requirements.txt

accelerate==0.16.0
bitsandbytes==0.37.0
flexgen==0.1.7
gradio==3.18.0
numpy
rwkv==0.0.8
safetensors==0.2.8
sentencepiece
git+https://github.com/oobabooga/transformers@llama_push
Bump accelerate version 2023-02-18 22:15:47 -03:00			`accelerate==0.16.0`
Bump bitsandbytes version 2023-02-03 09:29:31 -03:00			`bitsandbytes==0.37.0`
Bump flexgen version 2023-03-02 12:03:57 -03:00			`flexgen==0.1.7`
Downgrade gradio version (file uploads are broken in 3.19.1) 2023-02-25 22:59:02 -03:00			`gradio==3.18.0`
Add softprompt support (for real this time) Is this too much voodoo for our purposes? 2023-02-13 15:25:16 -03:00			`numpy`
Add --rwkv-cuda-on parameter, bump rwkv version 2023-03-06 20:12:54 -03:00			`rwkv==0.0.8`
Add safetensors version 2023-02-04 18:58:17 -03:00			`safetensors==0.2.8`
Add sentencepiece 2023-03-05 10:02:24 -03:00			`sentencepiece`
Move towards HF LLaMA implementation 2023-03-05 01:20:31 -03:00			`git+https://github.com/oobabooga/transformers@llama_push`