compose-vLLM.yml aktualisiert
This commit is contained in:
parent
241338c3e6
commit
1e245cb58b
|
|
@ -35,9 +35,9 @@ services:
|
|||
- "--dtype"
|
||||
- "half"
|
||||
- "--gpu-memory-utilization"
|
||||
- "0.95"
|
||||
- "0.90"
|
||||
- "--max-model-len"
|
||||
- "16384"
|
||||
- "65536"
|
||||
- "--max-num-seqs"
|
||||
- "32"
|
||||
- "--enable-prefix-caching"
|
||||
|
|
|
|||
Loading…
Reference in New Issue