timmeinhardt
When I was implementing python3 src/track.py with dataset_name=DEMO data_root_dir=data/snakeboard output_dir=data/snakeboard write_images=pretty,I encountered the following problems
0%| | 0/131 [00:02<?, ?it/s] ERROR - track - Failed after 0:00:06! Traceback (most recent calls WITHOUT Sacred internals): File "src/track.py", line 132, in main tracker.step(frame_data) File "/home/lihenian/trackformer-main/src/trackformer/models/tracker.py", line 306, in step outputs, _, features, _, _ = self.obj_detector(img, target, self._prev_features[0]) File "/home/lihenian/.conda/envs/py37/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "/home/lihenian/trackformer-main/src/trackformer/models/detr_tracking.py", line 275, in forward out, targets, features, memory, hs = super().forward(samples, targets, prev_features) File "/home/lihenian/trackformer-main/src/trackformer/models/deformable_detr.py", line 227, in forward self.transformer(src_list, mask_list, pos_list, query_embeds, targets) File "/home/lihenian/.conda/envs/py37/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "/home/lihenian/trackformer-main/src/trackformer/models/deformable_transformer.py", line 166, in forward mask_flatten[:, :src_flatten.shape[1] // 2]) File "/home/lihenian/.conda/envs/py37/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "/home/lihenian/trackformer-main/src/trackformer/models/deformable_transformer.py", line 325, in forward output = layer(output, pos, reference_points, spatial_shapes, padding_mask) File "/home/lihenian/.conda/envs/py37/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "/home/lihenian/trackformer-main/src/trackformer/models/deformable_transformer.py", line 290, in forward src2 = self.self_attn(self.with_pos_embed(src, pos), reference_points, src, spatial_shapes, padding_mask) File "/home/lihenian/.conda/envs/py37/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "/home/lihenian/trackformer-main/src/trackformer/models/ops/modules/ms_deform_attn.py", line 87, in forward value, input_spatial_shapes, sampling_locations, attention_weights, self.im2col_step) File "/home/lihenian/trackformer-main/src/trackformer/models/ops/functions/ms_deform_attn_func.py", line 19, in forward value, value_spatial_shapes, sampling_locations, attention_weights, ctx.im2col_step) RuntimeError: Unrecognized tensor type ID: PythonTLSSnapshot
I just checked the demo code and it works on my side. Did u install the correct PyTorch etc. versions? Can u run the regular MOT17 evaluation of our code or do you always get the same error?
