调用 google/gemma-2-9b-it 模型报 422 错误

[80imike]

调用 nvidia.com 官方的 google/gemma-2-9b-it 模型报 422 错误，这是报错日志

one-api  | 2024/08/06 - 17:26:55    INFO    GIN request {"status": 422, "request_id": "20240806172654824173977Fm2YBeMJ", "method": "POST", "path": "/v1/chat/completions", "query": "", "ip": "1.2.3.4", "user-agent": "sd/JS 4.54.0", "latency": "691.722531ms"}
one-api  | 2024/08/06 - 17:26:55    ERROR   20240806172654824173977Fm2YBeMJ | relay error (channel #36(api.nvidia.com)): Provider API error: bad response status code 422 (request id: 20240806172654824173977Fm2YBeMJ)

测速时，能通过

one-api  | 2024/08/06 - 17:29:39    INFO    [SYS] | 测试渠道 api.nvidia.com : google/gemma-2-27b-it 返回内容为：{"id":"c27086f7-046a-40a8-b50a-370f9e6762b3","object":"","created":null,"model":"","choices":[{"index":0,"message":{"role":"assistant","content":"hi"},"finish_reason":"stop"}],"usage":{"prompt_tokens":19,"completion_tokens":1,"total_tokens":20}}

渠道通过自定义渠道配置

[MartialBE]

没有使用过 api.nvidia.com， 如果它支持OpenAI API，那么应该是没有问题的