Spaces:

Dovakiins
/

qwerrwe

Build error

App Files Files Community

100 contributors

History: 24 commits

winglian's picture

casts the prepared data to int16 (doesn't help with training memory)

2db9436 over 1 year ago

configs
fix lora target module, require explicit flash attention, fix min logging steps, don't use adam8bit for int4, hash prepared datasets, support hf hub datasets over 1 year ago
data
WIP for axolotl trainer over 1 year ago
scripts
casts the prepared data to int16 (doesn't help with training memory) over 1 year ago
src
casts the prepared data to int16 (doesn't help with training memory) over 1 year ago
.editorconfig

186 Bytes

WIP for axolotl trainer over 1 year ago
.gitattributes

49 Bytes

make it work with pythia in the cloud over 1 year ago
.gitignore
41 Bytes

WIP for axolotl trainer over 1 year ago
README.md
2.38 kB

fix install to work with latest alpaca lora 4bit over 1 year ago
ds_config.json
844 Bytes

deepspeed doesn't work with flash-attn, and the gpu savings w flash attn are better than the deepspeed headaches over 1 year ago
requirements.txt
231 Bytes

fix install to work with latest alpaca lora 4bit over 1 year ago
setup.py
909 Bytes

fix install to work with latest alpaca lora 4bit over 1 year ago