p04798526
/
LLaMA-Factory-310P3
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix mask_history tiny bug

This commit is contained in:
“Wzw” 2024-08-08 10:09:33 +08:00
parent 18e455c232
commit b5ca86cc07
2 changed files with 13 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
src/llamafactory/data/processors/supervised.py
View File

@ -53,7 +53,7 @@ def _encode_supervised_example(
input_ids += [image_token_id] * getattr(processor, "image_seq_length")
labels += [IGNORE_INDEX] * getattr(processor, "image_seq_length")
encoded_pairs = template.encode_multiturn(tokenizer, messages, system, tools)
encoded_pairs = template.encode_multiturn(tokenizer, messages, system, tools, mask_history)
total_length = 1 if template.efficient_eos else 0
for turn_idx, (source_ids, target_ids) in enumerate(encoded_pairs):
if total_length >= cutoff_len:
@ -71,14 +71,15 @@ def _encode_supervised_example(
else:
source_label = [IGNORE_INDEX] * source_len
if mask_history and turn_idx != len(encoded_pairs) - 1:
target_label = [IGNORE_INDEX] * target_len
if mask_history:
target_label = target_ids if turn_idx==0 else [IGNORE_INDEX] * target_len
input_ids = source_ids + target_ids + input_ids
labels = source_label + target_label + labels
else:
target_label = target_ids
input_ids += source_ids + target_ids
labels += source_label + target_label
input_ids += source_ids + target_ids
labels += source_label + target_label
if template.efficient_eos:
input_ids += [tokenizer.eos_token_id]
labels += [tokenizer.eos_token_id]

6
src/llamafactory/data/template.py
View File

@ -69,12 +69,16 @@ class Template:
messages: Sequence[Dict[str, str]],
system: Optional[str] = None,
tools: Optional[str] = None,
mask_history: bool = False,
) -> List[Tuple[List[int], List[int]]]:
r"""
Returns multiple pairs of token ids representing prompts and responses respectively.
"""
encoded_messages = self._encode(tokenizer, messages, system, tools)
return [(encoded_messages[i], encoded_messages[i + 1]) for i in range(0, len(encoded_messages), 2)]
if not mask_history:
return [(encoded_messages[i], encoded_messages[i + 1]) for i in range(0, len(encoded_messages), 2)]
else:
return [(encoded_messages[i], encoded_messages[i + 1]) for i in range(len(encoded_messages)-2, -1, -2)]
def extract_tool(self, content: str) -> Union[str, List[Tuple[str, str]]]:
r"""