Param Thakkar
4c9f9f9be7
Made llama and mistral files mypy compatible ( #1359 )
...
* Made mypy compatible
* reformatted
* Added more fixes
* Added fixes to speculative-decoding
* Fixes
* fix circle
* revert some stuff
---------
Co-authored-by: Awni Hannun <awni@apple.com>
2025-04-23 14:23:46 -07:00
Jinwu Zhan
1a86d985d9
Support --add_eos_token argument within Lora training ( #760 )
...
* Support `--add_eos_token` argument to empower users to control the addition of the eos token during LoRA training, addressing issues like incomplete text generation.
* Support `--add_eos_token`， code format
---------
Co-authored-by: Zhan ChengLong <zhanchenglong@bytedance.com>
2024-05-13 17:17:42 -07:00
Javier de la Rosa
510d2bde49
Force multi_commits when uploading to HF ( #729 )
2024-04-28 19:07:17 -07:00
Awni Hannun
ecbc6ff1e3
one more quant fix ( #708 )
2024-04-22 18:12:52 -07:00
AlexandrosChrtn
f20e68fcc0
Load fused model with transformers ( #703 )
...
* save format for transformers compatibility
* save format for transformers compatibility arg
* hardcode mlx
* hardcode mlx format
2024-04-21 09:04:44 -07:00
Awni Hannun
b8a348c1b8
Switch to fast RMS/LN Norm ( #603 )
...
* use nn.RMSNorm, use sdpa, cleanup
* bump mlx versions
* minor update
* use fast layer norm
* version bump
* update requirement for whisper
* update requirement for gguf
2024-03-23 07:13:51 -07:00
Awni Hannun
5aa652d3c2
remove simplify ( #379 )
2024-01-26 13:54:49 -08:00
Anchen
8022083979
feat(lora): add de-quantized support for fuse.py ( #351 )
...
* feat(lora): add de-quantized support for fuse.py
* address comments
2024-01-22 17:32:24 -08:00
Anchen
1415595409
chore(lora): support mixtral in lora example ( #343 )
2024-01-20 06:07:45 -08:00
Yousif
7575125d5d
Added lora support for Phi-2 ( #302 )
...
* Added lora support for Phi-2
* Added Phi-2 support in fuse and convert
* format + readme
---------
Co-authored-by: Awni Hannun <awni@apple.com>
2024-01-12 13:45:30 -08:00
Awni Hannun
7b258f33ac
Move lora example to use the same model format / conversion as hf_llm ( #252 )
...
* huffing face the lora example to allow more models
* fixes
* comments
* more readme nits
* fusion + works better for qlora
* nits'
* comments
2024-01-09 11:14:52 -08:00
