AmelieSchreiber commited on
Commit
1d52dca
1 Parent(s): 64d929d

Upload 9 files

Browse files
adapter_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "facebook/esm2_t30_150M_UR50D",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "lora_alpha": 1,
12
+ "lora_dropout": 0.5,
13
+ "modules_to_save": null,
14
+ "peft_type": "LORA",
15
+ "r": 2,
16
+ "rank_pattern": {},
17
+ "revision": null,
18
+ "target_modules": [
19
+ "query",
20
+ "key",
21
+ "value",
22
+ "EsmSelfOutput.dense",
23
+ "EsmIntermediate.dense",
24
+ "EsmOutput.dense",
25
+ "EsmContactPredictionHead.regression",
26
+ "classifier"
27
+ ],
28
+ "task_type": "TOKEN_CLS"
29
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0814934d0faa6ad72b35e2b42e8c8e694a7ea22a7a8bedee8623562e725682a6
3
+ size 1005173
added_tokens.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "<cls>": 0,
3
+ "<eos>": 2,
4
+ "<mask>": 32,
5
+ "<pad>": 1,
6
+ "<unk>": 3
7
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd0adf0abd4a194d010de60f1c9fcee6c41a0f91f673810b8ef8c3306b7a97e2
3
+ size 1972101
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1d29d05640260032ada5358990143a726e529f2bc25ffcc229bf49f4c28a5d8
3
+ size 14575
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45b34c91f200cc98f7a62116ce2d407e234d647df7c68711d4ddf838d715c289
3
+ size 627
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "<cls>",
3
+ "eos_token": "<eos>",
4
+ "mask_token": "<mask>",
5
+ "pad_token": "<pad>",
6
+ "unk_token": "<unk>"
7
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<cls>",
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": "<eos>",
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
+ "32": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "additional_special_tokens": [],
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "<cls>",
47
+ "eos_token": "<eos>",
48
+ "mask_token": "<mask>",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "<pad>",
51
+ "tokenizer_class": "EsmTokenizer",
52
+ "tokenizer_file": null,
53
+ "unk_token": "<unk>"
54
+ }
trainer_state.json ADDED
@@ -0,0 +1,1497 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.15952083840732592,
3
+ "best_model_checkpoint": "esm2_t30_150M_qlora_ptm_sites_2023-10-19_22-31-34/checkpoint-48919",
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 48919,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 0.0003701416918864124,
14
+ "loss": 0.2635,
15
+ "step": 200
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 0.00037009604913071327,
20
+ "loss": 0.0714,
21
+ "step": 400
22
+ },
23
+ {
24
+ "epoch": 0.01,
25
+ "learning_rate": 0.00037001988446013227,
26
+ "loss": 0.056,
27
+ "step": 600
28
+ },
29
+ {
30
+ "epoch": 0.02,
31
+ "learning_rate": 0.00036991381967446495,
32
+ "loss": 0.0526,
33
+ "step": 800
34
+ },
35
+ {
36
+ "epoch": 0.02,
37
+ "learning_rate": 0.00036977680630794575,
38
+ "loss": 0.0411,
39
+ "step": 1000
40
+ },
41
+ {
42
+ "epoch": 0.02,
43
+ "learning_rate": 0.00036960932369126316,
44
+ "loss": 0.0415,
45
+ "step": 1200
46
+ },
47
+ {
48
+ "epoch": 0.03,
49
+ "learning_rate": 0.00036941139945362056,
50
+ "loss": 0.0338,
51
+ "step": 1400
52
+ },
53
+ {
54
+ "epoch": 0.03,
55
+ "learning_rate": 0.00036918306624610206,
56
+ "loss": 0.029,
57
+ "step": 1600
58
+ },
59
+ {
60
+ "epoch": 0.04,
61
+ "learning_rate": 0.0003689243617362863,
62
+ "loss": 0.0333,
63
+ "step": 1800
64
+ },
65
+ {
66
+ "epoch": 0.04,
67
+ "learning_rate": 0.0003686353286020322,
68
+ "loss": 0.0308,
69
+ "step": 2000
70
+ },
71
+ {
72
+ "epoch": 0.04,
73
+ "learning_rate": 0.0003683160145244388,
74
+ "loss": 0.0314,
75
+ "step": 2200
76
+ },
77
+ {
78
+ "epoch": 0.05,
79
+ "learning_rate": 0.00036796647217997937,
80
+ "loss": 0.0243,
81
+ "step": 2400
82
+ },
83
+ {
84
+ "epoch": 0.05,
85
+ "learning_rate": 0.00036758675923181147,
86
+ "loss": 0.027,
87
+ "step": 2600
88
+ },
89
+ {
90
+ "epoch": 0.06,
91
+ "learning_rate": 0.00036717693832026435,
92
+ "loss": 0.0283,
93
+ "step": 2800
94
+ },
95
+ {
96
+ "epoch": 0.06,
97
+ "learning_rate": 0.0003667370770525053,
98
+ "loss": 0.0235,
99
+ "step": 3000
100
+ },
101
+ {
102
+ "epoch": 0.07,
103
+ "learning_rate": 0.0003662672479913868,
104
+ "loss": 0.0223,
105
+ "step": 3200
106
+ },
107
+ {
108
+ "epoch": 0.07,
109
+ "learning_rate": 0.0003657701014579829,
110
+ "loss": 0.0208,
111
+ "step": 3400
112
+ },
113
+ {
114
+ "epoch": 0.07,
115
+ "learning_rate": 0.00036524072308681836,
116
+ "loss": 0.0251,
117
+ "step": 3600
118
+ },
119
+ {
120
+ "epoch": 0.08,
121
+ "learning_rate": 0.00036468162377219915,
122
+ "loss": 0.0198,
123
+ "step": 3800
124
+ },
125
+ {
126
+ "epoch": 0.08,
127
+ "learning_rate": 0.00036409289574739075,
128
+ "loss": 0.0231,
129
+ "step": 4000
130
+ },
131
+ {
132
+ "epoch": 0.09,
133
+ "learning_rate": 0.0003634746361334354,
134
+ "loss": 0.0212,
135
+ "step": 4200
136
+ },
137
+ {
138
+ "epoch": 0.09,
139
+ "learning_rate": 0.0003628269469231305,
140
+ "loss": 0.0203,
141
+ "step": 4400
142
+ },
143
+ {
144
+ "epoch": 0.09,
145
+ "learning_rate": 0.000362149934964203,
146
+ "loss": 0.021,
147
+ "step": 4600
148
+ },
149
+ {
150
+ "epoch": 0.1,
151
+ "learning_rate": 0.0003614437119416829,
152
+ "loss": 0.0241,
153
+ "step": 4800
154
+ },
155
+ {
156
+ "epoch": 0.1,
157
+ "learning_rate": 0.00036070839435947883,
158
+ "loss": 0.0205,
159
+ "step": 5000
160
+ },
161
+ {
162
+ "epoch": 0.11,
163
+ "learning_rate": 0.0003599441035211589,
164
+ "loss": 0.0197,
165
+ "step": 5200
166
+ },
167
+ {
168
+ "epoch": 0.11,
169
+ "learning_rate": 0.00035915096550993934,
170
+ "loss": 0.0201,
171
+ "step": 5400
172
+ },
173
+ {
174
+ "epoch": 0.11,
175
+ "learning_rate": 0.000358329111167885,
176
+ "loss": 0.0214,
177
+ "step": 5600
178
+ },
179
+ {
180
+ "epoch": 0.12,
181
+ "learning_rate": 0.00035747867607432445,
182
+ "loss": 0.0194,
183
+ "step": 5800
184
+ },
185
+ {
186
+ "epoch": 0.12,
187
+ "learning_rate": 0.0003565998005234839,
188
+ "loss": 0.0167,
189
+ "step": 6000
190
+ },
191
+ {
192
+ "epoch": 0.13,
193
+ "learning_rate": 0.0003556926295013433,
194
+ "loss": 0.0234,
195
+ "step": 6200
196
+ },
197
+ {
198
+ "epoch": 0.13,
199
+ "learning_rate": 0.00035475731266171827,
200
+ "loss": 0.0171,
201
+ "step": 6400
202
+ },
203
+ {
204
+ "epoch": 0.13,
205
+ "learning_rate": 0.00035379400430157195,
206
+ "loss": 0.0181,
207
+ "step": 6600
208
+ },
209
+ {
210
+ "epoch": 0.14,
211
+ "learning_rate": 0.00035280286333556107,
212
+ "loss": 0.0194,
213
+ "step": 6800
214
+ },
215
+ {
216
+ "epoch": 0.14,
217
+ "learning_rate": 0.00035178405326982047,
218
+ "loss": 0.0244,
219
+ "step": 7000
220
+ },
221
+ {
222
+ "epoch": 0.15,
223
+ "learning_rate": 0.0003507377421749893,
224
+ "loss": 0.0188,
225
+ "step": 7200
226
+ },
227
+ {
228
+ "epoch": 0.15,
229
+ "learning_rate": 0.00034966410265848546,
230
+ "loss": 0.0186,
231
+ "step": 7400
232
+ },
233
+ {
234
+ "epoch": 0.16,
235
+ "learning_rate": 0.00034856331183603055,
236
+ "loss": 0.0164,
237
+ "step": 7600
238
+ },
239
+ {
240
+ "epoch": 0.16,
241
+ "learning_rate": 0.0003474355513024315,
242
+ "loss": 0.0184,
243
+ "step": 7800
244
+ },
245
+ {
246
+ "epoch": 0.16,
247
+ "learning_rate": 0.0003462810071016236,
248
+ "loss": 0.0154,
249
+ "step": 8000
250
+ },
251
+ {
252
+ "epoch": 0.17,
253
+ "learning_rate": 0.00034509986969597905,
254
+ "loss": 0.0203,
255
+ "step": 8200
256
+ },
257
+ {
258
+ "epoch": 0.17,
259
+ "learning_rate": 0.0003438984369518822,
260
+ "loss": 0.0214,
261
+ "step": 8400
262
+ },
263
+ {
264
+ "epoch": 0.18,
265
+ "learning_rate": 0.00034266483253267676,
266
+ "loss": 0.0166,
267
+ "step": 8600
268
+ },
269
+ {
270
+ "epoch": 0.18,
271
+ "learning_rate": 0.0003414052314602316,
272
+ "loss": 0.017,
273
+ "step": 8800
274
+ },
275
+ {
276
+ "epoch": 0.18,
277
+ "learning_rate": 0.0003401198415278977,
278
+ "loss": 0.0182,
279
+ "step": 9000
280
+ },
281
+ {
282
+ "epoch": 0.19,
283
+ "learning_rate": 0.0003388088747833521,
284
+ "loss": 0.017,
285
+ "step": 9200
286
+ },
287
+ {
288
+ "epoch": 0.19,
289
+ "learning_rate": 0.0003374725474936167,
290
+ "loss": 0.0183,
291
+ "step": 9400
292
+ },
293
+ {
294
+ "epoch": 0.2,
295
+ "learning_rate": 0.0003361110801093815,
296
+ "loss": 0.0168,
297
+ "step": 9600
298
+ },
299
+ {
300
+ "epoch": 0.2,
301
+ "learning_rate": 0.0003347246972286373,
302
+ "loss": 0.0172,
303
+ "step": 9800
304
+ },
305
+ {
306
+ "epoch": 0.2,
307
+ "learning_rate": 0.000333313627559624,
308
+ "loss": 0.0197,
309
+ "step": 10000
310
+ },
311
+ {
312
+ "epoch": 0.21,
313
+ "learning_rate": 0.0003318781038831018,
314
+ "loss": 0.0139,
315
+ "step": 10200
316
+ },
317
+ {
318
+ "epoch": 0.21,
319
+ "learning_rate": 0.0003304183630139492,
320
+ "loss": 0.016,
321
+ "step": 10400
322
+ },
323
+ {
324
+ "epoch": 0.22,
325
+ "learning_rate": 0.0003289346457620967,
326
+ "loss": 0.0205,
327
+ "step": 10600
328
+ },
329
+ {
330
+ "epoch": 0.22,
331
+ "learning_rate": 0.00032742719689280103,
332
+ "loss": 0.0179,
333
+ "step": 10800
334
+ },
335
+ {
336
+ "epoch": 0.22,
337
+ "learning_rate": 0.00032589626508626625,
338
+ "loss": 0.0143,
339
+ "step": 11000
340
+ },
341
+ {
342
+ "epoch": 0.23,
343
+ "learning_rate": 0.00032434210289662013,
344
+ "loss": 0.0175,
345
+ "step": 11200
346
+ },
347
+ {
348
+ "epoch": 0.23,
349
+ "learning_rate": 0.0003227649667102507,
350
+ "loss": 0.0156,
351
+ "step": 11400
352
+ },
353
+ {
354
+ "epoch": 0.24,
355
+ "learning_rate": 0.0003211651167035108,
356
+ "loss": 0.0206,
357
+ "step": 11600
358
+ },
359
+ {
360
+ "epoch": 0.24,
361
+ "learning_rate": 0.0003195428167997974,
362
+ "loss": 0.0176,
363
+ "step": 11800
364
+ },
365
+ {
366
+ "epoch": 0.25,
367
+ "learning_rate": 0.00031789833462601304,
368
+ "loss": 0.0151,
369
+ "step": 12000
370
+ },
371
+ {
372
+ "epoch": 0.25,
373
+ "learning_rate": 0.0003162319414684155,
374
+ "loss": 0.0159,
375
+ "step": 12200
376
+ },
377
+ {
378
+ "epoch": 0.25,
379
+ "learning_rate": 0.00031455240573465427,
380
+ "loss": 0.0168,
381
+ "step": 12400
382
+ },
383
+ {
384
+ "epoch": 0.26,
385
+ "learning_rate": 0.0003128431249707856,
386
+ "loss": 0.0148,
387
+ "step": 12600
388
+ },
389
+ {
390
+ "epoch": 0.26,
391
+ "learning_rate": 0.0003111127671688482,
392
+ "loss": 0.0236,
393
+ "step": 12800
394
+ },
395
+ {
396
+ "epoch": 0.27,
397
+ "learning_rate": 0.00030937042477134576,
398
+ "loss": 0.0146,
399
+ "step": 13000
400
+ },
401
+ {
402
+ "epoch": 0.27,
403
+ "learning_rate": 0.00030759887447136833,
404
+ "loss": 0.0157,
405
+ "step": 13200
406
+ },
407
+ {
408
+ "epoch": 0.27,
409
+ "learning_rate": 0.0003058071122646787,
410
+ "loss": 0.0216,
411
+ "step": 13400
412
+ },
413
+ {
414
+ "epoch": 0.28,
415
+ "learning_rate": 0.00030399543373396665,
416
+ "loss": 0.0171,
417
+ "step": 13600
418
+ },
419
+ {
420
+ "epoch": 0.28,
421
+ "learning_rate": 0.0003021641377474699,
422
+ "loss": 0.0174,
423
+ "step": 13800
424
+ },
425
+ {
426
+ "epoch": 0.29,
427
+ "learning_rate": 0.0003003135264096706,
428
+ "loss": 0.0174,
429
+ "step": 14000
430
+ },
431
+ {
432
+ "epoch": 0.29,
433
+ "learning_rate": 0.0002984439050114576,
434
+ "loss": 0.0155,
435
+ "step": 14200
436
+ },
437
+ {
438
+ "epoch": 0.29,
439
+ "learning_rate": 0.00029655558197976386,
440
+ "loss": 0.0181,
441
+ "step": 14400
442
+ },
443
+ {
444
+ "epoch": 0.3,
445
+ "learning_rate": 0.0002946488688266855,
446
+ "loss": 0.0165,
447
+ "step": 14600
448
+ },
449
+ {
450
+ "epoch": 0.3,
451
+ "learning_rate": 0.00029272408009809267,
452
+ "loss": 0.0186,
453
+ "step": 14800
454
+ },
455
+ {
456
+ "epoch": 0.31,
457
+ "learning_rate": 0.00029078153332173963,
458
+ "loss": 0.0159,
459
+ "step": 15000
460
+ },
461
+ {
462
+ "epoch": 0.31,
463
+ "learning_rate": 0.00028882154895488294,
464
+ "loss": 0.0148,
465
+ "step": 15200
466
+ },
467
+ {
468
+ "epoch": 0.31,
469
+ "learning_rate": 0.00028684445033141644,
470
+ "loss": 0.0159,
471
+ "step": 15400
472
+ },
473
+ {
474
+ "epoch": 0.32,
475
+ "learning_rate": 0.00028485056360853155,
476
+ "loss": 0.015,
477
+ "step": 15600
478
+ },
479
+ {
480
+ "epoch": 0.32,
481
+ "learning_rate": 0.00028284021771291196,
482
+ "loss": 0.0146,
483
+ "step": 15800
484
+ },
485
+ {
486
+ "epoch": 0.33,
487
+ "learning_rate": 0.0002808137442864714,
488
+ "loss": 0.0156,
489
+ "step": 16000
490
+ },
491
+ {
492
+ "epoch": 0.33,
493
+ "learning_rate": 0.0002787817276943337,
494
+ "loss": 0.0154,
495
+ "step": 16200
496
+ },
497
+ {
498
+ "epoch": 0.34,
499
+ "learning_rate": 0.0002767240811582267,
500
+ "loss": 0.0146,
501
+ "step": 16400
502
+ },
503
+ {
504
+ "epoch": 0.34,
505
+ "learning_rate": 0.0002746513160555949,
506
+ "loss": 0.0138,
507
+ "step": 16600
508
+ },
509
+ {
510
+ "epoch": 0.34,
511
+ "learning_rate": 0.0002725637743254982,
512
+ "loss": 0.0159,
513
+ "step": 16800
514
+ },
515
+ {
516
+ "epoch": 0.35,
517
+ "learning_rate": 0.00027046180034466093,
518
+ "loss": 0.0149,
519
+ "step": 17000
520
+ },
521
+ {
522
+ "epoch": 0.35,
523
+ "learning_rate": 0.00026834574087066106,
524
+ "loss": 0.0131,
525
+ "step": 17200
526
+ },
527
+ {
528
+ "epoch": 0.36,
529
+ "learning_rate": 0.0002662159449847264,
530
+ "loss": 0.0143,
531
+ "step": 17400
532
+ },
533
+ {
534
+ "epoch": 0.36,
535
+ "learning_rate": 0.00026407276403414754,
536
+ "loss": 0.0209,
537
+ "step": 17600
538
+ },
539
+ {
540
+ "epoch": 0.36,
541
+ "learning_rate": 0.0002619165515743168,
542
+ "loss": 0.0136,
543
+ "step": 17800
544
+ },
545
+ {
546
+ "epoch": 0.37,
547
+ "learning_rate": 0.0002597476633104033,
548
+ "loss": 0.0148,
549
+ "step": 18000
550
+ },
551
+ {
552
+ "epoch": 0.37,
553
+ "learning_rate": 0.0002575664570386728,
554
+ "loss": 0.0169,
555
+ "step": 18200
556
+ },
557
+ {
558
+ "epoch": 0.38,
559
+ "learning_rate": 0.0002553732925874632,
560
+ "loss": 0.0116,
561
+ "step": 18400
562
+ },
563
+ {
564
+ "epoch": 0.38,
565
+ "learning_rate": 0.00025316853175782447,
566
+ "loss": 0.0164,
567
+ "step": 18600
568
+ },
569
+ {
570
+ "epoch": 0.38,
571
+ "learning_rate": 0.00025096364556859434,
572
+ "loss": 0.0154,
573
+ "step": 18800
574
+ },
575
+ {
576
+ "epoch": 0.39,
577
+ "learning_rate": 0.00024873683840049996,
578
+ "loss": 0.0141,
579
+ "step": 19000
580
+ },
581
+ {
582
+ "epoch": 0.39,
583
+ "learning_rate": 0.0002464995296538175,
584
+ "loss": 0.0144,
585
+ "step": 19200
586
+ },
587
+ {
588
+ "epoch": 0.4,
589
+ "learning_rate": 0.00024425208841197493,
590
+ "loss": 0.0122,
591
+ "step": 19400
592
+ },
593
+ {
594
+ "epoch": 0.4,
595
+ "learning_rate": 0.0002419948854299333,
596
+ "loss": 0.0129,
597
+ "step": 19600
598
+ },
599
+ {
600
+ "epoch": 0.4,
601
+ "learning_rate": 0.00023972829307302465,
602
+ "loss": 0.0133,
603
+ "step": 19800
604
+ },
605
+ {
606
+ "epoch": 0.41,
607
+ "learning_rate": 0.00023745268525552336,
608
+ "loss": 0.013,
609
+ "step": 20000
610
+ },
611
+ {
612
+ "epoch": 0.41,
613
+ "learning_rate": 0.00023516843737896262,
614
+ "loss": 0.0162,
615
+ "step": 20200
616
+ },
617
+ {
618
+ "epoch": 0.42,
619
+ "learning_rate": 0.00023287592627020543,
620
+ "loss": 0.0156,
621
+ "step": 20400
622
+ },
623
+ {
624
+ "epoch": 0.42,
625
+ "learning_rate": 0.00023057553011928013,
626
+ "loss": 0.0137,
627
+ "step": 20600
628
+ },
629
+ {
630
+ "epoch": 0.43,
631
+ "learning_rate": 0.0002282676284169916,
632
+ "loss": 0.0146,
633
+ "step": 20800
634
+ },
635
+ {
636
+ "epoch": 0.43,
637
+ "learning_rate": 0.00022595260189231734,
638
+ "loss": 0.0143,
639
+ "step": 21000
640
+ },
641
+ {
642
+ "epoch": 0.43,
643
+ "learning_rate": 0.00022363083244959957,
644
+ "loss": 0.0173,
645
+ "step": 21200
646
+ },
647
+ {
648
+ "epoch": 0.44,
649
+ "learning_rate": 0.00022131435893762175,
650
+ "loss": 0.0141,
651
+ "step": 21400
652
+ },
653
+ {
654
+ "epoch": 0.44,
655
+ "learning_rate": 0.00021899196729468126,
656
+ "loss": 0.0152,
657
+ "step": 21600
658
+ },
659
+ {
660
+ "epoch": 0.45,
661
+ "learning_rate": 0.00021665232533016668,
662
+ "loss": 0.0132,
663
+ "step": 21800
664
+ },
665
+ {
666
+ "epoch": 0.45,
667
+ "learning_rate": 0.00021430747469180223,
668
+ "loss": 0.0138,
669
+ "step": 22000
670
+ },
671
+ {
672
+ "epoch": 0.45,
673
+ "learning_rate": 0.00021195780220394222,
674
+ "loss": 0.0144,
675
+ "step": 22200
676
+ },
677
+ {
678
+ "epoch": 0.46,
679
+ "learning_rate": 0.00020960369548639,
680
+ "loss": 0.017,
681
+ "step": 22400
682
+ },
683
+ {
684
+ "epoch": 0.46,
685
+ "learning_rate": 0.00020724554289045297,
686
+ "loss": 0.0131,
687
+ "step": 22600
688
+ },
689
+ {
690
+ "epoch": 0.47,
691
+ "learning_rate": 0.00020488373343487728,
692
+ "loss": 0.0151,
693
+ "step": 22800
694
+ },
695
+ {
696
+ "epoch": 0.47,
697
+ "learning_rate": 0.0002025186567416726,
698
+ "loss": 0.0134,
699
+ "step": 23000
700
+ },
701
+ {
702
+ "epoch": 0.47,
703
+ "learning_rate": 0.00020015070297183673,
704
+ "loss": 0.015,
705
+ "step": 23200
706
+ },
707
+ {
708
+ "epoch": 0.48,
709
+ "learning_rate": 0.000197780262760992,
710
+ "loss": 0.0153,
711
+ "step": 23400
712
+ },
713
+ {
714
+ "epoch": 0.48,
715
+ "learning_rate": 0.00019540772715494277,
716
+ "loss": 0.0151,
717
+ "step": 23600
718
+ },
719
+ {
720
+ "epoch": 0.49,
721
+ "learning_rate": 0.00019303348754516576,
722
+ "loss": 0.0153,
723
+ "step": 23800
724
+ },
725
+ {
726
+ "epoch": 0.49,
727
+ "learning_rate": 0.00019065793560424322,
728
+ "loss": 0.0125,
729
+ "step": 24000
730
+ },
731
+ {
732
+ "epoch": 0.49,
733
+ "learning_rate": 0.00018828146322124917,
734
+ "loss": 0.0126,
735
+ "step": 24200
736
+ },
737
+ {
738
+ "epoch": 0.5,
739
+ "learning_rate": 0.00018590446243710106,
740
+ "loss": 0.0146,
741
+ "step": 24400
742
+ },
743
+ {
744
+ "epoch": 0.5,
745
+ "learning_rate": 0.0001835273253798852,
746
+ "loss": 0.0123,
747
+ "step": 24600
748
+ },
749
+ {
750
+ "epoch": 0.51,
751
+ "learning_rate": 0.00018115044420016874,
752
+ "loss": 0.0173,
753
+ "step": 24800
754
+ },
755
+ {
756
+ "epoch": 0.51,
757
+ "learning_rate": 0.00017877421100630705,
758
+ "loss": 0.0151,
759
+ "step": 25000
760
+ },
761
+ {
762
+ "epoch": 0.52,
763
+ "learning_rate": 0.00017639901779975916,
764
+ "loss": 0.0174,
765
+ "step": 25200
766
+ },
767
+ {
768
+ "epoch": 0.52,
769
+ "learning_rate": 0.00017402525641041958,
770
+ "loss": 0.0122,
771
+ "step": 25400
772
+ },
773
+ {
774
+ "epoch": 0.52,
775
+ "learning_rate": 0.00017166517293872442,
776
+ "loss": 0.0193,
777
+ "step": 25600
778
+ },
779
+ {
780
+ "epoch": 0.53,
781
+ "learning_rate": 0.00016929543761778923,
782
+ "loss": 0.0182,
783
+ "step": 25800
784
+ },
785
+ {
786
+ "epoch": 0.53,
787
+ "learning_rate": 0.00016692830597455233,
788
+ "loss": 0.0136,
789
+ "step": 26000
790
+ },
791
+ {
792
+ "epoch": 0.54,
793
+ "learning_rate": 0.0001645641685090116,
794
+ "loss": 0.0133,
795
+ "step": 26200
796
+ },
797
+ {
798
+ "epoch": 0.54,
799
+ "learning_rate": 0.000162203415227223,
800
+ "loss": 0.0107,
801
+ "step": 26400
802
+ },
803
+ {
804
+ "epoch": 0.54,
805
+ "learning_rate": 0.00015984643557696138,
806
+ "loss": 0.0128,
807
+ "step": 26600
808
+ },
809
+ {
810
+ "epoch": 0.55,
811
+ "learning_rate": 0.00015749361838347518,
812
+ "loss": 0.0134,
813
+ "step": 26800
814
+ },
815
+ {
816
+ "epoch": 0.55,
817
+ "learning_rate": 0.000155145351785342,
818
+ "loss": 0.0122,
819
+ "step": 27000
820
+ },
821
+ {
822
+ "epoch": 0.56,
823
+ "learning_rate": 0.000152802023170439,
824
+ "loss": 0.0145,
825
+ "step": 27200
826
+ },
827
+ {
828
+ "epoch": 0.56,
829
+ "learning_rate": 0.0001504640191120359,
830
+ "loss": 0.0145,
831
+ "step": 27400
832
+ },
833
+ {
834
+ "epoch": 0.56,
835
+ "learning_rate": 0.00014813172530502337,
836
+ "loss": 0.0163,
837
+ "step": 27600
838
+ },
839
+ {
840
+ "epoch": 0.57,
841
+ "learning_rate": 0.00014580552650228558,
842
+ "loss": 0.0106,
843
+ "step": 27800
844
+ },
845
+ {
846
+ "epoch": 0.57,
847
+ "learning_rate": 0.00014348580645122863,
848
+ "loss": 0.0132,
849
+ "step": 28000
850
+ },
851
+ {
852
+ "epoch": 0.58,
853
+ "learning_rate": 0.00014117294783047438,
854
+ "loss": 0.0117,
855
+ "step": 28200
856
+ },
857
+ {
858
+ "epoch": 0.58,
859
+ "learning_rate": 0.00013886733218673142,
860
+ "loss": 0.0129,
861
+ "step": 28400
862
+ },
863
+ {
864
+ "epoch": 0.58,
865
+ "learning_rate": 0.0001365693398718515,
866
+ "loss": 0.0134,
867
+ "step": 28600
868
+ },
869
+ {
870
+ "epoch": 0.59,
871
+ "learning_rate": 0.00013427934998008455,
872
+ "loss": 0.0127,
873
+ "step": 28800
874
+ },
875
+ {
876
+ "epoch": 0.59,
877
+ "learning_rate": 0.00013199774028553978,
878
+ "loss": 0.0112,
879
+ "step": 29000
880
+ },
881
+ {
882
+ "epoch": 0.6,
883
+ "learning_rate": 0.0001297248871798657,
884
+ "loss": 0.0137,
885
+ "step": 29200
886
+ },
887
+ {
888
+ "epoch": 0.6,
889
+ "learning_rate": 0.00012746116561015725,
890
+ "loss": 0.0167,
891
+ "step": 29400
892
+ },
893
+ {
894
+ "epoch": 0.61,
895
+ "learning_rate": 0.0001252069490171022,
896
+ "loss": 0.0135,
897
+ "step": 29600
898
+ },
899
+ {
900
+ "epoch": 0.61,
901
+ "learning_rate": 0.00012296260927337492,
902
+ "loss": 0.0103,
903
+ "step": 29800
904
+ },
905
+ {
906
+ "epoch": 0.61,
907
+ "learning_rate": 0.00012072851662228994,
908
+ "loss": 0.0121,
909
+ "step": 30000
910
+ },
911
+ {
912
+ "epoch": 0.62,
913
+ "learning_rate": 0.00011850503961672364,
914
+ "loss": 0.0134,
915
+ "step": 30200
916
+ },
917
+ {
918
+ "epoch": 0.62,
919
+ "learning_rate": 0.00011630357960773998,
920
+ "loss": 0.0126,
921
+ "step": 30400
922
+ },
923
+ {
924
+ "epoch": 0.63,
925
+ "learning_rate": 0.000114102374844354,
926
+ "loss": 0.0126,
927
+ "step": 30600
928
+ },
929
+ {
930
+ "epoch": 0.63,
931
+ "learning_rate": 0.00011191287882511448,
932
+ "loss": 0.0131,
933
+ "step": 30800
934
+ },
935
+ {
936
+ "epoch": 0.63,
937
+ "learning_rate": 0.0001097354527458991,
938
+ "loss": 0.0135,
939
+ "step": 31000
940
+ },
941
+ {
942
+ "epoch": 0.64,
943
+ "learning_rate": 0.00010757045581143638,
944
+ "loss": 0.0162,
945
+ "step": 31200
946
+ },
947
+ {
948
+ "epoch": 0.64,
949
+ "learning_rate": 0.00010541824517604906,
950
+ "loss": 0.012,
951
+ "step": 31400
952
+ },
953
+ {
954
+ "epoch": 0.65,
955
+ "learning_rate": 0.00010327917588473477,
956
+ "loss": 0.0134,
957
+ "step": 31600
958
+ },
959
+ {
960
+ "epoch": 0.65,
961
+ "learning_rate": 0.00010115360081459545,
962
+ "loss": 0.0144,
963
+ "step": 31800
964
+ },
965
+ {
966
+ "epoch": 0.65,
967
+ "learning_rate": 9.904187061662376e-05,
968
+ "loss": 0.0195,
969
+ "step": 32000
970
+ },
971
+ {
972
+ "epoch": 0.66,
973
+ "learning_rate": 9.69443336578574e-05,
974
+ "loss": 0.0157,
975
+ "step": 32200
976
+ },
977
+ {
978
+ "epoch": 0.66,
979
+ "learning_rate": 9.486133596390939e-05,
980
+ "loss": 0.0131,
981
+ "step": 32400
982
+ },
983
+ {
984
+ "epoch": 0.67,
985
+ "learning_rate": 9.279322116188555e-05,
986
+ "loss": 0.0146,
987
+ "step": 32600
988
+ },
989
+ {
990
+ "epoch": 0.67,
991
+ "learning_rate": 9.074033042369668e-05,
992
+ "loss": 0.0118,
993
+ "step": 32800
994
+ },
995
+ {
996
+ "epoch": 0.67,
997
+ "learning_rate": 8.870300240977657e-05,
998
+ "loss": 0.0153,
999
+ "step": 33000
1000
+ },
1001
+ {
1002
+ "epoch": 0.68,
1003
+ "learning_rate": 8.668157321321404e-05,
1004
+ "loss": 0.0107,
1005
+ "step": 33200
1006
+ },
1007
+ {
1008
+ "epoch": 0.68,
1009
+ "learning_rate": 8.467637630430823e-05,
1010
+ "loss": 0.0145,
1011
+ "step": 33400
1012
+ },
1013
+ {
1014
+ "epoch": 0.69,
1015
+ "learning_rate": 8.268774247555691e-05,
1016
+ "loss": 0.0124,
1017
+ "step": 33600
1018
+ },
1019
+ {
1020
+ "epoch": 0.69,
1021
+ "learning_rate": 8.07159997870866e-05,
1022
+ "loss": 0.0117,
1023
+ "step": 33800
1024
+ },
1025
+ {
1026
+ "epoch": 0.7,
1027
+ "learning_rate": 7.876147351253296e-05,
1028
+ "loss": 0.0153,
1029
+ "step": 34000
1030
+ },
1031
+ {
1032
+ "epoch": 0.7,
1033
+ "learning_rate": 7.682448608538163e-05,
1034
+ "loss": 0.0118,
1035
+ "step": 34200
1036
+ },
1037
+ {
1038
+ "epoch": 0.7,
1039
+ "learning_rate": 7.49149077432921e-05,
1040
+ "loss": 0.0133,
1041
+ "step": 34400
1042
+ },
1043
+ {
1044
+ "epoch": 0.71,
1045
+ "learning_rate": 7.301386202782465e-05,
1046
+ "loss": 0.0129,
1047
+ "step": 34600
1048
+ },
1049
+ {
1050
+ "epoch": 0.71,
1051
+ "learning_rate": 7.114066962139441e-05,
1052
+ "loss": 0.0105,
1053
+ "step": 34800
1054
+ },
1055
+ {
1056
+ "epoch": 0.72,
1057
+ "learning_rate": 6.927681374580455e-05,
1058
+ "loss": 0.0132,
1059
+ "step": 35000
1060
+ },
1061
+ {
1062
+ "epoch": 0.72,
1063
+ "learning_rate": 6.74320613790631e-05,
1064
+ "loss": 0.0148,
1065
+ "step": 35200
1066
+ },
1067
+ {
1068
+ "epoch": 0.72,
1069
+ "learning_rate": 6.560671684551868e-05,
1070
+ "loss": 0.0135,
1071
+ "step": 35400
1072
+ },
1073
+ {
1074
+ "epoch": 0.73,
1075
+ "learning_rate": 6.380108126785653e-05,
1076
+ "loss": 0.0147,
1077
+ "step": 35600
1078
+ },
1079
+ {
1080
+ "epoch": 0.73,
1081
+ "learning_rate": 6.20154525174229e-05,
1082
+ "loss": 0.0139,
1083
+ "step": 35800
1084
+ },
1085
+ {
1086
+ "epoch": 0.74,
1087
+ "learning_rate": 6.025012516508621e-05,
1088
+ "loss": 0.0125,
1089
+ "step": 36000
1090
+ },
1091
+ {
1092
+ "epoch": 0.74,
1093
+ "learning_rate": 5.850539043264194e-05,
1094
+ "loss": 0.0105,
1095
+ "step": 36200
1096
+ },
1097
+ {
1098
+ "epoch": 0.74,
1099
+ "learning_rate": 5.6781536144770985e-05,
1100
+ "loss": 0.0162,
1101
+ "step": 36400
1102
+ },
1103
+ {
1104
+ "epoch": 0.75,
1105
+ "learning_rate": 5.5078846681557415e-05,
1106
+ "loss": 0.0111,
1107
+ "step": 36600
1108
+ },
1109
+ {
1110
+ "epoch": 0.75,
1111
+ "learning_rate": 5.339760293157545e-05,
1112
+ "loss": 0.0111,
1113
+ "step": 36800
1114
+ },
1115
+ {
1116
+ "epoch": 0.76,
1117
+ "learning_rate": 5.173808224555153e-05,
1118
+ "loss": 0.0108,
1119
+ "step": 37000
1120
+ },
1121
+ {
1122
+ "epoch": 0.76,
1123
+ "learning_rate": 5.010055839061078e-05,
1124
+ "loss": 0.0124,
1125
+ "step": 37200
1126
+ },
1127
+ {
1128
+ "epoch": 0.76,
1129
+ "learning_rate": 4.849332195813418e-05,
1130
+ "loss": 0.013,
1131
+ "step": 37400
1132
+ },
1133
+ {
1134
+ "epoch": 0.77,
1135
+ "learning_rate": 4.690048518324095e-05,
1136
+ "loss": 0.0128,
1137
+ "step": 37600
1138
+ },
1139
+ {
1140
+ "epoch": 0.77,
1141
+ "learning_rate": 4.5338236363359446e-05,
1142
+ "loss": 0.0156,
1143
+ "step": 37800
1144
+ },
1145
+ {
1146
+ "epoch": 0.78,
1147
+ "learning_rate": 4.3791132442674044e-05,
1148
+ "loss": 0.0116,
1149
+ "step": 38000
1150
+ },
1151
+ {
1152
+ "epoch": 0.78,
1153
+ "learning_rate": 4.226733634221389e-05,
1154
+ "loss": 0.0131,
1155
+ "step": 38200
1156
+ },
1157
+ {
1158
+ "epoch": 0.78,
1159
+ "learning_rate": 4.0767099438946455e-05,
1160
+ "loss": 0.0115,
1161
+ "step": 38400
1162
+ },
1163
+ {
1164
+ "epoch": 0.79,
1165
+ "learning_rate": 3.929066922333545e-05,
1166
+ "loss": 0.013,
1167
+ "step": 38600
1168
+ },
1169
+ {
1170
+ "epoch": 0.79,
1171
+ "learning_rate": 3.78382892585125e-05,
1172
+ "loss": 0.0126,
1173
+ "step": 38800
1174
+ },
1175
+ {
1176
+ "epoch": 0.8,
1177
+ "learning_rate": 3.6410199140097565e-05,
1178
+ "loss": 0.0115,
1179
+ "step": 39000
1180
+ },
1181
+ {
1182
+ "epoch": 0.8,
1183
+ "learning_rate": 3.500663445667306e-05,
1184
+ "loss": 0.0137,
1185
+ "step": 39200
1186
+ },
1187
+ {
1188
+ "epoch": 0.81,
1189
+ "learning_rate": 3.3627826750919713e-05,
1190
+ "loss": 0.012,
1191
+ "step": 39400
1192
+ },
1193
+ {
1194
+ "epoch": 0.81,
1195
+ "learning_rate": 3.2274003481419156e-05,
1196
+ "loss": 0.0139,
1197
+ "step": 39600
1198
+ },
1199
+ {
1200
+ "epoch": 0.81,
1201
+ "learning_rate": 3.094538798513103e-05,
1202
+ "loss": 0.0124,
1203
+ "step": 39800
1204
+ },
1205
+ {
1206
+ "epoch": 0.82,
1207
+ "learning_rate": 2.964219944054926e-05,
1208
+ "loss": 0.0159,
1209
+ "step": 40000
1210
+ },
1211
+ {
1212
+ "epoch": 0.82,
1213
+ "learning_rate": 2.8364652831545036e-05,
1214
+ "loss": 0.0136,
1215
+ "step": 40200
1216
+ },
1217
+ {
1218
+ "epoch": 0.83,
1219
+ "learning_rate": 2.7112958911901083e-05,
1220
+ "loss": 0.0159,
1221
+ "step": 40400
1222
+ },
1223
+ {
1224
+ "epoch": 0.83,
1225
+ "learning_rate": 2.588732417054455e-05,
1226
+ "loss": 0.011,
1227
+ "step": 40600
1228
+ },
1229
+ {
1230
+ "epoch": 0.83,
1231
+ "learning_rate": 2.4693882010091674e-05,
1232
+ "loss": 0.0152,
1233
+ "step": 40800
1234
+ },
1235
+ {
1236
+ "epoch": 0.84,
1237
+ "learning_rate": 2.3520835081962702e-05,
1238
+ "loss": 0.0161,
1239
+ "step": 41000
1240
+ },
1241
+ {
1242
+ "epoch": 0.84,
1243
+ "learning_rate": 2.2374439916125642e-05,
1244
+ "loss": 0.0117,
1245
+ "step": 41200
1246
+ },
1247
+ {
1248
+ "epoch": 0.85,
1249
+ "learning_rate": 2.1254885630626012e-05,
1250
+ "loss": 0.0137,
1251
+ "step": 41400
1252
+ },
1253
+ {
1254
+ "epoch": 0.85,
1255
+ "learning_rate": 2.016235691563421e-05,
1256
+ "loss": 0.0105,
1257
+ "step": 41600
1258
+ },
1259
+ {
1260
+ "epoch": 0.85,
1261
+ "learning_rate": 1.909703400297725e-05,
1262
+ "loss": 0.0125,
1263
+ "step": 41800
1264
+ },
1265
+ {
1266
+ "epoch": 0.86,
1267
+ "learning_rate": 1.8059092636406624e-05,
1268
+ "loss": 0.0124,
1269
+ "step": 42000
1270
+ },
1271
+ {
1272
+ "epoch": 0.86,
1273
+ "learning_rate": 1.704870404260628e-05,
1274
+ "loss": 0.0119,
1275
+ "step": 42200
1276
+ },
1277
+ {
1278
+ "epoch": 0.87,
1279
+ "learning_rate": 1.6066034902945553e-05,
1280
+ "loss": 0.0096,
1281
+ "step": 42400
1282
+ },
1283
+ {
1284
+ "epoch": 0.87,
1285
+ "learning_rate": 1.5111247325982417e-05,
1286
+ "loss": 0.0114,
1287
+ "step": 42600
1288
+ },
1289
+ {
1290
+ "epoch": 0.87,
1291
+ "learning_rate": 1.4184498820720606e-05,
1292
+ "loss": 0.0129,
1293
+ "step": 42800
1294
+ },
1295
+ {
1296
+ "epoch": 0.88,
1297
+ "learning_rate": 1.3285942270625904e-05,
1298
+ "loss": 0.0143,
1299
+ "step": 43000
1300
+ },
1301
+ {
1302
+ "epoch": 0.88,
1303
+ "learning_rate": 1.2415725908405126e-05,
1304
+ "loss": 0.0135,
1305
+ "step": 43200
1306
+ },
1307
+ {
1308
+ "epoch": 0.89,
1309
+ "learning_rate": 1.1573993291552778e-05,
1310
+ "loss": 0.0111,
1311
+ "step": 43400
1312
+ },
1313
+ {
1314
+ "epoch": 0.89,
1315
+ "learning_rate": 1.0760883278668474e-05,
1316
+ "loss": 0.012,
1317
+ "step": 43600
1318
+ },
1319
+ {
1320
+ "epoch": 0.9,
1321
+ "learning_rate": 9.976530006549986e-06,
1322
+ "loss": 0.0152,
1323
+ "step": 43800
1324
+ },
1325
+ {
1326
+ "epoch": 0.9,
1327
+ "learning_rate": 9.221062868064821e-06,
1328
+ "loss": 0.0124,
1329
+ "step": 44000
1330
+ },
1331
+ {
1332
+ "epoch": 0.9,
1333
+ "learning_rate": 8.494606490804811e-06,
1334
+ "loss": 0.0163,
1335
+ "step": 44200
1336
+ },
1337
+ {
1338
+ "epoch": 0.91,
1339
+ "learning_rate": 7.800694690729383e-06,
1340
+ "loss": 0.012,
1341
+ "step": 44400
1342
+ },
1343
+ {
1344
+ "epoch": 0.91,
1345
+ "learning_rate": 7.132468049208598e-06,
1346
+ "loss": 0.0112,
1347
+ "step": 44600
1348
+ },
1349
+ {
1350
+ "epoch": 0.92,
1351
+ "learning_rate": 6.493596719362824e-06,
1352
+ "loss": 0.0106,
1353
+ "step": 44800
1354
+ },
1355
+ {
1356
+ "epoch": 0.92,
1357
+ "learning_rate": 5.884186094254458e-06,
1358
+ "loss": 0.0132,
1359
+ "step": 45000
1360
+ },
1361
+ {
1362
+ "epoch": 0.92,
1363
+ "learning_rate": 5.304336706884555e-06,
1364
+ "loss": 0.0137,
1365
+ "step": 45200
1366
+ },
1367
+ {
1368
+ "epoch": 0.93,
1369
+ "learning_rate": 4.754144213608023e-06,
1370
+ "loss": 0.0101,
1371
+ "step": 45400
1372
+ },
1373
+ {
1374
+ "epoch": 0.93,
1375
+ "learning_rate": 4.233699378353457e-06,
1376
+ "loss": 0.009,
1377
+ "step": 45600
1378
+ },
1379
+ {
1380
+ "epoch": 0.94,
1381
+ "learning_rate": 3.743088057650132e-06,
1382
+ "loss": 0.014,
1383
+ "step": 45800
1384
+ },
1385
+ {
1386
+ "epoch": 0.94,
1387
+ "learning_rate": 3.2823911864643436e-06,
1388
+ "loss": 0.0129,
1389
+ "step": 46000
1390
+ },
1391
+ {
1392
+ "epoch": 0.94,
1393
+ "learning_rate": 2.8516847648477812e-06,
1394
+ "loss": 0.0132,
1395
+ "step": 46200
1396
+ },
1397
+ {
1398
+ "epoch": 0.95,
1399
+ "learning_rate": 2.4510398454000396e-06,
1400
+ "loss": 0.0121,
1401
+ "step": 46400
1402
+ },
1403
+ {
1404
+ "epoch": 0.95,
1405
+ "learning_rate": 2.0805225215471503e-06,
1406
+ "loss": 0.0115,
1407
+ "step": 46600
1408
+ },
1409
+ {
1410
+ "epoch": 0.96,
1411
+ "learning_rate": 1.740193916638432e-06,
1412
+ "loss": 0.0121,
1413
+ "step": 46800
1414
+ },
1415
+ {
1416
+ "epoch": 0.96,
1417
+ "learning_rate": 1.430110173862991e-06,
1418
+ "loss": 0.0145,
1419
+ "step": 47000
1420
+ },
1421
+ {
1422
+ "epoch": 0.96,
1423
+ "learning_rate": 1.1503224469880126e-06,
1424
+ "loss": 0.013,
1425
+ "step": 47200
1426
+ },
1427
+ {
1428
+ "epoch": 0.97,
1429
+ "learning_rate": 9.008768919199934e-07,
1430
+ "loss": 0.0127,
1431
+ "step": 47400
1432
+ },
1433
+ {
1434
+ "epoch": 0.97,
1435
+ "learning_rate": 6.818146590905163e-07,
1436
+ "loss": 0.0138,
1437
+ "step": 47600
1438
+ },
1439
+ {
1440
+ "epoch": 0.98,
1441
+ "learning_rate": 4.931718866678032e-07,
1442
+ "loss": 0.0113,
1443
+ "step": 47800
1444
+ },
1445
+ {
1446
+ "epoch": 0.98,
1447
+ "learning_rate": 3.356948644792021e-07,
1448
+ "loss": 0.0169,
1449
+ "step": 48000
1450
+ },
1451
+ {
1452
+ "epoch": 0.99,
1453
+ "learning_rate": 2.0782690935592252e-07,
1454
+ "loss": 0.0109,
1455
+ "step": 48200
1456
+ },
1457
+ {
1458
+ "epoch": 0.99,
1459
+ "learning_rate": 1.1045660725531198e-07,
1460
+ "loss": 0.0119,
1461
+ "step": 48400
1462
+ },
1463
+ {
1464
+ "epoch": 0.99,
1465
+ "learning_rate": 4.360002112121604e-08,
1466
+ "loss": 0.0102,
1467
+ "step": 48600
1468
+ },
1469
+ {
1470
+ "epoch": 1.0,
1471
+ "learning_rate": 7.2681801234168665e-09,
1472
+ "loss": 0.0114,
1473
+ "step": 48800
1474
+ },
1475
+ {
1476
+ "epoch": 1.0,
1477
+ "eval_accuracy": 0.9751466987432912,
1478
+ "eval_auc": 0.9180107216187284,
1479
+ "eval_f1": 0.15952083840732592,
1480
+ "eval_loss": 0.2542668879032135,
1481
+ "eval_mcc": 0.27048987280764264,
1482
+ "eval_precision": 0.0879081257328702,
1483
+ "eval_recall": 0.8605598317425242,
1484
+ "eval_runtime": 27823.2385,
1485
+ "eval_samples_per_second": 16.313,
1486
+ "eval_steps_per_second": 0.453,
1487
+ "step": 48919
1488
+ }
1489
+ ],
1490
+ "logging_steps": 200,
1491
+ "max_steps": 48919,
1492
+ "num_train_epochs": 1,
1493
+ "save_steps": 500,
1494
+ "total_flos": 1.56321913484268e+18,
1495
+ "trial_name": null,
1496
+ "trial_params": null
1497
+ }