ci: Re-enable onnx test

[chtruong814]

ci: Re-enable onnx test

The onnx export only works with an older transformers. We'll enable it again and communicate via docs update that this is what is necessary for onnx export.

The onnx quantize export was failing with this error:

   File "/usr/local/lib/python3.12/dist-packages/torch/export/_trace.py", line 2122, in _export_for_training
      raise RuntimeError(error_msg)
  RuntimeError: We found a fake tensor in the exported program constant's list. This typically means our tracing system encountered an op that we can't trace through. For the potential source, you can refer to following model attribute: model.layers.0.self_attn.q_proj.input_quantizer.lifted_tensor_0. Please file an issue on github. 

I updated the export code to be similar to how modelopt is handling it. It seems that they default to dynamo=false?:
https://github.com/NVIDIA/Model-Optimizer/blob/main/modelopt/torch/_deploy/utils/torch_onnx.py#L527

tbh, unsure what should be appropriate there