@@ -93,6 +93,7 @@ Use explicit config when:
9393 apiKey: "${VLLM_API_KEY}",
9494 api: "openai-completions",
9595 request: { allowPrivateNetwork: true },
96+ timeoutSeconds: 300, // Optional: extend connect/header/body/request timeout for slow local models
9697 models: [
9798 {
9899 id: "your-model-id",
@@ -179,6 +180,7 @@ Use explicit config when:
179180 apiKey: "${VLLM_API_KEY}",
180181 api: "openai-completions",
181182 request: { allowPrivateNetwork: true },
183+ timeoutSeconds: 300,
182184 models: [
183185 {
184186 id: "my-custom-model",
@@ -201,6 +203,34 @@ Use explicit config when:
201203## Troubleshooting
202204203205<AccordionGroup>
206+<Accordion title="Slow first response or remote server timeout">
207+For large local models, remote LAN hosts, or tailnet links, set a
208+provider-scoped request timeout:
209+210+```json5
211+{
212+ models: {
213+ providers: {
214+ vllm: {
215+ baseUrl: "http://192.168.1.50:8000/v1",
216+ apiKey: "${VLLM_API_KEY}",
217+ api: "openai-completions",
218+ request: { allowPrivateNetwork: true },
219+ timeoutSeconds: 300,
220+ models: [{ id: "your-model-id", name: "Local vLLM Model" }],
221+ },
222+ },
223+ },
224+}
225+```
226+227+`timeoutSeconds` applies to vLLM model HTTP requests only, including
228+connection setup, response headers, body streaming, and the total
229+guarded-fetch abort. Prefer this before increasing
230+`agents.defaults.timeoutSeconds`, which controls the whole agent run.
231+232+</Accordion>
233+204234<Accordion title="Server not reachable">
205235Check that the vLLM server is running and accessible:
206236此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。