Merge pull request #258 from NanoCode012/fix/deprecate-push

README.md

@@ -341,7 +341,7 @@ dataset_prepared_path: data/last_run_prepared
 # push prepared dataset to hub
 push_dataset_to_hub: # repo path
 # push checkpoints to hub
-push_to_hub_model_id: # repo path
+hub_model_id: # repo path
 # whether to use hf `use_auth_token` for loading datasets. Useful for fetching private datasets
 # required to be true when used in combination with `push_dataset_to_hub`
 hf_use_auth_token: # boolean

src/axolotl/utils/trainer.py

@@ -124,8 +124,8 @@ def setup_trainer(cfg, train_dataset, eval_dataset, model, tokenizer):
     if cfg.max_grad_norm:
         training_arguments_kwargs["max_grad_norm"] = cfg.max_grad_norm
 
-    if cfg.push_to_hub_model_id:
-        training_arguments_kwargs["push_to_hub_model_id"] = cfg.push_to_hub_model_id
+    if cfg.hub_model_id:
+        training_arguments_kwargs["hub_model_id"] = cfg.hub_model_id
         training_arguments_kwargs["push_to_hub"] = True
 
     training_args = transformers.TrainingArguments(

src/axolotl/utils/validation.py

@@ -92,6 +92,11 @@ def validate_config(cfg):
     ):
         logging.warning("adamw hyperparameters found, but no adamw optimizer set")
 
+    if cfg.push_to_hub_model_id:
+        raise ValueError(
+            "push_to_hub_model_id is deprecated. Please use hub_model_id instead."
+        )
+
     # TODO
     # MPT 7b
     # https://github.com/facebookresearch/bitsandbytes/issues/25