w11wo's picture
Training in progress, epoch 1
f758e25 verified
raw
history blame
1.15 kB
{
"_name_or_path": "flax-community/indonesian-roberta-base",
"architectures": [
"RobertaForTokenClassification"
],
"attention_probs_dropout_prob": 0.1,
"bos_token_id": 0,
"classifier_dropout": null,
"eos_token_id": 2,
"gradient_checkpointing": false,
"hidden_act": "gelu",
"hidden_dropout_prob": 0.1,
"hidden_size": 768,
"id2label": {
"0": "I-PPL",
"1": "B-EVT",
"2": "B-PLC",
"3": "I-IND",
"4": "B-IND",
"5": "B-FNB",
"6": "I-EVT",
"7": "B-PPL",
"8": "I-PLC",
"9": "O",
"10": "I-FNB"
},
"initializer_range": 0.02,
"intermediate_size": 3072,
"label2id": {
"B-EVT": 1,
"B-FNB": 5,
"B-IND": 4,
"B-PLC": 2,
"B-PPL": 7,
"I-EVT": 6,
"I-FNB": 10,
"I-IND": 3,
"I-PLC": 8,
"I-PPL": 0,
"O": 9
},
"layer_norm_eps": 1e-05,
"max_position_embeddings": 514,
"model_type": "roberta",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"pad_token_id": 1,
"position_embedding_type": "absolute",
"torch_dtype": "float32",
"transformers_version": "4.37.2",
"type_vocab_size": 1,
"use_cache": true,
"vocab_size": 50265
}