mtmd: some small clean up by ngxson · Pull Request #17909 · ggml-org/llama.cpp

ngxson · 2025-12-10T15:14:36Z

Some small clean up that I feel can be regrouped into one PR

Use build_ffn when possible
Add support for fused qkv in build_vit
In single-turn mode, move the image to the beginning, ref: model : Fix marker placement for LFM2-VL in single turn llama-mtmd-cli #17616
Accept custom args in test script

Test results:

[vision] OK:   ggml-org/SmolVLM-500M-Instruct-GGUF:Q8_0
[vision] OK:   ggml-org/SmolVLM2-2.2B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/SmolVLM2-500M-Video-Instruct-GGUF:Q8_0
[vision] OK:   ggml-org/gemma-3-4b-it-GGUF:Q4_K_M
[vision] OK:   THUDM/glm-edge-v-5b-gguf:Q4_K_M
[vision] OK:   second-state/Llava-v1.5-7B-GGUF:Q2_K
[vision] OK:   cjpais/llava-1.6-mistral-7b-gguf:Q3_K_M
[vision] OK:   ibm-research/granite-vision-3.2-2b-GGUF:Q4_K_M
[vision] OK:   second-state/MiniCPM-Llama3-V-2_5-GGUF:Q2_K
[vision] OK:   openbmb/MiniCPM-V-2_6-gguf:Q2_K
[vision] OK:   openbmb/MiniCPM-o-2_6-gguf:Q4_0
[vision] OK:   bartowski/Qwen2-VL-2B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/Qwen2.5-VL-3B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/InternVL2_5-1B-GGUF:Q8_0
[vision] OK:   ggml-org/InternVL3-1B-Instruct-GGUF:Q8_0
[vision] OK:   ggml-org/Qwen2.5-Omni-3B-GGUF:Q4_K_M
[vision] OK:   ggml-org/LFM2-VL-450M-GGUF:Q8_0
[vision] OK:   ggml-org/granite-docling-258M-GGUF:Q8_0
[vision] OK:   ggml-org/LightOnOCR-1B-1025-GGUF:Q8_0
[audio]  OK:   ggml-org/ultravox-v0_5-llama-3_2-1b-GGUF:Q8_0
[audio]  OK:   ggml-org/Qwen2.5-Omni-3B-GGUF:Q4_K_M
[audio]  OK:   ggml-org/Voxtral-Mini-3B-2507-GGUF:Q4_K_M
[vision] OK:   ggml-org/pixtral-12b-GGUF:Q4_K_M
[vision] OK:   ggml-org/Mistral-Small-3.1-24B-Instruct-2503-GGUF
[vision] OK:   ggml-org/Qwen2-VL-2B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/Qwen2-VL-7B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/Qwen2.5-VL-3B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/Qwen2.5-VL-7B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/Qwen3-VL-2B-Instruct-GGUF:Q8_0
[vision] OK:   ggml-org/InternVL3-8B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/InternVL3-14B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/Qwen2.5-Omni-7B-GGUF:Q4_K_M
[audio]  OK:   ggml-org/ultravox-v0_5-llama-3_1-8b-GGUF:Q4_K_M
[audio]  OK:   ggml-org/Qwen2.5-Omni-7B-GGUF:Q4_K_M
[vision] OK:   ggml-org/Qwen2.5-VL-72B-Instruct-GGUF:Q4_K_M
[vision] OK:   ggml-org/Llama-4-Scout-17B-16E-Instruct-GGUF:IQ1_S

added new opt to tests.sh to disable flash-attn

amritsingh183 · 2025-12-11T10:03:11Z

where is build_vit() defined?

sfallah · 2025-12-11T10:06:55Z

where is build_vit() defined?

tools/mtmd/clip.cpp

* clip: add support for fused qkv in build_vit * use bulid_ffn whenever possible * fix internvl * mtmd-cli: move image to beginning * test script: support custom args

* mtmd: llama.cpp DeepSeekOCR support init commit * loading sam tensors * mtmd: fix vision model processing * deepseek-ocr clip-vit model impl * mtmd: add DeepSeek-OCR LM support with standard attention * mtmd: successfully runs DeepSeek-OCR LM in llama-cli * mtmd: Fix RoPE type for DeepSeek-OCR LM. * loading LM testing Vision model loading * sam warmup working * sam erroneous return corrected * clip-vit: corrected cls_embd concat * clip-vit: model convert qkv_proj split * corrected combining of image encoders' results * fix: update callback for ffn_moe_weighted and add callback for attn_out in deepseek2 model * concat image_newline and image_seperator tokens * visual_model warmup (technically) works * window partitioning using standard ggml ops * sam implementation without using CPU only ops * clip: fixed warnings * Merge branch 'sf/deepseek-ocr' of github.com:sfallah/llama.cpp into sf/deepseek-ocr * mtmd: fix get_rel_pos * mtmd: fixed the wrong scaler for get_rel_pos * image encoding technically works but the output can't be checked singe image decoding fails * mtmd: minor changed * mtmd: add native resolution support * - image encoding debugged - issues fixed mainly related wrong config like n_patches etc. - configs need to be corrected in the converter * mtmd: correct token order * - dynamic resizing - changes are concerning PR sfallah#4 * mtmd: quick fix token order * mtmd: fix danling pointer * mtmd: SAM numerically works * mtmd: debug CLIP-L (vit_pre_ln) * mtmd: debug CLIP-L & first working DeepSeek-OCR model * mtmd : add --dsocr-mode CLI argument for DeepSeek-OCR resolution control & all native resolution modes work * mtmd: simplify SAM patch embedding * mtmd: adapt Pillow image resizing function * mtmd: simplify DeepSeek-OCR dynamic resolution preprocessing * mtmd: remove --dsocr-mode argument * mtmd: refactor code & remove unused helper functions * mtmd: fix tensor names for image newlines and view separator * clean up * reverting automatically removed spaces * reverting automatically removed spaces * mtmd: fixed bad ocr check in Deepseek2 (LM) * mtmd: support combined QKV projection in buid_vit * using common build_attn in sam * corrected code-branch when flash-attn disabled enabling usage of --flash-attn option * mtmd: minor fix * minor formatting and style * fixed flake8 lint issues * minor editorconfig-check fixes * minor editorconfig-check fixes * mtmd: simplify get_rel_pos * mtmd: make sam hparams configurable * mtmd: add detailed comments for resize_bicubic_pillow * mtmd: fixed wrong input setting * mtmd: convert model in FP16 * mtmd: minor fix * mtmd: remove tweak to llama-mtmd-cli & deepseek-ocr template * fix: test-1.jpg ORC issue with small (640) resolution setting min-resolution base (1024) max large (1280) for dynamic-resolution * minor: editconfig-check fix * merge with changes from #17909 added new opt to tests.sh to disable flash-attn * minor: editconfig-check fix * testing deepseek-ocr quick and dirty test script comparing results of Qwen2.5-VL vs DeepSeek-OCR * quick and (potential) dirty merge with #17909 * refactoring, one single builder function and static helpers * added deepseek-ocr test to tests.sh * minor formatting fixes * check with fixed expected resutls * minor formatting * editorconfig-check fix * merge with changes from #18042 * minor - added GLM-4.6V to big tests - added missing deps for python test * convert: minor fix * mtmd: format code * convert: quick fix * convert: quick fix * minor python formatting * fixed merge build issue * merge resolved - fixed issues in convert - tested several deepseek models * minor fix * minor * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * - removed clip_is_deepseekocr - removed redundant RESIZE_ALGO_BICUBIC_PILLOW resize-algo - simplified image-preprocessing - removed/simplified debug functions * - cleaning commented out code * fixing instabilities issues reintroducing resize_bicubic_pillow * - use f16 model for deepseek-ocr test - ignore llama-arch test for deepseek-ocr * rename fc_w --> mm_fc_w * add links to OCR discussion * cleaner loading code * add missing .weight to some tensors * add default jinja template (to be used by server) * move test model to ggml-org * rolling back upscale change * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> --------- Co-authored-by: bluebread <hotbread70127@gmail.com> Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> Co-authored-by: Xuan Son Nguyen <son@huggingface.co> Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

* mtmd: llama.cpp DeepSeekOCR support init commit * loading sam tensors * mtmd: fix vision model processing * deepseek-ocr clip-vit model impl * mtmd: add DeepSeek-OCR LM support with standard attention * mtmd: successfully runs DeepSeek-OCR LM in llama-cli * mtmd: Fix RoPE type for DeepSeek-OCR LM. * loading LM testing Vision model loading * sam warmup working * sam erroneous return corrected * clip-vit: corrected cls_embd concat * clip-vit: model convert qkv_proj split * corrected combining of image encoders' results * fix: update callback for ffn_moe_weighted and add callback for attn_out in deepseek2 model * concat image_newline and image_seperator tokens * visual_model warmup (technically) works * window partitioning using standard ggml ops * sam implementation without using CPU only ops * clip: fixed warnings * Merge branch 'sf/deepseek-ocr' of github.com:sfallah/llama.cpp into sf/deepseek-ocr * mtmd: fix get_rel_pos * mtmd: fixed the wrong scaler for get_rel_pos * image encoding technically works but the output can't be checked singe image decoding fails * mtmd: minor changed * mtmd: add native resolution support * - image encoding debugged - issues fixed mainly related wrong config like n_patches etc. - configs need to be corrected in the converter * mtmd: correct token order * - dynamic resizing - changes are concerning PR sfallah/llama.cpp#4 * mtmd: quick fix token order * mtmd: fix danling pointer * mtmd: SAM numerically works * mtmd: debug CLIP-L (vit_pre_ln) * mtmd: debug CLIP-L & first working DeepSeek-OCR model * mtmd : add --dsocr-mode CLI argument for DeepSeek-OCR resolution control & all native resolution modes work * mtmd: simplify SAM patch embedding * mtmd: adapt Pillow image resizing function * mtmd: simplify DeepSeek-OCR dynamic resolution preprocessing * mtmd: remove --dsocr-mode argument * mtmd: refactor code & remove unused helper functions * mtmd: fix tensor names for image newlines and view separator * clean up * reverting automatically removed spaces * reverting automatically removed spaces * mtmd: fixed bad ocr check in Deepseek2 (LM) * mtmd: support combined QKV projection in buid_vit * using common build_attn in sam * corrected code-branch when flash-attn disabled enabling usage of --flash-attn option * mtmd: minor fix * minor formatting and style * fixed flake8 lint issues * minor editorconfig-check fixes * minor editorconfig-check fixes * mtmd: simplify get_rel_pos * mtmd: make sam hparams configurable * mtmd: add detailed comments for resize_bicubic_pillow * mtmd: fixed wrong input setting * mtmd: convert model in FP16 * mtmd: minor fix * mtmd: remove tweak to llama-mtmd-cli & deepseek-ocr template * fix: test-1.jpg ORC issue with small (640) resolution setting min-resolution base (1024) max large (1280) for dynamic-resolution * minor: editconfig-check fix * merge with changes from ggml-org/llama.cpp#17909 added new opt to tests.sh to disable flash-attn * minor: editconfig-check fix * testing deepseek-ocr quick and dirty test script comparing results of Qwen2.5-VL vs DeepSeek-OCR * quick and (potential) dirty merge with ggml-org/llama.cpp#17909 * refactoring, one single builder function and static helpers * added deepseek-ocr test to tests.sh * minor formatting fixes * check with fixed expected resutls * minor formatting * editorconfig-check fix * merge with changes from ggml-org/llama.cpp#18042 * minor - added GLM-4.6V to big tests - added missing deps for python test * convert: minor fix * mtmd: format code * convert: quick fix * convert: quick fix * minor python formatting * fixed merge build issue * merge resolved - fixed issues in convert - tested several deepseek models * minor fix * minor * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * - removed clip_is_deepseekocr - removed redundant RESIZE_ALGO_BICUBIC_PILLOW resize-algo - simplified image-preprocessing - removed/simplified debug functions * - cleaning commented out code * fixing instabilities issues reintroducing resize_bicubic_pillow * - use f16 model for deepseek-ocr test - ignore llama-arch test for deepseek-ocr * rename fc_w --> mm_fc_w * add links to OCR discussion * cleaner loading code * add missing .weight to some tensors * add default jinja template (to be used by server) * move test model to ggml-org * rolling back upscale change * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> --------- Co-authored-by: bluebread <hotbread70127@gmail.com> Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> Co-authored-by: Xuan Son Nguyen <son@huggingface.co> Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

* mtmd: llama.cpp DeepSeekOCR support init commit * loading sam tensors * mtmd: fix vision model processing * deepseek-ocr clip-vit model impl * mtmd: add DeepSeek-OCR LM support with standard attention * mtmd: successfully runs DeepSeek-OCR LM in llama-cli * mtmd: Fix RoPE type for DeepSeek-OCR LM. * loading LM testing Vision model loading * sam warmup working * sam erroneous return corrected * clip-vit: corrected cls_embd concat * clip-vit: model convert qkv_proj split * corrected combining of image encoders' results * fix: update callback for ffn_moe_weighted and add callback for attn_out in deepseek2 model * concat image_newline and image_seperator tokens * visual_model warmup (technically) works * window partitioning using standard ggml ops * sam implementation without using CPU only ops * clip: fixed warnings * Merge branch 'sf/deepseek-ocr' of github.com:sfallah/llama.cpp into sf/deepseek-ocr * mtmd: fix get_rel_pos * mtmd: fixed the wrong scaler for get_rel_pos * image encoding technically works but the output can't be checked singe image decoding fails * mtmd: minor changed * mtmd: add native resolution support * - image encoding debugged - issues fixed mainly related wrong config like n_patches etc. - configs need to be corrected in the converter * mtmd: correct token order * - dynamic resizing - changes are concerning PR sfallah#4 * mtmd: quick fix token order * mtmd: fix danling pointer * mtmd: SAM numerically works * mtmd: debug CLIP-L (vit_pre_ln) * mtmd: debug CLIP-L & first working DeepSeek-OCR model * mtmd : add --dsocr-mode CLI argument for DeepSeek-OCR resolution control & all native resolution modes work * mtmd: simplify SAM patch embedding * mtmd: adapt Pillow image resizing function * mtmd: simplify DeepSeek-OCR dynamic resolution preprocessing * mtmd: remove --dsocr-mode argument * mtmd: refactor code & remove unused helper functions * mtmd: fix tensor names for image newlines and view separator * clean up * reverting automatically removed spaces * reverting automatically removed spaces * mtmd: fixed bad ocr check in Deepseek2 (LM) * mtmd: support combined QKV projection in buid_vit * using common build_attn in sam * corrected code-branch when flash-attn disabled enabling usage of --flash-attn option * mtmd: minor fix * minor formatting and style * fixed flake8 lint issues * minor editorconfig-check fixes * minor editorconfig-check fixes * mtmd: simplify get_rel_pos * mtmd: make sam hparams configurable * mtmd: add detailed comments for resize_bicubic_pillow * mtmd: fixed wrong input setting * mtmd: convert model in FP16 * mtmd: minor fix * mtmd: remove tweak to llama-mtmd-cli & deepseek-ocr template * fix: test-1.jpg ORC issue with small (640) resolution setting min-resolution base (1024) max large (1280) for dynamic-resolution * minor: editconfig-check fix * merge with changes from ggml-org#17909 added new opt to tests.sh to disable flash-attn * minor: editconfig-check fix * testing deepseek-ocr quick and dirty test script comparing results of Qwen2.5-VL vs DeepSeek-OCR * quick and (potential) dirty merge with ggml-org#17909 * refactoring, one single builder function and static helpers * added deepseek-ocr test to tests.sh * minor formatting fixes * check with fixed expected resutls * minor formatting * editorconfig-check fix * merge with changes from ggml-org#18042 * minor - added GLM-4.6V to big tests - added missing deps for python test * convert: minor fix * mtmd: format code * convert: quick fix * convert: quick fix * minor python formatting * fixed merge build issue * merge resolved - fixed issues in convert - tested several deepseek models * minor fix * minor * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * - removed clip_is_deepseekocr - removed redundant RESIZE_ALGO_BICUBIC_PILLOW resize-algo - simplified image-preprocessing - removed/simplified debug functions * - cleaning commented out code * fixing instabilities issues reintroducing resize_bicubic_pillow * - use f16 model for deepseek-ocr test - ignore llama-arch test for deepseek-ocr * rename fc_w --> mm_fc_w * add links to OCR discussion * cleaner loading code * add missing .weight to some tensors * add default jinja template (to be used by server) * move test model to ggml-org * rolling back upscale change * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> --------- Co-authored-by: bluebread <hotbread70127@gmail.com> Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> Co-authored-by: Xuan Son Nguyen <son@huggingface.co> Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

* clip: add support for fused qkv in build_vit * use bulid_ffn whenever possible * fix internvl * mtmd-cli: move image to beginning * test script: support custom args

* mtmd: llama.cpp DeepSeekOCR support init commit * loading sam tensors * mtmd: fix vision model processing * deepseek-ocr clip-vit model impl * mtmd: add DeepSeek-OCR LM support with standard attention * mtmd: successfully runs DeepSeek-OCR LM in llama-cli * mtmd: Fix RoPE type for DeepSeek-OCR LM. * loading LM testing Vision model loading * sam warmup working * sam erroneous return corrected * clip-vit: corrected cls_embd concat * clip-vit: model convert qkv_proj split * corrected combining of image encoders' results * fix: update callback for ffn_moe_weighted and add callback for attn_out in deepseek2 model * concat image_newline and image_seperator tokens * visual_model warmup (technically) works * window partitioning using standard ggml ops * sam implementation without using CPU only ops * clip: fixed warnings * Merge branch 'sf/deepseek-ocr' of github.com:sfallah/llama.cpp into sf/deepseek-ocr * mtmd: fix get_rel_pos * mtmd: fixed the wrong scaler for get_rel_pos * image encoding technically works but the output can't be checked singe image decoding fails * mtmd: minor changed * mtmd: add native resolution support * - image encoding debugged - issues fixed mainly related wrong config like n_patches etc. - configs need to be corrected in the converter * mtmd: correct token order * - dynamic resizing - changes are concerning PR sfallah#4 * mtmd: quick fix token order * mtmd: fix danling pointer * mtmd: SAM numerically works * mtmd: debug CLIP-L (vit_pre_ln) * mtmd: debug CLIP-L & first working DeepSeek-OCR model * mtmd : add --dsocr-mode CLI argument for DeepSeek-OCR resolution control & all native resolution modes work * mtmd: simplify SAM patch embedding * mtmd: adapt Pillow image resizing function * mtmd: simplify DeepSeek-OCR dynamic resolution preprocessing * mtmd: remove --dsocr-mode argument * mtmd: refactor code & remove unused helper functions * mtmd: fix tensor names for image newlines and view separator * clean up * reverting automatically removed spaces * reverting automatically removed spaces * mtmd: fixed bad ocr check in Deepseek2 (LM) * mtmd: support combined QKV projection in buid_vit * using common build_attn in sam * corrected code-branch when flash-attn disabled enabling usage of --flash-attn option * mtmd: minor fix * minor formatting and style * fixed flake8 lint issues * minor editorconfig-check fixes * minor editorconfig-check fixes * mtmd: simplify get_rel_pos * mtmd: make sam hparams configurable * mtmd: add detailed comments for resize_bicubic_pillow * mtmd: fixed wrong input setting * mtmd: convert model in FP16 * mtmd: minor fix * mtmd: remove tweak to llama-mtmd-cli & deepseek-ocr template * fix: test-1.jpg ORC issue with small (640) resolution setting min-resolution base (1024) max large (1280) for dynamic-resolution * minor: editconfig-check fix * merge with changes from ggml-org#17909 added new opt to tests.sh to disable flash-attn * minor: editconfig-check fix * testing deepseek-ocr quick and dirty test script comparing results of Qwen2.5-VL vs DeepSeek-OCR * quick and (potential) dirty merge with ggml-org#17909 * refactoring, one single builder function and static helpers * added deepseek-ocr test to tests.sh * minor formatting fixes * check with fixed expected resutls * minor formatting * editorconfig-check fix * merge with changes from ggml-org#18042 * minor - added GLM-4.6V to big tests - added missing deps for python test * convert: minor fix * mtmd: format code * convert: quick fix * convert: quick fix * minor python formatting * fixed merge build issue * merge resolved - fixed issues in convert - tested several deepseek models * minor fix * minor * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * - removed clip_is_deepseekocr - removed redundant RESIZE_ALGO_BICUBIC_PILLOW resize-algo - simplified image-preprocessing - removed/simplified debug functions * - cleaning commented out code * fixing instabilities issues reintroducing resize_bicubic_pillow * - use f16 model for deepseek-ocr test - ignore llama-arch test for deepseek-ocr * rename fc_w --> mm_fc_w * add links to OCR discussion * cleaner loading code * add missing .weight to some tensors * add default jinja template (to be used by server) * move test model to ggml-org * rolling back upscale change * Update convert_hf_to_gguf.py Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> --------- Co-authored-by: bluebread <hotbread70127@gmail.com> Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> Co-authored-by: Xuan Son Nguyen <son@huggingface.co> Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

ngxson added 5 commits December 10, 2025 16:04

clip: add support for fused qkv in build_vit

35df7b8

use bulid_ffn whenever possible

05ab97b

fix internvl

ea005d6

mtmd-cli: move image to beginning

a811b0c

test script: support custom args

32e6a33

ngxson marked this pull request as ready for review December 10, 2025 15:53

ngxson requested a review from ggerganov December 10, 2025 15:53

ggerganov approved these changes Dec 10, 2025

View reviewed changes

CISC approved these changes Dec 10, 2025

View reviewed changes

github-actions Bot added the examples label Dec 10, 2025

ngxson merged commit c6b2c93 into ggml-org:master Dec 10, 2025
73 of 76 checks passed

ngxson mentioned this pull request Dec 10, 2025

mtmd: Add DeepSeekOCR Support #17400

Merged

sfallah added a commit to sfallah/llama.cpp that referenced this pull request Dec 11, 2025

merge with changes from ggml-org#17909

d70f171

added new opt to tests.sh to disable flash-attn

deadprogram mentioned this pull request Dec 11, 2025

chore: change media marker location to start of prompt to match behavior in llama.cpp hybridgroup/yzma#152

Merged

sfallah added a commit to sfallah/llama.cpp that referenced this pull request Dec 13, 2025

quick and (potential) dirty merge with ggml-org#17909

f95a6fe

sfallah mentioned this pull request Dec 13, 2025

Merged with PR #17965 sfallah/llama.cpp#11

Merged

blime4 referenced this pull request in blime4/llama.cpp Feb 5, 2026

mtmd: some small clean up (#17909)

bb66684

* clip: add support for fused qkv in build_vit * use bulid_ffn whenever possible * fix internvl * mtmd-cli: move image to beginning * test script: support custom args

wallentri88 mentioned this pull request Feb 24, 2026

Eval bug: qwen35 and qwen35moe graph split issues (Severe PP impact, crashes) #19864

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

mtmd: some small clean up#17909

mtmd: some small clean up#17909
ngxson merged 5 commits intoggml-org:masterfrom
ngxson:xsn/mtmd_improve_0

ngxson commented Dec 10, 2025 •

edited

Loading

Uh oh!

Uh oh!

amritsingh183 commented Dec 11, 2025

Uh oh!

sfallah commented Dec 11, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Conversation

ngxson commented Dec 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

amritsingh183 commented Dec 11, 2025

Uh oh!

sfallah commented Dec 11, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

ngxson commented Dec 10, 2025 •

edited

Loading