Update config.json
Browse files- config.json +67 -5
config.json
CHANGED
@@ -1,34 +1,96 @@
|
|
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 |
-
|
12 |
"initializer_factor": 1.0,
|
|
|
13 |
"is_encoder_decoder": true,
|
|
|
14 |
"layer_norm_epsilon": 1e-06,
|
|
|
15 |
"model_type": "t5",
|
|
|
16 |
"n_positions": 512,
|
17 |
-
|
18 |
"num_heads": 12,
|
|
|
19 |
"num_layers": 12,
|
|
|
20 |
"output_past": true,
|
|
|
21 |
"pad_token_id": 0,
|
|
|
22 |
"relative_attention_num_buckets": 32,
|
|
|
23 |
"task_specific_params": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
24 |
"translation_en_to_fr": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
"early_stopping": true,
|
|
|
26 |
"max_length": 32,
|
|
|
27 |
"num_beams": 4,
|
|
|
28 |
"prefix": ""
|
|
|
29 |
}
|
|
|
30 |
},
|
31 |
-
|
32 |
-
"
|
33 |
-
|
34 |
}
|
|
|
|
1 |
{
|
2 |
+
|
3 |
"architectures": [
|
4 |
+
|
5 |
"T5ForConditionalGeneration"
|
6 |
+
|
7 |
],
|
8 |
+
|
9 |
"d_ff": 3072,
|
10 |
+
|
11 |
"d_kv": 64,
|
12 |
+
|
13 |
"d_model": 768,
|
14 |
+
|
15 |
"decoder_start_token_id": 0,
|
16 |
+
|
17 |
"dropout_rate": 0.1,
|
18 |
+
|
19 |
"eos_token_id": 1,
|
20 |
+
|
21 |
"initializer_factor": 1.0,
|
22 |
+
|
23 |
"is_encoder_decoder": true,
|
24 |
+
|
25 |
"layer_norm_epsilon": 1e-06,
|
26 |
+
|
27 |
"model_type": "t5",
|
28 |
+
|
29 |
"n_positions": 512,
|
30 |
+
|
31 |
"num_heads": 12,
|
32 |
+
|
33 |
"num_layers": 12,
|
34 |
+
|
35 |
"output_past": true,
|
36 |
+
|
37 |
"pad_token_id": 0,
|
38 |
+
|
39 |
"relative_attention_num_buckets": 32,
|
40 |
+
|
41 |
"task_specific_params": {
|
42 |
+
|
43 |
+
"summarization": {
|
44 |
+
|
45 |
+
"early_stopping": true,
|
46 |
+
|
47 |
+
"max_length": 32,
|
48 |
+
|
49 |
+
"num_beams": 4,
|
50 |
+
|
51 |
+
"prefix": ""
|
52 |
+
|
53 |
+
},
|
54 |
+
|
55 |
"translation_en_to_fr": {
|
56 |
+
|
57 |
+
"early_stopping": true,
|
58 |
+
|
59 |
+
"max_length": 32,
|
60 |
+
|
61 |
+
"num_beams": 4,
|
62 |
+
|
63 |
+
"prefix": ""
|
64 |
+
|
65 |
+
},
|
66 |
+
|
67 |
+
"translation_en_to_de": {
|
68 |
+
|
69 |
+
"early_stopping": true,
|
70 |
+
|
71 |
+
"max_length": 32,
|
72 |
+
|
73 |
+
"num_beams": 4,
|
74 |
+
|
75 |
+
"prefix": ""
|
76 |
+
|
77 |
+
},
|
78 |
+
|
79 |
+
"translation_en_to_ro": {
|
80 |
+
|
81 |
"early_stopping": true,
|
82 |
+
|
83 |
"max_length": 32,
|
84 |
+
|
85 |
"num_beams": 4,
|
86 |
+
|
87 |
"prefix": ""
|
88 |
+
|
89 |
}
|
90 |
+
|
91 |
},
|
92 |
+
|
93 |
+
"vocab_size": 32101
|
94 |
+
|
95 |
}
|
96 |
+
|