You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
MEMIT is an interesting work.
When I run: python -m experiments.evaluate --alg_name=MEMIT --model_name=EleutherAI/gpt-j-6B --hparams_fname=EleutherAI_gpt-j-6B.json --num_edits=10 --use_cache
There is error about CUDA out of memory:
File "\memit-main\memit\memit_main.py", line 97, in
weights_copy = {k: v.detach().clone() for k, v in weights.items()}
RuntimeError: CUDA out of memory. Tried to allocate 256.00 MiB (GPU 0; 24.00 GiB total capacity; 23.15 GiB already allocated; 0 bytes free; 23.16 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
My local is 24GB 3090 GPU.
Would you help me how to run the MEMIT code ? How to revise the configure file (EleutherAI_gpt-j-6B.json) , in order to reduce memory?
Thank you very much.
MEMIT is very nice.
The text was updated successfully, but these errors were encountered:
Hi authors,
MEMIT is an interesting work.
When I run: python -m experiments.evaluate --alg_name=MEMIT --model_name=EleutherAI/gpt-j-6B --hparams_fname=EleutherAI_gpt-j-6B.json --num_edits=10 --use_cache
There is error about CUDA out of memory:
File "\memit-main\memit\memit_main.py", line 97, in
weights_copy = {k: v.detach().clone() for k, v in weights.items()}
RuntimeError: CUDA out of memory. Tried to allocate 256.00 MiB (GPU 0; 24.00 GiB total capacity; 23.15 GiB already allocated; 0 bytes free; 23.16 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
My local is 24GB 3090 GPU.
Would you help me how to run the MEMIT code ? How to revise the configure file (EleutherAI_gpt-j-6B.json) , in order to reduce memory?
Thank you very much.
MEMIT is very nice.
The text was updated successfully, but these errors were encountered: