[PP] Fix start_layer_id with pp in get kv_buffer_shape (#29887)
Co-authored-by: ybyang <10629930+whybeyoung@users.noreply.github.com> Co-authored-by: Shangming Cai <csmthu@gmail.com>
This commit is contained in:
co-authored by
ybyang
Shangming Cai
parent
fefc1743a9
commit
669fd4b8a5
@@ -278,7 +278,7 @@ class EagerRunner(BaseRunner):
|
||||
forward_batch.req_pool_indices,
|
||||
get_req_to_token_pool().req_to_token,
|
||||
forward_batch.seq_lens_sum,
|
||||
get_token_to_kv_pool().get_key_buffer(0).shape,
|
||||
get_token_to_kv_pool().get_kv_buffer_shape()[0],
|
||||
model_runner.kv_cache_dtype,
|
||||
model_runner.device,
|
||||
create_chunked_prefix_cache_kv_indices,
|
||||
|
||||
Reference in New Issue
Block a user