423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-12-16 20:00:36 +08:00
Code Issues Packages Projects Releases Wiki Activity

refactor adapter hparam

Browse Source
This commit is contained in:
hiyouga
2023-12-15 20:53:11 +08:00
parent d4c351f1ec
commit 0716f5e470
21 changed files with 302 additions and 311 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
src/llmtuner/hparams/finetuning_args.py
View File

@@ -15,7 +15,7 @@ class FreezeArguments:
LLaMA choices: [\"mlp\", \"self_attn\"], \
BLOOM & Falcon & ChatGLM choices: [\"mlp\", \"self_attention\"], \
Qwen choices: [\"mlp\", \"attn\"], \
Phi-1.5 choices: [\"mlp\", \"mixer\"], \
Phi choices: [\"mlp\", \"mixer\"], \
Others choices: the same as LLaMA."}
)
num_layer_trainable: Optional[int] = field(
@@ -33,9 +33,9 @@ class LoraArguments:
default=None,
metadata={"help": "Name(s) of modules apart from LoRA layers to be set as trainable and saved in the final checkpoint."}
)
lora_alpha: Optional[float] = field(
lora_alpha: Optional[int] = field(
default=None,
metadata={"help": "The scale factor for LoRA fine-tuning (default: lora_rank * 2.0)."}
metadata={"help": "The scale factor for LoRA fine-tuning (default: lora_rank * 2)."}
)
lora_dropout: Optional[float] = field(
default=0.1,
@@ -52,12 +52,12 @@ class LoraArguments:
BLOOM & Falcon & ChatGLM choices: [\"query_key_value\", \"dense\", \"dense_h_to_4h\", \"dense_4h_to_h\"], \
Baichuan choices: [\"W_pack\", \"o_proj\", \"gate_proj\", \"up_proj\", \"down_proj\"], \
Qwen choices: [\"c_attn\", \"attn.c_proj\", \"w1\", \"w2\", \"mlp.c_proj\"], \
Phi-1.5 choices: [\"Wqkv\", \"out_proj\", \"fc1\", \"fc2\"], \
Phi choices: [\"Wqkv\", \"out_proj\", \"fc1\", \"fc2\"], \
Others choices: the same as LLaMA."}
)
resume_lora_training: Optional[bool] = field(
default=True,
metadata={"help": "Whether to resume training from the last LoRA weights or create new weights after merging them."}
create_new_adapter: Optional[bool] = field(
default=False,
metadata={"help": "Whether to create a new adapter with randomly initialized weight or not."}
)
@@ -98,9 +98,9 @@ class RLHFArguments:
default=None,
metadata={"help": "Path to the reference model used for the PPO or DPO training."}
)
ref_model_checkpoint: Optional[str] = field(
ref_model_adapters: Optional[str] = field(
default=None,
metadata={"help": "Path to the directory(s) containing the model checkpoints of the reference model."}
metadata={"help": "Path to the adapters of the reference model."}
)
ref_model_quantization_bit: Optional[int] = field(
default=None,
@@ -108,11 +108,11 @@ class RLHFArguments:
)
reward_model: Optional[str] = field(
default=None,
metadata={"help": "Path to the directory containing the checkpoints of the reward model."}
metadata={"help": "Path to the reward model used for the PPO training."}
)
reward_model_checkpoint: Optional[str] = field(
reward_model_adapters: Optional[str] = field(
default=None,
metadata={"help": "Path to the directory(s) containing the model checkpoints of the reward model."}
metadata={"help": "Path to the adapters of the reward model."}
)
reward_model_quantization_bit: Optional[int] = field(
default=None,
@@ -161,11 +161,11 @@ class FinetuningArguments(FreezeArguments, LoraArguments, RLHFArguments):
return arg
self.name_module_trainable = split_arg(self.name_module_trainable)
self.lora_alpha = self.lora_alpha or float(self.lora_rank * 2.0)
self.lora_alpha = self.lora_alpha or self.lora_rank * 2
self.lora_target = split_arg(self.lora_target)
self.additional_target = split_arg(self.additional_target)
self.ref_model_checkpoint = split_arg(self.ref_model_checkpoint)
self.reward_model_checkpoint = split_arg(self.reward_model_checkpoint)
self.ref_model_adapters = split_arg(self.ref_model_adapters)
self.reward_model_adapters = split_arg(self.reward_model_adapters)
assert self.finetuning_type in ["lora", "freeze", "full"], "Invalid fine-tuning method."
assert self.ref_model_quantization_bit in [None, 8, 4], "We only accept 4-bit or 8-bit quantization."