RuntimeError: Input type (torch.FloatTensor) ...

[alexander-sony]

Issue Type

Bug

Source

pip (model-compression-toolkit)

MCT Version

2.2.0

OS Platform and Distribution

google colab

Python version

3.10.12

Describe the issue

imx500_notebooks/pytorch/pytorch_yolov8n_seg_for_imx500.ipynb

from tutorials.mct_model_garden.evaluation_metrics.coco_evaluation import evaluate_yolov8_segmentation
evaluate_yolov8_segmentation(quant_model, seg_model_predict, data_dir='coco', data_type='val2017', img_ids_limit=100, output_file='results_quant.json', iou_thresh=0.7, conf=0.001, max_dets=300,mask_thresh=0.55)

loading annotations into memory...
Done (t=0.44s)
creating index...
index created!
Processing Images:   0%|          | 0/100 [00:00<?, ?it/s]
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
[<ipython-input-13-837fbbfbddac>](https://localhost:8080/#) in <cell line: 2>()
      1 from tutorials.mct_model_garden.evaluation_metrics.coco_evaluation import evaluate_yolov8_segmentation
----> 2 evaluate_yolov8_segmentation(quant_model, seg_model_predict, data_dir='coco', data_type='val2017', img_ids_limit=100, output_file='results_quant.json', iou_thresh=0.7, conf=0.001, max_dets=300,mask_thresh=0.55)

12 frames
[/content/tutorials/mct_model_garden/evaluation_metrics/coco_evaluation.py](https://localhost:8080/#) in evaluate_yolov8_segmentation(model, model_predict_func, data_dir, data_type, img_ids_limit, output_file, iou_thresh, conf, max_dets, mask_thresh)
    539 
    540         # Run the model
--> 541         output = model_predict_func(model, input_img)
    542 
    543         #run post processing (nms)

[/content/tutorials/mct_model_garden/models_pytorch/yolov8/yolov8.py](https://localhost:8080/#) in seg_model_predict(model, inputs)
    540     # Run the model
    541     with torch.no_grad():
--> 542         outputs = model(input_tensor)
    543 
    544     return outputs

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py](https://localhost:8080/#) in _wrapped_call_impl(self, *args, **kwargs)
   1551             return self._compiled_call_impl(*args, **kwargs)  # type: ignore[misc]
   1552         else:
-> 1553             return self._call_impl(*args, **kwargs)
   1554 
   1555     def _call_impl(self, *args, **kwargs):

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py](https://localhost:8080/#) in _call_impl(self, *args, **kwargs)
   1560                 or _global_backward_pre_hooks or _global_backward_hooks
   1561                 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1562             return forward_call(*args, **kwargs)
   1563 
   1564         try:

[/usr/local/lib/python3.10/dist-packages/model_compression_toolkit/core/pytorch/back2framework/pytorch_model_builder.py](https://localhost:8080/#) in forward(self, *args)
    315 
    316             # Run node operation and fetch outputs
--> 317             out_tensors_of_n, out_tensors_of_n_float = _run_operation(node,
    318                                                                       input_tensors,
    319                                                                       op_func=op_func,

[/usr/local/lib/python3.10/dist-packages/model_compression_toolkit/core/pytorch/back2framework/pytorch_model_builder.py](https://localhost:8080/#) in _run_operation(n, input_tensors, op_func, quantize_node_activation_fn, use_activation_quantization)
    144         merged_inputs, functional_kwargs = _merge_inputs(n, input_tensors, op_call_args, functional_kwargs.copy(),
    145                                                          tensor_input_allocs=_tensor_input_allocs)
--> 146         out_tensors_of_n_float = op_func(*merged_inputs, **functional_kwargs)
    147 
    148     # Add a fake quant node if the node has an activation threshold.

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py](https://localhost:8080/#) in _wrapped_call_impl(self, *args, **kwargs)
   1551             return self._compiled_call_impl(*args, **kwargs)  # type: ignore[misc]
   1552         else:
-> 1553             return self._call_impl(*args, **kwargs)
   1554 
   1555     def _call_impl(self, *args, **kwargs):

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py](https://localhost:8080/#) in _call_impl(self, *args, **kwargs)
   1560                 or _global_backward_pre_hooks or _global_backward_hooks
   1561                 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1562             return forward_call(*args, **kwargs)
   1563 
   1564         try:

[/usr/local/lib/python3.10/dist-packages/mct_quantizers/pytorch/quantize_wrapper.py](https://localhost:8080/#) in forward(self, *args, **kwargs)
    254                 outputs = self.layer(args, *self.op_call_args, **_kwargs)
    255             else:
--> 256                 outputs = self.layer(*args, *self.op_call_args, **_kwargs)
    257 
    258             return outputs

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py](https://localhost:8080/#) in _wrapped_call_impl(self, *args, **kwargs)
   1551             return self._compiled_call_impl(*args, **kwargs)  # type: ignore[misc]
   1552         else:
-> 1553             return self._call_impl(*args, **kwargs)
   1554 
   1555     def _call_impl(self, *args, **kwargs):

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py](https://localhost:8080/#) in _call_impl(self, *args, **kwargs)
   1560                 or _global_backward_pre_hooks or _global_backward_hooks
   1561                 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1562             return forward_call(*args, **kwargs)
   1563 
   1564         try:

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/conv.py](https://localhost:8080/#) in forward(self, input)
    456 
    457     def forward(self, input: Tensor) -> Tensor:
--> 458         return self._conv_forward(input, self.weight, self.bias)
    459 
    460 class Conv3d(_ConvNd):

[/usr/local/lib/python3.10/dist-packages/torch/nn/modules/conv.py](https://localhost:8080/#) in _conv_forward(self, input, weight, bias)
    452                             weight, bias, self.stride,
    453                             _pair(0), self.dilation, self.groups)
--> 454         return F.conv2d(input, weight, bias, self.stride,
    455                         self.padding, self.dilation, self.groups)
    456 

RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same or input should be a MKLDNN tensor and weight is a dense tensor

Expected behaviour

No response

Code to reproduce the issue

see above

Log output

No response

[samuel-wj-chapman]

Thanks Alex, should be resolved with #1228.

[samuel-wj-chapman]

pr merged