CUDA_VISIBLE_DEVICES=0 accelerate launch --config_file configs/default_config.yaml distill.py \
    --teacher_generation_dataset_path /path/to/teacher/dataset \
    --student_name /path/to/model \
    --output_dir /path/to/output/dir/ \
    --method forward_kl_text2text \
    --use_norm linear \
    --norm_epsilon 1e-6 \

CUDA_VISIBLE_DEVICES=0 accelerate launch --config_file configs/default_config.yaml distill.py \
    --teacher_generation_dataset_path /path/to/teacher/dataset \
    --student_name /path/to/model \
    --output_dir /path/to/output/dir/ \
    --method forward_kl_text2text \
    --use_norm softmax \
    --student_temp 1.0 \
    --teacher_temp 1.0 \

CUDA_VISIBLE_DEVICES=0 accelerate launch --config_file configs/default_config.yaml distill.py \
    --teacher_generation_dataset_path /path/to/teacher/dataset \
    --student_name /path/to/model \
    --output_dir /path/to/output/dir/ \
    --method forward_kl_text2text \
    --use_norm linear \
    --norm_epsilon 1e-6 \
    --user_label_smoothing yes \
    --smoothing_factor 0.1 \

CUDA_VISIBLE_DEVICES=0 accelerate launch --config_file configs/default_config.yaml distill.py \
    --teacher_generation_dataset_path /path/to/teacher/dataset \
    --student_name /path/to/model \
    --output_dir /path/to/output/dir/ \
    --method forward_kl_text2text \
    --use_other_token yes \