RuntimeError: indices should be either on cpu or on the same device as the indexed tensor (cpu) #38
Description
When i ran model on imagenet: one error happened:
How can i tackle the problem about tensor? Dashen / masters help me , thanks!
Model unexpected keys:
['transformer.log_alpha', 'transformer.log_1_min_alpha', 'transformer.log_cumprod_alpha', 'transformer.log_1_min_cumprod_alpha']
Evaluate EMA model
/home/zvwang/miniconda3/envs/myenv/lib/python3.7/site-packages/torch/nn/functional.py:1967: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.
warnings.warn("nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.")
Traceback (most recent call last):
File "test_imagenet.py", line 5, in
VQ_Diffusion_model.inference_generate_sample_with_class(407, truncation_rate=0.86, save_root="RESULT", batch_size=4)
File "/home/zvwang/VQ-Diffusion/inference_VQ_Diffusion.py", line 93, in inference_generate_sample_with_class
sample_type="top"+str(truncation_rate)+'r',
File "/home/zvwang/miniconda3/envs/myenv/lib/python3.7/site-packages/torch/autograd/grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "/home/zvwang/VQ-Diffusion/image_synthesis/modeling/models/conditional_dalle.py", line 184, in generate_content
content = self.content_codec.decode(trans_out['content_token']) #(8,1024)->(8,3,256,256)
File "/home/zvwang/VQ-Diffusion/image_synthesis/modeling/codecs/image_codec/taming_gumbel_vqvae.py", line 202, in decode
img_seq=self.quantize_to_full[img_seq].type_as(img_seq)
RuntimeError: indices should be either on cpu or on the same device as the indexed tensor (cpu)