dumplingmiku/ubicloud
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
pykello/resolve-slice
ubicloud/config/ai_models.yml
Benjamin Satzger 8984f5ebe9 Double max_requests for vLLM V1 
vLLM V1 increases the default value for `max_num_seqs` allowing for
higher parallelism. Therefore, `max_requests` is doubled to 1000 for V1.
2025-02-14 16:35:59 +01:00

10 KiB
Raw Permalink Blame History

View Raw
Reference in New Issue View Git Blame Copy Permalink