Hugging Face's logo

drygegedsgryeuj
/
whis-small-bn

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
whisper
Inference Endpoints
Model card Files Metrics Community
drygegedsgryeuj commited on
Commit
84a6c34
•
1 Parent(s): 3c84b3e

Upload folder using huggingface_hub

Browse files
Files changed (39) hide show
  1. checkpoint-3000/config.json +52 -0
  2. checkpoint-3000/generation_config.json +264 -0
  3. checkpoint-3000/model.safetensors +3 -0
  4. checkpoint-3000/optimizer.pt +3 -0
  5. checkpoint-3000/preprocessor_config.json +14 -0
  6. checkpoint-3000/rng_state.pth +3 -0
  7. checkpoint-3000/scheduler.pt +3 -0
  8. checkpoint-3000/trainer_state.json +1098 -0
  9. checkpoint-3000/training_args.bin +3 -0
  10. checkpoint-4000/config.json +52 -0
  11. checkpoint-4000/generation_config.json +264 -0
  12. checkpoint-4000/model.safetensors +3 -0
  13. checkpoint-4000/optimizer.pt +3 -0
  14. checkpoint-4000/preprocessor_config.json +14 -0
  15. checkpoint-4000/rng_state.pth +3 -0
  16. checkpoint-4000/scheduler.pt +3 -0
  17. checkpoint-4000/trainer_state.json +1457 -0
  18. checkpoint-4000/training_args.bin +3 -0
  19. checkpoint-5000/config.json +52 -0
  20. checkpoint-5000/generation_config.json +264 -0
  21. checkpoint-5000/model.safetensors +3 -0
  22. checkpoint-5000/optimizer.pt +3 -0
  23. checkpoint-5000/preprocessor_config.json +14 -0
  24. checkpoint-5000/rng_state.pth +3 -0
  25. checkpoint-5000/scheduler.pt +3 -0
  26. checkpoint-5000/trainer_state.json +1816 -0
  27. checkpoint-5000/training_args.bin +3 -0
  28. checkpoint-6000/config.json +52 -0
  29. checkpoint-6000/generation_config.json +264 -0
  30. checkpoint-6000/model.safetensors +3 -0
  31. checkpoint-6000/optimizer.pt +3 -0
  32. checkpoint-6000/preprocessor_config.json +14 -0
  33. checkpoint-6000/rng_state.pth +3 -0
  34. checkpoint-6000/scheduler.pt +3 -0
  35. checkpoint-6000/trainer_state.json +2175 -0
  36. checkpoint-6000/training_args.bin +3 -0
  37. model.safetensors +1 -1
  38. runs/Mar14_08-03-53_841af74821ca/events.out.tfevents.1710403513.841af74821ca.172.0 +3 -0
  39. training_args.bin +1 -1
checkpoint-3000/config.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
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": 768,
17
+ "decoder_attention_heads": 12,
18
+ "decoder_ffn_dim": 3072,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 12,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 12,
24
+ "encoder_ffn_dim": 3072,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 12,
43
+ "num_mel_bins": 80,
44
+ "pad_token_id": 50257,
45
+ "scale_embedding": false,
46
+ "suppress_tokens": [],
47
+ "torch_dtype": "float32",
48
+ "transformers_version": "4.39.0.dev0",
49
+ "use_cache": true,
50
+ "use_weighted_layer_sum": false,
51
+ "vocab_size": 51865
52
+ }
checkpoint-3000/generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 5,
5
+ 3
6
+ ],
7
+ [
8
+ 5,
9
+ 9
10
+ ],
11
+ [
12
+ 8,
13
+ 0
14
+ ],
15
+ [
16
+ 8,
17
+ 4
18
+ ],
19
+ [
20
+ 8,
21
+ 7
22
+ ],
23
+ [
24
+ 8,
25
+ 8
26
+ ],
27
+ [
28
+ 9,
29
+ 0
30
+ ],
31
+ [
32
+ 9,
33
+ 7
34
+ ],
35
+ [
36
+ 9,
37
+ 9
38
+ ],
39
+ [
40
+ 10,
41
+ 5
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50359
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|zh|>": 50260
162
+ },
163
+ "max_initial_timestamp_index": 50,
164
+ "max_length": 448,
165
+ "no_timestamps_token_id": 50363,
166
+ "pad_token_id": 50257,
167
+ "prev_sot_token_id": 50361,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50358,
254
+ 50359,
255
+ 50360,
256
+ 50361,
257
+ 50362
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50359,
261
+ "translate": 50358
262
+ },
263
+ "transformers_version": "4.39.0.dev0"
264
+ }
checkpoint-3000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6457802c24a479e73a21d77b238378099e9dda2a2cd128ba56c6c8b4d7c3f454
3
+ size 966995080
checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d98942decd9ee22b110b9959b46ca39a146e6a353c9bdb0e127deb24b918d7a7
3
+ size 1925064044
checkpoint-3000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 80,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-3000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8b9899f3ff4a12889f25c3abf67d67db764fea482b3db241820e2dc8f90dec1
3
+ size 14244
checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20c5cca67e968287f62ef2bf481f5cdec5b883bd368efd34d1f1250365e22f78
3
+ size 1064
checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,1098 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 55.5672268907563,
3
+ "best_model_checkpoint": "./whisper-small-bn/checkpoint-3000",
4
+ "epoch": 2.5,
5
+ "eval_steps": 1000,
6
+ "global_step": 3000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.07,
13
+ "grad_norm": 27.802248001098633,
14
+ "learning_rate": 3.2e-07,
15
+ "loss": 2.2811,
16
+ "step": 20
17
+ },
18
+ {
19
+ "epoch": 0.13,
20
+ "grad_norm": 18.253421783447266,
21
+ "learning_rate": 7.2e-07,
22
+ "loss": 2.1964,
23
+ "step": 40
24
+ },
25
+ {
26
+ "epoch": 0.2,
27
+ "grad_norm": 13.585515975952148,
28
+ "learning_rate": 1.12e-06,
29
+ "loss": 1.8635,
30
+ "step": 60
31
+ },
32
+ {
33
+ "epoch": 0.27,
34
+ "grad_norm": 22.400362014770508,
35
+ "learning_rate": 1.52e-06,
36
+ "loss": 1.703,
37
+ "step": 80
38
+ },
39
+ {
40
+ "epoch": 0.33,
41
+ "grad_norm": 13.77535343170166,
42
+ "learning_rate": 1.9200000000000003e-06,
43
+ "loss": 1.5336,
44
+ "step": 100
45
+ },
46
+ {
47
+ "epoch": 0.4,
48
+ "grad_norm": 14.989215850830078,
49
+ "learning_rate": 2.3200000000000002e-06,
50
+ "loss": 1.4044,
51
+ "step": 120
52
+ },
53
+ {
54
+ "epoch": 0.47,
55
+ "grad_norm": 22.01858139038086,
56
+ "learning_rate": 2.7200000000000002e-06,
57
+ "loss": 1.2402,
58
+ "step": 140
59
+ },
60
+ {
61
+ "epoch": 0.53,
62
+ "grad_norm": 16.482311248779297,
63
+ "learning_rate": 3.12e-06,
64
+ "loss": 1.0922,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.6,
69
+ "grad_norm": 17.529874801635742,
70
+ "learning_rate": 3.52e-06,
71
+ "loss": 0.9583,
72
+ "step": 180
73
+ },
74
+ {
75
+ "epoch": 0.67,
76
+ "grad_norm": 18.252761840820312,
77
+ "learning_rate": 3.920000000000001e-06,
78
+ "loss": 0.8783,
79
+ "step": 200
80
+ },
81
+ {
82
+ "epoch": 0.73,
83
+ "grad_norm": 16.48067283630371,
84
+ "learning_rate": 4.32e-06,
85
+ "loss": 0.7643,
86
+ "step": 220
87
+ },
88
+ {
89
+ "epoch": 0.8,
90
+ "grad_norm": 15.062904357910156,
91
+ "learning_rate": 4.7200000000000005e-06,
92
+ "loss": 0.6991,
93
+ "step": 240
94
+ },
95
+ {
96
+ "epoch": 0.87,
97
+ "grad_norm": 14.231403350830078,
98
+ "learning_rate": 5.12e-06,
99
+ "loss": 0.6594,
100
+ "step": 260
101
+ },
102
+ {
103
+ "epoch": 0.93,
104
+ "grad_norm": 18.425230026245117,
105
+ "learning_rate": 5.5200000000000005e-06,
106
+ "loss": 0.5724,
107
+ "step": 280
108
+ },
109
+ {
110
+ "epoch": 1.0,
111
+ "grad_norm": 11.786844253540039,
112
+ "learning_rate": 5.92e-06,
113
+ "loss": 0.5687,
114
+ "step": 300
115
+ },
116
+ {
117
+ "epoch": 1.07,
118
+ "grad_norm": 15.346015930175781,
119
+ "learning_rate": 6.3200000000000005e-06,
120
+ "loss": 0.4696,
121
+ "step": 320
122
+ },
123
+ {
124
+ "epoch": 1.13,
125
+ "grad_norm": 7.936849594116211,
126
+ "learning_rate": 6.720000000000001e-06,
127
+ "loss": 0.4569,
128
+ "step": 340
129
+ },
130
+ {
131
+ "epoch": 1.2,
132
+ "grad_norm": 13.894779205322266,
133
+ "learning_rate": 7.1200000000000004e-06,
134
+ "loss": 0.4089,
135
+ "step": 360
136
+ },
137
+ {
138
+ "epoch": 1.27,
139
+ "grad_norm": 11.665081977844238,
140
+ "learning_rate": 7.520000000000001e-06,
141
+ "loss": 0.4168,
142
+ "step": 380
143
+ },
144
+ {
145
+ "epoch": 1.33,
146
+ "grad_norm": 16.7696475982666,
147
+ "learning_rate": 7.92e-06,
148
+ "loss": 0.4433,
149
+ "step": 400
150
+ },
151
+ {
152
+ "epoch": 1.4,
153
+ "grad_norm": 15.60254192352295,
154
+ "learning_rate": 8.32e-06,
155
+ "loss": 0.4013,
156
+ "step": 420
157
+ },
158
+ {
159
+ "epoch": 1.47,
160
+ "grad_norm": 13.72472858428955,
161
+ "learning_rate": 8.720000000000001e-06,
162
+ "loss": 0.3903,
163
+ "step": 440
164
+ },
165
+ {
166
+ "epoch": 1.53,
167
+ "grad_norm": 10.506444931030273,
168
+ "learning_rate": 9.12e-06,
169
+ "loss": 0.3581,
170
+ "step": 460
171
+ },
172
+ {
173
+ "epoch": 1.6,
174
+ "grad_norm": 13.643900871276855,
175
+ "learning_rate": 9.52e-06,
176
+ "loss": 0.3605,
177
+ "step": 480
178
+ },
179
+ {
180
+ "epoch": 1.67,
181
+ "grad_norm": 11.669689178466797,
182
+ "learning_rate": 9.920000000000002e-06,
183
+ "loss": 0.3548,
184
+ "step": 500
185
+ },
186
+ {
187
+ "epoch": 1.73,
188
+ "grad_norm": 9.907474517822266,
189
+ "learning_rate": 9.936000000000001e-06,
190
+ "loss": 0.3535,
191
+ "step": 520
192
+ },
193
+ {
194
+ "epoch": 1.8,
195
+ "grad_norm": 11.366904258728027,
196
+ "learning_rate": 9.856000000000002e-06,
197
+ "loss": 0.2943,
198
+ "step": 540
199
+ },
200
+ {
201
+ "epoch": 1.87,
202
+ "grad_norm": 7.5114240646362305,
203
+ "learning_rate": 9.776000000000001e-06,
204
+ "loss": 0.3057,
205
+ "step": 560
206
+ },
207
+ {
208
+ "epoch": 1.93,
209
+ "grad_norm": 11.717443466186523,
210
+ "learning_rate": 9.696000000000002e-06,
211
+ "loss": 0.2919,
212
+ "step": 580
213
+ },
214
+ {
215
+ "epoch": 2.0,
216
+ "grad_norm": 9.275129318237305,
217
+ "learning_rate": 9.616e-06,
218
+ "loss": 0.3115,
219
+ "step": 600
220
+ },
221
+ {
222
+ "epoch": 2.07,
223
+ "grad_norm": 9.388382911682129,
224
+ "learning_rate": 9.536000000000002e-06,
225
+ "loss": 0.1926,
226
+ "step": 620
227
+ },
228
+ {
229
+ "epoch": 2.13,
230
+ "grad_norm": 8.574433326721191,
231
+ "learning_rate": 9.456e-06,
232
+ "loss": 0.2193,
233
+ "step": 640
234
+ },
235
+ {
236
+ "epoch": 2.2,
237
+ "grad_norm": 5.512338638305664,
238
+ "learning_rate": 9.376000000000001e-06,
239
+ "loss": 0.18,
240
+ "step": 660
241
+ },
242
+ {
243
+ "epoch": 2.27,
244
+ "grad_norm": 9.644317626953125,
245
+ "learning_rate": 9.296e-06,
246
+ "loss": 0.1916,
247
+ "step": 680
248
+ },
249
+ {
250
+ "epoch": 2.33,
251
+ "grad_norm": 9.321474075317383,
252
+ "learning_rate": 9.216000000000001e-06,
253
+ "loss": 0.2055,
254
+ "step": 700
255
+ },
256
+ {
257
+ "epoch": 2.4,
258
+ "grad_norm": 4.868476390838623,
259
+ "learning_rate": 9.136e-06,
260
+ "loss": 0.1843,
261
+ "step": 720
262
+ },
263
+ {
264
+ "epoch": 2.47,
265
+ "grad_norm": 6.592445373535156,
266
+ "learning_rate": 9.056000000000001e-06,
267
+ "loss": 0.1713,
268
+ "step": 740
269
+ },
270
+ {
271
+ "epoch": 2.53,
272
+ "grad_norm": 8.138116836547852,
273
+ "learning_rate": 8.976e-06,
274
+ "loss": 0.1792,
275
+ "step": 760
276
+ },
277
+ {
278
+ "epoch": 2.6,
279
+ "grad_norm": 9.782510757446289,
280
+ "learning_rate": 8.896000000000001e-06,
281
+ "loss": 0.1796,
282
+ "step": 780
283
+ },
284
+ {
285
+ "epoch": 2.67,
286
+ "grad_norm": 5.867671012878418,
287
+ "learning_rate": 8.816000000000002e-06,
288
+ "loss": 0.1652,
289
+ "step": 800
290
+ },
291
+ {
292
+ "epoch": 2.73,
293
+ "grad_norm": 6.637588977813721,
294
+ "learning_rate": 8.736e-06,
295
+ "loss": 0.1735,
296
+ "step": 820
297
+ },
298
+ {
299
+ "epoch": 2.8,
300
+ "grad_norm": 11.652266502380371,
301
+ "learning_rate": 8.656000000000001e-06,
302
+ "loss": 0.1734,
303
+ "step": 840
304
+ },
305
+ {
306
+ "epoch": 2.87,
307
+ "grad_norm": 6.051203727722168,
308
+ "learning_rate": 8.576e-06,
309
+ "loss": 0.1557,
310
+ "step": 860
311
+ },
312
+ {
313
+ "epoch": 2.93,
314
+ "grad_norm": 11.018319129943848,
315
+ "learning_rate": 8.496000000000001e-06,
316
+ "loss": 0.1668,
317
+ "step": 880
318
+ },
319
+ {
320
+ "epoch": 3.0,
321
+ "grad_norm": 7.878730297088623,
322
+ "learning_rate": 8.416e-06,
323
+ "loss": 0.1706,
324
+ "step": 900
325
+ },
326
+ {
327
+ "epoch": 3.07,
328
+ "grad_norm": 6.223515510559082,
329
+ "learning_rate": 8.336000000000001e-06,
330
+ "loss": 0.0917,
331
+ "step": 920
332
+ },
333
+ {
334
+ "epoch": 3.13,
335
+ "grad_norm": 7.449841022491455,
336
+ "learning_rate": 8.256e-06,
337
+ "loss": 0.0853,
338
+ "step": 940
339
+ },
340
+ {
341
+ "epoch": 3.2,
342
+ "grad_norm": 3.824934244155884,
343
+ "learning_rate": 8.176000000000001e-06,
344
+ "loss": 0.1033,
345
+ "step": 960
346
+ },
347
+ {
348
+ "epoch": 3.27,
349
+ "grad_norm": 6.369926929473877,
350
+ "learning_rate": 8.096e-06,
351
+ "loss": 0.0997,
352
+ "step": 980
353
+ },
354
+ {
355
+ "epoch": 3.33,
356
+ "grad_norm": 6.938141822814941,
357
+ "learning_rate": 8.016e-06,
358
+ "loss": 0.095,
359
+ "step": 1000
360
+ },
361
+ {
362
+ "epoch": 3.33,
363
+ "eval_loss": 0.27550819516181946,
364
+ "eval_runtime": 345.1723,
365
+ "eval_samples_per_second": 0.29,
366
+ "eval_steps_per_second": 0.29,
367
+ "eval_wer": 60.18099547511312,
368
+ "step": 1000
369
+ },
370
+ {
371
+ "epoch": 2.18,
372
+ "grad_norm": 7.869150638580322,
373
+ "learning_rate": 7.936e-06,
374
+ "loss": 0.1443,
375
+ "step": 1020
376
+ },
377
+ {
378
+ "epoch": 2.23,
379
+ "grad_norm": 3.6447396278381348,
380
+ "learning_rate": 7.860000000000001e-06,
381
+ "loss": 0.1974,
382
+ "step": 1040
383
+ },
384
+ {
385
+ "epoch": 2.27,
386
+ "grad_norm": 9.588157653808594,
387
+ "learning_rate": 7.78e-06,
388
+ "loss": 0.1427,
389
+ "step": 1060
390
+ },
391
+ {
392
+ "epoch": 2.31,
393
+ "grad_norm": 10.796669006347656,
394
+ "learning_rate": 7.7e-06,
395
+ "loss": 0.187,
396
+ "step": 1080
397
+ },
398
+ {
399
+ "epoch": 2.36,
400
+ "grad_norm": 4.941339015960693,
401
+ "learning_rate": 7.620000000000001e-06,
402
+ "loss": 0.1568,
403
+ "step": 1100
404
+ },
405
+ {
406
+ "epoch": 2.4,
407
+ "grad_norm": 11.114888191223145,
408
+ "learning_rate": 7.540000000000001e-06,
409
+ "loss": 0.1697,
410
+ "step": 1120
411
+ },
412
+ {
413
+ "epoch": 2.44,
414
+ "grad_norm": 10.200749397277832,
415
+ "learning_rate": 7.4600000000000006e-06,
416
+ "loss": 0.1675,
417
+ "step": 1140
418
+ },
419
+ {
420
+ "epoch": 2.48,
421
+ "grad_norm": 5.67927360534668,
422
+ "learning_rate": 7.3800000000000005e-06,
423
+ "loss": 0.1557,
424
+ "step": 1160
425
+ },
426
+ {
427
+ "epoch": 2.53,
428
+ "grad_norm": 7.413832664489746,
429
+ "learning_rate": 7.304000000000001e-06,
430
+ "loss": 0.1629,
431
+ "step": 1180
432
+ },
433
+ {
434
+ "epoch": 2.57,
435
+ "grad_norm": 6.933262348175049,
436
+ "learning_rate": 7.224000000000001e-06,
437
+ "loss": 0.17,
438
+ "step": 1200
439
+ },
440
+ {
441
+ "epoch": 2.61,
442
+ "grad_norm": 4.847285747528076,
443
+ "learning_rate": 7.1440000000000005e-06,
444
+ "loss": 0.1389,
445
+ "step": 1220
446
+ },
447
+ {
448
+ "epoch": 2.66,
449
+ "grad_norm": 19.55495834350586,
450
+ "learning_rate": 7.0640000000000005e-06,
451
+ "loss": 0.1338,
452
+ "step": 1240
453
+ },
454
+ {
455
+ "epoch": 2.7,
456
+ "grad_norm": 7.148222923278809,
457
+ "learning_rate": 6.984e-06,
458
+ "loss": 0.1296,
459
+ "step": 1260
460
+ },
461
+ {
462
+ "epoch": 2.74,
463
+ "grad_norm": 6.323164939880371,
464
+ "learning_rate": 6.904e-06,
465
+ "loss": 0.163,
466
+ "step": 1280
467
+ },
468
+ {
469
+ "epoch": 2.78,
470
+ "grad_norm": 9.096046447753906,
471
+ "learning_rate": 6.824e-06,
472
+ "loss": 0.1942,
473
+ "step": 1300
474
+ },
475
+ {
476
+ "epoch": 2.83,
477
+ "grad_norm": 8.462915420532227,
478
+ "learning_rate": 6.744e-06,
479
+ "loss": 0.1576,
480
+ "step": 1320
481
+ },
482
+ {
483
+ "epoch": 2.87,
484
+ "grad_norm": 7.848359107971191,
485
+ "learning_rate": 6.664e-06,
486
+ "loss": 0.1219,
487
+ "step": 1340
488
+ },
489
+ {
490
+ "epoch": 2.91,
491
+ "grad_norm": 6.30383825302124,
492
+ "learning_rate": 6.584e-06,
493
+ "loss": 0.1404,
494
+ "step": 1360
495
+ },
496
+ {
497
+ "epoch": 2.96,
498
+ "grad_norm": 5.657530784606934,
499
+ "learning_rate": 6.504e-06,
500
+ "loss": 0.1449,
501
+ "step": 1380
502
+ },
503
+ {
504
+ "epoch": 3.0,
505
+ "grad_norm": 7.4172234535217285,
506
+ "learning_rate": 6.424e-06,
507
+ "loss": 0.1234,
508
+ "step": 1400
509
+ },
510
+ {
511
+ "epoch": 3.04,
512
+ "grad_norm": 6.550841331481934,
513
+ "learning_rate": 6.344e-06,
514
+ "loss": 0.0885,
515
+ "step": 1420
516
+ },
517
+ {
518
+ "epoch": 3.08,
519
+ "grad_norm": 6.212351322174072,
520
+ "learning_rate": 6.264e-06,
521
+ "loss": 0.0725,
522
+ "step": 1440
523
+ },
524
+ {
525
+ "epoch": 3.13,
526
+ "grad_norm": 5.2562575340271,
527
+ "learning_rate": 6.184e-06,
528
+ "loss": 0.0886,
529
+ "step": 1460
530
+ },
531
+ {
532
+ "epoch": 3.17,
533
+ "grad_norm": 7.921740531921387,
534
+ "learning_rate": 6.104000000000001e-06,
535
+ "loss": 0.0805,
536
+ "step": 1480
537
+ },
538
+ {
539
+ "epoch": 3.21,
540
+ "grad_norm": 3.5878069400787354,
541
+ "learning_rate": 6.024000000000001e-06,
542
+ "loss": 0.0628,
543
+ "step": 1500
544
+ },
545
+ {
546
+ "epoch": 3.25,
547
+ "grad_norm": 8.585208892822266,
548
+ "learning_rate": 5.944000000000001e-06,
549
+ "loss": 0.0812,
550
+ "step": 1520
551
+ },
552
+ {
553
+ "epoch": 3.3,
554
+ "grad_norm": 4.111368656158447,
555
+ "learning_rate": 5.868e-06,
556
+ "loss": 0.0694,
557
+ "step": 1540
558
+ },
559
+ {
560
+ "epoch": 3.34,
561
+ "grad_norm": 5.304944038391113,
562
+ "learning_rate": 5.788e-06,
563
+ "loss": 0.0739,
564
+ "step": 1560
565
+ },
566
+ {
567
+ "epoch": 3.38,
568
+ "grad_norm": 6.168578147888184,
569
+ "learning_rate": 5.708e-06,
570
+ "loss": 0.0821,
571
+ "step": 1580
572
+ },
573
+ {
574
+ "epoch": 3.43,
575
+ "grad_norm": 12.2472505569458,
576
+ "learning_rate": 5.628e-06,
577
+ "loss": 0.0885,
578
+ "step": 1600
579
+ },
580
+ {
581
+ "epoch": 3.47,
582
+ "grad_norm": 9.349952697753906,
583
+ "learning_rate": 5.548e-06,
584
+ "loss": 0.0607,
585
+ "step": 1620
586
+ },
587
+ {
588
+ "epoch": 3.51,
589
+ "grad_norm": 5.98253870010376,
590
+ "learning_rate": 5.468e-06,
591
+ "loss": 0.0619,
592
+ "step": 1640
593
+ },
594
+ {
595
+ "epoch": 3.55,
596
+ "grad_norm": 7.106723785400391,
597
+ "learning_rate": 5.388e-06,
598
+ "loss": 0.0992,
599
+ "step": 1660
600
+ },
601
+ {
602
+ "epoch": 3.6,
603
+ "grad_norm": 3.693020820617676,
604
+ "learning_rate": 5.308000000000001e-06,
605
+ "loss": 0.0855,
606
+ "step": 1680
607
+ },
608
+ {
609
+ "epoch": 3.64,
610
+ "grad_norm": 7.3649678230285645,
611
+ "learning_rate": 5.228000000000001e-06,
612
+ "loss": 0.081,
613
+ "step": 1700
614
+ },
615
+ {
616
+ "epoch": 3.68,
617
+ "grad_norm": 5.4454240798950195,
618
+ "learning_rate": 5.1480000000000005e-06,
619
+ "loss": 0.0872,
620
+ "step": 1720
621
+ },
622
+ {
623
+ "epoch": 3.73,
624
+ "grad_norm": 8.854460716247559,
625
+ "learning_rate": 5.0680000000000004e-06,
626
+ "loss": 0.0869,
627
+ "step": 1740
628
+ },
629
+ {
630
+ "epoch": 3.77,
631
+ "grad_norm": 3.8615269660949707,
632
+ "learning_rate": 4.988e-06,
633
+ "loss": 0.0821,
634
+ "step": 1760
635
+ },
636
+ {
637
+ "epoch": 3.81,
638
+ "grad_norm": 3.7190287113189697,
639
+ "learning_rate": 4.908e-06,
640
+ "loss": 0.0636,
641
+ "step": 1780
642
+ },
643
+ {
644
+ "epoch": 3.85,
645
+ "grad_norm": 7.683879375457764,
646
+ "learning_rate": 4.828e-06,
647
+ "loss": 0.0892,
648
+ "step": 1800
649
+ },
650
+ {
651
+ "epoch": 3.9,
652
+ "grad_norm": 3.645141839981079,
653
+ "learning_rate": 4.748e-06,
654
+ "loss": 0.0803,
655
+ "step": 1820
656
+ },
657
+ {
658
+ "epoch": 3.94,
659
+ "grad_norm": 3.8167316913604736,
660
+ "learning_rate": 4.668e-06,
661
+ "loss": 0.1012,
662
+ "step": 1840
663
+ },
664
+ {
665
+ "epoch": 3.98,
666
+ "grad_norm": 7.247605323791504,
667
+ "learning_rate": 4.588e-06,
668
+ "loss": 0.0728,
669
+ "step": 1860
670
+ },
671
+ {
672
+ "epoch": 4.03,
673
+ "grad_norm": 3.5495615005493164,
674
+ "learning_rate": 4.508e-06,
675
+ "loss": 0.0439,
676
+ "step": 1880
677
+ },
678
+ {
679
+ "epoch": 4.07,
680
+ "grad_norm": 4.105050086975098,
681
+ "learning_rate": 4.428000000000001e-06,
682
+ "loss": 0.0357,
683
+ "step": 1900
684
+ },
685
+ {
686
+ "epoch": 4.11,
687
+ "grad_norm": 9.399913787841797,
688
+ "learning_rate": 4.3480000000000006e-06,
689
+ "loss": 0.0363,
690
+ "step": 1920
691
+ },
692
+ {
693
+ "epoch": 4.15,
694
+ "grad_norm": 3.0376510620117188,
695
+ "learning_rate": 4.2680000000000005e-06,
696
+ "loss": 0.0221,
697
+ "step": 1940
698
+ },
699
+ {
700
+ "epoch": 4.2,
701
+ "grad_norm": 0.6908012628555298,
702
+ "learning_rate": 4.188e-06,
703
+ "loss": 0.0323,
704
+ "step": 1960
705
+ },
706
+ {
707
+ "epoch": 4.24,
708
+ "grad_norm": 6.255786895751953,
709
+ "learning_rate": 4.108e-06,
710
+ "loss": 0.0327,
711
+ "step": 1980
712
+ },
713
+ {
714
+ "epoch": 4.28,
715
+ "grad_norm": 7.1336669921875,
716
+ "learning_rate": 4.028e-06,
717
+ "loss": 0.0416,
718
+ "step": 2000
719
+ },
720
+ {
721
+ "epoch": 4.28,
722
+ "eval_loss": 0.27948442101478577,
723
+ "eval_runtime": 349.9899,
724
+ "eval_samples_per_second": 0.286,
725
+ "eval_steps_per_second": 0.286,
726
+ "eval_wer": 59.0146750524109,
727
+ "step": 2000
728
+ },
729
+ {
730
+ "epoch": 1.68,
731
+ "grad_norm": 13.214738845825195,
732
+ "learning_rate": 7.250909090909092e-06,
733
+ "loss": 0.2839,
734
+ "step": 2020
735
+ },
736
+ {
737
+ "epoch": 1.7,
738
+ "grad_norm": 8.846616744995117,
739
+ "learning_rate": 7.214545454545455e-06,
740
+ "loss": 0.2221,
741
+ "step": 2040
742
+ },
743
+ {
744
+ "epoch": 1.72,
745
+ "grad_norm": 11.778766632080078,
746
+ "learning_rate": 7.178181818181818e-06,
747
+ "loss": 0.2394,
748
+ "step": 2060
749
+ },
750
+ {
751
+ "epoch": 1.73,
752
+ "grad_norm": 4.484768867492676,
753
+ "learning_rate": 7.141818181818182e-06,
754
+ "loss": 0.2299,
755
+ "step": 2080
756
+ },
757
+ {
758
+ "epoch": 1.75,
759
+ "grad_norm": 4.385435104370117,
760
+ "learning_rate": 7.105454545454546e-06,
761
+ "loss": 0.2221,
762
+ "step": 2100
763
+ },
764
+ {
765
+ "epoch": 1.77,
766
+ "grad_norm": 8.751326560974121,
767
+ "learning_rate": 7.06909090909091e-06,
768
+ "loss": 0.1966,
769
+ "step": 2120
770
+ },
771
+ {
772
+ "epoch": 1.78,
773
+ "grad_norm": 8.837333679199219,
774
+ "learning_rate": 7.032727272727273e-06,
775
+ "loss": 0.1893,
776
+ "step": 2140
777
+ },
778
+ {
779
+ "epoch": 1.8,
780
+ "grad_norm": 9.331002235412598,
781
+ "learning_rate": 6.998181818181818e-06,
782
+ "loss": 0.2365,
783
+ "step": 2160
784
+ },
785
+ {
786
+ "epoch": 1.82,
787
+ "grad_norm": 6.599325180053711,
788
+ "learning_rate": 6.961818181818183e-06,
789
+ "loss": 0.2152,
790
+ "step": 2180
791
+ },
792
+ {
793
+ "epoch": 1.83,
794
+ "grad_norm": 11.716184616088867,
795
+ "learning_rate": 6.9254545454545464e-06,
796
+ "loss": 0.203,
797
+ "step": 2200
798
+ },
799
+ {
800
+ "epoch": 1.85,
801
+ "grad_norm": 6.506683349609375,
802
+ "learning_rate": 6.88909090909091e-06,
803
+ "loss": 0.2162,
804
+ "step": 2220
805
+ },
806
+ {
807
+ "epoch": 1.87,
808
+ "grad_norm": 4.64941930770874,
809
+ "learning_rate": 6.852727272727273e-06,
810
+ "loss": 0.214,
811
+ "step": 2240
812
+ },
813
+ {
814
+ "epoch": 1.88,
815
+ "grad_norm": 8.344452857971191,
816
+ "learning_rate": 6.816363636363637e-06,
817
+ "loss": 0.2106,
818
+ "step": 2260
819
+ },
820
+ {
821
+ "epoch": 1.9,
822
+ "grad_norm": 5.504006385803223,
823
+ "learning_rate": 6.780000000000001e-06,
824
+ "loss": 0.196,
825
+ "step": 2280
826
+ },
827
+ {
828
+ "epoch": 1.92,
829
+ "grad_norm": 9.84432601928711,
830
+ "learning_rate": 6.743636363636365e-06,
831
+ "loss": 0.2207,
832
+ "step": 2300
833
+ },
834
+ {
835
+ "epoch": 1.93,
836
+ "grad_norm": 7.112894535064697,
837
+ "learning_rate": 6.707272727272728e-06,
838
+ "loss": 0.2072,
839
+ "step": 2320
840
+ },
841
+ {
842
+ "epoch": 1.95,
843
+ "grad_norm": 7.10872745513916,
844
+ "learning_rate": 6.670909090909091e-06,
845
+ "loss": 0.1925,
846
+ "step": 2340
847
+ },
848
+ {
849
+ "epoch": 1.97,
850
+ "grad_norm": 8.612652778625488,
851
+ "learning_rate": 6.634545454545455e-06,
852
+ "loss": 0.2023,
853
+ "step": 2360
854
+ },
855
+ {
856
+ "epoch": 1.98,
857
+ "grad_norm": 6.083488941192627,
858
+ "learning_rate": 6.5981818181818195e-06,
859
+ "loss": 0.2111,
860
+ "step": 2380
861
+ },
862
+ {
863
+ "epoch": 2.0,
864
+ "grad_norm": 7.59145975112915,
865
+ "learning_rate": 6.561818181818182e-06,
866
+ "loss": 0.1934,
867
+ "step": 2400
868
+ },
869
+ {
870
+ "epoch": 2.02,
871
+ "grad_norm": 6.529219627380371,
872
+ "learning_rate": 6.525454545454546e-06,
873
+ "loss": 0.1756,
874
+ "step": 2420
875
+ },
876
+ {
877
+ "epoch": 2.03,
878
+ "grad_norm": 13.685791015625,
879
+ "learning_rate": 6.48909090909091e-06,
880
+ "loss": 0.1872,
881
+ "step": 2440
882
+ },
883
+ {
884
+ "epoch": 2.05,
885
+ "grad_norm": 6.508156776428223,
886
+ "learning_rate": 6.4527272727272725e-06,
887
+ "loss": 0.1424,
888
+ "step": 2460
889
+ },
890
+ {
891
+ "epoch": 2.07,
892
+ "grad_norm": 4.847611427307129,
893
+ "learning_rate": 6.416363636363637e-06,
894
+ "loss": 0.1588,
895
+ "step": 2480
896
+ },
897
+ {
898
+ "epoch": 2.08,
899
+ "grad_norm": 5.715017795562744,
900
+ "learning_rate": 6.380000000000001e-06,
901
+ "loss": 0.1896,
902
+ "step": 2500
903
+ },
904
+ {
905
+ "epoch": 2.1,
906
+ "grad_norm": 10.562045097351074,
907
+ "learning_rate": 6.3436363636363644e-06,
908
+ "loss": 0.2007,
909
+ "step": 2520
910
+ },
911
+ {
912
+ "epoch": 2.12,
913
+ "grad_norm": 3.2870736122131348,
914
+ "learning_rate": 6.307272727272727e-06,
915
+ "loss": 0.179,
916
+ "step": 2540
917
+ },
918
+ {
919
+ "epoch": 2.13,
920
+ "grad_norm": 11.021241188049316,
921
+ "learning_rate": 6.270909090909092e-06,
922
+ "loss": 0.1573,
923
+ "step": 2560
924
+ },
925
+ {
926
+ "epoch": 2.15,
927
+ "grad_norm": 7.566544055938721,
928
+ "learning_rate": 6.2345454545454555e-06,
929
+ "loss": 0.1832,
930
+ "step": 2580
931
+ },
932
+ {
933
+ "epoch": 2.17,
934
+ "grad_norm": 5.980641841888428,
935
+ "learning_rate": 6.198181818181819e-06,
936
+ "loss": 0.1498,
937
+ "step": 2600
938
+ },
939
+ {
940
+ "epoch": 2.18,
941
+ "grad_norm": 5.697175025939941,
942
+ "learning_rate": 6.161818181818182e-06,
943
+ "loss": 0.1378,
944
+ "step": 2620
945
+ },
946
+ {
947
+ "epoch": 2.2,
948
+ "grad_norm": 4.451406002044678,
949
+ "learning_rate": 6.125454545454546e-06,
950
+ "loss": 0.152,
951
+ "step": 2640
952
+ },
953
+ {
954
+ "epoch": 2.22,
955
+ "grad_norm": 5.854424476623535,
956
+ "learning_rate": 6.08909090909091e-06,
957
+ "loss": 0.1492,
958
+ "step": 2660
959
+ },
960
+ {
961
+ "epoch": 2.23,
962
+ "grad_norm": 6.480546951293945,
963
+ "learning_rate": 6.052727272727274e-06,
964
+ "loss": 0.1501,
965
+ "step": 2680
966
+ },
967
+ {
968
+ "epoch": 2.25,
969
+ "grad_norm": 3.3347442150115967,
970
+ "learning_rate": 6.016363636363637e-06,
971
+ "loss": 0.132,
972
+ "step": 2700
973
+ },
974
+ {
975
+ "epoch": 2.27,
976
+ "grad_norm": 5.9716081619262695,
977
+ "learning_rate": 5.98e-06,
978
+ "loss": 0.1483,
979
+ "step": 2720
980
+ },
981
+ {
982
+ "epoch": 2.28,
983
+ "grad_norm": 5.842201232910156,
984
+ "learning_rate": 5.943636363636364e-06,
985
+ "loss": 0.1688,
986
+ "step": 2740
987
+ },
988
+ {
989
+ "epoch": 2.3,
990
+ "grad_norm": 5.079952716827393,
991
+ "learning_rate": 5.9072727272727285e-06,
992
+ "loss": 0.167,
993
+ "step": 2760
994
+ },
995
+ {
996
+ "epoch": 2.32,
997
+ "grad_norm": 9.07481575012207,
998
+ "learning_rate": 5.870909090909091e-06,
999
+ "loss": 0.1581,
1000
+ "step": 2780
1001
+ },
1002
+ {
1003
+ "epoch": 2.33,
1004
+ "grad_norm": 9.30324935913086,
1005
+ "learning_rate": 5.834545454545455e-06,
1006
+ "loss": 0.1704,
1007
+ "step": 2800
1008
+ },
1009
+ {
1010
+ "epoch": 2.35,
1011
+ "grad_norm": 4.848841667175293,
1012
+ "learning_rate": 5.798181818181819e-06,
1013
+ "loss": 0.1894,
1014
+ "step": 2820
1015
+ },
1016
+ {
1017
+ "epoch": 2.37,
1018
+ "grad_norm": 7.52855110168457,
1019
+ "learning_rate": 5.7618181818181816e-06,
1020
+ "loss": 0.1702,
1021
+ "step": 2840
1022
+ },
1023
+ {
1024
+ "epoch": 2.38,
1025
+ "grad_norm": 6.624313831329346,
1026
+ "learning_rate": 5.725454545454546e-06,
1027
+ "loss": 0.1628,
1028
+ "step": 2860
1029
+ },
1030
+ {
1031
+ "epoch": 2.4,
1032
+ "grad_norm": 6.889793872833252,
1033
+ "learning_rate": 5.68909090909091e-06,
1034
+ "loss": 0.1705,
1035
+ "step": 2880
1036
+ },
1037
+ {
1038
+ "epoch": 2.42,
1039
+ "grad_norm": 5.635292053222656,
1040
+ "learning_rate": 5.6527272727272734e-06,
1041
+ "loss": 0.1531,
1042
+ "step": 2900
1043
+ },
1044
+ {
1045
+ "epoch": 2.43,
1046
+ "grad_norm": 4.1545729637146,
1047
+ "learning_rate": 5.616363636363636e-06,
1048
+ "loss": 0.1371,
1049
+ "step": 2920
1050
+ },
1051
+ {
1052
+ "epoch": 2.45,
1053
+ "grad_norm": 7.1985697746276855,
1054
+ "learning_rate": 5.580000000000001e-06,
1055
+ "loss": 0.1571,
1056
+ "step": 2940
1057
+ },
1058
+ {
1059
+ "epoch": 2.47,
1060
+ "grad_norm": 7.112518310546875,
1061
+ "learning_rate": 5.5436363636363645e-06,
1062
+ "loss": 0.1525,
1063
+ "step": 2960
1064
+ },
1065
+ {
1066
+ "epoch": 2.48,
1067
+ "grad_norm": 9.503728866577148,
1068
+ "learning_rate": 5.507272727272728e-06,
1069
+ "loss": 0.1426,
1070
+ "step": 2980
1071
+ },
1072
+ {
1073
+ "epoch": 2.5,
1074
+ "grad_norm": 5.497040271759033,
1075
+ "learning_rate": 5.470909090909091e-06,
1076
+ "loss": 0.1501,
1077
+ "step": 3000
1078
+ },
1079
+ {
1080
+ "epoch": 2.5,
1081
+ "eval_loss": 0.2098405510187149,
1082
+ "eval_runtime": 325.0771,
1083
+ "eval_samples_per_second": 0.308,
1084
+ "eval_steps_per_second": 0.308,
1085
+ "eval_wer": 55.5672268907563,
1086
+ "step": 3000
1087
+ }
1088
+ ],
1089
+ "logging_steps": 20,
1090
+ "max_steps": 6000,
1091
+ "num_input_tokens_seen": 0,
1092
+ "num_train_epochs": 5,
1093
+ "save_steps": 1000,
1094
+ "total_flos": 2.59669143207936e+18,
1095
+ "train_batch_size": 3,
1096
+ "trial_name": null,
1097
+ "trial_params": null
1098
+ }
checkpoint-3000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:152d1e9770ea3e9899a6c5c40e00cad1f52c4e2ae9aea55ed85a6bf2ed0a638f
3
+ size 5048
checkpoint-4000/config.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
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": 768,
17
+ "decoder_attention_heads": 12,
18
+ "decoder_ffn_dim": 3072,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 12,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 12,
24
+ "encoder_ffn_dim": 3072,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 12,
43
+ "num_mel_bins": 80,
44
+ "pad_token_id": 50257,
45
+ "scale_embedding": false,
46
+ "suppress_tokens": [],
47
+ "torch_dtype": "float32",
48
+ "transformers_version": "4.39.0.dev0",
49
+ "use_cache": true,
50
+ "use_weighted_layer_sum": false,
51
+ "vocab_size": 51865
52
+ }
checkpoint-4000/generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 5,
5
+ 3
6
+ ],
7
+ [
8
+ 5,
9
+ 9
10
+ ],
11
+ [
12
+ 8,
13
+ 0
14
+ ],
15
+ [
16
+ 8,
17
+ 4
18
+ ],
19
+ [
20
+ 8,
21
+ 7
22
+ ],
23
+ [
24
+ 8,
25
+ 8
26
+ ],
27
+ [
28
+ 9,
29
+ 0
30
+ ],
31
+ [
32
+ 9,
33
+ 7
34
+ ],
35
+ [
36
+ 9,
37
+ 9
38
+ ],
39
+ [
40
+ 10,
41
+ 5
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50359
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|zh|>": 50260
162
+ },
163
+ "max_initial_timestamp_index": 50,
164
+ "max_length": 448,
165
+ "no_timestamps_token_id": 50363,
166
+ "pad_token_id": 50257,
167
+ "prev_sot_token_id": 50361,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50358,
254
+ 50359,
255
+ 50360,
256
+ 50361,
257
+ 50362
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50359,
261
+ "translate": 50358
262
+ },
263
+ "transformers_version": "4.39.0.dev0"
264
+ }
checkpoint-4000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:914c7f805651d751363b4c6e184b27cbfc1697c66217d8d8067891f18531c45d
3
+ size 966995080
checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31da532c5ba4276ca3977556c649de6b88c9fbd6e38eb0ba2f09ad5058bfbe5e
3
+ size 1925064044
checkpoint-4000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 80,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-4000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:373f5daeea2274708f0e16b0e772da6cf436a4796973b28677cdd5711b2f0d22
3
+ size 14244
checkpoint-4000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48d67d553f2cba98b86e399bd267f11ee8c54067edace80afb250e82daaa48e3
3
+ size 1064
checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,1457 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 51.78571428571429,
3
+ "best_model_checkpoint": "./whisper-small-bn/checkpoint-4000",
4
+ "epoch": 3.3333333333333335,
5
+ "eval_steps": 1000,
6
+ "global_step": 4000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.07,
13
+ "grad_norm": 27.802248001098633,
14
+ "learning_rate": 3.2e-07,
15
+ "loss": 2.2811,
16
+ "step": 20
17
+ },
18
+ {
19
+ "epoch": 0.13,
20
+ "grad_norm": 18.253421783447266,
21
+ "learning_rate": 7.2e-07,
22
+ "loss": 2.1964,
23
+ "step": 40
24
+ },
25
+ {
26
+ "epoch": 0.2,
27
+ "grad_norm": 13.585515975952148,
28
+ "learning_rate": 1.12e-06,
29
+ "loss": 1.8635,
30
+ "step": 60
31
+ },
32
+ {
33
+ "epoch": 0.27,
34
+ "grad_norm": 22.400362014770508,
35
+ "learning_rate": 1.52e-06,
36
+ "loss": 1.703,
37
+ "step": 80
38
+ },
39
+ {
40
+ "epoch": 0.33,
41
+ "grad_norm": 13.77535343170166,
42
+ "learning_rate": 1.9200000000000003e-06,
43
+ "loss": 1.5336,
44
+ "step": 100
45
+ },
46
+ {
47
+ "epoch": 0.4,
48
+ "grad_norm": 14.989215850830078,
49
+ "learning_rate": 2.3200000000000002e-06,
50
+ "loss": 1.4044,
51
+ "step": 120
52
+ },
53
+ {
54
+ "epoch": 0.47,
55
+ "grad_norm": 22.01858139038086,
56
+ "learning_rate": 2.7200000000000002e-06,
57
+ "loss": 1.2402,
58
+ "step": 140
59
+ },
60
+ {
61
+ "epoch": 0.53,
62
+ "grad_norm": 16.482311248779297,
63
+ "learning_rate": 3.12e-06,
64
+ "loss": 1.0922,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.6,
69
+ "grad_norm": 17.529874801635742,
70
+ "learning_rate": 3.52e-06,
71
+ "loss": 0.9583,
72
+ "step": 180
73
+ },
74
+ {
75
+ "epoch": 0.67,
76
+ "grad_norm": 18.252761840820312,
77
+ "learning_rate": 3.920000000000001e-06,
78
+ "loss": 0.8783,
79
+ "step": 200
80
+ },
81
+ {
82
+ "epoch": 0.73,
83
+ "grad_norm": 16.48067283630371,
84
+ "learning_rate": 4.32e-06,
85
+ "loss": 0.7643,
86
+ "step": 220
87
+ },
88
+ {
89
+ "epoch": 0.8,
90
+ "grad_norm": 15.062904357910156,
91
+ "learning_rate": 4.7200000000000005e-06,
92
+ "loss": 0.6991,
93
+ "step": 240
94
+ },
95
+ {
96
+ "epoch": 0.87,
97
+ "grad_norm": 14.231403350830078,
98
+ "learning_rate": 5.12e-06,
99
+ "loss": 0.6594,
100
+ "step": 260
101
+ },
102
+ {
103
+ "epoch": 0.93,
104
+ "grad_norm": 18.425230026245117,
105
+ "learning_rate": 5.5200000000000005e-06,
106
+ "loss": 0.5724,
107
+ "step": 280
108
+ },
109
+ {
110
+ "epoch": 1.0,
111
+ "grad_norm": 11.786844253540039,
112
+ "learning_rate": 5.92e-06,
113
+ "loss": 0.5687,
114
+ "step": 300
115
+ },
116
+ {
117
+ "epoch": 1.07,
118
+ "grad_norm": 15.346015930175781,
119
+ "learning_rate": 6.3200000000000005e-06,
120
+ "loss": 0.4696,
121
+ "step": 320
122
+ },
123
+ {
124
+ "epoch": 1.13,
125
+ "grad_norm": 7.936849594116211,
126
+ "learning_rate": 6.720000000000001e-06,
127
+ "loss": 0.4569,
128
+ "step": 340
129
+ },
130
+ {
131
+ "epoch": 1.2,
132
+ "grad_norm": 13.894779205322266,
133
+ "learning_rate": 7.1200000000000004e-06,
134
+ "loss": 0.4089,
135
+ "step": 360
136
+ },
137
+ {
138
+ "epoch": 1.27,
139
+ "grad_norm": 11.665081977844238,
140
+ "learning_rate": 7.520000000000001e-06,
141
+ "loss": 0.4168,
142
+ "step": 380
143
+ },
144
+ {
145
+ "epoch": 1.33,
146
+ "grad_norm": 16.7696475982666,
147
+ "learning_rate": 7.92e-06,
148
+ "loss": 0.4433,
149
+ "step": 400
150
+ },
151
+ {
152
+ "epoch": 1.4,
153
+ "grad_norm": 15.60254192352295,
154
+ "learning_rate": 8.32e-06,
155
+ "loss": 0.4013,
156
+ "step": 420
157
+ },
158
+ {
159
+ "epoch": 1.47,
160
+ "grad_norm": 13.72472858428955,
161
+ "learning_rate": 8.720000000000001e-06,
162
+ "loss": 0.3903,
163
+ "step": 440
164
+ },
165
+ {
166
+ "epoch": 1.53,
167
+ "grad_norm": 10.506444931030273,
168
+ "learning_rate": 9.12e-06,
169
+ "loss": 0.3581,
170
+ "step": 460
171
+ },
172
+ {
173
+ "epoch": 1.6,
174
+ "grad_norm": 13.643900871276855,
175
+ "learning_rate": 9.52e-06,
176
+ "loss": 0.3605,
177
+ "step": 480
178
+ },
179
+ {
180
+ "epoch": 1.67,
181
+ "grad_norm": 11.669689178466797,
182
+ "learning_rate": 9.920000000000002e-06,
183
+ "loss": 0.3548,
184
+ "step": 500
185
+ },
186
+ {
187
+ "epoch": 1.73,
188
+ "grad_norm": 9.907474517822266,
189
+ "learning_rate": 9.936000000000001e-06,
190
+ "loss": 0.3535,
191
+ "step": 520
192
+ },
193
+ {
194
+ "epoch": 1.8,
195
+ "grad_norm": 11.366904258728027,
196
+ "learning_rate": 9.856000000000002e-06,
197
+ "loss": 0.2943,
198
+ "step": 540
199
+ },
200
+ {
201
+ "epoch": 1.87,
202
+ "grad_norm": 7.5114240646362305,
203
+ "learning_rate": 9.776000000000001e-06,
204
+ "loss": 0.3057,
205
+ "step": 560
206
+ },
207
+ {
208
+ "epoch": 1.93,
209
+ "grad_norm": 11.717443466186523,
210
+ "learning_rate": 9.696000000000002e-06,
211
+ "loss": 0.2919,
212
+ "step": 580
213
+ },
214
+ {
215
+ "epoch": 2.0,
216
+ "grad_norm": 9.275129318237305,
217
+ "learning_rate": 9.616e-06,
218
+ "loss": 0.3115,
219
+ "step": 600
220
+ },
221
+ {
222
+ "epoch": 2.07,
223
+ "grad_norm": 9.388382911682129,
224
+ "learning_rate": 9.536000000000002e-06,
225
+ "loss": 0.1926,
226
+ "step": 620
227
+ },
228
+ {
229
+ "epoch": 2.13,
230
+ "grad_norm": 8.574433326721191,
231
+ "learning_rate": 9.456e-06,
232
+ "loss": 0.2193,
233
+ "step": 640
234
+ },
235
+ {
236
+ "epoch": 2.2,
237
+ "grad_norm": 5.512338638305664,
238
+ "learning_rate": 9.376000000000001e-06,
239
+ "loss": 0.18,
240
+ "step": 660
241
+ },
242
+ {
243
+ "epoch": 2.27,
244
+ "grad_norm": 9.644317626953125,
245
+ "learning_rate": 9.296e-06,
246
+ "loss": 0.1916,
247
+ "step": 680
248
+ },
249
+ {
250
+ "epoch": 2.33,
251
+ "grad_norm": 9.321474075317383,
252
+ "learning_rate": 9.216000000000001e-06,
253
+ "loss": 0.2055,
254
+ "step": 700
255
+ },
256
+ {
257
+ "epoch": 2.4,
258
+ "grad_norm": 4.868476390838623,
259
+ "learning_rate": 9.136e-06,
260
+ "loss": 0.1843,
261
+ "step": 720
262
+ },
263
+ {
264
+ "epoch": 2.47,
265
+ "grad_norm": 6.592445373535156,
266
+ "learning_rate": 9.056000000000001e-06,
267
+ "loss": 0.1713,
268
+ "step": 740
269
+ },
270
+ {
271
+ "epoch": 2.53,
272
+ "grad_norm": 8.138116836547852,
273
+ "learning_rate": 8.976e-06,
274
+ "loss": 0.1792,
275
+ "step": 760
276
+ },
277
+ {
278
+ "epoch": 2.6,
279
+ "grad_norm": 9.782510757446289,
280
+ "learning_rate": 8.896000000000001e-06,
281
+ "loss": 0.1796,
282
+ "step": 780
283
+ },
284
+ {
285
+ "epoch": 2.67,
286
+ "grad_norm": 5.867671012878418,
287
+ "learning_rate": 8.816000000000002e-06,
288
+ "loss": 0.1652,
289
+ "step": 800
290
+ },
291
+ {
292
+ "epoch": 2.73,
293
+ "grad_norm": 6.637588977813721,
294
+ "learning_rate": 8.736e-06,
295
+ "loss": 0.1735,
296
+ "step": 820
297
+ },
298
+ {
299
+ "epoch": 2.8,
300
+ "grad_norm": 11.652266502380371,
301
+ "learning_rate": 8.656000000000001e-06,
302
+ "loss": 0.1734,
303
+ "step": 840
304
+ },
305
+ {
306
+ "epoch": 2.87,
307
+ "grad_norm": 6.051203727722168,
308
+ "learning_rate": 8.576e-06,
309
+ "loss": 0.1557,
310
+ "step": 860
311
+ },
312
+ {
313
+ "epoch": 2.93,
314
+ "grad_norm": 11.018319129943848,
315
+ "learning_rate": 8.496000000000001e-06,
316
+ "loss": 0.1668,
317
+ "step": 880
318
+ },
319
+ {
320
+ "epoch": 3.0,
321
+ "grad_norm": 7.878730297088623,
322
+ "learning_rate": 8.416e-06,
323
+ "loss": 0.1706,
324
+ "step": 900
325
+ },
326
+ {
327
+ "epoch": 3.07,
328
+ "grad_norm": 6.223515510559082,
329
+ "learning_rate": 8.336000000000001e-06,
330
+ "loss": 0.0917,
331
+ "step": 920
332
+ },
333
+ {
334
+ "epoch": 3.13,
335
+ "grad_norm": 7.449841022491455,
336
+ "learning_rate": 8.256e-06,
337
+ "loss": 0.0853,
338
+ "step": 940
339
+ },
340
+ {
341
+ "epoch": 3.2,
342
+ "grad_norm": 3.824934244155884,
343
+ "learning_rate": 8.176000000000001e-06,
344
+ "loss": 0.1033,
345
+ "step": 960
346
+ },
347
+ {
348
+ "epoch": 3.27,
349
+ "grad_norm": 6.369926929473877,
350
+ "learning_rate": 8.096e-06,
351
+ "loss": 0.0997,
352
+ "step": 980
353
+ },
354
+ {
355
+ "epoch": 3.33,
356
+ "grad_norm": 6.938141822814941,
357
+ "learning_rate": 8.016e-06,
358
+ "loss": 0.095,
359
+ "step": 1000
360
+ },
361
+ {
362
+ "epoch": 3.33,
363
+ "eval_loss": 0.27550819516181946,
364
+ "eval_runtime": 345.1723,
365
+ "eval_samples_per_second": 0.29,
366
+ "eval_steps_per_second": 0.29,
367
+ "eval_wer": 60.18099547511312,
368
+ "step": 1000
369
+ },
370
+ {
371
+ "epoch": 2.18,
372
+ "grad_norm": 7.869150638580322,
373
+ "learning_rate": 7.936e-06,
374
+ "loss": 0.1443,
375
+ "step": 1020
376
+ },
377
+ {
378
+ "epoch": 2.23,
379
+ "grad_norm": 3.6447396278381348,
380
+ "learning_rate": 7.860000000000001e-06,
381
+ "loss": 0.1974,
382
+ "step": 1040
383
+ },
384
+ {
385
+ "epoch": 2.27,
386
+ "grad_norm": 9.588157653808594,
387
+ "learning_rate": 7.78e-06,
388
+ "loss": 0.1427,
389
+ "step": 1060
390
+ },
391
+ {
392
+ "epoch": 2.31,
393
+ "grad_norm": 10.796669006347656,
394
+ "learning_rate": 7.7e-06,
395
+ "loss": 0.187,
396
+ "step": 1080
397
+ },
398
+ {
399
+ "epoch": 2.36,
400
+ "grad_norm": 4.941339015960693,
401
+ "learning_rate": 7.620000000000001e-06,
402
+ "loss": 0.1568,
403
+ "step": 1100
404
+ },
405
+ {
406
+ "epoch": 2.4,
407
+ "grad_norm": 11.114888191223145,
408
+ "learning_rate": 7.540000000000001e-06,
409
+ "loss": 0.1697,
410
+ "step": 1120
411
+ },
412
+ {
413
+ "epoch": 2.44,
414
+ "grad_norm": 10.200749397277832,
415
+ "learning_rate": 7.4600000000000006e-06,
416
+ "loss": 0.1675,
417
+ "step": 1140
418
+ },
419
+ {
420
+ "epoch": 2.48,
421
+ "grad_norm": 5.67927360534668,
422
+ "learning_rate": 7.3800000000000005e-06,
423
+ "loss": 0.1557,
424
+ "step": 1160
425
+ },
426
+ {
427
+ "epoch": 2.53,
428
+ "grad_norm": 7.413832664489746,
429
+ "learning_rate": 7.304000000000001e-06,
430
+ "loss": 0.1629,
431
+ "step": 1180
432
+ },
433
+ {
434
+ "epoch": 2.57,
435
+ "grad_norm": 6.933262348175049,
436
+ "learning_rate": 7.224000000000001e-06,
437
+ "loss": 0.17,
438
+ "step": 1200
439
+ },
440
+ {
441
+ "epoch": 2.61,
442
+ "grad_norm": 4.847285747528076,
443
+ "learning_rate": 7.1440000000000005e-06,
444
+ "loss": 0.1389,
445
+ "step": 1220
446
+ },
447
+ {
448
+ "epoch": 2.66,
449
+ "grad_norm": 19.55495834350586,
450
+ "learning_rate": 7.0640000000000005e-06,
451
+ "loss": 0.1338,
452
+ "step": 1240
453
+ },
454
+ {
455
+ "epoch": 2.7,
456
+ "grad_norm": 7.148222923278809,
457
+ "learning_rate": 6.984e-06,
458
+ "loss": 0.1296,
459
+ "step": 1260
460
+ },
461
+ {
462
+ "epoch": 2.74,
463
+ "grad_norm": 6.323164939880371,
464
+ "learning_rate": 6.904e-06,
465
+ "loss": 0.163,
466
+ "step": 1280
467
+ },
468
+ {
469
+ "epoch": 2.78,
470
+ "grad_norm": 9.096046447753906,
471
+ "learning_rate": 6.824e-06,
472
+ "loss": 0.1942,
473
+ "step": 1300
474
+ },
475
+ {
476
+ "epoch": 2.83,
477
+ "grad_norm": 8.462915420532227,
478
+ "learning_rate": 6.744e-06,
479
+ "loss": 0.1576,
480
+ "step": 1320
481
+ },
482
+ {
483
+ "epoch": 2.87,
484
+ "grad_norm": 7.848359107971191,
485
+ "learning_rate": 6.664e-06,
486
+ "loss": 0.1219,
487
+ "step": 1340
488
+ },
489
+ {
490
+ "epoch": 2.91,
491
+ "grad_norm": 6.30383825302124,
492
+ "learning_rate": 6.584e-06,
493
+ "loss": 0.1404,
494
+ "step": 1360
495
+ },
496
+ {
497
+ "epoch": 2.96,
498
+ "grad_norm": 5.657530784606934,
499
+ "learning_rate": 6.504e-06,
500
+ "loss": 0.1449,
501
+ "step": 1380
502
+ },
503
+ {
504
+ "epoch": 3.0,
505
+ "grad_norm": 7.4172234535217285,
506
+ "learning_rate": 6.424e-06,
507
+ "loss": 0.1234,
508
+ "step": 1400
509
+ },
510
+ {
511
+ "epoch": 3.04,
512
+ "grad_norm": 6.550841331481934,
513
+ "learning_rate": 6.344e-06,
514
+ "loss": 0.0885,
515
+ "step": 1420
516
+ },
517
+ {
518
+ "epoch": 3.08,
519
+ "grad_norm": 6.212351322174072,
520
+ "learning_rate": 6.264e-06,
521
+ "loss": 0.0725,
522
+ "step": 1440
523
+ },
524
+ {
525
+ "epoch": 3.13,
526
+ "grad_norm": 5.2562575340271,
527
+ "learning_rate": 6.184e-06,
528
+ "loss": 0.0886,
529
+ "step": 1460
530
+ },
531
+ {
532
+ "epoch": 3.17,
533
+ "grad_norm": 7.921740531921387,
534
+ "learning_rate": 6.104000000000001e-06,
535
+ "loss": 0.0805,
536
+ "step": 1480
537
+ },
538
+ {
539
+ "epoch": 3.21,
540
+ "grad_norm": 3.5878069400787354,
541
+ "learning_rate": 6.024000000000001e-06,
542
+ "loss": 0.0628,
543
+ "step": 1500
544
+ },
545
+ {
546
+ "epoch": 3.25,
547
+ "grad_norm": 8.585208892822266,
548
+ "learning_rate": 5.944000000000001e-06,
549
+ "loss": 0.0812,
550
+ "step": 1520
551
+ },
552
+ {
553
+ "epoch": 3.3,
554
+ "grad_norm": 4.111368656158447,
555
+ "learning_rate": 5.868e-06,
556
+ "loss": 0.0694,
557
+ "step": 1540
558
+ },
559
+ {
560
+ "epoch": 3.34,
561
+ "grad_norm": 5.304944038391113,
562
+ "learning_rate": 5.788e-06,
563
+ "loss": 0.0739,
564
+ "step": 1560
565
+ },
566
+ {
567
+ "epoch": 3.38,
568
+ "grad_norm": 6.168578147888184,
569
+ "learning_rate": 5.708e-06,
570
+ "loss": 0.0821,
571
+ "step": 1580
572
+ },
573
+ {
574
+ "epoch": 3.43,
575
+ "grad_norm": 12.2472505569458,
576
+ "learning_rate": 5.628e-06,
577
+ "loss": 0.0885,
578
+ "step": 1600
579
+ },
580
+ {
581
+ "epoch": 3.47,
582
+ "grad_norm": 9.349952697753906,
583
+ "learning_rate": 5.548e-06,
584
+ "loss": 0.0607,
585
+ "step": 1620
586
+ },
587
+ {
588
+ "epoch": 3.51,
589
+ "grad_norm": 5.98253870010376,
590
+ "learning_rate": 5.468e-06,
591
+ "loss": 0.0619,
592
+ "step": 1640
593
+ },
594
+ {
595
+ "epoch": 3.55,
596
+ "grad_norm": 7.106723785400391,
597
+ "learning_rate": 5.388e-06,
598
+ "loss": 0.0992,
599
+ "step": 1660
600
+ },
601
+ {
602
+ "epoch": 3.6,
603
+ "grad_norm": 3.693020820617676,
604
+ "learning_rate": 5.308000000000001e-06,
605
+ "loss": 0.0855,
606
+ "step": 1680
607
+ },
608
+ {
609
+ "epoch": 3.64,
610
+ "grad_norm": 7.3649678230285645,
611
+ "learning_rate": 5.228000000000001e-06,
612
+ "loss": 0.081,
613
+ "step": 1700
614
+ },
615
+ {
616
+ "epoch": 3.68,
617
+ "grad_norm": 5.4454240798950195,
618
+ "learning_rate": 5.1480000000000005e-06,
619
+ "loss": 0.0872,
620
+ "step": 1720
621
+ },
622
+ {
623
+ "epoch": 3.73,
624
+ "grad_norm": 8.854460716247559,
625
+ "learning_rate": 5.0680000000000004e-06,
626
+ "loss": 0.0869,
627
+ "step": 1740
628
+ },
629
+ {
630
+ "epoch": 3.77,
631
+ "grad_norm": 3.8615269660949707,
632
+ "learning_rate": 4.988e-06,
633
+ "loss": 0.0821,
634
+ "step": 1760
635
+ },
636
+ {
637
+ "epoch": 3.81,
638
+ "grad_norm": 3.7190287113189697,
639
+ "learning_rate": 4.908e-06,
640
+ "loss": 0.0636,
641
+ "step": 1780
642
+ },
643
+ {
644
+ "epoch": 3.85,
645
+ "grad_norm": 7.683879375457764,
646
+ "learning_rate": 4.828e-06,
647
+ "loss": 0.0892,
648
+ "step": 1800
649
+ },
650
+ {
651
+ "epoch": 3.9,
652
+ "grad_norm": 3.645141839981079,
653
+ "learning_rate": 4.748e-06,
654
+ "loss": 0.0803,
655
+ "step": 1820
656
+ },
657
+ {
658
+ "epoch": 3.94,
659
+ "grad_norm": 3.8167316913604736,
660
+ "learning_rate": 4.668e-06,
661
+ "loss": 0.1012,
662
+ "step": 1840
663
+ },
664
+ {
665
+ "epoch": 3.98,
666
+ "grad_norm": 7.247605323791504,
667
+ "learning_rate": 4.588e-06,
668
+ "loss": 0.0728,
669
+ "step": 1860
670
+ },
671
+ {
672
+ "epoch": 4.03,
673
+ "grad_norm": 3.5495615005493164,
674
+ "learning_rate": 4.508e-06,
675
+ "loss": 0.0439,
676
+ "step": 1880
677
+ },
678
+ {
679
+ "epoch": 4.07,
680
+ "grad_norm": 4.105050086975098,
681
+ "learning_rate": 4.428000000000001e-06,
682
+ "loss": 0.0357,
683
+ "step": 1900
684
+ },
685
+ {
686
+ "epoch": 4.11,
687
+ "grad_norm": 9.399913787841797,
688
+ "learning_rate": 4.3480000000000006e-06,
689
+ "loss": 0.0363,
690
+ "step": 1920
691
+ },
692
+ {
693
+ "epoch": 4.15,
694
+ "grad_norm": 3.0376510620117188,
695
+ "learning_rate": 4.2680000000000005e-06,
696
+ "loss": 0.0221,
697
+ "step": 1940
698
+ },
699
+ {
700
+ "epoch": 4.2,
701
+ "grad_norm": 0.6908012628555298,
702
+ "learning_rate": 4.188e-06,
703
+ "loss": 0.0323,
704
+ "step": 1960
705
+ },
706
+ {
707
+ "epoch": 4.24,
708
+ "grad_norm": 6.255786895751953,
709
+ "learning_rate": 4.108e-06,
710
+ "loss": 0.0327,
711
+ "step": 1980
712
+ },
713
+ {
714
+ "epoch": 4.28,
715
+ "grad_norm": 7.1336669921875,
716
+ "learning_rate": 4.028e-06,
717
+ "loss": 0.0416,
718
+ "step": 2000
719
+ },
720
+ {
721
+ "epoch": 4.28,
722
+ "eval_loss": 0.27948442101478577,
723
+ "eval_runtime": 349.9899,
724
+ "eval_samples_per_second": 0.286,
725
+ "eval_steps_per_second": 0.286,
726
+ "eval_wer": 59.0146750524109,
727
+ "step": 2000
728
+ },
729
+ {
730
+ "epoch": 1.68,
731
+ "grad_norm": 13.214738845825195,
732
+ "learning_rate": 7.250909090909092e-06,
733
+ "loss": 0.2839,
734
+ "step": 2020
735
+ },
736
+ {
737
+ "epoch": 1.7,
738
+ "grad_norm": 8.846616744995117,
739
+ "learning_rate": 7.214545454545455e-06,
740
+ "loss": 0.2221,
741
+ "step": 2040
742
+ },
743
+ {
744
+ "epoch": 1.72,
745
+ "grad_norm": 11.778766632080078,
746
+ "learning_rate": 7.178181818181818e-06,
747
+ "loss": 0.2394,
748
+ "step": 2060
749
+ },
750
+ {
751
+ "epoch": 1.73,
752
+ "grad_norm": 4.484768867492676,
753
+ "learning_rate": 7.141818181818182e-06,
754
+ "loss": 0.2299,
755
+ "step": 2080
756
+ },
757
+ {
758
+ "epoch": 1.75,
759
+ "grad_norm": 4.385435104370117,
760
+ "learning_rate": 7.105454545454546e-06,
761
+ "loss": 0.2221,
762
+ "step": 2100
763
+ },
764
+ {
765
+ "epoch": 1.77,
766
+ "grad_norm": 8.751326560974121,
767
+ "learning_rate": 7.06909090909091e-06,
768
+ "loss": 0.1966,
769
+ "step": 2120
770
+ },
771
+ {
772
+ "epoch": 1.78,
773
+ "grad_norm": 8.837333679199219,
774
+ "learning_rate": 7.032727272727273e-06,
775
+ "loss": 0.1893,
776
+ "step": 2140
777
+ },
778
+ {
779
+ "epoch": 1.8,
780
+ "grad_norm": 9.331002235412598,
781
+ "learning_rate": 6.998181818181818e-06,
782
+ "loss": 0.2365,
783
+ "step": 2160
784
+ },
785
+ {
786
+ "epoch": 1.82,
787
+ "grad_norm": 6.599325180053711,
788
+ "learning_rate": 6.961818181818183e-06,
789
+ "loss": 0.2152,
790
+ "step": 2180
791
+ },
792
+ {
793
+ "epoch": 1.83,
794
+ "grad_norm": 11.716184616088867,
795
+ "learning_rate": 6.9254545454545464e-06,
796
+ "loss": 0.203,
797
+ "step": 2200
798
+ },
799
+ {
800
+ "epoch": 1.85,
801
+ "grad_norm": 6.506683349609375,
802
+ "learning_rate": 6.88909090909091e-06,
803
+ "loss": 0.2162,
804
+ "step": 2220
805
+ },
806
+ {
807
+ "epoch": 1.87,
808
+ "grad_norm": 4.64941930770874,
809
+ "learning_rate": 6.852727272727273e-06,
810
+ "loss": 0.214,
811
+ "step": 2240
812
+ },
813
+ {
814
+ "epoch": 1.88,
815
+ "grad_norm": 8.344452857971191,
816
+ "learning_rate": 6.816363636363637e-06,
817
+ "loss": 0.2106,
818
+ "step": 2260
819
+ },
820
+ {
821
+ "epoch": 1.9,
822
+ "grad_norm": 5.504006385803223,
823
+ "learning_rate": 6.780000000000001e-06,
824
+ "loss": 0.196,
825
+ "step": 2280
826
+ },
827
+ {
828
+ "epoch": 1.92,
829
+ "grad_norm": 9.84432601928711,
830
+ "learning_rate": 6.743636363636365e-06,
831
+ "loss": 0.2207,
832
+ "step": 2300
833
+ },
834
+ {
835
+ "epoch": 1.93,
836
+ "grad_norm": 7.112894535064697,
837
+ "learning_rate": 6.707272727272728e-06,
838
+ "loss": 0.2072,
839
+ "step": 2320
840
+ },
841
+ {
842
+ "epoch": 1.95,
843
+ "grad_norm": 7.10872745513916,
844
+ "learning_rate": 6.670909090909091e-06,
845
+ "loss": 0.1925,
846
+ "step": 2340
847
+ },
848
+ {
849
+ "epoch": 1.97,
850
+ "grad_norm": 8.612652778625488,
851
+ "learning_rate": 6.634545454545455e-06,
852
+ "loss": 0.2023,
853
+ "step": 2360
854
+ },
855
+ {
856
+ "epoch": 1.98,
857
+ "grad_norm": 6.083488941192627,
858
+ "learning_rate": 6.5981818181818195e-06,
859
+ "loss": 0.2111,
860
+ "step": 2380
861
+ },
862
+ {
863
+ "epoch": 2.0,
864
+ "grad_norm": 7.59145975112915,
865
+ "learning_rate": 6.561818181818182e-06,
866
+ "loss": 0.1934,
867
+ "step": 2400
868
+ },
869
+ {
870
+ "epoch": 2.02,
871
+ "grad_norm": 6.529219627380371,
872
+ "learning_rate": 6.525454545454546e-06,
873
+ "loss": 0.1756,
874
+ "step": 2420
875
+ },
876
+ {
877
+ "epoch": 2.03,
878
+ "grad_norm": 13.685791015625,
879
+ "learning_rate": 6.48909090909091e-06,
880
+ "loss": 0.1872,
881
+ "step": 2440
882
+ },
883
+ {
884
+ "epoch": 2.05,
885
+ "grad_norm": 6.508156776428223,
886
+ "learning_rate": 6.4527272727272725e-06,
887
+ "loss": 0.1424,
888
+ "step": 2460
889
+ },
890
+ {
891
+ "epoch": 2.07,
892
+ "grad_norm": 4.847611427307129,
893
+ "learning_rate": 6.416363636363637e-06,
894
+ "loss": 0.1588,
895
+ "step": 2480
896
+ },
897
+ {
898
+ "epoch": 2.08,
899
+ "grad_norm": 5.715017795562744,
900
+ "learning_rate": 6.380000000000001e-06,
901
+ "loss": 0.1896,
902
+ "step": 2500
903
+ },
904
+ {
905
+ "epoch": 2.1,
906
+ "grad_norm": 10.562045097351074,
907
+ "learning_rate": 6.3436363636363644e-06,
908
+ "loss": 0.2007,
909
+ "step": 2520
910
+ },
911
+ {
912
+ "epoch": 2.12,
913
+ "grad_norm": 3.2870736122131348,
914
+ "learning_rate": 6.307272727272727e-06,
915
+ "loss": 0.179,
916
+ "step": 2540
917
+ },
918
+ {
919
+ "epoch": 2.13,
920
+ "grad_norm": 11.021241188049316,
921
+ "learning_rate": 6.270909090909092e-06,
922
+ "loss": 0.1573,
923
+ "step": 2560
924
+ },
925
+ {
926
+ "epoch": 2.15,
927
+ "grad_norm": 7.566544055938721,
928
+ "learning_rate": 6.2345454545454555e-06,
929
+ "loss": 0.1832,
930
+ "step": 2580
931
+ },
932
+ {
933
+ "epoch": 2.17,
934
+ "grad_norm": 5.980641841888428,
935
+ "learning_rate": 6.198181818181819e-06,
936
+ "loss": 0.1498,
937
+ "step": 2600
938
+ },
939
+ {
940
+ "epoch": 2.18,
941
+ "grad_norm": 5.697175025939941,
942
+ "learning_rate": 6.161818181818182e-06,
943
+ "loss": 0.1378,
944
+ "step": 2620
945
+ },
946
+ {
947
+ "epoch": 2.2,
948
+ "grad_norm": 4.451406002044678,
949
+ "learning_rate": 6.125454545454546e-06,
950
+ "loss": 0.152,
951
+ "step": 2640
952
+ },
953
+ {
954
+ "epoch": 2.22,
955
+ "grad_norm": 5.854424476623535,
956
+ "learning_rate": 6.08909090909091e-06,
957
+ "loss": 0.1492,
958
+ "step": 2660
959
+ },
960
+ {
961
+ "epoch": 2.23,
962
+ "grad_norm": 6.480546951293945,
963
+ "learning_rate": 6.052727272727274e-06,
964
+ "loss": 0.1501,
965
+ "step": 2680
966
+ },
967
+ {
968
+ "epoch": 2.25,
969
+ "grad_norm": 3.3347442150115967,
970
+ "learning_rate": 6.016363636363637e-06,
971
+ "loss": 0.132,
972
+ "step": 2700
973
+ },
974
+ {
975
+ "epoch": 2.27,
976
+ "grad_norm": 5.9716081619262695,
977
+ "learning_rate": 5.98e-06,
978
+ "loss": 0.1483,
979
+ "step": 2720
980
+ },
981
+ {
982
+ "epoch": 2.28,
983
+ "grad_norm": 5.842201232910156,
984
+ "learning_rate": 5.943636363636364e-06,
985
+ "loss": 0.1688,
986
+ "step": 2740
987
+ },
988
+ {
989
+ "epoch": 2.3,
990
+ "grad_norm": 5.079952716827393,
991
+ "learning_rate": 5.9072727272727285e-06,
992
+ "loss": 0.167,
993
+ "step": 2760
994
+ },
995
+ {
996
+ "epoch": 2.32,
997
+ "grad_norm": 9.07481575012207,
998
+ "learning_rate": 5.870909090909091e-06,
999
+ "loss": 0.1581,
1000
+ "step": 2780
1001
+ },
1002
+ {
1003
+ "epoch": 2.33,
1004
+ "grad_norm": 9.30324935913086,
1005
+ "learning_rate": 5.834545454545455e-06,
1006
+ "loss": 0.1704,
1007
+ "step": 2800
1008
+ },
1009
+ {
1010
+ "epoch": 2.35,
1011
+ "grad_norm": 4.848841667175293,
1012
+ "learning_rate": 5.798181818181819e-06,
1013
+ "loss": 0.1894,
1014
+ "step": 2820
1015
+ },
1016
+ {
1017
+ "epoch": 2.37,
1018
+ "grad_norm": 7.52855110168457,
1019
+ "learning_rate": 5.7618181818181816e-06,
1020
+ "loss": 0.1702,
1021
+ "step": 2840
1022
+ },
1023
+ {
1024
+ "epoch": 2.38,
1025
+ "grad_norm": 6.624313831329346,
1026
+ "learning_rate": 5.725454545454546e-06,
1027
+ "loss": 0.1628,
1028
+ "step": 2860
1029
+ },
1030
+ {
1031
+ "epoch": 2.4,
1032
+ "grad_norm": 6.889793872833252,
1033
+ "learning_rate": 5.68909090909091e-06,
1034
+ "loss": 0.1705,
1035
+ "step": 2880
1036
+ },
1037
+ {
1038
+ "epoch": 2.42,
1039
+ "grad_norm": 5.635292053222656,
1040
+ "learning_rate": 5.6527272727272734e-06,
1041
+ "loss": 0.1531,
1042
+ "step": 2900
1043
+ },
1044
+ {
1045
+ "epoch": 2.43,
1046
+ "grad_norm": 4.1545729637146,
1047
+ "learning_rate": 5.616363636363636e-06,
1048
+ "loss": 0.1371,
1049
+ "step": 2920
1050
+ },
1051
+ {
1052
+ "epoch": 2.45,
1053
+ "grad_norm": 7.1985697746276855,
1054
+ "learning_rate": 5.580000000000001e-06,
1055
+ "loss": 0.1571,
1056
+ "step": 2940
1057
+ },
1058
+ {
1059
+ "epoch": 2.47,
1060
+ "grad_norm": 7.112518310546875,
1061
+ "learning_rate": 5.5436363636363645e-06,
1062
+ "loss": 0.1525,
1063
+ "step": 2960
1064
+ },
1065
+ {
1066
+ "epoch": 2.48,
1067
+ "grad_norm": 9.503728866577148,
1068
+ "learning_rate": 5.507272727272728e-06,
1069
+ "loss": 0.1426,
1070
+ "step": 2980
1071
+ },
1072
+ {
1073
+ "epoch": 2.5,
1074
+ "grad_norm": 5.497040271759033,
1075
+ "learning_rate": 5.470909090909091e-06,
1076
+ "loss": 0.1501,
1077
+ "step": 3000
1078
+ },
1079
+ {
1080
+ "epoch": 2.5,
1081
+ "eval_loss": 0.2098405510187149,
1082
+ "eval_runtime": 325.0771,
1083
+ "eval_samples_per_second": 0.308,
1084
+ "eval_steps_per_second": 0.308,
1085
+ "eval_wer": 55.5672268907563,
1086
+ "step": 3000
1087
+ },
1088
+ {
1089
+ "epoch": 2.52,
1090
+ "grad_norm": 6.7488603591918945,
1091
+ "learning_rate": 5.434545454545455e-06,
1092
+ "loss": 0.1647,
1093
+ "step": 3020
1094
+ },
1095
+ {
1096
+ "epoch": 2.53,
1097
+ "grad_norm": 5.16122579574585,
1098
+ "learning_rate": 5.398181818181819e-06,
1099
+ "loss": 0.1617,
1100
+ "step": 3040
1101
+ },
1102
+ {
1103
+ "epoch": 2.55,
1104
+ "grad_norm": 4.19830322265625,
1105
+ "learning_rate": 5.361818181818183e-06,
1106
+ "loss": 0.1484,
1107
+ "step": 3060
1108
+ },
1109
+ {
1110
+ "epoch": 2.57,
1111
+ "grad_norm": 9.339705467224121,
1112
+ "learning_rate": 5.325454545454546e-06,
1113
+ "loss": 0.153,
1114
+ "step": 3080
1115
+ },
1116
+ {
1117
+ "epoch": 2.58,
1118
+ "grad_norm": 5.106137752532959,
1119
+ "learning_rate": 5.289090909090909e-06,
1120
+ "loss": 0.141,
1121
+ "step": 3100
1122
+ },
1123
+ {
1124
+ "epoch": 2.6,
1125
+ "grad_norm": 8.092510223388672,
1126
+ "learning_rate": 5.252727272727273e-06,
1127
+ "loss": 0.1443,
1128
+ "step": 3120
1129
+ },
1130
+ {
1131
+ "epoch": 2.62,
1132
+ "grad_norm": 5.475297451019287,
1133
+ "learning_rate": 5.2163636363636376e-06,
1134
+ "loss": 0.1638,
1135
+ "step": 3140
1136
+ },
1137
+ {
1138
+ "epoch": 2.63,
1139
+ "grad_norm": 8.545796394348145,
1140
+ "learning_rate": 5.18e-06,
1141
+ "loss": 0.1382,
1142
+ "step": 3160
1143
+ },
1144
+ {
1145
+ "epoch": 2.65,
1146
+ "grad_norm": 5.189711093902588,
1147
+ "learning_rate": 5.143636363636364e-06,
1148
+ "loss": 0.1492,
1149
+ "step": 3180
1150
+ },
1151
+ {
1152
+ "epoch": 2.67,
1153
+ "grad_norm": 7.558894157409668,
1154
+ "learning_rate": 5.107272727272728e-06,
1155
+ "loss": 0.1559,
1156
+ "step": 3200
1157
+ },
1158
+ {
1159
+ "epoch": 2.68,
1160
+ "grad_norm": 5.985530853271484,
1161
+ "learning_rate": 5.0709090909090906e-06,
1162
+ "loss": 0.1605,
1163
+ "step": 3220
1164
+ },
1165
+ {
1166
+ "epoch": 2.7,
1167
+ "grad_norm": 6.780598163604736,
1168
+ "learning_rate": 5.034545454545455e-06,
1169
+ "loss": 0.148,
1170
+ "step": 3240
1171
+ },
1172
+ {
1173
+ "epoch": 2.72,
1174
+ "grad_norm": 6.225440979003906,
1175
+ "learning_rate": 4.998181818181819e-06,
1176
+ "loss": 0.1451,
1177
+ "step": 3260
1178
+ },
1179
+ {
1180
+ "epoch": 2.73,
1181
+ "grad_norm": 5.706241607666016,
1182
+ "learning_rate": 4.9618181818181824e-06,
1183
+ "loss": 0.1814,
1184
+ "step": 3280
1185
+ },
1186
+ {
1187
+ "epoch": 2.75,
1188
+ "grad_norm": 7.860342502593994,
1189
+ "learning_rate": 4.925454545454546e-06,
1190
+ "loss": 0.1407,
1191
+ "step": 3300
1192
+ },
1193
+ {
1194
+ "epoch": 2.77,
1195
+ "grad_norm": 4.965442180633545,
1196
+ "learning_rate": 4.88909090909091e-06,
1197
+ "loss": 0.1494,
1198
+ "step": 3320
1199
+ },
1200
+ {
1201
+ "epoch": 2.78,
1202
+ "grad_norm": 9.562829971313477,
1203
+ "learning_rate": 4.8527272727272735e-06,
1204
+ "loss": 0.1577,
1205
+ "step": 3340
1206
+ },
1207
+ {
1208
+ "epoch": 2.8,
1209
+ "grad_norm": 6.261291027069092,
1210
+ "learning_rate": 4.816363636363637e-06,
1211
+ "loss": 0.1561,
1212
+ "step": 3360
1213
+ },
1214
+ {
1215
+ "epoch": 2.82,
1216
+ "grad_norm": 5.747939109802246,
1217
+ "learning_rate": 4.78e-06,
1218
+ "loss": 0.1344,
1219
+ "step": 3380
1220
+ },
1221
+ {
1222
+ "epoch": 2.83,
1223
+ "grad_norm": 8.154770851135254,
1224
+ "learning_rate": 4.7436363636363645e-06,
1225
+ "loss": 0.1419,
1226
+ "step": 3400
1227
+ },
1228
+ {
1229
+ "epoch": 2.85,
1230
+ "grad_norm": 5.976426601409912,
1231
+ "learning_rate": 4.707272727272727e-06,
1232
+ "loss": 0.1314,
1233
+ "step": 3420
1234
+ },
1235
+ {
1236
+ "epoch": 2.87,
1237
+ "grad_norm": 4.834322452545166,
1238
+ "learning_rate": 4.670909090909092e-06,
1239
+ "loss": 0.1128,
1240
+ "step": 3440
1241
+ },
1242
+ {
1243
+ "epoch": 2.88,
1244
+ "grad_norm": 8.85062026977539,
1245
+ "learning_rate": 4.634545454545455e-06,
1246
+ "loss": 0.1718,
1247
+ "step": 3460
1248
+ },
1249
+ {
1250
+ "epoch": 2.9,
1251
+ "grad_norm": 3.992954969406128,
1252
+ "learning_rate": 4.598181818181818e-06,
1253
+ "loss": 0.1341,
1254
+ "step": 3480
1255
+ },
1256
+ {
1257
+ "epoch": 2.92,
1258
+ "grad_norm": 6.488000392913818,
1259
+ "learning_rate": 4.561818181818182e-06,
1260
+ "loss": 0.1401,
1261
+ "step": 3500
1262
+ },
1263
+ {
1264
+ "epoch": 2.93,
1265
+ "grad_norm": 4.876968860626221,
1266
+ "learning_rate": 4.525454545454546e-06,
1267
+ "loss": 0.1319,
1268
+ "step": 3520
1269
+ },
1270
+ {
1271
+ "epoch": 2.95,
1272
+ "grad_norm": 7.407408237457275,
1273
+ "learning_rate": 4.489090909090909e-06,
1274
+ "loss": 0.1377,
1275
+ "step": 3540
1276
+ },
1277
+ {
1278
+ "epoch": 2.97,
1279
+ "grad_norm": 5.017177104949951,
1280
+ "learning_rate": 4.452727272727273e-06,
1281
+ "loss": 0.1732,
1282
+ "step": 3560
1283
+ },
1284
+ {
1285
+ "epoch": 2.98,
1286
+ "grad_norm": 8.190740585327148,
1287
+ "learning_rate": 4.416363636363637e-06,
1288
+ "loss": 0.1225,
1289
+ "step": 3580
1290
+ },
1291
+ {
1292
+ "epoch": 3.0,
1293
+ "grad_norm": 7.947383403778076,
1294
+ "learning_rate": 4.38e-06,
1295
+ "loss": 0.1454,
1296
+ "step": 3600
1297
+ },
1298
+ {
1299
+ "epoch": 3.02,
1300
+ "grad_norm": 5.404466152191162,
1301
+ "learning_rate": 4.343636363636364e-06,
1302
+ "loss": 0.0746,
1303
+ "step": 3620
1304
+ },
1305
+ {
1306
+ "epoch": 3.03,
1307
+ "grad_norm": 5.470146179199219,
1308
+ "learning_rate": 4.307272727272728e-06,
1309
+ "loss": 0.0725,
1310
+ "step": 3640
1311
+ },
1312
+ {
1313
+ "epoch": 3.05,
1314
+ "grad_norm": 4.404048919677734,
1315
+ "learning_rate": 4.2709090909090914e-06,
1316
+ "loss": 0.088,
1317
+ "step": 3660
1318
+ },
1319
+ {
1320
+ "epoch": 3.07,
1321
+ "grad_norm": 3.169548273086548,
1322
+ "learning_rate": 4.234545454545455e-06,
1323
+ "loss": 0.0811,
1324
+ "step": 3680
1325
+ },
1326
+ {
1327
+ "epoch": 3.08,
1328
+ "grad_norm": 2.710052967071533,
1329
+ "learning_rate": 4.198181818181819e-06,
1330
+ "loss": 0.0646,
1331
+ "step": 3700
1332
+ },
1333
+ {
1334
+ "epoch": 3.1,
1335
+ "grad_norm": 5.970617294311523,
1336
+ "learning_rate": 4.1618181818181825e-06,
1337
+ "loss": 0.0789,
1338
+ "step": 3720
1339
+ },
1340
+ {
1341
+ "epoch": 3.12,
1342
+ "grad_norm": 4.657149791717529,
1343
+ "learning_rate": 4.125454545454546e-06,
1344
+ "loss": 0.0814,
1345
+ "step": 3740
1346
+ },
1347
+ {
1348
+ "epoch": 3.13,
1349
+ "grad_norm": 4.4107666015625,
1350
+ "learning_rate": 4.089090909090909e-06,
1351
+ "loss": 0.0831,
1352
+ "step": 3760
1353
+ },
1354
+ {
1355
+ "epoch": 3.15,
1356
+ "grad_norm": 6.185523509979248,
1357
+ "learning_rate": 4.0527272727272735e-06,
1358
+ "loss": 0.0672,
1359
+ "step": 3780
1360
+ },
1361
+ {
1362
+ "epoch": 3.17,
1363
+ "grad_norm": 3.5439445972442627,
1364
+ "learning_rate": 4.016363636363636e-06,
1365
+ "loss": 0.0839,
1366
+ "step": 3800
1367
+ },
1368
+ {
1369
+ "epoch": 3.18,
1370
+ "grad_norm": 5.79164981842041,
1371
+ "learning_rate": 3.980000000000001e-06,
1372
+ "loss": 0.0692,
1373
+ "step": 3820
1374
+ },
1375
+ {
1376
+ "epoch": 3.2,
1377
+ "grad_norm": 6.883912086486816,
1378
+ "learning_rate": 3.943636363636364e-06,
1379
+ "loss": 0.0798,
1380
+ "step": 3840
1381
+ },
1382
+ {
1383
+ "epoch": 3.22,
1384
+ "grad_norm": 4.028432846069336,
1385
+ "learning_rate": 3.907272727272727e-06,
1386
+ "loss": 0.0808,
1387
+ "step": 3860
1388
+ },
1389
+ {
1390
+ "epoch": 3.23,
1391
+ "grad_norm": 4.477746486663818,
1392
+ "learning_rate": 3.870909090909091e-06,
1393
+ "loss": 0.0781,
1394
+ "step": 3880
1395
+ },
1396
+ {
1397
+ "epoch": 3.25,
1398
+ "grad_norm": 2.0900027751922607,
1399
+ "learning_rate": 3.834545454545455e-06,
1400
+ "loss": 0.073,
1401
+ "step": 3900
1402
+ },
1403
+ {
1404
+ "epoch": 3.27,
1405
+ "grad_norm": 2.609961748123169,
1406
+ "learning_rate": 3.798181818181819e-06,
1407
+ "loss": 0.0961,
1408
+ "step": 3920
1409
+ },
1410
+ {
1411
+ "epoch": 3.28,
1412
+ "grad_norm": 6.023408889770508,
1413
+ "learning_rate": 3.761818181818182e-06,
1414
+ "loss": 0.0806,
1415
+ "step": 3940
1416
+ },
1417
+ {
1418
+ "epoch": 3.3,
1419
+ "grad_norm": 5.091029167175293,
1420
+ "learning_rate": 3.725454545454546e-06,
1421
+ "loss": 0.0681,
1422
+ "step": 3960
1423
+ },
1424
+ {
1425
+ "epoch": 3.32,
1426
+ "grad_norm": 5.474444389343262,
1427
+ "learning_rate": 3.6890909090909094e-06,
1428
+ "loss": 0.0817,
1429
+ "step": 3980
1430
+ },
1431
+ {
1432
+ "epoch": 3.33,
1433
+ "grad_norm": 5.174444198608398,
1434
+ "learning_rate": 3.6527272727272727e-06,
1435
+ "loss": 0.0827,
1436
+ "step": 4000
1437
+ },
1438
+ {
1439
+ "epoch": 3.33,
1440
+ "eval_loss": 0.20179511606693268,
1441
+ "eval_runtime": 319.617,
1442
+ "eval_samples_per_second": 0.313,
1443
+ "eval_steps_per_second": 0.313,
1444
+ "eval_wer": 51.78571428571429,
1445
+ "step": 4000
1446
+ }
1447
+ ],
1448
+ "logging_steps": 20,
1449
+ "max_steps": 6000,
1450
+ "num_input_tokens_seen": 0,
1451
+ "num_train_epochs": 5,
1452
+ "save_steps": 1000,
1453
+ "total_flos": 3.46244763303936e+18,
1454
+ "train_batch_size": 3,
1455
+ "trial_name": null,
1456
+ "trial_params": null
1457
+ }
checkpoint-4000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:152d1e9770ea3e9899a6c5c40e00cad1f52c4e2ae9aea55ed85a6bf2ed0a638f
3
+ size 5048
checkpoint-5000/config.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
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": 768,
17
+ "decoder_attention_heads": 12,
18
+ "decoder_ffn_dim": 3072,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 12,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 12,
24
+ "encoder_ffn_dim": 3072,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 12,
43
+ "num_mel_bins": 80,
44
+ "pad_token_id": 50257,
45
+ "scale_embedding": false,
46
+ "suppress_tokens": [],
47
+ "torch_dtype": "float32",
48
+ "transformers_version": "4.39.0.dev0",
49
+ "use_cache": true,
50
+ "use_weighted_layer_sum": false,
51
+ "vocab_size": 51865
52
+ }
checkpoint-5000/generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 5,
5
+ 3
6
+ ],
7
+ [
8
+ 5,
9
+ 9
10
+ ],
11
+ [
12
+ 8,
13
+ 0
14
+ ],
15
+ [
16
+ 8,
17
+ 4
18
+ ],
19
+ [
20
+ 8,
21
+ 7
22
+ ],
23
+ [
24
+ 8,
25
+ 8
26
+ ],
27
+ [
28
+ 9,
29
+ 0
30
+ ],
31
+ [
32
+ 9,
33
+ 7
34
+ ],
35
+ [
36
+ 9,
37
+ 9
38
+ ],
39
+ [
40
+ 10,
41
+ 5
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50359
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|zh|>": 50260
162
+ },
163
+ "max_initial_timestamp_index": 50,
164
+ "max_length": 448,
165
+ "no_timestamps_token_id": 50363,
166
+ "pad_token_id": 50257,
167
+ "prev_sot_token_id": 50361,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50358,
254
+ 50359,
255
+ 50360,
256
+ 50361,
257
+ 50362
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50359,
261
+ "translate": 50358
262
+ },
263
+ "transformers_version": "4.39.0.dev0"
264
+ }
checkpoint-5000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4515162d961c3ad7cb5d3d1ed572a76593f8dec49fb548c50ba408d3968c054
3
+ size 966995080
checkpoint-5000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b53ed3835b0a3837b66cb40f69ef449cc47d38c8b5e881f6d0eb7f4800e0cd69
3
+ size 1925064044
checkpoint-5000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 80,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-5000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e5a1067820dd0dffb3568fe7f2c7edd673482a94825fd773fbebd20b471e1b5
3
+ size 14244
checkpoint-5000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cca49c083fbd235d0cebc0d04150eb712cd8c64c240865c95f9b8be683cda3f
3
+ size 1064
checkpoint-5000/trainer_state.json ADDED
@@ -0,0 +1,1816 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 47.37394957983193,
3
+ "best_model_checkpoint": "./whisper-small-bn/checkpoint-5000",
4
+ "epoch": 4.166666666666667,
5
+ "eval_steps": 1000,
6
+ "global_step": 5000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.07,
13
+ "grad_norm": 27.802248001098633,
14
+ "learning_rate": 3.2e-07,
15
+ "loss": 2.2811,
16
+ "step": 20
17
+ },
18
+ {
19
+ "epoch": 0.13,
20
+ "grad_norm": 18.253421783447266,
21
+ "learning_rate": 7.2e-07,
22
+ "loss": 2.1964,
23
+ "step": 40
24
+ },
25
+ {
26
+ "epoch": 0.2,
27
+ "grad_norm": 13.585515975952148,
28
+ "learning_rate": 1.12e-06,
29
+ "loss": 1.8635,
30
+ "step": 60
31
+ },
32
+ {
33
+ "epoch": 0.27,
34
+ "grad_norm": 22.400362014770508,
35
+ "learning_rate": 1.52e-06,
36
+ "loss": 1.703,
37
+ "step": 80
38
+ },
39
+ {
40
+ "epoch": 0.33,
41
+ "grad_norm": 13.77535343170166,
42
+ "learning_rate": 1.9200000000000003e-06,
43
+ "loss": 1.5336,
44
+ "step": 100
45
+ },
46
+ {
47
+ "epoch": 0.4,
48
+ "grad_norm": 14.989215850830078,
49
+ "learning_rate": 2.3200000000000002e-06,
50
+ "loss": 1.4044,
51
+ "step": 120
52
+ },
53
+ {
54
+ "epoch": 0.47,
55
+ "grad_norm": 22.01858139038086,
56
+ "learning_rate": 2.7200000000000002e-06,
57
+ "loss": 1.2402,
58
+ "step": 140
59
+ },
60
+ {
61
+ "epoch": 0.53,
62
+ "grad_norm": 16.482311248779297,
63
+ "learning_rate": 3.12e-06,
64
+ "loss": 1.0922,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.6,
69
+ "grad_norm": 17.529874801635742,
70
+ "learning_rate": 3.52e-06,
71
+ "loss": 0.9583,
72
+ "step": 180
73
+ },
74
+ {
75
+ "epoch": 0.67,
76
+ "grad_norm": 18.252761840820312,
77
+ "learning_rate": 3.920000000000001e-06,
78
+ "loss": 0.8783,
79
+ "step": 200
80
+ },
81
+ {
82
+ "epoch": 0.73,
83
+ "grad_norm": 16.48067283630371,
84
+ "learning_rate": 4.32e-06,
85
+ "loss": 0.7643,
86
+ "step": 220
87
+ },
88
+ {
89
+ "epoch": 0.8,
90
+ "grad_norm": 15.062904357910156,
91
+ "learning_rate": 4.7200000000000005e-06,
92
+ "loss": 0.6991,
93
+ "step": 240
94
+ },
95
+ {
96
+ "epoch": 0.87,
97
+ "grad_norm": 14.231403350830078,
98
+ "learning_rate": 5.12e-06,
99
+ "loss": 0.6594,
100
+ "step": 260
101
+ },
102
+ {
103
+ "epoch": 0.93,
104
+ "grad_norm": 18.425230026245117,
105
+ "learning_rate": 5.5200000000000005e-06,
106
+ "loss": 0.5724,
107
+ "step": 280
108
+ },
109
+ {
110
+ "epoch": 1.0,
111
+ "grad_norm": 11.786844253540039,
112
+ "learning_rate": 5.92e-06,
113
+ "loss": 0.5687,
114
+ "step": 300
115
+ },
116
+ {
117
+ "epoch": 1.07,
118
+ "grad_norm": 15.346015930175781,
119
+ "learning_rate": 6.3200000000000005e-06,
120
+ "loss": 0.4696,
121
+ "step": 320
122
+ },
123
+ {
124
+ "epoch": 1.13,
125
+ "grad_norm": 7.936849594116211,
126
+ "learning_rate": 6.720000000000001e-06,
127
+ "loss": 0.4569,
128
+ "step": 340
129
+ },
130
+ {
131
+ "epoch": 1.2,
132
+ "grad_norm": 13.894779205322266,
133
+ "learning_rate": 7.1200000000000004e-06,
134
+ "loss": 0.4089,
135
+ "step": 360
136
+ },
137
+ {
138
+ "epoch": 1.27,
139
+ "grad_norm": 11.665081977844238,
140
+ "learning_rate": 7.520000000000001e-06,
141
+ "loss": 0.4168,
142
+ "step": 380
143
+ },
144
+ {
145
+ "epoch": 1.33,
146
+ "grad_norm": 16.7696475982666,
147
+ "learning_rate": 7.92e-06,
148
+ "loss": 0.4433,
149
+ "step": 400
150
+ },
151
+ {
152
+ "epoch": 1.4,
153
+ "grad_norm": 15.60254192352295,
154
+ "learning_rate": 8.32e-06,
155
+ "loss": 0.4013,
156
+ "step": 420
157
+ },
158
+ {
159
+ "epoch": 1.47,
160
+ "grad_norm": 13.72472858428955,
161
+ "learning_rate": 8.720000000000001e-06,
162
+ "loss": 0.3903,
163
+ "step": 440
164
+ },
165
+ {
166
+ "epoch": 1.53,
167
+ "grad_norm": 10.506444931030273,
168
+ "learning_rate": 9.12e-06,
169
+ "loss": 0.3581,
170
+ "step": 460
171
+ },
172
+ {
173
+ "epoch": 1.6,
174
+ "grad_norm": 13.643900871276855,
175
+ "learning_rate": 9.52e-06,
176
+ "loss": 0.3605,
177
+ "step": 480
178
+ },
179
+ {
180
+ "epoch": 1.67,
181
+ "grad_norm": 11.669689178466797,
182
+ "learning_rate": 9.920000000000002e-06,
183
+ "loss": 0.3548,
184
+ "step": 500
185
+ },
186
+ {
187
+ "epoch": 1.73,
188
+ "grad_norm": 9.907474517822266,
189
+ "learning_rate": 9.936000000000001e-06,
190
+ "loss": 0.3535,
191
+ "step": 520
192
+ },
193
+ {
194
+ "epoch": 1.8,
195
+ "grad_norm": 11.366904258728027,
196
+ "learning_rate": 9.856000000000002e-06,
197
+ "loss": 0.2943,
198
+ "step": 540
199
+ },
200
+ {
201
+ "epoch": 1.87,
202
+ "grad_norm": 7.5114240646362305,
203
+ "learning_rate": 9.776000000000001e-06,
204
+ "loss": 0.3057,
205
+ "step": 560
206
+ },
207
+ {
208
+ "epoch": 1.93,
209
+ "grad_norm": 11.717443466186523,
210
+ "learning_rate": 9.696000000000002e-06,
211
+ "loss": 0.2919,
212
+ "step": 580
213
+ },
214
+ {
215
+ "epoch": 2.0,
216
+ "grad_norm": 9.275129318237305,
217
+ "learning_rate": 9.616e-06,
218
+ "loss": 0.3115,
219
+ "step": 600
220
+ },
221
+ {
222
+ "epoch": 2.07,
223
+ "grad_norm": 9.388382911682129,
224
+ "learning_rate": 9.536000000000002e-06,
225
+ "loss": 0.1926,
226
+ "step": 620
227
+ },
228
+ {
229
+ "epoch": 2.13,
230
+ "grad_norm": 8.574433326721191,
231
+ "learning_rate": 9.456e-06,
232
+ "loss": 0.2193,
233
+ "step": 640
234
+ },
235
+ {
236
+ "epoch": 2.2,
237
+ "grad_norm": 5.512338638305664,
238
+ "learning_rate": 9.376000000000001e-06,
239
+ "loss": 0.18,
240
+ "step": 660
241
+ },
242
+ {
243
+ "epoch": 2.27,
244
+ "grad_norm": 9.644317626953125,
245
+ "learning_rate": 9.296e-06,
246
+ "loss": 0.1916,
247
+ "step": 680
248
+ },
249
+ {
250
+ "epoch": 2.33,
251
+ "grad_norm": 9.321474075317383,
252
+ "learning_rate": 9.216000000000001e-06,
253
+ "loss": 0.2055,
254
+ "step": 700
255
+ },
256
+ {
257
+ "epoch": 2.4,
258
+ "grad_norm": 4.868476390838623,
259
+ "learning_rate": 9.136e-06,
260
+ "loss": 0.1843,
261
+ "step": 720
262
+ },
263
+ {
264
+ "epoch": 2.47,
265
+ "grad_norm": 6.592445373535156,
266
+ "learning_rate": 9.056000000000001e-06,
267
+ "loss": 0.1713,
268
+ "step": 740
269
+ },
270
+ {
271
+ "epoch": 2.53,
272
+ "grad_norm": 8.138116836547852,
273
+ "learning_rate": 8.976e-06,
274
+ "loss": 0.1792,
275
+ "step": 760
276
+ },
277
+ {
278
+ "epoch": 2.6,
279
+ "grad_norm": 9.782510757446289,
280
+ "learning_rate": 8.896000000000001e-06,
281
+ "loss": 0.1796,
282
+ "step": 780
283
+ },
284
+ {
285
+ "epoch": 2.67,
286
+ "grad_norm": 5.867671012878418,
287
+ "learning_rate": 8.816000000000002e-06,
288
+ "loss": 0.1652,
289
+ "step": 800
290
+ },
291
+ {
292
+ "epoch": 2.73,
293
+ "grad_norm": 6.637588977813721,
294
+ "learning_rate": 8.736e-06,
295
+ "loss": 0.1735,
296
+ "step": 820
297
+ },
298
+ {
299
+ "epoch": 2.8,
300
+ "grad_norm": 11.652266502380371,
301
+ "learning_rate": 8.656000000000001e-06,
302
+ "loss": 0.1734,
303
+ "step": 840
304
+ },
305
+ {
306
+ "epoch": 2.87,
307
+ "grad_norm": 6.051203727722168,
308
+ "learning_rate": 8.576e-06,
309
+ "loss": 0.1557,
310
+ "step": 860
311
+ },
312
+ {
313
+ "epoch": 2.93,
314
+ "grad_norm": 11.018319129943848,
315
+ "learning_rate": 8.496000000000001e-06,
316
+ "loss": 0.1668,
317
+ "step": 880
318
+ },
319
+ {
320
+ "epoch": 3.0,
321
+ "grad_norm": 7.878730297088623,
322
+ "learning_rate": 8.416e-06,
323
+ "loss": 0.1706,
324
+ "step": 900
325
+ },
326
+ {
327
+ "epoch": 3.07,
328
+ "grad_norm": 6.223515510559082,
329
+ "learning_rate": 8.336000000000001e-06,
330
+ "loss": 0.0917,
331
+ "step": 920
332
+ },
333
+ {
334
+ "epoch": 3.13,
335
+ "grad_norm": 7.449841022491455,
336
+ "learning_rate": 8.256e-06,
337
+ "loss": 0.0853,
338
+ "step": 940
339
+ },
340
+ {
341
+ "epoch": 3.2,
342
+ "grad_norm": 3.824934244155884,
343
+ "learning_rate": 8.176000000000001e-06,
344
+ "loss": 0.1033,
345
+ "step": 960
346
+ },
347
+ {
348
+ "epoch": 3.27,
349
+ "grad_norm": 6.369926929473877,
350
+ "learning_rate": 8.096e-06,
351
+ "loss": 0.0997,
352
+ "step": 980
353
+ },
354
+ {
355
+ "epoch": 3.33,
356
+ "grad_norm": 6.938141822814941,
357
+ "learning_rate": 8.016e-06,
358
+ "loss": 0.095,
359
+ "step": 1000
360
+ },
361
+ {
362
+ "epoch": 3.33,
363
+ "eval_loss": 0.27550819516181946,
364
+ "eval_runtime": 345.1723,
365
+ "eval_samples_per_second": 0.29,
366
+ "eval_steps_per_second": 0.29,
367
+ "eval_wer": 60.18099547511312,
368
+ "step": 1000
369
+ },
370
+ {
371
+ "epoch": 2.18,
372
+ "grad_norm": 7.869150638580322,
373
+ "learning_rate": 7.936e-06,
374
+ "loss": 0.1443,
375
+ "step": 1020
376
+ },
377
+ {
378
+ "epoch": 2.23,
379
+ "grad_norm": 3.6447396278381348,
380
+ "learning_rate": 7.860000000000001e-06,
381
+ "loss": 0.1974,
382
+ "step": 1040
383
+ },
384
+ {
385
+ "epoch": 2.27,
386
+ "grad_norm": 9.588157653808594,
387
+ "learning_rate": 7.78e-06,
388
+ "loss": 0.1427,
389
+ "step": 1060
390
+ },
391
+ {
392
+ "epoch": 2.31,
393
+ "grad_norm": 10.796669006347656,
394
+ "learning_rate": 7.7e-06,
395
+ "loss": 0.187,
396
+ "step": 1080
397
+ },
398
+ {
399
+ "epoch": 2.36,
400
+ "grad_norm": 4.941339015960693,
401
+ "learning_rate": 7.620000000000001e-06,
402
+ "loss": 0.1568,
403
+ "step": 1100
404
+ },
405
+ {
406
+ "epoch": 2.4,
407
+ "grad_norm": 11.114888191223145,
408
+ "learning_rate": 7.540000000000001e-06,
409
+ "loss": 0.1697,
410
+ "step": 1120
411
+ },
412
+ {
413
+ "epoch": 2.44,
414
+ "grad_norm": 10.200749397277832,
415
+ "learning_rate": 7.4600000000000006e-06,
416
+ "loss": 0.1675,
417
+ "step": 1140
418
+ },
419
+ {
420
+ "epoch": 2.48,
421
+ "grad_norm": 5.67927360534668,
422
+ "learning_rate": 7.3800000000000005e-06,
423
+ "loss": 0.1557,
424
+ "step": 1160
425
+ },
426
+ {
427
+ "epoch": 2.53,
428
+ "grad_norm": 7.413832664489746,
429
+ "learning_rate": 7.304000000000001e-06,
430
+ "loss": 0.1629,
431
+ "step": 1180
432
+ },
433
+ {
434
+ "epoch": 2.57,
435
+ "grad_norm": 6.933262348175049,
436
+ "learning_rate": 7.224000000000001e-06,
437
+ "loss": 0.17,
438
+ "step": 1200
439
+ },
440
+ {
441
+ "epoch": 2.61,
442
+ "grad_norm": 4.847285747528076,
443
+ "learning_rate": 7.1440000000000005e-06,
444
+ "loss": 0.1389,
445
+ "step": 1220
446
+ },
447
+ {
448
+ "epoch": 2.66,
449
+ "grad_norm": 19.55495834350586,
450
+ "learning_rate": 7.0640000000000005e-06,
451
+ "loss": 0.1338,
452
+ "step": 1240
453
+ },
454
+ {
455
+ "epoch": 2.7,
456
+ "grad_norm": 7.148222923278809,
457
+ "learning_rate": 6.984e-06,
458
+ "loss": 0.1296,
459
+ "step": 1260
460
+ },
461
+ {
462
+ "epoch": 2.74,
463
+ "grad_norm": 6.323164939880371,
464
+ "learning_rate": 6.904e-06,
465
+ "loss": 0.163,
466
+ "step": 1280
467
+ },
468
+ {
469
+ "epoch": 2.78,
470
+ "grad_norm": 9.096046447753906,
471
+ "learning_rate": 6.824e-06,
472
+ "loss": 0.1942,
473
+ "step": 1300
474
+ },
475
+ {
476
+ "epoch": 2.83,
477
+ "grad_norm": 8.462915420532227,
478
+ "learning_rate": 6.744e-06,
479
+ "loss": 0.1576,
480
+ "step": 1320
481
+ },
482
+ {
483
+ "epoch": 2.87,
484
+ "grad_norm": 7.848359107971191,
485
+ "learning_rate": 6.664e-06,
486
+ "loss": 0.1219,
487
+ "step": 1340
488
+ },
489
+ {
490
+ "epoch": 2.91,
491
+ "grad_norm": 6.30383825302124,
492
+ "learning_rate": 6.584e-06,
493
+ "loss": 0.1404,
494
+ "step": 1360
495
+ },
496
+ {
497
+ "epoch": 2.96,
498
+ "grad_norm": 5.657530784606934,
499
+ "learning_rate": 6.504e-06,
500
+ "loss": 0.1449,
501
+ "step": 1380
502
+ },
503
+ {
504
+ "epoch": 3.0,
505
+ "grad_norm": 7.4172234535217285,
506
+ "learning_rate": 6.424e-06,
507
+ "loss": 0.1234,
508
+ "step": 1400
509
+ },
510
+ {
511
+ "epoch": 3.04,
512
+ "grad_norm": 6.550841331481934,
513
+ "learning_rate": 6.344e-06,
514
+ "loss": 0.0885,
515
+ "step": 1420
516
+ },
517
+ {
518
+ "epoch": 3.08,
519
+ "grad_norm": 6.212351322174072,
520
+ "learning_rate": 6.264e-06,
521
+ "loss": 0.0725,
522
+ "step": 1440
523
+ },
524
+ {
525
+ "epoch": 3.13,
526
+ "grad_norm": 5.2562575340271,
527
+ "learning_rate": 6.184e-06,
528
+ "loss": 0.0886,
529
+ "step": 1460
530
+ },
531
+ {
532
+ "epoch": 3.17,
533
+ "grad_norm": 7.921740531921387,
534
+ "learning_rate": 6.104000000000001e-06,
535
+ "loss": 0.0805,
536
+ "step": 1480
537
+ },
538
+ {
539
+ "epoch": 3.21,
540
+ "grad_norm": 3.5878069400787354,
541
+ "learning_rate": 6.024000000000001e-06,
542
+ "loss": 0.0628,
543
+ "step": 1500
544
+ },
545
+ {
546
+ "epoch": 3.25,
547
+ "grad_norm": 8.585208892822266,
548
+ "learning_rate": 5.944000000000001e-06,
549
+ "loss": 0.0812,
550
+ "step": 1520
551
+ },
552
+ {
553
+ "epoch": 3.3,
554
+ "grad_norm": 4.111368656158447,
555
+ "learning_rate": 5.868e-06,
556
+ "loss": 0.0694,
557
+ "step": 1540
558
+ },
559
+ {
560
+ "epoch": 3.34,
561
+ "grad_norm": 5.304944038391113,
562
+ "learning_rate": 5.788e-06,
563
+ "loss": 0.0739,
564
+ "step": 1560
565
+ },
566
+ {
567
+ "epoch": 3.38,
568
+ "grad_norm": 6.168578147888184,
569
+ "learning_rate": 5.708e-06,
570
+ "loss": 0.0821,
571
+ "step": 1580
572
+ },
573
+ {
574
+ "epoch": 3.43,
575
+ "grad_norm": 12.2472505569458,
576
+ "learning_rate": 5.628e-06,
577
+ "loss": 0.0885,
578
+ "step": 1600
579
+ },
580
+ {
581
+ "epoch": 3.47,
582
+ "grad_norm": 9.349952697753906,
583
+ "learning_rate": 5.548e-06,
584
+ "loss": 0.0607,
585
+ "step": 1620
586
+ },
587
+ {
588
+ "epoch": 3.51,
589
+ "grad_norm": 5.98253870010376,
590
+ "learning_rate": 5.468e-06,
591
+ "loss": 0.0619,
592
+ "step": 1640
593
+ },
594
+ {
595
+ "epoch": 3.55,
596
+ "grad_norm": 7.106723785400391,
597
+ "learning_rate": 5.388e-06,
598
+ "loss": 0.0992,
599
+ "step": 1660
600
+ },
601
+ {
602
+ "epoch": 3.6,
603
+ "grad_norm": 3.693020820617676,
604
+ "learning_rate": 5.308000000000001e-06,
605
+ "loss": 0.0855,
606
+ "step": 1680
607
+ },
608
+ {
609
+ "epoch": 3.64,
610
+ "grad_norm": 7.3649678230285645,
611
+ "learning_rate": 5.228000000000001e-06,
612
+ "loss": 0.081,
613
+ "step": 1700
614
+ },
615
+ {
616
+ "epoch": 3.68,
617
+ "grad_norm": 5.4454240798950195,
618
+ "learning_rate": 5.1480000000000005e-06,
619
+ "loss": 0.0872,
620
+ "step": 1720
621
+ },
622
+ {
623
+ "epoch": 3.73,
624
+ "grad_norm": 8.854460716247559,
625
+ "learning_rate": 5.0680000000000004e-06,
626
+ "loss": 0.0869,
627
+ "step": 1740
628
+ },
629
+ {
630
+ "epoch": 3.77,
631
+ "grad_norm": 3.8615269660949707,
632
+ "learning_rate": 4.988e-06,
633
+ "loss": 0.0821,
634
+ "step": 1760
635
+ },
636
+ {
637
+ "epoch": 3.81,
638
+ "grad_norm": 3.7190287113189697,
639
+ "learning_rate": 4.908e-06,
640
+ "loss": 0.0636,
641
+ "step": 1780
642
+ },
643
+ {
644
+ "epoch": 3.85,
645
+ "grad_norm": 7.683879375457764,
646
+ "learning_rate": 4.828e-06,
647
+ "loss": 0.0892,
648
+ "step": 1800
649
+ },
650
+ {
651
+ "epoch": 3.9,
652
+ "grad_norm": 3.645141839981079,
653
+ "learning_rate": 4.748e-06,
654
+ "loss": 0.0803,
655
+ "step": 1820
656
+ },
657
+ {
658
+ "epoch": 3.94,
659
+ "grad_norm": 3.8167316913604736,
660
+ "learning_rate": 4.668e-06,
661
+ "loss": 0.1012,
662
+ "step": 1840
663
+ },
664
+ {
665
+ "epoch": 3.98,
666
+ "grad_norm": 7.247605323791504,
667
+ "learning_rate": 4.588e-06,
668
+ "loss": 0.0728,
669
+ "step": 1860
670
+ },
671
+ {
672
+ "epoch": 4.03,
673
+ "grad_norm": 3.5495615005493164,
674
+ "learning_rate": 4.508e-06,
675
+ "loss": 0.0439,
676
+ "step": 1880
677
+ },
678
+ {
679
+ "epoch": 4.07,
680
+ "grad_norm": 4.105050086975098,
681
+ "learning_rate": 4.428000000000001e-06,
682
+ "loss": 0.0357,
683
+ "step": 1900
684
+ },
685
+ {
686
+ "epoch": 4.11,
687
+ "grad_norm": 9.399913787841797,
688
+ "learning_rate": 4.3480000000000006e-06,
689
+ "loss": 0.0363,
690
+ "step": 1920
691
+ },
692
+ {
693
+ "epoch": 4.15,
694
+ "grad_norm": 3.0376510620117188,
695
+ "learning_rate": 4.2680000000000005e-06,
696
+ "loss": 0.0221,
697
+ "step": 1940
698
+ },
699
+ {
700
+ "epoch": 4.2,
701
+ "grad_norm": 0.6908012628555298,
702
+ "learning_rate": 4.188e-06,
703
+ "loss": 0.0323,
704
+ "step": 1960
705
+ },
706
+ {
707
+ "epoch": 4.24,
708
+ "grad_norm": 6.255786895751953,
709
+ "learning_rate": 4.108e-06,
710
+ "loss": 0.0327,
711
+ "step": 1980
712
+ },
713
+ {
714
+ "epoch": 4.28,
715
+ "grad_norm": 7.1336669921875,
716
+ "learning_rate": 4.028e-06,
717
+ "loss": 0.0416,
718
+ "step": 2000
719
+ },
720
+ {
721
+ "epoch": 4.28,
722
+ "eval_loss": 0.27948442101478577,
723
+ "eval_runtime": 349.9899,
724
+ "eval_samples_per_second": 0.286,
725
+ "eval_steps_per_second": 0.286,
726
+ "eval_wer": 59.0146750524109,
727
+ "step": 2000
728
+ },
729
+ {
730
+ "epoch": 1.68,
731
+ "grad_norm": 13.214738845825195,
732
+ "learning_rate": 7.250909090909092e-06,
733
+ "loss": 0.2839,
734
+ "step": 2020
735
+ },
736
+ {
737
+ "epoch": 1.7,
738
+ "grad_norm": 8.846616744995117,
739
+ "learning_rate": 7.214545454545455e-06,
740
+ "loss": 0.2221,
741
+ "step": 2040
742
+ },
743
+ {
744
+ "epoch": 1.72,
745
+ "grad_norm": 11.778766632080078,
746
+ "learning_rate": 7.178181818181818e-06,
747
+ "loss": 0.2394,
748
+ "step": 2060
749
+ },
750
+ {
751
+ "epoch": 1.73,
752
+ "grad_norm": 4.484768867492676,
753
+ "learning_rate": 7.141818181818182e-06,
754
+ "loss": 0.2299,
755
+ "step": 2080
756
+ },
757
+ {
758
+ "epoch": 1.75,
759
+ "grad_norm": 4.385435104370117,
760
+ "learning_rate": 7.105454545454546e-06,
761
+ "loss": 0.2221,
762
+ "step": 2100
763
+ },
764
+ {
765
+ "epoch": 1.77,
766
+ "grad_norm": 8.751326560974121,
767
+ "learning_rate": 7.06909090909091e-06,
768
+ "loss": 0.1966,
769
+ "step": 2120
770
+ },
771
+ {
772
+ "epoch": 1.78,
773
+ "grad_norm": 8.837333679199219,
774
+ "learning_rate": 7.032727272727273e-06,
775
+ "loss": 0.1893,
776
+ "step": 2140
777
+ },
778
+ {
779
+ "epoch": 1.8,
780
+ "grad_norm": 9.331002235412598,
781
+ "learning_rate": 6.998181818181818e-06,
782
+ "loss": 0.2365,
783
+ "step": 2160
784
+ },
785
+ {
786
+ "epoch": 1.82,
787
+ "grad_norm": 6.599325180053711,
788
+ "learning_rate": 6.961818181818183e-06,
789
+ "loss": 0.2152,
790
+ "step": 2180
791
+ },
792
+ {
793
+ "epoch": 1.83,
794
+ "grad_norm": 11.716184616088867,
795
+ "learning_rate": 6.9254545454545464e-06,
796
+ "loss": 0.203,
797
+ "step": 2200
798
+ },
799
+ {
800
+ "epoch": 1.85,
801
+ "grad_norm": 6.506683349609375,
802
+ "learning_rate": 6.88909090909091e-06,
803
+ "loss": 0.2162,
804
+ "step": 2220
805
+ },
806
+ {
807
+ "epoch": 1.87,
808
+ "grad_norm": 4.64941930770874,
809
+ "learning_rate": 6.852727272727273e-06,
810
+ "loss": 0.214,
811
+ "step": 2240
812
+ },
813
+ {
814
+ "epoch": 1.88,
815
+ "grad_norm": 8.344452857971191,
816
+ "learning_rate": 6.816363636363637e-06,
817
+ "loss": 0.2106,
818
+ "step": 2260
819
+ },
820
+ {
821
+ "epoch": 1.9,
822
+ "grad_norm": 5.504006385803223,
823
+ "learning_rate": 6.780000000000001e-06,
824
+ "loss": 0.196,
825
+ "step": 2280
826
+ },
827
+ {
828
+ "epoch": 1.92,
829
+ "grad_norm": 9.84432601928711,
830
+ "learning_rate": 6.743636363636365e-06,
831
+ "loss": 0.2207,
832
+ "step": 2300
833
+ },
834
+ {
835
+ "epoch": 1.93,
836
+ "grad_norm": 7.112894535064697,
837
+ "learning_rate": 6.707272727272728e-06,
838
+ "loss": 0.2072,
839
+ "step": 2320
840
+ },
841
+ {
842
+ "epoch": 1.95,
843
+ "grad_norm": 7.10872745513916,
844
+ "learning_rate": 6.670909090909091e-06,
845
+ "loss": 0.1925,
846
+ "step": 2340
847
+ },
848
+ {
849
+ "epoch": 1.97,
850
+ "grad_norm": 8.612652778625488,
851
+ "learning_rate": 6.634545454545455e-06,
852
+ "loss": 0.2023,
853
+ "step": 2360
854
+ },
855
+ {
856
+ "epoch": 1.98,
857
+ "grad_norm": 6.083488941192627,
858
+ "learning_rate": 6.5981818181818195e-06,
859
+ "loss": 0.2111,
860
+ "step": 2380
861
+ },
862
+ {
863
+ "epoch": 2.0,
864
+ "grad_norm": 7.59145975112915,
865
+ "learning_rate": 6.561818181818182e-06,
866
+ "loss": 0.1934,
867
+ "step": 2400
868
+ },
869
+ {
870
+ "epoch": 2.02,
871
+ "grad_norm": 6.529219627380371,
872
+ "learning_rate": 6.525454545454546e-06,
873
+ "loss": 0.1756,
874
+ "step": 2420
875
+ },
876
+ {
877
+ "epoch": 2.03,
878
+ "grad_norm": 13.685791015625,
879
+ "learning_rate": 6.48909090909091e-06,
880
+ "loss": 0.1872,
881
+ "step": 2440
882
+ },
883
+ {
884
+ "epoch": 2.05,
885
+ "grad_norm": 6.508156776428223,
886
+ "learning_rate": 6.4527272727272725e-06,
887
+ "loss": 0.1424,
888
+ "step": 2460
889
+ },
890
+ {
891
+ "epoch": 2.07,
892
+ "grad_norm": 4.847611427307129,
893
+ "learning_rate": 6.416363636363637e-06,
894
+ "loss": 0.1588,
895
+ "step": 2480
896
+ },
897
+ {
898
+ "epoch": 2.08,
899
+ "grad_norm": 5.715017795562744,
900
+ "learning_rate": 6.380000000000001e-06,
901
+ "loss": 0.1896,
902
+ "step": 2500
903
+ },
904
+ {
905
+ "epoch": 2.1,
906
+ "grad_norm": 10.562045097351074,
907
+ "learning_rate": 6.3436363636363644e-06,
908
+ "loss": 0.2007,
909
+ "step": 2520
910
+ },
911
+ {
912
+ "epoch": 2.12,
913
+ "grad_norm": 3.2870736122131348,
914
+ "learning_rate": 6.307272727272727e-06,
915
+ "loss": 0.179,
916
+ "step": 2540
917
+ },
918
+ {
919
+ "epoch": 2.13,
920
+ "grad_norm": 11.021241188049316,
921
+ "learning_rate": 6.270909090909092e-06,
922
+ "loss": 0.1573,
923
+ "step": 2560
924
+ },
925
+ {
926
+ "epoch": 2.15,
927
+ "grad_norm": 7.566544055938721,
928
+ "learning_rate": 6.2345454545454555e-06,
929
+ "loss": 0.1832,
930
+ "step": 2580
931
+ },
932
+ {
933
+ "epoch": 2.17,
934
+ "grad_norm": 5.980641841888428,
935
+ "learning_rate": 6.198181818181819e-06,
936
+ "loss": 0.1498,
937
+ "step": 2600
938
+ },
939
+ {
940
+ "epoch": 2.18,
941
+ "grad_norm": 5.697175025939941,
942
+ "learning_rate": 6.161818181818182e-06,
943
+ "loss": 0.1378,
944
+ "step": 2620
945
+ },
946
+ {
947
+ "epoch": 2.2,
948
+ "grad_norm": 4.451406002044678,
949
+ "learning_rate": 6.125454545454546e-06,
950
+ "loss": 0.152,
951
+ "step": 2640
952
+ },
953
+ {
954
+ "epoch": 2.22,
955
+ "grad_norm": 5.854424476623535,
956
+ "learning_rate": 6.08909090909091e-06,
957
+ "loss": 0.1492,
958
+ "step": 2660
959
+ },
960
+ {
961
+ "epoch": 2.23,
962
+ "grad_norm": 6.480546951293945,
963
+ "learning_rate": 6.052727272727274e-06,
964
+ "loss": 0.1501,
965
+ "step": 2680
966
+ },
967
+ {
968
+ "epoch": 2.25,
969
+ "grad_norm": 3.3347442150115967,
970
+ "learning_rate": 6.016363636363637e-06,
971
+ "loss": 0.132,
972
+ "step": 2700
973
+ },
974
+ {
975
+ "epoch": 2.27,
976
+ "grad_norm": 5.9716081619262695,
977
+ "learning_rate": 5.98e-06,
978
+ "loss": 0.1483,
979
+ "step": 2720
980
+ },
981
+ {
982
+ "epoch": 2.28,
983
+ "grad_norm": 5.842201232910156,
984
+ "learning_rate": 5.943636363636364e-06,
985
+ "loss": 0.1688,
986
+ "step": 2740
987
+ },
988
+ {
989
+ "epoch": 2.3,
990
+ "grad_norm": 5.079952716827393,
991
+ "learning_rate": 5.9072727272727285e-06,
992
+ "loss": 0.167,
993
+ "step": 2760
994
+ },
995
+ {
996
+ "epoch": 2.32,
997
+ "grad_norm": 9.07481575012207,
998
+ "learning_rate": 5.870909090909091e-06,
999
+ "loss": 0.1581,
1000
+ "step": 2780
1001
+ },
1002
+ {
1003
+ "epoch": 2.33,
1004
+ "grad_norm": 9.30324935913086,
1005
+ "learning_rate": 5.834545454545455e-06,
1006
+ "loss": 0.1704,
1007
+ "step": 2800
1008
+ },
1009
+ {
1010
+ "epoch": 2.35,
1011
+ "grad_norm": 4.848841667175293,
1012
+ "learning_rate": 5.798181818181819e-06,
1013
+ "loss": 0.1894,
1014
+ "step": 2820
1015
+ },
1016
+ {
1017
+ "epoch": 2.37,
1018
+ "grad_norm": 7.52855110168457,
1019
+ "learning_rate": 5.7618181818181816e-06,
1020
+ "loss": 0.1702,
1021
+ "step": 2840
1022
+ },
1023
+ {
1024
+ "epoch": 2.38,
1025
+ "grad_norm": 6.624313831329346,
1026
+ "learning_rate": 5.725454545454546e-06,
1027
+ "loss": 0.1628,
1028
+ "step": 2860
1029
+ },
1030
+ {
1031
+ "epoch": 2.4,
1032
+ "grad_norm": 6.889793872833252,
1033
+ "learning_rate": 5.68909090909091e-06,
1034
+ "loss": 0.1705,
1035
+ "step": 2880
1036
+ },
1037
+ {
1038
+ "epoch": 2.42,
1039
+ "grad_norm": 5.635292053222656,
1040
+ "learning_rate": 5.6527272727272734e-06,
1041
+ "loss": 0.1531,
1042
+ "step": 2900
1043
+ },
1044
+ {
1045
+ "epoch": 2.43,
1046
+ "grad_norm": 4.1545729637146,
1047
+ "learning_rate": 5.616363636363636e-06,
1048
+ "loss": 0.1371,
1049
+ "step": 2920
1050
+ },
1051
+ {
1052
+ "epoch": 2.45,
1053
+ "grad_norm": 7.1985697746276855,
1054
+ "learning_rate": 5.580000000000001e-06,
1055
+ "loss": 0.1571,
1056
+ "step": 2940
1057
+ },
1058
+ {
1059
+ "epoch": 2.47,
1060
+ "grad_norm": 7.112518310546875,
1061
+ "learning_rate": 5.5436363636363645e-06,
1062
+ "loss": 0.1525,
1063
+ "step": 2960
1064
+ },
1065
+ {
1066
+ "epoch": 2.48,
1067
+ "grad_norm": 9.503728866577148,
1068
+ "learning_rate": 5.507272727272728e-06,
1069
+ "loss": 0.1426,
1070
+ "step": 2980
1071
+ },
1072
+ {
1073
+ "epoch": 2.5,
1074
+ "grad_norm": 5.497040271759033,
1075
+ "learning_rate": 5.470909090909091e-06,
1076
+ "loss": 0.1501,
1077
+ "step": 3000
1078
+ },
1079
+ {
1080
+ "epoch": 2.5,
1081
+ "eval_loss": 0.2098405510187149,
1082
+ "eval_runtime": 325.0771,
1083
+ "eval_samples_per_second": 0.308,
1084
+ "eval_steps_per_second": 0.308,
1085
+ "eval_wer": 55.5672268907563,
1086
+ "step": 3000
1087
+ },
1088
+ {
1089
+ "epoch": 2.52,
1090
+ "grad_norm": 6.7488603591918945,
1091
+ "learning_rate": 5.434545454545455e-06,
1092
+ "loss": 0.1647,
1093
+ "step": 3020
1094
+ },
1095
+ {
1096
+ "epoch": 2.53,
1097
+ "grad_norm": 5.16122579574585,
1098
+ "learning_rate": 5.398181818181819e-06,
1099
+ "loss": 0.1617,
1100
+ "step": 3040
1101
+ },
1102
+ {
1103
+ "epoch": 2.55,
1104
+ "grad_norm": 4.19830322265625,
1105
+ "learning_rate": 5.361818181818183e-06,
1106
+ "loss": 0.1484,
1107
+ "step": 3060
1108
+ },
1109
+ {
1110
+ "epoch": 2.57,
1111
+ "grad_norm": 9.339705467224121,
1112
+ "learning_rate": 5.325454545454546e-06,
1113
+ "loss": 0.153,
1114
+ "step": 3080
1115
+ },
1116
+ {
1117
+ "epoch": 2.58,
1118
+ "grad_norm": 5.106137752532959,
1119
+ "learning_rate": 5.289090909090909e-06,
1120
+ "loss": 0.141,
1121
+ "step": 3100
1122
+ },
1123
+ {
1124
+ "epoch": 2.6,
1125
+ "grad_norm": 8.092510223388672,
1126
+ "learning_rate": 5.252727272727273e-06,
1127
+ "loss": 0.1443,
1128
+ "step": 3120
1129
+ },
1130
+ {
1131
+ "epoch": 2.62,
1132
+ "grad_norm": 5.475297451019287,
1133
+ "learning_rate": 5.2163636363636376e-06,
1134
+ "loss": 0.1638,
1135
+ "step": 3140
1136
+ },
1137
+ {
1138
+ "epoch": 2.63,
1139
+ "grad_norm": 8.545796394348145,
1140
+ "learning_rate": 5.18e-06,
1141
+ "loss": 0.1382,
1142
+ "step": 3160
1143
+ },
1144
+ {
1145
+ "epoch": 2.65,
1146
+ "grad_norm": 5.189711093902588,
1147
+ "learning_rate": 5.143636363636364e-06,
1148
+ "loss": 0.1492,
1149
+ "step": 3180
1150
+ },
1151
+ {
1152
+ "epoch": 2.67,
1153
+ "grad_norm": 7.558894157409668,
1154
+ "learning_rate": 5.107272727272728e-06,
1155
+ "loss": 0.1559,
1156
+ "step": 3200
1157
+ },
1158
+ {
1159
+ "epoch": 2.68,
1160
+ "grad_norm": 5.985530853271484,
1161
+ "learning_rate": 5.0709090909090906e-06,
1162
+ "loss": 0.1605,
1163
+ "step": 3220
1164
+ },
1165
+ {
1166
+ "epoch": 2.7,
1167
+ "grad_norm": 6.780598163604736,
1168
+ "learning_rate": 5.034545454545455e-06,
1169
+ "loss": 0.148,
1170
+ "step": 3240
1171
+ },
1172
+ {
1173
+ "epoch": 2.72,
1174
+ "grad_norm": 6.225440979003906,
1175
+ "learning_rate": 4.998181818181819e-06,
1176
+ "loss": 0.1451,
1177
+ "step": 3260
1178
+ },
1179
+ {
1180
+ "epoch": 2.73,
1181
+ "grad_norm": 5.706241607666016,
1182
+ "learning_rate": 4.9618181818181824e-06,
1183
+ "loss": 0.1814,
1184
+ "step": 3280
1185
+ },
1186
+ {
1187
+ "epoch": 2.75,
1188
+ "grad_norm": 7.860342502593994,
1189
+ "learning_rate": 4.925454545454546e-06,
1190
+ "loss": 0.1407,
1191
+ "step": 3300
1192
+ },
1193
+ {
1194
+ "epoch": 2.77,
1195
+ "grad_norm": 4.965442180633545,
1196
+ "learning_rate": 4.88909090909091e-06,
1197
+ "loss": 0.1494,
1198
+ "step": 3320
1199
+ },
1200
+ {
1201
+ "epoch": 2.78,
1202
+ "grad_norm": 9.562829971313477,
1203
+ "learning_rate": 4.8527272727272735e-06,
1204
+ "loss": 0.1577,
1205
+ "step": 3340
1206
+ },
1207
+ {
1208
+ "epoch": 2.8,
1209
+ "grad_norm": 6.261291027069092,
1210
+ "learning_rate": 4.816363636363637e-06,
1211
+ "loss": 0.1561,
1212
+ "step": 3360
1213
+ },
1214
+ {
1215
+ "epoch": 2.82,
1216
+ "grad_norm": 5.747939109802246,
1217
+ "learning_rate": 4.78e-06,
1218
+ "loss": 0.1344,
1219
+ "step": 3380
1220
+ },
1221
+ {
1222
+ "epoch": 2.83,
1223
+ "grad_norm": 8.154770851135254,
1224
+ "learning_rate": 4.7436363636363645e-06,
1225
+ "loss": 0.1419,
1226
+ "step": 3400
1227
+ },
1228
+ {
1229
+ "epoch": 2.85,
1230
+ "grad_norm": 5.976426601409912,
1231
+ "learning_rate": 4.707272727272727e-06,
1232
+ "loss": 0.1314,
1233
+ "step": 3420
1234
+ },
1235
+ {
1236
+ "epoch": 2.87,
1237
+ "grad_norm": 4.834322452545166,
1238
+ "learning_rate": 4.670909090909092e-06,
1239
+ "loss": 0.1128,
1240
+ "step": 3440
1241
+ },
1242
+ {
1243
+ "epoch": 2.88,
1244
+ "grad_norm": 8.85062026977539,
1245
+ "learning_rate": 4.634545454545455e-06,
1246
+ "loss": 0.1718,
1247
+ "step": 3460
1248
+ },
1249
+ {
1250
+ "epoch": 2.9,
1251
+ "grad_norm": 3.992954969406128,
1252
+ "learning_rate": 4.598181818181818e-06,
1253
+ "loss": 0.1341,
1254
+ "step": 3480
1255
+ },
1256
+ {
1257
+ "epoch": 2.92,
1258
+ "grad_norm": 6.488000392913818,
1259
+ "learning_rate": 4.561818181818182e-06,
1260
+ "loss": 0.1401,
1261
+ "step": 3500
1262
+ },
1263
+ {
1264
+ "epoch": 2.93,
1265
+ "grad_norm": 4.876968860626221,
1266
+ "learning_rate": 4.525454545454546e-06,
1267
+ "loss": 0.1319,
1268
+ "step": 3520
1269
+ },
1270
+ {
1271
+ "epoch": 2.95,
1272
+ "grad_norm": 7.407408237457275,
1273
+ "learning_rate": 4.489090909090909e-06,
1274
+ "loss": 0.1377,
1275
+ "step": 3540
1276
+ },
1277
+ {
1278
+ "epoch": 2.97,
1279
+ "grad_norm": 5.017177104949951,
1280
+ "learning_rate": 4.452727272727273e-06,
1281
+ "loss": 0.1732,
1282
+ "step": 3560
1283
+ },
1284
+ {
1285
+ "epoch": 2.98,
1286
+ "grad_norm": 8.190740585327148,
1287
+ "learning_rate": 4.416363636363637e-06,
1288
+ "loss": 0.1225,
1289
+ "step": 3580
1290
+ },
1291
+ {
1292
+ "epoch": 3.0,
1293
+ "grad_norm": 7.947383403778076,
1294
+ "learning_rate": 4.38e-06,
1295
+ "loss": 0.1454,
1296
+ "step": 3600
1297
+ },
1298
+ {
1299
+ "epoch": 3.02,
1300
+ "grad_norm": 5.404466152191162,
1301
+ "learning_rate": 4.343636363636364e-06,
1302
+ "loss": 0.0746,
1303
+ "step": 3620
1304
+ },
1305
+ {
1306
+ "epoch": 3.03,
1307
+ "grad_norm": 5.470146179199219,
1308
+ "learning_rate": 4.307272727272728e-06,
1309
+ "loss": 0.0725,
1310
+ "step": 3640
1311
+ },
1312
+ {
1313
+ "epoch": 3.05,
1314
+ "grad_norm": 4.404048919677734,
1315
+ "learning_rate": 4.2709090909090914e-06,
1316
+ "loss": 0.088,
1317
+ "step": 3660
1318
+ },
1319
+ {
1320
+ "epoch": 3.07,
1321
+ "grad_norm": 3.169548273086548,
1322
+ "learning_rate": 4.234545454545455e-06,
1323
+ "loss": 0.0811,
1324
+ "step": 3680
1325
+ },
1326
+ {
1327
+ "epoch": 3.08,
1328
+ "grad_norm": 2.710052967071533,
1329
+ "learning_rate": 4.198181818181819e-06,
1330
+ "loss": 0.0646,
1331
+ "step": 3700
1332
+ },
1333
+ {
1334
+ "epoch": 3.1,
1335
+ "grad_norm": 5.970617294311523,
1336
+ "learning_rate": 4.1618181818181825e-06,
1337
+ "loss": 0.0789,
1338
+ "step": 3720
1339
+ },
1340
+ {
1341
+ "epoch": 3.12,
1342
+ "grad_norm": 4.657149791717529,
1343
+ "learning_rate": 4.125454545454546e-06,
1344
+ "loss": 0.0814,
1345
+ "step": 3740
1346
+ },
1347
+ {
1348
+ "epoch": 3.13,
1349
+ "grad_norm": 4.4107666015625,
1350
+ "learning_rate": 4.089090909090909e-06,
1351
+ "loss": 0.0831,
1352
+ "step": 3760
1353
+ },
1354
+ {
1355
+ "epoch": 3.15,
1356
+ "grad_norm": 6.185523509979248,
1357
+ "learning_rate": 4.0527272727272735e-06,
1358
+ "loss": 0.0672,
1359
+ "step": 3780
1360
+ },
1361
+ {
1362
+ "epoch": 3.17,
1363
+ "grad_norm": 3.5439445972442627,
1364
+ "learning_rate": 4.016363636363636e-06,
1365
+ "loss": 0.0839,
1366
+ "step": 3800
1367
+ },
1368
+ {
1369
+ "epoch": 3.18,
1370
+ "grad_norm": 5.79164981842041,
1371
+ "learning_rate": 3.980000000000001e-06,
1372
+ "loss": 0.0692,
1373
+ "step": 3820
1374
+ },
1375
+ {
1376
+ "epoch": 3.2,
1377
+ "grad_norm": 6.883912086486816,
1378
+ "learning_rate": 3.943636363636364e-06,
1379
+ "loss": 0.0798,
1380
+ "step": 3840
1381
+ },
1382
+ {
1383
+ "epoch": 3.22,
1384
+ "grad_norm": 4.028432846069336,
1385
+ "learning_rate": 3.907272727272727e-06,
1386
+ "loss": 0.0808,
1387
+ "step": 3860
1388
+ },
1389
+ {
1390
+ "epoch": 3.23,
1391
+ "grad_norm": 4.477746486663818,
1392
+ "learning_rate": 3.870909090909091e-06,
1393
+ "loss": 0.0781,
1394
+ "step": 3880
1395
+ },
1396
+ {
1397
+ "epoch": 3.25,
1398
+ "grad_norm": 2.0900027751922607,
1399
+ "learning_rate": 3.834545454545455e-06,
1400
+ "loss": 0.073,
1401
+ "step": 3900
1402
+ },
1403
+ {
1404
+ "epoch": 3.27,
1405
+ "grad_norm": 2.609961748123169,
1406
+ "learning_rate": 3.798181818181819e-06,
1407
+ "loss": 0.0961,
1408
+ "step": 3920
1409
+ },
1410
+ {
1411
+ "epoch": 3.28,
1412
+ "grad_norm": 6.023408889770508,
1413
+ "learning_rate": 3.761818181818182e-06,
1414
+ "loss": 0.0806,
1415
+ "step": 3940
1416
+ },
1417
+ {
1418
+ "epoch": 3.3,
1419
+ "grad_norm": 5.091029167175293,
1420
+ "learning_rate": 3.725454545454546e-06,
1421
+ "loss": 0.0681,
1422
+ "step": 3960
1423
+ },
1424
+ {
1425
+ "epoch": 3.32,
1426
+ "grad_norm": 5.474444389343262,
1427
+ "learning_rate": 3.6890909090909094e-06,
1428
+ "loss": 0.0817,
1429
+ "step": 3980
1430
+ },
1431
+ {
1432
+ "epoch": 3.33,
1433
+ "grad_norm": 5.174444198608398,
1434
+ "learning_rate": 3.6527272727272727e-06,
1435
+ "loss": 0.0827,
1436
+ "step": 4000
1437
+ },
1438
+ {
1439
+ "epoch": 3.33,
1440
+ "eval_loss": 0.20179511606693268,
1441
+ "eval_runtime": 319.617,
1442
+ "eval_samples_per_second": 0.313,
1443
+ "eval_steps_per_second": 0.313,
1444
+ "eval_wer": 51.78571428571429,
1445
+ "step": 4000
1446
+ },
1447
+ {
1448
+ "epoch": 3.35,
1449
+ "grad_norm": 6.867219924926758,
1450
+ "learning_rate": 3.6163636363636368e-06,
1451
+ "loss": 0.0721,
1452
+ "step": 4020
1453
+ },
1454
+ {
1455
+ "epoch": 3.37,
1456
+ "grad_norm": 3.8965790271759033,
1457
+ "learning_rate": 3.58e-06,
1458
+ "loss": 0.0912,
1459
+ "step": 4040
1460
+ },
1461
+ {
1462
+ "epoch": 3.38,
1463
+ "grad_norm": 5.221298694610596,
1464
+ "learning_rate": 3.543636363636364e-06,
1465
+ "loss": 0.0674,
1466
+ "step": 4060
1467
+ },
1468
+ {
1469
+ "epoch": 3.4,
1470
+ "grad_norm": 6.95012092590332,
1471
+ "learning_rate": 3.5072727272727274e-06,
1472
+ "loss": 0.098,
1473
+ "step": 4080
1474
+ },
1475
+ {
1476
+ "epoch": 3.42,
1477
+ "grad_norm": 8.342905044555664,
1478
+ "learning_rate": 3.4709090909090915e-06,
1479
+ "loss": 0.0741,
1480
+ "step": 4100
1481
+ },
1482
+ {
1483
+ "epoch": 3.43,
1484
+ "grad_norm": 4.47432804107666,
1485
+ "learning_rate": 3.4345454545454547e-06,
1486
+ "loss": 0.0743,
1487
+ "step": 4120
1488
+ },
1489
+ {
1490
+ "epoch": 3.45,
1491
+ "grad_norm": 4.369394779205322,
1492
+ "learning_rate": 3.3981818181818184e-06,
1493
+ "loss": 0.0728,
1494
+ "step": 4140
1495
+ },
1496
+ {
1497
+ "epoch": 3.47,
1498
+ "grad_norm": 4.1392035484313965,
1499
+ "learning_rate": 3.361818181818182e-06,
1500
+ "loss": 0.0576,
1501
+ "step": 4160
1502
+ },
1503
+ {
1504
+ "epoch": 3.48,
1505
+ "grad_norm": 8.901311874389648,
1506
+ "learning_rate": 3.3254545454545458e-06,
1507
+ "loss": 0.0723,
1508
+ "step": 4180
1509
+ },
1510
+ {
1511
+ "epoch": 3.5,
1512
+ "grad_norm": 10.933469772338867,
1513
+ "learning_rate": 3.2890909090909094e-06,
1514
+ "loss": 0.0846,
1515
+ "step": 4200
1516
+ },
1517
+ {
1518
+ "epoch": 3.52,
1519
+ "grad_norm": 2.7586188316345215,
1520
+ "learning_rate": 3.252727272727273e-06,
1521
+ "loss": 0.0724,
1522
+ "step": 4220
1523
+ },
1524
+ {
1525
+ "epoch": 3.53,
1526
+ "grad_norm": 6.231330871582031,
1527
+ "learning_rate": 3.2163636363636364e-06,
1528
+ "loss": 0.0833,
1529
+ "step": 4240
1530
+ },
1531
+ {
1532
+ "epoch": 3.55,
1533
+ "grad_norm": 6.514349460601807,
1534
+ "learning_rate": 3.1800000000000005e-06,
1535
+ "loss": 0.0693,
1536
+ "step": 4260
1537
+ },
1538
+ {
1539
+ "epoch": 3.57,
1540
+ "grad_norm": 6.608529090881348,
1541
+ "learning_rate": 3.1436363636363637e-06,
1542
+ "loss": 0.0839,
1543
+ "step": 4280
1544
+ },
1545
+ {
1546
+ "epoch": 3.58,
1547
+ "grad_norm": 4.276981353759766,
1548
+ "learning_rate": 3.107272727272728e-06,
1549
+ "loss": 0.0783,
1550
+ "step": 4300
1551
+ },
1552
+ {
1553
+ "epoch": 3.6,
1554
+ "grad_norm": 3.6309165954589844,
1555
+ "learning_rate": 3.070909090909091e-06,
1556
+ "loss": 0.0704,
1557
+ "step": 4320
1558
+ },
1559
+ {
1560
+ "epoch": 3.62,
1561
+ "grad_norm": 4.466903209686279,
1562
+ "learning_rate": 3.034545454545455e-06,
1563
+ "loss": 0.0837,
1564
+ "step": 4340
1565
+ },
1566
+ {
1567
+ "epoch": 3.63,
1568
+ "grad_norm": 3.0481879711151123,
1569
+ "learning_rate": 2.9981818181818184e-06,
1570
+ "loss": 0.0789,
1571
+ "step": 4360
1572
+ },
1573
+ {
1574
+ "epoch": 3.65,
1575
+ "grad_norm": 4.183527946472168,
1576
+ "learning_rate": 2.9618181818181817e-06,
1577
+ "loss": 0.0679,
1578
+ "step": 4380
1579
+ },
1580
+ {
1581
+ "epoch": 3.67,
1582
+ "grad_norm": 7.014336585998535,
1583
+ "learning_rate": 2.9254545454545458e-06,
1584
+ "loss": 0.0754,
1585
+ "step": 4400
1586
+ },
1587
+ {
1588
+ "epoch": 3.68,
1589
+ "grad_norm": 3.5609641075134277,
1590
+ "learning_rate": 2.889090909090909e-06,
1591
+ "loss": 0.0645,
1592
+ "step": 4420
1593
+ },
1594
+ {
1595
+ "epoch": 3.7,
1596
+ "grad_norm": 7.09955358505249,
1597
+ "learning_rate": 2.852727272727273e-06,
1598
+ "loss": 0.0582,
1599
+ "step": 4440
1600
+ },
1601
+ {
1602
+ "epoch": 3.72,
1603
+ "grad_norm": 5.648538589477539,
1604
+ "learning_rate": 2.8163636363636364e-06,
1605
+ "loss": 0.0741,
1606
+ "step": 4460
1607
+ },
1608
+ {
1609
+ "epoch": 3.73,
1610
+ "grad_norm": 4.694089889526367,
1611
+ "learning_rate": 2.7800000000000005e-06,
1612
+ "loss": 0.0952,
1613
+ "step": 4480
1614
+ },
1615
+ {
1616
+ "epoch": 3.75,
1617
+ "grad_norm": 5.376272201538086,
1618
+ "learning_rate": 2.7436363636363637e-06,
1619
+ "loss": 0.0709,
1620
+ "step": 4500
1621
+ },
1622
+ {
1623
+ "epoch": 3.77,
1624
+ "grad_norm": 4.306270599365234,
1625
+ "learning_rate": 2.7072727272727274e-06,
1626
+ "loss": 0.0833,
1627
+ "step": 4520
1628
+ },
1629
+ {
1630
+ "epoch": 3.78,
1631
+ "grad_norm": 2.599487066268921,
1632
+ "learning_rate": 2.670909090909091e-06,
1633
+ "loss": 0.0794,
1634
+ "step": 4540
1635
+ },
1636
+ {
1637
+ "epoch": 3.8,
1638
+ "grad_norm": 4.195871353149414,
1639
+ "learning_rate": 2.6345454545454548e-06,
1640
+ "loss": 0.0833,
1641
+ "step": 4560
1642
+ },
1643
+ {
1644
+ "epoch": 3.82,
1645
+ "grad_norm": 5.1401543617248535,
1646
+ "learning_rate": 2.5981818181818184e-06,
1647
+ "loss": 0.0585,
1648
+ "step": 4580
1649
+ },
1650
+ {
1651
+ "epoch": 3.83,
1652
+ "grad_norm": 8.566014289855957,
1653
+ "learning_rate": 2.561818181818182e-06,
1654
+ "loss": 0.0821,
1655
+ "step": 4600
1656
+ },
1657
+ {
1658
+ "epoch": 3.85,
1659
+ "grad_norm": 4.7075419425964355,
1660
+ "learning_rate": 2.525454545454546e-06,
1661
+ "loss": 0.0791,
1662
+ "step": 4620
1663
+ },
1664
+ {
1665
+ "epoch": 3.87,
1666
+ "grad_norm": 6.129905700683594,
1667
+ "learning_rate": 2.4890909090909095e-06,
1668
+ "loss": 0.0779,
1669
+ "step": 4640
1670
+ },
1671
+ {
1672
+ "epoch": 3.88,
1673
+ "grad_norm": 3.7267708778381348,
1674
+ "learning_rate": 2.452727272727273e-06,
1675
+ "loss": 0.0612,
1676
+ "step": 4660
1677
+ },
1678
+ {
1679
+ "epoch": 3.9,
1680
+ "grad_norm": 4.743310928344727,
1681
+ "learning_rate": 2.416363636363637e-06,
1682
+ "loss": 0.0796,
1683
+ "step": 4680
1684
+ },
1685
+ {
1686
+ "epoch": 3.92,
1687
+ "grad_norm": 4.7333173751831055,
1688
+ "learning_rate": 2.38e-06,
1689
+ "loss": 0.073,
1690
+ "step": 4700
1691
+ },
1692
+ {
1693
+ "epoch": 3.93,
1694
+ "grad_norm": 4.569043159484863,
1695
+ "learning_rate": 2.3436363636363638e-06,
1696
+ "loss": 0.0677,
1697
+ "step": 4720
1698
+ },
1699
+ {
1700
+ "epoch": 3.95,
1701
+ "grad_norm": 4.0430755615234375,
1702
+ "learning_rate": 2.3072727272727274e-06,
1703
+ "loss": 0.0627,
1704
+ "step": 4740
1705
+ },
1706
+ {
1707
+ "epoch": 3.97,
1708
+ "grad_norm": 6.368178367614746,
1709
+ "learning_rate": 2.270909090909091e-06,
1710
+ "loss": 0.0789,
1711
+ "step": 4760
1712
+ },
1713
+ {
1714
+ "epoch": 3.98,
1715
+ "grad_norm": 3.39155912399292,
1716
+ "learning_rate": 2.234545454545455e-06,
1717
+ "loss": 0.0702,
1718
+ "step": 4780
1719
+ },
1720
+ {
1721
+ "epoch": 4.0,
1722
+ "grad_norm": 6.195110321044922,
1723
+ "learning_rate": 2.1981818181818185e-06,
1724
+ "loss": 0.062,
1725
+ "step": 4800
1726
+ },
1727
+ {
1728
+ "epoch": 4.02,
1729
+ "grad_norm": 1.523270845413208,
1730
+ "learning_rate": 2.1618181818181817e-06,
1731
+ "loss": 0.0333,
1732
+ "step": 4820
1733
+ },
1734
+ {
1735
+ "epoch": 4.03,
1736
+ "grad_norm": 2.800551414489746,
1737
+ "learning_rate": 2.1254545454545454e-06,
1738
+ "loss": 0.0425,
1739
+ "step": 4840
1740
+ },
1741
+ {
1742
+ "epoch": 4.05,
1743
+ "grad_norm": 2.0081329345703125,
1744
+ "learning_rate": 2.089090909090909e-06,
1745
+ "loss": 0.0352,
1746
+ "step": 4860
1747
+ },
1748
+ {
1749
+ "epoch": 4.07,
1750
+ "grad_norm": 2.3982510566711426,
1751
+ "learning_rate": 2.0527272727272727e-06,
1752
+ "loss": 0.0307,
1753
+ "step": 4880
1754
+ },
1755
+ {
1756
+ "epoch": 4.08,
1757
+ "grad_norm": 3.013807535171509,
1758
+ "learning_rate": 2.0163636363636364e-06,
1759
+ "loss": 0.0354,
1760
+ "step": 4900
1761
+ },
1762
+ {
1763
+ "epoch": 4.1,
1764
+ "grad_norm": 4.255500793457031,
1765
+ "learning_rate": 1.98e-06,
1766
+ "loss": 0.0372,
1767
+ "step": 4920
1768
+ },
1769
+ {
1770
+ "epoch": 4.12,
1771
+ "grad_norm": 4.95035982131958,
1772
+ "learning_rate": 1.9436363636363638e-06,
1773
+ "loss": 0.0342,
1774
+ "step": 4940
1775
+ },
1776
+ {
1777
+ "epoch": 4.13,
1778
+ "grad_norm": 3.9183099269866943,
1779
+ "learning_rate": 1.9072727272727272e-06,
1780
+ "loss": 0.0306,
1781
+ "step": 4960
1782
+ },
1783
+ {
1784
+ "epoch": 4.15,
1785
+ "grad_norm": 2.7411577701568604,
1786
+ "learning_rate": 1.870909090909091e-06,
1787
+ "loss": 0.0402,
1788
+ "step": 4980
1789
+ },
1790
+ {
1791
+ "epoch": 4.17,
1792
+ "grad_norm": 6.921283721923828,
1793
+ "learning_rate": 1.8345454545454546e-06,
1794
+ "loss": 0.0459,
1795
+ "step": 5000
1796
+ },
1797
+ {
1798
+ "epoch": 4.17,
1799
+ "eval_loss": 0.20849908888339996,
1800
+ "eval_runtime": 329.3813,
1801
+ "eval_samples_per_second": 0.304,
1802
+ "eval_steps_per_second": 0.304,
1803
+ "eval_wer": 47.37394957983193,
1804
+ "step": 5000
1805
+ }
1806
+ ],
1807
+ "logging_steps": 20,
1808
+ "max_steps": 6000,
1809
+ "num_input_tokens_seen": 0,
1810
+ "num_train_epochs": 5,
1811
+ "save_steps": 1000,
1812
+ "total_flos": 4.32820383399936e+18,
1813
+ "train_batch_size": 3,
1814
+ "trial_name": null,
1815
+ "trial_params": null
1816
+ }
checkpoint-5000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:152d1e9770ea3e9899a6c5c40e00cad1f52c4e2ae9aea55ed85a6bf2ed0a638f
3
+ size 5048
checkpoint-6000/config.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
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": 768,
17
+ "decoder_attention_heads": 12,
18
+ "decoder_ffn_dim": 3072,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 12,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 12,
24
+ "encoder_ffn_dim": 3072,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 12,
43
+ "num_mel_bins": 80,
44
+ "pad_token_id": 50257,
45
+ "scale_embedding": false,
46
+ "suppress_tokens": [],
47
+ "torch_dtype": "float32",
48
+ "transformers_version": "4.39.0.dev0",
49
+ "use_cache": true,
50
+ "use_weighted_layer_sum": false,
51
+ "vocab_size": 51865
52
+ }
checkpoint-6000/generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 5,
5
+ 3
6
+ ],
7
+ [
8
+ 5,
9
+ 9
10
+ ],
11
+ [
12
+ 8,
13
+ 0
14
+ ],
15
+ [
16
+ 8,
17
+ 4
18
+ ],
19
+ [
20
+ 8,
21
+ 7
22
+ ],
23
+ [
24
+ 8,
25
+ 8
26
+ ],
27
+ [
28
+ 9,
29
+ 0
30
+ ],
31
+ [
32
+ 9,
33
+ 7
34
+ ],
35
+ [
36
+ 9,
37
+ 9
38
+ ],
39
+ [
40
+ 10,
41
+ 5
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50359
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|zh|>": 50260
162
+ },
163
+ "max_initial_timestamp_index": 50,
164
+ "max_length": 448,
165
+ "no_timestamps_token_id": 50363,
166
+ "pad_token_id": 50257,
167
+ "prev_sot_token_id": 50361,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50358,
254
+ 50359,
255
+ 50360,
256
+ 50361,
257
+ 50362
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50359,
261
+ "translate": 50358
262
+ },
263
+ "transformers_version": "4.39.0.dev0"
264
+ }
checkpoint-6000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6f9893f0835513890479b8f540c1c3f0e680df780cd2d309f70bfc23a7a76ef
3
+ size 966995080
checkpoint-6000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:309748f86ee2dd7b7360de264135dda7d90c8fc755819b76da52ec47adc66018
3
+ size 1925064044
checkpoint-6000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 80,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-6000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81f8b7c9ac1681aaf95ef7bf51adbddbaecc5cd03ad76f2c1096d01e7c08ea72
3
+ size 14244
checkpoint-6000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b058c6790c7985bfd1bded5a9fa0b7ab2021d88dab248e7efec11d21e2a8f11
3
+ size 1064
checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,2175 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 47.37394957983193,
3
+ "best_model_checkpoint": "./whisper-small-bn/checkpoint-5000",
4
+ "epoch": 5.0,
5
+ "eval_steps": 1000,
6
+ "global_step": 6000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.07,
13
+ "grad_norm": 27.802248001098633,
14
+ "learning_rate": 3.2e-07,
15
+ "loss": 2.2811,
16
+ "step": 20
17
+ },
18
+ {
19
+ "epoch": 0.13,
20
+ "grad_norm": 18.253421783447266,
21
+ "learning_rate": 7.2e-07,
22
+ "loss": 2.1964,
23
+ "step": 40
24
+ },
25
+ {
26
+ "epoch": 0.2,
27
+ "grad_norm": 13.585515975952148,
28
+ "learning_rate": 1.12e-06,
29
+ "loss": 1.8635,
30
+ "step": 60
31
+ },
32
+ {
33
+ "epoch": 0.27,
34
+ "grad_norm": 22.400362014770508,
35
+ "learning_rate": 1.52e-06,
36
+ "loss": 1.703,
37
+ "step": 80
38
+ },
39
+ {
40
+ "epoch": 0.33,
41
+ "grad_norm": 13.77535343170166,
42
+ "learning_rate": 1.9200000000000003e-06,
43
+ "loss": 1.5336,
44
+ "step": 100
45
+ },
46
+ {
47
+ "epoch": 0.4,
48
+ "grad_norm": 14.989215850830078,
49
+ "learning_rate": 2.3200000000000002e-06,
50
+ "loss": 1.4044,
51
+ "step": 120
52
+ },
53
+ {
54
+ "epoch": 0.47,
55
+ "grad_norm": 22.01858139038086,
56
+ "learning_rate": 2.7200000000000002e-06,
57
+ "loss": 1.2402,
58
+ "step": 140
59
+ },
60
+ {
61
+ "epoch": 0.53,
62
+ "grad_norm": 16.482311248779297,
63
+ "learning_rate": 3.12e-06,
64
+ "loss": 1.0922,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.6,
69
+ "grad_norm": 17.529874801635742,
70
+ "learning_rate": 3.52e-06,
71
+ "loss": 0.9583,
72
+ "step": 180
73
+ },
74
+ {
75
+ "epoch": 0.67,
76
+ "grad_norm": 18.252761840820312,
77
+ "learning_rate": 3.920000000000001e-06,
78
+ "loss": 0.8783,
79
+ "step": 200
80
+ },
81
+ {
82
+ "epoch": 0.73,
83
+ "grad_norm": 16.48067283630371,
84
+ "learning_rate": 4.32e-06,
85
+ "loss": 0.7643,
86
+ "step": 220
87
+ },
88
+ {
89
+ "epoch": 0.8,
90
+ "grad_norm": 15.062904357910156,
91
+ "learning_rate": 4.7200000000000005e-06,
92
+ "loss": 0.6991,
93
+ "step": 240
94
+ },
95
+ {
96
+ "epoch": 0.87,
97
+ "grad_norm": 14.231403350830078,
98
+ "learning_rate": 5.12e-06,
99
+ "loss": 0.6594,
100
+ "step": 260
101
+ },
102
+ {
103
+ "epoch": 0.93,
104
+ "grad_norm": 18.425230026245117,
105
+ "learning_rate": 5.5200000000000005e-06,
106
+ "loss": 0.5724,
107
+ "step": 280
108
+ },
109
+ {
110
+ "epoch": 1.0,
111
+ "grad_norm": 11.786844253540039,
112
+ "learning_rate": 5.92e-06,
113
+ "loss": 0.5687,
114
+ "step": 300
115
+ },
116
+ {
117
+ "epoch": 1.07,
118
+ "grad_norm": 15.346015930175781,
119
+ "learning_rate": 6.3200000000000005e-06,
120
+ "loss": 0.4696,
121
+ "step": 320
122
+ },
123
+ {
124
+ "epoch": 1.13,
125
+ "grad_norm": 7.936849594116211,
126
+ "learning_rate": 6.720000000000001e-06,
127
+ "loss": 0.4569,
128
+ "step": 340
129
+ },
130
+ {
131
+ "epoch": 1.2,
132
+ "grad_norm": 13.894779205322266,
133
+ "learning_rate": 7.1200000000000004e-06,
134
+ "loss": 0.4089,
135
+ "step": 360
136
+ },
137
+ {
138
+ "epoch": 1.27,
139
+ "grad_norm": 11.665081977844238,
140
+ "learning_rate": 7.520000000000001e-06,
141
+ "loss": 0.4168,
142
+ "step": 380
143
+ },
144
+ {
145
+ "epoch": 1.33,
146
+ "grad_norm": 16.7696475982666,
147
+ "learning_rate": 7.92e-06,
148
+ "loss": 0.4433,
149
+ "step": 400
150
+ },
151
+ {
152
+ "epoch": 1.4,
153
+ "grad_norm": 15.60254192352295,
154
+ "learning_rate": 8.32e-06,
155
+ "loss": 0.4013,
156
+ "step": 420
157
+ },
158
+ {
159
+ "epoch": 1.47,
160
+ "grad_norm": 13.72472858428955,
161
+ "learning_rate": 8.720000000000001e-06,
162
+ "loss": 0.3903,
163
+ "step": 440
164
+ },
165
+ {
166
+ "epoch": 1.53,
167
+ "grad_norm": 10.506444931030273,
168
+ "learning_rate": 9.12e-06,
169
+ "loss": 0.3581,
170
+ "step": 460
171
+ },
172
+ {
173
+ "epoch": 1.6,
174
+ "grad_norm": 13.643900871276855,
175
+ "learning_rate": 9.52e-06,
176
+ "loss": 0.3605,
177
+ "step": 480
178
+ },
179
+ {
180
+ "epoch": 1.67,
181
+ "grad_norm": 11.669689178466797,
182
+ "learning_rate": 9.920000000000002e-06,
183
+ "loss": 0.3548,
184
+ "step": 500
185
+ },
186
+ {
187
+ "epoch": 1.73,
188
+ "grad_norm": 9.907474517822266,
189
+ "learning_rate": 9.936000000000001e-06,
190
+ "loss": 0.3535,
191
+ "step": 520
192
+ },
193
+ {
194
+ "epoch": 1.8,
195
+ "grad_norm": 11.366904258728027,
196
+ "learning_rate": 9.856000000000002e-06,
197
+ "loss": 0.2943,
198
+ "step": 540
199
+ },
200
+ {
201
+ "epoch": 1.87,
202
+ "grad_norm": 7.5114240646362305,
203
+ "learning_rate": 9.776000000000001e-06,
204
+ "loss": 0.3057,
205
+ "step": 560
206
+ },
207
+ {
208
+ "epoch": 1.93,
209
+ "grad_norm": 11.717443466186523,
210
+ "learning_rate": 9.696000000000002e-06,
211
+ "loss": 0.2919,
212
+ "step": 580
213
+ },
214
+ {
215
+ "epoch": 2.0,
216
+ "grad_norm": 9.275129318237305,
217
+ "learning_rate": 9.616e-06,
218
+ "loss": 0.3115,
219
+ "step": 600
220
+ },
221
+ {
222
+ "epoch": 2.07,
223
+ "grad_norm": 9.388382911682129,
224
+ "learning_rate": 9.536000000000002e-06,
225
+ "loss": 0.1926,
226
+ "step": 620
227
+ },
228
+ {
229
+ "epoch": 2.13,
230
+ "grad_norm": 8.574433326721191,
231
+ "learning_rate": 9.456e-06,
232
+ "loss": 0.2193,
233
+ "step": 640
234
+ },
235
+ {
236
+ "epoch": 2.2,
237
+ "grad_norm": 5.512338638305664,
238
+ "learning_rate": 9.376000000000001e-06,
239
+ "loss": 0.18,
240
+ "step": 660
241
+ },
242
+ {
243
+ "epoch": 2.27,
244
+ "grad_norm": 9.644317626953125,
245
+ "learning_rate": 9.296e-06,
246
+ "loss": 0.1916,
247
+ "step": 680
248
+ },
249
+ {
250
+ "epoch": 2.33,
251
+ "grad_norm": 9.321474075317383,
252
+ "learning_rate": 9.216000000000001e-06,
253
+ "loss": 0.2055,
254
+ "step": 700
255
+ },
256
+ {
257
+ "epoch": 2.4,
258
+ "grad_norm": 4.868476390838623,
259
+ "learning_rate": 9.136e-06,
260
+ "loss": 0.1843,
261
+ "step": 720
262
+ },
263
+ {
264
+ "epoch": 2.47,
265
+ "grad_norm": 6.592445373535156,
266
+ "learning_rate": 9.056000000000001e-06,
267
+ "loss": 0.1713,
268
+ "step": 740
269
+ },
270
+ {
271
+ "epoch": 2.53,
272
+ "grad_norm": 8.138116836547852,
273
+ "learning_rate": 8.976e-06,
274
+ "loss": 0.1792,
275
+ "step": 760
276
+ },
277
+ {
278
+ "epoch": 2.6,
279
+ "grad_norm": 9.782510757446289,
280
+ "learning_rate": 8.896000000000001e-06,
281
+ "loss": 0.1796,
282
+ "step": 780
283
+ },
284
+ {
285
+ "epoch": 2.67,
286
+ "grad_norm": 5.867671012878418,
287
+ "learning_rate": 8.816000000000002e-06,
288
+ "loss": 0.1652,
289
+ "step": 800
290
+ },
291
+ {
292
+ "epoch": 2.73,
293
+ "grad_norm": 6.637588977813721,
294
+ "learning_rate": 8.736e-06,
295
+ "loss": 0.1735,
296
+ "step": 820
297
+ },
298
+ {
299
+ "epoch": 2.8,
300
+ "grad_norm": 11.652266502380371,
301
+ "learning_rate": 8.656000000000001e-06,
302
+ "loss": 0.1734,
303
+ "step": 840
304
+ },
305
+ {
306
+ "epoch": 2.87,
307
+ "grad_norm": 6.051203727722168,
308
+ "learning_rate": 8.576e-06,
309
+ "loss": 0.1557,
310
+ "step": 860
311
+ },
312
+ {
313
+ "epoch": 2.93,
314
+ "grad_norm": 11.018319129943848,
315
+ "learning_rate": 8.496000000000001e-06,
316
+ "loss": 0.1668,
317
+ "step": 880
318
+ },
319
+ {
320
+ "epoch": 3.0,
321
+ "grad_norm": 7.878730297088623,
322
+ "learning_rate": 8.416e-06,
323
+ "loss": 0.1706,
324
+ "step": 900
325
+ },
326
+ {
327
+ "epoch": 3.07,
328
+ "grad_norm": 6.223515510559082,
329
+ "learning_rate": 8.336000000000001e-06,
330
+ "loss": 0.0917,
331
+ "step": 920
332
+ },
333
+ {
334
+ "epoch": 3.13,
335
+ "grad_norm": 7.449841022491455,
336
+ "learning_rate": 8.256e-06,
337
+ "loss": 0.0853,
338
+ "step": 940
339
+ },
340
+ {
341
+ "epoch": 3.2,
342
+ "grad_norm": 3.824934244155884,
343
+ "learning_rate": 8.176000000000001e-06,
344
+ "loss": 0.1033,
345
+ "step": 960
346
+ },
347
+ {
348
+ "epoch": 3.27,
349
+ "grad_norm": 6.369926929473877,
350
+ "learning_rate": 8.096e-06,
351
+ "loss": 0.0997,
352
+ "step": 980
353
+ },
354
+ {
355
+ "epoch": 3.33,
356
+ "grad_norm": 6.938141822814941,
357
+ "learning_rate": 8.016e-06,
358
+ "loss": 0.095,
359
+ "step": 1000
360
+ },
361
+ {
362
+ "epoch": 3.33,
363
+ "eval_loss": 0.27550819516181946,
364
+ "eval_runtime": 345.1723,
365
+ "eval_samples_per_second": 0.29,
366
+ "eval_steps_per_second": 0.29,
367
+ "eval_wer": 60.18099547511312,
368
+ "step": 1000
369
+ },
370
+ {
371
+ "epoch": 2.18,
372
+ "grad_norm": 7.869150638580322,
373
+ "learning_rate": 7.936e-06,
374
+ "loss": 0.1443,
375
+ "step": 1020
376
+ },
377
+ {
378
+ "epoch": 2.23,
379
+ "grad_norm": 3.6447396278381348,
380
+ "learning_rate": 7.860000000000001e-06,
381
+ "loss": 0.1974,
382
+ "step": 1040
383
+ },
384
+ {
385
+ "epoch": 2.27,
386
+ "grad_norm": 9.588157653808594,
387
+ "learning_rate": 7.78e-06,
388
+ "loss": 0.1427,
389
+ "step": 1060
390
+ },
391
+ {
392
+ "epoch": 2.31,
393
+ "grad_norm": 10.796669006347656,
394
+ "learning_rate": 7.7e-06,
395
+ "loss": 0.187,
396
+ "step": 1080
397
+ },
398
+ {
399
+ "epoch": 2.36,
400
+ "grad_norm": 4.941339015960693,
401
+ "learning_rate": 7.620000000000001e-06,
402
+ "loss": 0.1568,
403
+ "step": 1100
404
+ },
405
+ {
406
+ "epoch": 2.4,
407
+ "grad_norm": 11.114888191223145,
408
+ "learning_rate": 7.540000000000001e-06,
409
+ "loss": 0.1697,
410
+ "step": 1120
411
+ },
412
+ {
413
+ "epoch": 2.44,
414
+ "grad_norm": 10.200749397277832,
415
+ "learning_rate": 7.4600000000000006e-06,
416
+ "loss": 0.1675,
417
+ "step": 1140
418
+ },
419
+ {
420
+ "epoch": 2.48,
421
+ "grad_norm": 5.67927360534668,
422
+ "learning_rate": 7.3800000000000005e-06,
423
+ "loss": 0.1557,
424
+ "step": 1160
425
+ },
426
+ {
427
+ "epoch": 2.53,
428
+ "grad_norm": 7.413832664489746,
429
+ "learning_rate": 7.304000000000001e-06,
430
+ "loss": 0.1629,
431
+ "step": 1180
432
+ },
433
+ {
434
+ "epoch": 2.57,
435
+ "grad_norm": 6.933262348175049,
436
+ "learning_rate": 7.224000000000001e-06,
437
+ "loss": 0.17,
438
+ "step": 1200
439
+ },
440
+ {
441
+ "epoch": 2.61,
442
+ "grad_norm": 4.847285747528076,
443
+ "learning_rate": 7.1440000000000005e-06,
444
+ "loss": 0.1389,
445
+ "step": 1220
446
+ },
447
+ {
448
+ "epoch": 2.66,
449
+ "grad_norm": 19.55495834350586,
450
+ "learning_rate": 7.0640000000000005e-06,
451
+ "loss": 0.1338,
452
+ "step": 1240
453
+ },
454
+ {
455
+ "epoch": 2.7,
456
+ "grad_norm": 7.148222923278809,
457
+ "learning_rate": 6.984e-06,
458
+ "loss": 0.1296,
459
+ "step": 1260
460
+ },
461
+ {
462
+ "epoch": 2.74,
463
+ "grad_norm": 6.323164939880371,
464
+ "learning_rate": 6.904e-06,
465
+ "loss": 0.163,
466
+ "step": 1280
467
+ },
468
+ {
469
+ "epoch": 2.78,
470
+ "grad_norm": 9.096046447753906,
471
+ "learning_rate": 6.824e-06,
472
+ "loss": 0.1942,
473
+ "step": 1300
474
+ },
475
+ {
476
+ "epoch": 2.83,
477
+ "grad_norm": 8.462915420532227,
478
+ "learning_rate": 6.744e-06,
479
+ "loss": 0.1576,
480
+ "step": 1320
481
+ },
482
+ {
483
+ "epoch": 2.87,
484
+ "grad_norm": 7.848359107971191,
485
+ "learning_rate": 6.664e-06,
486
+ "loss": 0.1219,
487
+ "step": 1340
488
+ },
489
+ {
490
+ "epoch": 2.91,
491
+ "grad_norm": 6.30383825302124,
492
+ "learning_rate": 6.584e-06,
493
+ "loss": 0.1404,
494
+ "step": 1360
495
+ },
496
+ {
497
+ "epoch": 2.96,
498
+ "grad_norm": 5.657530784606934,
499
+ "learning_rate": 6.504e-06,
500
+ "loss": 0.1449,
501
+ "step": 1380
502
+ },
503
+ {
504
+ "epoch": 3.0,
505
+ "grad_norm": 7.4172234535217285,
506
+ "learning_rate": 6.424e-06,
507
+ "loss": 0.1234,
508
+ "step": 1400
509
+ },
510
+ {
511
+ "epoch": 3.04,
512
+ "grad_norm": 6.550841331481934,
513
+ "learning_rate": 6.344e-06,
514
+ "loss": 0.0885,
515
+ "step": 1420
516
+ },
517
+ {
518
+ "epoch": 3.08,
519
+ "grad_norm": 6.212351322174072,
520
+ "learning_rate": 6.264e-06,
521
+ "loss": 0.0725,
522
+ "step": 1440
523
+ },
524
+ {
525
+ "epoch": 3.13,
526
+ "grad_norm": 5.2562575340271,
527
+ "learning_rate": 6.184e-06,
528
+ "loss": 0.0886,
529
+ "step": 1460
530
+ },
531
+ {
532
+ "epoch": 3.17,
533
+ "grad_norm": 7.921740531921387,
534
+ "learning_rate": 6.104000000000001e-06,
535
+ "loss": 0.0805,
536
+ "step": 1480
537
+ },
538
+ {
539
+ "epoch": 3.21,
540
+ "grad_norm": 3.5878069400787354,
541
+ "learning_rate": 6.024000000000001e-06,
542
+ "loss": 0.0628,
543
+ "step": 1500
544
+ },
545
+ {
546
+ "epoch": 3.25,
547
+ "grad_norm": 8.585208892822266,
548
+ "learning_rate": 5.944000000000001e-06,
549
+ "loss": 0.0812,
550
+ "step": 1520
551
+ },
552
+ {
553
+ "epoch": 3.3,
554
+ "grad_norm": 4.111368656158447,
555
+ "learning_rate": 5.868e-06,
556
+ "loss": 0.0694,
557
+ "step": 1540
558
+ },
559
+ {
560
+ "epoch": 3.34,
561
+ "grad_norm": 5.304944038391113,
562
+ "learning_rate": 5.788e-06,
563
+ "loss": 0.0739,
564
+ "step": 1560
565
+ },
566
+ {
567
+ "epoch": 3.38,
568
+ "grad_norm": 6.168578147888184,
569
+ "learning_rate": 5.708e-06,
570
+ "loss": 0.0821,
571
+ "step": 1580
572
+ },
573
+ {
574
+ "epoch": 3.43,
575
+ "grad_norm": 12.2472505569458,
576
+ "learning_rate": 5.628e-06,
577
+ "loss": 0.0885,
578
+ "step": 1600
579
+ },
580
+ {
581
+ "epoch": 3.47,
582
+ "grad_norm": 9.349952697753906,
583
+ "learning_rate": 5.548e-06,
584
+ "loss": 0.0607,
585
+ "step": 1620
586
+ },
587
+ {
588
+ "epoch": 3.51,
589
+ "grad_norm": 5.98253870010376,
590
+ "learning_rate": 5.468e-06,
591
+ "loss": 0.0619,
592
+ "step": 1640
593
+ },
594
+ {
595
+ "epoch": 3.55,
596
+ "grad_norm": 7.106723785400391,
597
+ "learning_rate": 5.388e-06,
598
+ "loss": 0.0992,
599
+ "step": 1660
600
+ },
601
+ {
602
+ "epoch": 3.6,
603
+ "grad_norm": 3.693020820617676,
604
+ "learning_rate": 5.308000000000001e-06,
605
+ "loss": 0.0855,
606
+ "step": 1680
607
+ },
608
+ {
609
+ "epoch": 3.64,
610
+ "grad_norm": 7.3649678230285645,
611
+ "learning_rate": 5.228000000000001e-06,
612
+ "loss": 0.081,
613
+ "step": 1700
614
+ },
615
+ {
616
+ "epoch": 3.68,
617
+ "grad_norm": 5.4454240798950195,
618
+ "learning_rate": 5.1480000000000005e-06,
619
+ "loss": 0.0872,
620
+ "step": 1720
621
+ },
622
+ {
623
+ "epoch": 3.73,
624
+ "grad_norm": 8.854460716247559,
625
+ "learning_rate": 5.0680000000000004e-06,
626
+ "loss": 0.0869,
627
+ "step": 1740
628
+ },
629
+ {
630
+ "epoch": 3.77,
631
+ "grad_norm": 3.8615269660949707,
632
+ "learning_rate": 4.988e-06,
633
+ "loss": 0.0821,
634
+ "step": 1760
635
+ },
636
+ {
637
+ "epoch": 3.81,
638
+ "grad_norm": 3.7190287113189697,
639
+ "learning_rate": 4.908e-06,
640
+ "loss": 0.0636,
641
+ "step": 1780
642
+ },
643
+ {
644
+ "epoch": 3.85,
645
+ "grad_norm": 7.683879375457764,
646
+ "learning_rate": 4.828e-06,
647
+ "loss": 0.0892,
648
+ "step": 1800
649
+ },
650
+ {
651
+ "epoch": 3.9,
652
+ "grad_norm": 3.645141839981079,
653
+ "learning_rate": 4.748e-06,
654
+ "loss": 0.0803,
655
+ "step": 1820
656
+ },
657
+ {
658
+ "epoch": 3.94,
659
+ "grad_norm": 3.8167316913604736,
660
+ "learning_rate": 4.668e-06,
661
+ "loss": 0.1012,
662
+ "step": 1840
663
+ },
664
+ {
665
+ "epoch": 3.98,
666
+ "grad_norm": 7.247605323791504,
667
+ "learning_rate": 4.588e-06,
668
+ "loss": 0.0728,
669
+ "step": 1860
670
+ },
671
+ {
672
+ "epoch": 4.03,
673
+ "grad_norm": 3.5495615005493164,
674
+ "learning_rate": 4.508e-06,
675
+ "loss": 0.0439,
676
+ "step": 1880
677
+ },
678
+ {
679
+ "epoch": 4.07,
680
+ "grad_norm": 4.105050086975098,
681
+ "learning_rate": 4.428000000000001e-06,
682
+ "loss": 0.0357,
683
+ "step": 1900
684
+ },
685
+ {
686
+ "epoch": 4.11,
687
+ "grad_norm": 9.399913787841797,
688
+ "learning_rate": 4.3480000000000006e-06,
689
+ "loss": 0.0363,
690
+ "step": 1920
691
+ },
692
+ {
693
+ "epoch": 4.15,
694
+ "grad_norm": 3.0376510620117188,
695
+ "learning_rate": 4.2680000000000005e-06,
696
+ "loss": 0.0221,
697
+ "step": 1940
698
+ },
699
+ {
700
+ "epoch": 4.2,
701
+ "grad_norm": 0.6908012628555298,
702
+ "learning_rate": 4.188e-06,
703
+ "loss": 0.0323,
704
+ "step": 1960
705
+ },
706
+ {
707
+ "epoch": 4.24,
708
+ "grad_norm": 6.255786895751953,
709
+ "learning_rate": 4.108e-06,
710
+ "loss": 0.0327,
711
+ "step": 1980
712
+ },
713
+ {
714
+ "epoch": 4.28,
715
+ "grad_norm": 7.1336669921875,
716
+ "learning_rate": 4.028e-06,
717
+ "loss": 0.0416,
718
+ "step": 2000
719
+ },
720
+ {
721
+ "epoch": 4.28,
722
+ "eval_loss": 0.27948442101478577,
723
+ "eval_runtime": 349.9899,
724
+ "eval_samples_per_second": 0.286,
725
+ "eval_steps_per_second": 0.286,
726
+ "eval_wer": 59.0146750524109,
727
+ "step": 2000
728
+ },
729
+ {
730
+ "epoch": 1.68,
731
+ "grad_norm": 13.214738845825195,
732
+ "learning_rate": 7.250909090909092e-06,
733
+ "loss": 0.2839,
734
+ "step": 2020
735
+ },
736
+ {
737
+ "epoch": 1.7,
738
+ "grad_norm": 8.846616744995117,
739
+ "learning_rate": 7.214545454545455e-06,
740
+ "loss": 0.2221,
741
+ "step": 2040
742
+ },
743
+ {
744
+ "epoch": 1.72,
745
+ "grad_norm": 11.778766632080078,
746
+ "learning_rate": 7.178181818181818e-06,
747
+ "loss": 0.2394,
748
+ "step": 2060
749
+ },
750
+ {
751
+ "epoch": 1.73,
752
+ "grad_norm": 4.484768867492676,
753
+ "learning_rate": 7.141818181818182e-06,
754
+ "loss": 0.2299,
755
+ "step": 2080
756
+ },
757
+ {
758
+ "epoch": 1.75,
759
+ "grad_norm": 4.385435104370117,
760
+ "learning_rate": 7.105454545454546e-06,
761
+ "loss": 0.2221,
762
+ "step": 2100
763
+ },
764
+ {
765
+ "epoch": 1.77,
766
+ "grad_norm": 8.751326560974121,
767
+ "learning_rate": 7.06909090909091e-06,
768
+ "loss": 0.1966,
769
+ "step": 2120
770
+ },
771
+ {
772
+ "epoch": 1.78,
773
+ "grad_norm": 8.837333679199219,
774
+ "learning_rate": 7.032727272727273e-06,
775
+ "loss": 0.1893,
776
+ "step": 2140
777
+ },
778
+ {
779
+ "epoch": 1.8,
780
+ "grad_norm": 9.331002235412598,
781
+ "learning_rate": 6.998181818181818e-06,
782
+ "loss": 0.2365,
783
+ "step": 2160
784
+ },
785
+ {
786
+ "epoch": 1.82,
787
+ "grad_norm": 6.599325180053711,
788
+ "learning_rate": 6.961818181818183e-06,
789
+ "loss": 0.2152,
790
+ "step": 2180
791
+ },
792
+ {
793
+ "epoch": 1.83,
794
+ "grad_norm": 11.716184616088867,
795
+ "learning_rate": 6.9254545454545464e-06,
796
+ "loss": 0.203,
797
+ "step": 2200
798
+ },
799
+ {
800
+ "epoch": 1.85,
801
+ "grad_norm": 6.506683349609375,
802
+ "learning_rate": 6.88909090909091e-06,
803
+ "loss": 0.2162,
804
+ "step": 2220
805
+ },
806
+ {
807
+ "epoch": 1.87,
808
+ "grad_norm": 4.64941930770874,
809
+ "learning_rate": 6.852727272727273e-06,
810
+ "loss": 0.214,
811
+ "step": 2240
812
+ },
813
+ {
814
+ "epoch": 1.88,
815
+ "grad_norm": 8.344452857971191,
816
+ "learning_rate": 6.816363636363637e-06,
817
+ "loss": 0.2106,
818
+ "step": 2260
819
+ },
820
+ {
821
+ "epoch": 1.9,
822
+ "grad_norm": 5.504006385803223,
823
+ "learning_rate": 6.780000000000001e-06,
824
+ "loss": 0.196,
825
+ "step": 2280
826
+ },
827
+ {
828
+ "epoch": 1.92,
829
+ "grad_norm": 9.84432601928711,
830
+ "learning_rate": 6.743636363636365e-06,
831
+ "loss": 0.2207,
832
+ "step": 2300
833
+ },
834
+ {
835
+ "epoch": 1.93,
836
+ "grad_norm": 7.112894535064697,
837
+ "learning_rate": 6.707272727272728e-06,
838
+ "loss": 0.2072,
839
+ "step": 2320
840
+ },
841
+ {
842
+ "epoch": 1.95,
843
+ "grad_norm": 7.10872745513916,
844
+ "learning_rate": 6.670909090909091e-06,
845
+ "loss": 0.1925,
846
+ "step": 2340
847
+ },
848
+ {
849
+ "epoch": 1.97,
850
+ "grad_norm": 8.612652778625488,
851
+ "learning_rate": 6.634545454545455e-06,
852
+ "loss": 0.2023,
853
+ "step": 2360
854
+ },
855
+ {
856
+ "epoch": 1.98,
857
+ "grad_norm": 6.083488941192627,
858
+ "learning_rate": 6.5981818181818195e-06,
859
+ "loss": 0.2111,
860
+ "step": 2380
861
+ },
862
+ {
863
+ "epoch": 2.0,
864
+ "grad_norm": 7.59145975112915,
865
+ "learning_rate": 6.561818181818182e-06,
866
+ "loss": 0.1934,
867
+ "step": 2400
868
+ },
869
+ {
870
+ "epoch": 2.02,
871
+ "grad_norm": 6.529219627380371,
872
+ "learning_rate": 6.525454545454546e-06,
873
+ "loss": 0.1756,
874
+ "step": 2420
875
+ },
876
+ {
877
+ "epoch": 2.03,
878
+ "grad_norm": 13.685791015625,
879
+ "learning_rate": 6.48909090909091e-06,
880
+ "loss": 0.1872,
881
+ "step": 2440
882
+ },
883
+ {
884
+ "epoch": 2.05,
885
+ "grad_norm": 6.508156776428223,
886
+ "learning_rate": 6.4527272727272725e-06,
887
+ "loss": 0.1424,
888
+ "step": 2460
889
+ },
890
+ {
891
+ "epoch": 2.07,
892
+ "grad_norm": 4.847611427307129,
893
+ "learning_rate": 6.416363636363637e-06,
894
+ "loss": 0.1588,
895
+ "step": 2480
896
+ },
897
+ {
898
+ "epoch": 2.08,
899
+ "grad_norm": 5.715017795562744,
900
+ "learning_rate": 6.380000000000001e-06,
901
+ "loss": 0.1896,
902
+ "step": 2500
903
+ },
904
+ {
905
+ "epoch": 2.1,
906
+ "grad_norm": 10.562045097351074,
907
+ "learning_rate": 6.3436363636363644e-06,
908
+ "loss": 0.2007,
909
+ "step": 2520
910
+ },
911
+ {
912
+ "epoch": 2.12,
913
+ "grad_norm": 3.2870736122131348,
914
+ "learning_rate": 6.307272727272727e-06,
915
+ "loss": 0.179,
916
+ "step": 2540
917
+ },
918
+ {
919
+ "epoch": 2.13,
920
+ "grad_norm": 11.021241188049316,
921
+ "learning_rate": 6.270909090909092e-06,
922
+ "loss": 0.1573,
923
+ "step": 2560
924
+ },
925
+ {
926
+ "epoch": 2.15,
927
+ "grad_norm": 7.566544055938721,
928
+ "learning_rate": 6.2345454545454555e-06,
929
+ "loss": 0.1832,
930
+ "step": 2580
931
+ },
932
+ {
933
+ "epoch": 2.17,
934
+ "grad_norm": 5.980641841888428,
935
+ "learning_rate": 6.198181818181819e-06,
936
+ "loss": 0.1498,
937
+ "step": 2600
938
+ },
939
+ {
940
+ "epoch": 2.18,
941
+ "grad_norm": 5.697175025939941,
942
+ "learning_rate": 6.161818181818182e-06,
943
+ "loss": 0.1378,
944
+ "step": 2620
945
+ },
946
+ {
947
+ "epoch": 2.2,
948
+ "grad_norm": 4.451406002044678,
949
+ "learning_rate": 6.125454545454546e-06,
950
+ "loss": 0.152,
951
+ "step": 2640
952
+ },
953
+ {
954
+ "epoch": 2.22,
955
+ "grad_norm": 5.854424476623535,
956
+ "learning_rate": 6.08909090909091e-06,
957
+ "loss": 0.1492,
958
+ "step": 2660
959
+ },
960
+ {
961
+ "epoch": 2.23,
962
+ "grad_norm": 6.480546951293945,
963
+ "learning_rate": 6.052727272727274e-06,
964
+ "loss": 0.1501,
965
+ "step": 2680
966
+ },
967
+ {
968
+ "epoch": 2.25,
969
+ "grad_norm": 3.3347442150115967,
970
+ "learning_rate": 6.016363636363637e-06,
971
+ "loss": 0.132,
972
+ "step": 2700
973
+ },
974
+ {
975
+ "epoch": 2.27,
976
+ "grad_norm": 5.9716081619262695,
977
+ "learning_rate": 5.98e-06,
978
+ "loss": 0.1483,
979
+ "step": 2720
980
+ },
981
+ {
982
+ "epoch": 2.28,
983
+ "grad_norm": 5.842201232910156,
984
+ "learning_rate": 5.943636363636364e-06,
985
+ "loss": 0.1688,
986
+ "step": 2740
987
+ },
988
+ {
989
+ "epoch": 2.3,
990
+ "grad_norm": 5.079952716827393,
991
+ "learning_rate": 5.9072727272727285e-06,
992
+ "loss": 0.167,
993
+ "step": 2760
994
+ },
995
+ {
996
+ "epoch": 2.32,
997
+ "grad_norm": 9.07481575012207,
998
+ "learning_rate": 5.870909090909091e-06,
999
+ "loss": 0.1581,
1000
+ "step": 2780
1001
+ },
1002
+ {
1003
+ "epoch": 2.33,
1004
+ "grad_norm": 9.30324935913086,
1005
+ "learning_rate": 5.834545454545455e-06,
1006
+ "loss": 0.1704,
1007
+ "step": 2800
1008
+ },
1009
+ {
1010
+ "epoch": 2.35,
1011
+ "grad_norm": 4.848841667175293,
1012
+ "learning_rate": 5.798181818181819e-06,
1013
+ "loss": 0.1894,
1014
+ "step": 2820
1015
+ },
1016
+ {
1017
+ "epoch": 2.37,
1018
+ "grad_norm": 7.52855110168457,
1019
+ "learning_rate": 5.7618181818181816e-06,
1020
+ "loss": 0.1702,
1021
+ "step": 2840
1022
+ },
1023
+ {
1024
+ "epoch": 2.38,
1025
+ "grad_norm": 6.624313831329346,
1026
+ "learning_rate": 5.725454545454546e-06,
1027
+ "loss": 0.1628,
1028
+ "step": 2860
1029
+ },
1030
+ {
1031
+ "epoch": 2.4,
1032
+ "grad_norm": 6.889793872833252,
1033
+ "learning_rate": 5.68909090909091e-06,
1034
+ "loss": 0.1705,
1035
+ "step": 2880
1036
+ },
1037
+ {
1038
+ "epoch": 2.42,
1039
+ "grad_norm": 5.635292053222656,
1040
+ "learning_rate": 5.6527272727272734e-06,
1041
+ "loss": 0.1531,
1042
+ "step": 2900
1043
+ },
1044
+ {
1045
+ "epoch": 2.43,
1046
+ "grad_norm": 4.1545729637146,
1047
+ "learning_rate": 5.616363636363636e-06,
1048
+ "loss": 0.1371,
1049
+ "step": 2920
1050
+ },
1051
+ {
1052
+ "epoch": 2.45,
1053
+ "grad_norm": 7.1985697746276855,
1054
+ "learning_rate": 5.580000000000001e-06,
1055
+ "loss": 0.1571,
1056
+ "step": 2940
1057
+ },
1058
+ {
1059
+ "epoch": 2.47,
1060
+ "grad_norm": 7.112518310546875,
1061
+ "learning_rate": 5.5436363636363645e-06,
1062
+ "loss": 0.1525,
1063
+ "step": 2960
1064
+ },
1065
+ {
1066
+ "epoch": 2.48,
1067
+ "grad_norm": 9.503728866577148,
1068
+ "learning_rate": 5.507272727272728e-06,
1069
+ "loss": 0.1426,
1070
+ "step": 2980
1071
+ },
1072
+ {
1073
+ "epoch": 2.5,
1074
+ "grad_norm": 5.497040271759033,
1075
+ "learning_rate": 5.470909090909091e-06,
1076
+ "loss": 0.1501,
1077
+ "step": 3000
1078
+ },
1079
+ {
1080
+ "epoch": 2.5,
1081
+ "eval_loss": 0.2098405510187149,
1082
+ "eval_runtime": 325.0771,
1083
+ "eval_samples_per_second": 0.308,
1084
+ "eval_steps_per_second": 0.308,
1085
+ "eval_wer": 55.5672268907563,
1086
+ "step": 3000
1087
+ },
1088
+ {
1089
+ "epoch": 2.52,
1090
+ "grad_norm": 6.7488603591918945,
1091
+ "learning_rate": 5.434545454545455e-06,
1092
+ "loss": 0.1647,
1093
+ "step": 3020
1094
+ },
1095
+ {
1096
+ "epoch": 2.53,
1097
+ "grad_norm": 5.16122579574585,
1098
+ "learning_rate": 5.398181818181819e-06,
1099
+ "loss": 0.1617,
1100
+ "step": 3040
1101
+ },
1102
+ {
1103
+ "epoch": 2.55,
1104
+ "grad_norm": 4.19830322265625,
1105
+ "learning_rate": 5.361818181818183e-06,
1106
+ "loss": 0.1484,
1107
+ "step": 3060
1108
+ },
1109
+ {
1110
+ "epoch": 2.57,
1111
+ "grad_norm": 9.339705467224121,
1112
+ "learning_rate": 5.325454545454546e-06,
1113
+ "loss": 0.153,
1114
+ "step": 3080
1115
+ },
1116
+ {
1117
+ "epoch": 2.58,
1118
+ "grad_norm": 5.106137752532959,
1119
+ "learning_rate": 5.289090909090909e-06,
1120
+ "loss": 0.141,
1121
+ "step": 3100
1122
+ },
1123
+ {
1124
+ "epoch": 2.6,
1125
+ "grad_norm": 8.092510223388672,
1126
+ "learning_rate": 5.252727272727273e-06,
1127
+ "loss": 0.1443,
1128
+ "step": 3120
1129
+ },
1130
+ {
1131
+ "epoch": 2.62,
1132
+ "grad_norm": 5.475297451019287,
1133
+ "learning_rate": 5.2163636363636376e-06,
1134
+ "loss": 0.1638,
1135
+ "step": 3140
1136
+ },
1137
+ {
1138
+ "epoch": 2.63,
1139
+ "grad_norm": 8.545796394348145,
1140
+ "learning_rate": 5.18e-06,
1141
+ "loss": 0.1382,
1142
+ "step": 3160
1143
+ },
1144
+ {
1145
+ "epoch": 2.65,
1146
+ "grad_norm": 5.189711093902588,
1147
+ "learning_rate": 5.143636363636364e-06,
1148
+ "loss": 0.1492,
1149
+ "step": 3180
1150
+ },
1151
+ {
1152
+ "epoch": 2.67,
1153
+ "grad_norm": 7.558894157409668,
1154
+ "learning_rate": 5.107272727272728e-06,
1155
+ "loss": 0.1559,
1156
+ "step": 3200
1157
+ },
1158
+ {
1159
+ "epoch": 2.68,
1160
+ "grad_norm": 5.985530853271484,
1161
+ "learning_rate": 5.0709090909090906e-06,
1162
+ "loss": 0.1605,
1163
+ "step": 3220
1164
+ },
1165
+ {
1166
+ "epoch": 2.7,
1167
+ "grad_norm": 6.780598163604736,
1168
+ "learning_rate": 5.034545454545455e-06,
1169
+ "loss": 0.148,
1170
+ "step": 3240
1171
+ },
1172
+ {
1173
+ "epoch": 2.72,
1174
+ "grad_norm": 6.225440979003906,
1175
+ "learning_rate": 4.998181818181819e-06,
1176
+ "loss": 0.1451,
1177
+ "step": 3260
1178
+ },
1179
+ {
1180
+ "epoch": 2.73,
1181
+ "grad_norm": 5.706241607666016,
1182
+ "learning_rate": 4.9618181818181824e-06,
1183
+ "loss": 0.1814,
1184
+ "step": 3280
1185
+ },
1186
+ {
1187
+ "epoch": 2.75,
1188
+ "grad_norm": 7.860342502593994,
1189
+ "learning_rate": 4.925454545454546e-06,
1190
+ "loss": 0.1407,
1191
+ "step": 3300
1192
+ },
1193
+ {
1194
+ "epoch": 2.77,
1195
+ "grad_norm": 4.965442180633545,
1196
+ "learning_rate": 4.88909090909091e-06,
1197
+ "loss": 0.1494,
1198
+ "step": 3320
1199
+ },
1200
+ {
1201
+ "epoch": 2.78,
1202
+ "grad_norm": 9.562829971313477,
1203
+ "learning_rate": 4.8527272727272735e-06,
1204
+ "loss": 0.1577,
1205
+ "step": 3340
1206
+ },
1207
+ {
1208
+ "epoch": 2.8,
1209
+ "grad_norm": 6.261291027069092,
1210
+ "learning_rate": 4.816363636363637e-06,
1211
+ "loss": 0.1561,
1212
+ "step": 3360
1213
+ },
1214
+ {
1215
+ "epoch": 2.82,
1216
+ "grad_norm": 5.747939109802246,
1217
+ "learning_rate": 4.78e-06,
1218
+ "loss": 0.1344,
1219
+ "step": 3380
1220
+ },
1221
+ {
1222
+ "epoch": 2.83,
1223
+ "grad_norm": 8.154770851135254,
1224
+ "learning_rate": 4.7436363636363645e-06,
1225
+ "loss": 0.1419,
1226
+ "step": 3400
1227
+ },
1228
+ {
1229
+ "epoch": 2.85,
1230
+ "grad_norm": 5.976426601409912,
1231
+ "learning_rate": 4.707272727272727e-06,
1232
+ "loss": 0.1314,
1233
+ "step": 3420
1234
+ },
1235
+ {
1236
+ "epoch": 2.87,
1237
+ "grad_norm": 4.834322452545166,
1238
+ "learning_rate": 4.670909090909092e-06,
1239
+ "loss": 0.1128,
1240
+ "step": 3440
1241
+ },
1242
+ {
1243
+ "epoch": 2.88,
1244
+ "grad_norm": 8.85062026977539,
1245
+ "learning_rate": 4.634545454545455e-06,
1246
+ "loss": 0.1718,
1247
+ "step": 3460
1248
+ },
1249
+ {
1250
+ "epoch": 2.9,
1251
+ "grad_norm": 3.992954969406128,
1252
+ "learning_rate": 4.598181818181818e-06,
1253
+ "loss": 0.1341,
1254
+ "step": 3480
1255
+ },
1256
+ {
1257
+ "epoch": 2.92,
1258
+ "grad_norm": 6.488000392913818,
1259
+ "learning_rate": 4.561818181818182e-06,
1260
+ "loss": 0.1401,
1261
+ "step": 3500
1262
+ },
1263
+ {
1264
+ "epoch": 2.93,
1265
+ "grad_norm": 4.876968860626221,
1266
+ "learning_rate": 4.525454545454546e-06,
1267
+ "loss": 0.1319,
1268
+ "step": 3520
1269
+ },
1270
+ {
1271
+ "epoch": 2.95,
1272
+ "grad_norm": 7.407408237457275,
1273
+ "learning_rate": 4.489090909090909e-06,
1274
+ "loss": 0.1377,
1275
+ "step": 3540
1276
+ },
1277
+ {
1278
+ "epoch": 2.97,
1279
+ "grad_norm": 5.017177104949951,
1280
+ "learning_rate": 4.452727272727273e-06,
1281
+ "loss": 0.1732,
1282
+ "step": 3560
1283
+ },
1284
+ {
1285
+ "epoch": 2.98,
1286
+ "grad_norm": 8.190740585327148,
1287
+ "learning_rate": 4.416363636363637e-06,
1288
+ "loss": 0.1225,
1289
+ "step": 3580
1290
+ },
1291
+ {
1292
+ "epoch": 3.0,
1293
+ "grad_norm": 7.947383403778076,
1294
+ "learning_rate": 4.38e-06,
1295
+ "loss": 0.1454,
1296
+ "step": 3600
1297
+ },
1298
+ {
1299
+ "epoch": 3.02,
1300
+ "grad_norm": 5.404466152191162,
1301
+ "learning_rate": 4.343636363636364e-06,
1302
+ "loss": 0.0746,
1303
+ "step": 3620
1304
+ },
1305
+ {
1306
+ "epoch": 3.03,
1307
+ "grad_norm": 5.470146179199219,
1308
+ "learning_rate": 4.307272727272728e-06,
1309
+ "loss": 0.0725,
1310
+ "step": 3640
1311
+ },
1312
+ {
1313
+ "epoch": 3.05,
1314
+ "grad_norm": 4.404048919677734,
1315
+ "learning_rate": 4.2709090909090914e-06,
1316
+ "loss": 0.088,
1317
+ "step": 3660
1318
+ },
1319
+ {
1320
+ "epoch": 3.07,
1321
+ "grad_norm": 3.169548273086548,
1322
+ "learning_rate": 4.234545454545455e-06,
1323
+ "loss": 0.0811,
1324
+ "step": 3680
1325
+ },
1326
+ {
1327
+ "epoch": 3.08,
1328
+ "grad_norm": 2.710052967071533,
1329
+ "learning_rate": 4.198181818181819e-06,
1330
+ "loss": 0.0646,
1331
+ "step": 3700
1332
+ },
1333
+ {
1334
+ "epoch": 3.1,
1335
+ "grad_norm": 5.970617294311523,
1336
+ "learning_rate": 4.1618181818181825e-06,
1337
+ "loss": 0.0789,
1338
+ "step": 3720
1339
+ },
1340
+ {
1341
+ "epoch": 3.12,
1342
+ "grad_norm": 4.657149791717529,
1343
+ "learning_rate": 4.125454545454546e-06,
1344
+ "loss": 0.0814,
1345
+ "step": 3740
1346
+ },
1347
+ {
1348
+ "epoch": 3.13,
1349
+ "grad_norm": 4.4107666015625,
1350
+ "learning_rate": 4.089090909090909e-06,
1351
+ "loss": 0.0831,
1352
+ "step": 3760
1353
+ },
1354
+ {
1355
+ "epoch": 3.15,
1356
+ "grad_norm": 6.185523509979248,
1357
+ "learning_rate": 4.0527272727272735e-06,
1358
+ "loss": 0.0672,
1359
+ "step": 3780
1360
+ },
1361
+ {
1362
+ "epoch": 3.17,
1363
+ "grad_norm": 3.5439445972442627,
1364
+ "learning_rate": 4.016363636363636e-06,
1365
+ "loss": 0.0839,
1366
+ "step": 3800
1367
+ },
1368
+ {
1369
+ "epoch": 3.18,
1370
+ "grad_norm": 5.79164981842041,
1371
+ "learning_rate": 3.980000000000001e-06,
1372
+ "loss": 0.0692,
1373
+ "step": 3820
1374
+ },
1375
+ {
1376
+ "epoch": 3.2,
1377
+ "grad_norm": 6.883912086486816,
1378
+ "learning_rate": 3.943636363636364e-06,
1379
+ "loss": 0.0798,
1380
+ "step": 3840
1381
+ },
1382
+ {
1383
+ "epoch": 3.22,
1384
+ "grad_norm": 4.028432846069336,
1385
+ "learning_rate": 3.907272727272727e-06,
1386
+ "loss": 0.0808,
1387
+ "step": 3860
1388
+ },
1389
+ {
1390
+ "epoch": 3.23,
1391
+ "grad_norm": 4.477746486663818,
1392
+ "learning_rate": 3.870909090909091e-06,
1393
+ "loss": 0.0781,
1394
+ "step": 3880
1395
+ },
1396
+ {
1397
+ "epoch": 3.25,
1398
+ "grad_norm": 2.0900027751922607,
1399
+ "learning_rate": 3.834545454545455e-06,
1400
+ "loss": 0.073,
1401
+ "step": 3900
1402
+ },
1403
+ {
1404
+ "epoch": 3.27,
1405
+ "grad_norm": 2.609961748123169,
1406
+ "learning_rate": 3.798181818181819e-06,
1407
+ "loss": 0.0961,
1408
+ "step": 3920
1409
+ },
1410
+ {
1411
+ "epoch": 3.28,
1412
+ "grad_norm": 6.023408889770508,
1413
+ "learning_rate": 3.761818181818182e-06,
1414
+ "loss": 0.0806,
1415
+ "step": 3940
1416
+ },
1417
+ {
1418
+ "epoch": 3.3,
1419
+ "grad_norm": 5.091029167175293,
1420
+ "learning_rate": 3.725454545454546e-06,
1421
+ "loss": 0.0681,
1422
+ "step": 3960
1423
+ },
1424
+ {
1425
+ "epoch": 3.32,
1426
+ "grad_norm": 5.474444389343262,
1427
+ "learning_rate": 3.6890909090909094e-06,
1428
+ "loss": 0.0817,
1429
+ "step": 3980
1430
+ },
1431
+ {
1432
+ "epoch": 3.33,
1433
+ "grad_norm": 5.174444198608398,
1434
+ "learning_rate": 3.6527272727272727e-06,
1435
+ "loss": 0.0827,
1436
+ "step": 4000
1437
+ },
1438
+ {
1439
+ "epoch": 3.33,
1440
+ "eval_loss": 0.20179511606693268,
1441
+ "eval_runtime": 319.617,
1442
+ "eval_samples_per_second": 0.313,
1443
+ "eval_steps_per_second": 0.313,
1444
+ "eval_wer": 51.78571428571429,
1445
+ "step": 4000
1446
+ },
1447
+ {
1448
+ "epoch": 3.35,
1449
+ "grad_norm": 6.867219924926758,
1450
+ "learning_rate": 3.6163636363636368e-06,
1451
+ "loss": 0.0721,
1452
+ "step": 4020
1453
+ },
1454
+ {
1455
+ "epoch": 3.37,
1456
+ "grad_norm": 3.8965790271759033,
1457
+ "learning_rate": 3.58e-06,
1458
+ "loss": 0.0912,
1459
+ "step": 4040
1460
+ },
1461
+ {
1462
+ "epoch": 3.38,
1463
+ "grad_norm": 5.221298694610596,
1464
+ "learning_rate": 3.543636363636364e-06,
1465
+ "loss": 0.0674,
1466
+ "step": 4060
1467
+ },
1468
+ {
1469
+ "epoch": 3.4,
1470
+ "grad_norm": 6.95012092590332,
1471
+ "learning_rate": 3.5072727272727274e-06,
1472
+ "loss": 0.098,
1473
+ "step": 4080
1474
+ },
1475
+ {
1476
+ "epoch": 3.42,
1477
+ "grad_norm": 8.342905044555664,
1478
+ "learning_rate": 3.4709090909090915e-06,
1479
+ "loss": 0.0741,
1480
+ "step": 4100
1481
+ },
1482
+ {
1483
+ "epoch": 3.43,
1484
+ "grad_norm": 4.47432804107666,
1485
+ "learning_rate": 3.4345454545454547e-06,
1486
+ "loss": 0.0743,
1487
+ "step": 4120
1488
+ },
1489
+ {
1490
+ "epoch": 3.45,
1491
+ "grad_norm": 4.369394779205322,
1492
+ "learning_rate": 3.3981818181818184e-06,
1493
+ "loss": 0.0728,
1494
+ "step": 4140
1495
+ },
1496
+ {
1497
+ "epoch": 3.47,
1498
+ "grad_norm": 4.1392035484313965,
1499
+ "learning_rate": 3.361818181818182e-06,
1500
+ "loss": 0.0576,
1501
+ "step": 4160
1502
+ },
1503
+ {
1504
+ "epoch": 3.48,
1505
+ "grad_norm": 8.901311874389648,
1506
+ "learning_rate": 3.3254545454545458e-06,
1507
+ "loss": 0.0723,
1508
+ "step": 4180
1509
+ },
1510
+ {
1511
+ "epoch": 3.5,
1512
+ "grad_norm": 10.933469772338867,
1513
+ "learning_rate": 3.2890909090909094e-06,
1514
+ "loss": 0.0846,
1515
+ "step": 4200
1516
+ },
1517
+ {
1518
+ "epoch": 3.52,
1519
+ "grad_norm": 2.7586188316345215,
1520
+ "learning_rate": 3.252727272727273e-06,
1521
+ "loss": 0.0724,
1522
+ "step": 4220
1523
+ },
1524
+ {
1525
+ "epoch": 3.53,
1526
+ "grad_norm": 6.231330871582031,
1527
+ "learning_rate": 3.2163636363636364e-06,
1528
+ "loss": 0.0833,
1529
+ "step": 4240
1530
+ },
1531
+ {
1532
+ "epoch": 3.55,
1533
+ "grad_norm": 6.514349460601807,
1534
+ "learning_rate": 3.1800000000000005e-06,
1535
+ "loss": 0.0693,
1536
+ "step": 4260
1537
+ },
1538
+ {
1539
+ "epoch": 3.57,
1540
+ "grad_norm": 6.608529090881348,
1541
+ "learning_rate": 3.1436363636363637e-06,
1542
+ "loss": 0.0839,
1543
+ "step": 4280
1544
+ },
1545
+ {
1546
+ "epoch": 3.58,
1547
+ "grad_norm": 4.276981353759766,
1548
+ "learning_rate": 3.107272727272728e-06,
1549
+ "loss": 0.0783,
1550
+ "step": 4300
1551
+ },
1552
+ {
1553
+ "epoch": 3.6,
1554
+ "grad_norm": 3.6309165954589844,
1555
+ "learning_rate": 3.070909090909091e-06,
1556
+ "loss": 0.0704,
1557
+ "step": 4320
1558
+ },
1559
+ {
1560
+ "epoch": 3.62,
1561
+ "grad_norm": 4.466903209686279,
1562
+ "learning_rate": 3.034545454545455e-06,
1563
+ "loss": 0.0837,
1564
+ "step": 4340
1565
+ },
1566
+ {
1567
+ "epoch": 3.63,
1568
+ "grad_norm": 3.0481879711151123,
1569
+ "learning_rate": 2.9981818181818184e-06,
1570
+ "loss": 0.0789,
1571
+ "step": 4360
1572
+ },
1573
+ {
1574
+ "epoch": 3.65,
1575
+ "grad_norm": 4.183527946472168,
1576
+ "learning_rate": 2.9618181818181817e-06,
1577
+ "loss": 0.0679,
1578
+ "step": 4380
1579
+ },
1580
+ {
1581
+ "epoch": 3.67,
1582
+ "grad_norm": 7.014336585998535,
1583
+ "learning_rate": 2.9254545454545458e-06,
1584
+ "loss": 0.0754,
1585
+ "step": 4400
1586
+ },
1587
+ {
1588
+ "epoch": 3.68,
1589
+ "grad_norm": 3.5609641075134277,
1590
+ "learning_rate": 2.889090909090909e-06,
1591
+ "loss": 0.0645,
1592
+ "step": 4420
1593
+ },
1594
+ {
1595
+ "epoch": 3.7,
1596
+ "grad_norm": 7.09955358505249,
1597
+ "learning_rate": 2.852727272727273e-06,
1598
+ "loss": 0.0582,
1599
+ "step": 4440
1600
+ },
1601
+ {
1602
+ "epoch": 3.72,
1603
+ "grad_norm": 5.648538589477539,
1604
+ "learning_rate": 2.8163636363636364e-06,
1605
+ "loss": 0.0741,
1606
+ "step": 4460
1607
+ },
1608
+ {
1609
+ "epoch": 3.73,
1610
+ "grad_norm": 4.694089889526367,
1611
+ "learning_rate": 2.7800000000000005e-06,
1612
+ "loss": 0.0952,
1613
+ "step": 4480
1614
+ },
1615
+ {
1616
+ "epoch": 3.75,
1617
+ "grad_norm": 5.376272201538086,
1618
+ "learning_rate": 2.7436363636363637e-06,
1619
+ "loss": 0.0709,
1620
+ "step": 4500
1621
+ },
1622
+ {
1623
+ "epoch": 3.77,
1624
+ "grad_norm": 4.306270599365234,
1625
+ "learning_rate": 2.7072727272727274e-06,
1626
+ "loss": 0.0833,
1627
+ "step": 4520
1628
+ },
1629
+ {
1630
+ "epoch": 3.78,
1631
+ "grad_norm": 2.599487066268921,
1632
+ "learning_rate": 2.670909090909091e-06,
1633
+ "loss": 0.0794,
1634
+ "step": 4540
1635
+ },
1636
+ {
1637
+ "epoch": 3.8,
1638
+ "grad_norm": 4.195871353149414,
1639
+ "learning_rate": 2.6345454545454548e-06,
1640
+ "loss": 0.0833,
1641
+ "step": 4560
1642
+ },
1643
+ {
1644
+ "epoch": 3.82,
1645
+ "grad_norm": 5.1401543617248535,
1646
+ "learning_rate": 2.5981818181818184e-06,
1647
+ "loss": 0.0585,
1648
+ "step": 4580
1649
+ },
1650
+ {
1651
+ "epoch": 3.83,
1652
+ "grad_norm": 8.566014289855957,
1653
+ "learning_rate": 2.561818181818182e-06,
1654
+ "loss": 0.0821,
1655
+ "step": 4600
1656
+ },
1657
+ {
1658
+ "epoch": 3.85,
1659
+ "grad_norm": 4.7075419425964355,
1660
+ "learning_rate": 2.525454545454546e-06,
1661
+ "loss": 0.0791,
1662
+ "step": 4620
1663
+ },
1664
+ {
1665
+ "epoch": 3.87,
1666
+ "grad_norm": 6.129905700683594,
1667
+ "learning_rate": 2.4890909090909095e-06,
1668
+ "loss": 0.0779,
1669
+ "step": 4640
1670
+ },
1671
+ {
1672
+ "epoch": 3.88,
1673
+ "grad_norm": 3.7267708778381348,
1674
+ "learning_rate": 2.452727272727273e-06,
1675
+ "loss": 0.0612,
1676
+ "step": 4660
1677
+ },
1678
+ {
1679
+ "epoch": 3.9,
1680
+ "grad_norm": 4.743310928344727,
1681
+ "learning_rate": 2.416363636363637e-06,
1682
+ "loss": 0.0796,
1683
+ "step": 4680
1684
+ },
1685
+ {
1686
+ "epoch": 3.92,
1687
+ "grad_norm": 4.7333173751831055,
1688
+ "learning_rate": 2.38e-06,
1689
+ "loss": 0.073,
1690
+ "step": 4700
1691
+ },
1692
+ {
1693
+ "epoch": 3.93,
1694
+ "grad_norm": 4.569043159484863,
1695
+ "learning_rate": 2.3436363636363638e-06,
1696
+ "loss": 0.0677,
1697
+ "step": 4720
1698
+ },
1699
+ {
1700
+ "epoch": 3.95,
1701
+ "grad_norm": 4.0430755615234375,
1702
+ "learning_rate": 2.3072727272727274e-06,
1703
+ "loss": 0.0627,
1704
+ "step": 4740
1705
+ },
1706
+ {
1707
+ "epoch": 3.97,
1708
+ "grad_norm": 6.368178367614746,
1709
+ "learning_rate": 2.270909090909091e-06,
1710
+ "loss": 0.0789,
1711
+ "step": 4760
1712
+ },
1713
+ {
1714
+ "epoch": 3.98,
1715
+ "grad_norm": 3.39155912399292,
1716
+ "learning_rate": 2.234545454545455e-06,
1717
+ "loss": 0.0702,
1718
+ "step": 4780
1719
+ },
1720
+ {
1721
+ "epoch": 4.0,
1722
+ "grad_norm": 6.195110321044922,
1723
+ "learning_rate": 2.1981818181818185e-06,
1724
+ "loss": 0.062,
1725
+ "step": 4800
1726
+ },
1727
+ {
1728
+ "epoch": 4.02,
1729
+ "grad_norm": 1.523270845413208,
1730
+ "learning_rate": 2.1618181818181817e-06,
1731
+ "loss": 0.0333,
1732
+ "step": 4820
1733
+ },
1734
+ {
1735
+ "epoch": 4.03,
1736
+ "grad_norm": 2.800551414489746,
1737
+ "learning_rate": 2.1254545454545454e-06,
1738
+ "loss": 0.0425,
1739
+ "step": 4840
1740
+ },
1741
+ {
1742
+ "epoch": 4.05,
1743
+ "grad_norm": 2.0081329345703125,
1744
+ "learning_rate": 2.089090909090909e-06,
1745
+ "loss": 0.0352,
1746
+ "step": 4860
1747
+ },
1748
+ {
1749
+ "epoch": 4.07,
1750
+ "grad_norm": 2.3982510566711426,
1751
+ "learning_rate": 2.0527272727272727e-06,
1752
+ "loss": 0.0307,
1753
+ "step": 4880
1754
+ },
1755
+ {
1756
+ "epoch": 4.08,
1757
+ "grad_norm": 3.013807535171509,
1758
+ "learning_rate": 2.0163636363636364e-06,
1759
+ "loss": 0.0354,
1760
+ "step": 4900
1761
+ },
1762
+ {
1763
+ "epoch": 4.1,
1764
+ "grad_norm": 4.255500793457031,
1765
+ "learning_rate": 1.98e-06,
1766
+ "loss": 0.0372,
1767
+ "step": 4920
1768
+ },
1769
+ {
1770
+ "epoch": 4.12,
1771
+ "grad_norm": 4.95035982131958,
1772
+ "learning_rate": 1.9436363636363638e-06,
1773
+ "loss": 0.0342,
1774
+ "step": 4940
1775
+ },
1776
+ {
1777
+ "epoch": 4.13,
1778
+ "grad_norm": 3.9183099269866943,
1779
+ "learning_rate": 1.9072727272727272e-06,
1780
+ "loss": 0.0306,
1781
+ "step": 4960
1782
+ },
1783
+ {
1784
+ "epoch": 4.15,
1785
+ "grad_norm": 2.7411577701568604,
1786
+ "learning_rate": 1.870909090909091e-06,
1787
+ "loss": 0.0402,
1788
+ "step": 4980
1789
+ },
1790
+ {
1791
+ "epoch": 4.17,
1792
+ "grad_norm": 6.921283721923828,
1793
+ "learning_rate": 1.8345454545454546e-06,
1794
+ "loss": 0.0459,
1795
+ "step": 5000
1796
+ },
1797
+ {
1798
+ "epoch": 4.17,
1799
+ "eval_loss": 0.20849908888339996,
1800
+ "eval_runtime": 329.3813,
1801
+ "eval_samples_per_second": 0.304,
1802
+ "eval_steps_per_second": 0.304,
1803
+ "eval_wer": 47.37394957983193,
1804
+ "step": 5000
1805
+ },
1806
+ {
1807
+ "epoch": 4.18,
1808
+ "grad_norm": 4.997645378112793,
1809
+ "learning_rate": 1.7981818181818183e-06,
1810
+ "loss": 0.0299,
1811
+ "step": 5020
1812
+ },
1813
+ {
1814
+ "epoch": 4.2,
1815
+ "grad_norm": 2.389558792114258,
1816
+ "learning_rate": 1.761818181818182e-06,
1817
+ "loss": 0.032,
1818
+ "step": 5040
1819
+ },
1820
+ {
1821
+ "epoch": 4.22,
1822
+ "grad_norm": 4.187376022338867,
1823
+ "learning_rate": 1.7254545454545456e-06,
1824
+ "loss": 0.0276,
1825
+ "step": 5060
1826
+ },
1827
+ {
1828
+ "epoch": 4.23,
1829
+ "grad_norm": 4.168838977813721,
1830
+ "learning_rate": 1.689090909090909e-06,
1831
+ "loss": 0.0398,
1832
+ "step": 5080
1833
+ },
1834
+ {
1835
+ "epoch": 4.25,
1836
+ "grad_norm": 3.2597668170928955,
1837
+ "learning_rate": 1.6527272727272728e-06,
1838
+ "loss": 0.0269,
1839
+ "step": 5100
1840
+ },
1841
+ {
1842
+ "epoch": 4.27,
1843
+ "grad_norm": 5.048752307891846,
1844
+ "learning_rate": 1.6163636363636364e-06,
1845
+ "loss": 0.0294,
1846
+ "step": 5120
1847
+ },
1848
+ {
1849
+ "epoch": 4.28,
1850
+ "grad_norm": 2.3977174758911133,
1851
+ "learning_rate": 1.5800000000000001e-06,
1852
+ "loss": 0.033,
1853
+ "step": 5140
1854
+ },
1855
+ {
1856
+ "epoch": 4.3,
1857
+ "grad_norm": 1.9564763307571411,
1858
+ "learning_rate": 1.5436363636363638e-06,
1859
+ "loss": 0.0302,
1860
+ "step": 5160
1861
+ },
1862
+ {
1863
+ "epoch": 4.32,
1864
+ "grad_norm": 4.131407260894775,
1865
+ "learning_rate": 1.5072727272727275e-06,
1866
+ "loss": 0.0377,
1867
+ "step": 5180
1868
+ },
1869
+ {
1870
+ "epoch": 4.33,
1871
+ "grad_norm": 1.7391817569732666,
1872
+ "learning_rate": 1.470909090909091e-06,
1873
+ "loss": 0.0383,
1874
+ "step": 5200
1875
+ },
1876
+ {
1877
+ "epoch": 4.35,
1878
+ "grad_norm": 1.834222674369812,
1879
+ "learning_rate": 1.4345454545454546e-06,
1880
+ "loss": 0.0306,
1881
+ "step": 5220
1882
+ },
1883
+ {
1884
+ "epoch": 4.37,
1885
+ "grad_norm": 3.12608003616333,
1886
+ "learning_rate": 1.3981818181818183e-06,
1887
+ "loss": 0.0368,
1888
+ "step": 5240
1889
+ },
1890
+ {
1891
+ "epoch": 4.38,
1892
+ "grad_norm": 4.655078887939453,
1893
+ "learning_rate": 1.361818181818182e-06,
1894
+ "loss": 0.0329,
1895
+ "step": 5260
1896
+ },
1897
+ {
1898
+ "epoch": 4.4,
1899
+ "grad_norm": 5.355417251586914,
1900
+ "learning_rate": 1.3254545454545456e-06,
1901
+ "loss": 0.0319,
1902
+ "step": 5280
1903
+ },
1904
+ {
1905
+ "epoch": 4.42,
1906
+ "grad_norm": 3.4958527088165283,
1907
+ "learning_rate": 1.2890909090909093e-06,
1908
+ "loss": 0.027,
1909
+ "step": 5300
1910
+ },
1911
+ {
1912
+ "epoch": 4.43,
1913
+ "grad_norm": 2.762265920639038,
1914
+ "learning_rate": 1.252727272727273e-06,
1915
+ "loss": 0.0316,
1916
+ "step": 5320
1917
+ },
1918
+ {
1919
+ "epoch": 4.45,
1920
+ "grad_norm": 2.627501964569092,
1921
+ "learning_rate": 1.2163636363636365e-06,
1922
+ "loss": 0.0365,
1923
+ "step": 5340
1924
+ },
1925
+ {
1926
+ "epoch": 4.47,
1927
+ "grad_norm": 11.914468765258789,
1928
+ "learning_rate": 1.1800000000000001e-06,
1929
+ "loss": 0.037,
1930
+ "step": 5360
1931
+ },
1932
+ {
1933
+ "epoch": 4.48,
1934
+ "grad_norm": 1.0061908960342407,
1935
+ "learning_rate": 1.1436363636363638e-06,
1936
+ "loss": 0.0348,
1937
+ "step": 5380
1938
+ },
1939
+ {
1940
+ "epoch": 4.5,
1941
+ "grad_norm": 3.8775386810302734,
1942
+ "learning_rate": 1.1072727272727275e-06,
1943
+ "loss": 0.042,
1944
+ "step": 5400
1945
+ },
1946
+ {
1947
+ "epoch": 4.52,
1948
+ "grad_norm": 2.429129123687744,
1949
+ "learning_rate": 1.0709090909090912e-06,
1950
+ "loss": 0.036,
1951
+ "step": 5420
1952
+ },
1953
+ {
1954
+ "epoch": 4.53,
1955
+ "grad_norm": 8.489389419555664,
1956
+ "learning_rate": 1.0345454545454546e-06,
1957
+ "loss": 0.0516,
1958
+ "step": 5440
1959
+ },
1960
+ {
1961
+ "epoch": 4.55,
1962
+ "grad_norm": 1.5038552284240723,
1963
+ "learning_rate": 9.981818181818183e-07,
1964
+ "loss": 0.0401,
1965
+ "step": 5460
1966
+ },
1967
+ {
1968
+ "epoch": 4.57,
1969
+ "grad_norm": 2.365408182144165,
1970
+ "learning_rate": 9.61818181818182e-07,
1971
+ "loss": 0.0276,
1972
+ "step": 5480
1973
+ },
1974
+ {
1975
+ "epoch": 4.58,
1976
+ "grad_norm": 4.249473571777344,
1977
+ "learning_rate": 9.254545454545454e-07,
1978
+ "loss": 0.036,
1979
+ "step": 5500
1980
+ },
1981
+ {
1982
+ "epoch": 4.6,
1983
+ "grad_norm": 2.6517717838287354,
1984
+ "learning_rate": 8.890909090909091e-07,
1985
+ "loss": 0.0297,
1986
+ "step": 5520
1987
+ },
1988
+ {
1989
+ "epoch": 4.62,
1990
+ "grad_norm": 3.9568588733673096,
1991
+ "learning_rate": 8.527272727272728e-07,
1992
+ "loss": 0.0391,
1993
+ "step": 5540
1994
+ },
1995
+ {
1996
+ "epoch": 4.63,
1997
+ "grad_norm": 3.0504679679870605,
1998
+ "learning_rate": 8.163636363636364e-07,
1999
+ "loss": 0.0292,
2000
+ "step": 5560
2001
+ },
2002
+ {
2003
+ "epoch": 4.65,
2004
+ "grad_norm": 0.5098550319671631,
2005
+ "learning_rate": 7.8e-07,
2006
+ "loss": 0.0362,
2007
+ "step": 5580
2008
+ },
2009
+ {
2010
+ "epoch": 4.67,
2011
+ "grad_norm": 2.2305212020874023,
2012
+ "learning_rate": 7.436363636363637e-07,
2013
+ "loss": 0.039,
2014
+ "step": 5600
2015
+ },
2016
+ {
2017
+ "epoch": 4.68,
2018
+ "grad_norm": 2.1624348163604736,
2019
+ "learning_rate": 7.072727272727273e-07,
2020
+ "loss": 0.0322,
2021
+ "step": 5620
2022
+ },
2023
+ {
2024
+ "epoch": 4.7,
2025
+ "grad_norm": 2.08268141746521,
2026
+ "learning_rate": 6.70909090909091e-07,
2027
+ "loss": 0.0354,
2028
+ "step": 5640
2029
+ },
2030
+ {
2031
+ "epoch": 4.72,
2032
+ "grad_norm": 3.532118082046509,
2033
+ "learning_rate": 6.345454545454546e-07,
2034
+ "loss": 0.0276,
2035
+ "step": 5660
2036
+ },
2037
+ {
2038
+ "epoch": 4.73,
2039
+ "grad_norm": 1.8796314001083374,
2040
+ "learning_rate": 5.981818181818182e-07,
2041
+ "loss": 0.0226,
2042
+ "step": 5680
2043
+ },
2044
+ {
2045
+ "epoch": 4.75,
2046
+ "grad_norm": 5.283046245574951,
2047
+ "learning_rate": 5.618181818181819e-07,
2048
+ "loss": 0.0325,
2049
+ "step": 5700
2050
+ },
2051
+ {
2052
+ "epoch": 4.77,
2053
+ "grad_norm": 3.05381178855896,
2054
+ "learning_rate": 5.254545454545456e-07,
2055
+ "loss": 0.0366,
2056
+ "step": 5720
2057
+ },
2058
+ {
2059
+ "epoch": 4.78,
2060
+ "grad_norm": 3.8481531143188477,
2061
+ "learning_rate": 4.890909090909091e-07,
2062
+ "loss": 0.0228,
2063
+ "step": 5740
2064
+ },
2065
+ {
2066
+ "epoch": 4.8,
2067
+ "grad_norm": 2.535304307937622,
2068
+ "learning_rate": 4.527272727272727e-07,
2069
+ "loss": 0.0288,
2070
+ "step": 5760
2071
+ },
2072
+ {
2073
+ "epoch": 4.82,
2074
+ "grad_norm": 7.319835662841797,
2075
+ "learning_rate": 4.163636363636364e-07,
2076
+ "loss": 0.0341,
2077
+ "step": 5780
2078
+ },
2079
+ {
2080
+ "epoch": 4.83,
2081
+ "grad_norm": 4.280138969421387,
2082
+ "learning_rate": 3.8e-07,
2083
+ "loss": 0.0311,
2084
+ "step": 5800
2085
+ },
2086
+ {
2087
+ "epoch": 4.85,
2088
+ "grad_norm": 3.337953567504883,
2089
+ "learning_rate": 3.436363636363637e-07,
2090
+ "loss": 0.0313,
2091
+ "step": 5820
2092
+ },
2093
+ {
2094
+ "epoch": 4.87,
2095
+ "grad_norm": 2.2369048595428467,
2096
+ "learning_rate": 3.072727272727273e-07,
2097
+ "loss": 0.0403,
2098
+ "step": 5840
2099
+ },
2100
+ {
2101
+ "epoch": 4.88,
2102
+ "grad_norm": Infinity,
2103
+ "learning_rate": 2.7272727272727274e-07,
2104
+ "loss": 0.0428,
2105
+ "step": 5860
2106
+ },
2107
+ {
2108
+ "epoch": 4.9,
2109
+ "grad_norm": 3.3645973205566406,
2110
+ "learning_rate": 2.3636363636363637e-07,
2111
+ "loss": 0.0273,
2112
+ "step": 5880
2113
+ },
2114
+ {
2115
+ "epoch": 4.92,
2116
+ "grad_norm": 3.290832281112671,
2117
+ "learning_rate": 2.0000000000000002e-07,
2118
+ "loss": 0.0297,
2119
+ "step": 5900
2120
+ },
2121
+ {
2122
+ "epoch": 4.93,
2123
+ "grad_norm": 2.2660880088806152,
2124
+ "learning_rate": 1.6363636363636367e-07,
2125
+ "loss": 0.0278,
2126
+ "step": 5920
2127
+ },
2128
+ {
2129
+ "epoch": 4.95,
2130
+ "grad_norm": 3.304938793182373,
2131
+ "learning_rate": 1.272727272727273e-07,
2132
+ "loss": 0.0239,
2133
+ "step": 5940
2134
+ },
2135
+ {
2136
+ "epoch": 4.97,
2137
+ "grad_norm": 4.981977462768555,
2138
+ "learning_rate": 9.090909090909091e-08,
2139
+ "loss": 0.0303,
2140
+ "step": 5960
2141
+ },
2142
+ {
2143
+ "epoch": 4.98,
2144
+ "grad_norm": 5.8069071769714355,
2145
+ "learning_rate": 5.454545454545455e-08,
2146
+ "loss": 0.0354,
2147
+ "step": 5980
2148
+ },
2149
+ {
2150
+ "epoch": 5.0,
2151
+ "grad_norm": 4.917103290557861,
2152
+ "learning_rate": 1.8181818181818185e-08,
2153
+ "loss": 0.0375,
2154
+ "step": 6000
2155
+ },
2156
+ {
2157
+ "epoch": 5.0,
2158
+ "eval_loss": 0.21445466578006744,
2159
+ "eval_runtime": 328.2757,
2160
+ "eval_samples_per_second": 0.305,
2161
+ "eval_steps_per_second": 0.305,
2162
+ "eval_wer": 47.794117647058826,
2163
+ "step": 6000
2164
+ }
2165
+ ],
2166
+ "logging_steps": 20,
2167
+ "max_steps": 6000,
2168
+ "num_input_tokens_seen": 0,
2169
+ "num_train_epochs": 5,
2170
+ "save_steps": 1000,
2171
+ "total_flos": 5.19396003495936e+18,
2172
+ "train_batch_size": 3,
2173
+ "trial_name": null,
2174
+ "trial_params": null
2175
+ }
checkpoint-6000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:152d1e9770ea3e9899a6c5c40e00cad1f52c4e2ae9aea55ed85a6bf2ed0a638f
3
+ size 5048
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4d93661b01ee59de0e59fd7bd26f3b3183cdf01cad9d0ad11b3d71e5c28da35e
3
  size 966995080
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6f9893f0835513890479b8f540c1c3f0e680df780cd2d309f70bfc23a7a76ef
3
  size 966995080
runs/Mar14_08-03-53_841af74821ca/events.out.tfevents.1710403513.841af74821ca.172.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fd3516de94d11a9e246a22629c136b01c655828a34c25780ad3b1c32ad279da
3
+ size 49230
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:982969e542fa24b87a3035f03e0788a450b1a79cb7f7ed10b8083b221fa6713a
3
  size 5048
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:152d1e9770ea3e9899a6c5c40e00cad1f52c4e2ae9aea55ed85a6bf2ed0a638f
3
  size 5048