github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-09-20 17:10:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

[misc] update pre-commit and run all files (#4752)

Browse Source 
* [misc] update pre-commit

* [misc] run pre-commit

* [misc] remove useless configuration files

* [misc] ignore cuda for clang-format
This commit is contained in:
Hongxin Liu
2023-09-19 14:20:26 +08:00
committed by GitHub
parent 3c6b831c26
commit 079bf3cb26
1268 changed files with 50037 additions and 38444 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
applications/Chat/coati/kernels/__init__.py
View File

@@ -1,6 +1,6 @@
from .wrapper import convert_to_xformer_model, recover_from_xformer_model
__all__ = [
'convert_to_xformer_model',
'recover_from_xformer_model',
"convert_to_xformer_model",
"recover_from_xformer_model",
]

23
applications/Chat/coati/kernels/opt_attn.py
View File

@@ -21,11 +21,12 @@ class XOPTAttention(OPTAttention):
output_attentions: bool = False,
) -> Tuple[Tensor, Optional[Tensor], Optional[Tuple[Tensor]]]:
if not self.training:
return super().forward(hidden_states, key_value_states, past_key_value, attention_mask, layer_head_mask,
output_attentions)
return super().forward(
hidden_states, key_value_states, past_key_value, attention_mask, layer_head_mask, output_attentions
)
"""Input shape: Batch x Time x Channel"""
assert layer_head_mask is None, 'Xformers attention does not support layer_head_mask'
assert not output_attentions, 'Xformers attention does not support output_attentions'
assert layer_head_mask is None, "Xformers attention does not support layer_head_mask"
assert not output_attentions, "Xformers attention does not support output_attentions"
# if key_value_states are provided this layer is used as a cross-attention layer
# for the decoder
@@ -69,12 +70,14 @@ class XOPTAttention(OPTAttention):
key_states = key_states.transpose(1, 2)
value_states = value_states.transpose(1, 2)
attn_output = xops.memory_efficient_attention(query_states,
key_states,
value_states,
attn_bias=xops.LowerTriangularMask(),
p=self.dropout if self.training else 0.0,
scale=self.scaling)
attn_output = xops.memory_efficient_attention(
query_states,
key_states,
value_states,
attn_bias=xops.LowerTriangularMask(),
p=self.dropout if self.training else 0.0,
scale=self.scaling,
)
# Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be
# partitioned across GPUs when using tensor-parallelism.