mikhail-panzo commited on
Commit
978b80b
1 Parent(s): 3fd135e

Training in progress, step 3500, checkpoint

Browse files
checkpoint-3500/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<ctc_blank>": 80,
3
+ "<mask>": 79
4
+ }
checkpoint-3500/config.json ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/speecht5_tts",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "SpeechT5ForTextToSpeech"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 0,
10
+ "conv_bias": false,
11
+ "conv_dim": [
12
+ 512,
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512
19
+ ],
20
+ "conv_kernel": [
21
+ 10,
22
+ 3,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 2,
27
+ 2
28
+ ],
29
+ "conv_stride": [
30
+ 5,
31
+ 2,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2
37
+ ],
38
+ "decoder_attention_heads": 12,
39
+ "decoder_ffn_dim": 3072,
40
+ "decoder_layerdrop": 0.1,
41
+ "decoder_layers": 6,
42
+ "decoder_start_token_id": 2,
43
+ "encoder_attention_heads": 12,
44
+ "encoder_ffn_dim": 3072,
45
+ "encoder_layerdrop": 0.1,
46
+ "encoder_layers": 12,
47
+ "encoder_max_relative_position": 160,
48
+ "eos_token_id": 2,
49
+ "feat_extract_activation": "gelu",
50
+ "feat_extract_norm": "group",
51
+ "feat_proj_dropout": 0.0,
52
+ "guided_attention_loss_num_heads": 2,
53
+ "guided_attention_loss_scale": 10.0,
54
+ "guided_attention_loss_sigma": 0.4,
55
+ "hidden_act": "gelu",
56
+ "hidden_dropout": 0.1,
57
+ "hidden_size": 768,
58
+ "initializer_range": 0.02,
59
+ "is_encoder_decoder": true,
60
+ "layer_norm_eps": 1e-05,
61
+ "mask_feature_length": 10,
62
+ "mask_feature_min_masks": 0,
63
+ "mask_feature_prob": 0.0,
64
+ "mask_time_length": 10,
65
+ "mask_time_min_masks": 2,
66
+ "mask_time_prob": 0.05,
67
+ "max_length": 1876,
68
+ "max_speech_positions": 1876,
69
+ "max_text_positions": 600,
70
+ "model_type": "speecht5",
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_mel_bins": 80,
75
+ "pad_token_id": 1,
76
+ "positional_dropout": 0.1,
77
+ "reduction_factor": 2,
78
+ "scale_embedding": false,
79
+ "speaker_embedding_dim": 512,
80
+ "speech_decoder_postnet_dropout": 0.5,
81
+ "speech_decoder_postnet_kernel": 5,
82
+ "speech_decoder_postnet_layers": 5,
83
+ "speech_decoder_postnet_units": 256,
84
+ "speech_decoder_prenet_dropout": 0.5,
85
+ "speech_decoder_prenet_layers": 2,
86
+ "speech_decoder_prenet_units": 256,
87
+ "torch_dtype": "float32",
88
+ "transformers_version": "4.41.0.dev0",
89
+ "use_cache": false,
90
+ "use_guided_attention_loss": true,
91
+ "vocab_size": 81
92
+ }
checkpoint-3500/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "decoder_start_token_id": 2,
5
+ "eos_token_id": 2,
6
+ "max_length": 1876,
7
+ "pad_token_id": 1,
8
+ "transformers_version": "4.41.0.dev0"
9
+ }
checkpoint-3500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c23539bd1460813c8c4e081cced4ae9da13f325d39d2869d2602ccca94f150df
3
+ size 577789320
checkpoint-3500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b3da866d3b062482da0e647d41072de2f3a6136e08c11989e6bfaef36a2c1a4
3
+ size 1155772233
checkpoint-3500/preprocessor_config.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": false,
3
+ "feature_extractor_type": "SpeechT5FeatureExtractor",
4
+ "feature_size": 1,
5
+ "fmax": 7600,
6
+ "fmin": 80,
7
+ "frame_signal_scale": 1.0,
8
+ "hop_length": 16,
9
+ "mel_floor": 1e-10,
10
+ "num_mel_bins": 80,
11
+ "padding_side": "right",
12
+ "padding_value": 0.0,
13
+ "processor_class": "SpeechT5Processor",
14
+ "reduction_factor": 2,
15
+ "return_attention_mask": true,
16
+ "sampling_rate": 16000,
17
+ "win_function": "hann_window",
18
+ "win_length": 64
19
+ }
checkpoint-3500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09fabf5a21d82bc464522140f30f6c8c93915041bac1fde4a996b7c32edee4ef
3
+ size 14244
checkpoint-3500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f3ed572a87ec987aa1e9e8b51db6d752c78810e3be84d7ec6c3a5e266eaee75
3
+ size 1064
checkpoint-3500/special_tokens_map.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "mask_token": {
5
+ "content": "<mask>",
6
+ "lstrip": true,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false
10
+ },
11
+ "pad_token": "<pad>",
12
+ "unk_token": "<unk>"
13
+ }
checkpoint-3500/spm_char.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fcc48f3e225f627b1641db410ceb0c8649bd2b0c982e150b03f8be3728ab560
3
+ size 238473
checkpoint-3500/tokenizer_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "79": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": true,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "80": {
44
+ "content": "<ctc_blank>",
45
+ "lstrip": false,
46
+ "normalized": true,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": false
50
+ }
51
+ },
52
+ "bos_token": "<s>",
53
+ "clean_up_tokenization_spaces": true,
54
+ "eos_token": "</s>",
55
+ "mask_token": "<mask>",
56
+ "model_max_length": 600,
57
+ "normalize": false,
58
+ "pad_token": "<pad>",
59
+ "processor_class": "SpeechT5Processor",
60
+ "sp_model_kwargs": {},
61
+ "tokenizer_class": "SpeechT5Tokenizer",
62
+ "unk_token": "<unk>"
63
+ }
checkpoint-3500/trainer_state.json ADDED
@@ -0,0 +1,579 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.41453465819358826,
3
+ "best_model_checkpoint": "mikhail-panzo/fil_b128_le3_s4000/checkpoint-500",
4
+ "epoch": 311.1111111111111,
5
+ "eval_steps": 500,
6
+ "global_step": 3500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 4.444444444444445,
13
+ "grad_norm": 1.5276468992233276,
14
+ "learning_rate": 2.5e-05,
15
+ "loss": 0.7204,
16
+ "step": 50
17
+ },
18
+ {
19
+ "epoch": 8.88888888888889,
20
+ "grad_norm": 1.4177956581115723,
21
+ "learning_rate": 5e-05,
22
+ "loss": 0.5494,
23
+ "step": 100
24
+ },
25
+ {
26
+ "epoch": 13.333333333333334,
27
+ "grad_norm": 3.0381362438201904,
28
+ "learning_rate": 7.5e-05,
29
+ "loss": 0.4948,
30
+ "step": 150
31
+ },
32
+ {
33
+ "epoch": 17.77777777777778,
34
+ "grad_norm": 1.7661292552947998,
35
+ "learning_rate": 0.0001,
36
+ "loss": 0.4734,
37
+ "step": 200
38
+ },
39
+ {
40
+ "epoch": 22.22222222222222,
41
+ "grad_norm": 1.4447826147079468,
42
+ "learning_rate": 0.000125,
43
+ "loss": 0.473,
44
+ "step": 250
45
+ },
46
+ {
47
+ "epoch": 26.666666666666668,
48
+ "grad_norm": 2.745311975479126,
49
+ "learning_rate": 0.00015,
50
+ "loss": 0.464,
51
+ "step": 300
52
+ },
53
+ {
54
+ "epoch": 31.11111111111111,
55
+ "grad_norm": 2.4395411014556885,
56
+ "learning_rate": 0.000175,
57
+ "loss": 0.4593,
58
+ "step": 350
59
+ },
60
+ {
61
+ "epoch": 35.55555555555556,
62
+ "grad_norm": 2.1870596408843994,
63
+ "learning_rate": 0.0002,
64
+ "loss": 0.4452,
65
+ "step": 400
66
+ },
67
+ {
68
+ "epoch": 40.0,
69
+ "grad_norm": 2.9709341526031494,
70
+ "learning_rate": 0.00022500000000000002,
71
+ "loss": 0.4404,
72
+ "step": 450
73
+ },
74
+ {
75
+ "epoch": 44.44444444444444,
76
+ "grad_norm": 3.981621265411377,
77
+ "learning_rate": 0.00025,
78
+ "loss": 0.4367,
79
+ "step": 500
80
+ },
81
+ {
82
+ "epoch": 44.44444444444444,
83
+ "eval_loss": 0.41453465819358826,
84
+ "eval_runtime": 9.8143,
85
+ "eval_samples_per_second": 16.201,
86
+ "eval_steps_per_second": 2.038,
87
+ "step": 500
88
+ },
89
+ {
90
+ "epoch": 48.888888888888886,
91
+ "grad_norm": 9.63549518585205,
92
+ "learning_rate": 0.000275,
93
+ "loss": 0.4377,
94
+ "step": 550
95
+ },
96
+ {
97
+ "epoch": 53.333333333333336,
98
+ "grad_norm": 3.0573575496673584,
99
+ "learning_rate": 0.0003,
100
+ "loss": 0.4377,
101
+ "step": 600
102
+ },
103
+ {
104
+ "epoch": 57.77777777777778,
105
+ "grad_norm": 2.144299268722534,
106
+ "learning_rate": 0.00032500000000000004,
107
+ "loss": 0.4342,
108
+ "step": 650
109
+ },
110
+ {
111
+ "epoch": 62.22222222222222,
112
+ "grad_norm": 3.523078203201294,
113
+ "learning_rate": 0.00035,
114
+ "loss": 0.4375,
115
+ "step": 700
116
+ },
117
+ {
118
+ "epoch": 66.66666666666667,
119
+ "grad_norm": 3.0409202575683594,
120
+ "learning_rate": 0.000375,
121
+ "loss": 0.4382,
122
+ "step": 750
123
+ },
124
+ {
125
+ "epoch": 71.11111111111111,
126
+ "grad_norm": 7.679955959320068,
127
+ "learning_rate": 0.0004,
128
+ "loss": 0.4435,
129
+ "step": 800
130
+ },
131
+ {
132
+ "epoch": 75.55555555555556,
133
+ "grad_norm": 4.8538618087768555,
134
+ "learning_rate": 0.000425,
135
+ "loss": 0.4313,
136
+ "step": 850
137
+ },
138
+ {
139
+ "epoch": 80.0,
140
+ "grad_norm": 3.889315128326416,
141
+ "learning_rate": 0.00045000000000000004,
142
+ "loss": 0.4242,
143
+ "step": 900
144
+ },
145
+ {
146
+ "epoch": 84.44444444444444,
147
+ "grad_norm": 9.169000625610352,
148
+ "learning_rate": 0.000475,
149
+ "loss": 0.4703,
150
+ "step": 950
151
+ },
152
+ {
153
+ "epoch": 88.88888888888889,
154
+ "grad_norm": 3.6193642616271973,
155
+ "learning_rate": 0.0005,
156
+ "loss": 0.4709,
157
+ "step": 1000
158
+ },
159
+ {
160
+ "epoch": 88.88888888888889,
161
+ "eval_loss": 0.4963502883911133,
162
+ "eval_runtime": 9.099,
163
+ "eval_samples_per_second": 17.474,
164
+ "eval_steps_per_second": 2.198,
165
+ "step": 1000
166
+ },
167
+ {
168
+ "epoch": 93.33333333333333,
169
+ "grad_norm": 1.7234474420547485,
170
+ "learning_rate": 0.0005250000000000001,
171
+ "loss": 0.4563,
172
+ "step": 1050
173
+ },
174
+ {
175
+ "epoch": 97.77777777777777,
176
+ "grad_norm": 2.933211326599121,
177
+ "learning_rate": 0.00055,
178
+ "loss": 0.4449,
179
+ "step": 1100
180
+ },
181
+ {
182
+ "epoch": 102.22222222222223,
183
+ "grad_norm": 2.3853394985198975,
184
+ "learning_rate": 0.000575,
185
+ "loss": 0.4712,
186
+ "step": 1150
187
+ },
188
+ {
189
+ "epoch": 106.66666666666667,
190
+ "grad_norm": 0.20497646927833557,
191
+ "learning_rate": 0.000599,
192
+ "loss": 1.534,
193
+ "step": 1200
194
+ },
195
+ {
196
+ "epoch": 111.11111111111111,
197
+ "grad_norm": 0.0945451632142067,
198
+ "learning_rate": 0.000624,
199
+ "loss": 1.5935,
200
+ "step": 1250
201
+ },
202
+ {
203
+ "epoch": 115.55555555555556,
204
+ "grad_norm": 0.13917936384677887,
205
+ "learning_rate": 0.0006490000000000001,
206
+ "loss": 1.5906,
207
+ "step": 1300
208
+ },
209
+ {
210
+ "epoch": 120.0,
211
+ "grad_norm": 0.21496649086475372,
212
+ "learning_rate": 0.000674,
213
+ "loss": 1.5817,
214
+ "step": 1350
215
+ },
216
+ {
217
+ "epoch": 124.44444444444444,
218
+ "grad_norm": 0.7981905341148376,
219
+ "learning_rate": 0.000699,
220
+ "loss": 1.5851,
221
+ "step": 1400
222
+ },
223
+ {
224
+ "epoch": 128.88888888888889,
225
+ "grad_norm": 0.3589198589324951,
226
+ "learning_rate": 0.000724,
227
+ "loss": 1.5778,
228
+ "step": 1450
229
+ },
230
+ {
231
+ "epoch": 133.33333333333334,
232
+ "grad_norm": 0.15568029880523682,
233
+ "learning_rate": 0.000749,
234
+ "loss": 1.5891,
235
+ "step": 1500
236
+ },
237
+ {
238
+ "epoch": 133.33333333333334,
239
+ "eval_loss": 1.5470129251480103,
240
+ "eval_runtime": 10.6543,
241
+ "eval_samples_per_second": 14.924,
242
+ "eval_steps_per_second": 1.877,
243
+ "step": 1500
244
+ },
245
+ {
246
+ "epoch": 137.77777777777777,
247
+ "grad_norm": 0.12888091802597046,
248
+ "learning_rate": 0.0007740000000000001,
249
+ "loss": 1.5836,
250
+ "step": 1550
251
+ },
252
+ {
253
+ "epoch": 142.22222222222223,
254
+ "grad_norm": 0.17486007511615753,
255
+ "learning_rate": 0.000799,
256
+ "loss": 1.5765,
257
+ "step": 1600
258
+ },
259
+ {
260
+ "epoch": 146.66666666666666,
261
+ "grad_norm": 0.07815095037221909,
262
+ "learning_rate": 0.000824,
263
+ "loss": 1.5771,
264
+ "step": 1650
265
+ },
266
+ {
267
+ "epoch": 151.11111111111111,
268
+ "grad_norm": 0.8746868371963501,
269
+ "learning_rate": 0.000849,
270
+ "loss": 1.5796,
271
+ "step": 1700
272
+ },
273
+ {
274
+ "epoch": 155.55555555555554,
275
+ "grad_norm": 0.30329012870788574,
276
+ "learning_rate": 0.000874,
277
+ "loss": 1.5779,
278
+ "step": 1750
279
+ },
280
+ {
281
+ "epoch": 160.0,
282
+ "grad_norm": 0.3016453981399536,
283
+ "learning_rate": 0.0008990000000000001,
284
+ "loss": 1.5766,
285
+ "step": 1800
286
+ },
287
+ {
288
+ "epoch": 164.44444444444446,
289
+ "grad_norm": 0.12022041529417038,
290
+ "learning_rate": 0.000924,
291
+ "loss": 1.5775,
292
+ "step": 1850
293
+ },
294
+ {
295
+ "epoch": 168.88888888888889,
296
+ "grad_norm": 0.20793229341506958,
297
+ "learning_rate": 0.000949,
298
+ "loss": 1.5768,
299
+ "step": 1900
300
+ },
301
+ {
302
+ "epoch": 173.33333333333334,
303
+ "grad_norm": 0.03221210092306137,
304
+ "learning_rate": 0.000974,
305
+ "loss": 1.5779,
306
+ "step": 1950
307
+ },
308
+ {
309
+ "epoch": 177.77777777777777,
310
+ "grad_norm": 0.22706760466098785,
311
+ "learning_rate": 0.000999,
312
+ "loss": 1.5787,
313
+ "step": 2000
314
+ },
315
+ {
316
+ "epoch": 177.77777777777777,
317
+ "eval_loss": 1.5496832132339478,
318
+ "eval_runtime": 8.1644,
319
+ "eval_samples_per_second": 19.475,
320
+ "eval_steps_per_second": 2.45,
321
+ "step": 2000
322
+ },
323
+ {
324
+ "epoch": 182.22222222222223,
325
+ "grad_norm": 0.13995479047298431,
326
+ "learning_rate": 0.000976,
327
+ "loss": 1.5783,
328
+ "step": 2050
329
+ },
330
+ {
331
+ "epoch": 186.66666666666666,
332
+ "grad_norm": 0.19275836646556854,
333
+ "learning_rate": 0.000951,
334
+ "loss": 1.5748,
335
+ "step": 2100
336
+ },
337
+ {
338
+ "epoch": 191.11111111111111,
339
+ "grad_norm": 0.20744530856609344,
340
+ "learning_rate": 0.0009260000000000001,
341
+ "loss": 1.578,
342
+ "step": 2150
343
+ },
344
+ {
345
+ "epoch": 195.55555555555554,
346
+ "grad_norm": 0.08092151582241058,
347
+ "learning_rate": 0.000901,
348
+ "loss": 1.5765,
349
+ "step": 2200
350
+ },
351
+ {
352
+ "epoch": 200.0,
353
+ "grad_norm": 0.1601410210132599,
354
+ "learning_rate": 0.000876,
355
+ "loss": 1.5751,
356
+ "step": 2250
357
+ },
358
+ {
359
+ "epoch": 204.44444444444446,
360
+ "grad_norm": 0.20833778381347656,
361
+ "learning_rate": 0.000851,
362
+ "loss": 1.5723,
363
+ "step": 2300
364
+ },
365
+ {
366
+ "epoch": 208.88888888888889,
367
+ "grad_norm": 0.3090430498123169,
368
+ "learning_rate": 0.000826,
369
+ "loss": 1.5741,
370
+ "step": 2350
371
+ },
372
+ {
373
+ "epoch": 213.33333333333334,
374
+ "grad_norm": 0.049218811094760895,
375
+ "learning_rate": 0.0008010000000000001,
376
+ "loss": 1.5743,
377
+ "step": 2400
378
+ },
379
+ {
380
+ "epoch": 217.77777777777777,
381
+ "grad_norm": 0.10529515147209167,
382
+ "learning_rate": 0.000776,
383
+ "loss": 1.5721,
384
+ "step": 2450
385
+ },
386
+ {
387
+ "epoch": 222.22222222222223,
388
+ "grad_norm": 0.21685709059238434,
389
+ "learning_rate": 0.000751,
390
+ "loss": 1.5714,
391
+ "step": 2500
392
+ },
393
+ {
394
+ "epoch": 222.22222222222223,
395
+ "eval_loss": 1.5523545742034912,
396
+ "eval_runtime": 10.3314,
397
+ "eval_samples_per_second": 15.39,
398
+ "eval_steps_per_second": 1.936,
399
+ "step": 2500
400
+ },
401
+ {
402
+ "epoch": 226.66666666666666,
403
+ "grad_norm": 0.04025653377175331,
404
+ "learning_rate": 0.000726,
405
+ "loss": 1.5708,
406
+ "step": 2550
407
+ },
408
+ {
409
+ "epoch": 231.11111111111111,
410
+ "grad_norm": 0.136004239320755,
411
+ "learning_rate": 0.000701,
412
+ "loss": 1.5707,
413
+ "step": 2600
414
+ },
415
+ {
416
+ "epoch": 235.55555555555554,
417
+ "grad_norm": 0.0881115049123764,
418
+ "learning_rate": 0.0006760000000000001,
419
+ "loss": 1.5696,
420
+ "step": 2650
421
+ },
422
+ {
423
+ "epoch": 240.0,
424
+ "grad_norm": 0.14422671496868134,
425
+ "learning_rate": 0.000651,
426
+ "loss": 1.5722,
427
+ "step": 2700
428
+ },
429
+ {
430
+ "epoch": 244.44444444444446,
431
+ "grad_norm": 0.21736255288124084,
432
+ "learning_rate": 0.000626,
433
+ "loss": 1.5706,
434
+ "step": 2750
435
+ },
436
+ {
437
+ "epoch": 248.88888888888889,
438
+ "grad_norm": 0.0742669627070427,
439
+ "learning_rate": 0.000601,
440
+ "loss": 1.5697,
441
+ "step": 2800
442
+ },
443
+ {
444
+ "epoch": 253.33333333333334,
445
+ "grad_norm": 0.19210708141326904,
446
+ "learning_rate": 0.000576,
447
+ "loss": 1.5688,
448
+ "step": 2850
449
+ },
450
+ {
451
+ "epoch": 257.77777777777777,
452
+ "grad_norm": 0.04399856552481651,
453
+ "learning_rate": 0.0005510000000000001,
454
+ "loss": 1.5703,
455
+ "step": 2900
456
+ },
457
+ {
458
+ "epoch": 262.22222222222223,
459
+ "grad_norm": 0.09989127516746521,
460
+ "learning_rate": 0.000526,
461
+ "loss": 1.57,
462
+ "step": 2950
463
+ },
464
+ {
465
+ "epoch": 266.6666666666667,
466
+ "grad_norm": 0.025855202227830887,
467
+ "learning_rate": 0.000501,
468
+ "loss": 1.5715,
469
+ "step": 3000
470
+ },
471
+ {
472
+ "epoch": 266.6666666666667,
473
+ "eval_loss": 1.5577107667922974,
474
+ "eval_runtime": 9.4476,
475
+ "eval_samples_per_second": 16.83,
476
+ "eval_steps_per_second": 2.117,
477
+ "step": 3000
478
+ },
479
+ {
480
+ "epoch": 271.1111111111111,
481
+ "grad_norm": 0.05888780951499939,
482
+ "learning_rate": 0.00047599999999999997,
483
+ "loss": 1.5687,
484
+ "step": 3050
485
+ },
486
+ {
487
+ "epoch": 275.55555555555554,
488
+ "grad_norm": 0.18684841692447662,
489
+ "learning_rate": 0.000451,
490
+ "loss": 1.5693,
491
+ "step": 3100
492
+ },
493
+ {
494
+ "epoch": 280.0,
495
+ "grad_norm": 0.06679928302764893,
496
+ "learning_rate": 0.000426,
497
+ "loss": 1.5685,
498
+ "step": 3150
499
+ },
500
+ {
501
+ "epoch": 284.44444444444446,
502
+ "grad_norm": 0.19672997295856476,
503
+ "learning_rate": 0.00040100000000000004,
504
+ "loss": 1.5699,
505
+ "step": 3200
506
+ },
507
+ {
508
+ "epoch": 288.8888888888889,
509
+ "grad_norm": 0.06303097307682037,
510
+ "learning_rate": 0.00037600000000000003,
511
+ "loss": 1.5668,
512
+ "step": 3250
513
+ },
514
+ {
515
+ "epoch": 293.3333333333333,
516
+ "grad_norm": 0.04833269119262695,
517
+ "learning_rate": 0.00035099999999999997,
518
+ "loss": 1.5702,
519
+ "step": 3300
520
+ },
521
+ {
522
+ "epoch": 297.77777777777777,
523
+ "grad_norm": 0.11464402824640274,
524
+ "learning_rate": 0.000326,
525
+ "loss": 1.5663,
526
+ "step": 3350
527
+ },
528
+ {
529
+ "epoch": 302.22222222222223,
530
+ "grad_norm": 0.06809722632169724,
531
+ "learning_rate": 0.000301,
532
+ "loss": 1.569,
533
+ "step": 3400
534
+ },
535
+ {
536
+ "epoch": 306.6666666666667,
537
+ "grad_norm": 0.08639124035835266,
538
+ "learning_rate": 0.00027600000000000004,
539
+ "loss": 1.5694,
540
+ "step": 3450
541
+ },
542
+ {
543
+ "epoch": 311.1111111111111,
544
+ "grad_norm": 0.06652398407459259,
545
+ "learning_rate": 0.00025100000000000003,
546
+ "loss": 1.5689,
547
+ "step": 3500
548
+ },
549
+ {
550
+ "epoch": 311.1111111111111,
551
+ "eval_loss": 1.5573688745498657,
552
+ "eval_runtime": 9.3223,
553
+ "eval_samples_per_second": 17.056,
554
+ "eval_steps_per_second": 2.145,
555
+ "step": 3500
556
+ }
557
+ ],
558
+ "logging_steps": 50,
559
+ "max_steps": 4000,
560
+ "num_input_tokens_seen": 0,
561
+ "num_train_epochs": 364,
562
+ "save_steps": 500,
563
+ "stateful_callbacks": {
564
+ "TrainerControl": {
565
+ "args": {
566
+ "should_epoch_stop": false,
567
+ "should_evaluate": false,
568
+ "should_log": false,
569
+ "should_save": true,
570
+ "should_training_stop": false
571
+ },
572
+ "attributes": {}
573
+ }
574
+ },
575
+ "total_flos": 9.635141664709862e+16,
576
+ "train_batch_size": 16,
577
+ "trial_name": null,
578
+ "trial_params": null
579
+ }
checkpoint-3500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b92860c5badf59be96fc9bca70be2df1eea12cf3dbb359d406d83b5b23018b16
3
+ size 5304