Bofeng Huang commited on
Commit
f2c0457
·
unverified ·
1 Parent(s): 38c5ead

Update config

Browse files
Files changed (2) hide show
  1. config.json +11 -9
  2. generation_config.json +69 -17
config.json CHANGED
@@ -1,15 +1,17 @@
1
  {
2
- "_name_or_path": "/lustre/fswork/projects/rech/gkb/commun/models/pretrained/whisper/whisper-large-v3_dec2_init_cs",
3
  "activation_dropout": 0.0,
4
  "activation_function": "gelu",
5
- "apply_spec_augment": true,
6
  "architectures": [
7
  "WhisperForConditionalGeneration"
8
  ],
9
  "attention_dropout": 0.0,
10
- "begin_suppress_tokens": null,
 
 
 
11
  "bos_token_id": 50257,
12
- "bpe_dropout": 0.05,
13
  "classifier_proj_size": 256,
14
  "d_model": 1280,
15
  "decoder_attention_heads": 20,
@@ -25,13 +27,13 @@
25
  "eos_token_id": 50257,
26
  "init_std": 0.02,
27
  "is_encoder_decoder": true,
28
- "mask_feature_length": 14,
29
- "mask_feature_min_masks": 2,
30
- "mask_feature_prob": 0.22,
31
  "mask_time_length": 10,
32
  "mask_time_min_masks": 2,
33
- "mask_time_prob": 0.3,
34
- "max_length": null,
35
  "max_source_positions": 1500,
36
  "max_target_positions": 448,
37
  "median_filter_width": 7,
 
1
  {
2
+ "_name_or_path": "",
3
  "activation_dropout": 0.0,
4
  "activation_function": "gelu",
5
+ "apply_spec_augment": false,
6
  "architectures": [
7
  "WhisperForConditionalGeneration"
8
  ],
9
  "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": [
11
+ 220,
12
+ 50257
13
+ ],
14
  "bos_token_id": 50257,
 
15
  "classifier_proj_size": 256,
16
  "d_model": 1280,
17
  "decoder_attention_heads": 20,
 
27
  "eos_token_id": 50257,
28
  "init_std": 0.02,
29
  "is_encoder_decoder": true,
30
+ "mask_feature_length": 10,
31
+ "mask_feature_min_masks": 0,
32
+ "mask_feature_prob": 0.05,
33
  "mask_time_length": 10,
34
  "mask_time_min_masks": 2,
35
+ "mask_time_prob": 0.05,
36
+ "max_length": 448,
37
  "max_source_positions": 1500,
38
  "max_target_positions": 448,
39
  "median_filter_width": 7,
generation_config.json CHANGED
@@ -1,44 +1,84 @@
1
  {
2
  "alignment_heads": [
3
  [
4
- 7,
5
  0
6
  ],
7
  [
8
- 10,
9
- 17
10
  ],
11
  [
12
- 12,
13
- 18
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14
  ],
15
  [
16
- 13,
 
 
 
 
 
 
 
 
17
  12
18
  ],
19
  [
20
- 16,
21
- 1
22
  ],
23
  [
24
- 17,
25
  14
26
  ],
27
  [
28
- 19,
29
- 11
30
  ],
31
  [
32
- 21,
33
- 4
34
  ],
35
  [
36
- 24,
37
- 1
38
  ],
39
  [
40
- 25,
41
- 6
 
 
 
 
42
  ]
43
  ],
44
  "begin_suppress_tokens": [
@@ -48,6 +88,16 @@
48
  "bos_token_id": 50257,
49
  "decoder_start_token_id": 50258,
50
  "eos_token_id": 50257,
 
 
 
 
 
 
 
 
 
 
51
  "is_multilingual": true,
52
  "lang_to_id": {
53
  "<|af|>": 50327,
@@ -151,6 +201,7 @@
151
  "<|yo|>": 50325,
152
  "<|zh|>": 50260
153
  },
 
154
  "max_initial_timestamp_index": 50,
155
  "max_length": 448,
156
  "no_timestamps_token_id": 50364,
@@ -247,6 +298,7 @@
247
  50362,
248
  50363
249
  ],
 
250
  "task_to_id": {
251
  "transcribe": 50360,
252
  "translate": 50359
 
1
  {
2
  "alignment_heads": [
3
  [
4
+ 1,
5
  0
6
  ],
7
  [
8
+ 1,
9
+ 1
10
  ],
11
  [
12
+ 1,
13
+ 2
14
+ ],
15
+ [
16
+ 1,
17
+ 3
18
+ ],
19
+ [
20
+ 1,
21
+ 4
22
+ ],
23
+ [
24
+ 1,
25
+ 5
26
+ ],
27
+ [
28
+ 1,
29
+ 6
30
+ ],
31
+ [
32
+ 1,
33
+ 7
34
+ ],
35
+ [
36
+ 1,
37
+ 8
38
+ ],
39
+ [
40
+ 1,
41
+ 9
42
  ],
43
  [
44
+ 1,
45
+ 10
46
+ ],
47
+ [
48
+ 1,
49
+ 11
50
+ ],
51
+ [
52
+ 1,
53
  12
54
  ],
55
  [
56
+ 1,
57
+ 13
58
  ],
59
  [
60
+ 1,
61
  14
62
  ],
63
  [
64
+ 1,
65
+ 15
66
  ],
67
  [
68
+ 1,
69
+ 16
70
  ],
71
  [
72
+ 1,
73
+ 17
74
  ],
75
  [
76
+ 1,
77
+ 18
78
+ ],
79
+ [
80
+ 1,
81
+ 19
82
  ]
83
  ],
84
  "begin_suppress_tokens": [
 
88
  "bos_token_id": 50257,
89
  "decoder_start_token_id": 50258,
90
  "eos_token_id": 50257,
91
+ "forced_decoder_ids": [
92
+ [
93
+ 1,
94
+ 50358
95
+ ],
96
+ [
97
+ 2,
98
+ 50360
99
+ ]
100
+ ],
101
  "is_multilingual": true,
102
  "lang_to_id": {
103
  "<|af|>": 50327,
 
201
  "<|yo|>": 50325,
202
  "<|zh|>": 50260
203
  },
204
+ "language": "auto",
205
  "max_initial_timestamp_index": 50,
206
  "max_length": 448,
207
  "no_timestamps_token_id": 50364,
 
298
  50362,
299
  50363
300
  ],
301
+ "task": "transcribe",
302
  "task_to_id": {
303
  "transcribe": 50360,
304
  "translate": 50359