skittoo
skittoo
I am using 8 GPUs instance but the script uses only GPU:0 not all the GPUs . I get CUDA memory error when i try to make clip_guidance ofViT-L/14 model....
In "combine_docs" in "MapReduceDocumentsChain" class in "langchain/chains/combine_documents/map_reduce.py" num_tokens is defaulted to 3000 not flexible to the model type.
I think adding two example on save_every and evaluate_every in README could be more helpful, after all thanks for this great code
in train.py in line 158: smooth_real_labels = torch.FloatTensor(Utils.smooth_label(real_labels.numpy(), -0.1)) Where is Utils defined?
For CNN links from CNN RSSs, the newspaper library doesn't make any parsing for those! The problem with the newspaper library itself.