Simplify flashinfer indices update for prefill (#2074)
Co-authored-by: kavioyu <kavioyu@tencent.com> Co-authored-by: kavioyu <kavioyu@gmail.com>
This commit is contained in:
@@ -345,7 +345,7 @@ class LlavaBaseForCausalLM(nn.Module):
|
||||
|
||||
# Fill in the placeholder for the image
|
||||
extend_start_loc_cpu = forward_batch.extend_start_loc.cpu().numpy()
|
||||
prefix_lens_cpu = forward_batch.extend_prefix_lens.cpu().numpy()
|
||||
prefix_lens_cpu = forward_batch.extend_prefix_lens_cpu
|
||||
pt = 0
|
||||
for i in range(bs):
|
||||
if not need_vision[i]:
|
||||
|
||||
Reference in New Issue
Block a user