add model_init_kwargs to training_args (#1787)

examples/scripts/sft.py

@@ -94,23 +94,19 @@
         console = Console()
 
     ################
-    # Model & Tokenizer
+    # Model init kwargs & Tokenizer
     ################
-    torch_dtype = (
-        model_config.torch_dtype
-        if model_config.torch_dtype in ["auto", None]
-        else getattr(torch, model_config.torch_dtype)
-    )
     quantization_config = get_quantization_config(model_config)
     model_kwargs = dict(
         revision=model_config.model_revision,
         trust_remote_code=model_config.trust_remote_code,
         attn_implementation=model_config.attn_implementation,
-        torch_dtype=torch_dtype,
+        torch_dtype=model_config.torch_dtype,
         use_cache=False if training_args.gradient_checkpointing else True,
         device_map=get_kbit_device_map() if quantization_config is not None else None,
         quantization_config=quantization_config,
     )
+    training_args.model_init_kwargs = model_kwargs
     tokenizer = AutoTokenizer.from_pretrained(model_config.model_name_or_path, use_fast=True)
     tokenizer.pad_token = tokenizer.eos_token
 
@@ -138,7 +134,6 @@
     with init_context:
         trainer = SFTTrainer(
             model=model_config.model_name_or_path,
-            model_init_kwargs=model_kwargs,
             args=training_args,
             train_dataset=train_dataset,
             eval_dataset=eval_dataset,