ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update llama.py

Browse source
This commit is contained in:
Daniel Han 2024-11-05 21:07:02 -08:00
commit 4594ae98e3
1 changed files with 4 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

7
unsloth/models/llama.py
View file

@ -390,7 +390,7 @@ def LlamaAttention_fast_forward(
past_key_value = (K, V) if use_cache else None
# Attention module
if False:#(not HAS_FLASH_ATTENTION and attention_mask is None):
if (not HAS_FLASH_ATTENTION and attention_mask is None):
# Xformers memory efficient attention
# Also has Flash Attention v2 dispatching
Q = Q.transpose(1, 2)
@ -427,10 +427,10 @@ def LlamaAttention_fast_forward(
pass
# Must be contiguous or else results are False!
# https://github.com/pytorch/pytorch/issues/112577
# Q, K, V = Q.contiguous(), K.contiguous(), V.contiguous()
Q, K, V = Q.contiguous(), K.contiguous(), V.contiguous()
# Needs (batch_size, n_heads, seq_len, head_dim)
# is_casual and attention_mask must not be both set!
A = scaled_dot_product_attention(Q, K, V, is_causal = True)
A = scaled_dot_product_attention(Q, K, V, attn_mask = attention_mask, is_causal = False)
# Go back to (batch_size, seq_len, n_heads, head_dim)
A = A.transpose(1, 2).contiguous()
pass
@ -1028,6 +1028,7 @@ def CausalLM_fast_forward(fast_forward_inference):
pass
@torch._disable_dynamo
def PeftModelForCausalLM_fast_forward(
self,
input_ids=None,