Bug in running the file

[onjas-buidl]

I'm running the following script: CUDA_VISIBLE_DEVICES=4 python3 -i /net/scratch/jasonhu/legal_dec-sum/lex-glue/experiments/ecthr.py --model_name_or_path 'bert-base-uncased' --do_lower_case 'True' --task 'ecthr_a' --output_dir logs/'ecthr_a'/'bert-base-uncased'/seed_1 --do_train --do_eval --do_pred --overwrite_output_dir --load_best_model_at_end --metric_for_best_model micro-f1 --greater_is_better True --evaluation_strategy epoch --save_strategy epoch --save_total_limit 5 --num_train_epochs 20 --learning_rate 3e-5 --per_device_train_batch_size 2 --per_device_eval_batch_size 2 --seed 1 --gradient_accumulation_steps 4 --eval_accumulation_steps 4

And then the following bug occurs:

Tried many ways to solve it but failed, any idea how to tackle this problem? Thanks!

[onjas-buidl]

`max_seq_length: Optional[int] = field(

default=4096,

    default=128,
    metadata={
        "help": "The maximum total input sequence length after tokenization. Sequences longer "
        "than this will be truncated, sequences shorter will be padded."
    },
)`

I supressed the issue by changing the parameter from 4096 to 128.