xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-29 01:20:55 +08:00
Code Issues Packages Projects Releases Wiki Activity

[V1][Minor] Remove obsolete FIXME comment (#14304)

Browse Source 
Signed-off-by: Nick Hill <nhill@redhat.com>
This commit is contained in:
Nick Hill 2025-03-05 11:59:23 -08:00 committed by GitHub
parent ca2ca8de57
commit a32c8669ca
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 0 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/v1/worker/gpu_input_batch.py
View File

@ -298,11 +298,6 @@ class InputBatch:
if sampling_params.logit_bias is not None:
self.logit_bias[req_index] = sampling_params.logit_bias
# FIXME: this implementation is incorrect. We create this mask
# then apply -inf to these specific tokens, which means we never
# select the allowed tokens! We cannot do the reverse, since
# this will impact the requests that do not have allowed_token_ids.
# This feature is currently disabled on V1 (we reject in Processor).
if sampling_params.allowed_token_ids:
self.has_allowed_token_ids.add(req_id)
if self.allowed_token_ids_mask_cpu_tensor is None: