mirror of
https://github.com/huggingface/transformers.git
synced 2025-08-03 03:31:05 +06:00
chore(qwen2): display warning log only when sliding window attention … (#36316)
* chore(qwen2): display warning log only when sliding window attention is enabled * Align modeling_qwen2.py and modular_qwen2.py --------- Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>
This commit is contained in:
parent
8ea72d12a2
commit
31791b16a1
@ -232,7 +232,7 @@ class Qwen2DecoderLayer(GradientCheckpointingLayer):
|
||||
self.mlp = Qwen2MLP(config)
|
||||
self.input_layernorm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
|
||||
self.post_attention_layernorm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
|
||||
if config.sliding_window and config._attn_implementation != "flash_attention_2":
|
||||
if config.use_sliding_window and config._attn_implementation != "flash_attention_2":
|
||||
logger.warning_once(
|
||||
f"Sliding Window Attention is enabled but not implemented for `{config._attn_implementation}`; "
|
||||
"unexpected results may be encountered."
|
||||
|
@ -108,7 +108,7 @@ class Qwen2DecoderLayer(LlamaDecoderLayer):
|
||||
super().__init__()
|
||||
self.self_attn = Qwen2Attention(config=config, layer_idx=layer_idx)
|
||||
self.mlp = Qwen2MLP(config)
|
||||
if config.sliding_window and config._attn_implementation != "flash_attention_2":
|
||||
if config.use_sliding_window and config._attn_implementation != "flash_attention_2":
|
||||
logger.warning_once(
|
||||
f"Sliding Window Attention is enabled but not implemented for `{config._attn_implementation}`; "
|
||||
"unexpected results may be encountered."
|
||||
|
Loading…
Reference in New Issue
Block a user