Skip to content

Pull requests: unslothai/unsloth

New pull request New
67 Open 397 Closed
67 Open 397 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

feat: Support custom auto_model for wider model compatibility (Whisper, Bert,etc) & attn_implementation support
#2263 opened Apr 1, 2025 by Etherll Loading…
9
Fix Precision Mismatch in Continued Pretraining with FP16 Embeddings
#2259 opened Apr 1, 2025 by rupaut98 Loading…
loader.py: when dispatching to FastModel, use original model name
#2246 opened Mar 31, 2025 by ushakov Loading…
Fix Qwen2.5 'str object is not callable' error in generate()
#2239 opened Mar 30, 2025 by aditya0155 Loading…
3
Fix batched generation for prompts of different lengths
#2216 opened Mar 27, 2025 by RunFMe Loading…
5
[WIP] Initial support for Qwen3. Will udpate when the model is released
#2211 opened Mar 27, 2025 by Datta0 • Draft
Semantic Versioning for Future Releases
#2166 opened Mar 23, 2025 by ParagEkbote Loading…
1
4
Add flex attention with dynamic sequence length
#2162 opened Mar 23, 2025 by Andrew-Zhang Loading…
vLLM Windows CUDA support [tested]
#2158 opened Mar 23, 2025 by fenglui Loading…
2
Initial changes: Refactor Attention
#2156 opened Mar 22, 2025 by Itssshikhar Loading…
2
remove dead code from fast_rms_layernorm_inference
#2135 opened Mar 21, 2025 by KareemMusleh Loading…
1
VLM Data Collator - Make text & image mixing work efficiently
#2133 opened Mar 21, 2025 by mmathew23 Loading…
Optimize cohere contiguous
#2125 opened Mar 20, 2025 by NinoRisteski Loading…
1
Update README.md
#2119 opened Mar 20, 2025 by jackswl Loading…
1
Optimize get_executable func with list compr
#2083 opened Mar 18, 2025 by NinoRisteski Loading…
1
Optimize rl
#2065 opened Mar 17, 2025 by NinoRisteski Loading…
4
Improve code quality in llama.py and vision.py
#2061 opened Mar 17, 2025 by NinoRisteski Loading…
2
Add load_in_16bit Parameter and Fix 8-bit Quantization Config
#2022 opened Mar 14, 2025 by marcelodiaz558 Loading…
5
DynamicFlexAttention wrapper class for dynamic sequence lengths
#1960 opened Mar 9, 2025 by zyklotomic Loading…
7
Add automatic image resizing to prevent memory explosion
#1946 opened Mar 7, 2025 by issamarabi Loading…
fix/enable flex_attention
#1906 opened Mar 5, 2025 by notlober Loading…
1
[DRAFT]: Adding save to gguf support for qwen2_vl
#1904 opened Mar 5, 2025 by Captain-T2004 Loading…
2
patch vlm trainer to resize images
#1808 opened Feb 23, 2025 by oliveirabruno01 Loading…
1
Fix/load lora save lora
#1805 opened Feb 22, 2025 by Erland366 • Draft
Fix unwrapped old generate
#1784 opened Feb 21, 2025 by NinoRisteski Loading…
3
ProTip! Exclude everything labeled bug with -label:bug.