Hugging Face's logo

thak123
/
whisper-small-LDC-V1

Automatic Speech Recognition
Transformers
PyTorch
whisper
Inference Endpoints
Model card Files Community
thak123 commited on
Commit
bed066c
•
1 Parent(s): 05751fe

Upload 10 files

Browse files
Files changed (10) hide show
  1. config.json +51 -0
  2. generation_config.json +221 -0
  3. optimizer.pt +3 -0
  4. preprocessor_config.json +14 -0
  5. pytorch_model.bin +3 -0
  6. rng_state.pth +3 -0
  7. scaler.pt +3 -0
  8. scheduler.pt +3 -0
  9. trainer_state.json +2854 -0
  10. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 0.2,
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
+ "model_type": "whisper",
41
+ "num_hidden_layers": 12,
42
+ "num_mel_bins": 80,
43
+ "pad_token_id": 50257,
44
+ "scale_embedding": false,
45
+ "suppress_tokens": [],
46
+ "torch_dtype": "float32",
47
+ "transformers_version": "4.29.2",
48
+ "use_cache": true,
49
+ "use_weighted_layer_sum": false,
50
+ "vocab_size": 51865
51
+ }
generation_config.json ADDED
@@ -0,0 +1,221 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "begin_suppress_tokens": [
3
+ 220,
4
+ 50257
5
+ ],
6
+ "bos_token_id": 50257,
7
+ "decoder_start_token_id": 50258,
8
+ "eos_token_id": 50257,
9
+ "forced_decoder_ids": [
10
+ [
11
+ 1,
12
+ null
13
+ ],
14
+ [
15
+ 2,
16
+ 50359
17
+ ]
18
+ ],
19
+ "is_multilingual": true,
20
+ "lang_to_id": {
21
+ "<|af|>": 50327,
22
+ "<|am|>": 50334,
23
+ "<|ar|>": 50272,
24
+ "<|as|>": 50350,
25
+ "<|az|>": 50304,
26
+ "<|ba|>": 50355,
27
+ "<|be|>": 50330,
28
+ "<|bg|>": 50292,
29
+ "<|bn|>": 50302,
30
+ "<|bo|>": 50347,
31
+ "<|br|>": 50309,
32
+ "<|bs|>": 50315,
33
+ "<|ca|>": 50270,
34
+ "<|cs|>": 50283,
35
+ "<|cy|>": 50297,
36
+ "<|da|>": 50285,
37
+ "<|de|>": 50261,
38
+ "<|el|>": 50281,
39
+ "<|en|>": 50259,
40
+ "<|es|>": 50262,
41
+ "<|et|>": 50307,
42
+ "<|eu|>": 50310,
43
+ "<|fa|>": 50300,
44
+ "<|fi|>": 50277,
45
+ "<|fo|>": 50338,
46
+ "<|fr|>": 50265,
47
+ "<|gl|>": 50319,
48
+ "<|gu|>": 50333,
49
+ "<|haw|>": 50352,
50
+ "<|ha|>": 50354,
51
+ "<|he|>": 50279,
52
+ "<|hi|>": 50276,
53
+ "<|hr|>": 50291,
54
+ "<|ht|>": 50339,
55
+ "<|hu|>": 50286,
56
+ "<|hy|>": 50312,
57
+ "<|id|>": 50275,
58
+ "<|is|>": 50311,
59
+ "<|it|>": 50274,
60
+ "<|ja|>": 50266,
61
+ "<|jw|>": 50356,
62
+ "<|ka|>": 50329,
63
+ "<|kk|>": 50316,
64
+ "<|km|>": 50323,
65
+ "<|kn|>": 50306,
66
+ "<|ko|>": 50264,
67
+ "<|la|>": 50294,
68
+ "<|lb|>": 50345,
69
+ "<|ln|>": 50353,
70
+ "<|lo|>": 50336,
71
+ "<|lt|>": 50293,
72
+ "<|lv|>": 50301,
73
+ "<|mg|>": 50349,
74
+ "<|mi|>": 50295,
75
+ "<|mk|>": 50308,
76
+ "<|ml|>": 50296,
77
+ "<|mn|>": 50314,
78
+ "<|mr|>": 50320,
79
+ "<|ms|>": 50282,
80
+ "<|mt|>": 50343,
81
+ "<|my|>": 50346,
82
+ "<|ne|>": 50313,
83
+ "<|nl|>": 50271,
84
+ "<|nn|>": 50342,
85
+ "<|no|>": 50288,
86
+ "<|oc|>": 50328,
87
+ "<|pa|>": 50321,
88
+ "<|pl|>": 50269,
89
+ "<|ps|>": 50340,
90
+ "<|pt|>": 50267,
91
+ "<|ro|>": 50284,
92
+ "<|ru|>": 50263,
93
+ "<|sa|>": 50344,
94
+ "<|sd|>": 50332,
95
+ "<|si|>": 50322,
96
+ "<|sk|>": 50298,
97
+ "<|sl|>": 50305,
98
+ "<|sn|>": 50324,
99
+ "<|so|>": 50326,
100
+ "<|sq|>": 50317,
101
+ "<|sr|>": 50303,
102
+ "<|su|>": 50357,
103
+ "<|sv|>": 50273,
104
+ "<|sw|>": 50318,
105
+ "<|ta|>": 50287,
106
+ "<|te|>": 50299,
107
+ "<|tg|>": 50331,
108
+ "<|th|>": 50289,
109
+ "<|tk|>": 50341,
110
+ "<|tl|>": 50348,
111
+ "<|tr|>": 50268,
112
+ "<|tt|>": 50351,
113
+ "<|uk|>": 50280,
114
+ "<|ur|>": 50290,
115
+ "<|uz|>": 50337,
116
+ "<|vi|>": 50278,
117
+ "<|yi|>": 50335,
118
+ "<|yo|>": 50325,
119
+ "<|zh|>": 50260
120
+ },
121
+ "max_initial_timestamp_index": 1,
122
+ "max_length": 448,
123
+ "no_timestamps_token_id": 50363,
124
+ "pad_token_id": 50257,
125
+ "return_timestamps": false,
126
+ "suppress_tokens": [
127
+ 1,
128
+ 2,
129
+ 7,
130
+ 8,
131
+ 9,
132
+ 10,
133
+ 14,
134
+ 25,
135
+ 26,
136
+ 27,
137
+ 28,
138
+ 29,
139
+ 31,
140
+ 58,
141
+ 59,
142
+ 60,
143
+ 61,
144
+ 62,
145
+ 63,
146
+ 90,
147
+ 91,
148
+ 92,
149
+ 93,
150
+ 359,
151
+ 503,
152
+ 522,
153
+ 542,
154
+ 873,
155
+ 893,
156
+ 902,
157
+ 918,
158
+ 922,
159
+ 931,
160
+ 1350,
161
+ 1853,
162
+ 1982,
163
+ 2460,
164
+ 2627,
165
+ 3246,
166
+ 3253,
167
+ 3268,
168
+ 3536,
169
+ 3846,
170
+ 3961,
171
+ 4183,
172
+ 4667,
173
+ 6585,
174
+ 6647,
175
+ 7273,
176
+ 9061,
177
+ 9383,
178
+ 10428,
179
+ 10929,
180
+ 11938,
181
+ 12033,
182
+ 12331,
183
+ 12562,
184
+ 13793,
185
+ 14157,
186
+ 14635,
187
+ 15265,
188
+ 15618,
189
+ 16553,
190
+ 16604,
191
+ 18362,
192
+ 18956,
193
+ 20075,
194
+ 21675,
195
+ 22520,
196
+ 26130,
197
+ 26161,
198
+ 26435,
199
+ 28279,
200
+ 29464,
201
+ 31650,
202
+ 32302,
203
+ 32470,
204
+ 36865,
205
+ 42863,
206
+ 47425,
207
+ 49870,
208
+ 50254,
209
+ 50258,
210
+ 50358,
211
+ 50359,
212
+ 50360,
213
+ 50361,
214
+ 50362
215
+ ],
216
+ "task_to_id": {
217
+ "transcribe": 50359,
218
+ "translate": 50358
219
+ },
220
+ "transformers_version": "4.29.2"
221
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5065453ce039152a1322301c8d947a7069affc1cee9d3c44f635846b5ca4dd0f
3
+ size 1934161093
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
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7836ee53774de28698e49b6bdfb041fd3e19427b74e97e7cc28fb9c04b9851f
3
+ size 967102601
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3937f93fb3afc447ab36b9817c67f1be65ea8261776575adbff5a2bb5f371a58
3
+ size 14575
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77de831860f375b27ca1d9afc8b68d6ab7213dbeb8c1bb73beb1aaf3f5c714e6
3
+ size 557
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aaa82891a21a8a693a1ce386a3e5d389e91d833a498d5538655c5dd225c52ba9
3
+ size 627
trainer_state.json ADDED
@@ -0,0 +1,2854 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 17.449157344638696,
3
+ "best_model_checkpoint": "./openai/whisper-small-gom-LDC-v1.0/checkpoint-11000",
4
+ "epoch": 2.846790890269151,
5
+ "global_step": 11000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 3.36e-07,
13
+ "loss": 2.8574,
14
+ "step": 25
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 7.359999999999999e-07,
19
+ "loss": 2.3034,
20
+ "step": 50
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 1.1359999999999998e-06,
25
+ "loss": 1.5867,
26
+ "step": 75
27
+ },
28
+ {
29
+ "epoch": 0.03,
30
+ "learning_rate": 1.536e-06,
31
+ "loss": 1.0482,
32
+ "step": 100
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 1.9359999999999998e-06,
37
+ "loss": 0.8649,
38
+ "step": 125
39
+ },
40
+ {
41
+ "epoch": 0.04,
42
+ "learning_rate": 2.3359999999999997e-06,
43
+ "loss": 0.7442,
44
+ "step": 150
45
+ },
46
+ {
47
+ "epoch": 0.05,
48
+ "learning_rate": 2.736e-06,
49
+ "loss": 0.7494,
50
+ "step": 175
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 3.136e-06,
55
+ "loss": 0.6801,
56
+ "step": 200
57
+ },
58
+ {
59
+ "epoch": 0.06,
60
+ "learning_rate": 3.5359999999999997e-06,
61
+ "loss": 0.6158,
62
+ "step": 225
63
+ },
64
+ {
65
+ "epoch": 0.06,
66
+ "learning_rate": 3.936e-06,
67
+ "loss": 0.6206,
68
+ "step": 250
69
+ },
70
+ {
71
+ "epoch": 0.07,
72
+ "learning_rate": 4.3360000000000005e-06,
73
+ "loss": 0.5527,
74
+ "step": 275
75
+ },
76
+ {
77
+ "epoch": 0.08,
78
+ "learning_rate": 4.735999999999999e-06,
79
+ "loss": 0.5373,
80
+ "step": 300
81
+ },
82
+ {
83
+ "epoch": 0.08,
84
+ "learning_rate": 5.136e-06,
85
+ "loss": 0.4616,
86
+ "step": 325
87
+ },
88
+ {
89
+ "epoch": 0.09,
90
+ "learning_rate": 5.535999999999999e-06,
91
+ "loss": 0.4252,
92
+ "step": 350
93
+ },
94
+ {
95
+ "epoch": 0.1,
96
+ "learning_rate": 5.936e-06,
97
+ "loss": 0.3137,
98
+ "step": 375
99
+ },
100
+ {
101
+ "epoch": 0.1,
102
+ "learning_rate": 6.336e-06,
103
+ "loss": 0.2815,
104
+ "step": 400
105
+ },
106
+ {
107
+ "epoch": 0.11,
108
+ "learning_rate": 6.7359999999999995e-06,
109
+ "loss": 0.2738,
110
+ "step": 425
111
+ },
112
+ {
113
+ "epoch": 0.12,
114
+ "learning_rate": 7.136e-06,
115
+ "loss": 0.3058,
116
+ "step": 450
117
+ },
118
+ {
119
+ "epoch": 0.12,
120
+ "learning_rate": 7.5359999999999995e-06,
121
+ "loss": 0.2604,
122
+ "step": 475
123
+ },
124
+ {
125
+ "epoch": 0.13,
126
+ "learning_rate": 7.936e-06,
127
+ "loss": 0.2547,
128
+ "step": 500
129
+ },
130
+ {
131
+ "epoch": 0.13,
132
+ "eval_loss": 0.2495306134223938,
133
+ "eval_runtime": 4903.6311,
134
+ "eval_samples_per_second": 2.224,
135
+ "eval_steps_per_second": 0.278,
136
+ "eval_wer": 61.019581505656184,
137
+ "step": 500
138
+ },
139
+ {
140
+ "epoch": 0.14,
141
+ "learning_rate": 7.984853948791921e-06,
142
+ "loss": 0.2135,
143
+ "step": 525
144
+ },
145
+ {
146
+ "epoch": 0.14,
147
+ "learning_rate": 7.966822935448972e-06,
148
+ "loss": 0.2514,
149
+ "step": 550
150
+ },
151
+ {
152
+ "epoch": 0.15,
153
+ "learning_rate": 7.948791922106022e-06,
154
+ "loss": 0.238,
155
+ "step": 575
156
+ },
157
+ {
158
+ "epoch": 0.16,
159
+ "learning_rate": 7.930760908763071e-06,
160
+ "loss": 0.197,
161
+ "step": 600
162
+ },
163
+ {
164
+ "epoch": 0.16,
165
+ "learning_rate": 7.912729895420122e-06,
166
+ "loss": 0.2452,
167
+ "step": 625
168
+ },
169
+ {
170
+ "epoch": 0.17,
171
+ "learning_rate": 7.894698882077172e-06,
172
+ "loss": 0.2489,
173
+ "step": 650
174
+ },
175
+ {
176
+ "epoch": 0.17,
177
+ "learning_rate": 7.876667868734223e-06,
178
+ "loss": 0.2276,
179
+ "step": 675
180
+ },
181
+ {
182
+ "epoch": 0.18,
183
+ "learning_rate": 7.858636855391273e-06,
184
+ "loss": 0.2024,
185
+ "step": 700
186
+ },
187
+ {
188
+ "epoch": 0.19,
189
+ "learning_rate": 7.840605842048324e-06,
190
+ "loss": 0.2235,
191
+ "step": 725
192
+ },
193
+ {
194
+ "epoch": 0.19,
195
+ "learning_rate": 7.822574828705373e-06,
196
+ "loss": 0.1964,
197
+ "step": 750
198
+ },
199
+ {
200
+ "epoch": 0.2,
201
+ "learning_rate": 7.804543815362423e-06,
202
+ "loss": 0.205,
203
+ "step": 775
204
+ },
205
+ {
206
+ "epoch": 0.21,
207
+ "learning_rate": 7.786512802019472e-06,
208
+ "loss": 0.2044,
209
+ "step": 800
210
+ },
211
+ {
212
+ "epoch": 0.21,
213
+ "learning_rate": 7.768481788676523e-06,
214
+ "loss": 0.1805,
215
+ "step": 825
216
+ },
217
+ {
218
+ "epoch": 0.22,
219
+ "learning_rate": 7.750450775333573e-06,
220
+ "loss": 0.1725,
221
+ "step": 850
222
+ },
223
+ {
224
+ "epoch": 0.23,
225
+ "learning_rate": 7.732419761990624e-06,
226
+ "loss": 0.1888,
227
+ "step": 875
228
+ },
229
+ {
230
+ "epoch": 0.23,
231
+ "learning_rate": 7.714388748647674e-06,
232
+ "loss": 0.1867,
233
+ "step": 900
234
+ },
235
+ {
236
+ "epoch": 0.24,
237
+ "learning_rate": 7.696357735304725e-06,
238
+ "loss": 0.2047,
239
+ "step": 925
240
+ },
241
+ {
242
+ "epoch": 0.25,
243
+ "learning_rate": 7.678326721961773e-06,
244
+ "loss": 0.1637,
245
+ "step": 950
246
+ },
247
+ {
248
+ "epoch": 0.25,
249
+ "learning_rate": 7.660295708618824e-06,
250
+ "loss": 0.1941,
251
+ "step": 975
252
+ },
253
+ {
254
+ "epoch": 0.26,
255
+ "learning_rate": 7.642264695275874e-06,
256
+ "loss": 0.1726,
257
+ "step": 1000
258
+ },
259
+ {
260
+ "epoch": 0.26,
261
+ "eval_loss": 0.18191906809806824,
262
+ "eval_runtime": 5156.4322,
263
+ "eval_samples_per_second": 2.115,
264
+ "eval_steps_per_second": 0.265,
265
+ "eval_wer": 47.49931790039247,
266
+ "step": 1000
267
+ },
268
+ {
269
+ "epoch": 0.27,
270
+ "learning_rate": 7.624233681932924e-06,
271
+ "loss": 0.1846,
272
+ "step": 1025
273
+ },
274
+ {
275
+ "epoch": 0.27,
276
+ "learning_rate": 7.606202668589975e-06,
277
+ "loss": 0.1819,
278
+ "step": 1050
279
+ },
280
+ {
281
+ "epoch": 0.28,
282
+ "learning_rate": 7.588171655247024e-06,
283
+ "loss": 0.1714,
284
+ "step": 1075
285
+ },
286
+ {
287
+ "epoch": 0.28,
288
+ "learning_rate": 7.570140641904075e-06,
289
+ "loss": 0.1706,
290
+ "step": 1100
291
+ },
292
+ {
293
+ "epoch": 0.29,
294
+ "learning_rate": 7.552109628561125e-06,
295
+ "loss": 0.1503,
296
+ "step": 1125
297
+ },
298
+ {
299
+ "epoch": 0.3,
300
+ "learning_rate": 7.534078615218175e-06,
301
+ "loss": 0.1544,
302
+ "step": 1150
303
+ },
304
+ {
305
+ "epoch": 0.3,
306
+ "learning_rate": 7.5160476018752255e-06,
307
+ "loss": 0.1811,
308
+ "step": 1175
309
+ },
310
+ {
311
+ "epoch": 0.31,
312
+ "learning_rate": 7.498016588532275e-06,
313
+ "loss": 0.1715,
314
+ "step": 1200
315
+ },
316
+ {
317
+ "epoch": 0.32,
318
+ "learning_rate": 7.479985575189326e-06,
319
+ "loss": 0.1732,
320
+ "step": 1225
321
+ },
322
+ {
323
+ "epoch": 0.32,
324
+ "learning_rate": 7.461954561846375e-06,
325
+ "loss": 0.1624,
326
+ "step": 1250
327
+ },
328
+ {
329
+ "epoch": 0.33,
330
+ "learning_rate": 7.443923548503425e-06,
331
+ "loss": 0.1469,
332
+ "step": 1275
333
+ },
334
+ {
335
+ "epoch": 0.34,
336
+ "learning_rate": 7.4258925351604755e-06,
337
+ "loss": 0.1764,
338
+ "step": 1300
339
+ },
340
+ {
341
+ "epoch": 0.34,
342
+ "learning_rate": 7.407861521817526e-06,
343
+ "loss": 0.1604,
344
+ "step": 1325
345
+ },
346
+ {
347
+ "epoch": 0.35,
348
+ "learning_rate": 7.389830508474576e-06,
349
+ "loss": 0.1561,
350
+ "step": 1350
351
+ },
352
+ {
353
+ "epoch": 0.36,
354
+ "learning_rate": 7.371799495131626e-06,
355
+ "loss": 0.1661,
356
+ "step": 1375
357
+ },
358
+ {
359
+ "epoch": 0.36,
360
+ "learning_rate": 7.353768481788676e-06,
361
+ "loss": 0.1534,
362
+ "step": 1400
363
+ },
364
+ {
365
+ "epoch": 0.37,
366
+ "learning_rate": 7.335737468445726e-06,
367
+ "loss": 0.1423,
368
+ "step": 1425
369
+ },
370
+ {
371
+ "epoch": 0.38,
372
+ "learning_rate": 7.317706455102777e-06,
373
+ "loss": 0.1806,
374
+ "step": 1450
375
+ },
376
+ {
377
+ "epoch": 0.38,
378
+ "learning_rate": 7.299675441759827e-06,
379
+ "loss": 0.1488,
380
+ "step": 1475
381
+ },
382
+ {
383
+ "epoch": 0.39,
384
+ "learning_rate": 7.281644428416877e-06,
385
+ "loss": 0.1877,
386
+ "step": 1500
387
+ },
388
+ {
389
+ "epoch": 0.39,
390
+ "eval_loss": 0.1481647789478302,
391
+ "eval_runtime": 4672.6288,
392
+ "eval_samples_per_second": 2.334,
393
+ "eval_steps_per_second": 0.292,
394
+ "eval_wer": 40.243876844292394,
395
+ "step": 1500
396
+ },
397
+ {
398
+ "epoch": 0.39,
399
+ "learning_rate": 7.263613415073928e-06,
400
+ "loss": 0.1544,
401
+ "step": 1525
402
+ },
403
+ {
404
+ "epoch": 0.4,
405
+ "learning_rate": 7.2455824017309764e-06,
406
+ "loss": 0.1771,
407
+ "step": 1550
408
+ },
409
+ {
410
+ "epoch": 0.41,
411
+ "learning_rate": 7.227551388388027e-06,
412
+ "loss": 0.141,
413
+ "step": 1575
414
+ },
415
+ {
416
+ "epoch": 0.41,
417
+ "learning_rate": 7.209520375045077e-06,
418
+ "loss": 0.1359,
419
+ "step": 1600
420
+ },
421
+ {
422
+ "epoch": 0.42,
423
+ "learning_rate": 7.191489361702127e-06,
424
+ "loss": 0.1242,
425
+ "step": 1625
426
+ },
427
+ {
428
+ "epoch": 0.43,
429
+ "learning_rate": 7.173458348359178e-06,
430
+ "loss": 0.1281,
431
+ "step": 1650
432
+ },
433
+ {
434
+ "epoch": 0.43,
435
+ "learning_rate": 7.155427335016227e-06,
436
+ "loss": 0.1404,
437
+ "step": 1675
438
+ },
439
+ {
440
+ "epoch": 0.44,
441
+ "learning_rate": 7.137396321673278e-06,
442
+ "loss": 0.1502,
443
+ "step": 1700
444
+ },
445
+ {
446
+ "epoch": 0.45,
447
+ "learning_rate": 7.119365308330328e-06,
448
+ "loss": 0.1348,
449
+ "step": 1725
450
+ },
451
+ {
452
+ "epoch": 0.45,
453
+ "learning_rate": 7.101334294987378e-06,
454
+ "loss": 0.1473,
455
+ "step": 1750
456
+ },
457
+ {
458
+ "epoch": 0.46,
459
+ "learning_rate": 7.0833032816444285e-06,
460
+ "loss": 0.1401,
461
+ "step": 1775
462
+ },
463
+ {
464
+ "epoch": 0.47,
465
+ "learning_rate": 7.065272268301478e-06,
466
+ "loss": 0.1089,
467
+ "step": 1800
468
+ },
469
+ {
470
+ "epoch": 0.47,
471
+ "learning_rate": 7.047241254958528e-06,
472
+ "loss": 0.1416,
473
+ "step": 1825
474
+ },
475
+ {
476
+ "epoch": 0.48,
477
+ "learning_rate": 7.029210241615578e-06,
478
+ "loss": 0.1239,
479
+ "step": 1850
480
+ },
481
+ {
482
+ "epoch": 0.49,
483
+ "learning_rate": 7.011179228272628e-06,
484
+ "loss": 0.133,
485
+ "step": 1875
486
+ },
487
+ {
488
+ "epoch": 0.49,
489
+ "learning_rate": 6.993148214929679e-06,
490
+ "loss": 0.1305,
491
+ "step": 1900
492
+ },
493
+ {
494
+ "epoch": 0.5,
495
+ "learning_rate": 6.975117201586729e-06,
496
+ "loss": 0.1352,
497
+ "step": 1925
498
+ },
499
+ {
500
+ "epoch": 0.5,
501
+ "learning_rate": 6.957086188243779e-06,
502
+ "loss": 0.1567,
503
+ "step": 1950
504
+ },
505
+ {
506
+ "epoch": 0.51,
507
+ "learning_rate": 6.939055174900829e-06,
508
+ "loss": 0.1324,
509
+ "step": 1975
510
+ },
511
+ {
512
+ "epoch": 0.52,
513
+ "learning_rate": 6.921024161557879e-06,
514
+ "loss": 0.1354,
515
+ "step": 2000
516
+ },
517
+ {
518
+ "epoch": 0.52,
519
+ "eval_loss": 0.12738725543022156,
520
+ "eval_runtime": 4615.525,
521
+ "eval_samples_per_second": 2.363,
522
+ "eval_steps_per_second": 0.296,
523
+ "eval_wer": 34.36732637941528,
524
+ "step": 2000
525
+ },
526
+ {
527
+ "epoch": 0.52,
528
+ "learning_rate": 6.9029931482149295e-06,
529
+ "loss": 0.14,
530
+ "step": 2025
531
+ },
532
+ {
533
+ "epoch": 0.53,
534
+ "learning_rate": 6.88496213487198e-06,
535
+ "loss": 0.1316,
536
+ "step": 2050
537
+ },
538
+ {
539
+ "epoch": 0.54,
540
+ "learning_rate": 6.86693112152903e-06,
541
+ "loss": 0.1072,
542
+ "step": 2075
543
+ },
544
+ {
545
+ "epoch": 0.54,
546
+ "learning_rate": 6.84890010818608e-06,
547
+ "loss": 0.108,
548
+ "step": 2100
549
+ },
550
+ {
551
+ "epoch": 0.55,
552
+ "learning_rate": 6.83086909484313e-06,
553
+ "loss": 0.0992,
554
+ "step": 2125
555
+ },
556
+ {
557
+ "epoch": 0.56,
558
+ "learning_rate": 6.8128380815001795e-06,
559
+ "loss": 0.1377,
560
+ "step": 2150
561
+ },
562
+ {
563
+ "epoch": 0.56,
564
+ "learning_rate": 6.79480706815723e-06,
565
+ "loss": 0.1154,
566
+ "step": 2175
567
+ },
568
+ {
569
+ "epoch": 0.57,
570
+ "learning_rate": 6.77677605481428e-06,
571
+ "loss": 0.1272,
572
+ "step": 2200
573
+ },
574
+ {
575
+ "epoch": 0.58,
576
+ "learning_rate": 6.75874504147133e-06,
577
+ "loss": 0.1183,
578
+ "step": 2225
579
+ },
580
+ {
581
+ "epoch": 0.58,
582
+ "learning_rate": 6.740714028128381e-06,
583
+ "loss": 0.1131,
584
+ "step": 2250
585
+ },
586
+ {
587
+ "epoch": 0.59,
588
+ "learning_rate": 6.72268301478543e-06,
589
+ "loss": 0.1208,
590
+ "step": 2275
591
+ },
592
+ {
593
+ "epoch": 0.6,
594
+ "learning_rate": 6.704652001442481e-06,
595
+ "loss": 0.1224,
596
+ "step": 2300
597
+ },
598
+ {
599
+ "epoch": 0.6,
600
+ "learning_rate": 6.686620988099531e-06,
601
+ "loss": 0.1313,
602
+ "step": 2325
603
+ },
604
+ {
605
+ "epoch": 0.61,
606
+ "learning_rate": 6.668589974756581e-06,
607
+ "loss": 0.1284,
608
+ "step": 2350
609
+ },
610
+ {
611
+ "epoch": 0.61,
612
+ "learning_rate": 6.650558961413632e-06,
613
+ "loss": 0.1191,
614
+ "step": 2375
615
+ },
616
+ {
617
+ "epoch": 0.62,
618
+ "learning_rate": 6.6325279480706804e-06,
619
+ "loss": 0.1103,
620
+ "step": 2400
621
+ },
622
+ {
623
+ "epoch": 0.63,
624
+ "learning_rate": 6.614496934727731e-06,
625
+ "loss": 0.1218,
626
+ "step": 2425
627
+ },
628
+ {
629
+ "epoch": 0.63,
630
+ "learning_rate": 6.5964659213847815e-06,
631
+ "loss": 0.1555,
632
+ "step": 2450
633
+ },
634
+ {
635
+ "epoch": 0.64,
636
+ "learning_rate": 6.578434908041831e-06,
637
+ "loss": 0.1221,
638
+ "step": 2475
639
+ },
640
+ {
641
+ "epoch": 0.65,
642
+ "learning_rate": 6.560403894698882e-06,
643
+ "loss": 0.1186,
644
+ "step": 2500
645
+ },
646
+ {
647
+ "epoch": 0.65,
648
+ "eval_loss": 0.11205455660820007,
649
+ "eval_runtime": 4616.7031,
650
+ "eval_samples_per_second": 2.363,
651
+ "eval_steps_per_second": 0.295,
652
+ "eval_wer": 31.03028522257435,
653
+ "step": 2500
654
+ },
655
+ {
656
+ "epoch": 0.65,
657
+ "learning_rate": 6.542372881355932e-06,
658
+ "loss": 0.1156,
659
+ "step": 2525
660
+ },
661
+ {
662
+ "epoch": 0.66,
663
+ "learning_rate": 6.524341868012982e-06,
664
+ "loss": 0.1085,
665
+ "step": 2550
666
+ },
667
+ {
668
+ "epoch": 0.67,
669
+ "learning_rate": 6.506310854670032e-06,
670
+ "loss": 0.1327,
671
+ "step": 2575
672
+ },
673
+ {
674
+ "epoch": 0.67,
675
+ "learning_rate": 6.488279841327083e-06,
676
+ "loss": 0.1383,
677
+ "step": 2600
678
+ },
679
+ {
680
+ "epoch": 0.68,
681
+ "learning_rate": 6.4702488279841325e-06,
682
+ "loss": 0.1128,
683
+ "step": 2625
684
+ },
685
+ {
686
+ "epoch": 0.69,
687
+ "learning_rate": 6.452217814641183e-06,
688
+ "loss": 0.1116,
689
+ "step": 2650
690
+ },
691
+ {
692
+ "epoch": 0.69,
693
+ "learning_rate": 6.434186801298233e-06,
694
+ "loss": 0.1082,
695
+ "step": 2675
696
+ },
697
+ {
698
+ "epoch": 0.7,
699
+ "learning_rate": 6.416155787955282e-06,
700
+ "loss": 0.1235,
701
+ "step": 2700
702
+ },
703
+ {
704
+ "epoch": 0.71,
705
+ "learning_rate": 6.398124774612333e-06,
706
+ "loss": 0.1142,
707
+ "step": 2725
708
+ },
709
+ {
710
+ "epoch": 0.71,
711
+ "learning_rate": 6.3800937612693826e-06,
712
+ "loss": 0.1218,
713
+ "step": 2750
714
+ },
715
+ {
716
+ "epoch": 0.72,
717
+ "learning_rate": 6.362062747926433e-06,
718
+ "loss": 0.1019,
719
+ "step": 2775
720
+ },
721
+ {
722
+ "epoch": 0.72,
723
+ "learning_rate": 6.344031734583484e-06,
724
+ "loss": 0.1024,
725
+ "step": 2800
726
+ },
727
+ {
728
+ "epoch": 0.73,
729
+ "learning_rate": 6.326000721240533e-06,
730
+ "loss": 0.1283,
731
+ "step": 2825
732
+ },
733
+ {
734
+ "epoch": 0.74,
735
+ "learning_rate": 6.307969707897584e-06,
736
+ "loss": 0.0894,
737
+ "step": 2850
738
+ },
739
+ {
740
+ "epoch": 0.74,
741
+ "learning_rate": 6.2899386945546334e-06,
742
+ "loss": 0.1031,
743
+ "step": 2875
744
+ },
745
+ {
746
+ "epoch": 0.75,
747
+ "learning_rate": 6.271907681211684e-06,
748
+ "loss": 0.0935,
749
+ "step": 2900
750
+ },
751
+ {
752
+ "epoch": 0.76,
753
+ "learning_rate": 6.2538766678687345e-06,
754
+ "loss": 0.1256,
755
+ "step": 2925
756
+ },
757
+ {
758
+ "epoch": 0.76,
759
+ "learning_rate": 6.235845654525784e-06,
760
+ "loss": 0.0871,
761
+ "step": 2950
762
+ },
763
+ {
764
+ "epoch": 0.77,
765
+ "learning_rate": 6.217814641182835e-06,
766
+ "loss": 0.099,
767
+ "step": 2975
768
+ },
769
+ {
770
+ "epoch": 0.78,
771
+ "learning_rate": 6.199783627839884e-06,
772
+ "loss": 0.1367,
773
+ "step": 3000
774
+ },
775
+ {
776
+ "epoch": 0.78,
777
+ "eval_loss": 0.10175555944442749,
778
+ "eval_runtime": 4617.2532,
779
+ "eval_samples_per_second": 2.362,
780
+ "eval_steps_per_second": 0.295,
781
+ "eval_wer": 28.322874472684532,
782
+ "step": 3000
783
+ },
784
+ {
785
+ "epoch": 0.78,
786
+ "learning_rate": 6.181752614496934e-06,
787
+ "loss": 0.1023,
788
+ "step": 3025
789
+ },
790
+ {
791
+ "epoch": 0.79,
792
+ "learning_rate": 6.1637216011539845e-06,
793
+ "loss": 0.0861,
794
+ "step": 3050
795
+ },
796
+ {
797
+ "epoch": 0.8,
798
+ "learning_rate": 6.145690587811034e-06,
799
+ "loss": 0.1037,
800
+ "step": 3075
801
+ },
802
+ {
803
+ "epoch": 0.8,
804
+ "learning_rate": 6.127659574468085e-06,
805
+ "loss": 0.098,
806
+ "step": 3100
807
+ },
808
+ {
809
+ "epoch": 0.81,
810
+ "learning_rate": 6.109628561125135e-06,
811
+ "loss": 0.1019,
812
+ "step": 3125
813
+ },
814
+ {
815
+ "epoch": 0.82,
816
+ "learning_rate": 6.091597547782185e-06,
817
+ "loss": 0.0791,
818
+ "step": 3150
819
+ },
820
+ {
821
+ "epoch": 0.82,
822
+ "learning_rate": 6.073566534439235e-06,
823
+ "loss": 0.1013,
824
+ "step": 3175
825
+ },
826
+ {
827
+ "epoch": 0.83,
828
+ "learning_rate": 6.055535521096286e-06,
829
+ "loss": 0.0994,
830
+ "step": 3200
831
+ },
832
+ {
833
+ "epoch": 0.83,
834
+ "learning_rate": 6.037504507753336e-06,
835
+ "loss": 0.0995,
836
+ "step": 3225
837
+ },
838
+ {
839
+ "epoch": 0.84,
840
+ "learning_rate": 6.019473494410386e-06,
841
+ "loss": 0.1041,
842
+ "step": 3250
843
+ },
844
+ {
845
+ "epoch": 0.85,
846
+ "learning_rate": 6.001442481067435e-06,
847
+ "loss": 0.1038,
848
+ "step": 3275
849
+ },
850
+ {
851
+ "epoch": 0.85,
852
+ "learning_rate": 5.9834114677244854e-06,
853
+ "loss": 0.0966,
854
+ "step": 3300
855
+ },
856
+ {
857
+ "epoch": 0.86,
858
+ "learning_rate": 5.965380454381536e-06,
859
+ "loss": 0.0625,
860
+ "step": 3325
861
+ },
862
+ {
863
+ "epoch": 0.87,
864
+ "learning_rate": 5.947349441038586e-06,
865
+ "loss": 0.0919,
866
+ "step": 3350
867
+ },
868
+ {
869
+ "epoch": 0.87,
870
+ "learning_rate": 5.929318427695636e-06,
871
+ "loss": 0.0843,
872
+ "step": 3375
873
+ },
874
+ {
875
+ "epoch": 0.88,
876
+ "learning_rate": 5.911287414352687e-06,
877
+ "loss": 0.0989,
878
+ "step": 3400
879
+ },
880
+ {
881
+ "epoch": 0.89,
882
+ "learning_rate": 5.893256401009736e-06,
883
+ "loss": 0.0968,
884
+ "step": 3425
885
+ },
886
+ {
887
+ "epoch": 0.89,
888
+ "learning_rate": 5.875225387666787e-06,
889
+ "loss": 0.0817,
890
+ "step": 3450
891
+ },
892
+ {
893
+ "epoch": 0.9,
894
+ "learning_rate": 5.8571943743238365e-06,
895
+ "loss": 0.0778,
896
+ "step": 3475
897
+ },
898
+ {
899
+ "epoch": 0.91,
900
+ "learning_rate": 5.839163360980887e-06,
901
+ "loss": 0.1058,
902
+ "step": 3500
903
+ },
904
+ {
905
+ "epoch": 0.91,
906
+ "eval_loss": 0.09466591477394104,
907
+ "eval_runtime": 4606.7225,
908
+ "eval_samples_per_second": 2.368,
909
+ "eval_steps_per_second": 0.296,
910
+ "eval_wer": 27.83805905933217,
911
+ "step": 3500
912
+ },
913
+ {
914
+ "epoch": 0.91,
915
+ "learning_rate": 5.8211323476379375e-06,
916
+ "loss": 0.1071,
917
+ "step": 3525
918
+ },
919
+ {
920
+ "epoch": 0.92,
921
+ "learning_rate": 5.803101334294987e-06,
922
+ "loss": 0.088,
923
+ "step": 3550
924
+ },
925
+ {
926
+ "epoch": 0.93,
927
+ "learning_rate": 5.785070320952037e-06,
928
+ "loss": 0.0747,
929
+ "step": 3575
930
+ },
931
+ {
932
+ "epoch": 0.93,
933
+ "learning_rate": 5.767039307609087e-06,
934
+ "loss": 0.0716,
935
+ "step": 3600
936
+ },
937
+ {
938
+ "epoch": 0.94,
939
+ "learning_rate": 5.749008294266137e-06,
940
+ "loss": 0.0809,
941
+ "step": 3625
942
+ },
943
+ {
944
+ "epoch": 0.94,
945
+ "learning_rate": 5.730977280923188e-06,
946
+ "loss": 0.0824,
947
+ "step": 3650
948
+ },
949
+ {
950
+ "epoch": 0.95,
951
+ "learning_rate": 5.712946267580237e-06,
952
+ "loss": 0.0794,
953
+ "step": 3675
954
+ },
955
+ {
956
+ "epoch": 0.96,
957
+ "learning_rate": 5.694915254237288e-06,
958
+ "loss": 0.0896,
959
+ "step": 3700
960
+ },
961
+ {
962
+ "epoch": 0.96,
963
+ "learning_rate": 5.676884240894338e-06,
964
+ "loss": 0.0888,
965
+ "step": 3725
966
+ },
967
+ {
968
+ "epoch": 0.97,
969
+ "learning_rate": 5.658853227551388e-06,
970
+ "loss": 0.0932,
971
+ "step": 3750
972
+ },
973
+ {
974
+ "epoch": 0.98,
975
+ "learning_rate": 5.6408222142084385e-06,
976
+ "loss": 0.1099,
977
+ "step": 3775
978
+ },
979
+ {
980
+ "epoch": 0.98,
981
+ "learning_rate": 5.622791200865489e-06,
982
+ "loss": 0.0872,
983
+ "step": 3800
984
+ },
985
+ {
986
+ "epoch": 0.99,
987
+ "learning_rate": 5.604760187522539e-06,
988
+ "loss": 0.0971,
989
+ "step": 3825
990
+ },
991
+ {
992
+ "epoch": 1.0,
993
+ "learning_rate": 5.586729174179589e-06,
994
+ "loss": 0.0888,
995
+ "step": 3850
996
+ },
997
+ {
998
+ "epoch": 1.0,
999
+ "learning_rate": 5.568698160836638e-06,
1000
+ "loss": 0.074,
1001
+ "step": 3875
1002
+ },
1003
+ {
1004
+ "epoch": 1.01,
1005
+ "learning_rate": 5.5506671474936885e-06,
1006
+ "loss": 0.0521,
1007
+ "step": 3900
1008
+ },
1009
+ {
1010
+ "epoch": 1.02,
1011
+ "learning_rate": 5.532636134150739e-06,
1012
+ "loss": 0.0707,
1013
+ "step": 3925
1014
+ },
1015
+ {
1016
+ "epoch": 1.02,
1017
+ "learning_rate": 5.514605120807789e-06,
1018
+ "loss": 0.0581,
1019
+ "step": 3950
1020
+ },
1021
+ {
1022
+ "epoch": 1.03,
1023
+ "learning_rate": 5.496574107464839e-06,
1024
+ "loss": 0.0725,
1025
+ "step": 3975
1026
+ },
1027
+ {
1028
+ "epoch": 1.04,
1029
+ "learning_rate": 5.47854309412189e-06,
1030
+ "loss": 0.0807,
1031
+ "step": 4000
1032
+ },
1033
+ {
1034
+ "epoch": 1.04,
1035
+ "eval_loss": 0.08684225380420685,
1036
+ "eval_runtime": 4613.5658,
1037
+ "eval_samples_per_second": 2.364,
1038
+ "eval_steps_per_second": 0.296,
1039
+ "eval_wer": 24.383486893193695,
1040
+ "step": 4000
1041
+ },
1042
+ {
1043
+ "epoch": 1.04,
1044
+ "learning_rate": 5.460512080778939e-06,
1045
+ "loss": 0.0662,
1046
+ "step": 4025
1047
+ },
1048
+ {
1049
+ "epoch": 1.05,
1050
+ "learning_rate": 5.44248106743599e-06,
1051
+ "loss": 0.0821,
1052
+ "step": 4050
1053
+ },
1054
+ {
1055
+ "epoch": 1.05,
1056
+ "learning_rate": 5.4244500540930396e-06,
1057
+ "loss": 0.0688,
1058
+ "step": 4075
1059
+ },
1060
+ {
1061
+ "epoch": 1.06,
1062
+ "learning_rate": 5.40641904075009e-06,
1063
+ "loss": 0.0563,
1064
+ "step": 4100
1065
+ },
1066
+ {
1067
+ "epoch": 1.07,
1068
+ "learning_rate": 5.388388027407141e-06,
1069
+ "loss": 0.0651,
1070
+ "step": 4125
1071
+ },
1072
+ {
1073
+ "epoch": 1.07,
1074
+ "learning_rate": 5.370357014064189e-06,
1075
+ "loss": 0.0619,
1076
+ "step": 4150
1077
+ },
1078
+ {
1079
+ "epoch": 1.08,
1080
+ "learning_rate": 5.35232600072124e-06,
1081
+ "loss": 0.0745,
1082
+ "step": 4175
1083
+ },
1084
+ {
1085
+ "epoch": 1.09,
1086
+ "learning_rate": 5.3342949873782905e-06,
1087
+ "loss": 0.0664,
1088
+ "step": 4200
1089
+ },
1090
+ {
1091
+ "epoch": 1.09,
1092
+ "learning_rate": 5.31626397403534e-06,
1093
+ "loss": 0.098,
1094
+ "step": 4225
1095
+ },
1096
+ {
1097
+ "epoch": 1.1,
1098
+ "learning_rate": 5.298232960692391e-06,
1099
+ "loss": 0.0801,
1100
+ "step": 4250
1101
+ },
1102
+ {
1103
+ "epoch": 1.11,
1104
+ "learning_rate": 5.280201947349441e-06,
1105
+ "loss": 0.0575,
1106
+ "step": 4275
1107
+ },
1108
+ {
1109
+ "epoch": 1.11,
1110
+ "learning_rate": 5.262170934006491e-06,
1111
+ "loss": 0.0558,
1112
+ "step": 4300
1113
+ },
1114
+ {
1115
+ "epoch": 1.12,
1116
+ "learning_rate": 5.244139920663541e-06,
1117
+ "loss": 0.0637,
1118
+ "step": 4325
1119
+ },
1120
+ {
1121
+ "epoch": 1.13,
1122
+ "learning_rate": 5.226108907320591e-06,
1123
+ "loss": 0.0697,
1124
+ "step": 4350
1125
+ },
1126
+ {
1127
+ "epoch": 1.13,
1128
+ "learning_rate": 5.2080778939776415e-06,
1129
+ "loss": 0.0576,
1130
+ "step": 4375
1131
+ },
1132
+ {
1133
+ "epoch": 1.14,
1134
+ "learning_rate": 5.190046880634692e-06,
1135
+ "loss": 0.0714,
1136
+ "step": 4400
1137
+ },
1138
+ {
1139
+ "epoch": 1.15,
1140
+ "learning_rate": 5.172015867291742e-06,
1141
+ "loss": 0.0675,
1142
+ "step": 4425
1143
+ },
1144
+ {
1145
+ "epoch": 1.15,
1146
+ "learning_rate": 5.153984853948791e-06,
1147
+ "loss": 0.0555,
1148
+ "step": 4450
1149
+ },
1150
+ {
1151
+ "epoch": 1.16,
1152
+ "learning_rate": 5.135953840605842e-06,
1153
+ "loss": 0.0568,
1154
+ "step": 4475
1155
+ },
1156
+ {
1157
+ "epoch": 1.16,
1158
+ "learning_rate": 5.1179228272628916e-06,
1159
+ "loss": 0.0698,
1160
+ "step": 4500
1161
+ },
1162
+ {
1163
+ "epoch": 1.16,
1164
+ "eval_loss": 0.08182203769683838,
1165
+ "eval_runtime": 4681.2354,
1166
+ "eval_samples_per_second": 2.33,
1167
+ "eval_steps_per_second": 0.291,
1168
+ "eval_wer": 23.162003903708523,
1169
+ "step": 4500
1170
+ },
1171
+ {
1172
+ "epoch": 1.17,
1173
+ "learning_rate": 5.099891813919942e-06,
1174
+ "loss": 0.06,
1175
+ "step": 4525
1176
+ },
1177
+ {
1178
+ "epoch": 1.18,
1179
+ "learning_rate": 5.081860800576992e-06,
1180
+ "loss": 0.0637,
1181
+ "step": 4550
1182
+ },
1183
+ {
1184
+ "epoch": 1.18,
1185
+ "learning_rate": 5.063829787234042e-06,
1186
+ "loss": 0.0519,
1187
+ "step": 4575
1188
+ },
1189
+ {
1190
+ "epoch": 1.19,
1191
+ "learning_rate": 5.045798773891093e-06,
1192
+ "loss": 0.04,
1193
+ "step": 4600
1194
+ },
1195
+ {
1196
+ "epoch": 1.2,
1197
+ "learning_rate": 5.0277677605481424e-06,
1198
+ "loss": 0.0714,
1199
+ "step": 4625
1200
+ },
1201
+ {
1202
+ "epoch": 1.2,
1203
+ "learning_rate": 5.009736747205193e-06,
1204
+ "loss": 0.049,
1205
+ "step": 4650
1206
+ },
1207
+ {
1208
+ "epoch": 1.21,
1209
+ "learning_rate": 4.9917057338622435e-06,
1210
+ "loss": 0.0687,
1211
+ "step": 4675
1212
+ },
1213
+ {
1214
+ "epoch": 1.22,
1215
+ "learning_rate": 4.973674720519293e-06,
1216
+ "loss": 0.0469,
1217
+ "step": 4700
1218
+ },
1219
+ {
1220
+ "epoch": 1.22,
1221
+ "learning_rate": 4.955643707176344e-06,
1222
+ "loss": 0.073,
1223
+ "step": 4725
1224
+ },
1225
+ {
1226
+ "epoch": 1.23,
1227
+ "learning_rate": 4.9376126938333925e-06,
1228
+ "loss": 0.0583,
1229
+ "step": 4750
1230
+ },
1231
+ {
1232
+ "epoch": 1.24,
1233
+ "learning_rate": 4.919581680490443e-06,
1234
+ "loss": 0.0631,
1235
+ "step": 4775
1236
+ },
1237
+ {
1238
+ "epoch": 1.24,
1239
+ "learning_rate": 4.9015506671474935e-06,
1240
+ "loss": 0.0599,
1241
+ "step": 4800
1242
+ },
1243
+ {
1244
+ "epoch": 1.25,
1245
+ "learning_rate": 4.883519653804543e-06,
1246
+ "loss": 0.0597,
1247
+ "step": 4825
1248
+ },
1249
+ {
1250
+ "epoch": 1.26,
1251
+ "learning_rate": 4.865488640461594e-06,
1252
+ "loss": 0.0697,
1253
+ "step": 4850
1254
+ },
1255
+ {
1256
+ "epoch": 1.26,
1257
+ "learning_rate": 4.847457627118644e-06,
1258
+ "loss": 0.0638,
1259
+ "step": 4875
1260
+ },
1261
+ {
1262
+ "epoch": 1.27,
1263
+ "learning_rate": 4.829426613775694e-06,
1264
+ "loss": 0.0492,
1265
+ "step": 4900
1266
+ },
1267
+ {
1268
+ "epoch": 1.27,
1269
+ "learning_rate": 4.811395600432744e-06,
1270
+ "loss": 0.0576,
1271
+ "step": 4925
1272
+ },
1273
+ {
1274
+ "epoch": 1.28,
1275
+ "learning_rate": 4.793364587089794e-06,
1276
+ "loss": 0.0631,
1277
+ "step": 4950
1278
+ },
1279
+ {
1280
+ "epoch": 1.29,
1281
+ "learning_rate": 4.775333573746845e-06,
1282
+ "loss": 0.0616,
1283
+ "step": 4975
1284
+ },
1285
+ {
1286
+ "epoch": 1.29,
1287
+ "learning_rate": 4.757302560403895e-06,
1288
+ "loss": 0.0434,
1289
+ "step": 5000
1290
+ },
1291
+ {
1292
+ "epoch": 1.29,
1293
+ "eval_loss": 0.07633961737155914,
1294
+ "eval_runtime": 4650.315,
1295
+ "eval_samples_per_second": 2.345,
1296
+ "eval_steps_per_second": 0.293,
1297
+ "eval_wer": 22.19027430898063,
1298
+ "step": 5000
1299
+ },
1300
+ {
1301
+ "epoch": 1.3,
1302
+ "learning_rate": 4.739271547060944e-06,
1303
+ "loss": 0.0745,
1304
+ "step": 5025
1305
+ },
1306
+ {
1307
+ "epoch": 1.31,
1308
+ "learning_rate": 4.7212405337179944e-06,
1309
+ "loss": 0.0392,
1310
+ "step": 5050
1311
+ },
1312
+ {
1313
+ "epoch": 1.31,
1314
+ "learning_rate": 4.703209520375045e-06,
1315
+ "loss": 0.0609,
1316
+ "step": 5075
1317
+ },
1318
+ {
1319
+ "epoch": 1.32,
1320
+ "learning_rate": 4.685178507032095e-06,
1321
+ "loss": 0.0632,
1322
+ "step": 5100
1323
+ },
1324
+ {
1325
+ "epoch": 1.33,
1326
+ "learning_rate": 4.667147493689145e-06,
1327
+ "loss": 0.0586,
1328
+ "step": 5125
1329
+ },
1330
+ {
1331
+ "epoch": 1.33,
1332
+ "learning_rate": 4.649116480346195e-06,
1333
+ "loss": 0.0514,
1334
+ "step": 5150
1335
+ },
1336
+ {
1337
+ "epoch": 1.34,
1338
+ "learning_rate": 4.631085467003245e-06,
1339
+ "loss": 0.0732,
1340
+ "step": 5175
1341
+ },
1342
+ {
1343
+ "epoch": 1.35,
1344
+ "learning_rate": 4.613054453660296e-06,
1345
+ "loss": 0.0731,
1346
+ "step": 5200
1347
+ },
1348
+ {
1349
+ "epoch": 1.35,
1350
+ "learning_rate": 4.5950234403173455e-06,
1351
+ "loss": 0.0773,
1352
+ "step": 5225
1353
+ },
1354
+ {
1355
+ "epoch": 1.36,
1356
+ "learning_rate": 4.576992426974396e-06,
1357
+ "loss": 0.0475,
1358
+ "step": 5250
1359
+ },
1360
+ {
1361
+ "epoch": 1.37,
1362
+ "learning_rate": 4.5589614136314465e-06,
1363
+ "loss": 0.0695,
1364
+ "step": 5275
1365
+ },
1366
+ {
1367
+ "epoch": 1.37,
1368
+ "learning_rate": 4.540930400288496e-06,
1369
+ "loss": 0.0624,
1370
+ "step": 5300
1371
+ },
1372
+ {
1373
+ "epoch": 1.38,
1374
+ "learning_rate": 4.522899386945546e-06,
1375
+ "loss": 0.0516,
1376
+ "step": 5325
1377
+ },
1378
+ {
1379
+ "epoch": 1.38,
1380
+ "learning_rate": 4.5048683736025955e-06,
1381
+ "loss": 0.0582,
1382
+ "step": 5350
1383
+ },
1384
+ {
1385
+ "epoch": 1.39,
1386
+ "learning_rate": 4.486837360259646e-06,
1387
+ "loss": 0.0495,
1388
+ "step": 5375
1389
+ },
1390
+ {
1391
+ "epoch": 1.4,
1392
+ "learning_rate": 4.4688063469166966e-06,
1393
+ "loss": 0.0689,
1394
+ "step": 5400
1395
+ },
1396
+ {
1397
+ "epoch": 1.4,
1398
+ "learning_rate": 4.450775333573746e-06,
1399
+ "loss": 0.0634,
1400
+ "step": 5425
1401
+ },
1402
+ {
1403
+ "epoch": 1.41,
1404
+ "learning_rate": 4.432744320230797e-06,
1405
+ "loss": 0.0379,
1406
+ "step": 5450
1407
+ },
1408
+ {
1409
+ "epoch": 1.42,
1410
+ "learning_rate": 4.414713306887847e-06,
1411
+ "loss": 0.0701,
1412
+ "step": 5475
1413
+ },
1414
+ {
1415
+ "epoch": 1.42,
1416
+ "learning_rate": 4.396682293544897e-06,
1417
+ "loss": 0.0536,
1418
+ "step": 5500
1419
+ },
1420
+ {
1421
+ "epoch": 1.42,
1422
+ "eval_loss": 0.07404134422540665,
1423
+ "eval_runtime": 4631.2871,
1424
+ "eval_samples_per_second": 2.355,
1425
+ "eval_steps_per_second": 0.295,
1426
+ "eval_wer": 22.3413856066489,
1427
+ "step": 5500
1428
+ },
1429
+ {
1430
+ "epoch": 1.43,
1431
+ "learning_rate": 4.3786512802019475e-06,
1432
+ "loss": 0.0549,
1433
+ "step": 5525
1434
+ },
1435
+ {
1436
+ "epoch": 1.44,
1437
+ "learning_rate": 4.360620266858997e-06,
1438
+ "loss": 0.0664,
1439
+ "step": 5550
1440
+ },
1441
+ {
1442
+ "epoch": 1.44,
1443
+ "learning_rate": 4.342589253516048e-06,
1444
+ "loss": 0.0526,
1445
+ "step": 5575
1446
+ },
1447
+ {
1448
+ "epoch": 1.45,
1449
+ "learning_rate": 4.324558240173098e-06,
1450
+ "loss": 0.0581,
1451
+ "step": 5600
1452
+ },
1453
+ {
1454
+ "epoch": 1.46,
1455
+ "learning_rate": 4.306527226830147e-06,
1456
+ "loss": 0.0379,
1457
+ "step": 5625
1458
+ },
1459
+ {
1460
+ "epoch": 1.46,
1461
+ "learning_rate": 4.2884962134871975e-06,
1462
+ "loss": 0.079,
1463
+ "step": 5650
1464
+ },
1465
+ {
1466
+ "epoch": 1.47,
1467
+ "learning_rate": 4.270465200144248e-06,
1468
+ "loss": 0.061,
1469
+ "step": 5675
1470
+ },
1471
+ {
1472
+ "epoch": 1.48,
1473
+ "learning_rate": 4.252434186801298e-06,
1474
+ "loss": 0.0581,
1475
+ "step": 5700
1476
+ },
1477
+ {
1478
+ "epoch": 1.48,
1479
+ "learning_rate": 4.234403173458348e-06,
1480
+ "loss": 0.0695,
1481
+ "step": 5725
1482
+ },
1483
+ {
1484
+ "epoch": 1.49,
1485
+ "learning_rate": 4.216372160115398e-06,
1486
+ "loss": 0.047,
1487
+ "step": 5750
1488
+ },
1489
+ {
1490
+ "epoch": 1.49,
1491
+ "learning_rate": 4.198341146772448e-06,
1492
+ "loss": 0.0497,
1493
+ "step": 5775
1494
+ },
1495
+ {
1496
+ "epoch": 1.5,
1497
+ "learning_rate": 4.180310133429499e-06,
1498
+ "loss": 0.0632,
1499
+ "step": 5800
1500
+ },
1501
+ {
1502
+ "epoch": 1.51,
1503
+ "learning_rate": 4.1622791200865486e-06,
1504
+ "loss": 0.0566,
1505
+ "step": 5825
1506
+ },
1507
+ {
1508
+ "epoch": 1.51,
1509
+ "learning_rate": 4.144248106743599e-06,
1510
+ "loss": 0.0582,
1511
+ "step": 5850
1512
+ },
1513
+ {
1514
+ "epoch": 1.52,
1515
+ "learning_rate": 4.12621709340065e-06,
1516
+ "loss": 0.0624,
1517
+ "step": 5875
1518
+ },
1519
+ {
1520
+ "epoch": 1.53,
1521
+ "learning_rate": 4.108186080057699e-06,
1522
+ "loss": 0.0523,
1523
+ "step": 5900
1524
+ },
1525
+ {
1526
+ "epoch": 1.53,
1527
+ "learning_rate": 4.090155066714749e-06,
1528
+ "loss": 0.0498,
1529
+ "step": 5925
1530
+ },
1531
+ {
1532
+ "epoch": 1.54,
1533
+ "learning_rate": 4.072124053371799e-06,
1534
+ "loss": 0.0561,
1535
+ "step": 5950
1536
+ },
1537
+ {
1538
+ "epoch": 1.55,
1539
+ "learning_rate": 4.054093040028849e-06,
1540
+ "loss": 0.0548,
1541
+ "step": 5975
1542
+ },
1543
+ {
1544
+ "epoch": 1.55,
1545
+ "learning_rate": 4.0360620266859e-06,
1546
+ "loss": 0.0403,
1547
+ "step": 6000
1548
+ },
1549
+ {
1550
+ "epoch": 1.55,
1551
+ "eval_loss": 0.07143695652484894,
1552
+ "eval_runtime": 4576.9308,
1553
+ "eval_samples_per_second": 2.383,
1554
+ "eval_steps_per_second": 0.298,
1555
+ "eval_wer": 20.996075303796673,
1556
+ "step": 6000
1557
+ },
1558
+ {
1559
+ "epoch": 1.56,
1560
+ "learning_rate": 4.018031013342949e-06,
1561
+ "loss": 0.047,
1562
+ "step": 6025
1563
+ },
1564
+ {
1565
+ "epoch": 1.57,
1566
+ "learning_rate": 4e-06,
1567
+ "loss": 0.0445,
1568
+ "step": 6050
1569
+ },
1570
+ {
1571
+ "epoch": 1.57,
1572
+ "learning_rate": 3.9819689866570495e-06,
1573
+ "loss": 0.0459,
1574
+ "step": 6075
1575
+ },
1576
+ {
1577
+ "epoch": 1.58,
1578
+ "learning_rate": 3.9639379733141e-06,
1579
+ "loss": 0.042,
1580
+ "step": 6100
1581
+ },
1582
+ {
1583
+ "epoch": 1.59,
1584
+ "learning_rate": 3.9459069599711505e-06,
1585
+ "loss": 0.0513,
1586
+ "step": 6125
1587
+ },
1588
+ {
1589
+ "epoch": 1.59,
1590
+ "learning_rate": 3.9278759466282e-06,
1591
+ "loss": 0.055,
1592
+ "step": 6150
1593
+ },
1594
+ {
1595
+ "epoch": 1.6,
1596
+ "learning_rate": 3.909844933285251e-06,
1597
+ "loss": 0.0557,
1598
+ "step": 6175
1599
+ },
1600
+ {
1601
+ "epoch": 1.6,
1602
+ "learning_rate": 3.8918139199423e-06,
1603
+ "loss": 0.0646,
1604
+ "step": 6200
1605
+ },
1606
+ {
1607
+ "epoch": 1.61,
1608
+ "learning_rate": 3.873782906599351e-06,
1609
+ "loss": 0.0493,
1610
+ "step": 6225
1611
+ },
1612
+ {
1613
+ "epoch": 1.62,
1614
+ "learning_rate": 3.8557518932564006e-06,
1615
+ "loss": 0.0523,
1616
+ "step": 6250
1617
+ },
1618
+ {
1619
+ "epoch": 1.62,
1620
+ "learning_rate": 3.837720879913451e-06,
1621
+ "loss": 0.0515,
1622
+ "step": 6275
1623
+ },
1624
+ {
1625
+ "epoch": 1.63,
1626
+ "learning_rate": 3.819689866570502e-06,
1627
+ "loss": 0.054,
1628
+ "step": 6300
1629
+ },
1630
+ {
1631
+ "epoch": 1.64,
1632
+ "learning_rate": 3.8016588532275513e-06,
1633
+ "loss": 0.0643,
1634
+ "step": 6325
1635
+ },
1636
+ {
1637
+ "epoch": 1.64,
1638
+ "learning_rate": 3.7836278398846013e-06,
1639
+ "loss": 0.0699,
1640
+ "step": 6350
1641
+ },
1642
+ {
1643
+ "epoch": 1.65,
1644
+ "learning_rate": 3.7655968265416514e-06,
1645
+ "loss": 0.0474,
1646
+ "step": 6375
1647
+ },
1648
+ {
1649
+ "epoch": 1.66,
1650
+ "learning_rate": 3.747565813198702e-06,
1651
+ "loss": 0.0359,
1652
+ "step": 6400
1653
+ },
1654
+ {
1655
+ "epoch": 1.66,
1656
+ "learning_rate": 3.7295347998557516e-06,
1657
+ "loss": 0.0587,
1658
+ "step": 6425
1659
+ },
1660
+ {
1661
+ "epoch": 1.67,
1662
+ "learning_rate": 3.7115037865128017e-06,
1663
+ "loss": 0.0615,
1664
+ "step": 6450
1665
+ },
1666
+ {
1667
+ "epoch": 1.68,
1668
+ "learning_rate": 3.693472773169852e-06,
1669
+ "loss": 0.0482,
1670
+ "step": 6475
1671
+ },
1672
+ {
1673
+ "epoch": 1.68,
1674
+ "learning_rate": 3.6754417598269023e-06,
1675
+ "loss": 0.0526,
1676
+ "step": 6500
1677
+ },
1678
+ {
1679
+ "epoch": 1.68,
1680
+ "eval_loss": 0.06842584162950516,
1681
+ "eval_runtime": 4579.0646,
1682
+ "eval_samples_per_second": 2.382,
1683
+ "eval_steps_per_second": 0.298,
1684
+ "eval_wer": 20.52385249858333,
1685
+ "step": 6500
1686
+ },
1687
+ {
1688
+ "epoch": 1.69,
1689
+ "learning_rate": 3.6574107464839524e-06,
1690
+ "loss": 0.0636,
1691
+ "step": 6525
1692
+ },
1693
+ {
1694
+ "epoch": 1.7,
1695
+ "learning_rate": 3.6393797331410025e-06,
1696
+ "loss": 0.0671,
1697
+ "step": 6550
1698
+ },
1699
+ {
1700
+ "epoch": 1.7,
1701
+ "learning_rate": 3.621348719798052e-06,
1702
+ "loss": 0.0443,
1703
+ "step": 6575
1704
+ },
1705
+ {
1706
+ "epoch": 1.71,
1707
+ "learning_rate": 3.6033177064551027e-06,
1708
+ "loss": 0.0484,
1709
+ "step": 6600
1710
+ },
1711
+ {
1712
+ "epoch": 1.71,
1713
+ "learning_rate": 3.5852866931121528e-06,
1714
+ "loss": 0.0734,
1715
+ "step": 6625
1716
+ },
1717
+ {
1718
+ "epoch": 1.72,
1719
+ "learning_rate": 3.567255679769203e-06,
1720
+ "loss": 0.0705,
1721
+ "step": 6650
1722
+ },
1723
+ {
1724
+ "epoch": 1.73,
1725
+ "learning_rate": 3.549224666426253e-06,
1726
+ "loss": 0.0559,
1727
+ "step": 6675
1728
+ },
1729
+ {
1730
+ "epoch": 1.73,
1731
+ "learning_rate": 3.5311936530833035e-06,
1732
+ "loss": 0.0569,
1733
+ "step": 6700
1734
+ },
1735
+ {
1736
+ "epoch": 1.74,
1737
+ "learning_rate": 3.513162639740353e-06,
1738
+ "loss": 0.0457,
1739
+ "step": 6725
1740
+ },
1741
+ {
1742
+ "epoch": 1.75,
1743
+ "learning_rate": 3.4951316263974032e-06,
1744
+ "loss": 0.0558,
1745
+ "step": 6750
1746
+ },
1747
+ {
1748
+ "epoch": 1.75,
1749
+ "learning_rate": 3.4771006130544533e-06,
1750
+ "loss": 0.0449,
1751
+ "step": 6775
1752
+ },
1753
+ {
1754
+ "epoch": 1.76,
1755
+ "learning_rate": 3.459069599711504e-06,
1756
+ "loss": 0.0493,
1757
+ "step": 6800
1758
+ },
1759
+ {
1760
+ "epoch": 1.77,
1761
+ "learning_rate": 3.441038586368554e-06,
1762
+ "loss": 0.0608,
1763
+ "step": 6825
1764
+ },
1765
+ {
1766
+ "epoch": 1.77,
1767
+ "learning_rate": 3.423007573025604e-06,
1768
+ "loss": 0.0444,
1769
+ "step": 6850
1770
+ },
1771
+ {
1772
+ "epoch": 1.78,
1773
+ "learning_rate": 3.4049765596826537e-06,
1774
+ "loss": 0.0569,
1775
+ "step": 6875
1776
+ },
1777
+ {
1778
+ "epoch": 1.79,
1779
+ "learning_rate": 3.3869455463397042e-06,
1780
+ "loss": 0.0657,
1781
+ "step": 6900
1782
+ },
1783
+ {
1784
+ "epoch": 1.79,
1785
+ "learning_rate": 3.3689145329967543e-06,
1786
+ "loss": 0.0662,
1787
+ "step": 6925
1788
+ },
1789
+ {
1790
+ "epoch": 1.8,
1791
+ "learning_rate": 3.3508835196538044e-06,
1792
+ "loss": 0.0561,
1793
+ "step": 6950
1794
+ },
1795
+ {
1796
+ "epoch": 1.81,
1797
+ "learning_rate": 3.3328525063108545e-06,
1798
+ "loss": 0.0518,
1799
+ "step": 6975
1800
+ },
1801
+ {
1802
+ "epoch": 1.81,
1803
+ "learning_rate": 3.314821492967905e-06,
1804
+ "loss": 0.0505,
1805
+ "step": 7000
1806
+ },
1807
+ {
1808
+ "epoch": 1.81,
1809
+ "eval_loss": 0.06526884436607361,
1810
+ "eval_runtime": 4621.9984,
1811
+ "eval_samples_per_second": 2.36,
1812
+ "eval_steps_per_second": 0.295,
1813
+ "eval_wer": 19.052616114340882,
1814
+ "step": 7000
1815
+ },
1816
+ {
1817
+ "epoch": 1.82,
1818
+ "learning_rate": 3.2967904796249547e-06,
1819
+ "loss": 0.0534,
1820
+ "step": 7025
1821
+ },
1822
+ {
1823
+ "epoch": 1.82,
1824
+ "learning_rate": 3.2787594662820048e-06,
1825
+ "loss": 0.0494,
1826
+ "step": 7050
1827
+ },
1828
+ {
1829
+ "epoch": 1.83,
1830
+ "learning_rate": 3.260728452939055e-06,
1831
+ "loss": 0.036,
1832
+ "step": 7075
1833
+ },
1834
+ {
1835
+ "epoch": 1.84,
1836
+ "learning_rate": 3.2426974395961054e-06,
1837
+ "loss": 0.0541,
1838
+ "step": 7100
1839
+ },
1840
+ {
1841
+ "epoch": 1.84,
1842
+ "learning_rate": 3.2246664262531555e-06,
1843
+ "loss": 0.0473,
1844
+ "step": 7125
1845
+ },
1846
+ {
1847
+ "epoch": 1.85,
1848
+ "learning_rate": 3.206635412910205e-06,
1849
+ "loss": 0.0508,
1850
+ "step": 7150
1851
+ },
1852
+ {
1853
+ "epoch": 1.86,
1854
+ "learning_rate": 3.1886043995672552e-06,
1855
+ "loss": 0.0481,
1856
+ "step": 7175
1857
+ },
1858
+ {
1859
+ "epoch": 1.86,
1860
+ "learning_rate": 3.1705733862243058e-06,
1861
+ "loss": 0.057,
1862
+ "step": 7200
1863
+ },
1864
+ {
1865
+ "epoch": 1.87,
1866
+ "learning_rate": 3.152542372881356e-06,
1867
+ "loss": 0.0283,
1868
+ "step": 7225
1869
+ },
1870
+ {
1871
+ "epoch": 1.88,
1872
+ "learning_rate": 3.134511359538406e-06,
1873
+ "loss": 0.0436,
1874
+ "step": 7250
1875
+ },
1876
+ {
1877
+ "epoch": 1.88,
1878
+ "learning_rate": 3.116480346195456e-06,
1879
+ "loss": 0.0687,
1880
+ "step": 7275
1881
+ },
1882
+ {
1883
+ "epoch": 1.89,
1884
+ "learning_rate": 3.098449332852506e-06,
1885
+ "loss": 0.0441,
1886
+ "step": 7300
1887
+ },
1888
+ {
1889
+ "epoch": 1.9,
1890
+ "learning_rate": 3.0804183195095562e-06,
1891
+ "loss": 0.0504,
1892
+ "step": 7325
1893
+ },
1894
+ {
1895
+ "epoch": 1.9,
1896
+ "learning_rate": 3.0623873061666063e-06,
1897
+ "loss": 0.0384,
1898
+ "step": 7350
1899
+ },
1900
+ {
1901
+ "epoch": 1.91,
1902
+ "learning_rate": 3.0443562928236564e-06,
1903
+ "loss": 0.04,
1904
+ "step": 7375
1905
+ },
1906
+ {
1907
+ "epoch": 1.92,
1908
+ "learning_rate": 3.026325279480707e-06,
1909
+ "loss": 0.0482,
1910
+ "step": 7400
1911
+ },
1912
+ {
1913
+ "epoch": 1.92,
1914
+ "learning_rate": 3.008294266137757e-06,
1915
+ "loss": 0.0553,
1916
+ "step": 7425
1917
+ },
1918
+ {
1919
+ "epoch": 1.93,
1920
+ "learning_rate": 2.9902632527948067e-06,
1921
+ "loss": 0.0712,
1922
+ "step": 7450
1923
+ },
1924
+ {
1925
+ "epoch": 1.93,
1926
+ "learning_rate": 2.9722322394518568e-06,
1927
+ "loss": 0.0648,
1928
+ "step": 7475
1929
+ },
1930
+ {
1931
+ "epoch": 1.94,
1932
+ "learning_rate": 2.9542012261089073e-06,
1933
+ "loss": 0.053,
1934
+ "step": 7500
1935
+ },
1936
+ {
1937
+ "epoch": 1.94,
1938
+ "eval_loss": 0.06240825355052948,
1939
+ "eval_runtime": 4653.2425,
1940
+ "eval_samples_per_second": 2.344,
1941
+ "eval_steps_per_second": 0.293,
1942
+ "eval_wer": 18.697924318425084,
1943
+ "step": 7500
1944
+ },
1945
+ {
1946
+ "epoch": 1.95,
1947
+ "learning_rate": 2.9361702127659574e-06,
1948
+ "loss": 0.0613,
1949
+ "step": 7525
1950
+ },
1951
+ {
1952
+ "epoch": 1.95,
1953
+ "learning_rate": 2.9181391994230075e-06,
1954
+ "loss": 0.0492,
1955
+ "step": 7550
1956
+ },
1957
+ {
1958
+ "epoch": 1.96,
1959
+ "learning_rate": 2.9001081860800576e-06,
1960
+ "loss": 0.0372,
1961
+ "step": 7575
1962
+ },
1963
+ {
1964
+ "epoch": 1.97,
1965
+ "learning_rate": 2.8820771727371077e-06,
1966
+ "loss": 0.0474,
1967
+ "step": 7600
1968
+ },
1969
+ {
1970
+ "epoch": 1.97,
1971
+ "learning_rate": 2.8640461593941577e-06,
1972
+ "loss": 0.0472,
1973
+ "step": 7625
1974
+ },
1975
+ {
1976
+ "epoch": 1.98,
1977
+ "learning_rate": 2.846015146051208e-06,
1978
+ "loss": 0.0456,
1979
+ "step": 7650
1980
+ },
1981
+ {
1982
+ "epoch": 1.99,
1983
+ "learning_rate": 2.827984132708258e-06,
1984
+ "loss": 0.0496,
1985
+ "step": 7675
1986
+ },
1987
+ {
1988
+ "epoch": 1.99,
1989
+ "learning_rate": 2.8099531193653084e-06,
1990
+ "loss": 0.0517,
1991
+ "step": 7700
1992
+ },
1993
+ {
1994
+ "epoch": 2.0,
1995
+ "learning_rate": 2.7919221060223585e-06,
1996
+ "loss": 0.048,
1997
+ "step": 7725
1998
+ },
1999
+ {
2000
+ "epoch": 2.01,
2001
+ "learning_rate": 2.773891092679408e-06,
2002
+ "loss": 0.0342,
2003
+ "step": 7750
2004
+ },
2005
+ {
2006
+ "epoch": 2.01,
2007
+ "learning_rate": 2.7558600793364583e-06,
2008
+ "loss": 0.0259,
2009
+ "step": 7775
2010
+ },
2011
+ {
2012
+ "epoch": 2.02,
2013
+ "learning_rate": 2.737829065993509e-06,
2014
+ "loss": 0.0323,
2015
+ "step": 7800
2016
+ },
2017
+ {
2018
+ "epoch": 2.03,
2019
+ "learning_rate": 2.719798052650559e-06,
2020
+ "loss": 0.0338,
2021
+ "step": 7825
2022
+ },
2023
+ {
2024
+ "epoch": 2.03,
2025
+ "learning_rate": 2.701767039307609e-06,
2026
+ "loss": 0.0356,
2027
+ "step": 7850
2028
+ },
2029
+ {
2030
+ "epoch": 2.04,
2031
+ "learning_rate": 2.6837360259646595e-06,
2032
+ "loss": 0.0241,
2033
+ "step": 7875
2034
+ },
2035
+ {
2036
+ "epoch": 2.04,
2037
+ "learning_rate": 2.665705012621709e-06,
2038
+ "loss": 0.0359,
2039
+ "step": 7900
2040
+ },
2041
+ {
2042
+ "epoch": 2.05,
2043
+ "learning_rate": 2.6476739992787593e-06,
2044
+ "loss": 0.0336,
2045
+ "step": 7925
2046
+ },
2047
+ {
2048
+ "epoch": 2.06,
2049
+ "learning_rate": 2.6296429859358094e-06,
2050
+ "loss": 0.0395,
2051
+ "step": 7950
2052
+ },
2053
+ {
2054
+ "epoch": 2.06,
2055
+ "learning_rate": 2.61161197259286e-06,
2056
+ "loss": 0.0319,
2057
+ "step": 7975
2058
+ },
2059
+ {
2060
+ "epoch": 2.07,
2061
+ "learning_rate": 2.59358095924991e-06,
2062
+ "loss": 0.0234,
2063
+ "step": 8000
2064
+ },
2065
+ {
2066
+ "epoch": 2.07,
2067
+ "eval_loss": 0.0625641718506813,
2068
+ "eval_runtime": 4589.637,
2069
+ "eval_samples_per_second": 2.376,
2070
+ "eval_steps_per_second": 0.297,
2071
+ "eval_wer": 18.46076353180683,
2072
+ "step": 8000
2073
+ },
2074
+ {
2075
+ "epoch": 2.08,
2076
+ "learning_rate": 2.5755499459069596e-06,
2077
+ "loss": 0.0411,
2078
+ "step": 8025
2079
+ },
2080
+ {
2081
+ "epoch": 2.08,
2082
+ "learning_rate": 2.5575189325640097e-06,
2083
+ "loss": 0.0183,
2084
+ "step": 8050
2085
+ },
2086
+ {
2087
+ "epoch": 2.09,
2088
+ "learning_rate": 2.5394879192210603e-06,
2089
+ "loss": 0.0457,
2090
+ "step": 8075
2091
+ },
2092
+ {
2093
+ "epoch": 2.1,
2094
+ "learning_rate": 2.5214569058781103e-06,
2095
+ "loss": 0.0464,
2096
+ "step": 8100
2097
+ },
2098
+ {
2099
+ "epoch": 2.1,
2100
+ "learning_rate": 2.5034258925351604e-06,
2101
+ "loss": 0.0334,
2102
+ "step": 8125
2103
+ },
2104
+ {
2105
+ "epoch": 2.11,
2106
+ "learning_rate": 2.4853948791922105e-06,
2107
+ "loss": 0.0239,
2108
+ "step": 8150
2109
+ },
2110
+ {
2111
+ "epoch": 2.12,
2112
+ "learning_rate": 2.4673638658492606e-06,
2113
+ "loss": 0.0329,
2114
+ "step": 8175
2115
+ },
2116
+ {
2117
+ "epoch": 2.12,
2118
+ "learning_rate": 2.4493328525063107e-06,
2119
+ "loss": 0.0261,
2120
+ "step": 8200
2121
+ },
2122
+ {
2123
+ "epoch": 2.13,
2124
+ "learning_rate": 2.431301839163361e-06,
2125
+ "loss": 0.0294,
2126
+ "step": 8225
2127
+ },
2128
+ {
2129
+ "epoch": 2.14,
2130
+ "learning_rate": 2.413270825820411e-06,
2131
+ "loss": 0.0277,
2132
+ "step": 8250
2133
+ },
2134
+ {
2135
+ "epoch": 2.14,
2136
+ "learning_rate": 2.3952398124774614e-06,
2137
+ "loss": 0.037,
2138
+ "step": 8275
2139
+ },
2140
+ {
2141
+ "epoch": 2.15,
2142
+ "learning_rate": 2.3772087991345115e-06,
2143
+ "loss": 0.0258,
2144
+ "step": 8300
2145
+ },
2146
+ {
2147
+ "epoch": 2.15,
2148
+ "learning_rate": 2.359177785791561e-06,
2149
+ "loss": 0.0354,
2150
+ "step": 8325
2151
+ },
2152
+ {
2153
+ "epoch": 2.16,
2154
+ "learning_rate": 2.3411467724486113e-06,
2155
+ "loss": 0.0234,
2156
+ "step": 8350
2157
+ },
2158
+ {
2159
+ "epoch": 2.17,
2160
+ "learning_rate": 2.3231157591056618e-06,
2161
+ "loss": 0.0316,
2162
+ "step": 8375
2163
+ },
2164
+ {
2165
+ "epoch": 2.17,
2166
+ "learning_rate": 2.305084745762712e-06,
2167
+ "loss": 0.0358,
2168
+ "step": 8400
2169
+ },
2170
+ {
2171
+ "epoch": 2.18,
2172
+ "learning_rate": 2.287053732419762e-06,
2173
+ "loss": 0.0226,
2174
+ "step": 8425
2175
+ },
2176
+ {
2177
+ "epoch": 2.19,
2178
+ "learning_rate": 2.269022719076812e-06,
2179
+ "loss": 0.0265,
2180
+ "step": 8450
2181
+ },
2182
+ {
2183
+ "epoch": 2.19,
2184
+ "learning_rate": 2.25171294626758e-06,
2185
+ "loss": 0.0345,
2186
+ "step": 8475
2187
+ },
2188
+ {
2189
+ "epoch": 2.2,
2190
+ "learning_rate": 2.23368193292463e-06,
2191
+ "loss": 0.0268,
2192
+ "step": 8500
2193
+ },
2194
+ {
2195
+ "epoch": 2.2,
2196
+ "eval_loss": 0.0614020936191082,
2197
+ "eval_runtime": 4575.3897,
2198
+ "eval_samples_per_second": 2.384,
2199
+ "eval_steps_per_second": 0.298,
2200
+ "eval_wer": 19.62138224862006,
2201
+ "step": 8500
2202
+ },
2203
+ {
2204
+ "epoch": 2.21,
2205
+ "learning_rate": 2.2156509195816806e-06,
2206
+ "loss": 0.0329,
2207
+ "step": 8525
2208
+ },
2209
+ {
2210
+ "epoch": 2.21,
2211
+ "learning_rate": 2.1976199062387307e-06,
2212
+ "loss": 0.0471,
2213
+ "step": 8550
2214
+ },
2215
+ {
2216
+ "epoch": 2.22,
2217
+ "learning_rate": 2.1795888928957808e-06,
2218
+ "loss": 0.0366,
2219
+ "step": 8575
2220
+ },
2221
+ {
2222
+ "epoch": 2.23,
2223
+ "learning_rate": 2.161557879552831e-06,
2224
+ "loss": 0.0247,
2225
+ "step": 8600
2226
+ },
2227
+ {
2228
+ "epoch": 2.23,
2229
+ "learning_rate": 2.143526866209881e-06,
2230
+ "loss": 0.0257,
2231
+ "step": 8625
2232
+ },
2233
+ {
2234
+ "epoch": 2.24,
2235
+ "learning_rate": 2.125495852866931e-06,
2236
+ "loss": 0.0532,
2237
+ "step": 8650
2238
+ },
2239
+ {
2240
+ "epoch": 2.25,
2241
+ "learning_rate": 2.107464839523981e-06,
2242
+ "loss": 0.0343,
2243
+ "step": 8675
2244
+ },
2245
+ {
2246
+ "epoch": 2.25,
2247
+ "learning_rate": 2.0894338261810312e-06,
2248
+ "loss": 0.027,
2249
+ "step": 8700
2250
+ },
2251
+ {
2252
+ "epoch": 2.26,
2253
+ "learning_rate": 2.0714028128380817e-06,
2254
+ "loss": 0.031,
2255
+ "step": 8725
2256
+ },
2257
+ {
2258
+ "epoch": 2.26,
2259
+ "learning_rate": 2.053371799495132e-06,
2260
+ "loss": 0.0467,
2261
+ "step": 8750
2262
+ },
2263
+ {
2264
+ "epoch": 2.27,
2265
+ "learning_rate": 2.0353407861521815e-06,
2266
+ "loss": 0.029,
2267
+ "step": 8775
2268
+ },
2269
+ {
2270
+ "epoch": 2.28,
2271
+ "learning_rate": 2.0173097728092316e-06,
2272
+ "loss": 0.0422,
2273
+ "step": 8800
2274
+ },
2275
+ {
2276
+ "epoch": 2.28,
2277
+ "learning_rate": 1.9992787594662817e-06,
2278
+ "loss": 0.0328,
2279
+ "step": 8825
2280
+ },
2281
+ {
2282
+ "epoch": 2.29,
2283
+ "learning_rate": 1.981247746123332e-06,
2284
+ "loss": 0.0363,
2285
+ "step": 8850
2286
+ },
2287
+ {
2288
+ "epoch": 2.3,
2289
+ "learning_rate": 1.9632167327803823e-06,
2290
+ "loss": 0.031,
2291
+ "step": 8875
2292
+ },
2293
+ {
2294
+ "epoch": 2.3,
2295
+ "learning_rate": 1.9451857194374324e-06,
2296
+ "loss": 0.0302,
2297
+ "step": 8900
2298
+ },
2299
+ {
2300
+ "epoch": 2.31,
2301
+ "learning_rate": 1.9271547060944825e-06,
2302
+ "loss": 0.0248,
2303
+ "step": 8925
2304
+ },
2305
+ {
2306
+ "epoch": 2.32,
2307
+ "learning_rate": 1.9091236927515326e-06,
2308
+ "loss": 0.0307,
2309
+ "step": 8950
2310
+ },
2311
+ {
2312
+ "epoch": 2.32,
2313
+ "learning_rate": 1.8910926794085827e-06,
2314
+ "loss": 0.0344,
2315
+ "step": 8975
2316
+ },
2317
+ {
2318
+ "epoch": 2.33,
2319
+ "learning_rate": 1.873061666065633e-06,
2320
+ "loss": 0.0403,
2321
+ "step": 9000
2322
+ },
2323
+ {
2324
+ "epoch": 2.33,
2325
+ "eval_loss": 0.05997191742062569,
2326
+ "eval_runtime": 4588.7385,
2327
+ "eval_samples_per_second": 2.377,
2328
+ "eval_steps_per_second": 0.297,
2329
+ "eval_wer": 17.98014565450081,
2330
+ "step": 9000
2331
+ },
2332
+ {
2333
+ "epoch": 2.34,
2334
+ "learning_rate": 1.8550306527226828e-06,
2335
+ "loss": 0.0326,
2336
+ "step": 9025
2337
+ },
2338
+ {
2339
+ "epoch": 2.34,
2340
+ "learning_rate": 1.8369996393797331e-06,
2341
+ "loss": 0.0307,
2342
+ "step": 9050
2343
+ },
2344
+ {
2345
+ "epoch": 2.35,
2346
+ "learning_rate": 1.8189686260367832e-06,
2347
+ "loss": 0.0396,
2348
+ "step": 9075
2349
+ },
2350
+ {
2351
+ "epoch": 2.36,
2352
+ "learning_rate": 1.8009376126938333e-06,
2353
+ "loss": 0.044,
2354
+ "step": 9100
2355
+ },
2356
+ {
2357
+ "epoch": 2.36,
2358
+ "learning_rate": 1.7829065993508834e-06,
2359
+ "loss": 0.0292,
2360
+ "step": 9125
2361
+ },
2362
+ {
2363
+ "epoch": 2.37,
2364
+ "learning_rate": 1.7648755860079337e-06,
2365
+ "loss": 0.0436,
2366
+ "step": 9150
2367
+ },
2368
+ {
2369
+ "epoch": 2.37,
2370
+ "learning_rate": 1.7468445726649836e-06,
2371
+ "loss": 0.0318,
2372
+ "step": 9175
2373
+ },
2374
+ {
2375
+ "epoch": 2.38,
2376
+ "learning_rate": 1.728813559322034e-06,
2377
+ "loss": 0.0347,
2378
+ "step": 9200
2379
+ },
2380
+ {
2381
+ "epoch": 2.39,
2382
+ "learning_rate": 1.7107825459790838e-06,
2383
+ "loss": 0.0169,
2384
+ "step": 9225
2385
+ },
2386
+ {
2387
+ "epoch": 2.39,
2388
+ "learning_rate": 1.692751532636134e-06,
2389
+ "loss": 0.0287,
2390
+ "step": 9250
2391
+ },
2392
+ {
2393
+ "epoch": 2.4,
2394
+ "learning_rate": 1.6747205192931842e-06,
2395
+ "loss": 0.0552,
2396
+ "step": 9275
2397
+ },
2398
+ {
2399
+ "epoch": 2.41,
2400
+ "learning_rate": 1.6566895059502343e-06,
2401
+ "loss": 0.0326,
2402
+ "step": 9300
2403
+ },
2404
+ {
2405
+ "epoch": 2.41,
2406
+ "learning_rate": 1.6386584926072844e-06,
2407
+ "loss": 0.0239,
2408
+ "step": 9325
2409
+ },
2410
+ {
2411
+ "epoch": 2.42,
2412
+ "learning_rate": 1.6206274792643347e-06,
2413
+ "loss": 0.0329,
2414
+ "step": 9350
2415
+ },
2416
+ {
2417
+ "epoch": 2.43,
2418
+ "learning_rate": 1.6025964659213846e-06,
2419
+ "loss": 0.0286,
2420
+ "step": 9375
2421
+ },
2422
+ {
2423
+ "epoch": 2.43,
2424
+ "learning_rate": 1.5845654525784349e-06,
2425
+ "loss": 0.0298,
2426
+ "step": 9400
2427
+ },
2428
+ {
2429
+ "epoch": 2.44,
2430
+ "learning_rate": 1.566534439235485e-06,
2431
+ "loss": 0.0391,
2432
+ "step": 9425
2433
+ },
2434
+ {
2435
+ "epoch": 2.45,
2436
+ "learning_rate": 1.548503425892535e-06,
2437
+ "loss": 0.0173,
2438
+ "step": 9450
2439
+ },
2440
+ {
2441
+ "epoch": 2.45,
2442
+ "learning_rate": 1.5304724125495851e-06,
2443
+ "loss": 0.0329,
2444
+ "step": 9475
2445
+ },
2446
+ {
2447
+ "epoch": 2.46,
2448
+ "learning_rate": 1.5124413992066354e-06,
2449
+ "loss": 0.0419,
2450
+ "step": 9500
2451
+ },
2452
+ {
2453
+ "epoch": 2.46,
2454
+ "eval_loss": 0.05825132131576538,
2455
+ "eval_runtime": 4572.6194,
2456
+ "eval_samples_per_second": 2.385,
2457
+ "eval_steps_per_second": 0.298,
2458
+ "eval_wer": 17.92767645392155,
2459
+ "step": 9500
2460
+ },
2461
+ {
2462
+ "epoch": 2.47,
2463
+ "learning_rate": 1.4944103858636853e-06,
2464
+ "loss": 0.0519,
2465
+ "step": 9525
2466
+ },
2467
+ {
2468
+ "epoch": 2.47,
2469
+ "learning_rate": 1.4763793725207356e-06,
2470
+ "loss": 0.0319,
2471
+ "step": 9550
2472
+ },
2473
+ {
2474
+ "epoch": 2.48,
2475
+ "learning_rate": 1.4583483591777857e-06,
2476
+ "loss": 0.0206,
2477
+ "step": 9575
2478
+ },
2479
+ {
2480
+ "epoch": 2.48,
2481
+ "learning_rate": 1.4403173458348358e-06,
2482
+ "loss": 0.0339,
2483
+ "step": 9600
2484
+ },
2485
+ {
2486
+ "epoch": 2.49,
2487
+ "learning_rate": 1.422286332491886e-06,
2488
+ "loss": 0.0218,
2489
+ "step": 9625
2490
+ },
2491
+ {
2492
+ "epoch": 2.5,
2493
+ "learning_rate": 1.4042553191489362e-06,
2494
+ "loss": 0.0353,
2495
+ "step": 9650
2496
+ },
2497
+ {
2498
+ "epoch": 2.5,
2499
+ "learning_rate": 1.386224305805986e-06,
2500
+ "loss": 0.022,
2501
+ "step": 9675
2502
+ },
2503
+ {
2504
+ "epoch": 2.51,
2505
+ "learning_rate": 1.3681932924630364e-06,
2506
+ "loss": 0.0531,
2507
+ "step": 9700
2508
+ },
2509
+ {
2510
+ "epoch": 2.52,
2511
+ "learning_rate": 1.3501622791200865e-06,
2512
+ "loss": 0.0302,
2513
+ "step": 9725
2514
+ },
2515
+ {
2516
+ "epoch": 2.52,
2517
+ "learning_rate": 1.3321312657771366e-06,
2518
+ "loss": 0.0239,
2519
+ "step": 9750
2520
+ },
2521
+ {
2522
+ "epoch": 2.53,
2523
+ "learning_rate": 1.3141002524341867e-06,
2524
+ "loss": 0.0207,
2525
+ "step": 9775
2526
+ },
2527
+ {
2528
+ "epoch": 2.54,
2529
+ "learning_rate": 1.296069239091237e-06,
2530
+ "loss": 0.0271,
2531
+ "step": 9800
2532
+ },
2533
+ {
2534
+ "epoch": 2.54,
2535
+ "learning_rate": 1.2780382257482868e-06,
2536
+ "loss": 0.0245,
2537
+ "step": 9825
2538
+ },
2539
+ {
2540
+ "epoch": 2.55,
2541
+ "learning_rate": 1.2600072124053372e-06,
2542
+ "loss": 0.0217,
2543
+ "step": 9850
2544
+ },
2545
+ {
2546
+ "epoch": 2.56,
2547
+ "learning_rate": 1.2419761990623872e-06,
2548
+ "loss": 0.0297,
2549
+ "step": 9875
2550
+ },
2551
+ {
2552
+ "epoch": 2.56,
2553
+ "learning_rate": 1.2239451857194373e-06,
2554
+ "loss": 0.0313,
2555
+ "step": 9900
2556
+ },
2557
+ {
2558
+ "epoch": 2.57,
2559
+ "learning_rate": 1.2059141723764874e-06,
2560
+ "loss": 0.0269,
2561
+ "step": 9925
2562
+ },
2563
+ {
2564
+ "epoch": 2.58,
2565
+ "learning_rate": 1.1878831590335377e-06,
2566
+ "loss": 0.0273,
2567
+ "step": 9950
2568
+ },
2569
+ {
2570
+ "epoch": 2.58,
2571
+ "learning_rate": 1.1698521456905876e-06,
2572
+ "loss": 0.033,
2573
+ "step": 9975
2574
+ },
2575
+ {
2576
+ "epoch": 2.59,
2577
+ "learning_rate": 1.151821132347638e-06,
2578
+ "loss": 0.0333,
2579
+ "step": 10000
2580
+ },
2581
+ {
2582
+ "epoch": 2.59,
2583
+ "eval_loss": 0.057692721486091614,
2584
+ "eval_runtime": 4599.0725,
2585
+ "eval_samples_per_second": 2.372,
2586
+ "eval_steps_per_second": 0.297,
2587
+ "eval_wer": 17.71779965160451,
2588
+ "step": 10000
2589
+ },
2590
+ {
2591
+ "epoch": 2.59,
2592
+ "learning_rate": 1.133790119004688e-06,
2593
+ "loss": 0.0359,
2594
+ "step": 10025
2595
+ },
2596
+ {
2597
+ "epoch": 2.6,
2598
+ "learning_rate": 1.115759105661738e-06,
2599
+ "loss": 0.0413,
2600
+ "step": 10050
2601
+ },
2602
+ {
2603
+ "epoch": 2.61,
2604
+ "learning_rate": 1.0977280923187882e-06,
2605
+ "loss": 0.0304,
2606
+ "step": 10075
2607
+ },
2608
+ {
2609
+ "epoch": 2.61,
2610
+ "learning_rate": 1.0796970789758385e-06,
2611
+ "loss": 0.0297,
2612
+ "step": 10100
2613
+ },
2614
+ {
2615
+ "epoch": 2.62,
2616
+ "learning_rate": 1.0616660656328886e-06,
2617
+ "loss": 0.0365,
2618
+ "step": 10125
2619
+ },
2620
+ {
2621
+ "epoch": 2.63,
2622
+ "learning_rate": 1.0436350522899387e-06,
2623
+ "loss": 0.041,
2624
+ "step": 10150
2625
+ },
2626
+ {
2627
+ "epoch": 2.63,
2628
+ "learning_rate": 1.025604038946989e-06,
2629
+ "loss": 0.0273,
2630
+ "step": 10175
2631
+ },
2632
+ {
2633
+ "epoch": 2.64,
2634
+ "learning_rate": 1.0075730256040389e-06,
2635
+ "loss": 0.0293,
2636
+ "step": 10200
2637
+ },
2638
+ {
2639
+ "epoch": 2.65,
2640
+ "learning_rate": 9.89542012261089e-07,
2641
+ "loss": 0.0254,
2642
+ "step": 10225
2643
+ },
2644
+ {
2645
+ "epoch": 2.65,
2646
+ "learning_rate": 9.71510998918139e-07,
2647
+ "loss": 0.0233,
2648
+ "step": 10250
2649
+ },
2650
+ {
2651
+ "epoch": 2.66,
2652
+ "learning_rate": 9.534799855751892e-07,
2653
+ "loss": 0.0252,
2654
+ "step": 10275
2655
+ },
2656
+ {
2657
+ "epoch": 2.67,
2658
+ "learning_rate": 9.354489722322394e-07,
2659
+ "loss": 0.0246,
2660
+ "step": 10300
2661
+ },
2662
+ {
2663
+ "epoch": 2.67,
2664
+ "learning_rate": 9.174179588892895e-07,
2665
+ "loss": 0.0196,
2666
+ "step": 10325
2667
+ },
2668
+ {
2669
+ "epoch": 2.68,
2670
+ "learning_rate": 8.993869455463396e-07,
2671
+ "loss": 0.0354,
2672
+ "step": 10350
2673
+ },
2674
+ {
2675
+ "epoch": 2.69,
2676
+ "learning_rate": 8.813559322033897e-07,
2677
+ "loss": 0.0352,
2678
+ "step": 10375
2679
+ },
2680
+ {
2681
+ "epoch": 2.69,
2682
+ "learning_rate": 8.633249188604399e-07,
2683
+ "loss": 0.0279,
2684
+ "step": 10400
2685
+ },
2686
+ {
2687
+ "epoch": 2.7,
2688
+ "learning_rate": 8.4529390551749e-07,
2689
+ "loss": 0.0329,
2690
+ "step": 10425
2691
+ },
2692
+ {
2693
+ "epoch": 2.7,
2694
+ "learning_rate": 8.272628921745401e-07,
2695
+ "loss": 0.0247,
2696
+ "step": 10450
2697
+ },
2698
+ {
2699
+ "epoch": 2.71,
2700
+ "learning_rate": 8.092318788315903e-07,
2701
+ "loss": 0.0222,
2702
+ "step": 10475
2703
+ },
2704
+ {
2705
+ "epoch": 2.72,
2706
+ "learning_rate": 7.912008654886404e-07,
2707
+ "loss": 0.0264,
2708
+ "step": 10500
2709
+ },
2710
+ {
2711
+ "epoch": 2.72,
2712
+ "eval_loss": 0.05672454833984375,
2713
+ "eval_runtime": 4593.3082,
2714
+ "eval_samples_per_second": 2.375,
2715
+ "eval_steps_per_second": 0.297,
2716
+ "eval_wer": 17.501626545217956,
2717
+ "step": 10500
2718
+ },
2719
+ {
2720
+ "epoch": 2.72,
2721
+ "learning_rate": 7.731698521456905e-07,
2722
+ "loss": 0.0232,
2723
+ "step": 10525
2724
+ },
2725
+ {
2726
+ "epoch": 2.73,
2727
+ "learning_rate": 7.551388388027407e-07,
2728
+ "loss": 0.0227,
2729
+ "step": 10550
2730
+ },
2731
+ {
2732
+ "epoch": 2.74,
2733
+ "learning_rate": 7.371078254597908e-07,
2734
+ "loss": 0.0385,
2735
+ "step": 10575
2736
+ },
2737
+ {
2738
+ "epoch": 2.74,
2739
+ "learning_rate": 7.190768121168409e-07,
2740
+ "loss": 0.0323,
2741
+ "step": 10600
2742
+ },
2743
+ {
2744
+ "epoch": 2.75,
2745
+ "learning_rate": 7.010457987738911e-07,
2746
+ "loss": 0.0201,
2747
+ "step": 10625
2748
+ },
2749
+ {
2750
+ "epoch": 2.76,
2751
+ "learning_rate": 6.830147854309412e-07,
2752
+ "loss": 0.0336,
2753
+ "step": 10650
2754
+ },
2755
+ {
2756
+ "epoch": 2.76,
2757
+ "learning_rate": 6.649837720879913e-07,
2758
+ "loss": 0.0348,
2759
+ "step": 10675
2760
+ },
2761
+ {
2762
+ "epoch": 2.77,
2763
+ "learning_rate": 6.469527587450415e-07,
2764
+ "loss": 0.0238,
2765
+ "step": 10700
2766
+ },
2767
+ {
2768
+ "epoch": 2.78,
2769
+ "learning_rate": 6.289217454020915e-07,
2770
+ "loss": 0.0223,
2771
+ "step": 10725
2772
+ },
2773
+ {
2774
+ "epoch": 2.78,
2775
+ "learning_rate": 6.108907320591416e-07,
2776
+ "loss": 0.0294,
2777
+ "step": 10750
2778
+ },
2779
+ {
2780
+ "epoch": 2.79,
2781
+ "learning_rate": 5.928597187161918e-07,
2782
+ "loss": 0.0273,
2783
+ "step": 10775
2784
+ },
2785
+ {
2786
+ "epoch": 2.8,
2787
+ "learning_rate": 5.748287053732419e-07,
2788
+ "loss": 0.035,
2789
+ "step": 10800
2790
+ },
2791
+ {
2792
+ "epoch": 2.8,
2793
+ "learning_rate": 5.56797692030292e-07,
2794
+ "loss": 0.0388,
2795
+ "step": 10825
2796
+ },
2797
+ {
2798
+ "epoch": 2.81,
2799
+ "learning_rate": 5.387666786873421e-07,
2800
+ "loss": 0.0218,
2801
+ "step": 10850
2802
+ },
2803
+ {
2804
+ "epoch": 2.81,
2805
+ "learning_rate": 5.207356653443923e-07,
2806
+ "loss": 0.0347,
2807
+ "step": 10875
2808
+ },
2809
+ {
2810
+ "epoch": 2.82,
2811
+ "learning_rate": 5.027046520014424e-07,
2812
+ "loss": 0.0211,
2813
+ "step": 10900
2814
+ },
2815
+ {
2816
+ "epoch": 2.83,
2817
+ "learning_rate": 4.846736386584926e-07,
2818
+ "loss": 0.0265,
2819
+ "step": 10925
2820
+ },
2821
+ {
2822
+ "epoch": 2.83,
2823
+ "learning_rate": 4.666426253155427e-07,
2824
+ "loss": 0.0362,
2825
+ "step": 10950
2826
+ },
2827
+ {
2828
+ "epoch": 2.84,
2829
+ "learning_rate": 4.4861161197259284e-07,
2830
+ "loss": 0.0253,
2831
+ "step": 10975
2832
+ },
2833
+ {
2834
+ "epoch": 2.85,
2835
+ "learning_rate": 4.30580598629643e-07,
2836
+ "loss": 0.0274,
2837
+ "step": 11000
2838
+ },
2839
+ {
2840
+ "epoch": 2.85,
2841
+ "eval_loss": 0.055780503898859024,
2842
+ "eval_runtime": 4593.0706,
2843
+ "eval_samples_per_second": 2.375,
2844
+ "eval_steps_per_second": 0.297,
2845
+ "eval_wer": 17.449157344638696,
2846
+ "step": 11000
2847
+ }
2848
+ ],
2849
+ "max_steps": 11592,
2850
+ "num_train_epochs": 3,
2851
+ "total_flos": 5.078872177311744e+19,
2852
+ "trial_name": null,
2853
+ "trial_params": null
2854
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fe37312ed5591e9bbc8cf1ec984886da4ea3113f5dd7b1e43712a174423b4fe
3
+ size 4091