inference icon indicating copy to clipboard operation
inference copied to clipboard

Swap GPT for any LLM by changing a single line of code. Xinference lets you run open-source, speech, and multimodal models on cloud, on-prem, or your laptop — all through one unified, production-ready...

Results 526 inference issues
Sort by recently updated
recently updated
newest added

我在使用qwen-vl模型的时候无法使用gpu部署。选择N-GPU选择auto也是跑在cpu上。但是我如果跑qwen-1.5 chat模型,就可以跑在GPU上,这是为什么?

question
gpu

![image](https://github.com/xorbitsai/inference/assets/113818551/efccccff-345f-4e9d-8945-f2babcbe1cd0) ![QQ截图20240502203237](https://github.com/xorbitsai/inference/assets/113818551/27149e6e-50d7-411f-a694-d1959eb4764e)

The qwen model use ReAct to call tools, the token usage may larger than the expected. - Fixes: https://github.com/xorbitsai/inference/issues/1412

bug

### Describe the bug Chatglm3模型返回的token消耗数为-1 ### To Reproduce 请求参数: ```json { "model": "chatglm3", "messages": [ {"role": "user", "content": "你能做什么"} ], "temperature": 0.7, "max_tokens":1024, "tools": [{ "type": "function", "function": { "name":...

### Describe the bug 用xinference0.10.2.post1版本部署了bge-reranker-v2-m3模型,但经常出现如下错误,并且经常会出现服务无法响应的问题,需要重启才能解决: xinference.api.restful_api 22480 ERROR [address=0.0.0.0:44871, pid=22579] 'float' object is not subscriptable Traceback (most recent call last): File "/app/anaconda3/lib/python3.11/site-packages/xinference/api/restful_api.py", line 1067, in rerank scores = await model.rerank(...

- Fix temperature not set. - Fixes: https://github.com/xorbitsai/inference/issues/1387 - Fixes: https://github.com/xorbitsai/inference/issues/1253

feature

![image](https://github.com/xorbitsai/inference/assets/115398481/f5b61aee-a1ff-410f-949c-2c6c60415d19) ![image](https://github.com/xorbitsai/inference/assets/115398481/2b53d367-2a50-462f-8287-be220b71013c)

question

support pipeline parallel inference

question

### 如上述文字 自定义模型0.5B 是json文件这么写不对吗?请指教,json无法解析 ![image](https://github.com/xorbitsai/inference/assets/125621025/873e5d80-2f66-45f3-b234-97e729b81b32) ![image](https://github.com/xorbitsai/inference/assets/125621025/fcf8fb9b-d280-42bb-a812-2e3e822afd83)

对于xinference未官方支持的模型我们常需要自己手动注册,但是有时候并不确定注册的模型类型或者模型名是否合适,当部署失败时常常需要返回修改一些参数,但是目前平台仅支持删除后重建,无法快速修改部分参数,就很不方便,期望可以支持修改的功能

feature