Comments (5)
Closing, as I confirmed this is what is occurring. Please reference #4194 for ongoing refactoring of visual language models.
from vllm.
Hi y'all, I looked into this a bit more. I am using open-webui, and I believe what is occurring is that multiple images are being passed into the model as part of providing context from earlier conversations. I am not 100% sure about this though. If that is the case, then this condition is being hit every time and it is nothing more than that there is no support for multi-image inference right now in vLLM.
from vllm.
Hey @thealmightygrant! If you context does have multiple images (i.e, multiple image_url
) in the messages
, then this is indeed not supported currently by vLLM. (and in fact, I believe all the vision language models we support on vLLM today do not support multi-image themselves either). Thanks!
from vllm.
Oh btw, phi 3 vision is working great for us! Thanks for the hard work @ywang96!
from vllm.
Oh btw, phi 3 vision is working great for us! Thanks for the hard work @ywang96!
Very glad to hear that! :)
from vllm.
Related Issues (20)
- [Feature]: Testing - Use `torch.testing.assert_close` instead of `torch.allclose` as a Recommended Practice
- [Bug][0.5.4] Front-end server errors when overloaded with pending requests
- test
- [Bug]: TPU Dockerfile build fails HOT 1
- [Feature]: Support attention backend with FlexAttention HOT 1
- [Feature]: Improve the compile times of gptq_marlin.cu HOT 2
- [Usage]: Can we get count for current processing tokens from openai api?
- [Feature]: DeepSeek-Coder-V2-Instruct-FP8 on 8xA100
- [Usage]: How to config the parameters to support higher concurrency for deploying the qwen2-7b model as an API at 8-GPU A800 (80G) server? HOT 1
- [Bug]: Endless generation with fine tuned llama 3.1 model
- 为了便于交流,创建了一个多模态大模型交流群,欢迎大家入群交流学习~ HOT 2
- [Usage]: 流式输出前面几个字符为啥要设置成空字符?不能直接输出模型的生成吗 HOT 1
- [Bug]: Compiling FSM index high memory && subprocess OOM
- [Usage]: Does vllm support dynamic quantization
- [Feature]: support voice llm like cosyvoice HOT 1
- [Bug]: Extra body don't work when response_format is also sent for serving. HOT 6
- [Feature]: Small Model Large Latency Compared to SGLang and TensorRT-LLM HOT 1
- [Bug]: `ops.scaled_fp8_quant` returns wrong shape when input shape is () HOT 1
- [Bug]: LLama3 LoRA load failed HOT 1
- [Bug]:`vllm server` will get some error and `python3 -m vllm.entrypoints.openai.api_server` is correct HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from vllm.