YixuanWeng
commited on
Commit
•
112f102
1
Parent(s):
990307f
Upload config.json
Browse files- config.json +72 -0
config.json
ADDED
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"activation_dropout": 0.1,
|
3 |
+
"activation_function": "gelu",
|
4 |
+
"add_bias_logits": false,
|
5 |
+
"add_final_layer_norm": false,
|
6 |
+
"architectures": [
|
7 |
+
"BartModel"
|
8 |
+
],
|
9 |
+
"attention_dropout": 0.1,
|
10 |
+
"bos_token_id": 101,
|
11 |
+
"classif_dropout": 0.1,
|
12 |
+
"classifier_dropout": 0.0,
|
13 |
+
"d_model": 1024,
|
14 |
+
"decoder_attention_heads": 16,
|
15 |
+
"decoder_ffn_dim": 4096,
|
16 |
+
"decoder_layerdrop": 0.0,
|
17 |
+
"decoder_layers": 4,
|
18 |
+
"decoder_start_token_id": 102,
|
19 |
+
"dropout": 0.1,
|
20 |
+
"early_stopping": true,
|
21 |
+
"encoder_attention_heads": 16,
|
22 |
+
"encoder_ffn_dim": 4096,
|
23 |
+
"encoder_layerdrop": 0.0,
|
24 |
+
"encoder_layers": 24,
|
25 |
+
"eos_token_id": 102,
|
26 |
+
"forced_eos_token_id": 102,
|
27 |
+
"gradient_checkpointing": false,
|
28 |
+
"id2label": {
|
29 |
+
"0": "LABEL_0",
|
30 |
+
"1": "LABEL_1",
|
31 |
+
"2": "LABEL_2"
|
32 |
+
},
|
33 |
+
"init_std": 0.02,
|
34 |
+
"is_encoder_decoder": true,
|
35 |
+
"label2id": {
|
36 |
+
"LABEL_0": 0,
|
37 |
+
"LABEL_1": 1,
|
38 |
+
"LABEL_2": 2
|
39 |
+
},
|
40 |
+
"max_position_embeddings": 1024,
|
41 |
+
"model_type": "bart",
|
42 |
+
"no_repeat_ngram_size": 3,
|
43 |
+
"normalize_before": false,
|
44 |
+
"normalize_embedding": true,
|
45 |
+
"num_beams": 4,
|
46 |
+
"num_hidden_layers": 12,
|
47 |
+
"pad_token_id": 0,
|
48 |
+
"scale_embedding": false,
|
49 |
+
"task_specific_params": {
|
50 |
+
"summarization": {
|
51 |
+
"length_penalty": 1.0,
|
52 |
+
"max_length": 128,
|
53 |
+
"min_length": 12,
|
54 |
+
"num_beams": 4
|
55 |
+
},
|
56 |
+
"summarization_cnn": {
|
57 |
+
"length_penalty": 2.0,
|
58 |
+
"max_length": 142,
|
59 |
+
"min_length": 56,
|
60 |
+
"num_beams": 4
|
61 |
+
},
|
62 |
+
"summarization_xsum": {
|
63 |
+
"length_penalty": 1.0,
|
64 |
+
"max_length": 62,
|
65 |
+
"min_length": 11,
|
66 |
+
"num_beams": 6
|
67 |
+
}
|
68 |
+
},
|
69 |
+
"transformers_version": "4.4.1",
|
70 |
+
"use_cache": true,
|
71 |
+
"vocab_size": 21128
|
72 |
+
}
|