Skip to content

Pull requests: Blaizzy/mlx-vlm

New pull request New
24 Open 461 Closed
24 Open 461 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Add --max-tokens CLI argument to server
#936 opened Apr 5, 2026 by nnorris7 Loading…
1 of 3 tasks
Fix Gemma 4 quantized per-layer projection loading
#935 opened Apr 5, 2026 by spicyneuron Loading…
docs: add model guides for ERNIE 4.5 VL and PaddleOCR-VL
#934 opened Apr 5, 2026 by jimmyzhuu Loading…
test: add PaddleOCR-VL processor regression coverage
#933 opened Apr 5, 2026 by jimmyzhuu Loading…
fix: Gemma 4 audio — mel preprocessing, weight loading, feature extractor
#931 opened Apr 5, 2026 by stephencox-ict Loading…
2
fix(gemma4): guard against empty processed list for text-only requests
#929 opened Apr 5, 2026 by sindhura6 Loading…
docs: add Chinese documentation and restructure docs directory
#928 opened Apr 5, 2026 by Tsan1024 Loading…
3 tasks done
fix(grounded_reasoning): bump max_new_tokens 512 → 1024
#927 opened Apr 5, 2026 by lichengzhe Loading…
add grounded_reasoning: Falcon Perception + Gemma4 agentic demo
#926 opened Apr 4, 2026 by YasserdahouML Loading…
2
Fix duplicate docstring entries and minor issues in utils.py
#925 opened Apr 4, 2026 by sjhddh Loading…
Fix Gemma 4 'No text generated' when chat template is missing
#924 opened Apr 4, 2026 by nnorris7 Loading…
3 of 4 tasks
feat: Add SAM 3D Body — monocular 3D body mesh on Apple Silicon
#922 opened Apr 4, 2026 by shihwesley Loading…
1
fix: prevent crash when args.adapter_file is None in sft_trainer
#921 opened Apr 4, 2026 by JasonOA888 Loading…
1
fix: preserve LoRA adapter across requests in OpenAI responses endpoint
#920 opened Apr 4, 2026 by JasonOA888 Loading…
1
Centralize server config and add CLI flags
#918 opened Apr 4, 2026 by spicyneuron Loading…
Fix batch generation and adopt mlx-lm batch improvements
#911 opened Apr 4, 2026 by Blaizzy Loading…
3 tasks done
Optimize TurboQuant: O(d log d) Walsh-Hadamard Transform
#860 opened Mar 26, 2026 by Trucker2827 Loading…
1 of 3 tasks
2
fix: use alpha/rank scaling in LoRaLayer (standard LoRA convention)
#846 opened Mar 21, 2026 by kikoncuo Loading…
1
11
Fix preprocessing for image input for trainer
#826 opened Mar 15, 2026 by Goekdeniz-Guelmez Loading…
17
Add distributed infer for qwen3_vl_moe
#730 opened Feb 13, 2026 by Blaizzy Loading…
10
Distributed inference for Kimi K2.5
#689 opened Jan 27, 2026 by pcuenca Loading…
14
Implement Joycaption as a custom Llava model
#659 opened Jan 16, 2026 by nArn0 Loading…
3
[WIP] Token filtering + merging
#185 opened Jan 20, 2025 by Blaizzy Loading…
1
[WIP] Prompt caching and Vision token merging + filtering
#177 opened Jan 10, 2025 by Blaizzy Loading…
3
ProTip! no:milestone will show everything without a milestone.