coder4nlp

Results 23 comments of coder4nlp

Kosmos 2.5

+1

Kosmos 2.5

Please!

pre-processed data

pre-processed dat +1

DocOwl1.5 training code?

> > > Hi, @coder4nlp , the training code is scheduled for release at the end of this month. If you are urgent to finetune our model, you can refer...

DocOwl1.5 training code?

where are the schedules?

DocOwl1.5 training code?

@HAWLYQ Thank you very much!

The inference time of qwen2.5-vl is very slow.

@kzjeef dashinfer==2.0.0rc3 dashinfer-vlm==2.3.0 transformers==4.51.3 **When using qwen2-vl, the startup failed.** ``` dashinfer_vlm_serve --model /models/Qwen/Qwen2-VL-2B --host 127.0.0.1 ``` ``` Start converting ONNX model! Loading safetensors checkpoint shards: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:01

The inference time of qwen2.5-vl is very slow.

Hi, @kzjeef.Thanks for your response. When using qwen2.5-vl, the service can start normally, but it is extremely slow. Here are my startup commands. **However, when the input consists of multiple...

The inference time of qwen2.5-vl is very slow.

@kzjeef. I have fixed the qwen2-vl issue, but the inference time is still very slow.

The inference time of qwen2.5-vl is very slow.

> > [@kzjeef](https://github.com/kzjeef). I have fixed the qwen2-vl issue, but the inference time is still very slow. > > what's your --vision_engine parameter in qwen2-vl test? --vision_engine parameter has not...