Alisson990 commited on
Commit
04aeea6
·
verified ·
1 Parent(s): 53660c3

Upload training.log with huggingface_hub

Browse files
Files changed (1) hide show
  1. training.log +247 -0
training.log ADDED
@@ -0,0 +1,247 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [19:47:48] === PHASE 1 SUPERVISED PRE-TRAINING START ===
2
+ [19:47:48] Baixando dataset do HuggingFace...
3
+ [19:47:48] (isso pode levar 15-30 min para 10 GB)
4
+ [19:52:15] Dataset em: /kaggle/working/dataset
5
+ [19:52:15] Binance 1s: 2304 arquivos
6
+ [19:52:15] Binance 1m: 0 arquivos (fallback)
7
+ [19:52:15] NovaDax trades: 2304 arquivos
8
+ [19:52:15] NovaDax klines: 1 arquivos
9
+ [19:52:15] NovaDax indexado: 2304 datas de trades, 1 de klines
10
+ [19:52:15] Split: train=1461 | val=366 | test=477 dias
11
+ [19:52:15] Testando pipeline de dados com o primeiro arquivo...
12
+ [19:52:16] OK: 86400 rows | colunas: ['ts_ms', 'bin_ret', 'bin_ret_10', 'bin_ret_30', 'vol_z', 'trades_z']...
13
+ [19:52:16] y30 dist: flat=68476 up=8737 dn=9187
14
+ [19:52:16] Feats: mean=0.0361 | std=0.3507
15
+ [19:52:17] DataLoaders criados
16
+ [19:52:17] Modelo: 4,906,507 params (4.91M) | trainable: 4,906,507
17
+ [19:52:17] DataParallel em 2 GPUs
18
+ [19:52:17] Steps estimados: 4,400 | Steps/época: 24,654
19
+ [19:52:24] Tentando baixar checkpoint do HuggingFace...
20
+ [19:52:24] Nenhum checkpoint encontrado, iniciando do zero.
21
+ [19:52:24] Iniciando treino | step=0 | best_val=inf
22
+ [19:52:24] Máximo: 11h | Checkpoint a cada 1h
23
+ [19:52:24] === Época 1 | step=0 | elapsed=0.00h ===
24
+ [19:52:46] step= 100 | loss=2.5517 | acc30=0.184 | lr=2.61e-04 | 0.01h
25
+ [19:53:03] step= 200 | loss=2.1935 | acc30=0.260 | lr=3.00e-04 | 0.01h
26
+ [19:53:20] step= 300 | loss=2.3798 | acc30=0.439 | lr=2.99e-04 | 0.02h
27
+ [19:53:37] step= 400 | loss=2.7090 | acc30=0.989 | lr=2.97e-04 | 0.02h
28
+ [19:53:54] step= 500 | loss=2.8053 | acc30=0.583 | lr=2.95e-04 | 0.03h
29
+ [19:54:12] step= 600 | loss=2.5192 | acc30=0.263 | lr=2.91e-04 | 0.03h
30
+ [19:54:31] step= 700 | loss=2.2992 | acc30=0.212 | lr=2.87e-04 | 0.04h
31
+ [19:54:48] step= 800 | loss=2.2822 | acc30=0.213 | lr=2.82e-04 | 0.04h
32
+ [19:55:06] step= 900 | loss=2.3126 | acc30=0.226 | lr=2.77e-04 | 0.04h
33
+ [19:55:23] step= 1000 | loss=2.6052 | acc30=0.505 | lr=2.70e-04 | 0.05h
34
+ [19:55:23] Validando...
35
+ [19:55:40] VAL: loss=2.6031 acc30=0.601 acc60=0.277 acc120=0.342
36
+ [19:55:58] step= 1100 | loss=2.7995 | acc30=0.363 | lr=2.63e-04 | 0.06h
37
+ [19:56:15] step= 1200 | loss=2.9147 | acc30=0.927 | lr=2.56e-04 | 0.06h
38
+ [19:56:33] step= 1300 | loss=2.4326 | acc30=0.359 | lr=2.48e-04 | 0.07h
39
+ [19:56:51] step= 1400 | loss=2.3332 | acc30=0.215 | lr=2.39e-04 | 0.07h
40
+ [19:57:09] step= 1500 | loss=2.4511 | acc30=0.167 | lr=2.30e-04 | 0.08h
41
+ [19:57:26] step= 1600 | loss=2.4280 | acc30=0.173 | lr=2.21e-04 | 0.08h
42
+ [19:57:44] step= 1700 | loss=2.3409 | acc30=0.185 | lr=2.11e-04 | 0.09h
43
+ [19:58:02] step= 1800 | loss=2.3322 | acc30=0.207 | lr=2.00e-04 | 0.09h
44
+ [19:58:19] step= 1900 | loss=2.1135 | acc30=0.278 | lr=1.90e-04 | 0.10h
45
+ [19:58:36] step= 2000 | loss=2.1493 | acc30=0.271 | lr=1.79e-04 | 0.10h
46
+ [19:58:36] Validando...
47
+ [19:58:53] VAL: loss=2.3881 acc30=0.202 acc60=0.277 acc120=0.342
48
+ [19:59:11] step= 2100 | loss=2.1894 | acc30=0.257 | lr=1.68e-04 | 0.11h
49
+ [19:59:29] step= 2200 | loss=2.2146 | acc30=0.248 | lr=1.57e-04 | 0.12h
50
+ [19:59:47] step= 2300 | loss=2.3106 | acc30=0.216 | lr=1.46e-04 | 0.12h
51
+ [20:00:04] step= 2400 | loss=2.1249 | acc30=0.286 | lr=1.35e-04 | 0.13h
52
+ [20:00:22] step= 2500 | loss=2.4719 | acc30=0.178 | lr=1.24e-04 | 0.13h
53
+ [20:00:39] step= 2600 | loss=2.3971 | acc30=0.191 | lr=1.13e-04 | 0.14h
54
+ [20:00:57] step= 2700 | loss=2.2963 | acc30=0.210 | lr=1.03e-04 | 0.14h
55
+ [20:01:14] step= 2800 | loss=2.6621 | acc30=0.121 | lr=9.25e-05 | 0.15h
56
+ [20:01:32] step= 2900 | loss=2.8470 | acc30=0.359 | lr=8.24e-05 | 0.15h
57
+ [20:01:50] step= 3000 | loss=2.5497 | acc30=0.147 | lr=7.28e-05 | 0.16h
58
+ [20:01:50] Validando...
59
+ [20:02:06] VAL: loss=2.3954 acc30=0.197 acc60=0.268 acc120=0.324
60
+ [20:02:25] step= 3100 | loss=2.3512 | acc30=0.239 | lr=6.35e-05 | 0.17h
61
+ [20:02:42] step= 3200 | loss=2.0462 | acc30=0.324 | lr=5.47e-05 | 0.17h
62
+ [20:03:00] step= 3300 | loss=3.5350 | acc30=0.198 | lr=4.65e-05 | 0.18h
63
+ [20:03:18] step= 3400 | loss=2.7692 | acc30=0.233 | lr=3.88e-05 | 0.18h
64
+ [20:03:36] step= 3500 | loss=2.2814 | acc30=0.249 | lr=3.17e-05 | 0.19h
65
+ [20:03:53] step= 3600 | loss=2.8433 | acc30=0.103 | lr=2.52e-05 | 0.19h
66
+ [20:04:11] step= 3700 | loss=2.3994 | acc30=0.195 | lr=1.94e-05 | 0.20h
67
+ [20:04:28] step= 3800 | loss=2.4279 | acc30=0.175 | lr=1.44e-05 | 0.20h
68
+ [20:04:46] step= 3900 | loss=2.5048 | acc30=0.157 | lr=1.00e-05 | 0.21h
69
+ [20:05:04] step= 4000 | loss=2.4662 | acc30=0.160 | lr=6.43e-06 | 0.21h
70
+ [20:05:04] Validando...
71
+ [20:05:20] VAL: loss=2.3599 acc30=0.200 acc60=0.273 acc120=0.337
72
+ [20:05:38] step= 4100 | loss=2.3305 | acc30=0.229 | lr=3.63e-06 | 0.22h
73
+ [20:05:56] step= 4200 | loss=2.3756 | acc30=0.214 | lr=1.62e-06 | 0.23h
74
+ [20:06:14] step= 4300 | loss=2.6299 | acc30=0.112 | lr=4.08e-07 | 0.23h
75
+ [20:06:31] step= 4400 | loss=3.2240 | acc30=0.053 | lr=1.00e-08 | 0.24h
76
+ [20:06:49] step= 4500 | loss=2.4437 | acc30=0.290 | lr=1.00e-08 | 0.24h
77
+ [20:07:07] step= 4600 | loss=2.1073 | acc30=0.389 | lr=1.00e-08 | 0.25h
78
+ [20:07:24] step= 4700 | loss=2.2409 | acc30=0.248 | lr=1.00e-08 | 0.25h
79
+ [20:07:42] step= 4800 | loss=2.0776 | acc30=0.385 | lr=1.00e-08 | 0.26h
80
+ [20:08:00] step= 4900 | loss=2.3494 | acc30=0.258 | lr=1.00e-08 | 0.26h
81
+ [20:08:18] step= 5000 | loss=2.5491 | acc30=0.167 | lr=1.00e-08 | 0.27h
82
+ [20:08:18] Validando...
83
+ [20:08:34] VAL: loss=2.3569 acc30=0.200 acc60=0.272 acc120=0.337
84
+ [20:08:52] step= 5100 | loss=2.8098 | acc30=0.088 | lr=1.00e-08 | 0.27h
85
+ [20:09:10] step= 5200 | loss=2.3844 | acc30=0.192 | lr=1.00e-08 | 0.28h
86
+ [20:09:28] step= 5300 | loss=2.4574 | acc30=0.163 | lr=1.00e-08 | 0.28h
87
+ [20:09:45] step= 5400 | loss=2.4677 | acc30=0.173 | lr=1.00e-08 | 0.29h
88
+ [20:10:03] step= 5500 | loss=2.1160 | acc30=0.358 | lr=1.00e-08 | 0.29h
89
+ [20:10:21] step= 5600 | loss=2.2874 | acc30=0.226 | lr=1.00e-08 | 0.30h
90
+ [20:10:39] step= 5700 | loss=2.2475 | acc30=0.276 | lr=1.00e-08 | 0.30h
91
+ [20:10:57] step= 5800 | loss=2.4170 | acc30=0.293 | lr=1.00e-08 | 0.31h
92
+ [20:11:14] step= 5900 | loss=4.3855 | acc30=0.010 | lr=1.00e-08 | 0.31h
93
+ [20:11:32] step= 6000 | loss=3.4007 | acc30=0.065 | lr=1.00e-08 | 0.32h
94
+ [20:11:32] Validando...
95
+ [20:11:48] VAL: loss=2.3568 acc30=0.200 acc60=0.272 acc120=0.337
96
+ [20:12:06] step= 6100 | loss=2.9336 | acc30=0.081 | lr=1.00e-08 | 0.33h
97
+ [20:12:24] step= 6200 | loss=2.4307 | acc30=0.189 | lr=1.00e-08 | 0.33h
98
+ [20:12:41] step= 6300 | loss=3.0773 | acc30=0.060 | lr=1.00e-08 | 0.34h
99
+ [20:12:59] step= 6400 | loss=2.3944 | acc30=0.218 | lr=1.00e-08 | 0.34h
100
+ [20:13:17] step= 6500 | loss=2.7768 | acc30=0.123 | lr=1.00e-08 | 0.35h
101
+ [20:13:35] step= 6600 | loss=3.0216 | acc30=0.074 | lr=1.00e-08 | 0.35h
102
+ [20:13:52] step= 6700 | loss=2.3582 | acc30=0.162 | lr=1.00e-08 | 0.36h
103
+ [20:14:10] step= 6800 | loss=2.7971 | acc30=0.096 | lr=1.00e-08 | 0.36h
104
+ [20:14:28] step= 6900 | loss=2.4053 | acc30=0.172 | lr=1.00e-08 | 0.37h
105
+ [20:14:47] step= 7000 | loss=2.5172 | acc30=0.146 | lr=1.00e-08 | 0.37h
106
+ [20:14:47] Validando...
107
+ [20:15:03] VAL: loss=2.3568 acc30=0.200 acc60=0.272 acc120=0.337
108
+ [20:15:20] step= 7100 | loss=2.2227 | acc30=0.267 | lr=1.00e-08 | 0.38h
109
+ [20:15:38] step= 7200 | loss=2.1562 | acc30=0.295 | lr=1.00e-08 | 0.39h
110
+ [20:15:56] step= 7300 | loss=2.4936 | acc30=0.211 | lr=1.00e-08 | 0.39h
111
+ [20:16:14] step= 7400 | loss=2.2936 | acc30=0.218 | lr=1.00e-08 | 0.40h
112
+ [20:16:31] step= 7500 | loss=2.8560 | acc30=0.088 | lr=1.00e-08 | 0.40h
113
+ [20:16:49] step= 7600 | loss=2.4172 | acc30=0.177 | lr=1.00e-08 | 0.41h
114
+ [20:17:07] step= 7700 | loss=2.3952 | acc30=0.214 | lr=1.00e-08 | 0.41h
115
+ [20:17:25] step= 7800 | loss=2.7299 | acc30=0.109 | lr=1.00e-08 | 0.42h
116
+ [20:17:42] step= 7900 | loss=2.3964 | acc30=0.184 | lr=1.00e-08 | 0.42h
117
+ [20:18:00] step= 8000 | loss=3.2202 | acc30=0.096 | lr=1.00e-08 | 0.43h
118
+ [20:18:00] Validando...
119
+ [20:18:16] VAL: loss=2.3568 acc30=0.200 acc60=0.272 acc120=0.337
120
+ [20:18:34] step= 8100 | loss=2.9130 | acc30=0.117 | lr=1.00e-08 | 0.44h
121
+ [20:18:51] step= 8200 | loss=3.6459 | acc30=0.049 | lr=1.00e-08 | 0.44h
122
+ [20:19:09] step= 8300 | loss=2.1714 | acc30=0.289 | lr=1.00e-08 | 0.45h
123
+ [20:19:27] step= 8400 | loss=2.1737 | acc30=0.303 | lr=1.00e-08 | 0.45h
124
+ [20:19:45] step= 8500 | loss=2.3622 | acc30=0.230 | lr=1.00e-08 | 0.46h
125
+ [20:20:02] step= 8600 | loss=3.8097 | acc30=0.022 | lr=1.00e-08 | 0.46h
126
+ [20:20:20] step= 8700 | loss=4.7551 | acc30=0.011 | lr=1.00e-08 | 0.47h
127
+ [20:20:38] step= 8800 | loss=2.5744 | acc30=0.135 | lr=1.00e-08 | 0.47h
128
+ [20:20:56] step= 8900 | loss=2.2570 | acc30=0.266 | lr=1.00e-08 | 0.48h
129
+ [20:21:13] step= 9000 | loss=2.0862 | acc30=0.397 | lr=1.00e-08 | 0.48h
130
+ [20:21:13] Validando...
131
+ [20:21:31] VAL: loss=2.3567 acc30=0.200 acc60=0.272 acc120=0.337
132
+ [20:21:48] step= 9100 | loss=2.8084 | acc30=0.129 | lr=1.00e-08 | 0.49h
133
+ [20:22:06] step= 9200 | loss=2.4362 | acc30=0.148 | lr=1.00e-08 | 0.50h
134
+ [20:22:24] step= 9300 | loss=2.2917 | acc30=0.217 | lr=1.00e-08 | 0.50h
135
+ [20:22:41] step= 9400 | loss=2.0908 | acc30=0.378 | lr=1.00e-08 | 0.50h
136
+ [20:22:59] step= 9500 | loss=2.1227 | acc30=0.321 | lr=1.00e-08 | 0.51h
137
+ [20:23:17] step= 9600 | loss=2.8909 | acc30=0.176 | lr=1.00e-08 | 0.51h
138
+ [20:23:35] step= 9700 | loss=2.9897 | acc30=0.115 | lr=1.00e-08 | 0.52h
139
+ [20:23:52] step= 9800 | loss=2.1571 | acc30=0.311 | lr=1.00e-08 | 0.52h
140
+ [20:24:10] step= 9900 | loss=2.2028 | acc30=0.266 | lr=1.00e-08 | 0.53h
141
+ [20:24:28] step= 10000 | loss=2.2150 | acc30=0.270 | lr=1.00e-08 | 0.53h
142
+ [20:24:28] Validando...
143
+ [20:24:44] VAL: loss=2.3566 acc30=0.200 acc60=0.272 acc120=0.337
144
+ [20:25:02] step= 10100 | loss=2.2938 | acc30=0.205 | lr=1.00e-08 | 0.54h
145
+ [20:25:19] step= 10200 | loss=2.3379 | acc30=0.220 | lr=1.00e-08 | 0.55h
146
+ [20:25:37] step= 10300 | loss=2.2540 | acc30=0.234 | lr=1.00e-08 | 0.55h
147
+ [20:25:55] step= 10400 | loss=2.1680 | acc30=0.299 | lr=1.00e-08 | 0.56h
148
+ [20:26:13] step= 10500 | loss=2.1737 | acc30=0.288 | lr=1.00e-08 | 0.56h
149
+ [20:26:30] step= 10600 | loss=2.2804 | acc30=0.209 | lr=1.00e-08 | 0.57h
150
+ [20:26:48] step= 10700 | loss=2.2547 | acc30=0.218 | lr=1.00e-08 | 0.57h
151
+ [20:27:06] step= 10800 | loss=2.7201 | acc30=0.160 | lr=1.00e-08 | 0.58h
152
+ [20:27:24] step= 10900 | loss=3.1726 | acc30=0.065 | lr=1.00e-08 | 0.58h
153
+ [20:27:41] step= 11000 | loss=3.3902 | acc30=0.045 | lr=1.00e-08 | 0.59h
154
+ [20:27:41] Validando...
155
+ [20:27:57] VAL: loss=2.3565 acc30=0.200 acc60=0.272 acc120=0.337
156
+ [20:28:15] step= 11100 | loss=2.7160 | acc30=0.150 | lr=1.00e-08 | 0.60h
157
+ [20:28:33] step= 11200 | loss=2.3661 | acc30=0.193 | lr=1.00e-08 | 0.60h
158
+ [20:28:51] step= 11300 | loss=2.5008 | acc30=0.194 | lr=1.00e-08 | 0.61h
159
+ [20:29:09] step= 11400 | loss=3.8704 | acc30=0.027 | lr=1.00e-08 | 0.61h
160
+ [20:29:26] step= 11500 | loss=2.7116 | acc30=0.140 | lr=1.00e-08 | 0.62h
161
+ [20:29:44] step= 11600 | loss=2.2880 | acc30=0.243 | lr=1.00e-08 | 0.62h
162
+ [20:30:02] step= 11700 | loss=2.1555 | acc30=0.309 | lr=1.00e-08 | 0.63h
163
+ [20:30:19] step= 11800 | loss=2.2513 | acc30=0.252 | lr=1.00e-08 | 0.63h
164
+ [20:30:37] step= 11900 | loss=2.1663 | acc30=0.290 | lr=1.00e-08 | 0.64h
165
+ [20:30:55] step= 12000 | loss=2.1389 | acc30=0.330 | lr=1.00e-08 | 0.64h
166
+ [20:30:55] Validando...
167
+ [20:31:11] VAL: loss=2.3565 acc30=0.200 acc60=0.272 acc120=0.337
168
+ [20:31:29] step= 12100 | loss=2.2540 | acc30=0.261 | lr=1.00e-08 | 0.65h
169
+ [20:31:46] step= 12200 | loss=3.5529 | acc30=0.050 | lr=1.00e-08 | 0.66h
170
+ [20:32:04] step= 12300 | loss=2.2384 | acc30=0.274 | lr=1.00e-08 | 0.66h
171
+ [20:32:22] step= 12400 | loss=2.1568 | acc30=0.307 | lr=1.00e-08 | 0.67h
172
+ [20:32:40] step= 12500 | loss=2.0656 | acc30=0.405 | lr=1.00e-08 | 0.67h
173
+ [20:32:57] step= 12600 | loss=2.2727 | acc30=0.237 | lr=1.00e-08 | 0.68h
174
+ [20:33:15] step= 12700 | loss=2.2428 | acc30=0.243 | lr=1.00e-08 | 0.68h
175
+ [20:33:33] step= 12800 | loss=2.5467 | acc30=0.206 | lr=1.00e-08 | 0.69h
176
+ [20:33:51] step= 12900 | loss=3.3666 | acc30=0.058 | lr=1.00e-08 | 0.69h
177
+ [20:34:08] step= 13000 | loss=2.1594 | acc30=0.297 | lr=1.00e-08 | 0.70h
178
+ [20:34:08] Validando...
179
+ [20:34:25] VAL: loss=2.3564 acc30=0.200 acc60=0.272 acc120=0.337
180
+ [20:34:43] step= 13100 | loss=2.1094 | acc30=0.344 | lr=1.00e-08 | 0.71h
181
+ [20:35:01] step= 13200 | loss=2.2251 | acc30=0.279 | lr=1.00e-08 | 0.71h
182
+ [20:35:18] step= 13300 | loss=2.3688 | acc30=0.197 | lr=1.00e-08 | 0.72h
183
+ [20:35:36] step= 13400 | loss=2.2561 | acc30=0.242 | lr=1.00e-08 | 0.72h
184
+ [20:35:54] step= 13500 | loss=2.3765 | acc30=0.180 | lr=1.00e-08 | 0.73h
185
+ [20:36:12] step= 13600 | loss=2.4212 | acc30=0.171 | lr=1.00e-08 | 0.73h
186
+ [20:36:29] step= 13700 | loss=2.4924 | acc30=0.158 | lr=1.00e-08 | 0.73h
187
+ [20:36:47] step= 13800 | loss=2.3785 | acc30=0.198 | lr=1.00e-08 | 0.74h
188
+ [20:37:05] step= 13900 | loss=2.3181 | acc30=0.195 | lr=1.00e-08 | 0.74h
189
+ [20:37:23] step= 14000 | loss=2.1844 | acc30=0.263 | lr=1.00e-08 | 0.75h
190
+ [20:37:23] Validando...
191
+ [20:37:39] VAL: loss=2.3563 acc30=0.200 acc60=0.271 acc120=0.337
192
+ [20:37:57] step= 14100 | loss=2.2146 | acc30=0.258 | lr=1.00e-08 | 0.76h
193
+ [20:38:15] step= 14200 | loss=2.0913 | acc30=0.388 | lr=1.00e-08 | 0.76h
194
+ [20:38:32] step= 14300 | loss=3.8682 | acc30=0.156 | lr=1.00e-08 | 0.77h
195
+ [20:38:50] step= 14400 | loss=4.0869 | acc30=0.049 | lr=1.00e-08 | 0.77h
196
+ [20:39:07] step= 14500 | loss=2.3883 | acc30=0.184 | lr=1.00e-08 | 0.78h
197
+ [20:39:25] step= 14600 | loss=2.6680 | acc30=0.115 | lr=1.00e-08 | 0.78h
198
+ [20:39:43] step= 14700 | loss=2.5179 | acc30=0.137 | lr=1.00e-08 | 0.79h
199
+ [20:40:01] step= 14800 | loss=2.4570 | acc30=0.157 | lr=1.00e-08 | 0.79h
200
+ [20:40:18] step= 14900 | loss=2.5993 | acc30=0.135 | lr=1.00e-08 | 0.80h
201
+ [20:40:36] step= 15000 | loss=2.3919 | acc30=0.185 | lr=1.00e-08 | 0.80h
202
+ [20:40:36] Validando...
203
+ [20:40:53] VAL: loss=2.3562 acc30=0.200 acc60=0.272 acc120=0.337
204
+ [20:41:11] step= 15100 | loss=2.1487 | acc30=0.298 | lr=1.00e-08 | 0.81h
205
+ [20:41:29] step= 15200 | loss=2.1501 | acc30=0.312 | lr=1.00e-08 | 0.82h
206
+ [20:41:46] step= 15300 | loss=2.3261 | acc30=0.211 | lr=1.00e-08 | 0.82h
207
+ [20:42:04] step= 15400 | loss=2.1218 | acc30=0.383 | lr=1.00e-08 | 0.83h
208
+ [20:42:22] step= 15500 | loss=2.3623 | acc30=0.280 | lr=1.00e-08 | 0.83h
209
+ [20:42:40] step= 15600 | loss=2.5723 | acc30=0.124 | lr=1.00e-08 | 0.84h
210
+ [20:42:57] step= 15700 | loss=2.4892 | acc30=0.155 | lr=1.00e-08 | 0.84h
211
+ [20:43:15] step= 15800 | loss=2.4106 | acc30=0.187 | lr=1.00e-08 | 0.85h
212
+ [20:43:33] step= 15900 | loss=2.4061 | acc30=0.168 | lr=1.00e-08 | 0.85h
213
+ [20:43:51] step= 16000 | loss=2.2963 | acc30=0.221 | lr=1.00e-08 | 0.86h
214
+ [20:43:51] Validando...
215
+ [20:44:07] VAL: loss=2.3562 acc30=0.200 acc60=0.272 acc120=0.337
216
+ [20:44:24] step= 16100 | loss=2.1279 | acc30=0.334 | lr=1.00e-08 | 0.87h
217
+ [20:44:42] step= 16200 | loss=4.1668 | acc30=0.086 | lr=1.00e-08 | 0.87h
218
+ [20:45:00] step= 16300 | loss=3.8789 | acc30=0.062 | lr=1.00e-08 | 0.88h
219
+ [20:45:18] step= 16400 | loss=2.2729 | acc30=0.221 | lr=1.00e-08 | 0.88h
220
+ [20:45:35] step= 16500 | loss=2.1995 | acc30=0.280 | lr=1.00e-08 | 0.89h
221
+ [20:45:53] step= 16600 | loss=2.2381 | acc30=0.237 | lr=1.00e-08 | 0.89h
222
+ [20:46:11] step= 16700 | loss=2.1802 | acc30=0.273 | lr=1.00e-08 | 0.90h
223
+ [20:46:29] step= 16800 | loss=2.1204 | acc30=0.316 | lr=1.00e-08 | 0.90h
224
+ [20:46:46] step= 16900 | loss=2.3939 | acc30=0.181 | lr=1.00e-08 | 0.91h
225
+ [20:47:04] step= 17000 | loss=2.2572 | acc30=0.222 | lr=1.00e-08 | 0.91h
226
+ [20:47:04] Validando...
227
+ [20:47:20] VAL: loss=2.3561 acc30=0.200 acc60=0.272 acc120=0.337
228
+ [20:47:38] step= 17100 | loss=2.2856 | acc30=0.219 | lr=1.00e-08 | 0.92h
229
+ [20:47:56] step= 17200 | loss=2.4377 | acc30=0.175 | lr=1.00e-08 | 0.93h
230
+ [20:48:13] step= 17300 | loss=2.5572 | acc30=0.148 | lr=1.00e-08 | 0.93h
231
+ [20:48:31] step= 17400 | loss=2.1858 | acc30=0.307 | lr=1.00e-08 | 0.94h
232
+ [20:48:49] step= 17500 | loss=2.3156 | acc30=0.289 | lr=1.00e-08 | 0.94h
233
+ [20:49:06] step= 17600 | loss=2.6825 | acc30=0.131 | lr=1.00e-08 | 0.95h
234
+ [20:49:24] step= 17700 | loss=2.5440 | acc30=0.152 | lr=1.00e-08 | 0.95h
235
+ [20:49:42] step= 17800 | loss=2.2803 | acc30=0.222 | lr=1.00e-08 | 0.96h
236
+ [20:50:00] step= 17900 | loss=2.2970 | acc30=0.229 | lr=1.00e-08 | 0.96h
237
+ [20:50:17] step= 18000 | loss=2.1674 | acc30=0.288 | lr=1.00e-08 | 0.96h
238
+ [20:50:17] Validando...
239
+ [20:50:35] VAL: loss=2.3560 acc30=0.200 acc60=0.272 acc120=0.337
240
+ [20:50:52] step= 18100 | loss=3.2475 | acc30=0.071 | lr=1.00e-08 | 0.97h
241
+ [20:51:10] step= 18200 | loss=2.5619 | acc30=0.143 | lr=1.00e-08 | 0.98h
242
+ [20:51:28] step= 18300 | loss=2.1799 | acc30=0.278 | lr=1.00e-08 | 0.98h
243
+ [20:51:45] step= 18400 | loss=2.1331 | acc30=0.299 | lr=1.00e-08 | 0.99h
244
+ [20:52:03] step= 18500 | loss=2.2125 | acc30=0.269 | lr=1.00e-08 | 0.99h
245
+ [20:52:21] step= 18600 | loss=2.1454 | acc30=0.307 | lr=1.00e-08 | 1.00h
246
+ [20:52:24] --- CHECKPOINT (hora 1.0) ---
247
+ [20:52:24] Checkpoint salvo: checkpoint_step00018616.pt (59.0 MB)