Yang Fan

Hi @ywang96 @DarkLight1337 , I update some other examples [here](https://github.com/fyabc/vllm/tree/qwen2_omni_public_v1/examples/offline_inference/qwen2_5_omni), please check the code.

> Seems that you messed up the rebase, can you fix it? Sorry, I synced the master branch again.

@DarkLight1337 Hi, how can update typing to fix this error? ``` Error: vllm/multimodal/inputs.py:307: error: Invalid index type "slice | Sequence[slice]" for "list[NestedTensors] | list[Any] | Any | tuple[Any, ...]"; expected...

> Looks like this PR doesn't work with [huggingface/transformers#36752](https://github.com/huggingface/transformers/pull/36752) yet > > ``` > ERROR 03-31 07:57:21 [core.py:377] File "/tmp-nvme/myenv/lib/python3.12/site-packages/transformers/processing_utils.py", line 1082, in from_pretrained > ERROR 03-31 07:57:21 [core.py:377] return...

> @fyabc Sorry for the late response - could you please update this branch with main again? (I was busy in the past a few days) > > Ideally we...

> Is there anything else that needs to be done? Hi, I am adapting the latest version of transformers. Since [this comment](https://github.com/huggingface/transformers/pull/36752/files#r2032910045) was not accepted in final version, we need...

> Hi fyabc, > > Thank you for your excellent work on adding support for Qwen2.5-Omni in vLLM! > > I noticed that the current instructions don’t fully work for...

Hi @chunhuizng , I have updated the PR and fix the bug, please try again.

> Is there anything else that needs to be done? @DarkLight1337 Bug fixed and test all passed, I think this PR is able to merge.

> Can you resolve the failures in the basic models test? Hi @DarkLight1337, I have fixed the test registry, now the api timeout error seems raised outside of thie PR.