[6/N][warmup][DSv4] Migrate sampling, and DFlash JIT kernels (#56323)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com> Signed-off-by: LopezCastroRoberto <rocastro@redhat.com> Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com> Co-authored-by: OpenAI Codex <codex@openai.com>
This commit is contained in:
co-authored by
Lucas Wilkinson
OpenAI Codex
parent
e19a3e172e
commit
aed894c190
@@ -80,16 +80,15 @@ class CPUModelRunner(GPUModelRunner):
|
||||
|
||||
# Speculative decoding fallbacks
|
||||
import vllm.v1.sample.rejection_sampler
|
||||
import vllm.v1.spec_decode.llm_base_proposer
|
||||
import vllm.v1.spec_decode.utils as spec_decode_utils
|
||||
|
||||
vllm.v1.spec_decode.llm_base_proposer.eagle_prepare_inputs_padded_kernel = (
|
||||
spec_decode_utils._eagle_prepare_inputs_padded.kernel = (
|
||||
cpu_tl.eagle_prepare_inputs_padded_kernel
|
||||
)
|
||||
vllm.v1.spec_decode.llm_base_proposer.eagle_prepare_next_token_padded_kernel = (
|
||||
spec_decode_utils._eagle_prepare_next_token_padded.kernel = (
|
||||
cpu_tl.eagle_prepare_next_token_padded_kernel
|
||||
)
|
||||
vllm.v1.spec_decode.llm_base_proposer.copy_and_expand_eagle_inputs_kernel = (
|
||||
spec_decode_utils._copy_and_expand_eagle_inputs.kernel = (
|
||||
cpu_tl.copy_and_expand_eagle_inputs_kernel
|
||||
)
|
||||
spec_decode_utils.copy_and_expand_dflash_inputs_kernel = (
|
||||
|
||||
Reference in New Issue
Block a user