--lora_alpha 64 \ --mixed_precision fp16 \ --output_dir $output_dir \ --max_num_frames 49 \ --train_batch_size 1 \ --max_train_steps $steps \ --checkpointing_steps 1000 \ --gradient_accumulation_steps 1 \ $gradient_checkpointing \ --learning_rate $learning_rate \ --lr_schedul...
-`alpha`:LoRAscalingfactor. -`bias`:Specifiesifthe`bias`parametersshouldbetrained.Canbe`'none'`,`'all'`or`'lora_only'`. -`modules_to_save`:ListofmodulesapartfromLoRAlayerstobesetastrainableandsavedinthefinalcheckpoint.Thesetypicallyincludemodel'scustomheadthatisrandomlyinitializedforthefine-tuningtask...
\ --num_validation_videos 1 \ --validation_epochs 10 \ --seed 42 \ --rank 64 \ --lora_alpha 64 \ --mixed_precision bf16 \ --output_dir $output_dir \ --height 480 --width 720 --fps 8 --max_num_frames 49 --skip_frames_start 0 --skip_frames_end 0 \ --train_batch_size...
🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning. - peft/examples/lora_dreambooth/train_dreambooth.py at main · r-deo/peft