423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-23 14:22:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

tiny fix

Browse Source 
Former-commit-id: bda302fbfbdb114dee7782d405732600d2d73279
This commit is contained in:
hiyouga 2024-07-15 23:09:50 +08:00
parent 9d8e0f0837
commit ca40e42b3c
1 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
src/llamafactory/eval/evaluator.py
View File

@ -73,11 +73,11 @@ class Evaluator:
return [chr(ord("A") + offset.item()) for offset in torch.argmax(choice_probs, dim=-1)] return [chr(ord("A") + offset.item()) for offset in torch.argmax(choice_probs, dim=-1)]
def eval(self) -> None: def eval(self) -> None:
task = self.eval_args.task.split("_")[0] eval_task = self.eval_args.task.split("_")[0]
split = self.eval_args.task.split("_")[1] eval_split = self.eval_args.task.split("_")[1]
mapping = cached_file( mapping = cached_file(
path_or_repo_id=os.path.join(self.eval_args.task_dir, task), path_or_repo_id=os.path.join(self.eval_args.task_dir, eval_task),
filename="mapping.json", filename="mapping.json",
cache_dir=self.model_args.cache_dir, cache_dir=self.model_args.cache_dir,
token=self.model_args.hf_hub_token, token=self.model_args.hf_hub_token,
@ -91,7 +91,7 @@ class Evaluator:
results = {} results = {}
for subject in pbar: for subject in pbar:
dataset = load_dataset( dataset = load_dataset(
path=os.path.join(self.eval_args.task_dir, task), path=os.path.join(self.eval_args.task_dir, eval_task),
name=subject, name=subject,
cache_dir=self.model_args.cache_dir, cache_dir=self.model_args.cache_dir,
download_mode=self.eval_args.download_mode, download_mode=self.eval_args.download_mode,
@ -100,12 +100,12 @@ class Evaluator:
) )
pbar.set_postfix_str(categorys[subject]["name"]) pbar.set_postfix_str(categorys[subject]["name"])
inputs, outputs, labels = [], [], [] inputs, outputs, labels = [], [], []
for i in trange(len(dataset[split]), desc="Formatting batches", position=1, leave=False): for i in trange(len(dataset[eval_split]), desc="Formatting batches", position=1, leave=False):
support_set = ( support_set = (
dataset["train"].shuffle().select(range(min(self.eval_args.n_shot, len(dataset["train"])))) dataset["train"].shuffle().select(range(min(self.eval_args.n_shot, len(dataset["train"]))))
) )
messages = self.eval_template.format_example( messages = self.eval_template.format_example(
target_data=dataset[split][i], target_data=dataset[eval_split][i],
support_set=support_set, support_set=support_set,
subject_name=categorys[subject]["name"], subject_name=categorys[subject]["name"],
) )