sinopec issues

Results 19 issues of


                                            sinopec

Fix training api call failed because of mix use of async and sync func

as [1808](https://github.com/replicate/cog/issues/1808) described: because of training endpoint of fastapi is define as sync and in the sync function a async func is called. Something wrong happened. The curl test command...

locally call train interface failed in cog-sdxl

1. use cog-sdxl in https://github.com/replicate/cog-sdxl 2. run cog build to generage docker image 3. run docker locally use gpu 4. call the train interface usr curl ``` curl -X 'POST'...

details of deepseek-r1's reasoning_content should not return to model for infer

### Self Checks - [x] This is only for bug report, if you would like to ask a question, please head to [Discussions](https://github.com/langgenius/dify/discussions/categories/general). - [x] I have searched for existing...

🐞 bug

How to setup a proxy for marketplace access while offline docker host mode

[Feature Request]: Can we use a VLM to do document parser?

### Is there an existing issue for the same feature request? - [x] I have checked the existing issues. ### Is your feature request related to a problem? ```Markdown Unable...

[Question]: What does "positions" in the reference returned by the chat interface mean?

### Describe your problem I want to develop a flexible interface using the results returned by the chat API, but I don’t understand the meaning of the "positions" field in...

question

The childwork events send deadlock between the subthread write log and async predictor yield a large size output

cog version 0.14.4, async predict , enable concurrency.max = 8 and the async predict func, return a async generator, yield a base64 encoded image with high FPS (10). And I...

Bugfix: fix crash when put with the same id repeatly when use use concurrency and max > 1

When use async concurrency and max > 1, repeatedly put with the same prediction id ,will cause the cog server crash. when put the same id ,we should return its...

crash when put with the same prediction id twice

When I send two PUT requests to the same model using the same prediction ID, the Cog server crashes 100% of the time — of course, the second request is...