t5-base-e2e-qg / config.json
system
Update config.json 3118555
1
{
2
  "architectures": [
3
    "T5ForConditionalGeneration"
4
  ],
5
  "d_ff": 3072,
6
  "d_kv": 64,
7
  "d_model": 768,
8
  "decoder_start_token_id": 0,
9
  "dropout_rate": 0.1,
10
  "eos_token_id": 1,
11
  "initializer_factor": 1.0,
12
  "is_encoder_decoder": true,
13
  "layer_norm_epsilon": 1e-06,
14
  "model_type": "t5",
15
  "n_positions": 512,
16
  "num_heads": 12,
17
  "num_layers": 12,
18
  "output_past": true,
19
  "pad_token_id": 0,
20
  "relative_attention_num_buckets": 32,
21
  "task_specific_params": {
22
    "summarization": {
23
      "early_stopping": true,
24
      "length_penalty": 1.5,
25
      "max_length": 256,
26
      "no_repeat_ngram_size": 3,
27
      "num_beams": 4,
28
      "prefix": "generate questions: "
29
    },
30
    "translation_en_to_de": {
31
      "early_stopping": true,
32
      "length_penalty": 1.5,
33
      "max_length": 256,
34
      "no_repeat_ngram_size": 3,
35
      "num_beams": 4,
36
      "prefix": "generate questions: "
37
    },
38
    "translation_en_to_fr": {
39
      "early_stopping": true,
40
      "length_penalty": 1.5,
41
      "max_length": 256,
42
      "no_repeat_ngram_size": 3,
43
      "num_beams": 4,
44
      "prefix": "generate questions: "
45
    },
46
    "translation_en_to_ro": {
47
      "early_stopping": true,
48
      "length_penalty": 1.5,
49
      "max_length": 256,
50
      "no_repeat_ngram_size": 3,
51
      "num_beams": 4,
52
      "prefix": "generate questions: "
53
    }
54
  },
55
  "vocab_size": 32102
56
}
57