423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-04 04:32:50 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix #1659

Browse Source 
Former-commit-id: 475a3fa0f4c09d4cfd55ec66271a6d3c9eb5f4d2
This commit is contained in:
hiyouga 2023-11-28 20:52:28 +08:00
parent 33dc25e24f
commit ae1048db6d
4 changed files with 12 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
src/llmtuner/extras/constants.py
View File

@ -217,7 +217,11 @@ register_model_group(
"Qwen-7B": "Qwen/Qwen-7B", "Qwen-7B": "Qwen/Qwen-7B",
"Qwen-14B": "Qwen/Qwen-14B", "Qwen-14B": "Qwen/Qwen-14B",
"Qwen-7B-Chat": "Qwen/Qwen-7B-Chat", "Qwen-7B-Chat": "Qwen/Qwen-7B-Chat",
"Qwen-14B-Chat": "Qwen/Qwen-14B-Chat" "Qwen-14B-Chat": "Qwen/Qwen-14B-Chat",
"Qwen-7B-int8-Chat": "Qwen/Qwen-7B-Chat-Int8",
"Qwen-7B-int4-Chat": "Qwen/Qwen-7B-Chat-Int4",
"Qwen-14B-int8-Chat": "Qwen/Qwen-14B-Chat-Int8",
"Qwen-14B-int4-Chat": "Qwen/Qwen-14B-Chat-Int4"
}, },
module="c_attn", module="c_attn",
template="qwen" template="qwen"
@ -266,8 +270,7 @@ register_model_group(
"Yi-6B": "01-ai/Yi-6B", "Yi-6B": "01-ai/Yi-6B",
"Yi-34B": "01-ai/Yi-34B", "Yi-34B": "01-ai/Yi-34B",
"Yi-34B-Chat": "01-ai/Yi-34B-Chat", "Yi-34B-Chat": "01-ai/Yi-34B-Chat",
"Yi-34B-int8-Chat": "01-ai/Yi-34B-Chat-8bits", "Yi-34B-int8-Chat": "01-ai/Yi-34B-Chat-8bits"
"Yi-34B-int4-Chat": "01-ai/Yi-34B-Chat-4bits"
}, },
template="yi" template="yi"
) )

2
src/llmtuner/hparams/finetuning_args.py
View File

@ -179,7 +179,7 @@ class FinetuningArguments(FreezeArguments, LoraArguments, RLHFArguments):
raise ValueError("Reward model is necessary for PPO training.") raise ValueError("Reward model is necessary for PPO training.")
if self.stage == "ppo" and self.reward_model_type == "lora" and self.finetuning_type != "lora": if self.stage == "ppo" and self.reward_model_type == "lora" and self.finetuning_type != "lora":
raise ValueError("Lora reward model only supports lora training.") raise ValueError("Freeze/Full PPO training needs `reward_model_type=full`.")
def save_to_json(self, json_path: str): def save_to_json(self, json_path: str):
r"""Saves the content of this instance in JSON format inside `json_path`.""" r"""Saves the content of this instance in JSON format inside `json_path`."""

3
src/llmtuner/model/loader.py
View File

@ -146,6 +146,9 @@ def load_model_and_tokenizer(
# Quantization configurations (using bitsandbytes library) # Quantization configurations (using bitsandbytes library)
if model_args.quantization_bit is not None: if model_args.quantization_bit is not None:
if getattr(config, "quantization_config", None):
raise ValueError("Remove `quantization_bit` if you are using a quantized model.")
if is_deepspeed_zero3_enabled(): if is_deepspeed_zero3_enabled():
raise ValueError("DeepSpeed ZeRO-3 is incompatible with quantization.") raise ValueError("DeepSpeed ZeRO-3 is incompatible with quantization.")

4
src/llmtuner/train/tuner.py
View File

@ -38,8 +38,8 @@ def export_model(args: Optional[Dict[str, Any]] = None):
model_args, _, finetuning_args, _ = get_infer_args(args) model_args, _, finetuning_args, _ = get_infer_args(args)
model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args) model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args)
if getattr(model, "quantization_method", None) == "gptq": if getattr(model, "quantization_method", None) in ["gptq", "awq"]:
raise ValueError("Cannot export a GPTQ quantized model.") raise ValueError("Cannot export a GPTQ or AWQ quantized model.")
model.config.use_cache = True model.config.use_cache = True
model.save_pretrained(finetuning_args.export_dir, max_shard_size="{}GB".format(finetuning_args.export_size)) model.save_pretrained(finetuning_args.export_dir, max_shard_size="{}GB".format(finetuning_args.export_size))