Closed SkylerZheng closed 2 months ago
Yes, this is a standard issue in GroundingDINO. This happens if your $CUDA_HOME environment variable is not set properly. Please follow the steps here: https://github.com/IDEA-Research/GroundingDINO to resolve this issue.
Yes, this is a standard issue in GroundingDINO. This happens if your $CUDA_HOME environment variable is not set properly. Please follow the steps here: https://github.com/IDEA-Research/GroundingDINO to resolve this issue.
Hey, thank you so much for the quick reply! Many thanks!
I tried my best, I followed the instructions with the link, but I still see the same error. Do you have a docker file for this by any chance? I'm able to do inference with GroundingDINO locally, but cannot use lang_sam for composition purpose. lang_sam will re-install GroundingDINO and SAM, I removed them from the installation config, but still facing the same issue.
Sorry, I do not have a docker for this. You can either disable lang_sam in the code or refer to the web demo we support on Hugging Face that contains a docker file (https://huggingface.co/spaces/fffiloni/RB-Modulation/tree/main).
Hey LituRout, thanks so much!
Hi LituRout, I was able to installed the environment successfully with the huggingface docker. But when I was trying to test the gradio demo, this happened, did you see this before?
This error is due to data type mismatch. Check the dtype of your pipe and the inputs to the denoising network.
Hi, when I was trying to run the jupyter notebook script, I met this issue, this happened in the composition stage. Wondering if you have seen this before. If so, could you share how you resolved the issue? Many thanks!
File /opt/conda/envs/rbm/lib/python3.9/site-packages/groundingdino/models/GroundingDINO/ms_deform_attn.py:53, in MultiScaleDeformableAttnFunction.forward(ctx, value, value_spatial_shapes, value_level_start_index, sampling_locations, attention_weights, im2col_step) 42 @staticmethod 43 def forward( 44 ctx, (...) 50 im2col_step, 51 ): 52 ctx.im2col_step = im2col_step ---> 53 output = _C.ms_deform_attn_forward( 54 value, 55 value_spatial_shapes, 56 value_level_start_index, 57 sampling_locations, 58 attention_weights, 59 ctx.im2col_step, 60 ) 61 ctx.save_for_backward( 62 value, 63 value_spatial_shapes, (...) 66 attention_weights, 67 ) 68 return output
NameError: name '_C' is not defined