guanning's picture
add bs256_random run (init_model + ckpt-250..ckpt-4250)
d7b1a64 verified
{
"_from_model_config": true,
"bos_token_id": 1,
"eos_token_id": 2,
"output_attentions": false,
"output_hidden_states": false,
"pad_token_id": 0,
"transformers_version": "5.7.0",
"use_cache": true
}