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

[Refactor] Simplify BOS/EOS token handling (#34435)

Browse Source 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2026-02-13 10:18:24 +08:00
committed by GitHub
parent 04ea31baab
commit ea5ff3a1f6
29 changed files with 123 additions and 123 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/v1/core/test_prefix_caching.py
View File

@@ -75,13 +75,15 @@ def make_request(
)
mm_features.append(mm_feature)
sampling_params = SamplingParams(max_tokens=17, prompt_logprobs=prompt_logprobs)
sampling_params.update_from_generation_config({}, eos_token_id=100)
return Request(
request_id=request_id,
prompt_token_ids=prompt_token_ids,
mm_features=mm_features if mm_features else None,
sampling_params=SamplingParams(max_tokens=17, prompt_logprobs=prompt_logprobs),
sampling_params=sampling_params,
pooling_params=None,
eos_token_id=100,
lora_request=lora_request,
cache_salt=cache_salt,
block_hasher=get_request_block_hasher(block_size, hash_fn),