add parameter efficient finetuning pipeline
Why? this commit adds a generic finetuning pipeline with LoRA and DoRA support
This commit is contained in:
+3
-1
@@ -1,4 +1,4 @@
|
||||
name: tfm_env
|
||||
name: ok_tfm_env
|
||||
|
||||
channels:
|
||||
- conda-forge
|
||||
@@ -16,3 +16,5 @@ dependencies:
|
||||
- jax[cuda12]==0.4.26
|
||||
- einshape
|
||||
- scikit-learn
|
||||
- typer
|
||||
- wandb
|
||||
|
||||
Reference in New Issue
Block a user