tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Peft deepspeed resume (#1227)

Browse Source 
* import deepspeed integration

* monkeypatch peft adapater with deepspeed for resume from checkpoint

* fix patch

* fix patches attempt 2

* make sure to set lora_model_dir

* skip pylint for deepspeed.utils

* pick up upstream fix in transformers

* remove monkeypatch for deepspeed/peft fix

* no need to set the lora_model_dir on resume

* unset load_in_*bit when using quant config

* guard before del

* better handling of load_in* kwargs
This commit is contained in:
Wing Lian
2024-01-31 18:13:29 -05:00
committed by GitHub
parent 25e037fe2d
commit c67fb71583
4 changed files with 35 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
requirements.txt
View File

@@ -1,7 +1,7 @@
--extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/
packaging==23.2
peft @ git+https://github.com/huggingface/peft.git
transformers==4.37.0
transformers @ git+https://github.com/huggingface/transformers.git@bebeeee01275c32fccec3fa36d8b148d3813a7dc
tokenizers==0.15.0
bitsandbytes>=0.41.1
accelerate==0.26.1