21 lines
634 B
Bash
21 lines
634 B
Bash
|
./run_clm_flax.py \
|
||
|
|
--output_dir="." \
|
||
|
|
--model_type="gpt2" \
|
||
|
|
--config_name="." \
|
||
|
|
--tokenizer_name="." \
|
||
|
|
--dataset_name="oscar" \
|
||
|
|
--dataset_config_name="unshuffled_deduplicated_pl" \
|
||
|
|
--do_train --do_eval \
|
||
|
|
--block_size="512" \
|
||
|
|
--per_device_train_batch_size="4" \
|
||
|
|
--per_device_eval_batch_size="4" \
|
||
|
|
--learning_rate="3e-4" --warmup_steps="16000" \
|
||
|
|
--adam_beta1="0.9" --adam_beta2="0.98" --weight_decay="0.01" \
|
||
|
|
--overwrite_output_dir \
|
||
|
|
--num_train_epochs="2" \
|
||
|
|
--logging_steps="15000" \
|
||
|
|
--save_steps="30000" \
|
||
|
|
--eval_steps="30000" \
|
||
|
|
--adafactor \
|
||
|
|
--push_to_hub
|