-
This issue contains the test results for the upstream sync, develop PR, and release testing branches. Comment 'proceed with rebase' to approve. Close when maintenance is complete or there will be prob…
-
I used the DuckDuckGo search engine. I could not get Devika to run (updated 5 Apr 2024) with either Bing or Google.
I tried to emulate the demo video and get "Conway's Game of Life" running and I g…
-
### System Info
- `transformers` version: 4.39.1
- Platform: Linux-5.15.133.1-microsoft-standard-WSL2-x86_64-with-glibc2.35
- Python version: 3.10.12
- Huggingface_hub version: 0.19.4
- Safetenso…
-
In addition, does `Chat-UniVi-7B v1.5` use `clip-vit-large-patch14-336` and `vicuna-7b-v1.5` just as llava-1.5? It seems that the only difference is the usage of `VQAv2` data according to [this](https…
-
### System Info
v1.4.5
### Information
- [X] Docker
- [X] The CLI directly
### Tasks
- [X] An officially supported command
- [ ] My own modifications
### Reproduction
1. Deploy IDEFICS 9b or 80…
-
### Your current environment
I encountered a few issues while running phi-3-vision with the vllm built from current main branch.
1. Dependency:
`torchvision` is a dependency under [image_pro…
-
Thank you for a very useful add-on.
Would it be possible to grab the text and add it to the PDF, so it's text-searchable?
-
1. Do you have tried using llava-665K, whick is more robust than 158K dataset?
2. When will you release the pretrain score net?
Thanks~
-
Hi thanks for your interesting work. I have two questions.
**Question 1:**
In section 3.3, your conflict elimination loss is said to encourage the reassignment of conflicting tokens to different exp…
-
### 📚 The doc issue
Hey vLLM team it looks like there is added support for llava 1.5 but there are no docs or examples on how to use it via the api server. Are there any reference examples? For using…