Abu Qader
Abu Qader
Just ran into this issue today with image files (.png to be specific). It occurs every time I open the file in Preview.
30B LoRa adapters here https://huggingface.co/baseten/alpaca-30b
@zsc I used 1xA100 80GB to train the model. For inference, I used the same instance but it can work on a 40GB A100.
@baleksey Yes, definetly. ``` eval_loss: .8007 train_loss: .778 epochs: 3 steps: 1180 ``` As you can see here, and similar to other LoRa runs on this dataset, the train loss...
@baleksey Yes, absolutely will keep you updated :) @dnhkng Yes, you can use them from here https://huggingface.co/baseten/alpaca-30b
+1 here. This is happening for me with a medusa model.
you can reset by ``` rm -rf /dev/shm/nccl-* ```