Skip to content
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

Jump to bottom

refactor cache behaviour in training mode (reduce compute cost and me… #2473

Merged
merged 1 commit into from
Apr 14, 2024
Merged

refactor cache behaviour in training mode (reduce compute cost and me… #2473

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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: 9 additions & 9 deletions wenet/transformer/attention.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -226,15 +226,15 @@ def forward(
# >>> torch.equal(b, c) # True
# >>> d = torch.split(a, 2, dim=-1)
# >>> torch.equal(d[0], d[1]) # True
if cache.size(0) > 0:
if cache.size(0) > 0 and not self.training:
key_cache, value_cache = torch.split(cache,
cache.size(-1) // 2,
dim=-1)
k = torch.cat([key_cache, k], dim=2)
v = torch.cat([value_cache, v], dim=2)
# NOTE(xcsong): We do cache slicing in encoder.forward_chunk, since it's
# non-trivial to calculate `next_cache_start` here.
new_cache = torch.cat((k, v), dim=-1)
new_cache = torch.cat((k, v), dim=-1) if not self.training else cache

# for multi query or multi group attention
if self.h_kv != self.h:
Expand Down Expand Up @@ -370,7 +370,7 @@ def forward(
# >>> torch.equal(b, c) # True
# >>> d = torch.split(a, 2, dim=-1)
# >>> torch.equal(d[0], d[1]) # True
if cache.size(0) > 0:
if cache.size(0) > 0 and not self.training:
key_cache, value_cache = torch.split(cache,
cache.size(-1) // 2,
dim=-1)
Expand All @@ -379,7 +379,7 @@ def forward(

# NOTE(xcsong): We do cache slicing in encoder.forward_chunk, since it's
# non-trivial to calculate `next_cache_start` here.
new_cache = torch.cat((k, v), dim=-1)
new_cache = torch.cat((k, v), dim=-1) if not self.training else cache

# for multi query or multi groups attention
if self.h_kv != self.h:
Expand Down Expand Up @@ -472,7 +472,7 @@ def forward(

else:
q, k, v = self.forward_qkv(query, key, value)
new_cache = torch.cat((k, v), dim=-1)
new_cache = torch.cat((k, v), dim=-1) if not self.training else cache

# for multi query or multi groups attention
if self.h_kv != self.h:
Expand Down Expand Up @@ -563,13 +563,13 @@ def forward(
) -> Tuple[torch.Tensor, torch.Tensor]:
del pos_emb
q, k, v = self.forward_qkv(query, key, value)
if cache.size(0) > 0:
if cache.size(0) > 0 and not self.training:
key_cache, value_cache = torch.split(cache,
cache.size(-1) // 2,
dim=-1)
k = torch.cat([key_cache, k], dim=2)
v = torch.cat([value_cache, v], dim=2)
new_cache = torch.cat((k, v), dim=-1)
new_cache = torch.cat((k, v), dim=-1) if not self.training else cache

rel_k = self.rel_k_embed(
self._relative_indices(k.size(2), query.device)) # (t2, t2, d_k)
Expand Down Expand Up @@ -664,13 +664,13 @@ def forward(
q = llama_apply_rotary_emb(q, pos_emb)
k = llama_apply_rotary_emb(k, pos_emb)
# see above
if cache.size(0) > 0:
if cache.size(0) > 0 and not self.training:
key_cache, value_cache = torch.split(cache,
cache.size(-1) // 2,
dim=-1)
k = torch.cat([key_cache, k], dim=2)
v = torch.cat([value_cache, v], dim=2)
new_cache = torch.cat((k, v), dim=-1)
new_cache = torch.cat((k, v), dim=-1) if not self.training else cache

if self.h_kv != self.h:
k = torch.repeat_interleave(
Expand Down
Loading