I've downloaded your space of LocateAnything-3B and run on my vLLM 0.20.0 environment (transformers upgraded to v5), then it raised following error:
[transformers] The tokenizer you are loading from './LocateAnything-3B' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.
[transformers] `LocateAnythingProcessor` defines `image_processor_class = 'AutoImageProcessor'`, which is deprecated. Register the correct mapping in `AutoImageProcessor` instead.
[transformers] The tokenizer you are loading from './LocateAnything-3B' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.
[transformers] `torch_dtype` is deprecated! Use `dtype` instead!
Failed to load model: LocateAnythingPreTrainedModel._check_and_adjust_attn_implementation() got an unexpected keyword argument 'allow_all_kernels'. Will run in Mock Mode.
E:\models\app.py:1087: UserWarning: The parameters have been moved from the Blocks constructor to the launch() method in Gradio 6.0: theme. Please pass these parameters to launch() instead.
with gr.Blocks(
* Running on local URL: http://127.0.0.1:7860
* To create a public link, set `share=True` in `launch()`.
will you fix these soon? and any roadmap for integrate it into vLLM or make a out-of-tree vLLM plugin to let it able to serve on vLLM?
I've downloaded your space of LocateAnything-3B and run on my vLLM 0.20.0 environment (transformers upgraded to v5), then it raised following error:
will you fix these soon? and any roadmap for integrate it into vLLM or make a out-of-tree vLLM plugin to let it able to serve on vLLM?