biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[P/D][V1] KV Connector API V1 (#15960)

Browse Source 
Signed-off-by: ApostaC <yihua98@uchicago.edu>
Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>
Signed-off-by: remi <remi@mistral.ai>
Co-authored-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>
Co-authored-by: Rémi Delacourt <54138269+Flechman@users.noreply.github.com>
Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
This commit is contained in:
Yihua Cheng
2025-04-17 15:22:40 -05:00
committed by GitHub
parent 0377b8310b
commit 3408e47159
24 changed files with 1377 additions and 83 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
vllm/engine/arg_utils.py
View File

@@ -1527,12 +1527,6 @@ class EngineArgs:
recommend_to_remove=False)
return False
# No Disaggregated Prefill so far.
if self.kv_transfer_config != EngineArgs.kv_transfer_config:
_raise_or_fallback(feature_name="--kv-transfer-config",
recommend_to_remove=False)
return False
# No FlashInfer or XFormers so far.
V1_BACKENDS = [
"FLASH_ATTN_VLLM_V1", "FLASH_ATTN", "PALLAS", "PALLAS_VLLM_V1",