Bhimraj Yadav

**Temporary Workaround:** In the meantime, here's a quick hack to resolve the CORS issue (adding any middleware) until this feature is officially integrated: ```python server = LitServer(....) server.app.add_middleware( CORSMiddleware, allow_origins=["*"],...

Hi @Aceticia, [ParallelStreamingDataset #576](https://github.com/Lightning-AI/litData/pull/576) has just been merged into `litdata`. Feel free to give it a try! You can find more details in the README under the sections **Parallel Streaming**...

Closing this PR for now. Will reopen it if needed.

Hi @XuMayi, this guide might be helpful for your use case: 👉 https://github.com/Lightning-AI/litgpt/blob/main/tutorials/convert_lit_models.md It also includes steps for [merging LoRA weights](https://github.com/Lightning-AI/litgpt/blob/main/tutorials/convert_lit_models.md#merging-lora-weights), which might be relevant for your scenario. Let us...

Closing this issue for now. Please feel free to reopen in case you have any further questions. 😊

> (there was a short thought that I would like to contribute to LitServe, but I would have to rewrite 20% of it and provide proper OpenAI API implementation so...

We are currently experimenting with #467.

Hi @aniketmaurya, I’ve made some modifications to check at specific intervals rather than on each output, hoping this approach might minimize any impact. However, if this PR is affecting performance,...

For a global max limit, 10 to 20 minutes seems reasonable to me. And when needed, we can apply the timeout at the test level for specific functions using `@pytest.mark.timeout(...)`.

Closing this issue — the global max time limit is already handled in CI: https://github.com/Lightning-AI/litData/blob/1a0f48c4a9bcda6b0afe7f8d0d39dc678557f5cf/.github/workflows/ci-testing.yml#L24-L25