145cd4f475
fixed a typo
2026-09-14 11:09:24 +08:00
1c1cc3398b
handle dsv4 tool call bug
2026-09-14 10:41:18 +08:00
1c3cbac085
Merge commit 'ae9afff8d2c012ca760eb9c2adf41961cf6f6232' into concedo_experimental
...
# Conflicts:
# .github/workflows/build-openvino.yml
# .github/workflows/build-self-hosted.yml
# .github/workflows/build-vulkan.yml
# .github/workflows/build-wasm.yml
# .github/workflows/build-webgpu.yml
# .github/workflows/server-sanitize.yml
# CMakeLists.txt
# ci/run.sh
# common/CMakeLists.txt
# docs/backend/snapdragon/README.md
# docs/backend/snapdragon/developer.md
# docs/build.md
# examples/speculative-simple/speculative-simple.cpp
# ggml/src/ggml-cpu/CMakeLists.txt
# ggml/src/ggml-hexagon/ggml-hexagon.cpp
# ggml/src/ggml-hexagon/htp-opnode.h
# ggml/src/ggml-hexagon/htp/act-ops.c
# ggml/src/ggml-hexagon/htp/allreduce-ops.c
# ggml/src/ggml-hexagon/htp/allreduce-ops.h
# ggml/src/ggml-hexagon/htp/argsort-ops.c
# ggml/src/ggml-hexagon/htp/binary-ops.c
# ggml/src/ggml-hexagon/htp/concat-ops.c
# ggml/src/ggml-hexagon/htp/cpy-ops.c
# ggml/src/ggml-hexagon/htp/cumsum-ops.c
# ggml/src/ggml-hexagon/htp/diag-ops.c
# ggml/src/ggml-hexagon/htp/fill-ops.c
# ggml/src/ggml-hexagon/htp/flash-attn-ops.c
# ggml/src/ggml-hexagon/htp/flash-attn-ops.h
# ggml/src/ggml-hexagon/htp/gated-delta-net-ops.c
# ggml/src/ggml-hexagon/htp/get-rows-ops.c
# ggml/src/ggml-hexagon/htp/hex-common.h
# ggml/src/ggml-hexagon/htp/hex-utils.h
# ggml/src/ggml-hexagon/htp/hmx-utils.h
# ggml/src/ggml-hexagon/htp/htp-ctx.h
# ggml/src/ggml-hexagon/htp/htp-ops.h
# ggml/src/ggml-hexagon/htp/htp-tensor.c
# ggml/src/ggml-hexagon/htp/htp-tensor.h
# ggml/src/ggml-hexagon/htp/hvx-arith.h
# ggml/src/ggml-hexagon/htp/hvx-div.h
# ggml/src/ggml-hexagon/htp/hvx-inverse.h
# ggml/src/ggml-hexagon/htp/hvx-scale.h
# ggml/src/ggml-hexagon/htp/hvx-sigmoid.h
# ggml/src/ggml-hexagon/htp/im2col-ops.c
# ggml/src/ggml-hexagon/htp/main.c
# ggml/src/ggml-hexagon/htp/matmul-ops.c
# ggml/src/ggml-hexagon/htp/pad-ops.c
# ggml/src/ggml-hexagon/htp/repeat-ops.c
# ggml/src/ggml-hexagon/htp/rope-ops.c
# ggml/src/ggml-hexagon/htp/set-rows-ops.c
# ggml/src/ggml-hexagon/htp/softmax-ops.c
# ggml/src/ggml-hexagon/htp/solve-tri-ops.c
# ggml/src/ggml-hexagon/htp/ssm-conv.c
# ggml/src/ggml-hexagon/htp/sum-rows-ops.c
# ggml/src/ggml-hexagon/htp/unary-ops.c
# ggml/src/ggml-metal/CMakeLists.txt
# ggml/src/ggml-opencl/CMakeLists.txt
# ggml/src/ggml-opencl/ggml-opencl.cpp
# ggml/src/ggml-rpc/CMakeLists.txt
# ggml/src/ggml-sycl/ggml-sycl.cpp
# ggml/src/ggml-webgpu/CMakeLists.txt
# ggml/src/ggml-webgpu/ggml-webgpu.cpp
# scripts/snapdragon/run.py
# scripts/sync_vendor.py
# scripts/ui-assets.cmake
# src/CMakeLists.txt
# src/models/gemma4.cpp
# tests/.gitignore
# tests/CMakeLists.txt
# tests/test-backend-ops.cpp
# tests/test-jinja.cpp
# tests/test-llama-archs.cpp
# tests/test-quantize-fns.cpp
# tests/test-save-load-state.cpp
# tools/mtmd/CMakeLists.txt
# tools/server/CMakeLists.txt
2026-09-13 23:53:41 +08:00
874f833a54
stability and memory access fixes (codex generated/reviewed)
2026-09-13 22:27:35 +08:00
49b9132287
prevent segfault on load fail
2026-09-13 22:01:18 +08:00
9625b44b38
fixed adaptive p
2026-09-09 22:17:59 +08:00
8e6ba2630b
add ffn cpu flag
2026-09-06 18:08:03 +08:00
6085db16cf
fix streaming race condition
2026-08-30 18:04:49 +08:00
ae60e1d6e0
Merge commit 'deae5ee133a3c4c56fbd46c17c8c2103af3bd643' into concedo_experimental
...
# Conflicts:
# .devops/openvino.Dockerfile
# .github/workflows/build-apple.yml
# .github/workflows/build-cuda-ubuntu.yml
# .github/workflows/docker.yml
# .github/workflows/release.yml
# .github/workflows/server-sanitize.yml
# .github/workflows/ui-build-self-hosted.yml
# .github/workflows/ui-build.yml
# .github/workflows/ui-publish.yml
# .github/workflows/ui-self-hosted.yml
# .github/workflows/ui.yml
# CMakeLists.txt
# docs/backend/snapdragon/CMakeUserPresets.json
# docs/backend/snapdragon/README.md
# docs/backend/snapdragon/developer.md
# docs/backend/snapdragon/linux.md
# docs/backend/snapdragon/windows.md
# docs/ops.md
# docs/ops/Vulkan.csv
# ggml/cmake/ggml-config.cmake.in
# ggml/src/ggml-cpu/CMakeLists.txt
# ggml/src/ggml-cpu/kleidiai/kernels.cpp
# ggml/src/ggml-cpu/kleidiai/kernels.h
# ggml/src/ggml-cpu/kleidiai/kleidiai.cpp
# ggml/src/ggml-hexagon/ggml-hexagon.cpp
# ggml/src/ggml-hexagon/htp-opnode.h
# ggml/src/ggml-hexagon/htp/CMakeLists.txt
# ggml/src/ggml-hexagon/htp/act-ops.c
# ggml/src/ggml-hexagon/htp/cpy-ops.c
# ggml/src/ggml-hexagon/htp/dma-queue.h
# ggml/src/ggml-hexagon/htp/flash-attn-ops.c
# ggml/src/ggml-hexagon/htp/get-rows-ops.c
# ggml/src/ggml-hexagon/htp/hex-utils.h
# ggml/src/ggml-hexagon/htp/htp-ctx.h
# ggml/src/ggml-hexagon/htp/htp-ops.h
# ggml/src/ggml-hexagon/htp/htp-tensor.c
# ggml/src/ggml-hexagon/htp/htp-tensor.h
# ggml/src/ggml-hexagon/htp/hvx-arith.h
# ggml/src/ggml-hexagon/htp/main.c
# ggml/src/ggml-hexagon/htp/matmul-ops.c
# ggml/src/ggml-hexagon/htp/matmul-ops.h
# ggml/src/ggml-hexagon/htp/set-rows-ops.c
# ggml/src/ggml-rpc/CMakeLists.txt
# scripts/snapdragon/ggml-hexagon-profile.py
# scripts/snapdragon/ggml-hexagon-trace.py
# tests/test-backend-ops.cpp
# tools/cli/README.md
# tools/llama-bench/llama-bench.cpp
# tools/rpc/README.md
# tools/server/README.md
# tools/ui/tests/stories/a11y/ChatScreenForm.a11y.stories.svelte
2026-08-28 19:31:29 +08:00
4447017602
Merge branch 'upstream' into concedo_experimental
...
# Conflicts:
# .github/actions/ccache-clear/action.yml
# .github/workflows/build-apple.yml
# .github/workflows/build-cpu.yml
# .github/workflows/build-cuda-ubuntu.yml
# .github/workflows/build-opencl.yml
# .github/workflows/build-openvino.yml
# .github/workflows/build-sycl.yml
# .github/workflows/build-vulkan.yml
# .github/workflows/build-wasm.yml
# .github/workflows/build-webgpu.yml
# .github/workflows/hip-quality-check.yml
# .github/workflows/server.yml
# CONTRIBUTING.md
# README.md
# ci/run.sh
# common/CMakeLists.txt
# common/chat.cpp
# docs/autoparser.md
# ggml/src/ggml-webgpu/wgsl-shaders/flash_attn.wgsl
# ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_tile.wgsl
# ggml/src/ggml-webgpu/wgsl-shaders/flash_attn_vec_split.wgsl
# scripts/sync_vendor.py
# tests/CMakeLists.txt
# tests/peg-parser/test-json-serialization.cpp
# tests/peg-parser/tests.h
# tests/test-chat-auto-parser.cpp
# tests/test-chat-peg-parser.cpp
# tests/test-chat-template.cpp
# tests/test-chat.cpp
# tests/test-grammar-integration.cpp
# tests/test-jinja.cpp
# tests/test-json-schema-to-grammar.cpp
# tests/test-llama-archs.cpp
# tests/test-model-resolution.cpp
# tests/test-recurrent-state-rollback.cpp
# tools/CMakeLists.txt
2026-08-25 20:44:34 +08:00
3265983620
Merge commit '2fb989b9e79bf4da8159855e24892c8f4c20300f' into concedo_experimental
...
# Conflicts:
# .github/actions/ccache-clear/action.yml
# .github/workflows/build-cpu.yml
# .github/workflows/make-release.yml
# .github/workflows/release.yml
# .pi/gg/SYSTEM.md
# AGENTS.md
# README.md
# docs/build.md
# ggml/src/ggml-opencl/CMakeLists.txt
# ggml/src/ggml-opencl/ggml-opencl.cpp
# ggml/src/ggml-opencl/kernels/moe_combine.cl
# ggml/src/ggml-sycl/convert.cpp
# ggml/src/ggml-sycl/dequantize.hpp
# ggml/src/ggml-sycl/dmmv.cpp
# ggml/src/ggml-sycl/esimd.hpp
# ggml/src/ggml-sycl/ggml-sycl.cpp
# ggml/src/ggml-sycl/mmvq.cpp
# ggml/src/ggml-sycl/quants.hpp
# ggml/src/ggml-sycl/vecdotq.hpp
# src/CMakeLists.txt
# tests/test-llama-archs.cpp
# tools/llama-bench/llama-bench.cpp
# tools/mtmd/CMakeLists.txt
2026-08-25 20:30:15 +08:00
fe36aa1959
add directio support
2026-08-24 22:59:18 +08:00
b06f828505
mmap and mlock independent
2026-08-24 22:29:10 +08:00
2b78904b4a
bug fixes for backend code
2026-08-22 17:28:48 +08:00
b9b3bbcec1
refactor: add new module for backend-specific code ( #2308 )
...
* refactor: add new module for backend-specific code
* replace ifdefs on rwkv_v3
* replace ifdefs on llama_v2
* replace ifdefs on llama_v3
* replace ifdefs on gpt2_v3
* replace ifdefs on gptj_v3
* replace ifdefs on mpt_v3
* replace ifdefs on neox_v3
* replace ifdefs on whisper
* adjust build for kcpp_backend
Mostly a:
sed 's/ gpttype_adapter\([^ ]*\)\.o / gpttype_adapter_default.o kcpp_backend\1.o /g' Makefile
and include variant objects for kcpp_backend.
2026-08-22 10:00:00 +08:00
531cac9cd3
fixed dflash
2026-08-14 22:38:55 +08:00
22806c84d3
fixed prints
2026-08-14 22:27:12 +08:00
d135a116da
dflash works with muse glimmer now
2026-08-11 18:25:25 +08:00
db8b62dc9b
fix more debug printing
2026-08-11 18:18:09 +08:00
6473847b60
wip dflash support
2026-08-11 18:16:33 +08:00
7eb3a3405a
fixed some stdout prints
2026-08-11 18:11:51 +08:00
92c526d0e8
fix mtp
2026-08-11 17:59:16 +08:00
7fd4acc35c
reasoning budget for muse glimmer
2026-08-11 15:43:23 +08:00
b39ff27d6f
muse glimmer jinja and tool calls working
2026-08-11 15:06:30 +08:00
8a16f96307
Merge branch 'upstream' into concedo_experimental
...
# Conflicts:
# .github/workflows/build-apple.yml
# .github/workflows/build-self-hosted.yml
# .github/workflows/release.yml
# SECURITY.md
# build-xcframework.sh
# ci/run.sh
# docs/development/HOWTO-add-model.md
# examples/model-conversion/scripts/causal/convert-model.sh
# examples/model-conversion/scripts/embedding/convert-model.sh
# scripts/sync_vendor.py
# scripts/ui-assets.cmake
# tests/test-arg-parser.cpp
# tests/test-backend-sampler.cpp
# tests/test-grammar-parser.cpp
# tests/test-llama-archs.cpp
# tests/test-sampling.cpp
# tools/cli/README.md
# tools/completion/README.md
# tools/mtmd/CMakeLists.txt
# tools/mtmd/mtmd.h
# tools/mtmd/tests/test-deepseek-ocr.py
# tools/server/README.md
# tools/tts/CMakeLists.txt
# tools/tts/convert_pt_to_hf.py
2026-08-07 20:46:56 +08:00
152e080b6a
Add Mistral [THINK]/[/THINK] thinking format (mistral3 arch) ( #2380 )
...
The reasoning budget derived from reasoning_effort never applied to Mistral
models. gpttype_adapter.cpp picks the think delimiters from a switch on the
model architecture, and mistral3 has no case, so it falls back to <think> /
</think>. Those are not vocabulary tokens for Ministral-3, so TokenizeString
returns more than one token each, the expected_start/end_tokens guard clears
all three vectors, and apply_reasoning_budget() returns at its first if.
The parameter is accepted, converted and passed down to the sampler, then
dropped on a size check, with nothing logged.
Adding the mistral3 case arms the budget. [THINK] and [/THINK] are single
vocabulary tokens (ids 34 and 35 on Ministral-3), so the size guard passes.
The thinkformats entry is a separate fix for a separate defect: without it the
thinking block was never split out, so it leaked into content with its [THINK]
marker still in it, instead of going to reasoning_content.
Measured on Ministral-3-14B-Reasoning-2512 (IQ4_XS, ctx 8192, --jinja), 5 real
prompts x 3 samples per cell, max_tokens 3000 (so a 750-token budget at "low"):
reasoning_effort thinking words before thinking words after
none 311 - 2314 7 (the forced-close phrase)
low 340 - 2255 521 - 574
Forced closes: 0/15 before, 14/15 after at "low" and 15/15 at "none". Three
samples per cell because this model's variance at temperature 0.7 spans a
factor of 4 on an identical payload — a single sample per cell cannot tell an
effect from noise.
No regression on a non-reasoning mistral3 model: Ministral-3-8B-Instruct with
reasoning_effort "low" returns finish_reason "stop", a normal answer and zero
forced closes, since apply_reasoning_budget() bails out when the start marker
never appears.
2026-08-05 18:51:55 +08:00
304bd3d119
no host by default
2026-07-27 21:03:34 +08:00
49dbdaaab5
Merge branch 'upstream' into concedo_experimental
...
# Conflicts:
# AGENTS.md
# CODEOWNERS
# CONTRIBUTING.md
# docs/backend/OPENCL.md
# docs/development/HOWTO-add-model.md
# examples/training/finetune.cpp
# ggml/src/ggml-hexagon/ggml-hexagon.cpp
# ggml/src/ggml-hexagon/htp-drv.cpp
# ggml/src/ggml-hexagon/htp/act-ops.c
# ggml/src/ggml-hexagon/htp/dma-queue.c
# ggml/src/ggml-hexagon/htp/dma-queue.h
# ggml/src/ggml-hexagon/htp/flash-attn-ops.c
# ggml/src/ggml-hexagon/htp/flash-attn-ops.h
# ggml/src/ggml-hexagon/htp/hmx-mm-kernels-tiled.h
# ggml/src/ggml-hexagon/htp/htp-ctx.h
# ggml/src/ggml-hexagon/htp/htp-ops.h
# ggml/src/ggml-hexagon/htp/htp-tensor.c
# ggml/src/ggml-hexagon/htp/htp-tensor.h
# ggml/src/ggml-hexagon/htp/hvx-fa-kernels.h
# ggml/src/ggml-hexagon/htp/hvx-reduce.h
# ggml/src/ggml-hexagon/htp/main.c
# ggml/src/ggml-hexagon/htp/matmul-ops.c
# ggml/src/ggml-hexagon/htp/matmul-ops.h
# ggml/src/ggml-hexagon/htp/unary-ops.c
# ggml/src/ggml-hexagon/htp/unary-ops.h
# ggml/src/ggml-opencl/CMakeLists.txt
# ggml/src/ggml-opencl/ggml-opencl.cpp
# scripts/compare-llama-bench.py
# scripts/snapdragon/ggml-hexagon-profile.py
# scripts/snapdragon/ggml-hexagon-trace.py
# scripts/sync_vendor.py
# tests/test-arg-parser.cpp
# tests/test-chat.cpp
# tests/test-model-load-cancel.cpp
# tests/test-quantize-stats.cpp
# tools/cli/README.md
# tools/completion/README.md
# tools/llama-bench/llama-bench.cpp
# tools/server/README.md
# tools/ui/src/lib/constants/settings-registry.ts
2026-07-25 12:20:51 +08:00
917b379cb2
FIX: mtmd_tokenize: error ( #2360 )
...
DEEPSEEK4 SLOP Explanation. (Tested and is working correctly after build)
fix mtmd: missing text_len in gpttype_adapter.cpp aggregate init
Commit 4114ba18b#25548 )") added a `text_len` field to `mtmd_input_text`, changing the
struct layout from:
{ text, add_special, parse_special }
to:
{ text, text_len, add_special, parse_special }
The aggregate initialization in `gpttype_adapter.cpp` was never updated
to account for the new field. With C++ aggregate init, the positional
arguments shifted:
mtmd_input_text inp_txt = {
mtmd_default_marker(), // text = "<__media__>"
/* add_special */ false, // text_len = 0 <-- BUG
/* parse_special */ true, // add_special = true
};
// parse_special is zero-initialized
Because text_len was 0, `input_text.assign(text->text, text->text_len)`
in the mtmd_tokenizer constructor produced an empty string.
`split_text()` on an empty string returned an empty vector, yielding
zero media markers in the text -- but one bitmap was still provided,
triggering the error:
mtmd_tokenize: error: number of media markers in text (0) does not
match number of bitmaps (1)
Fix by explicitly capturing the marker pointer and passing its length.
2026-07-24 20:55:58 +08:00
20e5b61cf8
better timing info in batched mode
2026-07-18 19:14:01 +08:00
a25af27185
attempt to fix dsv4
2026-07-17 20:48:27 +08:00
91b1e1c89a
fix speculative logging
2026-07-13 18:29:21 +08:00
a4aa063153
vision fix
2026-07-02 22:30:21 +08:00
e2de771b2a
fixed a missing header
2026-07-01 20:30:25 +08:00
849ec89bad
restructure some compilation units
2026-07-01 18:51:25 +08:00
983dec9a54
prevent MTP drafting with batching
2026-07-01 16:17:13 +08:00
1365d11990
context shifting now works with images, images are now inserted inline instead of as placeholders
2026-06-29 23:40:43 +08:00
ae3c1b6a19
fix vision regression
2026-06-26 23:01:30 +08:00
145beb5744
try to use llama.cpp's tool call parser first
2026-06-25 18:11:44 +08:00
e975ad6854
Cap n_outputs_max on MTP draft contexts ( #2287 )
...
Co-authored-by: Pento95 <Pento95@users.noreply.github.com >
2026-06-25 13:37:22 +08:00
19064083bd
another fix for drafting
2026-06-23 22:34:34 +08:00
6df4ca13f1
drafting tweak
2026-06-23 20:11:25 +08:00
4a7d6dd8a0
alias for draft
2026-06-23 18:43:28 +08:00
7fe6fa6fb6
match draft defaults
2026-06-22 21:40:00 +08:00
a0f39fe0f5
allow drafting with vision
2026-06-22 20:50:39 +08:00
e4771e8e6b
restore draft state before main state, fixes reloading gemma4 assistant
2026-06-22 18:55:26 +08:00
f202c0a457
mtp init -2
2026-06-21 10:21:48 +08:00
6f4325ac87
support mp3 from api
2026-06-19 10:26:34 +08:00
635c45e1a0
fix incorrect mtp layers setting (+1 squashed commits)
...
Squashed commits:
[8dad1a5c0] fix incorrect mtp layers setting
2026-06-18 21:10:28 +08:00
197937aab0
cohere think budget
2026-06-16 01:03:24 +08:00
Concedo
Wagner Bruna
Julien BODIN
askmyteapot
Pento