system HF staff commited on
Commit
b2ec1a3
1 Parent(s): fc8b0d2

Update config.json

Browse files
Files changed (1) hide show
  1. config.json +29 -5
config.json CHANGED
@@ -1,14 +1,16 @@
1
  {
2
- "activation_dropout": 0.0,
3
  "activation_function": "gelu",
4
  "add_bias_logits": false,
5
  "add_final_layer_norm": false,
6
  "architectures": [
7
- "BartForConditionalGeneration"
 
 
8
  ],
9
- "attention_dropout": 0.0,
10
  "bos_token_id": 0,
11
- "classif_dropout": 0.0,
12
  "d_model": 768,
13
  "decoder_attention_heads": 12,
14
  "decoder_ffn_dim": 3072,
@@ -16,6 +18,7 @@
16
  "decoder_layers": 6,
17
  "decoder_start_token_id": 2,
18
  "dropout": 0.1,
 
19
  "encoder_attention_heads": 12,
20
  "encoder_ffn_dim": 3072,
21
  "encoder_layerdrop": 0.0,
@@ -40,6 +43,27 @@
40
  "num_hidden_layers": 6,
41
  "pad_token_id": 1,
42
  "scale_embedding": false,
43
- "static_position_embeddings": false,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44
  "vocab_size": 50265
45
  }
 
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
+ "BartForConditionalGeneration",
9
+ "BartForSequenceClassification"
10
  ],
11
+ "attention_dropout": 0.1,
12
  "bos_token_id": 0,
13
+ "classif_dropout": 0.1,
14
  "d_model": 768,
15
  "decoder_attention_heads": 12,
16
  "decoder_ffn_dim": 3072,
 
18
  "decoder_layers": 6,
19
  "decoder_start_token_id": 2,
20
  "dropout": 0.1,
21
+ "early_stopping": true,
22
  "encoder_attention_heads": 12,
23
  "encoder_ffn_dim": 3072,
24
  "encoder_layerdrop": 0.0,
 
43
  "num_hidden_layers": 6,
44
  "pad_token_id": 1,
45
  "scale_embedding": false,
46
+ "num_beams": 4,
47
+ "no_repeat_ngram_size": 3,
48
+ "task_specific_params": {
49
+ "summarization": {
50
+ "length_penalty": 1.0,
51
+ "max_length": 128,
52
+ "min_length": 12,
53
+ "num_beams": 4
54
+ },
55
+ "summarization_cnn": {
56
+ "length_penalty": 2.0,
57
+ "max_length": 142,
58
+ "min_length": 56,
59
+ "num_beams": 4
60
+ },
61
+ "summarization_xsum": {
62
+ "length_penalty": 1.0,
63
+ "max_length": 62,
64
+ "min_length": 11,
65
+ "num_beams": 6
66
+ }
67
+ },
68
  "vocab_size": 50265
69
  }