Closed michaeloo0 closed 1 year ago
We reproduced this problem by running the following script:
CUDA_VISIBLE_DEVICES=0 python src/finetune.py \
--do_train \
--dataset guanaco \
--finetuning_type lora \
--output_dir test \
--overwrite_cache \
--overwrite_output_dir \
--per_device_train_batch_size 4 \
--gradient_accumulation_steps 4 \
--lr_scheduler_type cosine \
--logging_steps 10 \
--save_steps 1000 \
--max_train_samples 10000 \
--learning_rate 1e-4 \
--num_train_epochs 1.0 \
--fp16
And everything seems good. Please check your internet connection, or log in with huggingface-cli login
.
We strongly recommend logging in with your HuggingFace account since this dataset requires confirmation before using it.