You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi thank you for pointing this but we usually use flash-attention in long-context settings. If you set attn_implementation="flash_attention_2",, the error should not happen.
code
if self._use_flash_attention_2:
# 2d mask is passed through the layers
attention_mask = attention_mask if (attention_mask is not None and 0 in attention_mask) else None
elif self._use_sdpa and not output_attentions:
# output_attentions=True can not be supported when using SDPA, and we fall back on
# the manual implementation that requires a 4D causal mask in all cases.
attention_mask = _prepare_4d_causal_attention_mask_for_sdpa(
attention_mask,
(batch_size, seq_length),
inputs_embeds,
past_key_values_length,
)
else:
# 4d mask is passed through the layers
attention_mask = _prepare_4d_causal_attention_mask(
attention_mask, (batch_size, seq_length), inputs_embeds, past_key_values_length
)
Have not from transformers.modeling_attn_mask_utils import _prepare_4d_causal_attention_mask_for_sdpa
flash_decoding_chunkllama.py:
The text was updated successfully, but these errors were encountered: