switch to Qwen2.5-7B-Instruct

This commit is contained in:
ViperEkura 2026-06-14 19:08:43 +08:00
parent b3caa7dd7b
commit 62db1a70aa
2 changed files with 8 additions and 9 deletions

View File

@ -8,12 +8,11 @@
"apiKey": "not-needed" "apiKey": "not-needed"
}, },
"models": { "models": {
"GLM-4.6V-Flash": { "Qwen2.5-7B-Instruct": {
"id": "GLM-4.6V-Flash", "id": "Qwen2.5-7B-Instruct",
"name": "GLM-4.6V-Flash", "name": "Qwen2.5-7B-Instruct",
"limit": { "context": 32768, "output": 8192 }, "limit": { "context": 32768, "output": 8192 },
"cost": { "input": 0, "output": 0 }, "cost": { "input": 0, "output": 0 }
"modalities": { "input": ["text", "image"], "output": ["text"] }
} }
} }
} }

View File

@ -20,7 +20,7 @@ def _pick_free_gpu():
if __name__ == "__main__": if __name__ == "__main__":
_pick_free_gpu() _pick_free_gpu()
model = str(Path(__file__).resolve().parent / "models/GLM-4.6V-Flash") model = str(Path(__file__).resolve().parent / "models/Qwen2.5-7B-Instruct")
os.environ["FLASHINFER_DISABLE_VERSION_CHECK"] = "1" os.environ["FLASHINFER_DISABLE_VERSION_CHECK"] = "1"
@ -29,14 +29,14 @@ if __name__ == "__main__":
model, model,
"--host", "0.0.0.0", "--host", "0.0.0.0",
"--port", "8000", "--port", "8000",
"--served-model-name", "GLM-4.6V-Flash", "--served-model-name", "Qwen2.5-7B-Instruct",
"--trust-remote-code", "--trust-remote-code",
"--enable-auto-tool-choice",
"--tool-call-parser", "glm47",
"--max-model-len", "32768", "--max-model-len", "32768",
"--dtype", "bfloat16", "--dtype", "bfloat16",
"--gpu-memory-utilization", "0.85", "--gpu-memory-utilization", "0.85",
"--enforce-eager", "--enforce-eager",
"--enable-auto-tool-choice",
"--tool-call-parser", "openai",
"--generation-config", "vllm", "--generation-config", "vllm",
"--override-generation-config", '{"temperature": 0.8, "top_k": 50, "top_p": 0.9}', "--override-generation-config", '{"temperature": 0.8, "top_k": 50, "top_p": 0.9}',
]) ])