Merge branch 'ggml-org:master' into master

Author: Thireus

common/console.cpp

@@ -700,13 +700,13 @@ namespace console {
         std::vector<std::string> entries;
         size_t viewing_idx = SIZE_MAX;
         std::string backup_line; // current line before viewing history
-        void add(const std::string & line) {
+        void add(std::string_view line) {
             if (line.empty()) {
                 return;
             }
             // avoid duplicates with the last entry
             if (entries.empty() || entries.back() != line) {
-                entries.push_back(line);
+                entries.emplace_back(line);
             }
             // also clear viewing state
             end_viewing();
@@ -1031,11 +1031,12 @@ namespace console {
 
         if (!end_of_stream && !line.empty()) {
             // remove the trailing newline for history storage
+            std::string_view hline = line;
             if (!line.empty() && line.back() == '\n') {
-                line.pop_back();
+                hline.remove_suffix(1);
             }
             // TODO: maybe support multiline history entries?
-            history.add(line);
+            history.add(hline);
         }
 
         fflush(out);

convert_hf_to_gguf.py

@@ -7472,7 +7472,7 @@ def set_vocab(self):
         special_vocab = gguf.SpecialVocab(self.dir_model, load_merges=True)
         special_vocab.add_to_gguf(self.gguf_writer)
         self.gguf_writer.add_add_space_prefix(False)
-        self.gguf_writer.add_add_bos_token(False) # already added via the chat template
+        self.gguf_writer.add_add_bos_token(True)
 
     def set_gguf_parameters(self):
         super().set_gguf_parameters()
@@ -11818,10 +11818,8 @@ class LFM2Model(TextModel):
     model_arch = gguf.MODEL_ARCH.LFM2
 
     def _add_feed_forward_length(self):
-        ff_dim = self.hparams["block_ff_dim"]
-
+        ff_dim = self.find_hparam(["block_ff_dim", "intermediate_size"])
         auto_adjust_ff_dim = self.hparams["block_auto_adjust_ff_dim"]
-        ff_dim = self.hparams["block_ff_dim"]
         ffn_dim_multiplier = self.hparams["block_ffn_dim_multiplier"]
         multiple_of = self.hparams["block_multiple_of"]
 

docs/build.md

@@ -741,7 +741,7 @@ cmake --build build --config Release
 
 WebGPU allows cross-platform access to the GPU from supported browsers. We utilize [Emscripten](https://emscripten.org/) to compile ggml's WebGPU backend to WebAssembly. Emscripten does not officially support WebGPU bindings yet, but Dawn currently maintains its own WebGPU bindings called emdawnwebgpu.
 
-Follow the instructions [here](https://dawn.googlesource.com/dawn/+/refs/heads/main/src/emdawnwebgpu/) to download or build the emdawnwebgpu package (Note that it might be safer to build the emdawbwebgpu package locally, so that it stays in sync with the version of Dawn you have installed above). When building using CMake, the path to the emdawnwebgpu port file needs to be set with the flag `EMDAWNWEBGPU_DIR`.
+Follow the instructions [here](https://dawn.googlesource.com/dawn/+/refs/heads/main/src/emdawnwebgpu/) to download or build the emdawnwebgpu package (Note that it might be safer to build the emdawnwebgpu package locally, so that it stays in sync with the version of Dawn you have installed above). When building using CMake, the path to the emdawnwebgpu port file needs to be set with the flag `EMDAWNWEBGPU_DIR`.
 
 ## IBM Z & LinuxONE
 

docs/multimodal.md

@@ -37,6 +37,7 @@ llama-server -hf ggml-org/gemma-3-4b-it-GGUF --no-mmproj-offload
 > - PaddleOCR-VL: https://github.com/ggml-org/llama.cpp/pull/18825
 > - GLM-OCR: https://github.com/ggml-org/llama.cpp/pull/19677
 > - Deepseek-OCR: https://github.com/ggml-org/llama.cpp/pull/17400
+> - HunyuanOCR: https://github.com/ggml-org/llama.cpp/pull/21395
 
 ## Pre-quantized models
 

docs/ops.md

@@ -68,7 +68,7 @@ Legend:
 |                             MEAN | ❌ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
 |                              MUL | ❌ | ✅ | ✅ | ✅ | 🟡 | ✅ | ✅ | ✅ | ✅ | ❌ | ❌ |
 |                          MUL_MAT | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 |
-|                       MUL_MAT_ID | ❌ | 🟡 | ✅ | ✅ | 🟡 | 🟡 | 🟡 | ✅ | ❌ | 🟡 | ❌ |
+|                       MUL_MAT_ID | ❌ | 🟡 | ✅ | ✅ | 🟡 | 🟡 | 🟡 | ✅ | 🟡 | 🟡 | ❌ |
 |                              NEG | ❌ | ✅ | ✅ | 🟡 | ✅ | ❌ | ✅ | 🟡 | ✅ | ❌ | ❌ |
 |                             NORM | ❌ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | 🟡 | ❌ | ❌ | ❌ |
 |                   OPT_STEP_ADAMW | ❌ | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ |