`model_max_length` why is it 2048?

[denadai2]

Hi everyoneee,

I anticipate that it might be a stupid question but why do we have model_max_length: 2048 in the tokenizer_config.json https://huggingface.co/LargeWorldModel/LWM-Text-Chat-1M/blob/main/tokenizer_config.json?

Thank you!

[LzhinFdu]

I encountered similar problem. When I use LWM-TEXT-512K (pytorch), warning that "Token indices sequence length is longer than the specified maximum sequence length for this model (42314 > 2048). Running this sequence through the model will result in indexing errors". When the input length exceeds 40k, the model will produce confusing output results："['\n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n### \n\n##']"