Skip to content

[shardformer] fix llama error when transformers upgraded. #5055

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
flybird11111 merged 2 commits into from
Nov 16, 2023
Merged

[shardformer] fix llama error when transformers upgraded. #5055

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
f68bcd0
fix-llama
flybird11111 Nov 16, 2023
3c46edb
Update llama.py
flybird11111 Nov 16, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 14 additions & 4 deletions colossalai/shardformer/modeling/llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
import warnings
from typing import List, Optional, Tuple
from typing import List, Optional, Tuple, Union

import torch
from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
Expand All @@ -13,6 +13,11 @@

from colossalai.pipeline.stage_manager import PipelineStageManager

try:
from transformers.models.llama.modeling_llama import _prepare_4d_causal_attention_mask
LATEST_VERSION = True
except ImportError:
LATEST_VERSION = False

class LlamaPipelineForwards:
"""
Expand Down Expand Up @@ -97,9 +102,14 @@ def llama_model_forward(
attention_mask = torch.ones(
(batch_size, seq_length_with_past), dtype=torch.bool, device=hidden_states.device
)
attention_mask = self._prepare_decoder_attention_mask(
attention_mask, (batch_size, seq_length), hidden_states, past_key_values_length
)
if LATEST_VERSION:
attention_mask = _prepare_4d_causal_attention_mask(
attention_mask, (batch_size, seq_length), hidden_states, past_key_values_length
)
else:
attention_mask = self._prepare_decoder_attention_mask(
attention_mask, (batch_size, seq_length), hidden_states, past_key_values_length
)

if self.gradient_checkpointing and self.training:
if use_cache:
Expand Down