mikhail-panzo commited on
Commit
93bf4e9
1 Parent(s): a3d2b6c

Training in progress, step 3000, checkpoint

Browse files
checkpoint-3000/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<ctc_blank>": 80,
3
+ "<mask>": 79
4
+ }
checkpoint-3000/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-3000/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-3000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6114d6f71b31641185ddd01c77a0517af34625affe672736bf6952533157d839
3
+ size 577789320
checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dcb0c8faa0431a31472ad29dc4dc301881337a211910efcdf52a5df8efc92eb1
3
+ size 1155772233
checkpoint-3000/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-3000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d3fc7c17d487270f27e2c3733adda12b96d6505fca14729858402b855f37bf4
3
+ size 14244
checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8250f51ac088c98aa252ef63c5068afe9e5763d01e79ba9eedd60c569b7a576d
3
+ size 1064
checkpoint-3000/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-3000/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-3000/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-3000/trainer_state.json ADDED
@@ -0,0 +1,501 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.3674115538597107,
3
+ "best_model_checkpoint": "mikhail-panzo/zlm_b32_le4_s8000/checkpoint-3000",
4
+ "epoch": 1.256544502617801,
5
+ "eval_steps": 500,
6
+ "global_step": 3000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.020942408376963352,
13
+ "grad_norm": 3.1775436401367188,
14
+ "learning_rate": 2.35e-06,
15
+ "loss": 1.1228,
16
+ "step": 50
17
+ },
18
+ {
19
+ "epoch": 0.041884816753926704,
20
+ "grad_norm": 4.18980073928833,
21
+ "learning_rate": 4.85e-06,
22
+ "loss": 0.8304,
23
+ "step": 100
24
+ },
25
+ {
26
+ "epoch": 0.06282722513089005,
27
+ "grad_norm": 4.5355024337768555,
28
+ "learning_rate": 7.35e-06,
29
+ "loss": 0.7701,
30
+ "step": 150
31
+ },
32
+ {
33
+ "epoch": 0.08376963350785341,
34
+ "grad_norm": 3.9590957164764404,
35
+ "learning_rate": 9.85e-06,
36
+ "loss": 0.7096,
37
+ "step": 200
38
+ },
39
+ {
40
+ "epoch": 0.10471204188481675,
41
+ "grad_norm": 5.419675350189209,
42
+ "learning_rate": 1.235e-05,
43
+ "loss": 0.6518,
44
+ "step": 250
45
+ },
46
+ {
47
+ "epoch": 0.1256544502617801,
48
+ "grad_norm": 3.1052777767181396,
49
+ "learning_rate": 1.485e-05,
50
+ "loss": 0.6326,
51
+ "step": 300
52
+ },
53
+ {
54
+ "epoch": 0.14659685863874344,
55
+ "grad_norm": 3.4649453163146973,
56
+ "learning_rate": 1.7349999999999998e-05,
57
+ "loss": 0.6241,
58
+ "step": 350
59
+ },
60
+ {
61
+ "epoch": 0.16753926701570682,
62
+ "grad_norm": 4.962624549865723,
63
+ "learning_rate": 1.985e-05,
64
+ "loss": 0.5723,
65
+ "step": 400
66
+ },
67
+ {
68
+ "epoch": 0.18848167539267016,
69
+ "grad_norm": 2.7586095333099365,
70
+ "learning_rate": 2.235e-05,
71
+ "loss": 0.5764,
72
+ "step": 450
73
+ },
74
+ {
75
+ "epoch": 0.2094240837696335,
76
+ "grad_norm": 3.5610299110412598,
77
+ "learning_rate": 2.485e-05,
78
+ "loss": 0.5552,
79
+ "step": 500
80
+ },
81
+ {
82
+ "epoch": 0.2094240837696335,
83
+ "eval_loss": 0.48825809359550476,
84
+ "eval_runtime": 277.5892,
85
+ "eval_samples_per_second": 30.581,
86
+ "eval_steps_per_second": 3.826,
87
+ "step": 500
88
+ },
89
+ {
90
+ "epoch": 0.23036649214659685,
91
+ "grad_norm": 3.3351809978485107,
92
+ "learning_rate": 2.7350000000000004e-05,
93
+ "loss": 0.5322,
94
+ "step": 550
95
+ },
96
+ {
97
+ "epoch": 0.2513089005235602,
98
+ "grad_norm": 3.7222373485565186,
99
+ "learning_rate": 2.985e-05,
100
+ "loss": 0.5186,
101
+ "step": 600
102
+ },
103
+ {
104
+ "epoch": 0.27225130890052357,
105
+ "grad_norm": 3.1811156272888184,
106
+ "learning_rate": 3.235e-05,
107
+ "loss": 0.5157,
108
+ "step": 650
109
+ },
110
+ {
111
+ "epoch": 0.2931937172774869,
112
+ "grad_norm": 2.2585642337799072,
113
+ "learning_rate": 3.485e-05,
114
+ "loss": 0.5089,
115
+ "step": 700
116
+ },
117
+ {
118
+ "epoch": 0.31413612565445026,
119
+ "grad_norm": 3.999460220336914,
120
+ "learning_rate": 3.735e-05,
121
+ "loss": 0.5117,
122
+ "step": 750
123
+ },
124
+ {
125
+ "epoch": 0.33507853403141363,
126
+ "grad_norm": 3.739990472793579,
127
+ "learning_rate": 3.9850000000000006e-05,
128
+ "loss": 0.5031,
129
+ "step": 800
130
+ },
131
+ {
132
+ "epoch": 0.35602094240837695,
133
+ "grad_norm": 4.251980781555176,
134
+ "learning_rate": 4.235e-05,
135
+ "loss": 0.5064,
136
+ "step": 850
137
+ },
138
+ {
139
+ "epoch": 0.3769633507853403,
140
+ "grad_norm": 2.770602226257324,
141
+ "learning_rate": 4.4850000000000006e-05,
142
+ "loss": 0.4921,
143
+ "step": 900
144
+ },
145
+ {
146
+ "epoch": 0.39790575916230364,
147
+ "grad_norm": 2.506974220275879,
148
+ "learning_rate": 4.735e-05,
149
+ "loss": 0.4839,
150
+ "step": 950
151
+ },
152
+ {
153
+ "epoch": 0.418848167539267,
154
+ "grad_norm": 2.2666189670562744,
155
+ "learning_rate": 4.9850000000000006e-05,
156
+ "loss": 0.4913,
157
+ "step": 1000
158
+ },
159
+ {
160
+ "epoch": 0.418848167539267,
161
+ "eval_loss": 0.4265913963317871,
162
+ "eval_runtime": 265.1599,
163
+ "eval_samples_per_second": 32.015,
164
+ "eval_steps_per_second": 4.005,
165
+ "step": 1000
166
+ },
167
+ {
168
+ "epoch": 0.4397905759162304,
169
+ "grad_norm": 3.1451058387756348,
170
+ "learning_rate": 5.235e-05,
171
+ "loss": 0.4722,
172
+ "step": 1050
173
+ },
174
+ {
175
+ "epoch": 0.4607329842931937,
176
+ "grad_norm": 3.197997570037842,
177
+ "learning_rate": 5.485e-05,
178
+ "loss": 0.4692,
179
+ "step": 1100
180
+ },
181
+ {
182
+ "epoch": 0.4816753926701571,
183
+ "grad_norm": 2.9112601280212402,
184
+ "learning_rate": 5.7350000000000005e-05,
185
+ "loss": 0.4738,
186
+ "step": 1150
187
+ },
188
+ {
189
+ "epoch": 0.5026178010471204,
190
+ "grad_norm": 3.036731243133545,
191
+ "learning_rate": 5.9850000000000005e-05,
192
+ "loss": 0.4521,
193
+ "step": 1200
194
+ },
195
+ {
196
+ "epoch": 0.5235602094240838,
197
+ "grad_norm": 5.021958351135254,
198
+ "learning_rate": 6.235000000000001e-05,
199
+ "loss": 0.4666,
200
+ "step": 1250
201
+ },
202
+ {
203
+ "epoch": 0.5445026178010471,
204
+ "grad_norm": 3.302204132080078,
205
+ "learning_rate": 6.485e-05,
206
+ "loss": 0.4625,
207
+ "step": 1300
208
+ },
209
+ {
210
+ "epoch": 0.5654450261780105,
211
+ "grad_norm": 3.2643635272979736,
212
+ "learning_rate": 6.735e-05,
213
+ "loss": 0.4683,
214
+ "step": 1350
215
+ },
216
+ {
217
+ "epoch": 0.5863874345549738,
218
+ "grad_norm": 1.7499467134475708,
219
+ "learning_rate": 6.985e-05,
220
+ "loss": 0.449,
221
+ "step": 1400
222
+ },
223
+ {
224
+ "epoch": 0.6073298429319371,
225
+ "grad_norm": 1.3616622686386108,
226
+ "learning_rate": 7.235000000000001e-05,
227
+ "loss": 0.4523,
228
+ "step": 1450
229
+ },
230
+ {
231
+ "epoch": 0.6282722513089005,
232
+ "grad_norm": 2.5826191902160645,
233
+ "learning_rate": 7.485e-05,
234
+ "loss": 0.446,
235
+ "step": 1500
236
+ },
237
+ {
238
+ "epoch": 0.6282722513089005,
239
+ "eval_loss": 0.3975289463996887,
240
+ "eval_runtime": 266.4222,
241
+ "eval_samples_per_second": 31.863,
242
+ "eval_steps_per_second": 3.986,
243
+ "step": 1500
244
+ },
245
+ {
246
+ "epoch": 0.6492146596858639,
247
+ "grad_norm": 3.6052303314208984,
248
+ "learning_rate": 7.735e-05,
249
+ "loss": 0.4449,
250
+ "step": 1550
251
+ },
252
+ {
253
+ "epoch": 0.6701570680628273,
254
+ "grad_norm": 3.4120566844940186,
255
+ "learning_rate": 7.985e-05,
256
+ "loss": 0.4477,
257
+ "step": 1600
258
+ },
259
+ {
260
+ "epoch": 0.6910994764397905,
261
+ "grad_norm": 2.187040090560913,
262
+ "learning_rate": 8.235000000000001e-05,
263
+ "loss": 0.4522,
264
+ "step": 1650
265
+ },
266
+ {
267
+ "epoch": 0.7120418848167539,
268
+ "grad_norm": 1.718518853187561,
269
+ "learning_rate": 8.485e-05,
270
+ "loss": 0.4431,
271
+ "step": 1700
272
+ },
273
+ {
274
+ "epoch": 0.7329842931937173,
275
+ "grad_norm": 1.8248894214630127,
276
+ "learning_rate": 8.735000000000001e-05,
277
+ "loss": 0.4496,
278
+ "step": 1750
279
+ },
280
+ {
281
+ "epoch": 0.7539267015706806,
282
+ "grad_norm": 2.733355760574341,
283
+ "learning_rate": 8.985e-05,
284
+ "loss": 0.4297,
285
+ "step": 1800
286
+ },
287
+ {
288
+ "epoch": 0.774869109947644,
289
+ "grad_norm": 2.5899884700775146,
290
+ "learning_rate": 9.235000000000001e-05,
291
+ "loss": 0.4588,
292
+ "step": 1850
293
+ },
294
+ {
295
+ "epoch": 0.7958115183246073,
296
+ "grad_norm": 1.5704914331436157,
297
+ "learning_rate": 9.485e-05,
298
+ "loss": 0.4345,
299
+ "step": 1900
300
+ },
301
+ {
302
+ "epoch": 0.8167539267015707,
303
+ "grad_norm": 5.18487548828125,
304
+ "learning_rate": 9.735000000000001e-05,
305
+ "loss": 0.4325,
306
+ "step": 1950
307
+ },
308
+ {
309
+ "epoch": 0.837696335078534,
310
+ "grad_norm": 2.4392499923706055,
311
+ "learning_rate": 9.985000000000001e-05,
312
+ "loss": 0.4222,
313
+ "step": 2000
314
+ },
315
+ {
316
+ "epoch": 0.837696335078534,
317
+ "eval_loss": 0.39489272236824036,
318
+ "eval_runtime": 267.0393,
319
+ "eval_samples_per_second": 31.789,
320
+ "eval_steps_per_second": 3.977,
321
+ "step": 2000
322
+ },
323
+ {
324
+ "epoch": 0.8586387434554974,
325
+ "grad_norm": 2.1138336658477783,
326
+ "learning_rate": 9.921666666666668e-05,
327
+ "loss": 0.4438,
328
+ "step": 2050
329
+ },
330
+ {
331
+ "epoch": 0.8795811518324608,
332
+ "grad_norm": 2.3158669471740723,
333
+ "learning_rate": 9.838333333333334e-05,
334
+ "loss": 0.421,
335
+ "step": 2100
336
+ },
337
+ {
338
+ "epoch": 0.900523560209424,
339
+ "grad_norm": 2.1935179233551025,
340
+ "learning_rate": 9.755000000000001e-05,
341
+ "loss": 0.4247,
342
+ "step": 2150
343
+ },
344
+ {
345
+ "epoch": 0.9214659685863874,
346
+ "grad_norm": 1.7845231294631958,
347
+ "learning_rate": 9.671666666666667e-05,
348
+ "loss": 0.4203,
349
+ "step": 2200
350
+ },
351
+ {
352
+ "epoch": 0.9424083769633508,
353
+ "grad_norm": 1.7287964820861816,
354
+ "learning_rate": 9.588333333333334e-05,
355
+ "loss": 0.4306,
356
+ "step": 2250
357
+ },
358
+ {
359
+ "epoch": 0.9633507853403142,
360
+ "grad_norm": 2.2276289463043213,
361
+ "learning_rate": 9.505e-05,
362
+ "loss": 0.4392,
363
+ "step": 2300
364
+ },
365
+ {
366
+ "epoch": 0.9842931937172775,
367
+ "grad_norm": 2.248389720916748,
368
+ "learning_rate": 9.421666666666668e-05,
369
+ "loss": 0.4304,
370
+ "step": 2350
371
+ },
372
+ {
373
+ "epoch": 1.0052356020942408,
374
+ "grad_norm": 1.5984878540039062,
375
+ "learning_rate": 9.338333333333333e-05,
376
+ "loss": 0.4125,
377
+ "step": 2400
378
+ },
379
+ {
380
+ "epoch": 1.0261780104712042,
381
+ "grad_norm": 2.6061534881591797,
382
+ "learning_rate": 9.255e-05,
383
+ "loss": 0.4229,
384
+ "step": 2450
385
+ },
386
+ {
387
+ "epoch": 1.0471204188481675,
388
+ "grad_norm": 2.127110481262207,
389
+ "learning_rate": 9.171666666666667e-05,
390
+ "loss": 0.4273,
391
+ "step": 2500
392
+ },
393
+ {
394
+ "epoch": 1.0471204188481675,
395
+ "eval_loss": 0.38292551040649414,
396
+ "eval_runtime": 263.8923,
397
+ "eval_samples_per_second": 32.168,
398
+ "eval_steps_per_second": 4.024,
399
+ "step": 2500
400
+ },
401
+ {
402
+ "epoch": 1.068062827225131,
403
+ "grad_norm": 1.9867545366287231,
404
+ "learning_rate": 9.088333333333334e-05,
405
+ "loss": 0.4164,
406
+ "step": 2550
407
+ },
408
+ {
409
+ "epoch": 1.0890052356020943,
410
+ "grad_norm": 2.1518051624298096,
411
+ "learning_rate": 9.005000000000001e-05,
412
+ "loss": 0.4038,
413
+ "step": 2600
414
+ },
415
+ {
416
+ "epoch": 1.1099476439790577,
417
+ "grad_norm": 2.1266932487487793,
418
+ "learning_rate": 8.921666666666668e-05,
419
+ "loss": 0.418,
420
+ "step": 2650
421
+ },
422
+ {
423
+ "epoch": 1.130890052356021,
424
+ "grad_norm": 1.9351108074188232,
425
+ "learning_rate": 8.838333333333334e-05,
426
+ "loss": 0.4213,
427
+ "step": 2700
428
+ },
429
+ {
430
+ "epoch": 1.1518324607329844,
431
+ "grad_norm": 1.4189810752868652,
432
+ "learning_rate": 8.755e-05,
433
+ "loss": 0.4201,
434
+ "step": 2750
435
+ },
436
+ {
437
+ "epoch": 1.1727748691099475,
438
+ "grad_norm": 3.590930223464966,
439
+ "learning_rate": 8.671666666666667e-05,
440
+ "loss": 0.4043,
441
+ "step": 2800
442
+ },
443
+ {
444
+ "epoch": 1.193717277486911,
445
+ "grad_norm": 3.049971103668213,
446
+ "learning_rate": 8.588333333333334e-05,
447
+ "loss": 0.4036,
448
+ "step": 2850
449
+ },
450
+ {
451
+ "epoch": 1.2146596858638743,
452
+ "grad_norm": 2.1481425762176514,
453
+ "learning_rate": 8.505000000000001e-05,
454
+ "loss": 0.4067,
455
+ "step": 2900
456
+ },
457
+ {
458
+ "epoch": 1.2356020942408377,
459
+ "grad_norm": 1.774927020072937,
460
+ "learning_rate": 8.421666666666666e-05,
461
+ "loss": 0.4126,
462
+ "step": 2950
463
+ },
464
+ {
465
+ "epoch": 1.256544502617801,
466
+ "grad_norm": 1.624089241027832,
467
+ "learning_rate": 8.338333333333333e-05,
468
+ "loss": 0.4028,
469
+ "step": 3000
470
+ },
471
+ {
472
+ "epoch": 1.256544502617801,
473
+ "eval_loss": 0.3674115538597107,
474
+ "eval_runtime": 263.8647,
475
+ "eval_samples_per_second": 32.172,
476
+ "eval_steps_per_second": 4.025,
477
+ "step": 3000
478
+ }
479
+ ],
480
+ "logging_steps": 50,
481
+ "max_steps": 8000,
482
+ "num_input_tokens_seen": 0,
483
+ "num_train_epochs": 4,
484
+ "save_steps": 500,
485
+ "stateful_callbacks": {
486
+ "TrainerControl": {
487
+ "args": {
488
+ "should_epoch_stop": false,
489
+ "should_evaluate": false,
490
+ "should_log": false,
491
+ "should_save": true,
492
+ "should_training_stop": false
493
+ },
494
+ "attributes": {}
495
+ }
496
+ },
497
+ "total_flos": 1.3431343807297152e+16,
498
+ "train_batch_size": 16,
499
+ "trial_name": null,
500
+ "trial_params": null
501
+ }
checkpoint-3000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f8ba8e957961bfac2918c828bd8fb1b1b5848bc9b1402a9fd5cd5021a38c66f
3
+ size 5304