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

Update gemma.py

Browse source
This commit is contained in:
Daniel Han-Chen 2024-02-26 04:27:55 +11:00
commit 686d9fa4c4
1 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
unsloth/models/gemma.py
View file

@ -77,8 +77,8 @@ class FastGemmaRotaryEmbedding(torch.nn.Module):
position_ids_expanded = position_ids[:, None, :].float()
freqs = (inv_freq_expanded @ position_ids_expanded).transpose(1, 2)
emb = torch.cat((freqs, freqs), dim=-1)
self.cos_cached = emb.cos().to(torch.bfloat16)
self.sin_cached = emb.sin().to(torch.bfloat16)
self.cos_cached2 = emb.cos().to(torch.bfloat16)
self.sin_cached2 = emb.sin().to(torch.bfloat16)
def forward(self, x, position_ids, seq_len=None):
# x: [bs, num_attention_heads, seq_len, head_size]
@ -107,7 +107,7 @@ class FastGemmaRotaryEmbedding(torch.nn.Module):
sin_cached = emb.sin().to(dtype=x.dtype)
print(cos_cached)
print(self.cos_cached)
print(self.cos_cached2)
# return self.cos_cached[:,:length], self.sin_cached[:,:length]
return self.cos_cached, self.sin_cached