20231088/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[V1] Improve Error Message for Unsupported Config (#12535)

Browse Source 
Co-authored-by: Michael Goin <michael@neuralmagic.com>
This commit is contained in:
Robert Shaw 2025-01-28 23:56:56 -05:00 committed by GitHub
parent bd02164cf9
commit 5f671cb4c3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 7 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
vllm/platforms/cuda.py
View File

@ -120,13 +120,18 @@ class CudaPlatformBase(Platform):
if parallel_config.worker_cls == "auto": if parallel_config.worker_cls == "auto":
if scheduler_config.is_multi_step: if scheduler_config.is_multi_step:
if envs.VLLM_USE_V1: if envs.VLLM_USE_V1:
raise NotImplementedError raise NotImplementedError(
"Multi-step scheduling is not supported (and not "
"needed) on VLLM V1. Please launch without "
"--num-scheduler-steps.")
else: else:
parallel_config.worker_cls = \ parallel_config.worker_cls = \
"vllm.worker.multi_step_worker.MultiStepWorker" "vllm.worker.multi_step_worker.MultiStepWorker"
elif vllm_config.speculative_config: elif vllm_config.speculative_config:
if envs.VLLM_USE_V1: if envs.VLLM_USE_V1:
raise NotImplementedError raise NotImplementedError(
"Speculative decoding is not yet supported on VLLM V1."
)
else: else:
parallel_config.worker_cls = \ parallel_config.worker_cls = \
"vllm.spec_decode.spec_decode_worker.create_spec_worker" "vllm.spec_decode.spec_decode_worker.create_spec_worker"