StephennFernandes commited on
Commit
3b18175
1 Parent(s): 7ec2993
.gitattributes CHANGED
@@ -25,3 +25,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
25
  *.zip filter=lfs diff=lfs merge=lfs -text
26
  *.zstandard filter=lfs diff=lfs merge=lfs -text
27
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
25
  *.zip filter=lfs diff=lfs merge=lfs -text
26
  *.zstandard filter=lfs diff=lfs merge=lfs -text
27
  *tfevents* filter=lfs diff=lfs merge=lfs -text
28
+ language_model/unigrams.txt filter=lfs diff=lfs merge=lfs -text
added_tokens.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"<s>": 88, "</s>": 89}
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 30.0,
3
+ "eval_loss": 0.5320505499839783,
4
+ "eval_runtime": 210.2131,
5
+ "eval_samples": 3401,
6
+ "eval_samples_per_second": 16.179,
7
+ "eval_steps_per_second": 1.013,
8
+ "eval_wer": 0.3199213685223138,
9
+ "train_loss": 0.5493880579162342,
10
+ "train_runtime": 54395.2991,
11
+ "train_samples": 30605,
12
+ "train_samples_per_second": 16.879,
13
+ "train_steps_per_second": 0.264
14
+ }
alphabet.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"labels": ["\u0925", "\u0947", "\u092d", "i", "5", "\u090b", "\u0924", "\u0940", "\u092c", "\u091c", "6", "\u092a", "\t", "7", "\u094a", "\u0910", "\u0916", "\u090d", "\u0921", "\u2019", "\u093d", "\u091a", "\u090f", "\u091e", "\u0911", "\u0905", "\u0945", "\u0926", "\u0902", "\u0909", "*", "]", "\u091d", "\u0901", "\u0920", "\u0949", "\u0907", "0", "\u0928", "\u091f", "\u0923", "\u094b", "\u0335", "1", "e", "\u0915", "\u0918", "\u0948", "\u091b", "\u092e", "\u0913", "\u0933", "\u0939", "\u0917", "\u0967", "\u0936", "\u0903", "9", "\u092b", "\u0943", "\u200d", "\u0930", "8", "\u092f", "\u0938", "\u093f", "\u094c", "\u0927", "\u0931", "\u0937", "\u0932", "2", "\u0941", "\u0935", "\u0942", "\u0919", "\u095f", " ", "[", "\u093e", "\u0906", "\u090a", "\u094d", "\u0914", "\u093c", "\u0908", "\u2047", "", "<s>", "</s>"], "is_bpe": false}
config.json ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-xls-r-300m",
3
+ "activation_dropout": 0.055,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "Wav2Vec2ForCTC"
7
+ ],
8
+ "attention_dropout": 0.094,
9
+ "bos_token_id": 1,
10
+ "codevector_dim": 768,
11
+ "contrastive_logits_temperature": 0.1,
12
+ "conv_bias": true,
13
+ "conv_dim": [
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512,
20
+ 512
21
+ ],
22
+ "conv_kernel": [
23
+ 10,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 3,
28
+ 2,
29
+ 2
30
+ ],
31
+ "conv_stride": [
32
+ 5,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2,
38
+ 2
39
+ ],
40
+ "ctc_loss_reduction": "mean",
41
+ "ctc_zero_infinity": true,
42
+ "diversity_loss_weight": 0.1,
43
+ "do_stable_layer_norm": true,
44
+ "eos_token_id": 2,
45
+ "feat_extract_activation": "gelu",
46
+ "feat_extract_dropout": 0.0,
47
+ "feat_extract_norm": "layer",
48
+ "feat_proj_dropout": 0.04,
49
+ "feat_quantizer_dropout": 0.0,
50
+ "final_dropout": 0.0,
51
+ "gradient_checkpointing": 1,
52
+ "hidden_act": "gelu",
53
+ "hidden_dropout": 0.1,
54
+ "hidden_size": 1024,
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 4096,
57
+ "layer_norm_eps": 1e-05,
58
+ "layerdrop": 0.04,
59
+ "mask_feature_length": 10,
60
+ "mask_feature_prob": 0.0,
61
+ "mask_time_length": 10,
62
+ "mask_time_prob": 0.05,
63
+ "model_type": "wav2vec2",
64
+ "num_attention_heads": 16,
65
+ "num_codevector_groups": 2,
66
+ "num_codevectors_per_group": 320,
67
+ "num_conv_pos_embedding_groups": 16,
68
+ "num_conv_pos_embeddings": 128,
69
+ "num_feat_extract_layers": 7,
70
+ "num_hidden_layers": 24,
71
+ "num_negatives": 100,
72
+ "pad_token_id": 87,
73
+ "proj_codevector_dim": 768,
74
+ "torch_dtype": "float32",
75
+ "transformers_version": "4.8.0",
76
+ "use_weighted_layer_sum": false,
77
+ "vocab_size": 90
78
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 30.0,
3
+ "eval_loss": 0.5320505499839783,
4
+ "eval_runtime": 210.2131,
5
+ "eval_samples": 3401,
6
+ "eval_samples_per_second": 16.179,
7
+ "eval_steps_per_second": 1.013,
8
+ "eval_wer": 0.3199213685223138
9
+ }
language_model/5gram.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba2177b37bb40a37bb36c7e4e2a9795e44b574883d38c883ba67c9ab3c560e78
3
+ size 334018687
language_model/attrs.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"alpha": 0.5, "beta": 1.5, "unk_score_offset": -10.0, "score_boundary": true}
language_model/unigrams.txt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9662db773aa10c111d5ad487a984909796d87450cdc8da5e0b4dcd89404ad208
3
+ size 10603812
preprocessor_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "processor_class": "Wav2Vec2ProcessorWithLM",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 16000
10
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cc9dc511453f4ec6e5ce4322bc53a8716c7b200d9f8051ceec31ef697110bc5
3
+ size 1262302807
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "special_tokens_map_file": "../../konkani_data/XLS-R-konkani/special_tokens_map.json", "tokenizer_file": null, "name_or_path": "../../konkani_data/XLS-R-konkani", "tokenizer_class": "Wav2Vec2CTCTokenizer", "processor_class": "Wav2Vec2ProcessorWithLM"}
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 30.0,
3
+ "train_loss": 0.5493880579162342,
4
+ "train_runtime": 54395.2991,
5
+ "train_samples": 30605,
6
+ "train_samples_per_second": 16.879,
7
+ "train_steps_per_second": 0.264
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,550 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 29.998955067920583,
5
+ "global_step": 14340,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.84,
12
+ "learning_rate": 0.000148875,
13
+ "loss": 6.9674,
14
+ "step": 400
15
+ },
16
+ {
17
+ "epoch": 0.84,
18
+ "eval_loss": 3.427687168121338,
19
+ "eval_runtime": 218.9995,
20
+ "eval_samples_per_second": 15.53,
21
+ "eval_steps_per_second": 0.973,
22
+ "eval_wer": 1.0,
23
+ "step": 400
24
+ },
25
+ {
26
+ "epoch": 1.67,
27
+ "learning_rate": 0.000298875,
28
+ "loss": 2.566,
29
+ "step": 800
30
+ },
31
+ {
32
+ "epoch": 1.67,
33
+ "eval_loss": 1.0569312572479248,
34
+ "eval_runtime": 218.6402,
35
+ "eval_samples_per_second": 15.555,
36
+ "eval_steps_per_second": 0.974,
37
+ "eval_wer": 0.7833242711949475,
38
+ "step": 800
39
+ },
40
+ {
41
+ "epoch": 2.51,
42
+ "learning_rate": 0.0002912038404726735,
43
+ "loss": 1.0118,
44
+ "step": 1200
45
+ },
46
+ {
47
+ "epoch": 2.51,
48
+ "eval_loss": 0.6927362680435181,
49
+ "eval_runtime": 220.9263,
50
+ "eval_samples_per_second": 15.394,
51
+ "eval_steps_per_second": 0.964,
52
+ "eval_wer": 0.5602283658873227,
53
+ "step": 1200
54
+ },
55
+ {
56
+ "epoch": 3.35,
57
+ "learning_rate": 0.000282341211225997,
58
+ "loss": 0.7536,
59
+ "step": 1600
60
+ },
61
+ {
62
+ "epoch": 3.35,
63
+ "eval_loss": 0.5688415765762329,
64
+ "eval_runtime": 210.7589,
65
+ "eval_samples_per_second": 16.137,
66
+ "eval_steps_per_second": 1.011,
67
+ "eval_wer": 0.5082395750554185,
68
+ "step": 1600
69
+ },
70
+ {
71
+ "epoch": 4.18,
72
+ "learning_rate": 0.00027347858197932054,
73
+ "loss": 0.6251,
74
+ "step": 2000
75
+ },
76
+ {
77
+ "epoch": 4.18,
78
+ "eval_loss": 0.5367330312728882,
79
+ "eval_runtime": 211.3794,
80
+ "eval_samples_per_second": 16.09,
81
+ "eval_steps_per_second": 1.008,
82
+ "eval_wer": 0.46095612530846125,
83
+ "step": 2000
84
+ },
85
+ {
86
+ "epoch": 5.02,
87
+ "learning_rate": 0.00026461595273264397,
88
+ "loss": 0.5453,
89
+ "step": 2400
90
+ },
91
+ {
92
+ "epoch": 5.02,
93
+ "eval_loss": 0.47355735301971436,
94
+ "eval_runtime": 211.7978,
95
+ "eval_samples_per_second": 16.058,
96
+ "eval_steps_per_second": 1.006,
97
+ "eval_wer": 0.4430758291856623,
98
+ "step": 2400
99
+ },
100
+ {
101
+ "epoch": 5.86,
102
+ "learning_rate": 0.0002557533234859675,
103
+ "loss": 0.4779,
104
+ "step": 2800
105
+ },
106
+ {
107
+ "epoch": 5.86,
108
+ "eval_loss": 0.4465107023715973,
109
+ "eval_runtime": 211.1636,
110
+ "eval_samples_per_second": 16.106,
111
+ "eval_steps_per_second": 1.009,
112
+ "eval_wer": 0.4200719394370321,
113
+ "step": 2800
114
+ },
115
+ {
116
+ "epoch": 6.69,
117
+ "learning_rate": 0.00024689069423929094,
118
+ "loss": 0.4458,
119
+ "step": 3200
120
+ },
121
+ {
122
+ "epoch": 6.69,
123
+ "eval_loss": 0.4270594120025635,
124
+ "eval_runtime": 210.6558,
125
+ "eval_samples_per_second": 16.145,
126
+ "eval_steps_per_second": 1.011,
127
+ "eval_wer": 0.4025262453469405,
128
+ "step": 3200
129
+ },
130
+ {
131
+ "epoch": 7.53,
132
+ "learning_rate": 0.00023802806499261444,
133
+ "loss": 0.4036,
134
+ "step": 3600
135
+ },
136
+ {
137
+ "epoch": 7.53,
138
+ "eval_loss": 0.4414581060409546,
139
+ "eval_runtime": 212.0272,
140
+ "eval_samples_per_second": 16.04,
141
+ "eval_steps_per_second": 1.005,
142
+ "eval_wer": 0.3957505541846167,
143
+ "step": 3600
144
+ },
145
+ {
146
+ "epoch": 8.37,
147
+ "learning_rate": 0.00022916543574593795,
148
+ "loss": 0.377,
149
+ "step": 4000
150
+ },
151
+ {
152
+ "epoch": 8.37,
153
+ "eval_loss": 0.4214448034763336,
154
+ "eval_runtime": 210.8625,
155
+ "eval_samples_per_second": 16.129,
156
+ "eval_steps_per_second": 1.01,
157
+ "eval_wer": 0.39071061106696225,
158
+ "step": 4000
159
+ },
160
+ {
161
+ "epoch": 9.2,
162
+ "learning_rate": 0.0002203028064992614,
163
+ "loss": 0.347,
164
+ "step": 4400
165
+ },
166
+ {
167
+ "epoch": 9.2,
168
+ "eval_loss": 0.42906612157821655,
169
+ "eval_runtime": 212.1547,
170
+ "eval_samples_per_second": 16.031,
171
+ "eval_steps_per_second": 1.004,
172
+ "eval_wer": 0.38757371701033083,
173
+ "step": 4400
174
+ },
175
+ {
176
+ "epoch": 10.04,
177
+ "learning_rate": 0.00021144017725258492,
178
+ "loss": 0.3322,
179
+ "step": 4800
180
+ },
181
+ {
182
+ "epoch": 10.04,
183
+ "eval_loss": 0.4414941370487213,
184
+ "eval_runtime": 210.3342,
185
+ "eval_samples_per_second": 16.17,
186
+ "eval_steps_per_second": 1.013,
187
+ "eval_wer": 0.3664728763227237,
188
+ "step": 4800
189
+ },
190
+ {
191
+ "epoch": 10.88,
192
+ "learning_rate": 0.00020257754800590837,
193
+ "loss": 0.311,
194
+ "step": 5200
195
+ },
196
+ {
197
+ "epoch": 10.88,
198
+ "eval_loss": 0.40909305214881897,
199
+ "eval_runtime": 211.7535,
200
+ "eval_samples_per_second": 16.061,
201
+ "eval_steps_per_second": 1.006,
202
+ "eval_wer": 0.37214019825170436,
203
+ "step": 5200
204
+ },
205
+ {
206
+ "epoch": 11.71,
207
+ "learning_rate": 0.00019371491875923188,
208
+ "loss": 0.2956,
209
+ "step": 5600
210
+ },
211
+ {
212
+ "epoch": 11.71,
213
+ "eval_loss": 0.46582677960395813,
214
+ "eval_runtime": 210.74,
215
+ "eval_samples_per_second": 16.138,
216
+ "eval_steps_per_second": 1.011,
217
+ "eval_wer": 0.3568530678823874,
218
+ "step": 5600
219
+ },
220
+ {
221
+ "epoch": 12.55,
222
+ "learning_rate": 0.0001848522895125554,
223
+ "loss": 0.2811,
224
+ "step": 6000
225
+ },
226
+ {
227
+ "epoch": 12.55,
228
+ "eval_loss": 0.44131794571876526,
229
+ "eval_runtime": 210.7883,
230
+ "eval_samples_per_second": 16.135,
231
+ "eval_steps_per_second": 1.01,
232
+ "eval_wer": 0.3576268350830231,
233
+ "step": 6000
234
+ },
235
+ {
236
+ "epoch": 13.39,
237
+ "learning_rate": 0.00017598966026587885,
238
+ "loss": 0.2732,
239
+ "step": 6400
240
+ },
241
+ {
242
+ "epoch": 13.39,
243
+ "eval_loss": 0.48606938123703003,
244
+ "eval_runtime": 210.0412,
245
+ "eval_samples_per_second": 16.192,
246
+ "eval_steps_per_second": 1.014,
247
+ "eval_wer": 0.3552846208540717,
248
+ "step": 6400
249
+ },
250
+ {
251
+ "epoch": 14.23,
252
+ "learning_rate": 0.00016712703101920236,
253
+ "loss": 0.2672,
254
+ "step": 6800
255
+ },
256
+ {
257
+ "epoch": 14.23,
258
+ "eval_loss": 0.4633455276489258,
259
+ "eval_runtime": 212.5303,
260
+ "eval_samples_per_second": 16.002,
261
+ "eval_steps_per_second": 1.002,
262
+ "eval_wer": 0.3534861349282697,
263
+ "step": 6800
264
+ },
265
+ {
266
+ "epoch": 15.06,
267
+ "learning_rate": 0.00015826440177252584,
268
+ "loss": 0.2497,
269
+ "step": 7200
270
+ },
271
+ {
272
+ "epoch": 15.06,
273
+ "eval_loss": 0.468420147895813,
274
+ "eval_runtime": 210.9196,
275
+ "eval_samples_per_second": 16.125,
276
+ "eval_steps_per_second": 1.01,
277
+ "eval_wer": 0.3575850098289347,
278
+ "step": 7200
279
+ },
280
+ {
281
+ "epoch": 15.9,
282
+ "learning_rate": 0.00014940177252584932,
283
+ "loss": 0.2334,
284
+ "step": 7600
285
+ },
286
+ {
287
+ "epoch": 15.9,
288
+ "eval_loss": 0.4702986776828766,
289
+ "eval_runtime": 211.8705,
290
+ "eval_samples_per_second": 16.052,
291
+ "eval_steps_per_second": 1.005,
292
+ "eval_wer": 0.34524655987285124,
293
+ "step": 7600
294
+ },
295
+ {
296
+ "epoch": 16.74,
297
+ "learning_rate": 0.0001405391432791728,
298
+ "loss": 0.2324,
299
+ "step": 8000
300
+ },
301
+ {
302
+ "epoch": 16.74,
303
+ "eval_loss": 0.4267388880252838,
304
+ "eval_runtime": 211.5039,
305
+ "eval_samples_per_second": 16.08,
306
+ "eval_steps_per_second": 1.007,
307
+ "eval_wer": 0.35070475553138986,
308
+ "step": 8000
309
+ },
310
+ {
311
+ "epoch": 17.57,
312
+ "learning_rate": 0.0001316765140324963,
313
+ "loss": 0.2166,
314
+ "step": 8400
315
+ },
316
+ {
317
+ "epoch": 17.57,
318
+ "eval_loss": 0.4422346353530884,
319
+ "eval_runtime": 213.9215,
320
+ "eval_samples_per_second": 15.898,
321
+ "eval_steps_per_second": 0.996,
322
+ "eval_wer": 0.3400393157388431,
323
+ "step": 8400
324
+ },
325
+ {
326
+ "epoch": 18.41,
327
+ "learning_rate": 0.00012281388478581977,
328
+ "loss": 0.2116,
329
+ "step": 8800
330
+ },
331
+ {
332
+ "epoch": 18.41,
333
+ "eval_loss": 0.4669197201728821,
334
+ "eval_runtime": 213.3022,
335
+ "eval_samples_per_second": 15.945,
336
+ "eval_steps_per_second": 0.999,
337
+ "eval_wer": 0.33359822660922667,
338
+ "step": 8800
339
+ },
340
+ {
341
+ "epoch": 19.25,
342
+ "learning_rate": 0.00011395125553914327,
343
+ "loss": 0.2055,
344
+ "step": 9200
345
+ },
346
+ {
347
+ "epoch": 19.25,
348
+ "eval_loss": 0.46659788489341736,
349
+ "eval_runtime": 209.6264,
350
+ "eval_samples_per_second": 16.224,
351
+ "eval_steps_per_second": 1.016,
352
+ "eval_wer": 0.3343510811828182,
353
+ "step": 9200
354
+ },
355
+ {
356
+ "epoch": 20.08,
357
+ "learning_rate": 0.00010508862629246675,
358
+ "loss": 0.2,
359
+ "step": 9600
360
+ },
361
+ {
362
+ "epoch": 20.08,
363
+ "eval_loss": 0.4791451096534729,
364
+ "eval_runtime": 212.2536,
365
+ "eval_samples_per_second": 16.023,
366
+ "eval_steps_per_second": 1.004,
367
+ "eval_wer": 0.3353967125350287,
368
+ "step": 9600
369
+ },
370
+ {
371
+ "epoch": 20.92,
372
+ "learning_rate": 9.622599704579024e-05,
373
+ "loss": 0.1851,
374
+ "step": 10000
375
+ },
376
+ {
377
+ "epoch": 20.92,
378
+ "eval_loss": 0.4670654237270355,
379
+ "eval_runtime": 209.9735,
380
+ "eval_samples_per_second": 16.197,
381
+ "eval_steps_per_second": 1.014,
382
+ "eval_wer": 0.33182065331046884,
383
+ "step": 10000
384
+ },
385
+ {
386
+ "epoch": 21.76,
387
+ "learning_rate": 8.736336779911373e-05,
388
+ "loss": 0.1768,
389
+ "step": 10400
390
+ },
391
+ {
392
+ "epoch": 21.76,
393
+ "eval_loss": 0.48615434765815735,
394
+ "eval_runtime": 211.2759,
395
+ "eval_samples_per_second": 16.097,
396
+ "eval_steps_per_second": 1.008,
397
+ "eval_wer": 0.33194612907273413,
398
+ "step": 10400
399
+ },
400
+ {
401
+ "epoch": 22.59,
402
+ "learning_rate": 7.850073855243721e-05,
403
+ "loss": 0.1759,
404
+ "step": 10800
405
+ },
406
+ {
407
+ "epoch": 22.59,
408
+ "eval_loss": 0.4796726107597351,
409
+ "eval_runtime": 210.8467,
410
+ "eval_samples_per_second": 16.13,
411
+ "eval_steps_per_second": 1.01,
412
+ "eval_wer": 0.32908109916767747,
413
+ "step": 10800
414
+ },
415
+ {
416
+ "epoch": 23.43,
417
+ "learning_rate": 6.96381093057607e-05,
418
+ "loss": 0.1697,
419
+ "step": 11200
420
+ },
421
+ {
422
+ "epoch": 23.43,
423
+ "eval_loss": 0.5016443133354187,
424
+ "eval_runtime": 210.1285,
425
+ "eval_samples_per_second": 16.185,
426
+ "eval_steps_per_second": 1.014,
427
+ "eval_wer": 0.32728261324187546,
428
+ "step": 11200
429
+ },
430
+ {
431
+ "epoch": 24.27,
432
+ "learning_rate": 6.077548005908419e-05,
433
+ "loss": 0.162,
434
+ "step": 11600
435
+ },
436
+ {
437
+ "epoch": 24.27,
438
+ "eval_loss": 0.4838166832923889,
439
+ "eval_runtime": 210.5105,
440
+ "eval_samples_per_second": 16.156,
441
+ "eval_steps_per_second": 1.012,
442
+ "eval_wer": 0.32224267012422103,
443
+ "step": 11600
444
+ },
445
+ {
446
+ "epoch": 25.1,
447
+ "learning_rate": 5.191285081240768e-05,
448
+ "loss": 0.1552,
449
+ "step": 12000
450
+ },
451
+ {
452
+ "epoch": 25.1,
453
+ "eval_loss": 0.4953179657459259,
454
+ "eval_runtime": 214.2232,
455
+ "eval_samples_per_second": 15.876,
456
+ "eval_steps_per_second": 0.994,
457
+ "eval_wer": 0.3248985737588356,
458
+ "step": 12000
459
+ },
460
+ {
461
+ "epoch": 25.94,
462
+ "learning_rate": 4.3050221565731165e-05,
463
+ "loss": 0.1505,
464
+ "step": 12400
465
+ },
466
+ {
467
+ "epoch": 25.94,
468
+ "eval_loss": 0.5147430300712585,
469
+ "eval_runtime": 213.2541,
470
+ "eval_samples_per_second": 15.948,
471
+ "eval_steps_per_second": 0.999,
472
+ "eval_wer": 0.320444184198419,
473
+ "step": 12400
474
+ },
475
+ {
476
+ "epoch": 26.78,
477
+ "learning_rate": 3.418759231905465e-05,
478
+ "loss": 0.1505,
479
+ "step": 12800
480
+ },
481
+ {
482
+ "epoch": 26.78,
483
+ "eval_loss": 0.5216009616851807,
484
+ "eval_runtime": 214.5736,
485
+ "eval_samples_per_second": 15.85,
486
+ "eval_steps_per_second": 0.993,
487
+ "eval_wer": 0.32184533021038103,
488
+ "step": 12800
489
+ },
490
+ {
491
+ "epoch": 27.61,
492
+ "learning_rate": 2.534711964549483e-05,
493
+ "loss": 0.1441,
494
+ "step": 13200
495
+ },
496
+ {
497
+ "epoch": 27.61,
498
+ "eval_loss": 0.5204435586929321,
499
+ "eval_runtime": 210.4984,
500
+ "eval_samples_per_second": 16.157,
501
+ "eval_steps_per_second": 1.012,
502
+ "eval_wer": 0.32084152411225897,
503
+ "step": 13200
504
+ },
505
+ {
506
+ "epoch": 28.45,
507
+ "learning_rate": 1.6506646971935004e-05,
508
+ "loss": 0.1432,
509
+ "step": 13600
510
+ },
511
+ {
512
+ "epoch": 28.45,
513
+ "eval_loss": 0.5269867777824402,
514
+ "eval_runtime": 210.8697,
515
+ "eval_samples_per_second": 16.128,
516
+ "eval_steps_per_second": 1.01,
517
+ "eval_wer": 0.31854113513739596,
518
+ "step": 13600
519
+ },
520
+ {
521
+ "epoch": 29.29,
522
+ "learning_rate": 7.644017725258493e-06,
523
+ "loss": 0.1379,
524
+ "step": 14000
525
+ },
526
+ {
527
+ "epoch": 29.29,
528
+ "eval_loss": 0.5423755049705505,
529
+ "eval_runtime": 212.2394,
530
+ "eval_samples_per_second": 16.024,
531
+ "eval_steps_per_second": 1.004,
532
+ "eval_wer": 0.3176209795474508,
533
+ "step": 14000
534
+ },
535
+ {
536
+ "epoch": 30.0,
537
+ "step": 14340,
538
+ "total_flos": 2.519939618068477e+20,
539
+ "train_loss": 0.5493880579162342,
540
+ "train_runtime": 54395.2991,
541
+ "train_samples_per_second": 16.879,
542
+ "train_steps_per_second": 0.264
543
+ }
544
+ ],
545
+ "max_steps": 14340,
546
+ "num_train_epochs": 30,
547
+ "total_flos": 2.519939618068477e+20,
548
+ "trial_name": null,
549
+ "trial_params": null
550
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d96ea3bb8c12990fc3033e3717e917a2e69392e05145af21c98f3996837104c
3
+ size 2607
vocab.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"थ": 0, "े": 1, "भ": 2, "i": 3, "5": 4, "ऋ": 5, "त": 6, "ी": 7, "ब": 8, "ज": 9, "6": 10, "प": 11, "\t": 12, "7": 13, "ॊ": 14, "ऐ": 15, "ख": 16, "ऍ": 17, "ड": 18, "’": 19, "ऽ": 20, "च": 21, "ए": 22, "ञ": 23, "ऑ": 24, "अ": 25, "ॅ": 26, "द": 27, "ं": 28, "उ": 29, "*": 30, "]": 31, "झ": 32, "ँ": 33, "ठ": 34, "ॉ": 35, "इ": 36, "0": 37, "न": 38, "ट": 39, "ण": 40, "ो": 41, "̵": 42, "1": 43, "e": 44, "क": 45, "घ": 46, "ै": 47, "छ": 48, "म": 49, "ओ": 50, "ळ": 51, "ह": 52, "ग": 53, "१": 54, "श": 55, "ः": 56, "9": 57, "फ": 58, "ृ": 59, "‍": 60, "र": 61, "8": 62, "य": 63, "स": 64, "ि": 65, "ौ": 66, "ध": 67, "ऱ": 68, "ष": 69, "ल": 70, "2": 71, "ु": 72, "व": 73, "ू": 74, "ङ": 75, "य़": 76, "[": 78, "ा": 79, "आ": 80, "ऊ": 81, "्": 82, "औ": 83, "़": 84, "ई": 85, "|": 77, "[UNK]": 86, "[PAD]": 87}