmax_ctx->octx;
+static void softmax_job_f32(unsigned int nth, unsigned int ith, void * data) {
+    struct htp_softmax_context * smctx = (struct htp_softmax_context *) data;
+    struct htp_ops_context * octx = smctx->octx;
 
     const struct htp_tensor * src0 = &octx->src0;
     const struct htp_tensor * src1 = &octx->src1;
@@ -268,7 +213,7 @@ static void softmax_job_f32_per_thread(struct softmax_th_ctx * softmax_ctx, int
     htp_softmax_preamble3;
 
     const uint32_t src0_nrows            = ne01 * ne02 * ne03;  // src0 rows
-    const uint32_t src0_nrows_per_thread = octx->src0_nrows_per_thread;
+    const uint32_t src0_nrows_per_thread = smctx->src0_nrows_per_thread;
 
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
@@ -291,20 +236,103 @@ static void softmax_job_f32_per_thread(struct softmax_th_ctx * softmax_ctx, int
         opt_path = 1;
     }
 
-    softmax_htp_f32(nth, ith, softmax_ctx, opt_path);
+    uint8_t * src0_spad_data = octx->src0_spad.data + (ith * smctx->spad_stride);
+    uint8_t * src1_spad_data = octx->src1_spad.data + (ith * smctx->spad_stride);
+    uint8_t * dst_spad_data  = octx->dst_spad.data + (ith * smctx->spad_stride);
+
+    float * wp0 = (float *) src0_spad_data;
+    float * wp1 = (float *) src1_spad_data;
+    float * wp2 = (float *) dst_spad_data;
+
+    uint32_t prev_i2 = (uint32_t)-1;
+    float slope = 1.0f;
+
+    for (uint32_t r = src0_start_row; r < src0_end_row; ++r) {
+        uint32_t i1 = fastmodulo(r, ne01, &smctx->fastdiv_ne01);
+        uint32_t r_div_ne01 = fastdiv(r, &smctx->fastdiv_ne01);
+        uint32_t i2 = fastmodulo(r_div_ne01, ne02, &smctx->fastdiv_ne02);
+        uint32_t i3 = fastdiv(r_div_ne01, &smctx->fastdiv_ne02);
+
+        // Map to original logic indices
+        // i01 = i1
+        // i02 = i2
+        // i03 = i3
+
+        const uint32_t i11 = i1;
+        // const uint32_t i12 = i2 % ne12;
+        // const uint32_t i13 = i3 % ne13;
+
+        uint32_t i12, i13;
+        if (ne12 == ne02) {
+             i12 = i2;
+        } else {
+             i12 = fastmodulo(i2, ne12, &smctx->fastdiv_ne12);
+        }
+
+        if (ne13 == ne03) {
+             i13 = i3;
+        } else {
+             i13 = fastmodulo(i3, ne13, &smctx->fastdiv_ne13);
+        }
+
+        // ALiBi
+        if (i2 != prev_i2) {
+            const uint32_t h = i2;  // head
+
+            slope = (smctx->max_bias > 0.0f) ?
+                        h < smctx->n_head_log2 ?
+                        powf(smctx->m0, h + 1) :
+                        powf(smctx->m1, 2 * (h - smctx->n_head_log2) + 1) :
+                        1.0f;
+            prev_i2 = i2;
+        }
+
+        float * sp = (float *) ((char *) octx->src0.data + i1 * nb01 + i2 * nb02 + i3 * nb03);
+        float * dp = (float *) ((char *) octx->dst.data + i1 * nb1 + i2 * nb2 + i3 * nb3);
+
+        // broadcast the mask across rows
+        __fp16 * mp_f16 = (smctx->use_src1) ?
+                              (__fp16 *) ((char *) octx->src1.data + i11 * nb11 + i12 * nb12 + i13 * nb13) :
+                              NULL;
+        float *  mp_f32 = (smctx->use_src1) ?
+                              (float *) ((char *) octx->src1.data + i11 * nb11 + i12 * nb12 + i13 * nb13) :
+                              NULL;
+
+        if ((1 == opt_path) && (mp_f32) && !(smctx->use_f16)) {
+            hvx_fast_softmax_prep_f32((const uint8_t *) sp, (uint8_t *) wp0, ne00, smctx->scale,
+                                      (const uint8_t *) mp_f32, slope);
+        } else {
+            hvx_scale_f32((uint8_t *) wp0, (const uint8_t *) sp, ne00, smctx->scale);
+            if (mp_f32) {
+                if (smctx->use_f16) {
+                    for (int i = 0; i < ne00; ++i) {
+                        wp0[i] += slope * (float) mp_f16[i];
+                    }
+                } else {
+                    for (int i = 0; i < ne00; ++i) {
+                        wp0[i] += slope * mp_f32[i];
+                    }
+                }
+            }
+        }
+
+        if (1 == opt_path) {
+            hvx_fast_softmax_f32((const uint8_t *) wp0, (uint8_t *) dp, (uint8_t *) wp1, ne00);
+        } else {
+            float max = hvx_reduce_max_f32((const uint8_t *) wp0, ne00);
+            float sum = hvx_softmax_f32((const uint8_t *) wp0, (uint8_t *) wp2, (uint8_t *) wp1, ne00, max);
+            sum       = sum > 0.0 ? (1.0 / sum) : 1;
+            hvx_scale_f32((uint8_t *) dp, (const uint8_t *) wp2, ne00, sum);
+        }
+    }
 
     t2 = HAP_perf_get_qtimer_count();
 
     FARF(HIGH, "softmax-f32 %d/%d/%d/%d: %ux%ux%ux%u (%u:%u) x %ux%ux%ux%u -> %ux%ux%ux%u usec %u\n", ith, nth,
-         softmax_ctx->use_f16, opt_path, ne00, ne01, ne02, ne03, src0_start_row, src0_end_row, ne10, ne11, ne12, ne13,
+         smctx->use_f16, opt_path, ne00, ne01, ne02, ne03, src0_start_row, src0_end_row, ne10, ne11, ne12, ne13,
          ne0, ne1, ne2, ne3, (unsigned) HAP_perf_qtimer_count_to_us(t2 - t1));
 }
 
-static void softmax_job_dispatcher_f32(unsigned int n, unsigned int i, void * p_data) {
-    struct softmax_th_ctx * p_softmax_ctx = (struct softmax_th_ctx *) p_data;
-    softmax_job_f32_per_thread(p_softmax_ctx, n, i);
-}
-
 static int execute_op_softmax_f32(struct htp_ops_context * octx) {
     int err = HTP_STATUS_OK;
 
@@ -312,17 +340,12 @@ static int execute_op_softmax_f32(struct htp_ops_context * octx) {
     const struct htp_tensor * src1 = &octx->src1;
     struct htp_tensor *       dst  = &octx->dst;
 
-    worker_callback_t op_func;
-    const char *      op_type = NULL;
-
-    struct softmax_th_ctx softmax_ctx;
+    struct htp_softmax_context smctx;
+    const char * op_type = "softmax-f32";
 
     switch (octx->op) {
         case HTP_OP_SOFTMAX:
-            op_func = softmax_job_dispatcher_f32;
-            op_type = "softmax-f32";
-
-            init_softmax_ctx(&softmax_ctx, octx);
+            init_softmax_ctx(&smctx, octx);
             break;
 
         default:
@@ -342,6 +365,9 @@ static int execute_op_softmax_f32(struct htp_ops_context * octx) {
     octx->src0_spad.size = hex_round_up(src0_row_size, 128) * n_threads;
     octx->src1_spad.size = hex_round_up(src1_row_size, 128) * n_threads;
 
+    // Use stride for calculating offset
+    smctx.spad_stride = hex_round_up(src0_row_size, 128);
+
     size_t spad_size = octx->src0_spad.size + octx->src1_spad.size + octx->dst_spad.size;
 
     if (src1->ne[0]) {
@@ -371,8 +397,8 @@ static int execute_op_softmax_f32(struct htp_ops_context * octx) {
 
     if (!(octx->flags & HTP_OPFLAGS_SKIP_COMPUTE)) {
         uint32_t n_jobs             = MIN(n_threads, src0_nrows);
-        octx->src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
-        worker_pool_run_func(octx->ctx->worker_pool, op_func, &softmax_ctx, n_jobs);
+        smctx.src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
+        worker_pool_run_func(octx->ctx->worker_pool, softmax_job_f32, &smctx, n_jobs);
     }
 
     return err;

ggml/src/ggml-hexagon/htp/sum-rows-ops.c

@@ -17,7 +17,6 @@
 #include "htp-msg.h"
 #include "htp-ops.h"
 
-
 #define sum_rows_preamble                       \
     struct htp_tensor *src0 =  &octx->src0;\
     struct htp_tensor *dst  = &octx->dst;  \
@@ -42,53 +41,54 @@
     const uint32_t  nb2 = dst->nb[2];      \
     const uint32_t  nb3 = dst->nb[3];      \
 
-static int sum_rows_thread_f32(struct htp_ops_context * octx, const int nth, const int ith) {
-    sum_rows_preamble;
+struct sum_rows_context {
+    const uint8_t * src_data;
+    uint8_t       * dst_data;
+    uint32_t        ne00;
+    size_t          src_stride;
+    size_t          dst_stride;
+    uint32_t        rows_per_thread;
+    uint32_t        total_rows;
+    bool            opt_path;
+};
 
-    const uint32_t src0_nrows_per_thread  = octx->src0_nrows_per_thread;
-    const size_t src0_row_size = nb01;
-    const size_t dst_row_size  = nb1;
+static void sum_rows_thread_f32(unsigned int nth, unsigned int ith, void *data) {
+    const struct sum_rows_context * smctx = (const struct sum_rows_context *) data;
 
-    const uint32_t src0_nrows = ne01 * ne02 * ne03;  // src0 rows
+    const uint32_t rows_per_thread = smctx->rows_per_thread;
+    const uint32_t total_rows      = smctx->total_rows;
 
-    const uint32_t src0_start_row = src0_nrows_per_thread * ith;
-    const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
+    const uint32_t start_row = rows_per_thread * ith;
+    const uint32_t end_row   = MIN(start_row + rows_per_thread, total_rows);
 
-    // no work for this thread
-    if (src0_start_row >= src0_end_row) {
-        return HTP_STATUS_OK;
+    if (start_row >= end_row) {
+        return;
     }
 
-    int opt_path   = 0;
-    if ((0 == hex_is_aligned((void *) src0->data, VLEN)) && !(nb01 & (VLEN - 1))) {
-        opt_path = 1;
-    }
+    const size_t   src_stride = smctx->src_stride;
+    const size_t   dst_stride = smctx->dst_stride;
+    const uint32_t ne00       = smctx->ne00;
+    const bool     opt_path   = smctx->opt_path;
 
-    const uint8_t * restrict data_src = (const uint8_t *) src0->data;
-    uint8_t * restrict data_dst       = (uint8_t *) dst->data;
+    const float * restrict src_th = (const float *) (smctx->src_data + (start_row * src_stride));
+    float       * restrict dst_th = (float *)       (smctx->dst_data + (start_row * dst_stride));
 
-    const float * restrict src_th = (float *) (data_src + (src0_start_row * src0_row_size));
-    float * restrict dst_th       = (float *) (data_dst + (src0_start_row * dst_row_size));
+    // Calculate actual number of rows for this thread
+    const uint32_t n_rows = end_row - start_row;
 
-    for (uint32_t ir = 0; ir < src0_nrows_per_thread; ir++) {
-        const float * restrict src_local = src_th + (ir * ne00);
+    for (uint32_t ir = 0; ir < n_rows; ir++) {
+        const float * restrict src_local = src_th + (ir * (src_stride / sizeof(float)));
 
-        if (ir + 1 < src0_nrows_per_thread) {
-            hex_l2fetch(src_local + ne00, src0_row_size, src0_row_size, 1);
+        if (ir + 1 < n_rows) {
+            hex_l2fetch(src_local + (src_stride / sizeof(float)), src_stride, src_stride, 1);
         }
 
-        if (1 == opt_path) {
+        if (opt_path) {
             dst_th[ir] = hvx_reduce_sum_f32_a((const uint8_t *) src_local, ne00);
         } else {
             dst_th[ir] = hvx_reduce_sum_f32((const uint8_t *) src_local, ne00);
         }
     }
-
-    return HTP_STATUS_OK;
-}
-
-static void sum_rows_work_f32(unsigned int n, unsigned int i, void *data) {
-    sum_rows_thread_f32((struct htp_ops_context *) data, n, i);
 }
 
 int op_sum_rows(struct htp_ops_context * octx) {
@@ -106,10 +106,25 @@ int op_sum_rows(struct htp_ops_context * octx) {
     const uint32_t src0_nrows = ne01 * ne02 * ne03;
 
     uint32_t n_jobs = MIN(n_threads, src0_nrows);
-    octx->src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
+    uint32_t rows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
 
-    worker_pool_run_func(octx->ctx->worker_pool, sum_rows_work_f32, octx, n_jobs);
+    bool opt_path = false;
+    if ((0 == hex_is_aligned((void *) src0->data, VLEN)) && !(nb01 & (VLEN - 1))) {
+        opt_path = true;
+    }
+
+    struct sum_rows_context smctx = {
+        .src_data        = (const uint8_t *) src0->data,
+        .dst_data        = (uint8_t *) dst->data,
+        .ne00            = ne00,
+        .src_stride      = nb01,
+        .dst_stride      = nb1,
+        .rows_per_thread = rows_per_thread,
+        .total_rows      = src0_nrows,
+        .opt_path        = opt_path,
+    };
+
+    worker_pool_run_func(octx->ctx->worker_pool, sum_rows_thread_f32, &smctx, n_jobs);
 
     return HTP_STATUS_OK;
 }
-

ggml/src/ggml-hexagon/htp/unary-ops.c

@@ -17,6 +17,28 @@
 #include "htp-msg.h"
 #include "htp-ops.h"
 
+struct htp_unary_context {
+    struct htp_ops_context * octx;
+
+    // Precomputed values
+    const uint8_t *           data_src0;
+    uint8_t *                 data_dst;
+
+    size_t                    src0_row_size;
+    size_t                    dst_row_size;
+
+    size_t                    src0_row_size_aligned;
+    size_t                    dst_row_size_aligned;
+
+    size_t                    src0_spad_half_size;
+    size_t                    dst_spad_half_size;
+
+    uint32_t                  block;
+    uint32_t                  src0_nrows;
+    uint32_t                  src0_nrows_per_thread;
+    uint32_t                  nc;
+};
+
 #define htp_unary_preamble            \
     const uint32_t ne00 = src->ne[0]; \
     const uint32_t ne01 = src->ne[1]; \
@@ -57,8 +79,7 @@ static void hvx_fast_rms_norm_f32(const uint8_t * restrict src,
         sum_v         = Q6_Vqf32_vadd_Vqf32Vqf32(sum_v, v2);
     }
 
-    HVX_Vector reduced_sum = hvx_vec_reduce_sum_f32(Q6_Vsf_equals_Vqf32(sum_v));
-    sum_v                  = hvx_vec_repl4(reduced_sum);
+    sum_v = hvx_vec_reduce_sum_f32(Q6_Vsf_equals_Vqf32(sum_v)); // replicated over all lanes
 
     HVX_Vector t_v            = hvx_vec_splat_f32((float) num_elems);
     HVX_Vector denom_v        = hvx_vec_inverse_f32(t_v);
@@ -75,128 +96,95 @@ static void hvx_fast_rms_norm_f32(const uint8_t * restrict src,
     }
 }
 
-static void scale_htp_f32(const float * restrict src,
-                          float * restrict dst,
-                          uint8_t * restrict spad,
-                          const uint32_t num_rows,
-                          const uint32_t row_elems,
-                          const size_t   row_size,
-                          int32_t *      op_params,
-                          int            opt_path) {
+static void scale_f32(const float * restrict src,
+                      float * restrict dst,
+                      uint8_t * restrict spad,
+                      const uint32_t num_rows,
+                      const uint32_t row_elems,
+                      const size_t   row_size,
+                      int32_t *      op_params) {
     float scale = 0.f;
     float bias  = 0.f;
     memcpy(&scale, &op_params[0], sizeof(float));
     memcpy(&bias,  &op_params[1], sizeof(float));
 
     for (uint32_t ir = 0; ir < num_rows; ir++) {
-        const float * restrict src_local = src + (ir * row_elems);
-        float * restrict dst_local       = dst + (ir * row_elems);
+        const uint8_t * restrict src_local = (const uint8_t *)src + (ir * row_size);
+        uint8_t * restrict dst_local       = (uint8_t *)dst + (ir * row_size);
 
-        if (ir + 1 < num_rows) {
-            hex_l2fetch(src_local + row_elems, row_size, row_size, 1);
-        }
-
-        hvx_scale_offset_f32((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems, scale, bias);
+        hvx_scale_offset_f32_aa((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems, scale, bias);
     }
 }
 
-static void rms_norm_htp_f32(const float * restrict src,
-                             float * restrict dst,
-                             uint8_t * restrict spad,
-                             const uint32_t num_rows,
-                             const uint32_t row_elems,
-                             const size_t   row_size,
-                             int32_t *      op_params,
-                             int            opt_path) {
+static void rms_norm_f32(const float * restrict src,
+                         float * restrict dst,
+                         uint8_t * restrict spad,
+                         const uint32_t num_rows,
+                         const uint32_t row_elems,
+                         const size_t   row_size,
+                         int32_t *      op_params) {
     float epsilon = 0.f;
     memcpy(&epsilon, op_params, sizeof(float));
 
     for (uint32_t ir = 0; ir < num_rows; ir++) {
-        const float * restrict src_local = src + (ir * row_elems);
-        float * restrict dst_local       = dst + (ir * row_elems);
+        const uint8_t * restrict src_local = (const uint8_t *)src + (ir * row_size);
+        uint8_t * restrict dst_local       = (uint8_t *)dst + (ir * row_size);
 
-        if (ir + 1 < num_rows) {
-            hex_l2fetch(src_local + row_elems, row_size, row_size, 1);
-        }
-
-        if (1 == opt_path) {
-            hvx_fast_rms_norm_f32((const uint8_t *) src_local, (uint8_t *) dst_local, spad, row_elems, epsilon);
-        } else {
-            float sum = hvx_sum_of_squares_f32((const uint8_t *) src_local, row_elems);
-
-            const float mean  = sum / row_elems;
-            const float scale = 1.0f / sqrtf(mean + epsilon);
-
-            hvx_scale_f32((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems, scale);
-        }
+        hvx_fast_rms_norm_f32((const uint8_t *) src_local, (uint8_t *) dst_local, spad, row_elems, epsilon);
     }
 }
 
-static void sqr_htp_f32(const float * restrict src,
-                          float * restrict dst,
-                          uint8_t * restrict spad,
-                          const uint32_t num_rows,
-                          const uint32_t row_elems,
-                          const size_t   row_size,
-                          int32_t *      op_params,
-                          int            opt_path) {
+static void sqr_f32(const float * restrict src,
+                    float * restrict dst,
+                    uint8_t * restrict spad,
+                    const uint32_t num_rows,
+                    const uint32_t row_elems,
+                    const size_t   row_size,
+                    int32_t *      op_params) {
 
     for (uint32_t ir = 0; ir < num_rows; ir++) {
-        const float * restrict src_local = src + (ir * row_elems);
-        float * restrict dst_local       = dst + (ir * row_elems);
+        const uint8_t * restrict src_local = (const uint8_t *)src + (ir * row_size);
+        uint8_t * restrict dst_local       = (uint8_t *)dst + (ir * row_size);
 
-        if (ir + 1 < num_rows) {
-            hex_l2fetch(src_local + row_elems, row_size, row_size, 1);
-        }
-
-        if (1 == opt_path) {
-            hvx_sqr_f32_aa((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems);
-        } else {
-            hvx_sqr_f32((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems);
-        }
+        hvx_sqr_f32_aa((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems);
     }
 }
 
-static void sqrt_htp_f32(const float * restrict src,
-                          float * restrict dst,
-                          uint8_t * restrict spad,
-                          const uint32_t num_rows,
-                          const uint32_t row_elems,
-                          const size_t   row_size,
-                          int32_t *      op_params,
-                          int            opt_path) {
+static void sqrt_f32(const float * restrict src,
+                     float * restrict dst,
+                     uint8_t * restrict spad,
+                     const uint32_t num_rows,
+                     const uint32_t row_elems,
+                     const size_t   row_size,
+                     int32_t *      op_params) {
 
     for (uint32_t ir = 0; ir < num_rows; ir++) {
-        const float * restrict src_local = src + (ir * row_elems);
-        float * restrict dst_local       = dst + (ir * row_elems);
+        const uint8_t * restrict src_local = (const uint8_t *)src + (ir * row_size);
+        uint8_t * restrict dst_local       = (uint8_t *)dst + (ir * row_size);
 
-        if (ir + 1 < num_rows) {
-            hex_l2fetch(src_local + row_elems, row_size, row_size, 1);
-        }
-
-        if (1 == opt_path) {
-            hvx_sqrt_f32_aa((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems);
-        } else {
-            hvx_sqrt_f32((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems);
-        }
+        hvx_sqrt_f32_aa((uint8_t *) dst_local, (const uint8_t *) src_local, row_elems);
     }
 }
 
-static void unary_job_f32_per_thread(const struct htp_tensor * src,
-                                     struct htp_tensor *       dst,
-                                     uint8_t *                 spad,
-                                     int                       htp_op,
-                                     int32_t *                 op_params,
-                                     uint32_t                  nth,
-                                     uint32_t                  ith,
-                                     uint32_t                  src0_nrows_per_thread) {
+static void unary_job_f32_per_thread(unsigned int nth, unsigned int ith, void * data) {
+    const struct htp_unary_context * uctx = (const struct htp_unary_context *) data;
+    struct htp_ops_context * octx = uctx->octx;
+    const struct htp_tensor * src = &octx->src0;
+    const struct htp_tensor * dst = &octx->dst;
+
     htp_unary_preamble;
 
-    const size_t src0_row_size = nb01;
-    const size_t dst_row_size  = nb1;
+    int                       htp_op = octx->op;
+    int32_t *                 op_params = octx->op_params;
+    uint32_t                  src0_nrows_per_thread = uctx->src0_nrows_per_thread;
 
-    const uint32_t src0_nrows = ne01 * ne02 * ne03;  // src0 rows
+    const size_t src0_row_size = uctx->src0_row_size;
+    const size_t dst_row_size  = uctx->dst_row_size;
 
+    const size_t src0_row_size_aligned = uctx->src0_row_size_aligned;
+    const size_t dst_row_size_aligned  = uctx->dst_row_size_aligned;
+
+    const uint32_t src0_nrows = uctx->src0_nrows;
     const uint32_t src0_start_row = src0_nrows_per_thread * ith;
     const uint32_t src0_end_row   = MIN(src0_start_row + src0_nrows_per_thread, src0_nrows);
 
@@ -208,79 +196,104 @@ static void unary_job_f32_per_thread(const struct htp_tensor * src,
     uint64_t t1, t2;
     t1 = HAP_perf_get_qtimer_count();
 
-    int is_aligned = 1;
-    int opt_path   = 0;
-    if ((0 == hex_is_aligned((void *) src->data, VLEN)) || (0 == hex_is_aligned((void *) dst->data, VLEN))) {
-        is_aligned = 0;
-    }
-    if ((1 == is_aligned) && !(nb01 & (VLEN - 1))) {
-        opt_path = 1;
+    const uint8_t * restrict data_src = uctx->data_src0;
+    uint8_t * restrict       data_dst = uctx->data_dst;
+
+    uint8_t * src0_spad_data = octx->src0_spad.data + (ith * octx->src0_spad.size_per_thread);
+    uint8_t * dst_spad_data  = octx->dst_spad.data  + (ith * octx->dst_spad.size_per_thread);
+
+    size_t src0_spad_half_size = uctx->src0_spad_half_size;
+    size_t dst_spad_half_size  = uctx->dst_spad_half_size;
+
+    const int BLOCK = uctx->block;
+    if (BLOCK == 0) {
+        FARF(ERROR, "unary-f32 : current VTCM reservation %zu is too small for even 1 row per thread, needed at least %zu\n",
+             octx->src0_spad.size_per_thread, src0_row_size_aligned);
+        return;
     }
 
-    const uint8_t * restrict data_src = (const uint8_t *) src->data;
-    uint8_t * restrict data_dst       = (uint8_t *) dst->data;
+    dma_queue * dma_queue = octx->ctx->dma[ith];
 
-    const float * restrict src_th = (float *) (data_src + (src0_start_row * src0_row_size));
-    float * restrict dst_th       = (float *) (data_dst + (src0_start_row * dst_row_size));
-    uint8_t * restrict spad_th    = (uint8_t *) spad + (ith * nb01);
+    for (uint32_t ir = src0_start_row, spad_idx = 0; ir < src0_end_row && spad_idx < 2; ir += BLOCK, spad_idx++) {
+        const uint32_t block_size = MIN(BLOCK, src0_end_row - ir);
 
-    switch (htp_op) {
-        case HTP_OP_RMS_NORM:
-            rms_norm_htp_f32(src_th, dst_th, spad_th, src0_end_row - src0_start_row, ne0, nb1, op_params, opt_path);
-            break;
-        case HTP_OP_SCALE:
-            scale_htp_f32(src_th, dst_th, spad_th, src0_end_row - src0_start_row, ne0, nb1, op_params, opt_path);
-            break;
-        case HTP_OP_SQR:
-            sqr_htp_f32(src_th, dst_th, spad_th, src0_end_row - src0_start_row, ne0, nb1, op_params, opt_path);
-            break;
-        case HTP_OP_SQRT:
-            sqrt_htp_f32(src_th, dst_th, spad_th, src0_end_row - src0_start_row, ne0, nb1, op_params, opt_path);
-            break;
+        // Dummy DMA transation for sequencing (interleaving dst,src,dst,...)
+        dma_queue_push_vtcm_to_ddr(dma_queue,
+            dma_make_ptr(data_dst, dst_spad_data + (spad_idx * dst_spad_half_size)),
+            dst_row_size, dst_row_size_aligned, 0);
 
-        default:
-            break;
+        dma_queue_push_ddr_to_vtcm(dma_queue,
+            dma_make_ptr(src0_spad_data + (spad_idx * src0_spad_half_size), data_src + (ir * src0_row_size)),
+            src0_row_size_aligned, src0_row_size, block_size);
     }
 
+    for (uint32_t ir = src0_start_row; ir < src0_end_row; ir += BLOCK) {
+        const uint32_t block_size = MIN(BLOCK, src0_end_row - ir);
+
+        float * dst_spad  = (float *) dma_queue_pop(dma_queue).src;
+        float * src0_spad = (float *) dma_queue_pop(dma_queue).dst;
+
+        // Process block in VTCM
+        switch (htp_op) {
+            case HTP_OP_RMS_NORM:
+                rms_norm_f32(src0_spad, dst_spad, NULL, block_size, ne0, src0_row_size_aligned, op_params);
+                break;
+            case HTP_OP_SCALE:
+                scale_f32(src0_spad, dst_spad, NULL, block_size, ne0, src0_row_size_aligned, op_params);
+                break;
+            case HTP_OP_SQR:
+                sqr_f32(src0_spad, dst_spad, NULL, block_size, ne0, src0_row_size_aligned, op_params);
+                break;
+            case HTP_OP_SQRT:
+                sqrt_f32(src0_spad, dst_spad, NULL, block_size, ne0, src0_row_size_aligned, op_params);
+                break;
+            default:
+                break;
+        }
+
+        dma_queue_push_vtcm_to_ddr(dma_queue,
+            dma_make_ptr(data_dst + (ir * dst_row_size), dst_spad),
+            dst_row_size, dst_row_size_aligned, block_size);
+
+        // prefetch N+2 loop iteration if any
+        const uint32_t pref_block = (ir + BLOCK * 2);
+        if (pref_block < src0_end_row) {
+            const uint32_t pref_block_size = MIN(BLOCK, src0_end_row - pref_block);
+            dma_queue_push_ddr_to_vtcm(dma_queue,
+                dma_make_ptr(src0_spad, data_src + (pref_block * src0_row_size)),
+                src0_row_size_aligned, src0_row_size, pref_block_size);
+        }
+    }
+
+    dma_queue_flush(dma_queue);
+
     t2 = HAP_perf_get_qtimer_count();
 
-    FARF(HIGH, "unary-f32 %d/%d/%d: %ux%ux%ux%u (%u:%u) -> %ux%ux%ux%u usec %u\n", ith, nth, opt_path, src->ne[0],
+    FARF(HIGH, "unary-f32 %d/%d: %ux%ux%ux%u (%u:%u) -> %ux%ux%ux%u usec %u\n", ith, nth, src->ne[0],
          src->ne[1], src->ne[2], src->ne[3], src0_start_row, src0_end_row, dst->ne[0], dst->ne[1], dst->ne[2],
          dst->ne[3], (unsigned) HAP_perf_qtimer_count_to_us(t2 - t1));
 }
 
-static void unary_job_dispatcher_f32(unsigned int n, unsigned int i, void * data) {
-    struct htp_ops_context * octx = (struct htp_ops_context *) data;
-
-    unary_job_f32_per_thread(&octx->src0, &octx->dst, octx->src0_spad.data, octx->op, octx->op_params, n, i,
-                             octx->src0_nrows_per_thread);
-}
-
 static int execute_op_unary_f32(struct htp_ops_context * octx) {
     int err = HTP_STATUS_OK;
 
     const struct htp_tensor * src0 = &octx->src0;
     struct htp_tensor *       dst  = &octx->dst;
 
-    worker_callback_t unary_op_func;
-    const char *      op_type = NULL;
+    const char * op_type = NULL;
 
     switch (octx->op) {
         case HTP_OP_RMS_NORM:
-            unary_op_func = unary_job_dispatcher_f32;
-            op_type       = "rmsnorm-f32";
+            op_type = "rmsnorm-f32";
             break;
         case HTP_OP_SCALE:
-            unary_op_func = unary_job_dispatcher_f32;
-            op_type       = "scale-f32";
+            op_type = "scale-f32";
             break;
         case HTP_OP_SQR:
-            unary_op_func = unary_job_dispatcher_f32;
-            op_type       = "sqr-f32";
+            op_type = "sqr-f32";
             break;
         case HTP_OP_SQRT:
-            unary_op_func = unary_job_dispatcher_f32;
-            op_type       = "sqrt-f32";
+            op_type = "sqrt-f32";
             break;
 
         default:
@@ -294,32 +307,61 @@ static int execute_op_unary_f32(struct htp_ops_context * octx) {
     const size_t src0_row_size = src0->nb[1];
     const size_t dst_row_size  = dst->nb[1];
 
-    // VTCM scratchpads for all tensors
-    octx->dst_spad.size  = hex_round_up(dst_row_size, 128) * n_threads;
-    octx->src0_spad.size = hex_round_up(src0_row_size, 128) * n_threads;
+    const size_t src0_row_size_aligned = hex_round_up(src0_row_size, VLEN);
+    const size_t dst_row_size_aligned  = hex_round_up(dst_row_size, VLEN);
 
-    size_t spad_size = octx->src0_spad.size + octx->dst_spad.size;
+    // VTCM scratchpads for all tensors
+    // N rows per thread, padded to HVX vector size
+    // Double buffering requires 2x size per buffer
+
+    size_t spad_size_per_row   = 2 * (src0_row_size_aligned + dst_row_size_aligned);
+    size_t vtcm_row_per_thread = (octx->ctx->vtcm_size)/ (n_threads * spad_size_per_row);
+
+    // Make sure the reserved vtcm size is sufficient
+    if (vtcm_row_per_thread == 0) {
+        FARF(ERROR, "unary-%s : current VTCM reservation %zu is too small, needed %zu\n", op_type, octx->ctx->vtcm_size,
+             spad_size_per_row * n_threads);
+        return HTP_STATUS_VTCM_TOO_SMALL;
+    }
+
+    octx->src0_spad.size_per_thread = src0_row_size_aligned * vtcm_row_per_thread * 2;
+    octx->dst_spad.size_per_thread  = dst_row_size_aligned * vtcm_row_per_thread * 2;
+
+    octx->src0_spad.size = n_threads * octx->src0_spad.size_per_thread;
+    octx->dst_spad.size  = n_threads * octx->dst_spad.size_per_thread;
+
+    octx->src0_spad.data = octx->ctx->vtcm_base;
+    octx->dst_spad.data  = octx->src0_spad.data + octx->src0_spad.size;
 
     FARF(HIGH, "%s: (%ux%ux%ux%u) -> (%ux%ux%ux%u) : src0-spad-size %u src1-spad-size %u dst-spad-size %u\n", op_type,
          src0->ne[0], src0->ne[1], src0->ne[2], src0->ne[3], dst->ne[0], dst->ne[1], dst->ne[2], dst->ne[3],
          octx->src0_spad.size, octx->src1_spad.size, octx->dst_spad.size);
 
-    // Make sure the reserved vtcm size is sufficient
-    if (octx->ctx->vtcm_size < spad_size) {
-        FARF(ERROR, "unary-%s : current VTCM reservation %zu is too small, needed %zu\n", op_type, octx->ctx->vtcm_size,
-             spad_size);
-        return HTP_STATUS_VTCM_TOO_SMALL;
-    }
-
-    octx->src0_spad.data = octx->ctx->vtcm_base;
-    octx->dst_spad.data  = octx->src0_spad.data + octx->src0_spad.size;
-
     if (!(octx->flags & HTP_OPFLAGS_SKIP_COMPUTE)) {
         uint32_t n_jobs = MIN(n_threads, src0_nrows);
 
-        octx->src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs;
+        struct htp_unary_context uctx = {
+            .octx                  = octx,
+            .src0_nrows_per_thread = (src0_nrows + n_jobs - 1) / n_jobs,
+            .src0_nrows            = src0_nrows,
 
-        worker_pool_run_func(octx->ctx->worker_pool, unary_op_func, octx, n_jobs);
+            .data_src0             = (const uint8_t *)src0->data,
+            .data_dst              = (uint8_t *)dst->data,
+
+            .src0_row_size         = src0_row_size,
+            .dst_row_size          = dst_row_size,
+
+            .src0_row_size_aligned = src0_row_size_aligned,
+            .dst_row_size_aligned  = dst_row_size_aligned,
+
+            .src0_spad_half_size   = octx->src0_spad.size_per_thread / 2,
+            .dst_spad_half_size    = octx->dst_spad.size_per_thread / 2,
+
+            .block                 = (octx->src0_spad.size_per_thread / 2) / src0_row_size_aligned,
+            .nc                    = src0->ne[0],
+        };
+
+        worker_pool_run_func(octx->ctx->worker_pool, unary_job_f32_per_thread, &uctx, n_jobs);
     }
 
     return err;

ggml/src/ggml-vulkan/ggml-vulkan.cpp

@@ -403,19 +403,20 @@ enum FaCodePath {
 };
 
 struct vk_fa_pipeline_state {
-    vk_fa_pipeline_state(uint32_t HSK, uint32_t HSV, bool small_rows, bool small_cache, FaCodePath path, bool aligned, bool f32acc, uint32_t flags)
-        : HSK(HSK), HSV(HSV), small_rows(small_rows), small_cache(small_cache), path(path), aligned(aligned), f32acc(f32acc), flags(flags) {}
-
     uint32_t HSK, HSV;
-    bool small_rows, small_cache;
+    uint32_t Br, Bc;
+    uint32_t D_split, row_split;
+    bool shmem_staging;
     FaCodePath path;
+    uint32_t workgroup_size, subgroup_size;
     bool aligned;
     bool f32acc;
     uint32_t flags;
+    uint32_t limit_occupancy_shmem;
 
     bool operator<(const vk_fa_pipeline_state &b) const {
-        return std::tie(HSK, HSV, small_rows, small_cache, path, aligned, f32acc, flags) <
-               std::tie(b.HSK, b.HSV, b.small_rows, b.small_cache, b.path, b.aligned, b.f32acc, b.flags);
+        return std::tie(HSK, HSV, Br, Bc, D_split, row_split, shmem_staging, path, workgroup_size, subgroup_size, aligned, f32acc, flags, limit_occupancy_shmem) <
+               std::tie(b.HSK, b.HSV, b.Br, b.Bc, b.D_split, b.row_split, b.shmem_staging, b.path, b.workgroup_size, b.subgroup_size, b.aligned, b.f32acc, b.flags, b.limit_occupancy_shmem);
     }
 };
 
@@ -623,6 +624,8 @@ struct vk_device_struct {
     // floor(log2(maxComputeWorkGroupInvocations))
     uint32_t max_workgroup_size_log2 {};
 
+    bool flash_attention_fp16;
+
     bool coopmat_support;
     bool coopmat_acc_f32_support {};
     bool coopmat_acc_f16_support {};
@@ -1656,6 +1659,7 @@ static bool vk_perf_logger_concurrent = false;
 static bool vk_enable_sync_logger = false;
 // number of calls between perf logger prints
 static uint32_t vk_perf_logger_frequency = 1;
+static std::string vk_pipeline_stats_filter;
 
 class vk_perf_logger {
   public:
@@ -2172,7 +2176,32 @@ static void ggml_vk_create_pipeline_func(vk_device& device, vk_pipeline& pipelin
         executableInfo.pipeline = pipeline->pipeline;
 
         auto statistics = device->device.getPipelineExecutableStatisticsKHR(executableInfo);
+
+        bool print_stats = !vk_pipeline_stats_filter.empty() &&
+                           pipeline->name.find(vk_pipeline_stats_filter) != std::string::npos;
+        if (print_stats) {
+            std::cerr << "ggml_vulkan: pipeline stats for " << pipeline->name << ":" << std::endl;
+        }
+
         for (auto & s : statistics) {
+            if (print_stats) {
+                std::cerr << "ggml_vulkan:   " << s.name.data() << ": ";
+                switch (s.format) {
+                    case vk::PipelineExecutableStatisticFormatKHR::eBool32:
+                        std::cerr << (s.value.b32 ? "true" : "false");
+                        break;
+                    case vk::PipelineExecutableStatisticFormatKHR::eInt64:
+                        std::cerr << s.value.i64;
+                        break;
+                    case vk::PipelineExecutableStatisticFormatKHR::eUint64:
+                        std::cerr << s.value.u64;
+                        break;
+                    case vk::PipelineExecutableStatisticFormatKHR::eFloat64:
+                        std::cerr << s.value.f64;
+                        break;
+                }
+                std::cerr << std::endl;
+            }
             // "Register Count" is reported by NVIDIA drivers.
             if (strcmp(s.name, "Register Count") == 0) {
                 VK_LOG_DEBUG(pipeline->name << " " << s.name << ": " << s.value.u64 << " registers");
@@ -2755,78 +2784,214 @@ static void ggml_vk_wait_events(vk_context& ctx, std::vector<vk::Event>&& events
     );
 }
 
-// number of rows/cols for flash attention shader
-static constexpr uint32_t flash_attention_num_small_rows = 32;
-static constexpr uint32_t scalar_flash_attention_num_small_rows = 1;
+struct vk_fa_tuning_params {
+    FaCodePath path;
+    uint32_t workgroup_size;
+    uint32_t subgroup_size;
+    uint32_t block_rows;
+    uint32_t block_cols;
+    uint32_t d_split;
+    uint32_t row_split;
+    bool shmem_staging;
+    bool disable_subgroups;
+    uint32_t limit_occupancy_shmem;
 
-static uint32_t get_fa_scalar_num_large_rows(uint32_t hsk, uint32_t hsv, bool small_cache) {
-    if (hsv >= 192) {
-        return 2;
-    } else if ((hsv | hsk) & 8 || small_cache) {
-        return 4;
-    } else {
-        return 8;
+    void print() const {
+        std::cerr << "path=" << path << " workgroup_size=" << workgroup_size << " subgroup_size=" << subgroup_size <<
+                     " block_rows=" << block_rows << " block_cols=" << block_cols << " d_split=" << d_split <<
+                     " row_split=" << row_split << " shmem_staging=" << shmem_staging << " disable_subgroups=" << disable_subgroups <<
+                     " limit_occupancy_shmem=" << limit_occupancy_shmem << std::endl;
     }
-}
+};
 
-// The FA coopmat1 shader assumes 16x16x16 matrix multiply support.
-// 128 threads split into four subgroups, each subgroup does 1/4
-// of the Bc dimension.
-static constexpr uint32_t coopmat1_flash_attention_num_large_rows = 16;
-static constexpr uint32_t scalar_flash_attention_Bc = 64;
-static constexpr uint32_t scalar_flash_attention_workgroup_size = 128;
+static bool ggml_vk_flash_attn_scalar_shmem_support(const vk_device& device, const vk_fa_tuning_params& params, uint32_t hsk, uint32_t hsv, bool f32acc);
+static bool ggml_vk_flash_attn_coopmat_shmem_support(const vk_device& device, const vk_fa_tuning_params& params, uint32_t hsk, uint32_t hsv, bool f32acc);
 
-static uint32_t get_fa_num_small_rows(FaCodePath path) {
-    if (path == FA_COOPMAT2) {
-        return flash_attention_num_small_rows;
+static vk_fa_tuning_params get_fa_tuning_params_scalar(const vk_device& device, uint32_t hsk, uint32_t hsv, uint32_t n_rows, uint32_t n_kv, ggml_type kv_type, bool f32acc) {
+    GGML_UNUSED(kv_type);
+
+    vk_fa_tuning_params result{};
+    result.path = FA_SCALAR;
+
+    if (device->vendor_id == VK_VENDOR_ID_INTEL) {
+        // Disable subgroup use due to performance issues when enforcing subgroup sizes
+        result.subgroup_size = 32;
+        result.disable_subgroups = true;
+    } else if (device->vendor_id == VK_VENDOR_ID_AMD && device->architecture != AMD_GCN) {
+        result.subgroup_size = n_rows < 4 ? 32 : device->subgroup_size;
     } else {
-        return scalar_flash_attention_num_small_rows;
+        result.subgroup_size = device->subgroup_size;
     }
-}
 
-static std::array<uint32_t, 2> fa_rows_cols(FaCodePath path, uint32_t hsk, uint32_t hsv, uint32_t clamp, ggml_type type, bool small_rows, bool small_cache) {
-    GGML_UNUSED(clamp);
+    // Row split splits the workgroup so that synchronization only has to happen within subgroups, which avoids barriers
+    uint32_t row_split_max_hsk = 64;
+    if (device->vendor_id == VK_VENDOR_ID_AMD && device->architecture != AMD_GCN && !device->uma) {
+        row_split_max_hsk = n_rows <= 8 ? 64 : 128;
+    }
+    result.row_split = (n_rows < 4 || hsk <= row_split_max_hsk) ? 1 : 4;
 
-    if (path == FA_SCALAR) {
-        if (small_rows) {
-            return {scalar_flash_attention_num_small_rows, 64};
+    if (result.subgroup_size > 32 && (n_rows < 4 || hsk < (result.row_split == 1 ? 128 : 64))) {
+        result.workgroup_size = result.subgroup_size * 2;
+    } else {
+        result.workgroup_size = result.subgroup_size * 4;
+    }
+
+    const uint32_t D = hsk | hsv;
+
+    const bool reduce_block_rows = D & 8 || n_kv < 1024 || device->vendor_id == VK_VENDOR_ID_INTEL;
+
+    if (n_rows == 1) {
+        result.block_rows = 1;
+        result.block_cols = 64;
+    } else {
+        // row_split 1 means higher register use per row, so block size has to be adjusted
+        if (result.row_split == 1) {
+            result.block_rows = n_rows == 2 ? 2 : ((n_rows <= 4 || reduce_block_rows) ? 4 : 8);
         } else {
-            if ((hsv | hsk) & 8) {
-                // HSV/HSK not being a multiple of 16 makes D_split smaller, which makes cols_per_iter
-                // larger, and Bc needs to be >= cols_per_thread. 64 is large enough, 32 is not.
-                return {get_fa_scalar_num_large_rows(hsk, hsv, small_cache), 64};
-            } else {
-                return {get_fa_scalar_num_large_rows(hsk, hsv, small_cache), 32};
-            }
+            result.block_rows = n_rows <= 4 ? 4 : ((n_rows <= 8 || reduce_block_rows) ? 8 : 16);
         }
+
+        result.block_cols = (D & 8) ? 64 : 32;
+    }
+
+    const uint32_t D_lsb = D ^ (D & (D-1));  // extract lowest set bit
+
+    result.d_split = std::min(std::min(result.subgroup_size, 8u), D_lsb / 4);
+
+    result.shmem_staging = (device->vendor_id == VK_VENDOR_ID_NVIDIA && hsk < 256 && hsv < 256) ? 1 : 0;
+
+    if (!reduce_block_rows && !ggml_vk_flash_attn_scalar_shmem_support(device, result, hsk, hsv, f32acc)) {
+        result.block_rows /= 2;
+    }
+
+    // On AMD RDNA, for small head sizes and big batch size the shader uses few registers, so too many subgroups get scheduled
+    // at once and end up thrashing the cache. Fix this by setting a large (unused) shmem buffer that reduces occupancy.
+    // This targets an occupancy of 4 subgroups per SIMD.
+    if (device->vendor_id == VK_VENDOR_ID_AMD && device->properties.limits.maxComputeSharedMemorySize == 65536) {
+        if (device->architecture != AMD_GCN && n_rows >= 64 && hsk <= 128) {
+            // 30kb target for hsk > 64, 26kb for <= 64 due to smaller workgroup size
+            // Values are guessed, tested on RDNA2
+            result.limit_occupancy_shmem = (hsk <= 64 ? 26 : 30) * 1024 / 4 / 4;
+        } else if (device->architecture == AMD_GCN && n_rows <= 8 && hsk >= 256) {
+            // Same thing for GCN, with an occupancy target of 2 subgroups per SIMD.
+            // Here low-batch FA with large head size is affected.
+            // n_rows < 4 switch because workgroup size switches from 128 to 256 there.
+            result.limit_occupancy_shmem = (n_rows < 4 ? 14 : 26) * 1024 / 4 / 4;
+        }
+    }
+
+    return result;
+}
+
+static vk_fa_tuning_params get_fa_tuning_params_coopmat1(const vk_device& device, uint32_t hsk, uint32_t hsv, uint32_t n_rows, uint32_t n_kv, ggml_type kv_type, bool f32acc) {
+    GGML_UNUSED(n_rows);
+    GGML_UNUSED(n_kv);
+    GGML_UNUSED(kv_type);
+    GGML_UNUSED(f32acc);
+
+    vk_fa_tuning_params result{};
+    result.path = FA_COOPMAT1;
+
+    const uint32_t D = hsk | hsv;
+
+    const uint32_t coopmat_block_rows = 16;
+    const uint32_t coopmat_block_cols = 16;
+
+    const uint32_t num_subgroups = 4;
+
+    result.block_rows = coopmat_block_rows;
+    result.block_cols = coopmat_block_cols * num_subgroups;
+    result.row_split = num_subgroups;
+    result.subgroup_size = device->subgroup_size;
+    result.workgroup_size = num_subgroups * result.subgroup_size;
+
+    const uint32_t D_lsb = D ^ (D & (D-1));  // extract lowest set bit
+    result.d_split = std::min(std::min(result.subgroup_size, 8u), D_lsb / 4);
+
+    result.shmem_staging = (device->vendor_id == VK_VENDOR_ID_NVIDIA && hsk < 256 && hsv < 256) ? 1 : 0;
+
+    return result;
+}
+
+static vk_fa_tuning_params get_fa_tuning_params_coopmat2(const vk_device& device, uint32_t hsk, uint32_t hsv, uint32_t n_rows, uint32_t n_kv, ggml_type kv_type, bool f32acc) {
+    GGML_UNUSED(n_kv);
+    GGML_UNUSED(f32acc);
+
+    vk_fa_tuning_params result{};
+    result.path = FA_COOPMAT2;
+
+    const uint32_t D = hsk | hsv;
+
+    const bool small_rows = n_rows < 32;
+
+    if (small_rows) {
+        result.block_rows = 32;
+        result.block_cols = 32;
+    } else if (ggml_is_quantized(kv_type) || hsk >= 256 || hsv >= 256) {
+        result.block_rows = (hsk >= 512 || hsv >= 512) ? 32 : 64;
+        result.block_cols = 32;
+    } else {
+        result.block_rows = 64;
+        result.block_cols = 64;
+    }
+
+    result.subgroup_size = device->subgroup_size;
+    result.workgroup_size = (small_rows && (D % 32) == 0) ? 256 : 128;
+
+    return result;
+}
+
+static vk_fa_tuning_params get_fa_tuning_params(const vk_device& device, uint32_t hsk, uint32_t hsv, uint32_t n_rows, uint32_t n_kv, ggml_type kv_type, bool f32acc) {
+    FaCodePath path = device->coopmat2 ? FA_COOPMAT2 :
+                      device->coopmat1_fa_support ? FA_COOPMAT1 : FA_SCALAR;
+
+    if (path == FA_COOPMAT1 && device->architecture == vk_device_architecture::NVIDIA_TURING) {
+        // Nvidia compiler bug, see https://github.com/ggml-org/llama.cpp/pull/19075#issuecomment-3820716090
+        path = FA_SCALAR;
     }
 
     if (path == FA_COOPMAT1) {
-        if (small_rows) {
-            return {scalar_flash_attention_num_small_rows, scalar_flash_attention_Bc};
-        } else {
-            return {coopmat1_flash_attention_num_large_rows, scalar_flash_attention_Bc};
+        bool shape_ok = (f32acc && device->coopmat_support_16x16x16_f32acc) ||
+                        (!f32acc && device->coopmat_support_16x16x16_f16acc);
+        const vk_fa_tuning_params params = get_fa_tuning_params_coopmat1(device, hsk, hsv, n_rows, n_kv, kv_type, f32acc);
+        bool shmem_ok = ggml_vk_flash_attn_coopmat_shmem_support(device, params, hsk, hsv, f32acc);
+
+        if (!shape_ok || !shmem_ok) {
+            path = FA_SCALAR;
         }
     }
 
-    // small rows, large cols
-    if (small_rows) {
-        return {get_fa_num_small_rows(FA_COOPMAT2), 32};
+    // scalar is faster than coopmat when N==1
+    if (n_rows == 1 && (path == FA_COOPMAT1 || path == FA_COOPMAT2)) {
+        path = FA_SCALAR;
     }
 
-    // small cols to reduce register count
-    if (ggml_is_quantized(type) || hsk >= 256 || hsv >= 256) {
-        if (hsk >= 512 || hsv >= 512) {
-            return {32, 32};
-        } else {
-            return {64, 32};
-        }
+    switch (path) {
+    case FA_SCALAR:
+        return get_fa_tuning_params_scalar(device, hsk, hsv, n_rows, n_kv, kv_type, f32acc);
+    case FA_COOPMAT1:
+        return get_fa_tuning_params_coopmat1(device, hsk, hsv, n_rows, n_kv, kv_type, f32acc);
+    case FA_COOPMAT2:
+        return get_fa_tuning_params_coopmat2(device, hsk, hsv, n_rows, n_kv, kv_type, f32acc);
+    default:
+        throw std::runtime_error("unsupported FaCodePath");
     }
-    return {64, 64};
 }
 
-static uint32_t fa_align(FaCodePath path, uint32_t hsk, uint32_t hsv, ggml_type type, bool small_rows, bool small_cache) {
-    return fa_rows_cols(path, hsk, hsv, 0, type, small_rows, small_cache)[1];
+static vk_fa_pipeline_state get_fa_pipeline_state(const vk_fa_tuning_params& params, uint32_t hsk, uint32_t hsv, bool aligned, bool f32acc,
+                                                  bool use_mask, bool use_mask_opt, bool use_logit_softcap) {
+    uint32_t flags = (use_mask_opt      ? 1 : 0) |
+                     (use_mask          ? 2 : 0) |
+                     (use_logit_softcap ? 4 : 0);
+
+    const uint32_t subgroup_size = params.disable_subgroups ? 0 : params.subgroup_size;
+
+    return vk_fa_pipeline_state{hsk, hsv, params.block_rows, params.block_cols, params.d_split, params.row_split, params.shmem_staging, params.path, params.workgroup_size, subgroup_size, aligned, f32acc, flags, params.limit_occupancy_shmem};
+}
+
+static std::vector<uint32_t> get_fa_spec_constants(const vk_fa_pipeline_state& state) {
+    return {state.workgroup_size, state.Br, state.Bc, state.HSK, state.HSV, !state.aligned, state.D_split,
+            state.row_split, state.subgroup_size, state.shmem_staging ? 1u : 0u, state.flags, state.limit_occupancy_shmem};
 }
 
 static bool ggml_vk_matmul_shmem_support(const vk_device& device, const std::vector<uint32_t>& warptile, bool mul_mat_id, ggml_type src0_type) {
@@ -3193,76 +3358,43 @@ static void ggml_vk_load_shaders(vk_device& device) {
                                        align, disable_robustness, require_full_subgroups, required_subgroup_size);
     };
 
-    auto const &fa_wg_denoms = [&](FaCodePath path, uint32_t hsk, uint32_t hsv, uint32_t clamp, ggml_type type, bool small_rows, bool small_cache) -> std::array<uint32_t, 3> {
-        return {fa_rows_cols(path, hsk, hsv, clamp, type, small_rows, small_cache)[0], 1, 1};
-    };
-
-    auto const &fa_spec_constants = [&](FaCodePath path, uint32_t hsk, uint32_t hsv, uint32_t clamp, ggml_type type, bool small_rows, bool small_cache, uint32_t flags) -> std::vector<uint32_t> {
-        // For large number of rows, 128 invocations seems to work best.
-        // For small number of rows (e.g. N==1), 256 works better. But matrix granularity for 256 is 32, so we
-        // can't use 256 for D==80.
-        // For scalar, use 128 (arbitrary)
-        // The same D_split value is used for both HSK and HSV, so just base it on the union of the LSBs.
-        const uint32_t D = (hsk|hsv);
-        auto rows_cols = fa_rows_cols(path, hsk, hsv, clamp, type, small_rows, small_cache);
-
-        uint32_t wg_size;
-        switch (path) {
-        case FA_COOPMAT2:
-            wg_size = ((small_rows && (D % 32) == 0) ? 256 : 128);
-            break;
-        case FA_COOPMAT1:
-            wg_size = (rows_cols[1] / 16) * device->subgroup_size; // enough subgroups for Bc/MatBc
-            break;
-        default:
-            wg_size = scalar_flash_attention_workgroup_size;
-            break;
-        }
-
-        // D_split can't be larger than a subgroup because we use subgroupShuffle to reduce it.
-        // D_split can't be larger than the LSB of D divided by 4 due to vectorization in the shader.
-        const uint32_t D_lsb = D ^ (D & (D-1));
-        uint32_t D_split = std::min(std::min(device->subgroup_size, 8u), D_lsb / 4);
-
-        // Nvidia prefers shared memory use to load large tiles of K.
-        // Switch to loading from global memory when it would use too much shared memory.
-        // AMD prefers loading K directly from global memory
-        const uint32_t k_load_shmem = device->vendor_id == VK_VENDOR_ID_NVIDIA && hsk < 256 ? 1 : 0;
-
-        return {wg_size, rows_cols[0], rows_cols[1], hsk, hsv, clamp, D_split, device->subgroup_size, k_load_shmem, flags};
-    };
-
 #define CREATE_FA(TYPE, NAMELC, FAPATH, SUFFIX) \
         for (auto &fa : device->pipeline_flash_attn_f32_f16[TYPE]) { \
-            uint32_t HSK = fa.first.HSK; \
-            uint32_t HSV = fa.first.HSV; \
-            bool small_rows = fa.first.small_rows; \
-            bool small_cache = fa.first.small_cache; \
             FaCodePath path = fa.first.path; \
+            uint32_t Br = fa.first.Br; \
+            uint32_t Bc = fa.first.Bc; \
             bool aligned = fa.first.aligned; \
             bool f32acc = fa.first.f32acc; \
-            uint32_t flags = fa.first.flags; \
+            uint32_t fa_sgs = fa.first.subgroup_size; \
+            bool fa_ds = fa.first.subgroup_size == 0; \
             if (path == FAPATH) { \
                 if (aligned) { \
                     if (f32acc) { \
-                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_aligned_f32acc" #NAMELC, flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), fa_wg_denoms(FAPATH, HSK,HSV,0,TYPE,small_rows,small_cache), fa_spec_constants(FAPATH, HSK,HSV,0,TYPE,small_rows,small_cache,flags), fa_align(FAPATH,HSK,HSV,TYPE,small_rows,small_cache), true, FAPATH==FA_COOPMAT1, (FAPATH==FA_COOPMAT1 ? device->subgroup_size : 0));     \
+                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_aligned_f32acc" #NAMELC, flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), {Br, 1, 1}, get_fa_spec_constants(fa.first), Bc, true, (!fa_ds && (FAPATH!=FA_COOPMAT2)), ((!fa_ds && (FAPATH!=FA_COOPMAT2)) ? fa_sgs : 0));     \
                     } else { \
-                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_aligned_f16acc" #NAMELC, flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), fa_wg_denoms(FAPATH, HSK,HSV,0,TYPE,small_rows,small_cache), fa_spec_constants(FAPATH, HSK,HSV,0,TYPE,small_rows,small_cache,flags), fa_align(FAPATH,HSK,HSV,TYPE,small_rows,small_cache), true, FAPATH==FA_COOPMAT1, (FAPATH==FA_COOPMAT1 ? device->subgroup_size : 0));     \
+                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_aligned_f16acc" #NAMELC, flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), {Br, 1, 1}, get_fa_spec_constants(fa.first), Bc, true, (!fa_ds && (FAPATH!=FA_COOPMAT2)), ((!fa_ds && (FAPATH!=FA_COOPMAT2)) ? fa_sgs : 0));     \
                     } \
                 } else { \
                     if (f32acc) { \
-                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_f32acc"         #NAMELC, flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), fa_wg_denoms(FAPATH, HSK,HSV,1,TYPE,small_rows,small_cache), fa_spec_constants(FAPATH, HSK,HSV,1,TYPE,small_rows,small_cache,flags), 1,                                        true, FAPATH==FA_COOPMAT1, (FAPATH==FA_COOPMAT1 ? device->subgroup_size : 0));     \
+                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_f32acc"         #NAMELC, flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ##            SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), {Br, 1, 1}, get_fa_spec_constants(fa.first), 1,  true, (!fa_ds && (FAPATH!=FA_COOPMAT2)), ((!fa_ds && (FAPATH!=FA_COOPMAT2)) ? fa_sgs : 0));     \
                     } else { \
-                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_f16acc"         #NAMELC, flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), fa_wg_denoms(FAPATH, HSK,HSV,1,TYPE,small_rows,small_cache), fa_spec_constants(FAPATH, HSK,HSV,1,TYPE,small_rows,small_cache,flags), 1,                                        true, FAPATH==FA_COOPMAT1, (FAPATH==FA_COOPMAT1 ? device->subgroup_size : 0));     \
+                        ggml_vk_create_pipeline(device, fa.second, "flash_attn_f32_f16_f16acc"         #NAMELC, flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _len,  flash_attn_f32_f16_ ## NAMELC ## _f16acc ## SUFFIX ## _data,  "main", 7, sizeof(vk_flash_attn_push_constants), {Br, 1, 1}, get_fa_spec_constants(fa.first), 1,  true, (!fa_ds && (FAPATH!=FA_COOPMAT2)), ((!fa_ds && (FAPATH!=FA_COOPMAT2)) ? fa_sgs : 0));     \
                     } \
                 } \
             } \
         }
 
-    CREATE_FA(GGML_TYPE_F32, f32, FA_SCALAR, )
-    CREATE_FA(GGML_TYPE_F16, f16, FA_SCALAR, )
-    CREATE_FA(GGML_TYPE_Q4_0, q4_0, FA_SCALAR, )
-    CREATE_FA(GGML_TYPE_Q8_0, q8_0, FA_SCALAR, )
+    if (device->flash_attention_fp16) {
+        CREATE_FA(GGML_TYPE_F32, f32, FA_SCALAR, )
+        CREATE_FA(GGML_TYPE_F16, f16, FA_SCALAR, )
+        CREATE_FA(GGML_TYPE_Q4_0, q4_0, FA_SCALAR, )
+        CREATE_FA(GGML_TYPE_Q8_0, q8_0, FA_SCALAR, )
+    } else {
+        CREATE_FA(GGML_TYPE_F32, f32, FA_SCALAR, _fp32)
+        CREATE_FA(GGML_TYPE_F16, f16, FA_SCALAR, _fp32)
+        CREATE_FA(GGML_TYPE_Q4_0, q4_0, FA_SCALAR, _fp32)
+        CREATE_FA(GGML_TYPE_Q8_0, q8_0, FA_SCALAR, _fp32)
+    }
 #if defined(VK_KHR_cooperative_matrix) && defined(GGML_VULKAN_COOPMAT_GLSLC_SUPPORT)
     if (device->coopmat1_fa_support) {
         CREATE_FA(GGML_TYPE_F32, f32, FA_COOPMAT1, _cm1)
@@ -3780,10 +3912,12 @@ static void ggml_vk_load_shaders(vk_device& device) {
         && !device->coopmat_bf16_support
 #endif
         ) {
+        const uint32_t s_warptile_wm = device->subgroup_size == 8 ? 8 : 32;
+
         // use scalar tile sizes
         l_warptile = { 128, 128, 128, 16, subgroup_size_8 * 2, 64, 2, 4, 4, 1, subgroup_size_8 };
         m_warptile = { 128,  64,  64, 16, subgroup_size_8, 32, 2, 4, 2, 1, subgroup_size_8 };
-        s_warptile = { subgroup_size_16, 32, 32, 16, 32, 32, 2, 2, 2, 1, subgroup_size_8 };
+        s_warptile = { subgroup_size_32, 32, 32, 16, s_warptile_wm, 32, 2, 2, 2, 1, subgroup_size_8 };
 
         l_wg_denoms = {128, 128, 1 };
         m_wg_denoms = { 64,  64, 1 };
@@ -4533,6 +4667,7 @@ static void ggml_vk_load_shaders(vk_device& device) {
 }
 
 static bool ggml_vk_khr_cooperative_matrix_support(const vk::PhysicalDeviceProperties& props, const vk::PhysicalDeviceDriverProperties& driver_props, vk_device_architecture arch);
+static uint32_t ggml_vk_intel_shader_core_count(const vk::PhysicalDevice& vkdev);
 
 static vk_device ggml_vk_get_device(size_t idx) {
     VK_LOG_DEBUG("ggml_vk_get_device(" << idx << ")");
@@ -4749,6 +4884,8 @@ static vk_device ggml_vk_get_device(size_t idx) {
             device->shader_core_count = sm_props.shaderSMCount;
         } else if (amd_shader_core_properties2) {
             device->shader_core_count = amd_shader_core_properties2_props.activeComputeUnitCount;
+        } else if (device->vendor_id == VK_VENDOR_ID_INTEL) {
+            device->shader_core_count = ggml_vk_intel_shader_core_count(device->physical_device);
         } else {
             device->shader_core_count = 0;
         }
@@ -4968,11 +5105,7 @@ static vk_device ggml_vk_get_device(size_t idx) {
 
 #if defined(VK_KHR_cooperative_matrix)
         device->coopmat_support = device->coopmat_support && coopmat_features.cooperativeMatrix;
-
-        // coopmat1 fa shader currently assumes 32 invocations per subgroup
-        device->coopmat1_fa_support = device->coopmat_support && device->subgroup_require_full_support &&
-                                      device->subgroup_size_control && device->subgroup_min_size <= 32 &&
-                                      device->subgroup_max_size >= 32;
+        device->coopmat1_fa_support = device->coopmat_support && device->subgroup_require_full_support;
 #endif
 
         if (coopmat2_support) {
@@ -5290,6 +5423,10 @@ static vk_device ggml_vk_get_device(size_t idx) {
             device->mmvq_mode = 1;
         }
 
+        // Driver issues with older AMD GPUs on Windows, see https://github.com/ggml-org/llama.cpp/pull/19625#issuecomment-3940840613
+        const bool is_amd_proprietary_gcn = device->vendor_id == VK_VENDOR_ID_AMD && device->architecture == AMD_GCN && device->driver_id == vk::DriverId::eAmdProprietary;
+        device->flash_attention_fp16 = device->fp16 && !is_amd_proprietary_gcn;
+
         return device;
     }
 
@@ -5540,6 +5677,10 @@ static void ggml_vk_instance_init() {
     vk_perf_logger_concurrent = getenv("GGML_VK_PERF_LOGGER_CONCURRENT") != nullptr;
     vk_enable_sync_logger = getenv("GGML_VK_SYNC_LOGGER") != nullptr;
     vk_memory_logger_enabled = getenv("GGML_VK_MEMORY_LOGGER") != nullptr;
+    const char* GGML_VK_PIPELINE_STATS = getenv("GGML_VK_PIPELINE_STATS");
+    if (GGML_VK_PIPELINE_STATS != nullptr) {
+        vk_pipeline_stats_filter = GGML_VK_PIPELINE_STATS;
+    }
     const char* GGML_VK_PERF_LOGGER_FREQUENCY = getenv("GGML_VK_PERF_LOGGER_FREQUENCY");
 
     if (GGML_VK_PERF_LOGGER_FREQUENCY != nullptr) {
@@ -8419,21 +8560,27 @@ static void ggml_vk_mul_mat_id(ggml_backend_vk_context * ctx, vk_context& subctx
     }
 }
 
-static bool ggml_vk_flash_attn_scalar_shmem_support(const vk_device& device, const uint32_t hsk, uint32_t hsv, bool small_cache) {
+static bool ggml_vk_flash_attn_scalar_shmem_support(const vk_device& device, const vk_fa_tuning_params& params, uint32_t hsk, uint32_t hsv, bool f32acc) {
+    GGML_UNUSED(f32acc);
     // Needs to be kept up to date on shader changes
-    GGML_UNUSED(hsv);
-    const uint32_t wg_size = scalar_flash_attention_workgroup_size;
-    const uint32_t Br = get_fa_scalar_num_large_rows(hsk, hsv, small_cache);
-    const uint32_t Bc = scalar_flash_attention_Bc;
+    const uint32_t wg_size = params.workgroup_size;
+    const uint32_t Br = params.block_rows;
+    const uint32_t Bc = params.block_cols;
 
+    const uint32_t float_type_size = device->flash_attention_fp16 ? sizeof(ggml_fp16_t) : sizeof(float);
+
+    // tmpsh is overestimated slightly
     const uint32_t tmpsh = wg_size * sizeof(float);
-    const uint32_t tmpshv4 = wg_size * 4 * sizeof(float);
+    const uint32_t tmpshv4 = wg_size * 4 * float_type_size;
 
-    const uint32_t masksh = Bc * Br * sizeof(float);
+    const uint32_t masksh = Bc * (Br + 1) * float_type_size;
 
-    const uint32_t Qf = Br * (hsk / 4 + 2) * 4 * sizeof(float);
+    const uint32_t Qf = Br * (hsk / 4 + 1) * 4 * float_type_size;
 
-    const uint32_t total_size = tmpsh + tmpshv4 + masksh + Qf;
+    const uint32_t D = std::max(hsk, hsv);
+    const uint32_t kvsh = params.shmem_staging ? Bc * (D / 4 + 1) * 4 * float_type_size : 4 * float_type_size;
+
+    const uint32_t total_size = tmpsh + tmpshv4 + masksh + Qf + kvsh;
     const bool supported = total_size <= device->properties.limits.maxComputeSharedMemorySize;
 
     VK_LOG_DEBUG("ggml_vk_flash_attn_scalar_shmem_support(HSK=" << hsk << ", HSV=" << hsv << ", total_size=" << total_size << ", supported=" << supported);
@@ -8441,18 +8588,17 @@ static bool ggml_vk_flash_attn_scalar_shmem_support(const vk_device& device, con
     return supported;
 }
 
-static bool ggml_vk_flash_attn_coopmat_shmem_support(const vk_device& device, const uint32_t hsk, uint32_t hsv, bool f32acc, ggml_type kv_type) {
+static bool ggml_vk_flash_attn_coopmat_shmem_support(const vk_device& device, const vk_fa_tuning_params& params, uint32_t hsk, uint32_t hsv, bool f32acc) {
     // Needs to be kept up to date on shader changes
-    GGML_UNUSED(hsv);
-    const auto rows_cols = fa_rows_cols(FA_COOPMAT1, hsk, hsv, 0, kv_type, false, false);
-    const uint32_t Br = rows_cols[0];
-    const uint32_t Bc = rows_cols[1];
+    const uint32_t Br = params.block_rows;
+    const uint32_t Bc = params.block_cols;
 
     const uint32_t MatBr = 16, MatBc = 16;
 
     const uint32_t row_split = Bc / MatBc;
 
     const uint32_t hsk_pad = ROUNDUP_POW2(hsk, 16);
+    const uint32_t hsv_pad = ROUNDUP_POW2(hsv, 16);
 
     const uint32_t acctype = f32acc ? 4 : 2;
     const uint32_t f16vec4 = 8;
@@ -8468,17 +8614,19 @@ static bool ggml_vk_flash_attn_coopmat_shmem_support(const vk_device& device, co
     const uint32_t sfshstride = (hsk <= 128) ? (Br + 8) : Br;
     const uint32_t sfsh = Bc * sfshstride * acctype;
 
-    const bool k_load_shmem = device->vendor_id == VK_VENDOR_ID_NVIDIA && hsk < 256;
-    const uint32_t kshstride = (k_load_shmem ? hsk_pad : MatBr) / 4 + 2;
+    const uint32_t kvshstride = (params.shmem_staging ? std::max(hsk_pad, hsv_pad) : MatBr) / 4 + 2;
     const uint32_t vsh_stride = MatBc / 4 * row_split;
-    const uint32_t ksh = ((kshstride >= vsh_stride) ? (Bc * kshstride) : (Bc * vsh_stride)) * f16vec4;
+    const uint32_t ksh = ((kvshstride >= vsh_stride) ? (Bc * kvshstride) : (Bc * vsh_stride)) * f16vec4;
+
+    const uint32_t osh_stride = params.row_split * MatBr / 4;
+    const uint32_t pvsh = MatBc * osh_stride * f16vec4;
 
     const uint32_t slope = Br * acctype;
 
-    const uint32_t total_size = tmpsh + Qf + Psh + sfsh + ksh + slope;
+    const uint32_t total_size = tmpsh + Qf + Psh + sfsh + ksh + pvsh + slope;
     const bool supported = total_size <= device->properties.limits.maxComputeSharedMemorySize;
 
-    VK_LOG_DEBUG("ggml_vk_flash_attn_coopmat_shmem_support(HSK=" << hsk << ", HSV=" << hsv << ", f32acc=" << f32acc << ", kv_type=" << kv_type << ", total_size=" << total_size << ", supported=" << supported);
+    VK_LOG_DEBUG("ggml_vk_flash_attn_coopmat_shmem_support(HSK=" << hsk << ", HSV=" << hsv << ", f32acc=" << f32acc << ", total_size=" << total_size << ", supported=" << supported);
 
     return supported;
 }
@@ -8536,48 +8684,18 @@ static void ggml_vk_flash_attn(ggml_backend_vk_context * ctx, vk_context& subctx
     assert(q->type == GGML_TYPE_F32);
     assert(k->type == v->type);
 
-    FaCodePath path = ctx->device->coopmat2 ? FA_COOPMAT2 :
-                      ctx->device->coopmat1_fa_support ? FA_COOPMAT1 : FA_SCALAR;
-
-    if (path == FA_COOPMAT1 && ctx->device->architecture == vk_device_architecture::NVIDIA_TURING) {
-        // Nvidia compiler bug, see https://github.com/ggml-org/llama.cpp/pull/19075#issuecomment-3820716090
-        path = FA_SCALAR;
-    }
-
-    if (path == FA_COOPMAT1) {
-        const bool coopmat_shape_supported = (dst->op_params[3] == GGML_PREC_F32 && ctx->device->coopmat_support_16x16x16_f32acc) ||
-                                             (dst->op_params[3] != GGML_PREC_F32 && ctx->device->coopmat_support_16x16x16_f16acc);
-
-        const bool coopmat_shmem_supported = ggml_vk_flash_attn_coopmat_shmem_support(ctx->device, HSK, HSV, dst->op_params[3] == GGML_PREC_F32, k->type);
-
-        if (!coopmat_shape_supported || !coopmat_shmem_supported) {
-            path = FA_SCALAR;
-        }
-    }
-
     uint32_t gqa_ratio = 1;
     uint32_t qk_ratio = neq2 / nek2;
     uint32_t workgroups_x = (uint32_t)neq1;
     uint32_t workgroups_y = (uint32_t)neq2;
     uint32_t workgroups_z = (uint32_t)neq3;
 
-    const bool small_cache = nek1 < 1024;
+    const bool f32acc = !ctx->device->flash_attention_fp16 || dst->op_params[3] == GGML_PREC_F32;
 
     // For scalar/coopmat1 FA, we can use the "large" size to accommodate qga.
     // For coopmat2 FA, we always use the small size (which is still pretty large for gqa).
-    uint32_t max_gqa;
-    switch (path) {
-    case FA_SCALAR:
-    case FA_COOPMAT1:
-        // We may switch from coopmat1 to scalar, so use the scalar limit for both
-        max_gqa = get_fa_scalar_num_large_rows(HSK, HSV, small_cache);
-        break;
-    case FA_COOPMAT2:
-        max_gqa = get_fa_num_small_rows(FA_COOPMAT2);
-        break;
-    default:
-        GGML_ASSERT(0);
-    }
+    vk_fa_tuning_params tuning_params = get_fa_tuning_params(ctx->device, HSK, HSV, 512, KV, k->type, f32acc);
+    const uint32_t max_gqa = std::min(tuning_params.block_rows, 32u);
 
     if (N <= 8 && qk_ratio > 1 && qk_ratio <= max_gqa &&
         qk_ratio * nek2 == neq2 && nek2 == nev2 && nem2 <= 1) {
@@ -8589,24 +8707,7 @@ static void ggml_vk_flash_attn(ggml_backend_vk_context * ctx, vk_context& subctx
         workgroups_y /= gqa_ratio;
     }
 
-    bool small_rows = N <= get_fa_num_small_rows(path);
-
-    // coopmat1 does not actually support "small rows" (it needs 16 rows).
-    // So use scalar instead.
-    if (small_rows && path == FA_COOPMAT1) {
-        path = FA_SCALAR;
-    }
-
-    // scalar is faster than coopmat2 when N==1
-    if (N == 1 && path == FA_COOPMAT2) {
-        path = FA_SCALAR;
-    }
-
-    // with large hsk/hsv, scalar path may need to use small_rows to fit in shared memory
-    if (path == FA_SCALAR &&
-        !ggml_vk_flash_attn_scalar_shmem_support(ctx->device, HSK, HSV, small_cache)) {
-        small_rows = true;
-    }
+    tuning_params = get_fa_tuning_params(ctx->device, HSK, HSV, N, KV, k->type, f32acc);
 
     const uint32_t q_stride = (uint32_t)(nbq1 / ggml_type_size(q->type));
     uint32_t k_stride = (uint32_t)(nbk1 / ggml_type_size(k->type));
@@ -8620,18 +8721,16 @@ static void ggml_vk_flash_attn(ggml_backend_vk_context * ctx, vk_context& subctx
         v_stride /= 4;
     }
 
-    uint32_t alignment = fa_align(path, HSK, HSV, k->type, small_rows, small_cache);
+    const uint32_t alignment = tuning_params.block_cols;
     bool aligned = (KV % alignment) == 0 &&
                    // the "aligned" shader variant will forcibly align strides, for performance
                    (q_stride & 7) == 0 && (k_stride & 7) == 0 && (v_stride & 7) == 0;
 
     // Need to use the coopmat2 variant that clamps loads when HSK/HSV aren't sufficiently aligned.
-    if (((HSK | HSV) % 16) != 0 && path == FA_COOPMAT2) {
+    if (((HSK | HSV) % 16) != 0 && tuning_params.path == FA_COOPMAT2) {
         aligned = false;
     }
 
-    bool f32acc = path == FA_SCALAR || dst->op_params[3] == GGML_PREC_F32;
-
     float scale         = 1.0f;
     float max_bias      = 0.0f;
     float logit_softcap = 0.0f;
@@ -8646,12 +8745,8 @@ static void ggml_vk_flash_attn(ggml_backend_vk_context * ctx, vk_context& subctx
 
     // Only use mask opt when the mask is fairly large. This hasn't been tuned extensively.
     bool use_mask_opt = mask && nem1 >= 32 && nem0 * nem1 > 32768;
-
-    uint32_t flags = (use_mask_opt       ? 1 : 0) |
-                     (mask != nullptr    ? 2 : 0) |
-                     (logit_softcap != 0 ? 4 : 0);
-
-    vk_fa_pipeline_state fa_pipeline_state(HSK, HSV, small_rows, small_cache, path, aligned, f32acc, flags);
+    vk_fa_pipeline_state fa_pipeline_state = get_fa_pipeline_state(tuning_params, HSK, HSV, aligned, f32acc,
+                                                                   mask != nullptr, use_mask_opt, logit_softcap != 0);
 
     vk_pipeline pipeline = nullptr;
 
@@ -8673,22 +8768,35 @@ static void ggml_vk_flash_attn(ggml_backend_vk_context * ctx, vk_context& subctx
     uint32_t split_kv = KV;
     uint32_t split_k = 1;
 
+    // Intel Alchemist prefers more workgroups
+    const uint32_t shader_core_count_multiplier = (ctx->device->vendor_id == VK_VENDOR_ID_INTEL && ctx->device->architecture != INTEL_XE2) ? 2 : 1;
+
     // Use a placeholder core count if one isn't available. split_k is a big help for perf.
-    const uint32_t shader_core_count = ctx->device->shader_core_count ? ctx->device->shader_core_count : 16;
+    const uint32_t shader_core_count = ctx->device->shader_core_count ? ctx->device->shader_core_count * shader_core_count_multiplier : 16;
+
+    const uint32_t Br = fa_pipeline_state.Br;
+    const uint32_t Bc = fa_pipeline_state.Bc;
+
+    GGML_ASSERT(Br == pipeline->wg_denoms[0]);
+    const uint32_t Tr = CEIL_DIV(N, Br);
 
     // Try to use split_k when KV is large enough to be worth the overhead.
-    // Must either be a single batch or be using gqa, we can't mix the two.
-    if (workgroups_x <= pipeline->wg_denoms[0] && (workgroups_x == 1 || gqa_ratio > 1)) {
-        // Try to run two workgroups per SM.
+    if (gqa_ratio > 1 && workgroups_x <= Br) {
         split_k = shader_core_count * 2 / (workgroups_x * workgroups_y * workgroups_z);
-        if (split_k > 1) {
-            // Try to evenly split KV into split_k chunks, but it needs to be a multiple
-            // of "align", so recompute split_k based on that.
-            split_kv = ROUNDUP_POW2(std::max(1u, KV / split_k), alignment);
-            split_k = CEIL_DIV(KV, split_kv);
+    } else if (gqa_ratio <= 1) {
+        uint32_t total_wgs_no_split = Tr * workgroups_y * workgroups_z;
+        if (total_wgs_no_split < shader_core_count * 2) {
+            split_k = shader_core_count * 2 / total_wgs_no_split;
         }
     }
 
+    if (split_k > 1) {
+        // Try to evenly split KV into split_k chunks, but it needs to be a multiple
+        // of "align", so recompute split_k based on that.
+        split_kv = ROUNDUP_POW2(std::max(1u, KV / split_k), alignment);
+        split_k = CEIL_DIV(KV, split_kv);
+    }
+
     // Reserve space for split_k temporaries. For each split x batch, we need to store the O matrix (D x ne1)
     // and the per-row m and L values (ne1 rows). We store all the matrices first, followed by the rows.
     // For matrices, the order is (inner to outer) [HSV, ne1, k, ne2, ne3].
@@ -8702,10 +8810,6 @@ static void ggml_vk_flash_attn(ggml_backend_vk_context * ctx, vk_context& subctx
         ggml_vk_preallocate_buffers(ctx, subctx);
     }
 
-    auto rows_cols = fa_rows_cols(path, HSK, HSV, !aligned, k->type, small_rows, small_cache);
-    const uint32_t Br = rows_cols[0];
-    const uint32_t Bc = rows_cols[1];
-
     const uint32_t mask_opt_num_dwords = CEIL_DIV(nem0, 16 * Bc);
     const uint64_t mask_opt_size = sizeof(uint32_t) * mask_opt_num_dwords * CEIL_DIV(nem1, Br) * nem2 * nem3;
 
@@ -8785,15 +8889,21 @@ static void ggml_vk_flash_attn(ggml_backend_vk_context * ctx, vk_context& subctx
         if (ctx->prealloc_split_k_need_sync) {
             ggml_vk_sync_buffers(ctx, subctx);
         }
-        workgroups_x *= pipeline->wg_denoms[0];
+
+        // We reuse workgroups_x to mean the number of splits, so we need to
+        // cancel out the divide by wg_denoms[0].
+        uint32_t dispatch_x;
+        if (gqa_ratio > 1) {
+            workgroups_x *= pipeline->wg_denoms[0];
+            dispatch_x = split_k * workgroups_x;
+        } else {
+            dispatch_x = Tr * split_k * pipeline->wg_denoms[0];
+        }
+
         vk_subbuffer split_k_buf = ggml_vk_subbuffer(ctx, ctx->prealloc_split_k, 0);
         ggml_vk_dispatch_pipeline(ctx, subctx, pipeline,
                                     {q_buf, k_buf, v_buf, mask_buf, sinks_buf, split_k_buf, mask_opt_buf},
-                                    // We only use split_k when group query attention is enabled, which means
-                                    // there's no more than one tile of rows (i.e. workgroups_x would have been
-                                    // one). We reuse workgroups_x to mean the number of splits, so we need to
-                                    // cancel out the divide by wg_denoms[0].
-                                    pc, { split_k * workgroups_x, workgroups_y, workgroups_z });
+                                    pc, { dispatch_x, workgroups_y, workgroups_z });
 
         ggml_vk_sync_buffers(ctx, subctx);
         const vk_op_flash_attn_split_k_reduce_push_constants pc2 = { HSV, (uint32_t)ne1, (uint32_t)ne2, (uint32_t)ne3, split_k, (sinks != nullptr) };
@@ -15418,6 +15528,46 @@ static bool ggml_vk_khr_cooperative_matrix_support(const vk::PhysicalDevicePrope
     }
 }
 
+static uint32_t ggml_vk_intel_shader_core_count(const vk::PhysicalDevice& vkdev) {
+    VkPhysicalDeviceProperties2 props = vkdev.getProperties2();
+
+    if (props.properties.vendorID != VK_VENDOR_ID_INTEL) {
+        return 0;
+    }
+
+    const uint32_t device_id = props.properties.deviceID;
+
+    switch (device_id) {
+    case 0x56A6:  // A310
+        return 6;
+    case 0x5693:  // A370M
+    case 0x56A5:  // A380
+    case 0x56B1:  // Pro A40/A50
+        return 8;
+    case 0x5697:  // A530M
+        return 12;
+    case 0x5692:  // A550M
+    case 0x56B3:  // Pro A60
+        return 16;
+    case 0x56A2:  // A580
+        return 24;
+    case 0x5691:  // A730M
+    case 0x56A1:  // A750
+        return 28;
+    case 0x56A0:  // A770
+    case 0x5690:  // A770M
+        return 32;
+    case 0xE212:  // Pro B50
+        return 16;
+    case 0xE20C:  // B570
+        return 18;
+    case 0xE20B:  // B580
+        return 20;
+    default:
+        return 0;
+    }
+}
+
 // checks
 
 #ifdef GGML_VULKAN_CHECK_RESULTS
@@ -16094,7 +16244,7 @@ static void ggml_vk_check_results_1(ggml_backend_vk_context * ctx, ggml_cgraph *
         ggml_vk_print_graph_origin(tensor, done);
     }
 
-    if (avg_err > 0.5 || std::isnan(avg_err)) {
+    if (avg_err > 0.01 || std::isnan(avg_err)) {
         std::cerr << "ERROR: avg_err=" << avg_err << " in " << ggml_op_name(tensor->op) << " (check " << check_counter << ")" << std::endl;
         std::cerr << "tensor=" << tensor << " tensor->name=" << tensor->name << " tensor->type: " << ggml_type_name(tensor->type) << " ne0=" << tensor->ne[0] << " nb0=" << tensor->nb[0] << " ne1=" << tensor->ne[1] << " nb1=" << tensor->nb[1] << " ne2=" << tensor->ne[2] << " nb2=" << tensor->nb[2] << " ne3=" << tensor->ne[3] << " nb3=" << tensor->nb[3] << " offset=" << tensor->view_offs << std::endl;
         if (src0 != nullptr) {

ggml/src/ggml-vulkan/vulkan-shaders/flash_attn.comp

@@ -3,9 +3,13 @@
 #extension GL_EXT_control_flow_attributes : enable
 #extension GL_EXT_shader_16bit_storage : require
 
-#extension GL_EXT_shader_explicit_arithmetic_types_float16 : require
 #extension GL_EXT_shader_explicit_arithmetic_types_int32 : require
 
+#ifdef FLOAT16
+#extension GL_EXT_shader_explicit_arithmetic_types_float16 : require
+#extension GL_EXT_shader_subgroup_extended_types_float16 : require
+#endif
+
 #extension GL_KHR_shader_subgroup_shuffle : enable
 #extension GL_KHR_shader_subgroup_vote : enable
 
@@ -15,8 +19,10 @@
 const uint32_t HSK_per_thread = HSK / D_split;
 const uint32_t HSV_per_thread = HSV / D_split;
 
-const uint32_t cols_per_iter = WorkGroupSize / D_split;
+const uint32_t rows_per_thread = Br / row_split;
+const uint32_t cols_per_iter = WorkGroupSize / D_split / row_split;
 const uint32_t cols_per_thread = Bc / cols_per_iter;
+const uint32_t num_subgroups = SubGroupSize == 0 ? 0 : WorkGroupSize / SubGroupSize;
 
 
 layout (binding = 0) readonly buffer Q {float data_q[];};
@@ -27,20 +33,22 @@ layout (binding = 2) readonly buffer V {float16_t data_v[];};
 layout (binding = 2) readonly buffer VV4 {f16vec4 data_vv4[];};
 layout (binding = 3) readonly buffer M {float16_t data_m[];};
 
-// Store the output when doing grouped query attention.
-// Rows index by Q's dimension 2, and the first N rows are valid.
-D_TYPE perElemOpGqaStore(const in uint32_t r, const in uint32_t c, const in D_TYPE elem, const in uint32_t o_offset, const in uint32_t iq2, const in uint32_t N)
-{
-    uint32_t offset = (iq2 + r) * HSV + c;
-    data_o[o_offset + offset] = D_TYPE(elem);
-    return elem;
-}
+// If SubGroupSize is set to 0 then only use shmem reductions
+const uint32_t tmpsh_size = (SubGroupSize > 0) ? (row_split == 1 ? num_subgroups * D_split : num_subgroups) : WorkGroupSize;
+shared float tmpsh[tmpsh_size];
+shared FLOAT_TYPEV4 tmpshv4[tmpsh_size];
 
-shared FLOAT_TYPE tmpsh[WorkGroupSize];
-shared vec4 tmpshv4[WorkGroupSize];
+const uint32_t masksh_stride = Br + 1;
+shared FLOAT_TYPE masksh[Bc * masksh_stride];
 
-shared float masksh[Bc][Br];
-shared vec4 Qf[Br][HSK / 4];
+const uint32_t qf_stride = HSK / 4 + 1;
+shared FLOAT_TYPEV4 Qf[Br * qf_stride];
+
+const uint32_t D = HSK > HSV ? HSK : HSV;
+const uint32_t kvsh_stride = D / 4 + 1;
+shared FLOAT_TYPEV4 kvsh[SHMEM_STAGING != 0 ? Bc * kvsh_stride : 1];
+
+shared vec4 occupancy_limiter[LIMIT_OCCUPANCY_SHMEM > 0 ? LIMIT_OCCUPANCY_SHMEM : 1];
 
 void main() {
 #ifdef NEEDS_INIT_IQ_SHMEM
@@ -50,8 +58,24 @@ void main() {
     init_indices();
 
     const uint32_t tid = gl_LocalInvocationIndex;
+    const uint32_t threads_per_rowgroup = gl_WorkGroupSize.x / row_split;
+    const uint32_t row_tid = gl_LocalInvocationIndex / threads_per_rowgroup;
+    const uint32_t rowgroup_tid = gl_LocalInvocationIndex % threads_per_rowgroup;
     const uint32_t d_tid = gl_LocalInvocationIndex % D_split;
-    const uint32_t col_tid = gl_LocalInvocationIndex / D_split;
+    const uint32_t col_tid = (gl_LocalInvocationIndex % threads_per_rowgroup) / D_split;
+
+    if (LIMIT_OCCUPANCY_SHMEM > 0) {
+        // This just exists to avoid the occupancy_limiter array getting optimized out
+        occupancy_limiter[tid] = vec4(tid);
+
+        barrier();
+
+        if (occupancy_limiter[tid] == vec4(99999.0)) {
+            data_ov4[0] = D_TYPEV4(occupancy_limiter[tid]);
+        }
+    }
+
+#define tile_row(r) (row_tid * rows_per_thread + (r))
 
     uint32_t q_offset = gqa_iq1*p.nb01 + (iq2*p.nb02 + iq3*p.nb03) / 4;
 
@@ -60,37 +84,37 @@ void main() {
         uint32_t r = (idx + tid) / (HSK / 4);
         if (r < Br && d < HSK / 4 &&
             i * Br + r < N) {
-            Qf[r][d] = vec4(data_qv4[q_offset / 4 + (i * Br + r) * q_stride / 4 + d]) * p.scale;
+            Qf[r * qf_stride + d] = FLOAT_TYPEV4(data_qv4[q_offset / 4 + (i * Br + r) * q_stride / 4 + d] * p.scale);
         }
     }
     barrier();
 
-    vec4 Of[Br][HSV_per_thread / 4];
+    FLOAT_TYPEV4 Of[rows_per_thread][HSV_per_thread / 4];
     [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            Of[r][d] = vec4(0.0);
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            Of[r][d] = FLOAT_TYPEV4(0.0);
         }
     }
 
-    float Lf[Br], Mf[Br];
+    float Lf[rows_per_thread], Mf[rows_per_thread];
 
     // Use -FLT_MAX/2 rather than -inf to reduce the possibility of NaNs, e.g. when computing Mold-M.
     const float NEG_FLT_MAX_OVER_2 = uintBitsToFloat(0xFEFFFFFF);
 
-    [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
+    [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
         Lf[r] = 0;
         Mf[r] = NEG_FLT_MAX_OVER_2;
     }
 
-    float slope[Br];
-    [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-        slope[r] = 1.0;
+    ACC_TYPE slope[rows_per_thread];
+    [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+        slope[r] = ACC_TYPE(1.0);
     }
 
     // ALiBi
     if (p.max_bias > 0.0f) {
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            slope[r] = perElemOpComputeSlope(r, col_tid, ACC_TYPE(0), iq2);
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            slope[r] = perElemOpComputeSlope(tile_row(r), col_tid, ACC_TYPE(0), iq2);
         }
     }
 
@@ -113,75 +137,141 @@ void main() {
 
     uint32_t mask_opt = 0;
     uint32_t mask_opt_idx = ~0;
+    uint32_t mask_opt_bits = 0;
 
     [[dont_unroll]]
     for (uint32_t j = start_j; j < end_j; ++j) {
+        if (MASK_ENABLE) {
+            if (USE_MASK_OPT && mask_opt_idx != j / 16) {
+                mask_opt_idx = j / 16;
+                mask_opt = data_mask_opt[mo_offset + mask_opt_idx];
+            }
+            mask_opt_bits = (mask_opt >> ((j % 16) * 2)) & 0x3;
+            if (mask_opt_bits == MASK_OPT_ALL_NEG_INF) {
+                // skip this block
+                continue;
+            }
+            // Only load if the block is not all zeros
+            if (mask_opt_bits != MASK_OPT_ALL_ZERO) {
+                bool nem1_bounds_check = !(p.gqa_ratio > 1) && (p.nem1 % Br) != 0;
 
-        if (USE_MASK_OPT && mask_opt_idx != j / 16) {
-            mask_opt_idx = j / 16;
-            mask_opt = data_mask_opt[mo_offset + mask_opt_idx];
+                float max_mask = NEG_FLT_MAX_OVER_2;
+                barrier();
+                [[unroll]] for (uint32_t idx = 0; idx < Bc * Br; idx += gl_WorkGroupSize.x) {
+                    uint32_t c = (idx + tid) % Bc;
+                    uint32_t r = (idx + tid) / Bc;
+                    if (idx + tid < Bc * Br) {
+                        if ((!KV_bounds_check || j * Bc + c < KV) && (!nem1_bounds_check || i * Br + r < p.nem1)) {
+                            FLOAT_TYPE m = FLOAT_TYPE(data_m[m_offset + (i * Br + r) * m_stride + (j * Bc + c)]);
+                            masksh[c * masksh_stride + r] = m;
+                            max_mask = max(max_mask, float(m));
+                        } else {
+                            masksh[c * masksh_stride + r] = FLOAT_TYPE(0);
+                        }
+                    }
+                }
+                // skip the block if the mask is entirely -inf
+                bool all_less = subgroupAll(max_mask <= NEG_FLT_MAX_OVER_2);
+                barrier();
+                if (gl_SubgroupInvocationID == 0) {
+                    tmpsh[gl_SubgroupID] = all_less ? NEG_FLT_MAX_OVER_2 : 0.0f;
+                }
+                barrier();
+                [[unroll]] for (uint s = 0; s < gl_NumSubgroups; ++s) {
+                    max_mask = max(max_mask, tmpsh[s]);
+                }
+                if (max_mask <= NEG_FLT_MAX_OVER_2) {
+                    continue;
+                }
+            }
         }
-        uint32_t mask_opt_bits = (mask_opt >> ((j % 16) * 2)) & 0x3;
-        if (mask_opt_bits == MASK_OPT_ALL_NEG_INF) {
-            // skip this block
-            continue;
-        }
-        // Only load if the block is not all zeros
-        if (MASK_ENABLE && mask_opt_bits != MASK_OPT_ALL_ZERO) {
-            bool nem1_bounds_check = !(p.gqa_ratio > 1) && (p.nem1 % Br) != 0;
 
-            float max_mask = NEG_FLT_MAX_OVER_2;
-            [[unroll]] for (uint32_t idx = 0; idx < Bc * Br; idx += gl_WorkGroupSize.x) {
-                uint32_t c = (idx + tid) % Bc;
-                uint32_t r = (idx + tid) / Bc;
-                if (idx + tid < Bc * Br) {
-                    if ((!KV_bounds_check || j * Bc + c < KV) && (!nem1_bounds_check || i * Br + r < p.nem1)) {
-                        float m = float(data_m[m_offset + (i * Br + r) * m_stride + (j * Bc + c)]);
-                        masksh[c][r] = m;
-                        max_mask = max(max_mask, m);
+        ACC_TYPE Sf[rows_per_thread][cols_per_thread];
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
+                Sf[r][c] = ACC_TYPE(0.0);
+            }
+        }
+
+        if (SHMEM_STAGING != 0) {
+            barrier();
+            [[unroll]] for (uint32_t idx = 0; idx < Bc * HSK / 4; idx += gl_WorkGroupSize.x) {
+                uint32_t d = (idx + tid) % (HSK / 4);
+                uint32_t c = (idx + tid) / (HSK / 4);
+                if (idx + gl_WorkGroupSize.x <= Bc * HSK / 4 || c < Bc) {
+                    FLOAT_TYPEV4 K_Tf = FLOAT_TYPEV4(0);
+                    if (!KV_bounds_check || j * Bc + c < KV) {
+#if BLOCK_SIZE > 1
+                        uint coord = (j * Bc + c) * k_stride * BLOCK_SIZE + 4 * d;
+                        uint ib = coord / BLOCK_SIZE;
+                        uint iqs = (coord % BLOCK_SIZE);
+                        K_Tf = dequantize4(ib, iqs, k_offset, BINDING_IDX_K);
+#else
+                        K_Tf = FLOAT_TYPEV4(data_kv4[k_offset / 4 + (j * Bc + c) * k_stride / 4 + d]);
+#endif
+                    }
+
+                    kvsh[c * kvsh_stride + d] = K_Tf;
+                }
+            }
+            barrier();
+        }
+
+        // More d iterations means Q register caching becomes relevant
+        // Few iterations means the additional registers needed are worse than the speed-up from caching
+        if (HSK_per_thread / 4 > 4) {
+            [[unroll]] for (uint32_t d = 0; d < HSK_per_thread / 4; ++d) {
+                FLOAT_TYPEV4 Q_cache[rows_per_thread];
+                [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                    Q_cache[r] = Qf[tile_row(r) * qf_stride + d * D_split + d_tid];
+                }
+
+                [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
+                    if (KV_bounds_check && j * Bc + c * cols_per_iter + col_tid >= KV) {
+                        continue;
+                    }
+
+                    FLOAT_TYPEV4 K_Tf;
+                    if (SHMEM_STAGING != 0) {
+                        K_Tf = kvsh[(c * cols_per_iter + col_tid) * kvsh_stride + (d * D_split + d_tid)];
                     } else {
-                        masksh[c][r] = float(0);
+#if BLOCK_SIZE > 1
+                        uint coord = (j * Bc + c * cols_per_iter + col_tid) * k_stride * BLOCK_SIZE + 4 * (d * D_split + d_tid);
+                        uint ib = coord / BLOCK_SIZE;
+                        uint iqs = (coord % BLOCK_SIZE);
+                        K_Tf = dequantize4(ib, iqs, k_offset, BINDING_IDX_K);
+#else
+                        K_Tf = FLOAT_TYPEV4(data_kv4[k_offset / 4 + (j * Bc + c * cols_per_iter + col_tid) * k_stride / 4 + d * D_split + d_tid]);
+#endif
+                    }
+                    [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                        Sf[r][c] += ACC_TYPE(dot(Q_cache[r], K_Tf));
                     }
                 }
             }
-            // skip the block if the mask is entirely -inf
-            bool all_less = subgroupAll(max_mask <= NEG_FLT_MAX_OVER_2);
-            barrier();
-            if (gl_SubgroupInvocationID == 0) {
-                tmpsh[gl_SubgroupID] = all_less ? NEG_FLT_MAX_OVER_2 : 0.0f;
-            }
-            barrier();
-            [[unroll]] for (uint s = 0; s < gl_NumSubgroups; ++s) {
-                max_mask = max(max_mask, tmpsh[s]);
-            }
-            if (max_mask <= NEG_FLT_MAX_OVER_2) {
-                continue;
-            }
-        }
-
-        float Sf[Br][cols_per_thread];
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
+        } else {
             [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
-                Sf[r][c] = 0.0;
-            }
-        }
+                if (KV_bounds_check && j * Bc + c * cols_per_iter + col_tid >= KV) {
+                    continue;
+                }
 
-
-        [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
-            if (KV_bounds_check && j * Bc + c * cols_per_iter + col_tid >= KV) {
-                continue;
-            }
-            [[unroll]] for (uint32_t d = 0; d < HSK_per_thread / 4; ++d) {
+                [[unroll]] for (uint32_t d = 0; d < HSK_per_thread / 4; ++d) {
+                    FLOAT_TYPEV4 K_Tf;
+                    if (SHMEM_STAGING != 0) {
+                        K_Tf = kvsh[(c * cols_per_iter + col_tid) * kvsh_stride + (d * D_split + d_tid)];
+                    } else {
 #if BLOCK_SIZE > 1
-                uint coord = (j * Bc + c * cols_per_iter + col_tid) * k_stride * BLOCK_SIZE + 4 * (d * D_split + d_tid);
-                uint ib = coord / BLOCK_SIZE;
-                uint iqs = (coord % BLOCK_SIZE);
-                vec4 K_Tf = dequantize4(ib, iqs, k_offset, BINDING_IDX_K);
+                        uint coord = (j * Bc + c * cols_per_iter + col_tid) * k_stride * BLOCK_SIZE + 4 * (d * D_split + d_tid);
+                        uint ib = coord / BLOCK_SIZE;
+                        uint iqs = (coord % BLOCK_SIZE);
+                        K_Tf = dequantize4(ib, iqs, k_offset, BINDING_IDX_K);
 #else
-                vec4 K_Tf = vec4(data_kv4[k_offset / 4 + (j * Bc + c * cols_per_iter + col_tid) * k_stride / 4 + d * D_split + d_tid]);
+                        K_Tf = FLOAT_TYPEV4(data_kv4[k_offset / 4 + (j * Bc + c * cols_per_iter + col_tid) * k_stride / 4 + d * D_split + d_tid]);
 #endif
-                [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-                    Sf[r][c] += dot(Qf[r][d * D_split + d_tid], K_Tf);
+                    }
+                    [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                        Sf[r][c] += ACC_TYPE(dot(Qf[tile_row(r) * qf_stride + d * D_split + d_tid], K_Tf));
+                    }
                 }
             }
         }
@@ -189,89 +279,109 @@ void main() {
         [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
             // Compute sum across the D_split
             [[unroll]] for (uint s = D_split / 2; s > 0; s >>= 1) {
-                [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
+                [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
                     Sf[r][c] += subgroupShuffleXor(Sf[r][c], s);
                 }
             }
         }
 
         if (LOGIT_SOFTCAP) {
-            [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
                 [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
-                    Sf[r][c] = p.logit_softcap * tanh(Sf[r][c]);
+                    Sf[r][c] = ACC_TYPE(p.logit_softcap * tanh(Sf[r][c]));
                 }
             }
         }
 
         if (MASK_ENABLE && mask_opt_bits != MASK_OPT_ALL_ZERO) {
             [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
-                [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-                    float mvf = masksh[c * cols_per_iter + col_tid][r];
+                [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                    FLOAT_TYPE mvf = masksh[(c * cols_per_iter + col_tid) * masksh_stride + tile_row(r)];
 
                     Sf[r][c] += slope[r]*mvf;
                 }
             }
-            barrier();
         }
 
-        float rowmaxf[Br], Pf[Br][cols_per_thread], rowsumf[Br], eMf[Br], Moldf[Br];
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            rowmaxf[r] = NEG_FLT_MAX_OVER_2;
+        float eMf[rows_per_thread];
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            float rowmaxf = NEG_FLT_MAX_OVER_2;
             [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
                 if (KV_bounds_check && j * Bc + c * cols_per_iter + col_tid >= KV) {
                     continue;
                 }
-                rowmaxf[r] = max(rowmaxf[r], Sf[r][c]);
+                rowmaxf = max(rowmaxf, float(Sf[r][c]));
             }
-            Moldf[r] = Mf[r];
+            float Moldf = Mf[r];
 
             // M = max(rowmax, Mold)
             // P = e^(S - M)
             // eM = e^(Mold - M)
-            Mf[r] = max(rowmaxf[r], Moldf[r]);
-            [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
-                Pf[r][c] = exp(Sf[r][c] - Mf[r]);
-            }
-            eMf[r] = exp(Moldf[r] - Mf[r]);
-
-            // Compute sum across row of P
-            rowsumf[r] = 0.0;
-            [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
-                if (KV_bounds_check && j * Bc + c * cols_per_iter + col_tid >= KV) {
-                    continue;
-                }
-                rowsumf[r] += Pf[r][c];
-            }
-
-            Lf[r] = eMf[r]*Lf[r] + rowsumf[r];
+            Mf[r] = max(rowmaxf, Moldf);
+            eMf[r] = exp(Moldf - Mf[r]);
+            Lf[r] = eMf[r]*Lf[r];
         }
 
         [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-            [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-                Of[r][d] = eMf[r] * Of[r][d];
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                Of[r][d] = FLOAT_TYPE(eMf[r]) * Of[r][d];
             }
         }
 
+        if (SHMEM_STAGING != 0) {
+            barrier();
+            [[unroll]] for (uint32_t idx = 0; idx < Bc * HSV / 4; idx += gl_WorkGroupSize.x) {
+                uint32_t d = (idx + tid) % (HSV / 4);
+                uint32_t c = (idx + tid) / (HSV / 4);
+                if (idx + gl_WorkGroupSize.x <= Bc * HSV / 4 || c < Bc) {
+                    FLOAT_TYPEV4 V_Tf = FLOAT_TYPEV4(0);
+                    if (!KV_bounds_check || j * Bc + c < KV) {
+#if BLOCK_SIZE > 1
+                        uint coord = (j * Bc + c) * v_stride * BLOCK_SIZE + 4 * d;
+                        uint ib = coord / BLOCK_SIZE;
+                        uint iqs = (coord % BLOCK_SIZE);
+                        V_Tf = dequantize4(ib, iqs, v_offset, BINDING_IDX_V);
+#else
+                        V_Tf = FLOAT_TYPEV4(data_vv4[v_offset / 4 + (j * Bc + c) * v_stride / 4 + d]);
+#endif
+                    }
+
+                    kvsh[c * kvsh_stride + d] = V_Tf;
+                }
+            }
+            barrier();
+        }
+
         [[unroll]] for (uint32_t c = 0; c < cols_per_thread; ++c) {
             if (KV_bounds_check && j * Bc + c * cols_per_iter + col_tid >= KV) {
                 continue;
             }
+
+            FLOAT_TYPE Pf[rows_per_thread];
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                Pf[r] = FLOAT_TYPE(exp(float(Sf[r][c]) - Mf[r]));
+                Lf[r] += Pf[r];
+            }
+
             [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
+                FLOAT_TYPEV4 Vf;
+                if (SHMEM_STAGING != 0) {
+                    Vf = kvsh[(c * cols_per_iter + col_tid) * kvsh_stride + (d * D_split + d_tid)];
+                } else {
 #if BLOCK_SIZE > 1
-                uint coord = (j * Bc + c * cols_per_iter + col_tid) * v_stride * BLOCK_SIZE + 4 * (d * D_split + d_tid);
-                uint ib = coord / BLOCK_SIZE;
-                uint iqs = (coord % BLOCK_SIZE);
-                vec4 Vf = dequantize4(ib, iqs, v_offset, BINDING_IDX_V);
+                    uint coord = (j * Bc + c * cols_per_iter + col_tid) * v_stride * BLOCK_SIZE + 4 * (d * D_split + d_tid);
+                    uint ib = coord / BLOCK_SIZE;
+                    uint iqs = (coord % BLOCK_SIZE);
+                    Vf = dequantize4(ib, iqs, v_offset, BINDING_IDX_V);
 #else
-                vec4 Vf = vec4(data_vv4[v_offset / 4 + (j * Bc + c * cols_per_iter + col_tid) * v_stride / 4 + d * D_split + d_tid]);
+                    Vf = FLOAT_TYPEV4(data_vv4[v_offset / 4 + (j * Bc + c * cols_per_iter + col_tid) * v_stride / 4 + d * D_split + d_tid]);
 #endif
-                [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-                    Of[r][d] += Pf[r][c] * Vf;
+                }
+                [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                    Of[r][d] += FLOAT_TYPEV4(Pf[r] * Vf);
                 }
             }
         }
-
-        barrier();
     }
 
     // prevent race on tmpsh
@@ -279,58 +389,108 @@ void main() {
 
     // reduce across threads
 
-    [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-        float rowmaxf, eMf;
+    [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+        float rowmaxf = Mf[r];
 
-        tmpsh[tid] = Mf[r];
         // Compute max across the row
-        barrier();
-        [[unroll]] for (int s = int(gl_WorkGroupSize.x) / 2; s >= D_split; s >>= 1) {
-            if (tid < s) {
-                tmpsh[tid] = max(tmpsh[tid], tmpsh[tid + s]);
+        if (SubGroupSize > 0) {
+            [[unroll]] for (uint s = D_split; s < SubGroupSize; s *= 2) {
+                rowmaxf = max(rowmaxf, subgroupShuffleXor(rowmaxf, s));
             }
+            if (row_split == 1) {
+                // Reduce inside workgroup with shmem
+                barrier();
+                if (gl_SubgroupInvocationID == d_tid) {
+                    tmpsh[gl_SubgroupID * D_split + d_tid] = rowmaxf;
+                }
+                barrier();
+                rowmaxf = tmpsh[d_tid];
+                [[unroll]] for (uint32_t s = 1; s < num_subgroups; ++s) {
+                    rowmaxf = max(rowmaxf, tmpsh[s * D_split + d_tid]);
+                }
+            }
+        } else {
             barrier();
+            tmpsh[tid] = rowmaxf;
+            barrier();
+            [[unroll]] for (int s = int(threads_per_rowgroup) / 2; s >= D_split; s >>= 1) {
+                if (rowgroup_tid < s) {
+                    tmpsh[tid] = max(tmpsh[tid], tmpsh[tid ^ s]);
+                }
+                barrier();
+            }
+            rowmaxf = tmpsh[row_tid * threads_per_rowgroup + d_tid];
         }
-        rowmaxf = tmpsh[d_tid];
-        barrier();
 
         float Moldf = Mf[r];
 
         // M = max(rowmax, Mold)
         // eM = e^(Mold - M)
         Mf[r] = max(rowmaxf, Moldf);
-        eMf = exp(Moldf - Mf[r]);
+        float eMf = exp(Moldf - Mf[r]);
 
         Lf[r] = eMf*Lf[r];
 
-        tmpsh[tid] = Lf[r];
-
         // Compute sum across the row
-        barrier();
-        [[unroll]] for (int s = int(gl_WorkGroupSize.x) / 2; s >= D_split; s >>= 1) {
-            if (tid < s) {
-                tmpsh[tid] = tmpsh[tid] + tmpsh[tid + s];
+        if (SubGroupSize > 0) {
+            [[unroll]] for (uint s = D_split; s < SubGroupSize; s *= 2) {
+                Lf[r] += subgroupShuffleXor(Lf[r], s);
             }
+            if (row_split == 1) {
+                barrier();
+                if (gl_SubgroupInvocationID == d_tid) {
+                    tmpsh[gl_SubgroupID * D_split + d_tid] = Lf[r];
+                }
+                barrier();
+                Lf[r] = tmpsh[d_tid];
+                [[unroll]] for (uint32_t s = 1; s < num_subgroups; ++s) {
+                    Lf[r] += tmpsh[s * D_split + d_tid];
+                }
+            }
+        } else {
             barrier();
-        }
-        Lf[r] = tmpsh[d_tid];
-        barrier();
-
-        [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-
-            Of[r][d] = eMf * Of[r][d];
-            tmpshv4[tid] = Of[r][d];
-
+            tmpsh[tid] = Lf[r];
             barrier();
-            [[unroll]] for (int s = int(gl_WorkGroupSize.x) / 2; s >= D_split; s >>= 1) {
-                if (tid < s) {
-                    Of[r][d] += tmpshv4[tid + s];
-                    tmpshv4[tid] = Of[r][d];
+            [[unroll]] for (int s = int(threads_per_rowgroup) / 2; s >= D_split; s >>= 1) {
+                if (rowgroup_tid < s) {
+                    tmpsh[tid] = tmpsh[tid] + tmpsh[tid ^ s];
                 }
                 barrier();
             }
-            Of[r][d] = tmpshv4[d_tid];
-            barrier();
+            Lf[r] = tmpsh[row_tid * threads_per_rowgroup + d_tid];
+        }
+
+        [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
+            Of[r][d] = FLOAT_TYPE(eMf) * Of[r][d];
+
+            if (SubGroupSize > 0) {
+                [[unroll]] for (uint s = D_split; s < SubGroupSize; s *= 2) {
+                    Of[r][d] += subgroupShuffleXor(Of[r][d], s);
+                }
+                if (row_split == 1) {
+                    barrier();
+                    if (gl_SubgroupInvocationID == d_tid) {
+                        tmpshv4[gl_SubgroupID * D_split + d_tid] = Of[r][d];
+                    }
+                    barrier();
+                    Of[r][d] = tmpshv4[d_tid];
+                    [[unroll]] for (uint32_t s = 1; s < num_subgroups; ++s) {
+                        Of[r][d] += tmpshv4[s * D_split + d_tid];
+                    }
+                }
+            } else {
+                barrier();
+                tmpshv4[tid] = Of[r][d];
+                barrier();
+                [[unroll]] for (int s = int(threads_per_rowgroup) / 2; s >= D_split; s >>= 1) {
+                    if (rowgroup_tid < s) {
+                        Of[r][d] += tmpshv4[tid ^ s];
+                        tmpshv4[tid] = Of[r][d];
+                    }
+                    barrier();
+                }
+                Of[r][d] = tmpshv4[row_tid * threads_per_rowgroup + d_tid];
+            }
         }
     }
 
@@ -338,33 +498,53 @@ void main() {
     // If there is split_k, then the split_k resolve shader does the final
     // division by L. Store the intermediate O value and per-row m and L values.
     if (p.k_num > 1) {
-        // note: O and Q have swapped coord 1,2.
-        uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
+        if (p.gqa_ratio > 1) {
+            // note: O and Q have swapped coord 1,2.
+            uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3)) / 4;
 
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            if (r < N) {
-                [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-                    [[unroll]] for (uint32_t comp = 0; comp < 4; ++comp) {
-                        perElemOpGqaStore(r, 4*(d * D_split + d_tid) + comp, Of[r][d][comp], o_offset, iq2, N);
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                const uint row = tile_row(r);
+                if (row < N) {
+                    [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
+                        gqaStore(row, d * D_split + d_tid, Of[r][d], o_offset, iq2, N);
                     }
                 }
             }
-        }
 
-        o_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            if (r < N) {
-                perElemOpStoreCol0(r, 0u, ACC_TYPE(Lf[r]), o_offset, iq2, N);
-                perElemOpStoreCol0(r, 0u, ACC_TYPE(Mf[r]), o_offset + p.ne1, iq2, N);
+            o_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                const uint row = tile_row(r);
+                if (row < N) {
+                    perElemOpStoreCol0(row, 0u, ACC_TYPE(Lf[r]), o_offset, iq2, N);
+                    perElemOpStoreCol0(row, 0u, ACC_TYPE(Mf[r]), o_offset + p.ne1, iq2, N);
+                }
+            }
+        } else {
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                const uint row = tile_row(r);
+                const uint global_row = i * Br + row;
+
+                if (global_row < N) {
+                    uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (global_row + p.ne2 * iq3)) / 4;
+
+                    [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
+                        data_ov4[o_offset + iq2 * HSV/4 + d * D_split + d_tid] = D_TYPEV4(Of[r][d]);
+                    }
+                }
+
+                if (global_row < N && d_tid == 0 && col_tid == 0) {
+                    uint32_t lm_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (global_row + p.ne2 * iq3));
+                    data_o[lm_offset + iq2] = D_TYPE(Lf[r]);
+                    data_o[lm_offset + p.ne1 + iq2] = D_TYPE(Mf[r]);
+                }
             }
         }
-
         return;
     }
 
     if ((p.mask_n_head_log2 & SINK_ENABLE_BIT) != 0) {
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            float sink = perElemOpGetSink(r, 0u, ACC_TYPE(0), iq2);
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            float sink = perElemOpGetSink(tile_row(r), 0u, ACC_TYPE(0), iq2);
 
             float ms = 1.0f;
             float vs = 1.0f;
@@ -373,7 +553,7 @@ void main() {
                 ms = exp(Mf[r] - sink);
 
                 [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-                    Of[r][d] *= ms;
+                    Of[r][d] *= FLOAT_TYPE(ms);
                 }
             } else {
                 vs = exp(sink - Mf[r]);
@@ -383,39 +563,37 @@ void main() {
         }
     }
 
-    float Lfrcp[Br];
-    [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
+    float Lfrcp[rows_per_thread];
+    [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
         Lfrcp[r] = (Lf[r] == 0.0) ? 0.0 : (1.0 / Lf[r]);
     }
 
     [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            Of[r][d] *= Lfrcp[r];
-#if defined(ACC_TYPE_MAX)
-            Of[r][d] = clamp(Of[r][d], -vec4(ACC_TYPE_MAX), vec4(ACC_TYPE_MAX));
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            Of[r][d] *= FLOAT_TYPE(Lfrcp[r]);
+#if defined(FLOAT_TYPE_MAX)
+            Of[r][d] = clamp(Of[r][d], -FLOAT_TYPE_MAX, FLOAT_TYPE_MAX);
 #endif
         }
     }
 
-    uint32_t o_offset = gqa_iq1*p.ne1*HSV + iq3*p.ne2*p.ne1*HSV;
+    uint32_t o_offset = (gqa_iq1*p.ne1*HSV + iq3*p.ne2*p.ne1*HSV) / 4;
 
     if (p.gqa_ratio > 1) {
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            if (r < N) {
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            const uint row = tile_row(r);
+            if (row < N) {
                 [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-                    [[unroll]] for (uint32_t comp = 0; comp < 4; ++comp) {
-                        perElemOpGqaStore(r, 4*(d * D_split + d_tid) + comp, Of[r][d][comp], o_offset, iq2, N);
-                    }
+                    gqaStore(row, d * D_split + d_tid, Of[r][d], o_offset, iq2, N);
                 }
             }
         }
     } else {
-        [[unroll]] for (uint32_t r = 0; r < Br; ++r) {
-            if (i * Br + r < N) {
+        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+            const uint row = tile_row(r);
+            if (i * Br + row < N) {
                 [[unroll]] for (uint32_t d = 0; d < HSV_per_thread / 4; ++d) {
-                    [[unroll]] for (uint32_t comp = 0; comp < 4; ++comp) {
-                        data_o[o_offset + iq2 * HSV + (i * Br + r) * p.ne1 * HSV + 4*(d * D_split + d_tid) + comp] = D_TYPE(Of[r][d][comp]);
-                    }
+                    data_ov4[o_offset + (iq2 * HSV + (i * Br + row) * p.ne1 * HSV) / 4 + d * D_split + d_tid] = D_TYPEV4(Of[r][d]);
                 }
             }
         }

ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.glsl

@@ -1,16 +1,18 @@
 
 layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
 
-layout (constant_id = 0) const uint32_t WorkGroupSize = 128;
-layout (constant_id = 1) const uint32_t Br = 1;
-layout (constant_id = 2) const uint32_t Bc = 32;
-layout (constant_id = 3) const uint32_t HSK = 32;
-layout (constant_id = 4) const uint32_t HSV = 32;
-layout (constant_id = 5) const uint32_t Clamp = 0;
-layout (constant_id = 6) const uint32_t D_split = 16;
-layout (constant_id = 7) const uint32_t SubGroupSize = 32;
-layout (constant_id = 8) const uint32_t K_LOAD_SHMEM = 0;
-layout (constant_id = 9) const uint32_t Flags = 0;
+layout (constant_id =  0) const uint32_t WorkGroupSize = 128;
+layout (constant_id =  1) const uint32_t Br = 1;
+layout (constant_id =  2) const uint32_t Bc = 32;
+layout (constant_id =  3) const uint32_t HSK = 32;
+layout (constant_id =  4) const uint32_t HSV = 32;
+layout (constant_id =  5) const uint32_t Clamp = 0;
+layout (constant_id =  6) const uint32_t D_split = 16;
+layout (constant_id =  7) const uint32_t row_split = 1;
+layout (constant_id =  8) const uint32_t SubGroupSize = 32;
+layout (constant_id =  9) const uint32_t SHMEM_STAGING = 0;
+layout (constant_id = 10) const uint32_t Flags = 0;
+layout (constant_id = 11) const uint32_t LIMIT_OCCUPANCY_SHMEM = 0;
 
 const bool USE_MASK_OPT  = (Flags & 1) != 0;
 const bool MASK_ENABLE   = (Flags & 2) != 0;
@@ -69,6 +71,7 @@ layout (push_constant) uniform parameter {
 layout (binding = 4) readonly buffer S {float data_s[];};
 
 layout (binding = 5) writeonly buffer O {D_TYPE data_o[];};
+layout (binding = 5) writeonly buffer OV4 {D_TYPEV4 data_ov4[];};
 
 layout (binding = 6) readonly buffer MO {uint32_t data_mask_opt[];};
 
@@ -94,12 +97,12 @@ layout (binding = 2) readonly buffer V_PACKED16 {A_TYPE_PACKED16 v_data_packed16
 #define BLOCK_SIZE 4
 #define BLOCK_BYTE_SIZE 16
 
-vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
+FLOAT_TYPEV4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
     // iqs is currently always zero in the flash attention shaders
     if (binding_idx == BINDING_IDX_K) {
-        return k_packed.k_data_packed[a_offset + ib];
+        return FLOAT_TYPEV4(k_packed.k_data_packed[a_offset + ib]);
     } else {
-        return v_packed.v_data_packed[a_offset + ib];
+        return FLOAT_TYPEV4(v_packed.v_data_packed[a_offset + ib]);
     }
 }
 #endif
@@ -107,7 +110,7 @@ vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
 #if defined(DATA_A_Q4_0)
 #define BLOCK_BYTE_SIZE 18
 
-vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
+FLOAT_TYPEV4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
     if (binding_idx == BINDING_IDX_K) {
         uint vui_lo = uint(k_packed.k_data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 0]);
         uint vui_hi = uint(k_packed.k_data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 1]);
@@ -115,7 +118,7 @@ vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
         vui_lo >>= shift;
         vui_hi >>= shift;
 
-        return float(k_packed.k_data_packed16[a_offset + ib].d) * (vec4(vui_lo & 0xF, (vui_lo >> 8) & 0xF, vui_hi & 0xF, (vui_hi >> 8) & 0xF) - 8.0f);
+        return FLOAT_TYPE(k_packed.k_data_packed16[a_offset + ib].d) * (FLOAT_TYPEV4(vui_lo & 0xF, (vui_lo >> 8) & 0xF, vui_hi & 0xF, (vui_hi >> 8) & 0xF) - FLOAT_TYPE(8.0f));
     } else {
         uint vui_lo = uint(v_packed.v_data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 0]);
         uint vui_hi = uint(v_packed.v_data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 1]);
@@ -123,24 +126,24 @@ vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
         vui_lo >>= shift;
         vui_hi >>= shift;
 
-        return float(v_packed.v_data_packed16[a_offset + ib].d) * (vec4(vui_lo & 0xF, (vui_lo >> 8) & 0xF, vui_hi & 0xF, (vui_hi >> 8) & 0xF) - 8.0f);
+        return FLOAT_TYPE(v_packed.v_data_packed16[a_offset + ib].d) * (FLOAT_TYPEV4(vui_lo & 0xF, (vui_lo >> 8) & 0xF, vui_hi & 0xF, (vui_hi >> 8) & 0xF) - FLOAT_TYPE(8.0f));
     }
 }
 #endif
 
 #if defined(DATA_A_Q8_0)
 #define BLOCK_BYTE_SIZE 34
-vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
+FLOAT_TYPEV4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
     if (binding_idx == BINDING_IDX_K) {
         const i8vec2 v0 = unpack8(int32_t(k_packed.k_data_packed16[a_offset + ib].qs[iqs / 2])).xy; // vec4 used due to #12147
         const i8vec2 v1 = unpack8(int32_t(k_packed.k_data_packed16[a_offset + ib].qs[iqs / 2 + 1])).xy;
 
-        return float(k_packed.k_data_packed16[a_offset + ib].d) * vec4(v0.x, v0.y, v1.x, v1.y);
+        return FLOAT_TYPE(k_packed.k_data_packed16[a_offset + ib].d) * FLOAT_TYPEV4(v0.x, v0.y, v1.x, v1.y);
     } else {
         const i8vec2 v0 = unpack8(int32_t(v_packed.v_data_packed16[a_offset + ib].qs[iqs / 2])).xy; // vec4 used due to #12147
         const i8vec2 v1 = unpack8(int32_t(v_packed.v_data_packed16[a_offset + ib].qs[iqs / 2 + 1])).xy;
 
-        return float(v_packed.v_data_packed16[a_offset + ib].d) * vec4(v0.x, v0.y, v1.x, v1.y);
+        return FLOAT_TYPE(v_packed.v_data_packed16[a_offset + ib].d) * FLOAT_TYPEV4(v0.x, v0.y, v1.x, v1.y);
     }
 }
 #endif
@@ -189,10 +192,16 @@ void init_indices()
     KV = p.KV;
 
     if (p.k_num > 1) {
-        i = 0;
-        // batch and split_k share gl_WorkGroupID.x
-        gqa_iq1 = gl_WorkGroupID.x / p.k_num;
-        split_k_index = gl_WorkGroupID.x % p.k_num;
+        if (p.gqa_ratio > 1) {
+            i = 0;
+            // batch and split_k share gl_WorkGroupID.x
+            gqa_iq1 = gl_WorkGroupID.x / p.k_num;
+            split_k_index = gl_WorkGroupID.x % p.k_num;
+        } else {
+            gqa_iq1 = 0;
+            split_k_index = gl_WorkGroupID.x % p.k_num;
+            i = gl_WorkGroupID.x / p.k_num;
+        }
     } else if (p.gqa_ratio > 1) {
         i = 0;
         gqa_iq1 = gl_WorkGroupID.x;
@@ -244,3 +253,11 @@ void init_indices()
 // Bias applied to softmax to stay in fp16 range.
 // Based on ggml-cuda issue https://github.com/ggml-org/llama.cpp/issues/18606
 const float FATTN_KQ_MAX_OFFSET = 3.0f*0.6931f;
+
+// Store the output when doing grouped query attention.
+// Rows index by Q's dimension 2, and the first N rows are valid.
+void gqaStore(const in uint32_t r, const in uint32_t c, const in FLOAT_TYPEV4 elems, const in uint32_t o_offset, const in uint32_t iq2, const in uint32_t N)
+{
+    uint32_t offset = (iq2 + r) * HSV / 4 + c;
+    data_ov4[o_offset + offset] = D_TYPEV4(elems);
+}

ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp

@@ -19,7 +19,6 @@
 const uint32_t MatBr = 16;
 const uint32_t MatBc = 16;
 
-const uint32_t row_split = Bc / MatBc;
 const uint32_t rows_per_thread = Br / row_split;
 const uint32_t cols_per_iter = gl_WorkGroupSize.x / row_split;
 const uint32_t cols_per_thread = Bc / cols_per_iter;
@@ -33,15 +32,6 @@ layout (binding = 2) readonly buffer V {float16_t data_v[];};
 layout (binding = 2) readonly buffer VV4 {f16vec4 data_vv4[];};
 layout (binding = 3) readonly buffer M {float16_t data_m[];};
 
-// Store the output when doing grouped query attention.
-// Rows index by Q's dimension 2, and the first N rows are valid.
-D_TYPE perElemOpGqaStore(const in uint32_t r, const in uint32_t c, const in D_TYPE elem, const in uint32_t o_offset, const in uint32_t iq2, const in uint32_t N)
-{
-    uint32_t offset = (iq2 + r) * HSV + c;
-    data_o[o_offset + offset] = D_TYPE(elem);
-    return elem;
-}
-
 shared float tmpsh[row_split];
 
 const uint32_t qstride = HSK_pad / 4 + 2; // in units of f16vec4
@@ -54,10 +44,14 @@ shared f16vec4 Psh[Bc * psh_stride];
 const uint32_t sfshstride = (HSK <= 128) ? (Br / 4 + 2) : Br / 4;
 shared ACC_TYPEV4 sfsh[Bc * sfshstride];
 
-const uint32_t kshstride = (K_LOAD_SHMEM != 0 ? HSK_pad : MatBr) / 4 + 2; // in units of f16vec4
+const uint32_t D_pad = HSK_pad > HSV_pad ? HSK_pad : HSV_pad;
+const uint32_t kvsh_stride = (SHMEM_STAGING != 0 ? D_pad : MatBr) / 4 + 2; // in units of f16vec4
 const uint v_cols = MatBc / 4 * row_split; // total cols, 4 vec4s per MatBc * number of subgroups
 const uint vsh_stride = v_cols;
-shared f16vec4 ksh[(kshstride >= vsh_stride) ? (Bc * kshstride) : (Bc * vsh_stride)];
+shared f16vec4 kvsh[(kvsh_stride >= vsh_stride) ? (Bc * kvsh_stride) : (Bc * vsh_stride)];
+
+const uint32_t osh_stride = row_split * MatBr / 4;
+shared f16vec4 pvsh[MatBc * osh_stride];
 
 shared ACC_TYPE slope[Br];
 
@@ -84,11 +78,6 @@ void main() {
                 Qf[i + tid] = f16vec4(0);
             }
         }
-        [[unroll]] for (uint i = 0; i < Bc * kshstride; i += gl_WorkGroupSize.x) {
-            if (i + tid < Bc * kshstride) {
-                ksh[i + tid] = f16vec4(0);
-            }
-        }
         barrier();
     }
 
@@ -104,10 +93,10 @@ void main() {
     }
     barrier();
 
-    ACC_TYPEV4 Of[rows_per_thread][d_per_thread];
+    f16vec4 Of[rows_per_thread][d_per_thread];
     [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
         [[unroll]] for (uint32_t d = 0; d < d_per_thread; ++d) {
-            Of[r][d] = ACC_TYPEV4(0.0);
+            Of[r][d] = f16vec4(0.0);
         }
     }
 
@@ -153,22 +142,22 @@ void main() {
 
     uint32_t mask_opt = 0;
     uint32_t mask_opt_idx = ~0;
+    uint32_t mask_opt_bits = 0;
+    f16vec4 mask_cache[Bc * Br / 4 / WorkGroupSize];
 
     [[dont_unroll]]
     for (uint32_t j = start_j; j < end_j; ++j) {
 
-        f16vec4 mask_cache[Bc * Br / 4 / WorkGroupSize];
         [[unroll]] for (uint32_t idx = 0; idx < mask_cache.length(); ++idx) {
             mask_cache[idx] = f16vec4(0);
         }
 
         if (MASK_ENABLE) {
-
             if (USE_MASK_OPT && mask_opt_idx != j / 16) {
                 mask_opt_idx = j / 16;
                 mask_opt = data_mask_opt[mo_offset + mask_opt_idx];
             }
-            uint32_t mask_opt_bits = (mask_opt >> ((j % 16) * 2)) & 0x3;
+            mask_opt_bits = (mask_opt >> ((j % 16) * 2)) & 0x3;
             if (mask_opt_bits == MASK_OPT_ALL_NEG_INF) {
                 // skip this block
                 continue;
@@ -231,24 +220,24 @@ void main() {
             }
         }
 
-        if (K_LOAD_SHMEM != 0) {
-            [[unroll]] for (uint32_t idx = 0; idx < Bc * HSK / 4; idx += gl_WorkGroupSize.x) {
-                uint32_t d = (idx + tid) % (HSK / 4);
-                uint32_t c = (idx + tid) / (HSK / 4);
-                if (c < Bc && d < HSK / 4) {
+        if (SHMEM_STAGING != 0) {
+            [[unroll]] for (uint32_t idx = 0; idx < Bc * HSK_pad / 4; idx += gl_WorkGroupSize.x) {
+                uint32_t d = (idx + tid) % (HSK_pad / 4);
+                uint32_t c = (idx + tid) / (HSK_pad / 4);
+                if (idx + gl_WorkGroupSize.x <= Bc * HSK_pad / 4 || c < Bc) {
                     f16vec4 K_Tf = f16vec4(0);
-                    if (!KV_bounds_check || j * Bc + c < KV) {
+                    if ((!KV_bounds_check || j * Bc + c < KV) && (HSK == HSK_pad || d < HSK / 4)) {
 #if BLOCK_SIZE > 1
                         uint coord = (j * Bc + c) * k_stride * BLOCK_SIZE + 4 * d;
                         uint ib = coord / BLOCK_SIZE;
                         uint iqs = (coord % BLOCK_SIZE);
-                        K_Tf = f16vec4(dequantize4(ib, iqs, k_offset, BINDING_IDX_K));
+                        K_Tf = dequantize4(ib, iqs, k_offset, BINDING_IDX_K);
 #else
                         K_Tf = f16vec4(data_kv4[k_offset / 4 + (j * Bc + c) * k_stride / 4 + d]);
 #endif
                     }
 
-                    ksh[c * kshstride + d] = K_Tf;
+                    kvsh[c * kvsh_stride + d] = K_Tf;
                 }
             }
             barrier();
@@ -262,7 +251,11 @@ void main() {
         coopmat<float16_t, gl_ScopeSubgroup, 16, MatBr, gl_MatrixUseB> QMat;
 
         [[unroll]] for (uint32_t d = 0; d < HSK_pad / 16; ++d) {
-            if (K_LOAD_SHMEM == 0) {
+            // If SHMEM_STAGING is set, a Bc * HSK_pad size tile of K is loaded to shmem
+            // If not, f16 K is loaded directly from global memory if aligned, otherwise
+            // staged through a Bc * MatBr size staging buffer.
+            // If K is not type f16, then it is always staged for dequantization.
+            if (SHMEM_STAGING == 0) {
 #if BLOCK_SIZE == 1
             if (KV_bounds_check || d * 16 + 16 > HSK) {
 #endif
@@ -277,13 +270,13 @@ void main() {
                         uint coord = (j * Bc + row) * k_stride * BLOCK_SIZE + d * 16 + col_vec * 4;
                         uint ib = coord / BLOCK_SIZE;
                         uint iqs = (coord % BLOCK_SIZE);
-                        K_Tf = f16vec4(dequantize4(ib, iqs, k_offset, BINDING_IDX_K));
+                        K_Tf = dequantize4(ib, iqs, k_offset, BINDING_IDX_K);
 #else
                         K_Tf = f16vec4(data_kv4[k_offset / 4 + (j * Bc + row) * k_stride / 4 + d * 16 / 4 + col_vec]);
 #endif
                     }
 
-                    ksh[row * kshstride + col_vec] = K_Tf;
+                    kvsh[row * kvsh_stride + col_vec] = K_Tf;
                 }
             }
             barrier();
@@ -295,8 +288,8 @@ void main() {
             if (KV_bounds_check || d * 16 + 16 > HSK)
 #endif
             {
-                uint coord = (gl_SubgroupID * MatBc) * kshstride;
-                coopMatLoad(KMat, ksh, coord, kshstride, gl_CooperativeMatrixLayoutRowMajor);
+                uint coord = (gl_SubgroupID * MatBc) * kvsh_stride;
+                coopMatLoad(KMat, kvsh, coord, kvsh_stride, gl_CooperativeMatrixLayoutRowMajor);
             }
 #if BLOCK_SIZE == 1
             else {
@@ -305,8 +298,8 @@ void main() {
             }
 #endif
             } else {
-                uint coord = (gl_SubgroupID * MatBc) * kshstride + d * 16 / 4;
-                coopMatLoad(KMat, ksh, coord, kshstride, gl_CooperativeMatrixLayoutRowMajor);
+                uint coord = (gl_SubgroupID * MatBc) * kvsh_stride + d * 16 / 4;
+                coopMatLoad(KMat, kvsh, coord, kvsh_stride, gl_CooperativeMatrixLayoutRowMajor);
             }
 
             coopMatLoad(QMat, Qf, d * 16 / 4, qstride, gl_CooperativeMatrixLayoutColumnMajor);
@@ -329,7 +322,7 @@ void main() {
             barrier();
         }
 
-        if (MASK_ENABLE) {
+        if (MASK_ENABLE && mask_opt_bits != MASK_OPT_ALL_ZERO) {
             [[unroll]] for (uint32_t idx = 0; idx < Bc * Br / 4; idx += gl_WorkGroupSize.x) {
                 uint32_t c = (idx + tid) / (Br / 4);
                 uint32_t r = (idx + tid) % (Br / 4);
@@ -374,7 +367,7 @@ void main() {
         [[unroll]] for (uint32_t d0 = 0; d0 < HSV / 4; d0 += threads_per_rowgroup) {
             const uint d_local = d0 / threads_per_rowgroup;
             [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
-                Of[r][d_local] = ACC_TYPE(eMf[r]) * Of[r][d_local];
+                Of[r][d_local] = float16_t(eMf[r]) * Of[r][d_local];
             }
         }
 
@@ -397,19 +390,47 @@ void main() {
             }
         }
 
+        if (SHMEM_STAGING != 0) {
+            [[unroll]] for (uint32_t idx = 0; idx < Bc * HSV_pad / 4; idx += gl_WorkGroupSize.x) {
+                uint32_t d = (idx + tid) % (HSV_pad / 4);
+                uint32_t c = (idx + tid) / (HSV_pad / 4);
+                if (idx + gl_WorkGroupSize.x <= Bc * HSV_pad / 4 || c < Bc) {
+                    f16vec4 V_Tf = f16vec4(0);
+                    if ((!KV_bounds_check || j * Bc + c < KV) && (HSV == HSV_pad || d < HSV / 4)) {
+#if BLOCK_SIZE > 1
+                        uint coord = (j * Bc + c) * v_stride * BLOCK_SIZE + 4 * d;
+                        uint ib = coord / BLOCK_SIZE;
+                        uint iqs = (coord % BLOCK_SIZE);
+                        V_Tf = dequantize4(ib, iqs, v_offset, BINDING_IDX_V);
+#else
+                        V_Tf = f16vec4(data_vv4[v_offset / 4 + (j * Bc + c) * v_stride / 4 + d]);
+#endif
+                    }
+
+                    kvsh[c * kvsh_stride + d] = V_Tf;
+                }
+            }
+        }
+        barrier();
+
         const uint num_hsv_tiles = (HSV + MatBc * row_split - 1) / (MatBc * row_split); // round up
 
         // Each subgroup handles HSV/4 columns
         [[unroll]] for (uint32_t hsv_tile = 0; hsv_tile < num_hsv_tiles; ++hsv_tile) {
             const uint hsv_offset = (hsv_tile * row_split + gl_SubgroupID) * 16;
 
-            SfMat = coopmat<ACC_TYPE, gl_ScopeSubgroup, MatBc, MatBr, gl_MatrixUseAccumulator>(0);
+            coopmat<float16_t, gl_ScopeSubgroup, MatBc, MatBr, gl_MatrixUseAccumulator> PVMat = coopmat<float16_t, gl_ScopeSubgroup, MatBc, MatBr, gl_MatrixUseAccumulator>(0);
 
             // Preload V tiles for [Bc, 16 * num subgroups]
             const uint v_rows = Bc;
             const uint v_total = v_rows * v_cols;
             const uint v_loads_per_thread = v_total / gl_WorkGroupSize.x;
 
+            // If SHMEM_STAGING is set, a Bc * HSV_pad size tile of V is loaded to shmem.
+            // If not, f16 V is loaded directly from global memory if aligned, otherwise
+            // staged through a Bc * MatBr size staging buffer.
+            // If V is not type f16, then it is always staged for dequantization.
+            if (SHMEM_STAGING == 0) {
 #if BLOCK_SIZE == 1
             // For f16, only preload if not aligned
             if (KV_bounds_check) {
@@ -428,44 +449,52 @@ void main() {
 
                 if (!KV_bounds_check || (v_row < KV && v_col < HSV)) {
 #if BLOCK_SIZE > 1
-                    ksh[row * vsh_stride + col] = f16vec4(dequantize4(ib, iqs, v_offset, BINDING_IDX_V));
+                    kvsh[row * vsh_stride + col] = dequantize4(ib, iqs, v_offset, BINDING_IDX_V);
 #else
-                    ksh[row * vsh_stride + col] = data_vv4[(v_offset + v_row * v_stride + v_col) / 4];
+                    kvsh[row * vsh_stride + col] = data_vv4[(v_offset + v_row * v_stride + v_col) / 4];
 #endif
                 } else {
-                    ksh[row * vsh_stride + col] = f16vec4(0.0f);
+                    kvsh[row * vsh_stride + col] = f16vec4(0.0f);
                 }
             }
+
 #if BLOCK_SIZE == 1
             }
 #endif
-
+            }
             barrier();
 
-            [[unroll]] for (uint32_t bc_chunk = 0; bc_chunk < Bc / MatBc; ++bc_chunk) {
-                coopMatLoad(KMat, Psh, bc_chunk * MatBc * psh_stride, psh_stride, gl_CooperativeMatrixLayoutColumnMajor);
+            const uint o_offset = gl_SubgroupID * MatBr / 4;
 
+            if (hsv_offset < HSV_pad) {
+                [[unroll]] for (uint32_t bc_chunk = 0; bc_chunk < Bc / MatBc; ++bc_chunk) {
+                    coopMatLoad(KMat, Psh, bc_chunk * MatBc * psh_stride, psh_stride, gl_CooperativeMatrixLayoutColumnMajor);
+
+                    if (SHMEM_STAGING == 0) {
 #if BLOCK_SIZE == 1
-                if (!KV_bounds_check) {
-                    // F16 values can be loaded directly from global memory
-                    const uint v_tile_row = j * Bc + bc_chunk * MatBc;
-                    const uint v_tile_offset = v_offset / 4 + v_tile_row * v_stride / 4 + hsv_offset / 4;
-                    coopMatLoad(QMat, data_vv4, v_tile_offset, v_stride / 4, gl_CooperativeMatrixLayoutRowMajor);
-                } else
+                    if (!KV_bounds_check) {
+                        // F16 values can be loaded directly from global memory
+                        const uint v_tile_row = j * Bc + bc_chunk * MatBc;
+                        const uint v_tile_offset = v_offset / 4 + v_tile_row * v_stride / 4 + hsv_offset / 4;
+                        coopMatLoad(QMat, data_vv4, v_tile_offset, v_stride / 4, gl_CooperativeMatrixLayoutRowMajor);
+                    } else
 #endif
-                {
-                    const uint v_tile_offset = bc_chunk * MatBr * v_cols + gl_SubgroupID * (MatBc / 4);
-                    coopMatLoad(QMat, ksh, v_tile_offset, vsh_stride, gl_CooperativeMatrixLayoutRowMajor);
+                    {
+                        const uint v_tile_offset = bc_chunk * MatBr * v_cols + gl_SubgroupID * (MatBc / 4);
+                        coopMatLoad(QMat, kvsh, v_tile_offset, vsh_stride, gl_CooperativeMatrixLayoutRowMajor);
+                    }
+                    } else {
+                        const uint v_tile_offset = bc_chunk * MatBc * kvsh_stride + (hsv_tile * row_split + gl_SubgroupID) * (MatBc / 4);
+                        coopMatLoad(QMat, kvsh, v_tile_offset, kvsh_stride, gl_CooperativeMatrixLayoutRowMajor);
+                    }
+
+                    PVMat = coopMatMulAdd(KMat, QMat, PVMat);
                 }
 
-                SfMat = coopMatMulAdd(KMat, QMat, SfMat);
+                // Store PVMat to pvsh and load into Of
+                coopMatStore(PVMat, pvsh, o_offset, osh_stride, gl_CooperativeMatrixLayoutRowMajor);
             }
 
-            // Store SfMat to sfsh and load into Of
-            const uint osh_stride = row_split * MatBc / 4;
-            const uint o_offset = gl_SubgroupID * MatBc / 4;
-            coopMatStore(SfMat, sfsh, o_offset, osh_stride, gl_CooperativeMatrixLayoutRowMajor);
-
             barrier();
 
             const uint hsv_per_tile = row_split * MatBc;
@@ -484,7 +513,7 @@ void main() {
 
                     if (hsv_col >= hsv_base && hsv_col < hsv_base + hsv_per_tile && hsv_col < HSV) {
                         const uint local_hsv = (hsv_col - hsv_base) / 4;
-                        Of[r][d_local] += ACC_TYPEV4(sfsh[row * osh_stride + local_hsv]);
+                        Of[r][d_local] += pvsh[row * osh_stride + local_hsv];
                     }
                 }
             }
@@ -500,27 +529,48 @@ void main() {
     // If there is split_k, then the split_k resolve shader does the final
     // division by L. Store the intermediate O value and per-row m and L values.
     if (p.k_num > 1) {
-        // note: O and Q have swapped coord 1,2.
-        uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
+        if (p.gqa_ratio > 1) {
+            // note: O and Q have swapped coord 1,2.
+            uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3)) / 4;
 
-        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
-            if (tile_row(r) < N) {
-                [[unroll]] for (uint32_t d0 = 0; d0 < HSV / 4; d0 += threads_per_rowgroup) {
-                    const uint d = d0 + col_tid;
-                    if (d >= HSV/4) break;
-                    const uint d_local = d0 / threads_per_rowgroup;
-                    [[unroll]] for (uint32_t comp = 0; comp < 4; ++comp) {
-                        perElemOpGqaStore(tile_row(r), 4 * d + comp, float(Of[r][d_local][comp]), o_offset, iq2, N);
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                if (tile_row(r) < N) {
+                    [[unroll]] for (uint32_t d0 = 0; d0 < HSV / 4; d0 += threads_per_rowgroup) {
+                        const uint d = d0 + col_tid;
+                        if (d >= HSV/4) break;
+                        const uint d_local = d0 / threads_per_rowgroup;
+                        gqaStore(tile_row(r), d, Of[r][d_local], o_offset, iq2, N);
                     }
                 }
             }
-        }
 
-        o_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
-        [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
-            if (tile_row(r) < N) {
-                perElemOpStoreCol0(tile_row(r), 0u, ACC_TYPE(Lf[r]), o_offset, iq2, N);
-                perElemOpStoreCol0(tile_row(r), 0u, ACC_TYPE(Mf[r]), o_offset + p.ne1, iq2, N);
+            o_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                if (tile_row(r) < N) {
+                    perElemOpStoreCol0(tile_row(r), 0u, ACC_TYPE(Lf[r]), o_offset, iq2, N);
+                    perElemOpStoreCol0(tile_row(r), 0u, ACC_TYPE(Mf[r]), o_offset + p.ne1, iq2, N);
+                }
+            }
+        } else {
+            [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
+                const uint row = tile_row(r);
+                const uint global_row = i * Br + row;
+
+                if (global_row < N) {
+                    uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (global_row + p.ne2 * iq3)) / 4;
+
+                    [[unroll]] for (uint32_t d0 = 0; d0 < HSV / 4; d0 += threads_per_rowgroup) {
+                        const uint d = d0 + col_tid;
+                        if (d >= HSV/4) break;
+                        data_ov4[o_offset + iq2 * HSV/4 + d] = D_TYPEV4(Of[r][d/threads_per_rowgroup]);
+                    }
+                }
+
+                if (global_row < N && col_tid == 0) {
+                    uint32_t lm_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (global_row + p.ne2 * iq3));
+                    data_o[lm_offset + iq2] = D_TYPE(Lf[r]);
+                    data_o[lm_offset + p.ne1 + iq2] = D_TYPE(Mf[r]);
+                }
             }
         }
 
@@ -539,7 +589,7 @@ void main() {
 
                 [[unroll]] for (uint32_t d0 = 0; d0 < HSV / 4; d0 += threads_per_rowgroup) {
                     const uint d_local = d0 / threads_per_rowgroup;
-                    Of[r][d_local] *= ACC_TYPE(ms);
+                    Of[r][d_local] *= float16_t(ms);
                 }
             } else {
                 vs = exp(sink - Mf[r]);
@@ -557,14 +607,14 @@ void main() {
     [[unroll]] for (uint32_t d0 = 0; d0 < HSV / 4; d0 += threads_per_rowgroup) {
         const uint d_local = d0 / threads_per_rowgroup;
         [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
-            Of[r][d_local] *= ACC_TYPE(Lfrcp[r]);
-#if defined(ACC_TYPE_MAX)
-            Of[r][d_local] = clamp(Of[r][d_local], -ACC_TYPE_MAX, ACC_TYPE_MAX);
+            Of[r][d_local] *= float16_t(Lfrcp[r]);
+#if defined(FLOAT_TYPE_MAX)
+            Of[r][d_local] = clamp(Of[r][d_local], -FLOAT_TYPE_MAX, FLOAT_TYPE_MAX);
 #endif
         }
     }
 
-    uint32_t o_offset = gqa_iq1*p.ne1*HSV + iq3*p.ne2*p.ne1*HSV;
+    uint32_t o_offset = (gqa_iq1*p.ne1*HSV + iq3*p.ne2*p.ne1*HSV) / 4;
 
     if (p.gqa_ratio > 1) {
         [[unroll]] for (uint32_t r = 0; r < rows_per_thread; ++r) {
@@ -573,9 +623,7 @@ void main() {
                     const uint d = d0 + col_tid;
                     if (d >= HSV / 4) break;
                     const uint d_local = d0 / threads_per_rowgroup;
-                    [[unroll]] for (uint32_t comp = 0; comp < 4; ++comp) {
-                        perElemOpGqaStore(tile_row(r), 4 * d + comp, float(Of[r][d_local][comp]), o_offset, iq2, N);
-                    }
+                    gqaStore(tile_row(r), d, Of[r][d_local], o_offset, iq2, N);
                 }
             }
         }
@@ -586,9 +634,7 @@ void main() {
                     const uint d = d0 + col_tid;
                     if (d >= HSV / 4) break;
                     const uint d_local = d0 / threads_per_rowgroup;
-                    [[unroll]] for (uint32_t comp = 0; comp < 4; ++comp) {
-                        data_o[o_offset + iq2 * HSV + (i * Br + tile_row(r)) * p.ne1 * HSV + 4 * d + comp] = D_TYPE(Of[r][d_local][comp]);
-                    }
+                    data_ov4[o_offset + (iq2 * HSV + (i * Br + tile_row(r)) * p.ne1 * HSV) / 4 + d] = D_TYPEV4(Of[r][d_local]);
                 }
             }
         }

ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp

@@ -72,6 +72,28 @@ D_TYPE perElemOpGqaStore(const in uint32_t r, const in uint32_t c, const in D_TY
     return elem;
 }
 
+// Store O values for non-GQA split_k. Rows are tokens, not heads.
+D_TYPE perElemOpNonGqaSplitKStore(const in uint32_t r, const in uint32_t c, const in D_TYPE elem, const in uint32_t unused, const in uint32_t iq2, const in uint32_t N) {
+    uint32_t global_row = i * Br + r;
+    if (global_row < N && c < HSV) {
+        uint32_t o_off = HSV * p.ne1
+            * (split_k_index + p.k_num * (global_row + p.ne2 * iq3));
+        data_o[o_off + iq2 * HSV + c] = D_TYPE(elem);
+    }
+    return elem;
+}
+
+// Store L/M values for non-GQA split_k.
+ACC_TYPE perElemOpNonGqaSplitKStoreCol0(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t lm_base, const in uint32_t iq2, const in uint32_t N) {
+    uint32_t global_row = i * Br + r;
+    if (global_row < N && c == 0) {
+        uint32_t lm_off = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3
+            + p.ne1 * 2 * (split_k_index + p.k_num * (global_row + p.ne2 * iq3));
+        data_o[lm_off + lm_base + iq2] = D_TYPE(elem);
+    }
+    return elem;
+}
+
 void main() {
 #ifdef NEEDS_INIT_IQ_SHMEM
     init_iq_shmem(gl_WorkGroupSize);
@@ -290,13 +312,19 @@ void main() {
     if (p.k_num > 1) {
         coopmat<D_TYPE, gl_ScopeWorkgroup, Br, HSV_pad, gl_MatrixUseAccumulator> O_D = coopmat<D_TYPE, gl_ScopeWorkgroup, Br, HSV_pad, gl_MatrixUseAccumulator>(O);
 
-        // note: O and Q have swapped coord 1,2.
-        uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
-        coopMatPerElementNV(O_D, O_D, perElemOpGqaStore, o_offset, iq2, N);
+        if (p.gqa_ratio > 1) {
+            // note: O and Q have swapped coord 1,2.
+            uint32_t o_offset = HSV * p.ne1 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
+            coopMatPerElementNV(O_D, O_D, perElemOpGqaStore, o_offset, iq2, N);
 
-        o_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
-        coopMatPerElementNV(L, L, perElemOpStoreCol0, o_offset, iq2, N);
-        coopMatPerElementNV(M, M, perElemOpStoreCol0, o_offset + p.ne1, iq2, N);
+            o_offset = HSV * p.ne1 * p.k_num * p.ne2 * p.ne3 + p.ne1 * 2 * (split_k_index + p.k_num * (gqa_iq1 + p.ne2 * iq3));
+            coopMatPerElementNV(L, L, perElemOpStoreCol0, o_offset, iq2, N);
+            coopMatPerElementNV(M, M, perElemOpStoreCol0, o_offset + p.ne1, iq2, N);
+        } else {
+            coopMatPerElementNV(O_D, O_D, perElemOpNonGqaSplitKStore, 0u, iq2, N);
+            coopMatPerElementNV(L, L, perElemOpNonGqaSplitKStoreCol0, 0u, iq2, N);
+            coopMatPerElementNV(M, M, perElemOpNonGqaSplitKStoreCol0, p.ne1, iq2, N);
+        }
         return;
     }
 

ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_cm2.comp

@@ -167,7 +167,9 @@ void load_row_ids(uint expert_idx, bool nei0_is_pow2, uint ic) {
         uint id = ids[iter++];
         uvec4 ballot = subgroupBallot(in_range && id == expert_idx);
 
-        ballots_sh[gl_SubgroupID] = ballot;
+        if (gl_SubgroupInvocationID == 0) {
+            ballots_sh[gl_SubgroupID] = ballot;
+        }
         barrier();
 
         uint subgroup_base = 0;

ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_id_funcs.glsl

@@ -43,7 +43,9 @@ void load_row_ids(uint expert_idx, bool nei0_is_pow2, uint ic) {
         uint id = ids[iter++];
         uvec4 ballot = subgroupBallot(in_range && id == expert_idx);
 
-        ballots_sh[gl_SubgroupID] = ballot;
+        if (gl_SubgroupInvocationID == 0) {
+            ballots_sh[gl_SubgroupID] = ballot;
+        }
         barrier();
 
         uint subgroup_base = 0;

ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq.comp

@@ -57,6 +57,8 @@ layout (push_constant) uniform parameter
     uint nbi1;
     uint ne11;
 #else
+    uint base_work_group_z;
+    uint num_batches;
     uint k_split;
     uint ne02;
     uint ne12;
@@ -108,7 +110,7 @@ void main() {
     const uint ic = gl_WorkGroupID.y;
 
 #ifdef MUL_MAT_ID
-    const uint expert_idx = gl_GlobalInvocationID.z;
+    const uint expert_idx = gl_WorkGroupID.z;
     if (ic * BN >= data_expert_count[expert_idx]) {
         return;
     }
@@ -118,7 +120,7 @@ void main() {
 #endif
 
 #ifndef MUL_MAT_ID
-    const uint batch_idx = gl_GlobalInvocationID.z;
+    const uint batch_idx = gl_WorkGroupID.z + p.base_work_group_z;
 
     const uint i13 = batch_idx / p.ne12;
     const uint i12 = batch_idx % p.ne12;
@@ -276,7 +278,7 @@ void main() {
     const uint dc = ic * BN + warp_c * WN;
 
 #ifndef MUL_MAT_ID
-    const uint offsets = batch_idx * p.batch_stride_d + ik * p.batch_stride_d * gl_NumWorkGroups.z;
+    const uint offsets = batch_idx * p.batch_stride_d + ik * p.batch_stride_d * p.num_batches;
 #endif
 
     [[unroll]] for (uint wsic = 0; wsic < WNITER; wsic++) {

ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp

@@ -595,8 +595,6 @@ void matmul_shaders(bool fp16, MatMulIdType matmul_id_type, bool coopmat, bool c
 }
 
 void process_shaders() {
-    std::map<std::string, std::string> base_dict = {{"FLOAT_TYPE", "float"}, {"FLOAT_TYPE_VEC2", "vec2"}};
-
     // matmul
     for (const MatMulIdType& matmul_id_type : {MatMulIdType::NONE, MatMulIdType::DEFAULT, MatMulIdType::SUBGROUP}) {
         // No coopmats
@@ -622,49 +620,63 @@ void process_shaders() {
         }
     }
 
-    // flash attention
-    for (const auto& f16acc : {false, true}) {
-        std::map<std::string, std::string> fa_base_dict = base_dict;
-        fa_base_dict["ACC_TYPE"] = f16acc ? "float16_t" : "float";
-        fa_base_dict["ACC_TYPEV4"] = f16acc ? "f16vec4" : "vec4";
-        if (f16acc) {
-            fa_base_dict["ACC_TYPE_MAX"] = "float16_t(65504.0)";
+    for (const bool& fp16 : {false, true}) {
+        std::map<std::string, std::string> base_dict;
+        if (fp16) {
+            base_dict = {{"FLOAT_TYPE", "float16_t"}, {"FLOAT_TYPEV4", "f16vec4"}, {"FLOAT16", "1"}, {"FLOAT_TYPE_MAX", "float16_t(65504.0)"}};
+        } else {
+            base_dict = {{"FLOAT_TYPE", "float"}, {"FLOAT_TYPEV4", "vec4"}};
         }
 
-        for (const auto& tname : type_names) {
-            if (tname == "bf16") continue;
-
-#if defined(GGML_VULKAN_COOPMAT2_GLSLC_SUPPORT)
-            if (tname == "f16") {
-                string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm2.comp",
-                    merge_maps(fa_base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}}), true, false, true, f16acc);
-            } else {
-                std::string data_a_key = "DATA_A_" + to_uppercase(tname);
-                string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm2.comp",
-                    merge_maps(fa_base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"DEQUANTFUNC", "dequantFunc"+to_uppercase(tname) }, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname) }}), true, false, true, f16acc);
+        // flash attention
+        for (const bool& f16acc : {false, true}) {
+            std::map<std::string, std::string> fa_base_dict = base_dict;
+            fa_base_dict["ACC_TYPE"] = fp16 && f16acc ? "float16_t" : "float";
+            fa_base_dict["ACC_TYPEV4"] = fp16 && f16acc ? "f16vec4" : "vec4";
+            if (fp16 && f16acc) {
+                fa_base_dict["ACC_TYPE_MAX"] = "float16_t(65504.0)";
             }
+
+            for (const auto& tname : type_names) {
+                if (tname == "bf16") continue;
+
+                if (fp16) {
+#if defined(GGML_VULKAN_COOPMAT2_GLSLC_SUPPORT)
+                if (tname == "f16") {
+                    string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm2.comp",
+                        merge_maps(fa_base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"D_TYPEV4", "vec4"}}), fp16, false, true, f16acc);
+                } else {
+                    std::string data_a_key = "DATA_A_" + to_uppercase(tname);
+                    string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm2.comp",
+                        merge_maps(fa_base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"D_TYPEV4", "vec4"}, {"DEQUANTFUNC", "dequantFunc"+to_uppercase(tname) }, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname) }}), fp16, false, true, f16acc);
+                }
 #endif
 #if defined(GGML_VULKAN_COOPMAT_GLSLC_SUPPORT)
-            if (tname == "f16") {
-                string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm1.comp",
-                    merge_maps(fa_base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"COOPMAT", "1"}}), true, true, false, f16acc);
-            } else if (tname == "q4_0" || tname == "q8_0" || tname == "f32") {
-                std::string data_a_key = "DATA_A_" + to_uppercase(tname);
-                string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm1.comp",
-                    merge_maps(fa_base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname)}, {"COOPMAT", "1"}}), true, true, false, f16acc);
-            }
+                if (tname == "f16") {
+                    string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm1.comp",
+                        merge_maps(fa_base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"D_TYPEV4", "vec4"}, {"COOPMAT", "1"}}), fp16, true, false, f16acc);
+                } else if (tname == "q4_0" || tname == "q8_0" || tname == "f32") {
+                    std::string data_a_key = "DATA_A_" + to_uppercase(tname);
+                    string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm1.comp",
+                        merge_maps(fa_base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"D_TYPEV4", "vec4"}, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname)}, {"COOPMAT", "1"}}), fp16, true, false, f16acc);
+                }
 #endif
-            if (tname == "f16") {
-                string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn.comp",
-                    merge_maps(fa_base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}}), true, false, false, f16acc);
-            } else if (tname == "q4_0" || tname == "q8_0" || tname == "f32") {
-                std::string data_a_key = "DATA_A_" + to_uppercase(tname);
-                string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn.comp",
-                    merge_maps(fa_base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname) }}), true, false, false, f16acc);
+                }
+
+                if (tname == "f16") {
+                    string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn.comp",
+                        merge_maps(fa_base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"D_TYPEV4", "vec4"}}), fp16, false, false, f16acc);
+                } else if (tname == "q4_0" || tname == "q8_0" || tname == "f32") {
+                    std::string data_a_key = "DATA_A_" + to_uppercase(tname);
+                    string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn.comp",
+                        merge_maps(fa_base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"D_TYPEV4", "vec4"}, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname) }}), fp16, false, false, f16acc);
+                }
             }
         }
     }
 
+    std::map<std::string, std::string> base_dict = {{"FLOAT_TYPE", "float"}, {"FLOAT_TYPE_VEC2", "vec2"}};
+
     for (const auto& tname : type_names) {
         // mul mat vec
         std::string data_a_key = "DATA_A_" + to_uppercase(tname);

ggml/src/ggml-webgpu/ggml-webgpu.cpp

@@ -2008,6 +2008,14 @@ static std::optional<webgpu_command> ggml_webgpu_encode_node(webgpu_context ctx,
             return ggml_webgpu_unary_op(ctx, src0, node);
         case GGML_OP_LOG:
             return ggml_webgpu_unary_op(ctx, src0, node);
+        case GGML_OP_SQR:
+            return ggml_webgpu_unary_op(ctx, src0, node);
+        case GGML_OP_SQRT:
+            return ggml_webgpu_unary_op(ctx, src0, node);
+        case GGML_OP_SIN:
+            return ggml_webgpu_unary_op(ctx, src0, node);
+        case GGML_OP_COS:
+            return ggml_webgpu_unary_op(ctx, src0, node);
         case GGML_OP_PAD:
             return ggml_webgpu_pad(ctx, src0, node);
         case GGML_OP_ARGMAX:
@@ -2967,6 +2975,18 @@ static bool ggml_backend_webgpu_device_supports_op(ggml_backend_dev_t dev, const
         case GGML_OP_LOG:
             supports_op = (op->type == GGML_TYPE_F32 || op->type == GGML_TYPE_F16) && (src0->type == op->type);
             break;
+        case GGML_OP_SQR:
+            supports_op = (op->type == GGML_TYPE_F32 || op->type == GGML_TYPE_F16) && (src0->type == op->type);
+            break;
+        case GGML_OP_SQRT:
+            supports_op = (op->type == GGML_TYPE_F32 || op->type == GGML_TYPE_F16) && (src0->type == op->type);
+            break;
+        case GGML_OP_SIN:
+            supports_op = (op->type == GGML_TYPE_F32 || op->type == GGML_TYPE_F16) && (src0->type == op->type);
+            break;
+        case GGML_OP_COS:
+            supports_op = (op->type == GGML_TYPE_F32 || op->type == GGML_TYPE_F16) && (src0->type == op->type);
+            break;
         case GGML_OP_PAD:
             supports_op = op->type == GGML_TYPE_F32 && src0->type == GGML_TYPE_F32;
             break;

ggml/src/ggml-webgpu/wgsl-shaders/unary.wgsl

@@ -170,6 +170,20 @@ fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
 #ifdef TRUNC
     let res = trunc(src[params.offset_src + src_idx]);
 #endif
+#ifdef SQR
+    let res = src[params.offset_src + src_idx] * src[params.offset_src + src_idx];
+#endif
+#ifdef SQRT
+    let res = sqrt(src[params.offset_src + src_idx]);
+#endif
+#ifdef SIN
+    let res_f32 = sin(f32(src[params.offset_src + src_idx]));
+    let res = TYPE(res_f32);
+#endif
+#ifdef COS
+    let res_f32 = cos(f32(src[params.offset_src + src_idx]));
+    let res = TYPE(res_f32);
+#endif
 
 #ifdef INPLACE
     src[params.offset_src + src_idx] = res;

ggml/src/ggml.c

@@ -899,8 +899,7 @@ static const struct ggml_type_traits type_traits[GGML_TYPE_COUNT] = {
 };
 
 const struct ggml_type_traits * ggml_get_type_traits(enum ggml_type type) {
-    assert(type >= 0);
-    assert(type < GGML_TYPE_COUNT);
+    GGML_ASSERT(type < GGML_TYPE_COUNT);
     return &type_traits[type];
 }
 
@@ -1266,33 +1265,27 @@ size_t ggml_nbytes_pad(const struct ggml_tensor * tensor) {
 }
 
 int64_t ggml_blck_size(enum ggml_type type) {
-    assert(type >= 0);
-    assert(type < GGML_TYPE_COUNT);
     return type_traits[type].blck_size;
 }
 
 size_t ggml_type_size(enum ggml_type type) {
-    assert(type >= 0);
-    assert(type < GGML_TYPE_COUNT);
     return type_traits[type].type_size;
 }
 
 size_t ggml_row_size(enum ggml_type type, int64_t ne) {
-    assert(type >= 0);
-    assert(type < GGML_TYPE_COUNT);
     assert(ne % ggml_blck_size(type) == 0);
     return ggml_type_size(type)*ne/ggml_blck_size(type);
 }
 
+double ggml_type_sizef(enum ggml_type type) {
+    return ((double)(type_traits[type].type_size))/type_traits[type].blck_size;
+}
+
 const char * ggml_type_name(enum ggml_type type) {
-    assert(type >= 0);
-    assert(type < GGML_TYPE_COUNT);
-    return type_traits[type].type_name;
+    return type < GGML_TYPE_COUNT ? type_traits[type].type_name : "NONE";
 }
 
 bool ggml_is_quantized(enum ggml_type type) {
-    assert(type >= 0);
-    assert(type < GGML_TYPE_COUNT);
     return type_traits[type].is_quantized;
 }
 
@@ -1636,23 +1629,11 @@ static struct ggml_object * ggml_new_object(struct ggml_context * ctx, enum ggml
     const size_t cur_end  = cur_offs + cur_size;
 
     // align to GGML_MEM_ALIGN
-    GGML_ASSERT(size <= SIZE_MAX - (GGML_MEM_ALIGN - 1));
     size_t size_needed = GGML_PAD(size, GGML_MEM_ALIGN);
 
     char * const mem_buffer = ctx->mem_buffer;
     struct ggml_object * const obj_new = (struct ggml_object *)(mem_buffer + cur_end);
 
-    // integer overflow checks
-    if (cur_end > SIZE_MAX - size_needed) {
-        GGML_LOG_WARN("%s: overflow detected in cur_end (%zu) + size_needed (%zu)\n", __func__, cur_end, size_needed);
-        return NULL;
-    }
-    if (cur_end + size_needed > SIZE_MAX - GGML_OBJECT_SIZE) {
-        GGML_LOG_WARN("%s: overflow detected in cur_end (%zu) + size_needed (%zu) + GGML_OBJECT_SIZE (%zu)\n", __func__,
-                cur_end, size_needed, (size_t) GGML_OBJECT_SIZE);
-        return NULL;
-    }
-
     if (cur_end + size_needed + GGML_OBJECT_SIZE > ctx->mem_size) {
         GGML_LOG_WARN("%s: not enough space in the context's memory pool (needed %zu, available %zu)\n",
                 __func__, cur_end + size_needed + GGML_OBJECT_SIZE, ctx->mem_size);
@@ -1721,8 +1702,6 @@ static struct ggml_tensor * ggml_new_tensor_impl(
         obj_alloc_size = data_size;
     }
 
-    GGML_ASSERT(GGML_TENSOR_SIZE <= SIZE_MAX - obj_alloc_size);
-
     struct ggml_object * const obj_new = ggml_new_object(ctx, GGML_OBJECT_TYPE_TENSOR, GGML_TENSOR_SIZE + obj_alloc_size);
     GGML_ASSERT(obj_new);
 

ggml/src/gguf.cpp

@@ -15,9 +15,6 @@
 #include <string>
 #include <vector>
 
-#define GGUF_MAX_STRING_LENGTH  (1024*1024*1024)
-#define GGUF_MAX_ARRAY_ELEMENTS (1024*1024*1024)
-
 template <typename T>
 struct type_to_gguf_type;
 
@@ -231,26 +228,6 @@ struct gguf_reader {
 
     template <typename T>
     bool read(std::vector<T> & dst, const size_t n) const {
-        if (n > GGUF_MAX_ARRAY_ELEMENTS) {
-            return false;
-        }
-        const uint64_t nbytes = nbytes_remain();
-        if constexpr (std::is_same<T, std::string>::value) {
-            // strings are prefixed with their length, so we need to account for that
-            if (n > SIZE_MAX / sizeof(uint64_t)) {
-                return false;
-            }
-            if (nbytes < n * sizeof(uint64_t)) {
-                return false;
-            }
-        } else {
-            if (n > SIZE_MAX / sizeof(T)) {
-                return false;
-            }
-            if (nbytes < n * sizeof(T)) {
-                return false;
-            }
-        }
         dst.resize(n);
         for (size_t i = 0; i < dst.size(); ++i) {
             if constexpr (std::is_same<T, bool>::value) {
@@ -300,43 +277,13 @@ struct gguf_reader {
         if (!read(size)) {
             return false;
         }
-        if (size > GGUF_MAX_STRING_LENGTH) {
-            GGML_LOG_ERROR("%s: string length %" PRIu64 " exceeds maximum %" PRIu64 "\n", __func__, size, (uint64_t) GGUF_MAX_STRING_LENGTH);
-            return false;
-        }
-        const uint64_t nbytes = nbytes_remain();
-        if (size > nbytes) {
-            GGML_LOG_ERROR("%s: string length %" PRIu64 " exceeds remaining file size %" PRIu64 " bytes\n", __func__, size, nbytes);
-            return false;
-        }
-        dst.resize(static_cast<size_t>(size));
+        dst.resize(size);
         return fread(dst.data(), 1, dst.length(), file) == dst.length();
     }
 
     bool read(void * dst, const size_t size) const {
         return fread(dst, 1, size, file) == size;
     }
-
-    // remaining bytes in the file
-    uint64_t nbytes_remain() const {
-        const long cur = ftell(file);
-        if (cur < 0) {
-            return 0;
-        }
-        if (fseek(file, 0, SEEK_END) != 0) {
-            fseek(file, cur, SEEK_SET);
-
-            return 0;
-        }
-        const long end = ftell(file);
-        if (end < 0) {
-            fseek(file, cur, SEEK_SET);
-
-            return 0;
-        }
-        fseek(file, cur, SEEK_SET);
-        return static_cast<uint64_t>(end - cur);
-    }
 };
 
 struct gguf_context * gguf_init_empty(void) {
@@ -621,8 +568,8 @@ struct gguf_context * gguf_init_from_file_impl(FILE * file, struct gguf_init_par
 
             // check that tensor type is within defined range
             if (info.t.type < 0 || info.t.type >= GGML_TYPE_COUNT) {
-                GGML_LOG_ERROR("%s: tensor '%s' has invalid ggml type %d. should be in [0, %d)\n",
-                    __func__, info.t.name, info.t.type, GGML_TYPE_COUNT);
+                GGML_LOG_ERROR("%s: tensor '%s' has invalid ggml type %d (%s)\n",
+                    __func__, info.t.name, info.t.type, ggml_type_name(info.t.type));
                 ok = false;
                 break;
             }
@@ -710,34 +657,10 @@ struct gguf_context * gguf_init_from_file_impl(FILE * file, struct gguf_init_par
         //   the ggml_tensor structs to the appropriate locations in the binary blob
 
         // compute the exact size needed for the new ggml_context
-        size_t mem_size = 0;
-        if (params.no_alloc) {
-            if (n_tensors != 0 && SIZE_MAX / n_tensors < ggml_tensor_overhead()) {
-                GGML_LOG_ERROR("%s: memory size overflow while allocating ggml context\n", __func__);
-                gguf_free(ctx);
-                return nullptr;
-            }
-
-            const size_t overhead = n_tensors * ggml_tensor_overhead();
-
-            mem_size = overhead;
-        } else {
-            if ((n_tensors + 1) != 0 && SIZE_MAX / (n_tensors + 1) < ggml_tensor_overhead()) {
-                GGML_LOG_ERROR("%s: memory size overflow while allocating ggml context\n", __func__);
-                gguf_free(ctx);
-                return nullptr;
-            }
-
-            const size_t overhead = (n_tensors + 1) * ggml_tensor_overhead();
-
-            if (SIZE_MAX - overhead < ctx->size) {
-                GGML_LOG_ERROR("%s: memory size overflow while allocating ggml context\n", __func__);
-                gguf_free(ctx);
-                return nullptr;
-            }
-
-            mem_size = overhead + ctx->size;
-        }
+        const size_t mem_size =
+            params.no_alloc ?
+            (n_tensors    )*ggml_tensor_overhead() :
+            (n_tensors + 1)*ggml_tensor_overhead() + ctx->size;
 
         struct ggml_init_params pdata = {
             /*mem_size   =*/ mem_size,

gguf-py/gguf/constants.py

@@ -473,6 +473,7 @@ class MODEL_ARCH(IntEnum):
     RND1             = auto()
     PANGU_EMBED      = auto()
     MISTRAL3         = auto()
+    PADDLEOCR        = auto()
     MIMO2            = auto()
     STEP35           = auto()
     LLAMA_EMBED      = auto()
@@ -914,6 +915,7 @@ MODEL_ARCH_NAMES: dict[MODEL_ARCH, str] = {
     MODEL_ARCH.RND1:             "rnd1",
     MODEL_ARCH.PANGU_EMBED:      "pangu-embedded",
     MODEL_ARCH.MISTRAL3:         "mistral3",
+    MODEL_ARCH.PADDLEOCR:        "paddleocr",
     MODEL_ARCH.MIMO2:            "mimo2",
     MODEL_ARCH.STEP35:           "step35",
     MODEL_ARCH.LLAMA_EMBED:      "llama-embed",
@@ -3186,6 +3188,20 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
         MODEL_TENSOR.FFN_DOWN,
         MODEL_TENSOR.FFN_UP,
     ],
+    MODEL_ARCH.PADDLEOCR: [
+        MODEL_TENSOR.TOKEN_EMBD,
+        MODEL_TENSOR.OUTPUT_NORM,
+        MODEL_TENSOR.OUTPUT,
+        MODEL_TENSOR.ATTN_NORM,
+        MODEL_TENSOR.ATTN_Q,
+        MODEL_TENSOR.ATTN_K,
+        MODEL_TENSOR.ATTN_V,
+        MODEL_TENSOR.ATTN_OUT,
+        MODEL_TENSOR.FFN_NORM,
+        MODEL_TENSOR.FFN_GATE,
+        MODEL_TENSOR.FFN_DOWN,
+        MODEL_TENSOR.FFN_UP,
+    ],
     MODEL_ARCH.FALCON_H1: [
         # Token embedding
         MODEL_TENSOR.TOKEN_EMBD,
@@ -3847,6 +3863,7 @@ class VisionProjectorType:
     VOXTRAL = "voxtral"
     LFM2 = "lfm2"
     KIMIVL = "kimivl"
+    PADDLEOCR = "paddleocr"
     KIMIK25 = "kimik25"
     LIGHTONOCR = "lightonocr"
     COGVLM = "cogvlm"

gguf-py/gguf/gguf_reader.py

@@ -175,9 +175,6 @@ class GGUFReader:
             if new_align.types != [GGUFValueType.UINT32]:
                 raise ValueError('Bad type for general.alignment field')
             self.alignment = new_align.parts[-1][0]
-            # Ensure alignment is a non-zero power of two
-            if self.alignment == 0 or (self.alignment & (self.alignment - 1)) != 0:
-                raise ValueError('Invalid alignment: must be a non-zero power of two')
         padding = offs % self.alignment
         if padding != 0:
             offs += self.alignment - padding
@@ -205,11 +202,11 @@ class GGUFReader:
 
     def _push_field(self, field: ReaderField, skip_sum: bool = False) -> int:
         if field.name in self.fields:
-            # TODO: add option to make this a warning and accept duplicate keys like below
-            raise KeyError(f'Duplicate {field.name} already in list at offset {field.offset}')
+            # TODO: add option to generate error on duplicate keys
+            # raise KeyError(f'Duplicate {field.name} already in list at offset {field.offset}')
 
-            # logger.warning(f'Duplicate key {field.name} at offset {field.offset}')
-            # self.fields[field.name + '_{}'.format(field.offset)] = field
+            logger.warning(f'Duplicate key {field.name} at offset {field.offset}')
+            self.fields[field.name + '_{}'.format(field.offset)] = field
         else:
             self.fields[field.name] = field
         return 0 if skip_sum else sum(int(part.nbytes) for part in field.parts)

gguf-py/gguf/gguf_writer.py

@@ -501,8 +501,6 @@ class GGUFWriter:
         self.add_uint32(Keys.General.QUANTIZATION_VERSION, quantization_version)
 
     def add_custom_alignment(self, alignment: int) -> None:
-        if alignment <= 0 or (alignment & (alignment - 1)) != 0:
-            raise ValueError('Invalid alignment: must be a non-zero power of two')
         self.data_alignment = alignment
         self.add_uint32(Keys.General.ALIGNMENT, alignment)
 

gguf-py/gguf/tensor_mapping.py

@@ -1325,6 +1325,7 @@ class TensorNameMap:
             "multi_modal_projector.linear_{bid}",
             "mm_projector.proj.linear_{bid}", # Kimi-K2.5
             "visual.merger.mlp.{bid}", # qwen2vl
+            "mlp_AR.linear_{bid}", # PaddleOCR-VL
             "merger.mlp.{bid}",
         ),
 
@@ -1574,6 +1575,7 @@ class TensorNameMap:
             "mm_projector.pre_norm", # Kimi-K2.5
             "pre_mm_projector_norm",
             "model.vision.linear_proj.norm1", # cogvlm
+            "mlp_AR.pre_norm", # PaddleOCR-VL
             "merger.ln_q",
         ),
 
@@ -1599,6 +1601,7 @@ class TensorNameMap:
 
         MODEL_TENSOR.V_RESMPL_ATTN_OUT: (
             "resampler.attn.out_proj",
+            "model.vision_model.head.attention.out_proj",
         ),
 
         MODEL_TENSOR.V_RESMPL_KV: (

include/llama.h

@@ -389,6 +389,7 @@ extern "C" {
         bool only_copy;                       // only copy tensors - ftype, allow_requantize and quantize_output_tensor are ignored
         bool pure;                            // quantize all tensors to the default type
         bool keep_split;                      // quantize to the same number of shards
+        bool dry_run;                         // calculate and show the final quantization size without performing quantization
         void * imatrix;                       // pointer to importance matrix data
         void * kv_overrides;                  // pointer to vector containing overrides
         void * tensor_types;                  // pointer to vector containing tensor types

models/templates/stepfun-ai-Step-3.5-Flash.jinja

@@ -0,0 +1,80 @@
+{% macro render_content(content) %}{% if content is none %}{{- '' }}{% elif content is string %}{{- content }}{% elif content is mapping %}{{- content['value'] if 'value' in content else content['text'] }}{% elif content is iterable %}{% for item in content %}{% if item.type == 'text' %}{{- item['value'] if 'value' in item else item['text'] }}{% elif item.type == 'image' %}<im_patch>{% endif %}{% endfor %}{% endif %}{% endmacro %}
+{{bos_token}}{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- render_content(messages[0].content) + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou have access to the following functions in JSONSchema format:\n\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson(ensure_ascii=False) }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...>\n...\n</function> block must be nested within <tool_call>\n...\n</tool_call> XML tags\n- Required parameters MUST be specified\n</IMPORTANT><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + render_content(messages[0].content) + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and render_content(message.content) is string and not(render_content(message.content).startswith('<tool_response>') and render_content(message.content).endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- set content = render_content(message.content) %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {%- set role_name = 'observation' if (message.role == "system" and not loop.first and message.name == 'observation') else message.role %}
+        {{- '<|im_start|>' + role_name + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = render_content(message.reasoning_content) %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- else %}
+                {%- set reasoning_content = '' %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n' + content }}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if tool_call.function is defined %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                {%- if tool_call.arguments is defined %}
+                    {%- set arguments = tool_call.arguments %}
+                    {%- for args_name, args_value in arguments|items %}
+                        {{- '<parameter=' + args_name + '>\n' }}
+                        {%- set args_value = args_value | tojson(ensure_ascii=False) | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
+                        {{- args_value }}
+                        {{- '\n</parameter>\n' }}
+                    {%- endfor %}
+                {%- endif %}
+                {{- '</function>\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>tool_response\n' }}
+        {%- endif %}
+        {{- '<tool_response>' }}
+        {{- content }}
+        {{- '</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n<think>\n' }}
+{%- endif %}

scripts/snapdragon/adb/run-cli.sh

@@ -54,6 +54,6 @@ adb $adbserial $adbhost shell " \
     $verbose $experimental $sched $opmask $profile $nhvx $ndev $hb \
       ./$branch/bin/llama-cli --no-mmap -m $basedir/../gguf/$model \
          --poll 1000 -t 6 --cpu-mask 0xfc --cpu-strict 1           \
-         --ctx-size 8192 --batch-size 128 -fa on \
-         -ngl 99 --device $device $cli_opts $@   \
+         --ctx-size 8192 --ubatch-size 256 -fa on                  \
+         -ngl 99 --device $device $cli_opts $@                     \
 "

scripts/snapdragon/adb/run-completion.sh

@@ -54,6 +54,6 @@ adb $adbserial $adbhost shell " \
     $verbose $experimental $sched $opmask $profile $nhvx $ndev $hb        \
       ./$branch/bin/llama-completion --no-mmap -m $basedir/../gguf/$model \
          --poll 1000 -t 6 --cpu-mask 0xfc --cpu-strict 1                  \
-         --ctx-size 8192 --batch-size 128 -fa on \
-         -ngl 99 -no-cnv --device $device $cli_opts $@   \
+         --ctx-size 8192 --ubatch-size 256 -fa on                         \
+         -ngl 99 -no-cnv --device $device $cli_opts $@                    \
 "

scripts/snapdragon/adb/run-mtmd.sh

@@ -58,11 +58,11 @@ adb $adbserial $adbhost shell " \
   cd $basedir; ulimit -c unlimited;        \
     LD_LIBRARY_PATH=$basedir/$branch/lib   \
     ADSP_LIBRARY_PATH=$basedir/$branch/lib \
-    $verbose $experimental $sched $opmask $profile $nhvx $ndev $mtmd_backend       \
-      ./$branch/bin/llama-mtmd-cli --no-mmap -m $basedir/../gguf/$model   \
-         --mmproj $basedir/../gguf/$mmproj \
-         --image $basedir/../gguf/$image \
-         --poll 1000 -t 6 --cpu-mask 0xfc --cpu-strict 1             \
-         --ctx-size 8192 --batch-size 128 -ctk q8_0 -ctv q8_0 -fa on \
-         -ngl 99 --device $device -v $cli_opts $@ \
+    $verbose $experimental $sched $opmask $profile $nhvx $ndev $mtmd_backend \
+      ./$branch/bin/llama-mtmd-cli --no-mmap -m $basedir/../gguf/$model      \
+         --mmproj $basedir/../gguf/$mmproj                                   \
+         --image $basedir/../gguf/$image                                     \
+         --poll 1000 -t 6 --cpu-mask 0xfc --cpu-strict 1                     \
+         --ctx-size 8192 --ubatch-size 256 -fa on                            \
+         -ngl 99 --device $device -v $cli_opts $@                            \
 "

scripts/snapdragon/windows/run-cli.ps1

@@ -49,5 +49,5 @@ $env:ADSP_LIBRARY_PATH="$basedir\lib"
 & "$basedir\bin\llama-completion.exe" `
     --no-mmap -no-cnv -m $basedir\..\..\gguf\$model `
     --poll 1000 -t 6 --cpu-mask 0xfc --cpu-strict 1 `
-    --ctx-size 8192 --batch-size 128 -ctk q8_0 -ctv q8_0 -fa on `
+    --ctx-size 8192 --ubatch-size 128 -fa on `
     -ngl 99 --device $device $cli_opts

scripts/sync_vendor.py

@@ -5,7 +5,7 @@ import os
 import sys
 import subprocess
 
-HTTPLIB_VERSION = "d4180e923f846b44a3d30acd938438d6e64fc9f6"
+HTTPLIB_VERSION = "refs/tags/v0.34.0"
 
 vendor = {
     "https://github.com/nlohmann/json/releases/latest/download/json.hpp":     "vendor/nlohmann/json.hpp",

src/CMakeLists.txt

@@ -110,6 +110,7 @@ add_library(llama
             models/openai-moe-iswa.cpp
             models/openelm.cpp
             models/orion.cpp
+            models/paddleocr.cpp
             models/pangu-embedded.cpp
             models/phi2.cpp
             models/phi3.cpp

src/llama-arch.cpp

@@ -121,6 +121,7 @@ static const std::map<llm_arch, const char *> LLM_ARCH_NAMES = {
     { LLM_ARCH_RND1,             "rnd1"             },
     { LLM_ARCH_PANGU_EMBED,      "pangu-embedded"   },
     { LLM_ARCH_MISTRAL3,         "mistral3"         },
+    { LLM_ARCH_PADDLEOCR,        "paddleocr"        },
     { LLM_ARCH_MIMO2,            "mimo2"            },
     { LLM_ARCH_STEP35,           "step35"           },
     { LLM_ARCH_LLAMA_EMBED,      "llama-embed"      },
@@ -739,6 +740,7 @@ static std::set<llm_tensor> llm_get_tensor_names(llm_arch arch) {
         case LLM_ARCH_INTERNLM2:
         case LLM_ARCH_GRANITE:
         case LLM_ARCH_ERNIE4_5:
+        case LLM_ARCH_PADDLEOCR:
         case LLM_ARCH_SMOLLM3:
         case LLM_ARCH_DREAM:
         case LLM_ARCH_LLADA:

src/llama-arch.h

@@ -125,6 +125,7 @@ enum llm_arch {
     LLM_ARCH_RND1,
     LLM_ARCH_PANGU_EMBED,
     LLM_ARCH_MISTRAL3,
+    LLM_ARCH_PADDLEOCR,
     LLM_ARCH_MIMO2,
     LLM_ARCH_STEP35,
     LLM_ARCH_LLAMA_EMBED,

src/llama-context.cpp

@@ -2440,64 +2440,6 @@ size_t llama_context::state_write_data(llama_io_write_i & io) {
         // TODO: add more model-specific info which should prevent loading the session file if not identical
     }
 
-    // write output ids
-    {
-        LLAMA_LOG_DEBUG("%s: - writing output ids\n", __func__);
-
-        const auto n_outputs    = this->n_outputs;
-        const auto & output_ids = this->output_ids;
-
-        std::vector<int32_t> w_output_pos;
-
-        w_output_pos.resize(n_outputs);
-
-        // build a more compact representation of the output ids
-        for (size_t i = 0; i < n_batch(); ++i) {
-            // map an output id to a position in the batch
-            int64_t pos = output_ids[i];
-            if (pos >= 0) {
-                GGML_ASSERT(pos < n_outputs);
-                w_output_pos[pos] = i;
-            }
-        }
-
-        io.write(&n_outputs, sizeof(n_outputs));
-
-        if (n_outputs) {
-            io.write(w_output_pos.data(), n_outputs * sizeof(int32_t));
-        }
-    }
-
-    // [TAG_CONTEXT_STATE_LOGITS]
-    // write logits
-    {
-        LLAMA_LOG_DEBUG("%s: - writing logits\n", __func__);
-
-        const uint64_t logits_size = std::min((uint64_t) this->logits.size, (uint64_t) n_outputs * model.vocab.n_tokens());
-
-        io.write(&logits_size, sizeof(logits_size));
-
-        if (logits_size) {
-            io.write(logits.data, logits_size * sizeof(float));
-        }
-    }
-
-    // write embeddings
-    {
-        LLAMA_LOG_DEBUG("%s: - writing embeddings\n", __func__);
-
-        const uint64_t embd_size = std::min((uint64_t) this->embd.size, (uint64_t) n_outputs * model.hparams.n_embd);
-
-        io.write(&embd_size, sizeof(embd_size));
-
-        if (embd_size) {
-            io.write(embd.data, embd_size * sizeof(float));
-        }
-    }
-
-    // TODO: handle sampling buffers and samplers state ?
-    //       https://github.com/ggml-org/llama.cpp/pull/17004
-
     if (memory != nullptr) {
         LLAMA_LOG_DEBUG("%s: - writing memory module\n", __func__);
         memory->state_write(io);
@@ -2523,70 +2465,6 @@ size_t llama_context::state_read_data(llama_io_read_i & io) {
         // TODO: add more info which needs to be identical but which is not verified otherwise
     }
 
-    // read output ids
-    {
-        LLAMA_LOG_DEBUG("%s: - reading output ids\n", __func__);
-
-        auto n_outputs = this->n_outputs;
-        io.read_to(&n_outputs, sizeof(n_outputs));
-
-        if (n_outputs > output_reserve(n_outputs)) {
-            throw std::runtime_error("could not reserve outputs");
-        }
-
-        std::vector<int32_t> output_pos;
-
-        if (n_outputs) {
-            output_pos.resize(n_outputs);
-            io.read_to(output_pos.data(), n_outputs * sizeof(int32_t));
-
-            for (int32_t i = 0; i < (int32_t) output_pos.size(); ++i) {
-                int32_t id = output_pos[i];
-                if ((uint32_t) id >= n_batch()) {
-                    throw std::runtime_error(format("invalid output id, %d does not fit in batch size of %u", id, n_batch()));
-                }
-                this->output_ids[id] = i;
-            }
-
-            this->n_outputs = n_outputs;
-        }
-    }
-
-    // read logits
-    {
-        LLAMA_LOG_DEBUG("%s: - reading logits\n", __func__);
-
-        uint64_t logits_size;
-        io.read_to(&logits_size, sizeof(logits_size));
-
-        if (this->logits.size < logits_size) {
-            throw std::runtime_error("logits buffer too small");
-        }
-
-        if (logits_size) {
-            io.read_to(this->logits.data, logits_size * sizeof(float));
-        }
-    }
-
-    // read embeddings
-    {
-        LLAMA_LOG_DEBUG("%s: - reading embeddings\n", __func__);
-
-        uint64_t embd_size;
-        io.read_to(&embd_size, sizeof(embd_size));
-
-        if (this->embd.size < embd_size) {
-            throw std::runtime_error("embeddings buffer too small");
-        }
-
-        if (embd_size) {
-            io.read_to(this->embd.data, embd_size * sizeof(float));
-        }
-    }
-
-    // TODO: handle sampling buffers and samplers state ?
-    //       https://github.com/ggml-org/llama.cpp/pull/17004
-
     if (memory) {
         LLAMA_LOG_DEBUG("%s: - reading memory module\n", __func__);
 

src/llama-impl.cpp

@@ -109,9 +109,9 @@ std::string llama_format_tensor_shape(const std::vector<int64_t> & ne) {
 
 std::string llama_format_tensor_shape(const struct ggml_tensor * t) {
     char buf[256];
-    snprintf(buf, sizeof(buf), "%5" PRId64, t->ne[0]);
+    snprintf(buf, sizeof(buf), "%6" PRId64, t->ne[0]);
     for (int i = 1; i < GGML_MAX_DIMS; i++) {
-        snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf), ", %5" PRId64, t->ne[i]);
+        snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf), ", %6" PRId64, t->ne[i]);
     }
     return buf;
 }

src/llama-model.cpp

@@ -123,6 +123,7 @@ const char * llm_type_name(llm_type type) {
         case LLM_TYPE_8B_A1B:        return "8B.A1B";
         case LLM_TYPE_16B_A1B:       return "16B.A1B";
         case LLM_TYPE_21B_A3B:       return "21B.A3B";
+        case LLM_TYPE_24B_A2B:       return "24B.A2B";
         case LLM_TYPE_30B_A3B:       return "30B.A3B";
         case LLM_TYPE_31B_A3_5B:     return "31B.A3.5B";
         case LLM_TYPE_35B_A3B:       return "35B.A3B";
@@ -1703,8 +1704,8 @@ void llama_model::load_hparams(llama_model_loader & ml) {
             } break;
         case LLM_ARCH_DEEPSEEK2:
             {
-                // lite variants include DeepSeek-V2-Lite, GigaChat3-10B-A1.8B
-                const bool is_lite = (hparams.n_layer == 27 || hparams.n_layer == 26);
+                // lite variants include DeepSeek-V2-Lite, GigaChat3-10B-A1.8B, Kanana-2-30B-A3B
+                const bool is_lite = (hparams.n_layer == 27 || hparams.n_layer == 26 || (hparams.n_layer == 48 && n_vocab == 128256));
 
                 ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
                 ml.get_key(LLM_KV_LEADING_DENSE_BLOCK_COUNT,   hparams.n_layer_dense_lead);
@@ -2244,7 +2245,11 @@ void llama_model::load_hparams(llama_model_loader & ml) {
             } break;
         case LLM_ARCH_ERNIE4_5:
         case LLM_ARCH_ERNIE4_5_MOE:
+        case LLM_ARCH_PADDLEOCR:
             {
+                // paddleocr need mrope_section
+                ml.get_key_or_arr(LLM_KV_ROPE_DIMENSION_SECTIONS, hparams.rope_sections, 4, false);
+
                 ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
                 if (arch == LLM_ARCH_ERNIE4_5_MOE) {
                     ml.get_key(LLM_KV_EXPERT_FEED_FORWARD_LENGTH,        hparams.n_ff_exp);
@@ -2377,7 +2382,11 @@ void llama_model::load_hparams(llama_model_loader & ml) {
                     hparams.recurrent_layer_arr[il] = hparams.n_head_kv(il) == 0;
                 }
 
-                type = LLM_TYPE_8B_A1B;
+                switch (hparams.n_layer) {
+                    case 24: type = LLM_TYPE_8B_A1B;  break;
+                    case 40: type = LLM_TYPE_24B_A2B; break;
+                    default: type = LLM_TYPE_UNKNOWN;
+                }
             } break;
         case LLM_ARCH_SMALLTHINKER:
             {
@@ -6631,6 +6640,7 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
                 } break;
             case LLM_ARCH_ERNIE4_5:
             case LLM_ARCH_ERNIE4_5_MOE:
+            case LLM_ARCH_PADDLEOCR:
                 {
                     tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, 0);
 
@@ -8709,6 +8719,10 @@ ggml_cgraph * llama_model::build_graph(const llm_graph_params & params) const {
             {
                 llm = std::make_unique<llm_build_ernie4_5_moe>(*this, params);
             } break;
+        case LLM_ARCH_PADDLEOCR:
+            {
+                llm = std::make_unique<llm_build_paddleocr>(*this, params);
+            } break;
         case LLM_ARCH_HUNYUAN_MOE:
             {
                 llm = std::make_unique<llm_build_hunyuan_moe>(*this, params);
@@ -9045,6 +9059,7 @@ llama_rope_type llama_model_rope_type(const llama_model * model) {
             return LLAMA_ROPE_TYPE_NEOX;
 
         case LLM_ARCH_QWEN2VL:
+        case LLM_ARCH_PADDLEOCR:
             return LLAMA_ROPE_TYPE_MROPE;
         case LLM_ARCH_QWEN3VL:
         case LLM_ARCH_QWEN3VLMOE:

src/llama-model.h

@@ -116,6 +116,7 @@ enum llm_type {
     LLM_TYPE_8B_A1B, // lfm2moe
     LLM_TYPE_16B_A1B,
     LLM_TYPE_21B_A3B, // Ernie MoE small
+    LLM_TYPE_24B_A2B, // lfm2moe
     LLM_TYPE_30B_A3B,
     LLM_TYPE_31B_A3_5B,
     LLM_TYPE_35B_A3B, // Qwen3.5

src/llama-quant.cpp

@@ -479,6 +479,17 @@ static size_t llama_tensor_quantize_impl(enum ggml_type new_type, const float *
     return new_size;
 }
 
+static bool tensor_type_requires_imatrix(const ggml_tensor * t, const ggml_type dst_type, const llama_ftype ftype) {
+    return (
+        dst_type == GGML_TYPE_IQ2_XXS || dst_type == GGML_TYPE_IQ2_XS ||
+        dst_type == GGML_TYPE_IQ3_XXS || dst_type == GGML_TYPE_IQ1_S  ||
+        dst_type == GGML_TYPE_IQ2_S   || dst_type == GGML_TYPE_IQ1_M  ||
+        (   // Q2_K_S is the worst k-quant type - only allow it without imatrix for token embeddings
+            dst_type == GGML_TYPE_Q2_K && ftype == LLAMA_FTYPE_MOSTLY_Q2_K_S && strcmp(t->name, "token_embd.weight") != 0
+        )
+    );
+}
+
 static void llama_model_quantize_impl(const std::string & fname_inp, const std::string & fname_out, const llama_model_quantize_params * params) {
     ggml_type default_type;
     llama_ftype ftype = params->ftype;
@@ -735,24 +746,36 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
     };
 
     const auto tn = LLM_TN(model.arch);
-    new_ofstream(0);
+
+    // no output file for --dry-run
+    if (!params->dry_run) {
+        new_ofstream(0);
+    }
+
+    // flag for `--dry-run`, to let the user know if imatrix will be required for a real
+    // quantization, as a courtesy
+    bool will_require_imatrix = false;
+
     for (const auto * it : tensors) {
         const auto & weight = *it;
         ggml_tensor * tensor = weight.tensor;
-        if (weight.idx != cur_split && params->keep_split) {
+        if (!params->dry_run && (weight.idx != cur_split && params->keep_split)) {
             close_ofstream();
             new_ofstream(weight.idx);
         }
 
         const std::string name = ggml_get_name(tensor);
+        const size_t tensor_size = ggml_nbytes(tensor);
 
-        if (!ml.use_mmap) {
-            if (read_data.size() < ggml_nbytes(tensor)) {
-                read_data.resize(ggml_nbytes(tensor));
+        if (!params->dry_run) {
+            if (!ml.use_mmap) {
+                if (read_data.size() < tensor_size) {
+                    read_data.resize(tensor_size);
+                }
+                tensor->data = read_data.data();
             }
-            tensor->data = read_data.data();
+            ml.load_data_for(tensor);
         }
-        ml.load_data_for(tensor);
 
         LLAMA_LOG_INFO("[%4d/%4d] %36s - [%s], type = %6s, ",
                ++idx, ml.n_tensors,
@@ -900,129 +923,155 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
             quantize = tensor->type != new_type;
         }
 
-        if (!quantize) {
-            new_type = tensor->type;
-            new_data = tensor->data;
-            new_size = ggml_nbytes(tensor);
-            LLAMA_LOG_INFO("size = %8.3f MiB\n", ggml_nbytes(tensor)/1024.0/1024.0);
-        } else {
-            const int64_t nelements = ggml_nelements(tensor);
-
-            const float * imatrix = nullptr;
-            if (imatrix_data) {
-                auto it = imatrix_data->find(remap_imatrix(tensor->name, mapped));
-                if (it == imatrix_data->end()) {
-                    LLAMA_LOG_INFO("\n====== %s: did not find weights for %s\n", __func__, tensor->name);
-                } else {
-                    if (it->second.size() == (size_t)tensor->ne[0]*tensor->ne[2]) {
-                        imatrix = it->second.data();
-                    } else {
-                        LLAMA_LOG_INFO("\n====== %s: imatrix size %d is different from tensor size %d for %s\n", __func__,
-                                int(it->second.size()), int(tensor->ne[0]*tensor->ne[2]), tensor->name);
-
-                        // this can happen when quantizing an old mixtral model with split tensors with a new incompatible imatrix
-                        // this is a significant error and it may be good idea to abort the process if this happens,
-                        // since many people will miss the error and not realize that most of the model is being quantized without an imatrix
-                        // tok_embd should be ignored in this case, since it always causes this warning
-                        if (name != tn(LLM_TENSOR_TOKEN_EMBD, "weight")) {
-                            throw std::runtime_error(format("imatrix size %d is different from tensor size %d for %s",
-                                    int(it->second.size()), int(tensor->ne[0]*tensor->ne[2]), tensor->name));
-                        }
-                    }
+        // we have now decided on the target type for this tensor
+        if (params->dry_run) {
+            // the --dry-run option calculates the final quantization size without quantizting
+            if (quantize) {
+                new_size = ggml_nrows(tensor) * ggml_row_size(new_type, tensor->ne[0]);
+                LLAMA_LOG_INFO("size = %8.2f MiB -> %8.2f MiB (%s)\n",
+                               tensor_size/1024.0/1024.0,
+                               new_size/1024.0/1024.0,
+                               ggml_type_name(new_type));
+                if (!will_require_imatrix && tensor_type_requires_imatrix(tensor, new_type, params->ftype)) {
+                    will_require_imatrix = true;
                 }
-            }
-            if ((new_type == GGML_TYPE_IQ2_XXS ||
-                 new_type == GGML_TYPE_IQ2_XS  ||
-                 new_type == GGML_TYPE_IQ2_S   ||
-                 new_type == GGML_TYPE_IQ1_S   ||
-                (new_type == GGML_TYPE_IQ1_M && strcmp(tensor->name, "token_embd.weight") && strcmp(tensor->name, "output.weight"))  ||
-                (new_type == GGML_TYPE_Q2_K && params->ftype == LLAMA_FTYPE_MOSTLY_Q2_K_S && strcmp(tensor->name, "token_embd.weight") != 0)) && !imatrix) {
-                LLAMA_LOG_ERROR("\n\n============================================================\n");
-                LLAMA_LOG_ERROR("Missing importance matrix for tensor %s in a very low-bit quantization\n", tensor->name);
-                LLAMA_LOG_ERROR("The result will be garbage, so bailing out\n");
-                LLAMA_LOG_ERROR("============================================================\n\n");
-                throw std::runtime_error(format("Missing importance matrix for tensor %s in a very low-bit quantization", tensor->name));
-            }
-
-            float * f32_data;
-
-            if (tensor->type == GGML_TYPE_F32) {
-                f32_data = (float *) tensor->data;
-            } else if (ggml_is_quantized(tensor->type) && !params->allow_requantize) {
-                throw std::runtime_error(format("requantizing from type %s is disabled", ggml_type_name(tensor->type)));
             } else {
-                llama_tensor_dequantize_impl(tensor, f32_conv_buf, workers, nelements, nthread);
-                f32_data = (float *) f32_conv_buf.data();
+                new_size = tensor_size;
+                LLAMA_LOG_INFO("size = %8.3f MiB\n", new_size/1024.0/1024.0);
             }
+            total_size_org += tensor_size;
+            total_size_new += new_size;
+            continue;
+        } else {
+            // no --dry-run, perform quantization
+            if (!quantize) {
+                new_type = tensor->type;
+                new_data = tensor->data;
+                new_size = tensor_size;
+                LLAMA_LOG_INFO("size = %8.3f MiB\n", tensor_size/1024.0/1024.0);
+            } else {
+                const int64_t nelements = ggml_nelements(tensor);
 
-            LLAMA_LOG_INFO("converting to %s .. ", ggml_type_name(new_type));
-            fflush(stdout);
+                const float * imatrix = nullptr;
+                if (imatrix_data) {
+                    auto it = imatrix_data->find(remap_imatrix(tensor->name, mapped));
+                    if (it == imatrix_data->end()) {
+                        LLAMA_LOG_INFO("\n====== %s: did not find weights for %s\n", __func__, tensor->name);
+                    } else {
+                        if (it->second.size() == (size_t)tensor->ne[0]*tensor->ne[2]) {
+                            imatrix = it->second.data();
+                        } else {
+                            LLAMA_LOG_INFO("\n====== %s: imatrix size %d is different from tensor size %d for %s\n", __func__,
+                                    int(it->second.size()), int(tensor->ne[0]*tensor->ne[2]), tensor->name);
 
-            if (work.size() < (size_t)nelements * 4) {
-                work.resize(nelements * 4); // upper bound on size
-            }
-            new_data = work.data();
-
-            const int64_t n_per_row = tensor->ne[0];
-            const int64_t nrows = tensor->ne[1];
-
-            static const int64_t min_chunk_size = 32 * 512;
-            const int64_t chunk_size = (n_per_row >= min_chunk_size ? n_per_row : n_per_row * ((min_chunk_size + n_per_row - 1)/n_per_row));
-
-            const int64_t nelements_matrix = tensor->ne[0] * tensor->ne[1];
-            const int64_t nchunk = (nelements_matrix + chunk_size - 1)/chunk_size;
-            const int64_t nthread_use = nthread > 1 ? std::max((int64_t)1, std::min((int64_t)nthread, nchunk)) : 1;
-
-            // quantize each expert separately since they have different importance matrices
-            new_size = 0;
-            for (int64_t i03 = 0; i03 < tensor->ne[2]; ++i03) {
-                const float * f32_data_03 = f32_data + i03 * nelements_matrix;
-                void * new_data_03 = (char *)new_data + ggml_row_size(new_type, n_per_row) * i03 * nrows;
-                const float * imatrix_03 = imatrix ? imatrix + i03 * n_per_row : nullptr;
-
-                new_size += llama_tensor_quantize_impl(new_type, f32_data_03, new_data_03, chunk_size, nrows, n_per_row, imatrix_03, workers, nthread_use);
-
-                // TODO: temporary sanity check that the F16 -> MXFP4 is lossless
-#if 0
-                if (new_type == GGML_TYPE_MXFP4) {
-                    auto * x = f32_data_03;
-
-                    //LLAMA_LOG_INFO("nrows = %d, n_per_row = %d\n", nrows, n_per_row);
-                    std::vector<float> deq(nrows*n_per_row);
-                    const ggml_type_traits * qtype = ggml_get_type_traits(new_type);
-                    qtype->to_float(new_data_03, deq.data(), deq.size());
-
-                    double err = 0.0f;
-                    for (int i = 0; i < (int) deq.size(); ++i) {
-                        err += fabsf(deq[i] - x[i]);
-                        //if (fabsf(deq[i] - x[i]) > 0.00001 && i < 256) {
-                        if (deq[i] != x[i]) {
-                            LLAMA_LOG_INFO("deq[%d] = %f, x[%d] = %f\n", i, deq[i], i, x[i]);
+                            // this can happen when quantizing an old mixtral model with split tensors with a new incompatible imatrix
+                            // this is a significant error and it may be good idea to abort the process if this happens,
+                            // since many people will miss the error and not realize that most of the model is being quantized without an imatrix
+                            // tok_embd should be ignored in this case, since it always causes this warning
+                            if (name != tn(LLM_TENSOR_TOKEN_EMBD, "weight")) {
+                                throw std::runtime_error(format("imatrix size %d is different from tensor size %d for %s",
+                                        int(it->second.size()), int(tensor->ne[0]*tensor->ne[2]), tensor->name));
+                            }
                         }
                     }
-                    //LLAMA_LOG_INFO("err = %f\n", err);
-                    GGML_ASSERT(err == 0.00000);
                 }
+                if (!imatrix && tensor_type_requires_imatrix(tensor, new_type, params->ftype)) {
+                    LLAMA_LOG_ERROR("\n\n============================================================\n");
+                    LLAMA_LOG_ERROR("Missing importance matrix for tensor %s in a very low-bit quantization\n", tensor->name);
+                    LLAMA_LOG_ERROR("The result will be garbage, so bailing out\n");
+                    LLAMA_LOG_ERROR("============================================================\n\n");
+                    throw std::runtime_error(format("Missing importance matrix for tensor %s in a very low-bit quantization", tensor->name));
+                }
+
+                float * f32_data;
+
+                if (tensor->type == GGML_TYPE_F32) {
+                    f32_data = (float *) tensor->data;
+                } else if (ggml_is_quantized(tensor->type) && !params->allow_requantize) {
+                    throw std::runtime_error(format("requantizing from type %s is disabled", ggml_type_name(tensor->type)));
+                } else {
+                    llama_tensor_dequantize_impl(tensor, f32_conv_buf, workers, nelements, nthread);
+                    f32_data = (float *) f32_conv_buf.data();
+                }
+
+                LLAMA_LOG_INFO("converting to %s .. ", ggml_type_name(new_type));
+                fflush(stdout);
+
+                if (work.size() < (size_t)nelements * 4) {
+                    work.resize(nelements * 4); // upper bound on size
+                }
+                new_data = work.data();
+
+                const int64_t n_per_row = tensor->ne[0];
+                const int64_t nrows = tensor->ne[1];
+
+                static const int64_t min_chunk_size = 32 * 512;
+                const int64_t chunk_size = (n_per_row >= min_chunk_size ? n_per_row : n_per_row * ((min_chunk_size + n_per_row - 1)/n_per_row));
+
+                const int64_t nelements_matrix = tensor->ne[0] * tensor->ne[1];
+                const int64_t nchunk = (nelements_matrix + chunk_size - 1)/chunk_size;
+                const int64_t nthread_use = nthread > 1 ? std::max((int64_t)1, std::min((int64_t)nthread, nchunk)) : 1;
+
+                // quantize each expert separately since they have different importance matrices
+                new_size = 0;
+                for (int64_t i03 = 0; i03 < tensor->ne[2]; ++i03) {
+                    const float * f32_data_03 = f32_data + i03 * nelements_matrix;
+                    void * new_data_03 = (char *)new_data + ggml_row_size(new_type, n_per_row) * i03 * nrows;
+                    const float * imatrix_03 = imatrix ? imatrix + i03 * n_per_row : nullptr;
+
+                    new_size += llama_tensor_quantize_impl(new_type, f32_data_03, new_data_03, chunk_size, nrows, n_per_row, imatrix_03, workers, nthread_use);
+
+                    // TODO: temporary sanity check that the F16 -> MXFP4 is lossless
+#if 0
+                    if (new_type == GGML_TYPE_MXFP4) {
+                        auto * x = f32_data_03;
+
+                        //LLAMA_LOG_INFO("nrows = %d, n_per_row = %d\n", nrows, n_per_row);
+                        std::vector<float> deq(nrows*n_per_row);
+                        const ggml_type_traits * qtype = ggml_get_type_traits(new_type);
+                        qtype->to_float(new_data_03, deq.data(), deq.size());
+
+                        double err = 0.0f;
+                        for (int i = 0; i < (int) deq.size(); ++i) {
+                            err += fabsf(deq[i] - x[i]);
+                            //if (fabsf(deq[i] - x[i]) > 0.00001 && i < 256) {
+                            if (deq[i] != x[i]) {
+                                LLAMA_LOG_INFO("deq[%d] = %f, x[%d] = %f\n", i, deq[i], i, x[i]);
+                            }
+                        }
+                        //LLAMA_LOG_INFO("err = %f\n", err);
+                        GGML_ASSERT(err == 0.00000);
+                    }
 #endif
+                }
+                LLAMA_LOG_INFO("size = %8.2f MiB -> %8.2f MiB\n", tensor_size/1024.0/1024.0, new_size/1024.0/1024.0);
             }
-            LLAMA_LOG_INFO("size = %8.2f MiB -> %8.2f MiB\n", ggml_nbytes(tensor)/1024.0/1024.0, new_size/1024.0/1024.0);
-        }
-        total_size_org += ggml_nbytes(tensor);
-        total_size_new += new_size;
+            total_size_org += tensor_size;
+            total_size_new += new_size;
 
-        // update the gguf meta data as we go
-        gguf_set_tensor_type(ctx_outs[cur_split].get(), name.c_str(), new_type);
-        GGML_ASSERT(gguf_get_tensor_size(ctx_outs[cur_split].get(), gguf_find_tensor(ctx_outs[cur_split].get(), name.c_str())) == new_size);
-        gguf_set_tensor_data(ctx_outs[cur_split].get(), name.c_str(), new_data);
+            // update the gguf meta data as we go
+            gguf_set_tensor_type(ctx_outs[cur_split].get(), name.c_str(), new_type);
+            GGML_ASSERT(gguf_get_tensor_size(ctx_outs[cur_split].get(), gguf_find_tensor(ctx_outs[cur_split].get(), name.c_str())) == new_size);
+            gguf_set_tensor_data(ctx_outs[cur_split].get(), name.c_str(), new_data);
 
-        // write tensor data + padding
-        fout.write((const char *) new_data, new_size);
-        zeros(fout, GGML_PAD(new_size, align) - new_size);
+            // write tensor data + padding
+            fout.write((const char *) new_data, new_size);
+            zeros(fout, GGML_PAD(new_size, align) - new_size);
+        } // no --dry-run
+    } // iterate over tensors
+
+    if (!params->dry_run) {
+        close_ofstream();
     }
-    close_ofstream();
 
-    LLAMA_LOG_INFO("%s: model size  = %8.2f MiB\n", __func__, total_size_org/1024.0/1024.0);
-    LLAMA_LOG_INFO("%s: quant size  = %8.2f MiB\n", __func__, total_size_new/1024.0/1024.0);
+    LLAMA_LOG_INFO("%s: model size  = %8.2f MiB (%.2f BPW)\n", __func__, total_size_org/1024.0/1024.0, total_size_org*8.0/ml.n_elements);
+    LLAMA_LOG_INFO("%s: quant size  = %8.2f MiB (%.2f BPW)\n", __func__, total_size_new/1024.0/1024.0, total_size_new*8.0/ml.n_elements);
+
+    if (!params->imatrix && params->dry_run && will_require_imatrix) {
+        LLAMA_LOG_WARN("%s: WARNING: dry run completed successfully, but actually completing this quantization will require an imatrix!\n",
+                       __func__
+        );
+    }
 
     if (qs.n_fallback > 0) {
         LLAMA_LOG_WARN("%s: WARNING: %d of %d tensor(s) required fallback quantization\n",
@@ -1045,6 +1094,7 @@ llama_model_quantize_params llama_model_quantize_default_params() {
         /*.only_copy                   =*/ false,
         /*.pure                        =*/ false,
         /*.keep_split                  =*/ false,
+        /*.dry_run                     =*/ false,
         /*.imatrix                     =*/ nullptr,
         /*.kv_overrides                =*/ nullptr,
         /*.tensor_type                 =*/ nullptr,

src/llama-vocab.cpp

@@ -2027,7 +2027,8 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {
                 pre_type = LLAMA_VOCAB_PRE_TYPE_QWEN2;
             } else if (
                 tokenizer_pre == "gpt-4o" ||
-                tokenizer_pre == "llama4") {
+                tokenizer_pre == "llama4" ||
+                tokenizer_pre == "kanana2") {
                 pre_type = LLAMA_VOCAB_PRE_TYPE_GPT4O;
                 clean_spaces = false;
             } else if (
@@ -2470,6 +2471,7 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {
                     || t.first == "<|calls|>"  // solar-open
                     || t.first == "<end_of_turn>"
                     || t.first == "<|endoftext|>"
+                    || t.first == "</s>"      // paddleocr
                     || t.first == "<|eom_id|>"
                     || t.first == "<EOT>"
                     || t.first == "_<EOT>"

src/models/models.h

@@ -190,6 +190,10 @@ struct llm_build_ernie4_5_moe : public llm_graph_context {
     llm_build_ernie4_5_moe(const llama_model & model, const llm_graph_params & params);
 };
 
+struct llm_build_paddleocr : public llm_graph_context {
+    llm_build_paddleocr(const llama_model & model, const llm_graph_params & params);
+};
+
 template <bool iswa>
 struct llm_build_exaone4 : public llm_graph_context {
     llm_build_exaone4(const llama_model & model, const llm_graph_params & params);

src/models/paddleocr.cpp

@@ -0,0 +1,122 @@
+#include "models.h"
+
+llm_build_paddleocr::llm_build_paddleocr(const llama_model & model, const llm_graph_params & params) :
+    llm_graph_context(params) {
+
+    // NOTE: same with qwen2vl.cpp, but bias tensors are optional
+
+    const int64_t n_embd_head = hparams.n_embd_head_v;
+
+    GGML_ASSERT(n_embd_head == hparams.n_embd_head_k);
+    GGML_ASSERT(n_embd_head == hparams.n_rot);
+
+    ggml_tensor * cur;
+    ggml_tensor * inpL;
+
+    inpL = build_inp_embd(model.tok_embd);
+
+    int sections[4];
+    std::copy(std::begin(hparams.rope_sections), std::begin(hparams.rope_sections) + 4, sections);
+
+    // inp_pos - contains the positions
+    ggml_tensor * inp_pos = build_inp_pos();
+
+    auto * inp_attn = build_attn_inp_kv();
+
+    ggml_tensor * inp_out_ids = build_inp_out_ids();
+
+    for (int il = 0; il < n_layer; ++il) {
+        ggml_tensor * inpSA = inpL;
+
+        // norm
+        {
+            cur = build_norm(inpL, model.layers[il].attn_norm, NULL, LLM_NORM_RMS, il);
+            cb(cur, "attn_norm", il);
+        }
+        // self-attention
+        {
+            ggml_tensor * Qcur = build_lora_mm(model.layers[il].wq, cur);
+            cb(Qcur, "Qcur", il);
+            if (model.layers[il].bq) {
+                Qcur = ggml_add(ctx0, Qcur, model.layers[il].bq);
+                cb(Qcur, "Qcur", il);
+            }
+            ggml_tensor * Kcur = build_lora_mm(model.layers[il].wk, cur);
+            cb(Kcur, "Kcur", il);
+            if (model.layers[il].bk) {
+                Kcur = ggml_add(ctx0, Kcur, model.layers[il].bk);
+                cb(Kcur, "Kcur", il);
+            }
+            ggml_tensor * Vcur = build_lora_mm(model.layers[il].wv, cur);
+            cb(Vcur, "Vcur", il);
+            if (model.layers[il].bv) {
+                Vcur = ggml_add(ctx0, Vcur, model.layers[il].bv);
+                cb(Vcur, "Vcur", il);
+            }
+            Qcur = ggml_reshape_3d(ctx0, Qcur, n_embd_head, n_head, n_tokens);
+            Kcur = ggml_reshape_3d(ctx0, Kcur, n_embd_head, n_head_kv, n_tokens);
+            Vcur = ggml_reshape_3d(ctx0, Vcur, n_embd_head, n_head_kv, n_tokens);
+
+            Qcur = ggml_rope_multi(
+                    ctx0, Qcur, inp_pos, nullptr,
+                    n_rot, sections, rope_type, n_ctx_orig, freq_base, freq_scale,
+                    ext_factor, attn_factor, beta_fast, beta_slow
+                    );
+
+            Kcur = ggml_rope_multi(
+                    ctx0, Kcur, inp_pos, nullptr,
+                    n_rot, sections, rope_type, n_ctx_orig, freq_base, freq_scale,
+                    ext_factor, attn_factor, beta_fast, beta_slow
+                    );
+
+            cb(Qcur, "Qcur", il);
+            cb(Kcur, "Kcur", il);
+            cb(Vcur, "Vcur", il);
+
+            cur = build_attn(inp_attn,
+                    model.layers[il].wo, model.layers[il].bo,
+                    Qcur, Kcur, Vcur, nullptr, nullptr, nullptr, 1.0f/sqrtf(float(n_embd_head)), il);
+        }
+        if (il == n_layer - 1) {
+            // skip computing output for unused tokens
+            cur   = ggml_get_rows(ctx0, cur, inp_out_ids);
+            inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);
+        }
+        ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);
+        cb(ffn_inp, "ffn_inp", il);
+
+        // feed-forward network
+        {
+            cur = build_norm(ffn_inp, model.layers[il].ffn_norm, NULL, LLM_NORM_RMS, il);
+            cb(cur, "ffn_norm", il);
+
+            cur = build_ffn(cur,
+                    model.layers[il].ffn_up, NULL, NULL,
+                    model.layers[il].ffn_gate, NULL, NULL,
+                    model.layers[il].ffn_down, NULL, NULL,
+                    NULL, LLM_FFN_SILU, LLM_FFN_PAR, il);
+            cb(cur, "ffn_out", il);
+        }
+        cur = ggml_add(ctx0, cur, ffn_inp);
+
+        cur = build_cvec(cur, il);
+        cb(cur, "l_out", il);
+
+        // input for next layer
+        inpL = cur;
+    }
+    cur = inpL;
+
+    cur = build_norm(cur, model.output_norm, NULL, LLM_NORM_RMS, -1);
+
+    cb(cur, "result_norm", -1);
+    res->t_embd = cur;
+
+    // lm_head
+    cur = build_lora_mm(model.output, cur);
+
+    cb(cur, "result_output", -1);
+    res->t_logits = cur;
+
+    ggml_build_forward_expand(gf, cur);
+}

tests/test-backend-ops.cpp

@@ -7791,6 +7791,10 @@ static std::vector<std::unique_ptr<test_case>> make_test_cases_eval() {
     test_cases.emplace_back(new test_mul_mat(GGML_TYPE_F16, GGML_TYPE_F32, 1056, 1, 67,  {1,  1}, {4, 1}, {0, 2, 1, 3}));
     test_cases.emplace_back(new test_mul_mat(GGML_TYPE_F32, GGML_TYPE_F32, 16, 32, 32, { 1,  1}, {1, 1}, {0, 1, 2, 3}, 64, 3));
     test_cases.emplace_back(new test_mul_mat(GGML_TYPE_F32, GGML_TYPE_F32, 64, 77, 77, {12,1}, {1,1}));
+    test_cases.emplace_back(new test_mul_mat(GGML_TYPE_F16, GGML_TYPE_F32, 2, 1, 3, {128, 1024}, {1, 1}));
+    test_cases.emplace_back(new test_mul_mat(GGML_TYPE_F16, GGML_TYPE_F32, 2, 3, 4, {128, 1024}, {1, 1}));
+    test_cases.emplace_back(new test_mul_mat(GGML_TYPE_F16, GGML_TYPE_F32, 2, 1, 3, {128*1024, 1}, {1, 1}, {0, 2, 1, 3}));
+    test_cases.emplace_back(new test_mul_mat(GGML_TYPE_F16, GGML_TYPE_F32, 2, 1, 3, {128*1024, 1}, {1, 1}, {0, 1, 2, 3}, 64));
 
     test_cases.emplace_back(new test_mul_mat(GGML_TYPE_Q4_0, GGML_TYPE_F32, 576, 512, 576, {1,1}, {1,1}));
     test_cases.emplace_back(new test_mul_mat(GGML_TYPE_Q4_0, GGML_TYPE_F32, 1, 2048, 8192, {1,  1}, {1, 1}));

tests/test-backend-sampler.cpp

@@ -361,7 +361,7 @@ static void test_backend_temp_sampling(const test_params & params) {
             GGML_ASSERT(false && "Failed to decode token");
         }
 
-        // Verfify sequence 0
+        // Verify sequence 0
         {
             int32_t batch_idx = test_ctx.idx_for_seq(0);
             int n_logits = llama_get_sampled_logits_count_ith(test_ctx.ctx.get(), batch_idx);
@@ -379,7 +379,7 @@ static void test_backend_temp_sampling(const test_params & params) {
         }
 
 
-        // Verfify sequence 1
+        // Verify sequence 1
         {
             int32_t batch_idx = test_ctx.idx_for_seq(1);
 
@@ -395,7 +395,7 @@ static void test_backend_temp_sampling(const test_params & params) {
         }
     }
 
-    // lambda to testing non-positive temperature values.
+    // lambda for testing non-positive temperature values.
     auto test_argmax_temp = [&](float temp) {
         printf("\nTesting temperature = %.1f\n", temp);
 
@@ -454,7 +454,7 @@ static void test_backend_temp_ext_sampling(const test_params & params) {
         }
     }
 
-    // lambda to testing non-positive temp/delta/exponent values.
+    // lambda for testing non-positive temp/delta/exponent values.
     auto test_argmax_temp = [&](float temp, float delta, float exponent) {
         printf("\nTesting temperature = %.1f, delta = %1.f, exponent = %1.f\n", temp, delta, exponent);
 
@@ -530,7 +530,7 @@ static void test_backend_min_p_sampling(const test_params & params) {
     printf("min-p cpu sampled token id:%d, string: '%s'\n", token, token_str.c_str());
     GGML_ASSERT(token >= 0 && token < test_ctx.n_vocab);
 
-    // Decode and sampler 10 more tokens
+    // Decode and sample 10 more tokens
     for (int i = 0; i < 10; i++) {
         int32_t loop_idx = test_ctx.idx_for_seq(seq_id);
         llama_token token = llama_sampler_sample(chain.get(), test_ctx.ctx.get(), loop_idx);
@@ -582,7 +582,7 @@ static void test_backend_top_p_sampling(const test_params & params) {
     printf("top-p cpu sampled token id:%d, string: '%s'\n", token, token_str.c_str());
     GGML_ASSERT(token >= 0 && token < test_ctx.n_vocab);
 
-    // Decode and sampler 10 more tokens
+    // Decode and sample 10 more tokens
     for (int i = 0; i < 10; i++) {
         int32_t loop_idx = test_ctx.idx_for_seq(seq_id);
         llama_token token = llama_sampler_sample(chain.get(), test_ctx.ctx.get(), loop_idx);
@@ -619,7 +619,7 @@ static void test_backend_multi_sequence_sampling(const test_params & params) {
         GGML_ASSERT(false && "Failed to decode token");
     }
 
-    // Verfiy sequence 0
+    // Verify sequence 0
     {
         int32_t batch_idx = test_ctx.idx_for_seq(0);
         llama_token token = llama_get_sampled_token_ith(test_ctx.ctx.get(), batch_idx);
@@ -763,7 +763,7 @@ static void test_backend_logit_bias_sampling(const test_params & params) {
     printf("backend logit bias sampling test PASSED\n");
 }
 
-// This test verifies that it is possible to have two different backend sampler,
+// This test verifies that it is possible to have two different backend samplers,
 // one that uses the backend dist sampler, and another that uses CPU dist sampler.
 static void test_backend_mixed_sampling(const test_params & params) {
     struct llama_sampler_chain_params chain_params_0 = llama_sampler_chain_default_params();
@@ -791,7 +791,7 @@ static void test_backend_mixed_sampling(const test_params & params) {
         GGML_ASSERT(false && "Failed to decode token");
     }
 
-    // Verfiy sequence 0 that used the dist backend sampler.
+    // Verify sequence 0 that used the dist backend sampler.
     {
         int32_t batch_idx = test_ctx.idx_for_seq(0);
         llama_token token = llama_get_sampled_token_ith(test_ctx.ctx.get(), batch_idx);
@@ -802,7 +802,7 @@ static void test_backend_mixed_sampling(const test_params & params) {
         //GGML_ASSERT(llama_get_sampled_logits_count_ith(test_ctx.ctx.get(), batch_idx) == 0);
     }
 
-    // Verfiy sequence 1 that used the top-k backend sampler.
+    // Verify sequence 1 that used the top-k backend sampler.
     {
         int32_t batch_idx = test_ctx.idx_for_seq(1);
         float * logits = llama_get_sampled_logits_ith(test_ctx.ctx.get(), batch_idx);
@@ -934,7 +934,7 @@ static void test_backend_cpu_mixed_batch(const test_params & params) {
         // samplers.
         llama_set_sampler(test_ctx.ctx.get(), 0, nullptr);
 
-        // Create a CPU sampler and verify we can sampler from it.
+        // Create a CPU sampler and verify we can sample from it.
         struct llama_sampler_chain_params chain_params = llama_sampler_chain_default_params();
         llama_sampler_ptr chain(llama_sampler_chain_init(chain_params));
         llama_sampler_chain_add(chain.get(), llama_sampler_init_greedy());

tests/test-chat.cpp

@@ -229,6 +229,20 @@ common_chat_tool python_tool {
         "required": ["code"]
     })",
 };
+common_chat_tool todo_list_tool {
+    /* .name = */ "todo_list",
+    /* .description = */ "Create or update the todo list",
+    /* .parameters = */ R"({
+        "type": "object",
+        "properties": {
+            "todos": {
+                "type": "array",
+                "description": "List of TODO list items"
+            }
+        },
+        "required": ["todos"]
+    })",
+};
 common_chat_tool code_interpreter_tool {
     /* .name = */ "code_interpreter",
     /* .description = */ "an ipython interpreter",
@@ -3018,540 +3032,26 @@ Hey there!<|im_end|>
         );
     }
 
-    // Test Qwen3-Coder XML format
     {
-        // Basic XML tool call parsing
-        assert_msg_equals(
-            message_assist_call,
-            test_chat_parse(
-                "<tool_call>\n"
-                "  <function=special_function>\n"
-                "    <parameter=arg1>\n"
-                "      1\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-                /* is_partial= */ false,
-                {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}));
+        // Step-3.5-Flash template: uses same XML output format as Qwen3-Coder and Nemotron v3,
+        // but with <think> support. Routes to the Nemotron v3 PEG parser for streaming and
+        // schema-aware parameter parsing.
+        auto tmpls = read_templates("models/templates/stepfun-ai-Step-3.5-Flash.jinja");
+        assert_equals(COMMON_CHAT_FORMAT_PEG_CONSTRUCTED, common_chat_templates_apply(tmpls.get(), inputs_tools).format);
 
-        // Multiple parameters with different types
-        common_chat_msg expected_multi_param;
-        expected_multi_param.role = "assistant";
-        expected_multi_param.tool_calls = {
-            { "complex_function", "{\"name\":\"John Doe\",\"age\":30,\"active\":true,\"score\":95.5}", "" }
-        };
-
-        test_parser_with_streaming(expected_multi_param,
-                "<tool_call>\n"
-                "  <function=complex_function>\n"
-                "    <parameter=name>\n"
-                "      John Doe\n"
-                "    </parameter>\n"
-                "    <parameter=age>\n"
-                "      30\n"
-                "    </parameter>\n"
-                "    <parameter=active>\n"
-                "      true\n"
-                "    </parameter>\n"
-                "    <parameter=score>\n"
-                "      95.5\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Special characters and Unicode
-        common_chat_msg expected_special_chars;
-        expected_special_chars.role = "assistant";
-        expected_special_chars.tool_calls = {
-            { "unicode_function", "{\"message\":\"Hello 世界! 🌍 Special chars: @#$%^&*()\"}", "" }
-        };
-
-        test_parser_with_streaming(expected_special_chars,
-                "<tool_call>\n"
-                "  <function=unicode_function>\n"
-                "    <parameter=message>\n"
-                "      Hello 世界! 🌍 Special chars: @#$%^&*()\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Multiline content with newlines and indentation
-        common_chat_msg expected_multiline;
-        expected_multiline.role = "assistant";
-        expected_multiline.tool_calls = {
-            { "code_function", "{\"code\":\"def hello():\\n    print(\\\"Hello, World!\\\")\\n    return True\"}", "" }
-        };
-
-        test_parser_with_streaming(expected_multiline,
-                "<tool_call>\n"
-                "  <function=code_function>\n"
-                "    <parameter=code>\n"
-                "def hello():\n"
-                "    print(\"Hello, World!\")\n"
-                "    return True\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // JSON object as parameter value
-        common_chat_msg expected_json_param;
-        expected_json_param.role = "assistant";
-        expected_json_param.tool_calls = {
-            { "json_function", "{\"config\":{\"host\":\"localhost\",\"port\":8080,\"ssl\":false}}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_json_param,
-                "<tool_call>\n"
-                "  <function=json_function>\n"
-                "    <parameter=config>\n"
-                "      {\"host\": \"localhost\", \"port\": 8080, \"ssl\": false}\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Array as parameter value
-        common_chat_msg expected_array_param;
-        expected_array_param.role = "assistant";
-        expected_array_param.tool_calls = {
-            { "array_function", "{\"items\":[\"apple\",\"banana\",\"cherry\"]}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_array_param,
-                "<tool_call>\n"
-                "  <function=array_function>\n"
-                "    <parameter=items>\n"
-                "      [\"apple\", \"banana\", \"cherry\"]\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Empty parameter
-        common_chat_msg expected_empty_param;
-        expected_empty_param.role = "assistant";
-        expected_empty_param.tool_calls = {
-            { "empty_function", "{\"empty_param\":\"\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_empty_param,
-                "<tool_call>\n"
-                "  <function=empty_function>\n"
-                "    <parameter=empty_param>\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Boolean values (true/false)
-        common_chat_msg expected_boolean;
-        expected_boolean.role = "assistant";
-        expected_boolean.tool_calls = {
-            { "boolean_function", "{\"enabled\":true,\"debug\":false}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_boolean,
-                "<tool_call>\n"
-                "  <function=boolean_function>\n"
-                "    <parameter=enabled>\n"
-                "      true\n"
-                "    </parameter>\n"
-                "    <parameter=debug>\n"
-                "      false\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Null value
-        common_chat_msg expected_null;
-        expected_null.role = "assistant";
-        expected_null.tool_calls = {
-            { "null_function", "{\"optional_param\":null}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_null,
-                "<tool_call>\n"
-                "  <function=null_function>\n"
-                "    <parameter=optional_param>\n"
-                "      null\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Negative numbers and scientific notation
-        common_chat_msg expected_numbers;
-        expected_numbers.role = "assistant";
-        expected_numbers.tool_calls = {
-            { "math_function", "{\"negative\":-42,\"decimal\":-3.14,\"scientific\":1.23e-4}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_numbers,
-                "<tool_call>\n"
-                "  <function=math_function>\n"
-                "    <parameter=negative>\n"
-                "      -42\n"
-                "    </parameter>\n"
-                "    <parameter=decimal>\n"
-                "      -3.14\n"
-                "    </parameter>\n"
-                "    <parameter=scientific>\n"
-                "      1.23e-4\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // XML-like content in parameters (should be escaped)
-        common_chat_msg expected_xml_content;
-        expected_xml_content.role = "assistant";
-        expected_xml_content.tool_calls = {
-            { "xml_function", "{\"xml_content\":\"<root><item>value</item></root>\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_xml_content,
-                "<tool_call>\n"
-                "  <function=xml_function>\n"
-                "    <parameter=xml_content>\n"
-                "      <root><item>value</item></root>\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Quotes and escape characters
-        common_chat_msg expected_quotes;
-        expected_quotes.role = "assistant";
-        expected_quotes.tool_calls = {
-            { "quote_function", "{\"message\":\"She said \\\"Hello!\\\" and left.\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_quotes,
-                "<tool_call>\n"
-                "  <function=quote_function>\n"
-                "    <parameter=message>\n"
-                "      She said \"Hello!\" and left.\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Long parameter value (simplified)
-        std::string long_text = "This is a long text parameter that should test the parser's ability to handle larger amounts of text data.";
-
-        common_chat_msg expected_long_text;
-        expected_long_text.role = "assistant";
-        expected_long_text.tool_calls = {
-            { "long_function", "{\"long_text\":\"" + long_text + "\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_long_text,
-                "<tool_call>\n"
-                "  <function=long_function>\n"
-                "    <parameter=long_text>\n"
-                "      " + long_text + "\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Mixed content with text before and after tool call
-        common_chat_msg expected_mixed_content;
-        expected_mixed_content.role = "assistant";
-        expected_mixed_content.content = "I'll help you search for products. ";
-        expected_mixed_content.tool_calls = {
-            { "search_function", "{\"query\":\"laptops\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_mixed_content,
-                "I'll help you search for products. <tool_call>\n"
-                "  <function=search_function>\n"
-                "    <parameter=query>\n"
-                "      laptops\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Compact format (no extra whitespace)
-        common_chat_msg expected_compact;
-        expected_compact.role = "assistant";
-        expected_compact.tool_calls = {
-            { "compact_function", "{\"param\":\"value\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_compact,
-                "<tool_call><function=compact_function><parameter=param>value</parameter></function></tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Function name with underscores and numbers
-        common_chat_msg expected_complex_name;
-        expected_complex_name.role = "assistant";
-        expected_complex_name.tool_calls = {
-            { "get_user_data_v2", "{\"user_id\":12345}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_complex_name,
-                "<tool_call>\n"
-                "  <function=get_user_data_v2>\n"
-                "    <parameter=user_id>\n"
-                "      12345\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Parameter names with underscores and numbers
-        common_chat_msg expected_complex_params;
-        expected_complex_params.role = "assistant";
-        expected_complex_params.tool_calls = {
-            { "test_function", "{\"param_1\":\"value1\",\"param_2_name\":\"value2\",\"param3\":123}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_complex_params,
-                "<tool_call>\n"
-                "  <function=test_function>\n"
-                "    <parameter=param_1>\n"
-                "      value1\n"
-                "    </parameter>\n"
-                "    <parameter=param_2_name>\n"
-                "      value2\n"
-                "    </parameter>\n"
-                "    <parameter=param3>\n"
-                "      123\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Very deeply nested XML content in parameter
-        common_chat_msg expected_deep_xml;
-        expected_deep_xml.role = "assistant";
-        expected_deep_xml.tool_calls = {
-            { "xml_parser", "{\"xml\":\"<root><level1><level2><level3>deep content</level3></level2></level1></root>\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_deep_xml,
-                "<tool_call>\n"
-                "  <function=xml_parser>\n"
-                "    <parameter=xml>\n"
-                "      <root><level1><level2><level3>deep content</level3></level2></level1></root>\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Parameter with only whitespace
-        common_chat_msg expected_whitespace_param;
-        expected_whitespace_param.role = "assistant";
-        expected_whitespace_param.tool_calls = {
-            { "whitespace_function", "{\"spaces\":\"\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_whitespace_param,
-                "<tool_call>\n"
-                "  <function=whitespace_function>\n"
-                "    <parameter=spaces>\n"
-                "      \n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Parameter with tabs and mixed whitespace
-        common_chat_msg expected_mixed_whitespace;
-        expected_mixed_whitespace.role = "assistant";
-        expected_mixed_whitespace.tool_calls = {
-            { "tab_function", "{\"content\":\"line1\\n\\tindented line\\n    spaces\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_mixed_whitespace,
-                "<tool_call>\n"
-                "  <function=tab_function>\n"
-                "    <parameter=content>\n"
-                "line1\n"
-                "\tindented line\n"
-                "    spaces\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Control characters and special Unicode
-        common_chat_msg expected_control_chars;
-        expected_control_chars.role = "assistant";
-        expected_control_chars.tool_calls = {
-            { "control_function", "{\"text\":\"Line1\\nLine2\\tTabbed\\rCarriage return\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_control_chars,
-                "<tool_call>\n"
-                "  <function=control_function>\n"
-                "    <parameter=text>\n"
-                "Line1\nLine2\tTabbed\rCarriage return\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Emoji and extended Unicode characters
-        common_chat_msg expected_emoji;
-        expected_emoji.role = "assistant";
-        expected_emoji.tool_calls = {
-            { "emoji_function", "{\"message\":\"Hello! 👋 🌟 🚀 Testing emojis: 😀😃😄😁 and symbols: ∑∏∆∇\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_emoji,
-                "<tool_call>\n"
-                "  <function=emoji_function>\n"
-                "    <parameter=message>\n"
-                "      Hello! 👋 🌟 🚀 Testing emojis: 😀😃😄😁 and symbols: ∑∏∆∇\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Mathematical expressions and formulas
-        common_chat_msg expected_math;
-        expected_math.role = "assistant";
-        expected_math.tool_calls = {
-            { "math_function", "{\"formula\":\"E = mc² and ∫f(x)dx = F(x) + C\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_math,
-                "<tool_call>\n"
-                "  <function=math_function>\n"
-                "    <parameter=formula>\n"
-                "      E = mc² and ∫f(x)dx = F(x) + C\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // SQL injection-like content (should be safely escaped)
-        common_chat_msg expected_sql;
-        expected_sql.role = "assistant";
-        expected_sql.tool_calls = {
-            { "sql_function", "{\"query\":\"SELECT * FROM users WHERE id = 1; DROP TABLE users; --\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_sql,
-                "<tool_call>\n"
-                "  <function=sql_function>\n"
-                "    <parameter=query>\n"
-                "      SELECT * FROM users WHERE id = 1; DROP TABLE users; --\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // HTML/XML injection content
-        common_chat_msg expected_html;
-        expected_html.role = "assistant";
-        expected_html.tool_calls = {
-            { "html_function", "{\"content\":\"<script>alert('xss')</script><img src=x onerror=alert(1)>\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_html,
-                "<tool_call>\n"
-                "  <function=html_function>\n"
-                "    <parameter=content>\n"
-                "      <script>alert('xss')</script><img src=x onerror=alert(1)>\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Binary-like content (base64)
-        common_chat_msg expected_binary;
-        expected_binary.role = "assistant";
-        expected_binary.tool_calls = {
-            { "binary_function", "{\"data\":\"SGVsbG8gV29ybGQhIFRoaXMgaXMgYmFzZTY0IGVuY29kZWQgdGV4dC4=\"}", "" }
-        };
-
-        test_parser_with_streaming(
-            expected_binary,
-                "<tool_call>\n"
-                "  <function=binary_function>\n"
-                "    <parameter=data>\n"
-                "      SGVsbG8gV29ybGQhIFRoaXMgaXMgYmFzZTY0IGVuY29kZWQgdGV4dC4=\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-
-        // Very large numbers (should be parsed as scientific notation)
-        common_chat_msg expected_large_numbers;
-        expected_large_numbers.role = "assistant";
-        expected_large_numbers.tool_calls = {
-            { "number_function", "{\"big_int\":1e+60}", "" }  // Large number becomes scientific notation
-        };
-
-        test_parser_with_streaming(
-            expected_large_numbers,
-                "<tool_call>\n"
-                "  <function=number_function>\n"
-                "    <parameter=big_int>\n"
-                "      999999999999999999999999999999999999999999999999999999999999\n"
-                "    </parameter>\n"
-                "  </function>\n"
-                "</tool_call>",
-            [&](const std::string &msg) { return test_chat_parse(msg, /* is_partial= */ true, {COMMON_CHAT_FORMAT_QWEN3_CODER_XML}); });
-    }
-
-    {
-        // Qwen3-Coder template
-        auto tmpls = read_templates("models/templates/Qwen3-Coder.jinja");
-        common_chat_templates_inputs inputs;
-        inputs.messages = { message_user };
-
-        common_chat_tool qwen_union_tool {
-            /* .name = */ "qwen_union",
-            /* .description = */ "Test tool for union/anyOf handling",
-            /* .parameters = */ R"({
-                "type": "object",
-                "properties": {
-                    "priority": { "type": ["number", "null"] },
-                    "maybe_text": { "anyOf": [ { "type": "string" } ] },
-                    "config": { "anyOf": [ { "type": "object" }, { "type": "null" } ] }
-                },
-                "required": []
-            })",
-        };
-        inputs.tools = { qwen_union_tool };
-
-        auto params = common_chat_templates_apply(tmpls.get(), inputs);
-        assert_equals(COMMON_CHAT_FORMAT_QWEN3_CODER_XML, params.format);
-        assert_equals(false, params.grammar.empty());
-
-        // Grammar should compile successfully
-        auto grammar = build_grammar(params.grammar);
-        GGML_ASSERT(grammar && "Failed to build Qwen3-Coder grammar with union types");
+        // Grammar and PEG parser should be generated with thinking_forced_open
+        {
+            common_chat_templates_inputs inputs;
+            inputs.messages = { message_user };
+            inputs.tools = { special_function_tool };
+            auto params = common_chat_templates_apply(tmpls.get(), inputs);
+            assert_equals(COMMON_CHAT_FORMAT_PEG_CONSTRUCTED, params.format);
+            assert_equals(true, params.thinking_forced_open);
+            assert_equals(false, params.grammar.empty());
+            assert_equals(false, params.parser.empty());
+            auto grammar = build_grammar(params.grammar);
+            GGML_ASSERT(grammar && "Failed to build Step-3.5-Flash grammar");
+        }
     }
 }
 
@@ -3643,6 +3143,135 @@ static void test_template_output_peg_parsers() {
         });
     }
 
+    {
+        // Qwen3-Coder
+        auto tmpls = read_templates("models/templates/Qwen3-Coder.jinja");
+
+        // Test basic message
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input = "Hello, world!\nWhat's up?";
+            t.expect = message_assist;
+        });
+
+        // Test tool call
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "<tool_call>\n"
+                "<function=special_function>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.tools = {special_function_tool};
+            t.expect = message_assist_call;
+        });
+
+        // Test parallel tool calls
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "<tool_call>\n"
+                "<function=special_function>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>\n"
+                "<tool_call>\n"
+                "<function=special_function_with_opt>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "<parameter=arg2>\n"
+                "2\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.parallel_tool_calls = true;
+            t.params.tools = {special_function_tool, special_function_tool_with_optional_param};
+
+            t.expect.tool_calls = {{
+                /* .name = */      "special_function",
+                /* .arguments = */ R"({"arg1": 1})",
+                /* .id = */        {},
+            }, {
+                /* .name = */      "special_function_with_opt",
+                /* .arguments = */ R"({"arg1": 1, "arg2": 2})",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test tool call with string parameter
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "<tool_call>\n"
+                "<function=python>\n"
+                "<parameter=code>\n"
+                "def hello():\n"
+                "    print(\"Hello, world!\")\n"
+                "\n"
+                "hello()\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.tools = {python_tool};
+
+            t.expect.tool_calls = {{
+                /* .name = */      "python",
+                /* .arguments = */ "{\"code\": \"def hello():\\n    print(\\\"Hello, world!\\\")\\n\\nhello()\"}",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test tool call with JSON parameter
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "<tool_call>\n"
+                "<function=todo_list>\n"
+                "<parameter=todos>\n"
+                "[{\"item\": \"Check stuff\", \"selected\": false}, {\"item\": \"Prepare stuff\", \"selected\": true}]\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.tools = {todo_list_tool};
+
+            t.expect.tool_calls = {{
+                /* .name = */      "todo_list",
+                /* .arguments = */ "{\"todos\": [{\"item\": \"Check stuff\", \"selected\": false}, {\"item\": \"Prepare stuff\", \"selected\": true}]}",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test tool call with string parameter and no closing </parameter> tag
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "<tool_call>\n"
+                "<function=python>\n"
+                "<parameter=code>\n"
+                "def hello():\n"
+                "    print(\"Hello, world!\")\n"
+                "\n"
+                "hello()\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.tools = {python_tool};
+
+            t.expect.tool_calls = {{
+                /* .name = */      "python",
+                /* .arguments = */ "{\"code\": \"def hello():\\n    print(\\\"Hello, world!\\\")\\n\\nhello()\"}",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test response format
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input = R"({"amount": 123.45, "date": "2025-12-03"})";
+            t.params.json_schema = invoice_schema;
+
+            t.expect.content = R"({"amount": 123.45, "date": "2025-12-03"})";
+        });
+    }
+
     {
         // NVIDIA Nemotron-3 Nano
         auto tmpls = read_templates("models/templates/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16.jinja");
@@ -3799,6 +3428,196 @@ static void test_template_output_peg_parsers() {
         });
     }
 
+    {
+        // Step-3.5-Flash (uses Nemotron v3 PEG parser with thinking_forced_open)
+        // Unlike Nemotron, Step-3.5-Flash always emits <think> regardless of enable_thinking,
+        // so all inputs must include a </think> delimiter.
+        auto tmpls = read_templates("models/templates/stepfun-ai-Step-3.5-Flash.jinja");
+
+        // Test basic message with reasoning
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input = "I'm\nthinking\n</think>\nHello, world!\nWhat's up?";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+
+            t.expect = message_assist_thoughts;
+        });
+
+        // Test basic message without thinking content
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input = "</think>\nHello, world!\nWhat's up?";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+
+            t.expect = message_assist;
+        });
+
+        // Test tool call without thinking content
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "</think>\n"
+                "<tool_call>\n"
+                "<function=special_function>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+            t.params.tools = {special_function_tool};
+
+            t.expect = message_assist_call;
+        });
+
+        // Test tool call with thinking
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "I'm\nthinking\n</think>\n"
+                "<tool_call>\n"
+                "<function=special_function>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+            t.params.tools = {special_function_tool};
+
+            t.expect = message_assist_call_thoughts;
+        });
+
+        // Test parallel tool calls with thinking
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "I'm\nthinking\n</think>\n"
+                "<tool_call>\n"
+                "<function=special_function>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>\n"
+                "<tool_call>\n"
+                "<function=special_function_with_opt>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "<parameter=arg2>\n"
+                "2\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+            t.params.parallel_tool_calls = true;
+            t.params.tools = {special_function_tool, special_function_tool_with_optional_param};
+
+            t.expect.reasoning_content = "I'm\nthinking";
+            t.expect.tool_calls = {{
+                /* .name = */      "special_function",
+                /* .arguments = */ R"({"arg1": 1})",
+                /* .id = */        {},
+            }, {
+                /* .name = */      "special_function_with_opt",
+                /* .arguments = */ R"({"arg1": 1, "arg2": 2})",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test parallel tool calls without thinking content
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "</think>\n"
+                "<tool_call>\n"
+                "<function=special_function>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>\n"
+                "<tool_call>\n"
+                "<function=special_function_with_opt>\n"
+                "<parameter=arg1>\n"
+                "1\n"
+                "</parameter>\n"
+                "<parameter=arg2>\n"
+                "2\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+            t.params.parallel_tool_calls = true;
+            t.params.tools = {special_function_tool, special_function_tool_with_optional_param};
+
+            t.expect.tool_calls = {{
+                /* .name = */      "special_function",
+                /* .arguments = */ R"({"arg1": 1})",
+                /* .id = */        {},
+            }, {
+                /* .name = */      "special_function_with_opt",
+                /* .arguments = */ R"({"arg1": 1, "arg2": 2})",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test tool call with code string parameter
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "</think>\n"
+                "<tool_call>\n"
+                "<function=python>\n"
+                "<parameter=code>\n"
+                "def hello():\n"
+                "    print(\"Hello, world!\")\n"
+                "\n"
+                "hello()\n"
+                "</parameter>\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+            t.params.tools = {python_tool};
+
+            t.expect.tool_calls = {{
+                /* .name = */      "python",
+                /* .arguments = */ "{\"code\": \"def hello():\\n    print(\\\"Hello, world!\\\")\\n\\nhello()\"}",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test tool call with string parameter and no closing </parameter> tag
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+                "</think>\n"
+                "<tool_call>\n"
+                "<function=python>\n"
+                "<parameter=code>\n"
+                "def hello():\n"
+                "    print(\"Hello, world!\")\n"
+                "\n"
+                "hello()\n"
+                "</function>\n"
+                "</tool_call>";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+            t.params.tools = {python_tool};
+
+            t.expect.tool_calls = {{
+                /* .name = */      "python",
+                /* .arguments = */ "{\"code\": \"def hello():\\n    print(\\\"Hello, world!\\\")\\n\\nhello()\"}",
+                /* .id = */        {},
+            }};
+        });
+
+        // Test response format (JSON schema with thinking)
+        test_peg_parser(tmpls.get(), [&](auto & t) {
+            t.input =
+              "I need to output the invoice details in JSON\n"
+              "</think>\n"
+              R"({"amount": 123.45, "date": "2025-12-03"})";
+            t.params.reasoning_format = COMMON_REASONING_FORMAT_AUTO;
+            t.params.json_schema = invoice_schema;
+
+            t.expect.reasoning_content = "I need to output the invoice details in JSON";
+            t.expect.content = R"({"amount": 123.45, "date": "2025-12-03"})";
+        });
+    }
+
     {
         // Solar-Open-100B
         auto tmpls = read_templates("models/templates/upstage-Solar-Open-100B.jinja");

tests/test-gguf.cpp

@@ -48,7 +48,6 @@ enum handcrafted_file_type {
     HANDCRAFTED_DATA_NOT_ENOUGH_DATA       =  10 + offset_has_data,
     HANDCRAFTED_DATA_BAD_ALIGN             =  15 + offset_has_data,
     HANDCRAFTED_DATA_INCONSISTENT_ALIGN    =  20 + offset_has_data,
-    HANDCRAFTED_DATA_MEM_SIZE_OVERFLOW     =  30 + offset_has_data,
     HANDCRAFTED_DATA_SUCCESS               = 800 + offset_has_data,
     HANDCRAFTED_DATA_CUSTOM_ALIGN          = 810 + offset_has_data,
 };
@@ -85,7 +84,6 @@ static std::string handcrafted_file_type_name(const enum handcrafted_file_type h
         case HANDCRAFTED_DATA_NOT_ENOUGH_DATA:       return "DATA_NOT_ENOUGH_DATA";
         case HANDCRAFTED_DATA_BAD_ALIGN:             return "DATA_BAD_ALIGN";
         case HANDCRAFTED_DATA_INCONSISTENT_ALIGN:    return "DATA_INCONSISTENT_ALIGN";
-        case HANDCRAFTED_DATA_MEM_SIZE_OVERFLOW:     return "DATA_MEM_SIZE_OVERFLOW";
         case HANDCRAFTED_DATA_SUCCESS:               return "DATA_SUCCESS";
         case HANDCRAFTED_DATA_CUSTOM_ALIGN:          return "DATA_CUSTOM_ALIGN";
     }
@@ -198,13 +196,6 @@ static FILE * get_handcrafted_file(const unsigned int seed, const enum handcraft
         tensor_configs = get_tensor_configs(rng);
     }
 
-    if (hft == HANDCRAFTED_DATA_MEM_SIZE_OVERFLOW) {
-        tensor_configs.resize(2);
-
-        tensor_configs[0] = { GGML_TYPE_I8, { 0x7FFFFFFFFFFFFFC0, 1, 1, 1 } };
-        tensor_configs[1] = { GGML_TYPE_I8, { 0x7FFFFFFFFFFFFFC0, 1, 1, 1 } };
-    }
-
     if (hft == HANDCRAFTED_HEADER_BAD_N_TENSORS) {
         const uint64_t n_tensors = -1;
         helper_write(file, n_tensors);
@@ -406,8 +397,7 @@ static FILE * get_handcrafted_file(const unsigned int seed, const enum handcraft
         for (uint32_t i = 1; i < n_dims; ++i) {
             ne *= shape[i];
         }
-
-        offset += GGML_PAD(ggml_row_size(type, ne), (uint64_t) alignment);
+        offset += GGML_PAD(ggml_row_size(type, ne), alignment);
     }
 
     while (ftell(file) % alignment != 0) {
@@ -421,9 +411,6 @@ static FILE * get_handcrafted_file(const unsigned int seed, const enum handcraft
         if (hft == HANDCRAFTED_DATA_NOT_ENOUGH_DATA) {
             nbytes -= 1;
         }
-        if (hft == HANDCRAFTED_DATA_MEM_SIZE_OVERFLOW) {
-            nbytes = 32;
-        }
         for (uint64_t i = 0; i < nbytes; ++i) {
             const uint8_t random_byte = i % 256;
             helper_write(file, random_byte);
@@ -717,7 +704,6 @@ static std::pair<int, int> test_handcrafted_file(const unsigned int seed) {
         HANDCRAFTED_DATA_NOT_ENOUGH_DATA,
         HANDCRAFTED_DATA_BAD_ALIGN,
         HANDCRAFTED_DATA_INCONSISTENT_ALIGN,
-        HANDCRAFTED_DATA_MEM_SIZE_OVERFLOW,
         HANDCRAFTED_DATA_SUCCESS,
         HANDCRAFTED_DATA_CUSTOM_ALIGN,
     };

tests/test-jinja.cpp

@@ -32,6 +32,7 @@ static void test_string_methods(testing & t);
 static void test_array_methods(testing & t);
 static void test_object_methods(testing & t);
 static void test_hasher(testing & t);
+static void test_stats(testing & t);
 static void test_fuzzing(testing & t);
 
 static bool g_python_mode = false;
@@ -70,6 +71,7 @@ int main(int argc, char *argv[]) {
     t.test("object methods", test_object_methods);
     if (!g_python_mode) {
         t.test("hasher", test_hasher);
+        t.test("stats", test_stats);
         t.test("fuzzing", test_fuzzing);
     }
 
@@ -1795,6 +1797,63 @@ static void test_hasher(testing & t) {
     });
 }
 
+static void test_stats(testing & t) {
+    static auto get_stats = [](const std::string & tmpl, const json & vars) -> jinja::value {
+        jinja::lexer lexer;
+        auto lexer_res = lexer.tokenize(tmpl);
+
+        jinja::program prog = jinja::parse_from_tokens(lexer_res);
+
+        jinja::context ctx(tmpl);
+        jinja::global_from_json(ctx, json{{ "val", vars }}, true);
+        ctx.is_get_stats = true;
+
+        jinja::runtime runtime(ctx);
+        runtime.execute(prog);
+
+        return ctx.get_val("val");
+    };
+
+    t.test("stats", [](testing & t) {
+        jinja::value val = get_stats(
+            "{{val.num}} "
+            "{{val.str}} "
+            "{{val.arr[0]}} "
+            "{{val.obj.key1}} "
+            "{{val.nested | tojson}}",
+            // Note: the json below will be wrapped inside "val" in the context
+            json{
+                {"num", 1},
+                {"str", "abc"},
+                {"arr", json::array({1, 2, 3})},
+                {"obj", json::object({{"key1", 1}, {"key2", 2}, {"key3", 3}})},
+                {"nested", json::object({
+                    {"inner_key1", json::array({1, 2})},
+                    {"inner_key2", json::object({{"a", "x"}, {"b", "y"}})}
+                })},
+                {"mixed", json::object({
+                    {"used", 1},
+                    {"unused", 2},
+                })},
+            }
+        );
+
+        t.assert_true("num is used", val->at("num")->stats.used);
+        t.assert_true("str is used", val->at("str")->stats.used);
+
+        t.assert_true("arr is used", val->at("arr")->stats.used);
+        t.assert_true("arr[0] is used", val->at("arr")->at(0)->stats.used);
+        t.assert_true("arr[1] is not used", !val->at("arr")->at(1)->stats.used);
+
+        t.assert_true("obj is used", val->at("obj")->stats.used);
+        t.assert_true("obj.key1 is used", val->at("obj")->at("key1")->stats.used);
+        t.assert_true("obj.key2 is not used", !val->at("obj")->at("key2")->stats.used);
+
+        t.assert_true("inner_key1[0] is used", val->at("nested")->at("inner_key1")->at(0)->stats.used);
+        t.assert_true("inner_key2.a is used", val->at("nested")->at("inner_key2")->at("a")->stats.used);
+    });
+}
+
 static void test_template_cpp(testing & t, const std::string & name, const std::string & tmpl, const json & vars, const std::string & expect) {
     t.test(name, [&tmpl, &vars, &expect](testing & t) {
         jinja::lexer lexer;

tools/cli/cli.cpp

@@ -380,6 +380,15 @@ int main(int argc, char ** argv) {
                 console::error("file does not exist or cannot be opened: '%s'\n", fname.c_str());
                 continue;
             }
+            if (inf.fim_sep_token != LLAMA_TOKEN_NULL) {
+                cur_msg += common_token_to_piece(ctx_cli.ctx_server.get_llama_context(), inf.fim_sep_token, true);
+                cur_msg += fname;
+                cur_msg.push_back('\n');
+            } else {
+                cur_msg += "--- File: ";
+                cur_msg += fname;
+                cur_msg += " ---\n";
+            }
             cur_msg += marker;
             console::log("Loaded text from '%s'\n", fname.c_str());
             continue;

tools/completion/completion.cpp

@@ -387,6 +387,17 @@ int main(int argc, char ** argv) {
         }
 
         session_do_save = !path_session.empty() && n_match < embd_inp.size() && !params.prompt_cache_ro;
+
+        // Logits are not stored as part of the session state so we need to
+        // "replay" the last token to get logits for sampling.
+        if (!session_tokens.empty() && n_match > 0 && n_match == session_tokens.size()) {
+            if (!common_replay_last_token(ctx, session_tokens.back(), n_match)) {
+                return 1;
+            }
+
+            session_do_save = false;
+            LOG_INF("%s: replayed last token from session\n", __func__);
+        }
     }
 
     // number of tokens to keep when resetting context
@@ -675,40 +686,27 @@ int main(int argc, char ** argv) {
             }
 
             if (!embd.empty()) {
-                int n_eval = (int) embd.size();
-                LOG_DBG("eval: %s\n", string_from(ctx, embd).c_str());
-
-                GGML_ASSERT(n_eval <= params.n_batch);
-                if (llama_decode(ctx, llama_batch_get_one(embd.data(), n_eval))) {
-                    LOG_ERR("%s : failed to eval\n", __func__);
+                const bool is_last_batch = (n_consumed >= (int) embd_inp.size());
+                const bool save_now = session_do_save && is_last_batch;
+                if (!common_prompt_batch_decode(ctx, embd, n_past, params.n_batch, path_session, save_now)) {
                     return 1;
                 }
-
-                n_past += n_eval;
+                session_tokens.insert(session_tokens.end(), embd.begin(), embd.begin());
+                n_session_consumed = session_tokens.size();
+                session_do_save = false;
 
                 LOG_DBG("n_past = %d\n", n_past);
+
                 // Display total tokens alongside total time
                 if (params.n_print > 0 && n_past % params.n_print == 0) {
                     LOG_DBG("\n\033[31mTokens consumed so far = %d / %d \033[0m\n", n_past, n_ctx);
                 }
             }
-
-            if (!embd.empty() && !path_session.empty()) {
-                session_tokens.insert(session_tokens.end(), embd.begin(), embd.end());
-                n_session_consumed = session_tokens.size();
-            }
         }
 
         embd.clear();
 
         if ((int) embd_inp.size() <= n_consumed && !is_interacting) {
-            // optionally save the session on first sample (for faster prompt loading next time)
-            if (session_do_save) {
-                session_do_save = false;
-                llama_state_save_file(ctx, path_session.c_str(), session_tokens.data(), session_tokens.size());
-
-                LOG_DBG("saved session to %s\n", path_session.c_str());
-            }
 
             const llama_token id = common_sampler_sample(smpl, ctx, -1);
 

tools/mtmd/CMakeLists.txt

@@ -24,6 +24,7 @@ add_library(mtmd
             models/llama4.cpp
             models/llava.cpp
             models/minicpmv.cpp
+            models/paddleocr.cpp
             models/pixtral.cpp
             models/qwen2vl.cpp
             models/qwen3vl.cpp

tools/mtmd/clip-impl.h

@@ -229,6 +229,7 @@ enum projector_type {
     PROJECTOR_TYPE_MUSIC_FLAMINGO,
     PROJECTOR_TYPE_LFM2,
     PROJECTOR_TYPE_KIMIVL,
+    PROJECTOR_TYPE_PADDLEOCR,
     PROJECTOR_TYPE_LIGHTONOCR,
     PROJECTOR_TYPE_COGVLM,
     PROJECTOR_TYPE_JANUS_PRO,
@@ -264,6 +265,7 @@ static std::map<projector_type, std::string> PROJECTOR_TYPE_NAMES = {
     { PROJECTOR_TYPE_MUSIC_FLAMINGO, "musicflamingo"},
     { PROJECTOR_TYPE_LFM2,      "lfm2"},
     { PROJECTOR_TYPE_KIMIVL,    "kimivl"},
+    { PROJECTOR_TYPE_PADDLEOCR, "paddleocr"},
     { PROJECTOR_TYPE_LIGHTONOCR,"lightonocr"},
     { PROJECTOR_TYPE_COGVLM,    "cogvlm"},
     { PROJECTOR_TYPE_JANUS_PRO, "janus_pro"},

tools/mtmd/clip.cpp

@@ -841,6 +841,10 @@ static ggml_cgraph * clip_image_build_graph(clip_ctx * ctx, const clip_image_f32
             {
                 builder = std::make_unique<clip_graph_kimivl>(ctx, img);
             } break;
+        case PROJECTOR_TYPE_PADDLEOCR:
+            {
+                builder = std::make_unique<clip_graph_paddleocr>(ctx, img);
+            } break;
         case PROJECTOR_TYPE_KIMIK25:
             {
                 builder = std::make_unique<clip_graph_kimik25>(ctx, img);
@@ -1256,6 +1260,14 @@ struct clip_model_loader {
                         hparams.audio_window_len   = 400;
                         hparams.audio_hop_len      = 160;
                     } break;
+                case PROJECTOR_TYPE_PADDLEOCR:
+                    {
+                        hparams.n_merge = 2;
+                        get_u32(KEY_IMAGE_MIN_PIXELS, hparams.image_min_pixels);
+                        get_u32(KEY_IMAGE_MAX_PIXELS, hparams.image_max_pixels);
+
+                        hparams.set_warmup_n_tokens(28*28); // avoid OOM on warmup
+                    } break;
                 case PROJECTOR_TYPE_LFM2A:
                     {
                         // audio preprocessing params
@@ -1704,6 +1716,7 @@ struct clip_model_loader {
                     model.mm_2_b = get_tensor(string_format(TN_LLAVA_PROJ, 2, "bias"));
                 } break;
             case PROJECTOR_TYPE_KIMIVL:
+            case PROJECTOR_TYPE_PADDLEOCR:
             case PROJECTOR_TYPE_KIMIK25:
                 {
                     model.mm_input_norm_w = get_tensor(TN_MM_INP_NORM);
@@ -2990,6 +3003,7 @@ bool clip_image_preprocess(struct clip_ctx * ctx, const clip_image_u8 * img, str
         case PROJECTOR_TYPE_QWEN25VL:
         case PROJECTOR_TYPE_QWEN3VL:
         case PROJECTOR_TYPE_GLM4V:
+        case PROJECTOR_TYPE_PADDLEOCR:
             {
                 GGML_ASSERT(params.image_min_pixels > 0 && params.image_max_pixels > 0);
                 clip_image_u8 resized;
@@ -3330,6 +3344,7 @@ int clip_n_output_tokens_x(const struct clip_ctx * ctx, struct clip_image_f32 *
         case PROJECTOR_TYPE_QWEN25VL:
         case PROJECTOR_TYPE_QWEN3VL:
         case PROJECTOR_TYPE_GLM4V:
+        case PROJECTOR_TYPE_PADDLEOCR:
         case PROJECTOR_TYPE_YOUTUVL:
             return (img->nx / params.patch_size) / 2;
         default:
@@ -3346,6 +3361,7 @@ int clip_n_output_tokens_y(const struct clip_ctx * ctx, struct clip_image_f32 *
         case PROJECTOR_TYPE_QWEN25VL:
         case PROJECTOR_TYPE_QWEN3VL:
         case PROJECTOR_TYPE_GLM4V:
+        case PROJECTOR_TYPE_PADDLEOCR:
         case PROJECTOR_TYPE_YOUTUVL:
             return (img->ny / params.patch_size) / 2;
         default:
@@ -3443,6 +3459,13 @@ int clip_n_output_tokens(const struct clip_ctx * ctx, struct clip_image_f32 * im
                 int y_patch = CLIP_ALIGN(img->ny, out_patch_size) / out_patch_size;
                 n_patches = x_patch * y_patch;
             } break;
+        case PROJECTOR_TYPE_PADDLEOCR:
+            {
+                // dynamic size
+                int n_merge = ctx->model.hparams.n_merge;
+                int stride = n_merge * n_merge;
+                n_patches = CLIP_ALIGN(n_patches, stride) / stride;
+            } break;
         case PROJECTOR_TYPE_PIXTRAL:
         case PROJECTOR_TYPE_LIGHTONOCR:
             {
@@ -3690,6 +3713,30 @@ bool clip_image_batch_encode(clip_ctx * ctx, const int n_threads, const clip_ima
                     }
                 }
 
+                set_input_i32("positions", positions);
+            } break;
+        case PROJECTOR_TYPE_PADDLEOCR:
+            {
+                const int merge_ratio = hparams.n_merge;
+                const int pw = image_size_width  / patch_size;
+                const int ph = image_size_height / patch_size;
+                std::vector<int> positions(n_pos * 4);
+                int ptr = 0;
+                // NOTE: same as Qwen-VL, but x and y are swapped
+                for (int y = 0; y < ph; y += merge_ratio) {
+                    for (int dy = 0; dy < 2; dy++) {
+                        for (int x = 0; x < pw; x += merge_ratio) {
+                            for (int dx = 0; dx < 2; dx++) {
+                                positions[                  ptr] = y + dy;
+                                positions[    num_patches + ptr] = x + dx;
+                                positions[2 * num_patches + ptr] = y + dy;
+                                positions[3 * num_patches + ptr] = x + dx;
+                                ptr++;
+                            }
+                        }
+                    }
+                }
+
                 set_input_i32("positions", positions);
             } break;
         case PROJECTOR_TYPE_QWEN25VL:
@@ -4003,6 +4050,7 @@ int clip_n_mmproj_embd(const struct clip_ctx * ctx) {
             return ctx->model.mm_2_w->ne[1];
         case PROJECTOR_TYPE_LFM2:
         case PROJECTOR_TYPE_KIMIVL:
+        case PROJECTOR_TYPE_PADDLEOCR:
         case PROJECTOR_TYPE_KIMIK25:
             return ctx->model.mm_2_w->ne[1];
         case PROJECTOR_TYPE_COGVLM:

tools/mtmd/models/models.h

@@ -57,6 +57,11 @@ struct clip_graph_kimivl : clip_graph {
     ggml_cgraph * build() override;
 };
 
+struct clip_graph_paddleocr : clip_graph {
+    clip_graph_paddleocr(clip_ctx * ctx, const clip_image_f32 & img) : clip_graph(ctx, img) {}
+    ggml_cgraph * build() override;
+};
+
 struct clip_graph_cogvlm : clip_graph {
     clip_graph_cogvlm(clip_ctx * ctx, const clip_image_f32 & img) : clip_graph(ctx, img) {}
     ggml_cgraph * build() override;

tools/mtmd/models/paddleocr.cpp

@@ -0,0 +1,52 @@
+#include "models.h"
+
+ggml_cgraph * clip_graph_paddleocr::build() {
+    const int n_pos            = n_patches;
+    const int num_position_ids = n_pos * 4; // m-rope requires 4 dim per position
+
+    int mrope_sections[4] = {d_head/4, d_head/4, d_head/4, d_head/4};
+
+    ggml_tensor * positions = ggml_new_tensor_1d(ctx0, GGML_TYPE_I32, num_position_ids);
+    ggml_set_name(positions, "positions");
+    ggml_set_input(positions);
+
+    auto add_pos = [&](ggml_tensor * cur, const clip_layer &) {
+        return ggml_rope_multi(
+                    ctx0, cur, positions, nullptr,
+                    d_head/2, mrope_sections, GGML_ROPE_TYPE_VISION,
+                    32768, 10000, 1, 0, 1, 32, 1);
+    };
+
+    ggml_tensor * learned_pos_embd = resize_position_embeddings();
+    ggml_tensor * inp = build_inp();
+    ggml_tensor * cur = build_vit(
+                            inp, n_patches,
+                            NORM_TYPE_NORMAL,
+                            hparams.ffn_op,
+                            learned_pos_embd,
+                            add_pos);
+
+    cb(cur, "vit_out", -1);
+
+    {
+        // mlp_AR paddleocr projector
+        float proj_norm_eps = 1e-5;
+        cur = build_norm(cur,
+                    model.mm_input_norm_w, model.mm_input_norm_b,
+                    NORM_TYPE_NORMAL, proj_norm_eps, -1);
+
+        const int scale_factor = model.hparams.n_merge;
+        cur = build_patch_merge_permute(cur, scale_factor);
+        cur = build_ffn(cur,
+                    model.mm_1_w, model.mm_1_b,
+                    nullptr, nullptr,
+                    model.mm_2_w, model.mm_2_b,
+                    hparams.ffn_op, -1);
+        cb(cur, "mlp_out", -1);
+    }
+
+    // build the graph
+    ggml_build_forward_expand(gf, cur);
+
+    return gf;
+}

tools/mtmd/mtmd.cpp

@@ -325,6 +325,10 @@ struct mtmd_context {
             img_beg = "<|begin_of_image|>";
             img_end = "<|end_of_image|>";
 
+        } else if (proj == PROJECTOR_TYPE_PADDLEOCR) {
+            // <|IMAGE_START|> ... (image embeddings) ... <|IMAGE_END|>
+            img_beg = "<|IMAGE_START|>";
+            img_end = "<|IMAGE_END|>";
         }
     }
 
@@ -890,6 +894,7 @@ bool mtmd_decode_use_mrope(mtmd_context * ctx) {
         case PROJECTOR_TYPE_QWEN25VL:
         case PROJECTOR_TYPE_QWEN3VL:
         case PROJECTOR_TYPE_GLM4V:
+        case PROJECTOR_TYPE_PADDLEOCR:
             return true;
         default:
             return false;

tools/quantize/quantize.cpp

@@ -120,7 +120,7 @@ static bool try_parse_ftype(const std::string & ftype_str_in, llama_ftype & ftyp
 static void usage(const char * executable) {
     printf("usage: %s [--help] [--allow-requantize] [--leave-output-tensor] [--pure] [--imatrix] [--include-weights]\n", executable);
     printf("       [--exclude-weights] [--output-tensor-type] [--token-embedding-type] [--tensor-type] [--tensor-type-file]\n");
-    printf("       [--prune-layers] [--keep-split] [--override-kv]\n");
+    printf("       [--prune-layers] [--keep-split] [--override-kv] [--dry-run]\n");
     printf("       model-f32.gguf [model-quant.gguf] type [nthreads]\n\n");
     printf("  --allow-requantize\n");
     printf("                                      allow requantizing tensors that have already been quantized\n");
@@ -156,7 +156,10 @@ static void usage(const char * executable) {
     printf("                                      generate quantized model in the same shards as input\n");
     printf("  --override-kv KEY=TYPE:VALUE\n");
     printf("                                      override model metadata by key in the quantized model. may be specified multiple times.\n");
-    printf("                                      WARNING: this is an advanced option, use with care.\n\n");
+    printf("                                      WARNING: this is an advanced option, use with care.\n");
+    printf("  --dry-run\n");
+    printf("                                      calculate and show the final quantization size without performing quantization\n");
+    printf("                                      example: llama-quantize --dry-run model-f32.gguf Q4_K\n\n");
     printf("note: --include-weights and --exclude-weights cannot be used together\n\n");
     printf("-----------------------------------------------------------------------------\n");
     printf(" allowed quantization types\n");
@@ -532,6 +535,8 @@ int main(int argc, char ** argv) {
             if (arg_idx == argc-1 || !string_parse_kv_override(argv[++arg_idx], kv_overrides)) {
                 usage(argv[0]);
             }
+        } else if (strcmp(argv[arg_idx], "--dry-run") == 0) {
+            params.dry_run = true;
         } else if (strcmp(argv[arg_idx], "--allow-requantize") == 0) {
             params.allow_requantize = true;
         } else if (strcmp(argv[arg_idx], "--pure") == 0) {
@@ -630,22 +635,26 @@ int main(int argc, char ** argv) {
     std::string ftype_str;
     std::string suffix = ".gguf";
     if (try_parse_ftype(argv[arg_idx], params.ftype, ftype_str)) {
-        std::string fpath;
-        const size_t pos = fname_inp.find_last_of("/\\");
-        if (pos != std::string::npos) {
-            fpath = fname_inp.substr(0, pos + 1);
-        }
+        // argv[arg_idx] is the ftype directly: <input> <ftype>
+        if (!params.dry_run) {
+            std::string fpath;
+            const size_t pos = fname_inp.find_last_of("/\\");
+            if (pos != std::string::npos) {
+                fpath = fname_inp.substr(0, pos + 1);
+            }
 
-        // export as [inp path]/ggml-model-[ftype]. Only add extension if there is no splitting
-        fname_out = fpath + "ggml-model-" + ftype_str;
-        if (!params.keep_split) {
-            fname_out += suffix;
+            // export as [inp path]/ggml-model-[ftype]. Only add extension if there is no splitting
+            fname_out = fpath + "ggml-model-" + ftype_str;
+            if (!params.keep_split) {
+                fname_out += suffix;
+            }
         }
         arg_idx++;
         if (ftype_str == "COPY") {
             params.only_copy = true;
         }
     } else {
+        // argv[arg_idx] is not a valid ftype, so treat it as output path: <input> <output> <ftype>
         fname_out = argv[arg_idx];
         if (params.keep_split && fname_out.find(suffix) != std::string::npos) {
             fname_out = fname_out.substr(0, fname_out.length() - suffix.length());
@@ -677,25 +686,33 @@ int main(int argc, char ** argv) {
         }
     }
 
-    if ((params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XS || params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XXS ||
-         params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_S  ||
-         params.ftype == LLAMA_FTYPE_MOSTLY_Q2_K_S ||
-         params.ftype == LLAMA_FTYPE_MOSTLY_IQ1_S  ||
-         params.ftype == LLAMA_FTYPE_MOSTLY_IQ1_M) && imatrix_data.empty()) {
+    if (!params.dry_run &&
+        (
+            params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XS  || params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XXS ||
+            params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_S   || params.ftype == LLAMA_FTYPE_MOSTLY_Q2_K_S  ||
+            params.ftype == LLAMA_FTYPE_MOSTLY_IQ1_S   || params.ftype == LLAMA_FTYPE_MOSTLY_IQ1_M
+        ) && imatrix_data.empty()) {
         fprintf(stderr, "\n==========================================================================================================\n");
         fprintf(stderr, "Please do not use IQ1_S, IQ1_M, IQ2_S, IQ2_XXS, IQ2_XS or Q2_K_S quantization without an importance matrix\n");
         fprintf(stderr, "==========================================================================================================\n\n\n");
         return 1;
     }
 
-    if (std::error_code ec; std::filesystem::equivalent(fname_inp, fname_out, ec)) {
-        fprintf(stderr, "%s: error: input and output files are the same: '%s'\n", __func__, fname_inp.c_str());
-        return 1;
+    if (!params.dry_run) {
+        if (std::error_code ec; std::filesystem::equivalent(fname_inp, fname_out, ec)) {
+            fprintf(stderr, "%s: error: input and output files are the same: '%s'\n", __func__, fname_inp.c_str());
+            return 1;
+        }
     }
 
     print_build_info();
 
-    fprintf(stderr, "%s: quantizing '%s' to '%s' as %s", __func__, fname_inp.c_str(), fname_out.c_str(), ftype_str.c_str());
+    if (params.dry_run) {
+        fprintf(stderr, "%s: calculating quantization size for '%s' as %s", __func__, fname_inp.c_str(), ftype_str.c_str());
+    } else {
+        fprintf(stderr, "%s: quantizing '%s' to '%s' as %s", __func__, fname_inp.c_str(), fname_out.c_str(), ftype_str.c_str());
+    }
+
     if (params.nthread > 0) {
         fprintf(stderr, " using %d threads", params.nthread);
     }

tools/server/server-common.cpp

@@ -1105,6 +1105,8 @@ json convert_responses_to_chatcmpl(const json & response_body) {
         };
 
         for (json item : input_value) {
+            bool merge_prev = !chatcmpl_messages.empty() && chatcmpl_messages.back().value("role", "") == "assistant";
+
             if (exists_and_is_string(item, "content")) {
                 // #responses_create-input-input_item_list-input_message-content-text_input
                 // Only "Input message" contains item["content"]::string
@@ -1193,7 +1195,7 @@ json convert_responses_to_chatcmpl(const json & response_body) {
                 item.at("type") == "message"
             ) {
                 // #responses_create-input-input_item_list-item-output_message
-                std::vector<json> chatcmpl_content;
+                auto chatcmpl_content = json::array();
 
                 for (const auto & output_text : item.at("content")) {
                     const std::string type = json_value(output_text, "type", std::string());
@@ -1210,10 +1212,19 @@ json convert_responses_to_chatcmpl(const json & response_body) {
                     });
                 }
 
-                item.erase("status");
-                item.erase("type");
-                item["content"] = chatcmpl_content;
-                chatcmpl_messages.push_back(item);
+                if (merge_prev) {
+                    auto & prev_msg = chatcmpl_messages.back();
+                    if (!exists_and_is_array(prev_msg, "content")) {
+                        prev_msg["content"] = json::array();
+                    }
+                    auto & prev_content = prev_msg["content"];
+                    prev_content.insert(prev_content.end(), chatcmpl_content.begin(), chatcmpl_content.end());
+                } else {
+                    item.erase("status");
+                    item.erase("type");
+                    item["content"] = chatcmpl_content;
+                    chatcmpl_messages.push_back(item);
+                }
             } else if (exists_and_is_string(item, "arguments") &&
                 exists_and_is_string(item, "call_id") &&
                 exists_and_is_string(item, "name") &&
@@ -1221,24 +1232,27 @@ json convert_responses_to_chatcmpl(const json & response_body) {
                 item.at("type") == "function_call"
             ) {
                 // #responses_create-input-input_item_list-item-function_tool_call
-                json msg = json {
-                    {"role", "assistant"},
-                    {"tool_calls", json::array({ json {
-                        {"function", json {
-                            {"arguments", item.at("arguments")},
-                            {"name",      item.at("name")},
-                        }},
-                        {"id",   item.at("call_id")},
-                        {"type", "function"},
-                    }})},
+                json tool_call = {
+                    {"function", json {
+                        {"arguments", item.at("arguments")},
+                        {"name",      item.at("name")},
+                    }},
+                    {"id",   item.at("call_id")},
+                    {"type", "function"},
                 };
 
-                if (!chatcmpl_messages.empty() && chatcmpl_messages.back().contains("reasoning_content")) {
-                    // Move reasoning content from dummy message to tool call message
-                    msg["reasoning_content"] = chatcmpl_messages.back().at("reasoning_content");
-                    chatcmpl_messages.pop_back();
+                if (merge_prev) {
+                    auto & prev_msg = chatcmpl_messages.back();
+                    if (!exists_and_is_array(prev_msg, "tool_calls")) {
+                        prev_msg["tool_calls"] = json::array();
+                    }
+                    prev_msg["tool_calls"].push_back(tool_call);
+                } else {
+                    chatcmpl_messages.push_back(json {
+                        {"role",       "assistant"},
+                        {"tool_calls", json::array({tool_call})}
+                    });
                 }
-                chatcmpl_messages.push_back(msg);
             } else if (exists_and_is_string(item, "call_id") &&
                 (exists_and_is_string(item, "output") || exists_and_is_array(item, "output")) &&
                 exists_and_is_string(item, "type") &&
@@ -1282,12 +1296,16 @@ json convert_responses_to_chatcmpl(const json & response_body) {
                     throw std::invalid_argument("item['content']['text'] is not a string");
                 }
 
-                // Pack reasoning content in dummy message
-                chatcmpl_messages.push_back(json {
-                    {"role", "assistant"},
-                    {"content", json::array()},
-                    {"reasoning_content", item.at("content")[0].at("text")},
-                });
+                if (merge_prev) {
+                    auto & prev_msg = chatcmpl_messages.back();
+                    prev_msg["reasoning_content"] = item.at("content")[0].at("text");
+                } else {
+                    chatcmpl_messages.push_back(json {
+                        {"role", "assistant"},
+                        {"content", json::array()},
+                        {"reasoning_content", item.at("content")[0].at("text")},
+                    });
+                }
             } else {
                 throw std::invalid_argument("Cannot determine type of 'item'");
             }
@@ -1296,20 +1314,6 @@ json convert_responses_to_chatcmpl(const json & response_body) {
         throw std::invalid_argument("'input' must be a string or array of objects");
     }
 
-    // Remove unused dummy message which contains
-    // reasoning content not followed by tool call
-    chatcmpl_messages.erase(std::remove_if(
-        chatcmpl_messages.begin(),
-        chatcmpl_messages.end(),
-        [](const json & x){ return x.contains("role") &&
-            x.at("role") == "assistant" &&
-            x.contains("content") &&
-            x.at("content") == json::array() &&
-            x.contains("reasoning_content");
-        }),
-        chatcmpl_messages.end()
-    );
-
     chatcmpl_body["messages"] = chatcmpl_messages;
 
     if (response_body.contains("tools")) {

tools/server/server-context.cpp

@@ -2911,6 +2911,9 @@ server_context_meta server_context::get_meta() const {
         /* fim_pre_token          */ llama_vocab_fim_pre(impl->vocab),
         /* fim_sub_token          */ llama_vocab_fim_suf(impl->vocab),
         /* fim_mid_token          */ llama_vocab_fim_mid(impl->vocab),
+        /* fim_pad_token          */ llama_vocab_fim_pad(impl->vocab),
+        /* fim_rep_token          */ llama_vocab_fim_rep(impl->vocab),
+        /* fim_sep_token          */ llama_vocab_fim_sep(impl->vocab),
 
         /* model_vocab_type       */ llama_vocab_type(impl->vocab),
         /* model_vocab_n_tokens   */ llama_vocab_n_tokens(impl->vocab),

tools/server/server-context.h

@@ -30,6 +30,9 @@ struct server_context_meta {
     llama_token fim_pre_token;
     llama_token fim_sub_token;
     llama_token fim_mid_token;
+    llama_token fim_pad_token;
+    llama_token fim_rep_token;
+    llama_token fim_sep_token;
 
     // model meta
     enum llama_vocab_type model_vocab_type;

tools/server/server-task.cpp

@@ -204,7 +204,8 @@ task_params server_task::params_from_json_cmpl(
     params.cache_prompt     = json_value(data,       "cache_prompt",       defaults.cache_prompt);
     params.return_tokens    = json_value(data,       "return_tokens",      false);
     params.return_progress  = json_value(data,       "return_progress",    false);
-    params.n_predict        = json_value(data,       "n_predict",          json_value(data, "max_tokens", defaults.n_predict));
+    auto max_tokens         = json_value(data,       "max_tokens",         defaults.n_predict);
+    params.n_predict        = json_value(data,       "n_predict",          json_value(data, "max_completion_tokens", max_tokens));
     params.n_indent         = json_value(data,       "n_indent",           defaults.n_indent);
     params.n_keep           = json_value(data,       "n_keep",             defaults.n_keep);
     params.n_discard        = json_value(data,       "n_discard",          defaults.n_discard);

tools/server/webui/README.md

@@ -101,7 +101,7 @@ In a separate terminal, start the backend server:
 ./llama-server -m model.gguf
 
 # Multi-model (ROUTER mode)
-./llama-server --model-store /path/to/models
+./llama-server --models-dir /path/to/models
 ```
 
 ### 3. Start Development Servers

tools/server/webui/src/lib/components/app/chat/ChatSettings/ChatSettings.svelte

@@ -114,6 +114,11 @@
 					label: 'Render user content as Markdown',
 					type: SettingsFieldType.CHECKBOX
 				},
+				{
+					key: SETTINGS_KEYS.FULL_HEIGHT_CODE_BLOCKS,
+					label: 'Use full height code blocks',
+					type: SettingsFieldType.CHECKBOX
+				},
 				{
 					key: SETTINGS_KEYS.DISABLE_AUTO_SCROLL,
 					label: 'Disable automatic scroll',

tools/server/webui/src/lib/components/app/content/MarkdownContent.svelte

@@ -38,6 +38,8 @@
 	import { ActionIconsCodeBlock, DialogCodePreview } from '$lib/components/app';
 	import { createAutoScrollController } from '$lib/hooks/use-auto-scroll.svelte';
 	import type { DatabaseMessageExtra } from '$lib/types/database';
+	import { config } from '$lib/stores/settings.svelte';
+	import { SETTINGS_KEYS } from '$lib/constants/settings-keys';
 
 	interface Props {
 		attachments?: DatabaseMessageExtra[];
@@ -593,7 +595,12 @@
 	});
 </script>
 
-<div bind:this={containerRef} class={className}>
+<div
+	bind:this={containerRef}
+	class="{className}{config()[SETTINGS_KEYS.FULL_HEIGHT_CODE_BLOCKS]
+		? ' full-height-code-blocks'
+		: ''}"
+>
 	{#each renderedBlocks as block (block.id)}
 		<div class="markdown-block" data-block-id={block.id}>
 			<!-- eslint-disable-next-line no-at-html-tags -->
@@ -914,6 +921,16 @@
 		line-height: 1.3;
 	}
 
+	.full-height-code-blocks :global(.code-block-wrapper) {
+		max-height: none;
+	}
+
+	.full-height-code-blocks :global(.code-block-scroll-container),
+	.full-height-code-blocks .streaming-code-scroll-container {
+		max-height: none;
+		overflow-y: visible;
+	}
+
 	div :global(.code-block-header) {
 		display: flex;
 		justify-content: space-between;

tools/server/webui/src/lib/components/app/models/ModelsSelector.svelte

@@ -42,7 +42,13 @@
 		useGlobalSelection = false
 	}: Props = $props();
 
-	let options = $derived(modelOptions());
+	let options = $derived(
+		modelOptions().filter((option) => {
+			const modelProps = modelsStore.getModelProps(option.model);
+
+			return modelProps?.webui !== false;
+		})
+	);
 	let loading = $derived(modelsLoading());
 	let updating = $derived(modelsUpdating());
 	let activeId = $derived(selectedModelId());

tools/server/webui/src/lib/constants/settings-config.ts

@@ -22,6 +22,7 @@ export const SETTING_CONFIG_DEFAULT: Record<string, string | number | boolean> =
 	alwaysShowSidebarOnDesktop: false,
 	autoShowSidebarOnNewChat: true,
 	autoMicOnEmpty: false,
+	fullHeightCodeBlocks: false,
 	// make sure these default values are in sync with `common.h`
 	samplers: 'top_k;typ_p;top_p;min_p;temperature',
 	backend_sampling: false,
@@ -113,6 +114,8 @@ export const SETTING_CONFIG_INFO: Record<string, string> = {
 		'Automatically show sidebar when starting a new chat. Disable to keep the sidebar hidden until you click on it.',
 	autoMicOnEmpty:
 		'Automatically show microphone button instead of send button when textarea is empty for models with audio modality support.',
+	fullHeightCodeBlocks:
+		'Always display code blocks at their full natural height, overriding any height limits.',
 	pyInterpreterEnabled:
 		'Enable Python interpreter using Pyodide. Allows running Python code in markdown code blocks.',
 	enableContinueGeneration:

tools/server/webui/src/lib/constants/settings-keys.ts

@@ -23,6 +23,7 @@ export const SETTINGS_KEYS = {
 	DISABLE_AUTO_SCROLL: 'disableAutoScroll',
 	ALWAYS_SHOW_SIDEBAR_ON_DESKTOP: 'alwaysShowSidebarOnDesktop',
 	AUTO_SHOW_SIDEBAR_ON_NEW_CHAT: 'autoShowSidebarOnNewChat',
+	FULL_HEIGHT_CODE_BLOCKS: 'fullHeightCodeBlocks',
 	// Sampling
 	TEMPERATURE: 'temperature',
 	DYNATEMP_RANGE: 'dynatemp_range',

tools/server/webui/src/lib/services/parameter-sync.service.ts

@@ -153,6 +153,12 @@ export const SYNCABLE_PARAMETERS: SyncableParameter[] = [
 		serverKey: 'enableContinueGeneration',
 		type: SyncableParameterType.BOOLEAN,
 		canSync: true
+	},
+	{
+		key: 'fullHeightCodeBlocks',
+		serverKey: 'fullHeightCodeBlocks',
+		type: SyncableParameterType.BOOLEAN,
+		canSync: true
 	}
 ];