|
{ |
|
"attn_type": "bi", |
|
"bi_data": false, |
|
"bos_token_id": 1, |
|
"clamp_len": -1, |
|
"d_head": 8, |
|
"d_inner": 128, |
|
"d_model": 32, |
|
"dropout": 0.1, |
|
"end_n_top": 5, |
|
"eos_token_id": 2, |
|
"ff_activation": "gelu", |
|
"initializer_range": 0.05, |
|
"layer_norm_eps": 1e-12, |
|
"mem_len": 10, |
|
"model_type": "xlnet", |
|
"n_head": 4, |
|
"n_layer": 5, |
|
"pad_token_id": 5, |
|
"reuse_len": 15, |
|
"same_length": false, |
|
"start_n_top": 5, |
|
"summary_activation": "tanh", |
|
"summary_last_dropout": 0.1, |
|
"summary_type": "last", |
|
"summary_use_proj": true, |
|
"transformers_version": "4.11.0.dev0", |
|
"untie_r": true, |
|
"use_mems_eval": true, |
|
"use_mems_train": false, |
|
"vocab_size": 1009 |
|
} |
|
|