ch_text_001 / config.json
lixugang's picture
Training in progress, epoch 2
10bff31 verified
{
"_name_or_path": "shibing624/text2vec-base-chinese",
"architectures": [
"BertForSequenceClassification"
],
"attention_probs_dropout_prob": 0.1,
"classifier_dropout": null,
"directionality": "bidi",
"gradient_checkpointing": false,
"hidden_act": "gelu",
"hidden_dropout_prob": 0.1,
"hidden_size": 768,
"id2label": {
"0": "\u8fdd\u6cd5\u60ac\u6302\u5e7f\u544a\u724c\u533e",
"1": "\u5e7f\u544a\u62db\u724c\u7834\u635f",
"2": "\u975e\u6cd5\u6d82\u5199\u5f20\u8d34\u5c0f\u5e7f\u544a",
"3": "\u9053\u8def\u7834\u635f",
"4": "\u4ea4\u901a\u62a4\u680f",
"5": "\u673a\u52a8\u8f66\u4e71\u505c\u653e",
"6": "\u8def\u706f",
"7": "\u5176\u5b83\u8857\u9762\u79e9\u5e8f\u95ee\u9898",
"8": "\u8fdd\u6cd5\u5360\u9053\u5e7f\u544a\u724c",
"9": "\u65bd\u5de5\u5360\u9053(\u65bd\u5de5\u7ef4\u62a4\u810f\u4e71\u5dee)",
"10": "\u5e97\u5916\u7ecf\u8425",
"11": "\u64c5\u81ea\u9972\u517b\u5bb6\u79bd\u5bb6\u755c",
"12": "\u65bd\u5de5\u5e9f\u5f03\u6599",
"13": "\u65e0\u7167\u7ecf\u8425\u6e38\u5546",
"14": "\u6cbf\u8857\u667e\u6302",
"15": "\u66b4\u9732\u5783\u573e",
"16": "\u4e71\u5806\u7269\u5806\u6599"
},
"initializer_range": 0.02,
"intermediate_size": 3072,
"label2id": {
"\u4e71\u5806\u7269\u5806\u6599": 16,
"\u4ea4\u901a\u62a4\u680f": 4,
"\u5176\u5b83\u8857\u9762\u79e9\u5e8f\u95ee\u9898": 7,
"\u5e7f\u544a\u62db\u724c\u7834\u635f": 1,
"\u5e97\u5916\u7ecf\u8425": 10,
"\u64c5\u81ea\u9972\u517b\u5bb6\u79bd\u5bb6\u755c": 11,
"\u65bd\u5de5\u5360\u9053(\u65bd\u5de5\u7ef4\u62a4\u810f\u4e71\u5dee)": 9,
"\u65bd\u5de5\u5e9f\u5f03\u6599": 12,
"\u65e0\u7167\u7ecf\u8425\u6e38\u5546": 13,
"\u66b4\u9732\u5783\u573e": 15,
"\u673a\u52a8\u8f66\u4e71\u505c\u653e": 5,
"\u6cbf\u8857\u667e\u6302": 14,
"\u8def\u706f": 6,
"\u8fdd\u6cd5\u5360\u9053\u5e7f\u544a\u724c": 8,
"\u8fdd\u6cd5\u60ac\u6302\u5e7f\u544a\u724c\u533e": 0,
"\u9053\u8def\u7834\u635f": 3,
"\u975e\u6cd5\u6d82\u5199\u5f20\u8d34\u5c0f\u5e7f\u544a": 2
},
"layer_norm_eps": 1e-12,
"max_position_embeddings": 512,
"model_type": "bert",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"pad_token_id": 0,
"pooler_fc_size": 768,
"pooler_num_attention_heads": 12,
"pooler_num_fc_layers": 3,
"pooler_size_per_head": 128,
"pooler_type": "first_token_transform",
"position_embedding_type": "absolute",
"problem_type": "single_label_classification",
"torch_dtype": "float32",
"transformers_version": "4.36.2",
"type_vocab_size": 2,
"use_cache": true,
"vocab_size": 21128
}