p12475398
/
LLaMA-Factory-Mirror
forked from p04798526/LLaMA-Factory-Mirror
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix #1668

This commit is contained in:
hiyouga 2023-11-30 21:02:00 +08:00
parent a38dbf55e3
commit 1585962eb7
2 changed files with 4 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/llmtuner/train/ppo/trainer.py
View File

@ -298,7 +298,8 @@ class CustomPPOTrainer(PPOTrainer, Trainer):
with torch.cuda.amp.autocast(dtype=self.model_args.compute_dtype): # support bf16
logits, _, values = model(**input_kwargs)
if getattr(model.config, "model_type", None) == "chatglm":
unwrapped_model: "AutoModelForCausalLMWithValueHead" = self.accelerator.unwrap_model(self.model)
if getattr(unwrapped_model.config, "model_type", None) == "chatglm":
values = torch.transpose(values, 0, 1)
logprobs = logprobs_from_logits(logits[:, :-1, :], input_ids[:, 1:])

3
src/llmtuner/train/rm/trainer.py
View File

@ -40,7 +40,8 @@ class PairwiseTrainer(Trainer):
# Compute rewards
_, _, values = model(**inputs, output_hidden_states=True, return_dict=True)
if getattr(model.config, "model_type", None) == "chatglm":
unwrapped_model: "PreTrainedModel" = self.accelerator.unwrap_model(self.model)
if getattr(unwrapped_model.config, "model_type", None) == "chatglm":
values = torch.transpose(values, 0, 1)
# Split the inputs and rewards into two parts, chosen and rejected