github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-09-02 01:28:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

[misc] resolve code factor issues (#4433)

Browse Source
This commit is contained in:
Hongxin Liu
2023-08-14 17:43:33 +08:00
parent 328a791d10
commit 172f7fa3cf
20 changed files with 31 additions and 205 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
colossalai/shardformer/modeling/llama.py
View File

@@ -65,7 +65,7 @@ class LlamaPipelineForwards:
seq_length_with_past = seq_length
past_key_values_length = 0
# TODO: left the recording kv-value tensors as () or None type, this feature may be added in the future.
# TODO(jianghai): left the recording kv-value tensors as () or None type, this feature may be added in the future.
if output_attentions:
logger.warning_once('output_attentions=True is not supported for pipeline models at the moment.')
output_attentions = False
@@ -216,7 +216,7 @@ class LlamaPipelineForwards:
if output_hidden_states is not None else self.config.output_hidden_states)
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
# TODO: left the recording kv-value tensors as () or None type, this feature may be added in the future.
# TODO(jianghai): left the recording kv-value tensors as () or None type, this feature may be added in the future.
if output_attentions:
logger.warning_once('output_attentions=True is not supported for pipeline models at the moment.')
output_attentions = False
@@ -301,7 +301,7 @@ class LlamaPipelineForwards:
logger = logging.get_logger(__name__)
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
# TODO: left the recording kv-value tensors as () or None type, this feature may be added in the future.
# TODO(jianghai): left the recording kv-value tensors as () or None type, this feature may be added in the future.
if output_attentions:
logger.warning_once('output_attentions=True is not supported for pipeline models at the moment.')
output_attentions = False