feat(*):增加了一些模型

2025-04-18 12:38:06 +08:00
parent a513dd8a91
commit d4c43315fb
19 changed files with 398 additions and 44 deletions
--- a/vllm-qwq-32b/compose.yaml
+++ b/vllm-qwq-32b/compose.yaml
@@ -18,12 +18,12 @@ services:
    runtime: nvidia
    volumes:
      - /home/deepgeek/data/data_local/server/vllm/models:/models
-    restart: always
-    command: --served-model-name QwQ-32B --model /models/Qwen/QwQ-32B
+    restart: unless-stopped
+    command: --served-model-name QwQ:32b --model /models/Qwen/QwQ-32B
      --trust-remote-code --host 0.0.0.0 --port 8080 --max-model-len 8192
      --tensor-parallel-size 2 --gpu_memory_utilization 0.9 --enforce-eager
-      --dtype auto --swap-space 8 --enable-auto-tool-choice
-      --tool-call-parser llama3_json
+      --dtype auto --swap-space 8 --enable-auto-tool-choice --tool-call-parser
+      llama3_json
 x-dockge:
  urls:
    - http://local.citory.tech:10580