andrew-healey commited on
Commit
f2b5a92
·
verified ·
1 Parent(s): 3852543

Upload folder using huggingface_hub

Browse files
lr1.5e-3_total_batch_size40960_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr1.5e-3_total_batch_size40960_seq_len128", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 128, "batch_size": null, "total_batch_size": 40960, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0015, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1.5e-3_total_batch_size40960_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cec289292cf51a5462e2631af59109e64df3eca2d9e62e1fbd1c53a776a0f874
3
+ size 964
lr1.5e-3_total_batch_size40960_seq_len128/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.6870
3
+ 0 val perplexity 119016.5000
4
+ 0 train 11.697288 (lr=5.2448e-06) (hash(x)=5534142)
5
+ 1 train 11.686671 (lr=1.0490e-05) (hash(x)=6328012)
6
+ 2 train 11.642699 (lr=1.5734e-05) (hash(x)=5556098)
7
+ 3 train 11.575415 (lr=2.0979e-05) (hash(x)=6219138)
8
+ 4 train 11.483486 (lr=2.6224e-05) (hash(x)=5491972)
9
+ 5 train 11.388724 (lr=3.1469e-05) (hash(x)=5917741)
10
+ 6 train 11.279057 (lr=3.6713e-05) (hash(x)=6112458)
11
+ 7 train 11.155971 (lr=4.1958e-05) (hash(x)=6568379)
12
+ 8 train 10.990664 (lr=4.7203e-05) (hash(x)=7386082)
13
+ 9 train 10.888305 (lr=5.2448e-05) (hash(x)=6151674)
14
+ 10 train 10.788390 (lr=5.7692e-05) (hash(x)=7859060)
15
+ 11 train 10.679645 (lr=6.2937e-05) (hash(x)=6391506)
16
+ 12 train 10.621133 (lr=6.8182e-05) (hash(x)=5985675)
17
+ 13 train 10.569341 (lr=7.3427e-05) (hash(x)=6984772)
18
+ 14 train 10.510788 (lr=7.8671e-05) (hash(x)=6603717)
19
+ 15 train 10.403092 (lr=8.3916e-05) (hash(x)=6190579)
20
+ 16 train 10.391287 (lr=8.9161e-05) (hash(x)=7276303)
21
+ 17 train 10.431553 (lr=9.4406e-05) (hash(x)=6630378)
22
+ 18 train 10.445994 (lr=9.9650e-05) (hash(x)=9989744)
23
+ 19 train 10.321763 (lr=1.0490e-04) (hash(x)=6317894)
24
+ 20 train 10.243794 (lr=1.1014e-04) (hash(x)=7947703)
25
+ 21 train 10.203502 (lr=1.1538e-04) (hash(x)=5777498)
26
+ 22 train 10.178701 (lr=1.2063e-04) (hash(x)=7723226)
27
+ 23 train 10.122749 (lr=1.2587e-04) (hash(x)=6491229)
28
+ 24 train 10.196757 (lr=1.3112e-04) (hash(x)=6019389)
29
+ 25 train 10.119077 (lr=1.3636e-04) (hash(x)=4648609)
30
+ 26 train 10.038429 (lr=1.4161e-04) (hash(x)=5558355)
31
+ 27 train 10.028001 (lr=1.4685e-04) (hash(x)=6369326)
32
+ 28 train 9.927476 (lr=1.5210e-04) (hash(x)=4906853)
33
+ 29 train 9.791779 (lr=1.5734e-04) (hash(x)=6437376)
34
+ 30 train 9.851848 (lr=1.6259e-04) (hash(x)=7084093)
35
+ 31 train 9.732794 (lr=1.6783e-04) (hash(x)=7048804)
36
+ 32 train 9.692970 (lr=1.7308e-04) (hash(x)=5752594)
37
+ 33 train 9.419542 (lr=1.7832e-04) (hash(x)=6393520)
38
+ 34 train 9.583248 (lr=1.8357e-04) (hash(x)=6793550)
39
+ 35 train 9.424244 (lr=1.8881e-04) (hash(x)=6581348)
40
+ 36 train 9.319365 (lr=1.9406e-04) (hash(x)=6235837)
41
+ 37 train 9.301547 (lr=1.9930e-04) (hash(x)=7083021)
42
+ 38 train 9.238060 (lr=2.0455e-04) (hash(x)=6155821)
43
+ 39 train 9.037781 (lr=2.0979e-04) (hash(x)=5657123)
44
+ 40 train 8.871748 (lr=2.1503e-04) (hash(x)=5312267)
45
+ 41 train 8.838330 (lr=2.2028e-04) (hash(x)=6830813)
46
+ 42 train 8.773199 (lr=2.2552e-04) (hash(x)=6424933)
47
+ 43 train 8.698863 (lr=2.3077e-04) (hash(x)=6680023)
48
+ 44 train 8.607744 (lr=2.3601e-04) (hash(x)=7523279)
49
+ 45 train 8.475797 (lr=2.4126e-04) (hash(x)=6635925)
50
+ 46 train 8.428854 (lr=2.4650e-04) (hash(x)=5869441)
51
+ 47 train 8.376323 (lr=2.5175e-04) (hash(x)=6349004)
52
+ 48 train 8.204035 (lr=2.5699e-04) (hash(x)=6008454)
53
+ 49 train 8.187551 (lr=2.6224e-04) (hash(x)=7682741)
54
+ 50 val loss 8.1678
55
+ 50 val perplexity 3525.6497
56
+ 50 train 8.216325 (lr=2.6748e-04) (hash(x)=7794028)
57
+ 51 train 8.131626 (lr=2.7273e-04) (hash(x)=6166062)
58
+ 52 train 8.050996 (lr=2.7797e-04) (hash(x)=4811192)
59
+ 53 train 7.942014 (lr=2.8322e-04) (hash(x)=7822680)
60
+ 54 train 7.926376 (lr=2.8846e-04) (hash(x)=6431409)
61
+ 55 train 7.938275 (lr=2.9371e-04) (hash(x)=5323032)
62
+ 56 train 7.762260 (lr=2.9895e-04) (hash(x)=7363286)
63
+ 57 train 7.960917 (lr=3.0420e-04) (hash(x)=5616075)
64
+ 58 train 7.902378 (lr=3.0944e-04) (hash(x)=5672608)
65
+ 59 train 7.760784 (lr=3.1469e-04) (hash(x)=6839867)
66
+ 60 train 7.709024 (lr=3.1993e-04) (hash(x)=4916928)
67
+ 61 train 7.815749 (lr=3.2517e-04) (hash(x)=8353379)
68
+ 62 train 7.743651 (lr=3.3042e-04) (hash(x)=4838871)
69
+ 63 train 7.545825 (lr=3.3566e-04) (hash(x)=6795665)
70
+ 64 train 7.728905 (lr=3.4091e-04) (hash(x)=6678038)
71
+ 65 train 7.791191 (lr=3.4615e-04) (hash(x)=6112215)
72
+ 66 train 7.806372 (lr=3.5140e-04) (hash(x)=5915361)
73
+ 67 train 7.720731 (lr=3.5664e-04) (hash(x)=5771588)
74
+ 68 train 7.678614 (lr=3.6189e-04) (hash(x)=5688829)
75
+ 69 train 7.733504 (lr=3.6713e-04) (hash(x)=5443305)
76
+ 70 train 7.784967 (lr=3.7238e-04) (hash(x)=6901933)
77
+ 71 train 7.668831 (lr=3.7762e-04) (hash(x)=4294425)
78
+ 72 train 7.605977 (lr=3.8287e-04) (hash(x)=4363941)
79
+ 73 train 7.893797 (lr=3.8811e-04) (hash(x)=5486335)
80
+ 74 train 7.890447 (lr=3.9336e-04) (hash(x)=7295165)
81
+ 75 train 7.774928 (lr=3.9860e-04) (hash(x)=6267424)
82
+ 76 train 7.850296 (lr=4.0385e-04) (hash(x)=5270452)
83
+ 77 train 7.661283 (lr=4.0909e-04) (hash(x)=6305297)
84
+ 78 train 7.955189 (lr=4.1434e-04) (hash(x)=6040077)
85
+ 79 train 7.920587 (lr=4.1958e-04) (hash(x)=7130251)
86
+ 80 train 7.605008 (lr=4.2483e-04) (hash(x)=6508350)
87
+ 81 train 7.763794 (lr=4.3007e-04) (hash(x)=5859576)
88
+ 82 train 7.431453 (lr=4.3531e-04) (hash(x)=4892065)
89
+ 83 train 7.574790 (lr=4.4056e-04) (hash(x)=7317289)
90
+ 84 train 7.713534 (lr=4.4580e-04) (hash(x)=5510218)
91
+ 85 train 7.868369 (lr=4.5105e-04) (hash(x)=6752265)
92
+ 86 train 7.618994 (lr=4.5629e-04) (hash(x)=6530867)
93
+ 87 train 7.678219 (lr=4.6154e-04) (hash(x)=5790078)
94
+ 88 train 7.703485 (lr=4.6678e-04) (hash(x)=7614023)
95
+ 89 train 7.734940 (lr=4.7203e-04) (hash(x)=5245146)
96
+ 90 train 7.647908 (lr=4.7727e-04) (hash(x)=5786972)
97
+ 91 train 7.705590 (lr=4.8252e-04) (hash(x)=6346459)
98
+ 92 train 7.738087 (lr=4.8776e-04) (hash(x)=5209234)
99
+ 93 train 7.640960 (lr=4.9301e-04) (hash(x)=5676768)
100
+ 94 train 7.780929 (lr=4.9825e-04) (hash(x)=5863092)
101
+ 95 train 7.805684 (lr=5.0350e-04) (hash(x)=7019489)
102
+ 96 train 7.654411 (lr=5.0874e-04) (hash(x)=5483019)
103
+ 97 train 7.649938 (lr=5.1399e-04) (hash(x)=5968716)
104
+ 98 train 7.530699 (lr=5.1923e-04) (hash(x)=7780194)
105
+ 99 train 7.599035 (lr=5.2448e-04) (hash(x)=3784321)
106
+ 100 val loss 7.6126
107
+ 100 val perplexity 2023.4713
108
+ 100 train 7.634463 (lr=5.2972e-04) (hash(x)=7774109)
109
+ 101 train 7.803634 (lr=5.3497e-04) (hash(x)=6576749)
110
+ 102 train 7.694294 (lr=5.4021e-04) (hash(x)=7072404)
111
+ 103 train 7.596482 (lr=5.4545e-04) (hash(x)=6556025)
112
+ 104 train 7.686661 (lr=5.5070e-04) (hash(x)=7325027)
113
+ 105 train 7.685046 (lr=5.5594e-04) (hash(x)=6373412)
114
+ 106 train 7.576042 (lr=5.6119e-04) (hash(x)=5262868)
115
+ 107 train 7.503265 (lr=5.6643e-04) (hash(x)=8055563)
116
+ 108 train 7.657665 (lr=5.7168e-04) (hash(x)=5932790)
117
+ 109 train 7.603575 (lr=5.7692e-04) (hash(x)=5680154)
118
+ 110 train 7.640508 (lr=5.8217e-04) (hash(x)=5914217)
119
+ 111 train 7.549636 (lr=5.8741e-04) (hash(x)=6424128)
120
+ 112 train 7.496486 (lr=5.9266e-04) (hash(x)=7804089)
121
+ 113 train 7.624385 (lr=5.9790e-04) (hash(x)=6490149)
122
+ 114 train 7.565671 (lr=6.0315e-04) (hash(x)=6283540)
123
+ 115 train 7.522608 (lr=6.0839e-04) (hash(x)=5728337)
124
+ 116 train 7.535039 (lr=6.1364e-04) (hash(x)=7400272)
125
+ 117 train 7.541674 (lr=6.1888e-04) (hash(x)=6473432)
126
+ 118 train 7.644533 (lr=6.2413e-04) (hash(x)=6922646)
127
+ 119 train 7.502877 (lr=6.2937e-04) (hash(x)=7277141)
128
+ 120 train 7.570019 (lr=6.3462e-04) (hash(x)=7443124)
129
+ 121 train 7.553487 (lr=6.3986e-04) (hash(x)=5632735)
130
+ 122 train 7.421946 (lr=6.4510e-04) (hash(x)=5881015)
131
+ 123 train 7.537175 (lr=6.5035e-04) (hash(x)=5809411)
132
+ 124 train 7.529128 (lr=6.5559e-04) (hash(x)=7232964)
133
+ 125 train 7.494824 (lr=6.6084e-04) (hash(x)=6171829)
134
+ 126 train 7.585102 (lr=6.6608e-04) (hash(x)=5884456)
135
+ 127 train 7.541535 (lr=6.7133e-04) (hash(x)=5077031)
136
+ 128 train 7.593821 (lr=6.7657e-04) (hash(x)=6248895)
137
+ 129 train 7.506536 (lr=6.8182e-04) (hash(x)=6918004)
138
+ 130 train 7.581892 (lr=6.8706e-04) (hash(x)=6482466)
139
+ 131 train 7.501937 (lr=6.9231e-04) (hash(x)=6588573)
140
+ 132 train 7.551303 (lr=6.9755e-04) (hash(x)=6136939)
141
+ 133 train 7.442265 (lr=7.0280e-04) (hash(x)=7040157)
142
+ 134 train 7.556200 (lr=7.0804e-04) (hash(x)=6455595)
143
+ 135 train 7.526120 (lr=7.1329e-04) (hash(x)=4122983)
144
+ 136 train 7.462646 (lr=7.1853e-04) (hash(x)=4724052)
145
+ 137 train 7.415959 (lr=7.2378e-04) (hash(x)=6820797)
146
+ 138 train 7.666915 (lr=7.2902e-04) (hash(x)=7477594)
147
+ 139 train 7.393244 (lr=7.3427e-04) (hash(x)=6146305)
148
+ 140 train 7.408115 (lr=7.3951e-04) (hash(x)=6814831)
149
+ 141 train 7.411650 (lr=7.4476e-04) (hash(x)=5548298)
150
+ 142 train 7.596254 (lr=7.5000e-04) (hash(x)=7609416)
151
+ 143 train 7.517482 (lr=7.5524e-04) (hash(x)=7574342)
152
+ 144 train 7.363398 (lr=7.6049e-04) (hash(x)=5906937)
153
+ 145 train 7.414183 (lr=7.6573e-04) (hash(x)=8013471)
154
+ 146 train 7.412988 (lr=7.7098e-04) (hash(x)=6146213)
155
+ 147 train 7.479443 (lr=7.7622e-04) (hash(x)=6709902)
156
+ 148 train 7.567240 (lr=7.8147e-04) (hash(x)=6402571)
157
+ 149 train 7.392032 (lr=7.8671e-04) (hash(x)=7683614)
158
+ 150 val loss 7.4234
159
+ 150 val perplexity 1674.7784
160
+ 150 train 7.387026 (lr=7.9196e-04) (hash(x)=5453026)
161
+ 151 train 7.347255 (lr=7.9720e-04) (hash(x)=6260017)
162
+ 152 train 7.470717 (lr=8.0245e-04) (hash(x)=6591466)
163
+ 153 train 7.456794 (lr=8.0769e-04) (hash(x)=5501472)
164
+ 154 train 7.483077 (lr=8.1294e-04) (hash(x)=5871698)
165
+ 155 train 7.311157 (lr=8.1818e-04) (hash(x)=4282052)
166
+ 156 train 7.415814 (lr=8.2343e-04) (hash(x)=7824660)
167
+ 157 train 7.343049 (lr=8.2867e-04) (hash(x)=5954950)
168
+ 158 train 7.387298 (lr=8.3392e-04) (hash(x)=5484656)
169
+ 159 train 7.496413 (lr=8.3916e-04) (hash(x)=7031328)
170
+ 160 train 7.442774 (lr=8.4441e-04) (hash(x)=9157069)
171
+ 161 train 7.542655 (lr=8.4965e-04) (hash(x)=5996883)
172
+ 162 train 7.603206 (lr=8.5490e-04) (hash(x)=7645447)
173
+ 163 train 7.642731 (lr=8.6014e-04) (hash(x)=6336164)
174
+ 164 train 7.360022 (lr=8.6538e-04) (hash(x)=5224240)
175
+ 165 train 7.725898 (lr=8.7063e-04) (hash(x)=5447801)
176
+ 166 train 7.400049 (lr=8.7587e-04) (hash(x)=4155598)
177
+ 167 train 7.383247 (lr=8.8112e-04) (hash(x)=7482744)
178
+ 168 train 7.414576 (lr=8.8636e-04) (hash(x)=4733414)
179
+ 169 train 7.380687 (lr=8.9161e-04) (hash(x)=7806284)
180
+ 170 train 7.388601 (lr=8.9685e-04) (hash(x)=7236478)
181
+ 171 train 7.395342 (lr=9.0210e-04) (hash(x)=4656255)
182
+ 172 train 7.511054 (lr=9.0734e-04) (hash(x)=9298325)
183
+ 173 train 7.418988 (lr=9.1259e-04) (hash(x)=6287524)
184
+ 174 train 7.468686 (lr=9.1783e-04) (hash(x)=7895970)
185
+ 175 train 7.429417 (lr=9.2308e-04) (hash(x)=8847992)
186
+ 176 train 7.658670 (lr=9.2832e-04) (hash(x)=6338998)
187
+ 177 train 7.624251 (lr=9.3357e-04) (hash(x)=7104011)
188
+ 178 train 7.440904 (lr=9.3881e-04) (hash(x)=5573214)
189
+ 179 train 7.466913 (lr=9.4406e-04) (hash(x)=6882897)
190
+ 180 train 7.441818 (lr=9.4930e-04) (hash(x)=5854242)
191
+ 181 train 7.269040 (lr=9.5455e-04) (hash(x)=6972494)
192
+ 182 train 7.324336 (lr=9.5979e-04) (hash(x)=6796805)
193
+ 183 train 7.226832 (lr=9.6503e-04) (hash(x)=6139120)
194
+ 184 train 7.414863 (lr=9.7028e-04) (hash(x)=6486796)
195
+ 185 train 7.295081 (lr=9.7552e-04) (hash(x)=5578797)
196
+ 186 train 7.375923 (lr=9.8077e-04) (hash(x)=6895326)
197
+ 187 train 7.409684 (lr=9.8601e-04) (hash(x)=7783160)
198
+ 188 train 7.436715 (lr=9.9126e-04) (hash(x)=4867625)
199
+ 189 train 7.283149 (lr=9.9650e-04) (hash(x)=6631123)
200
+ 190 train 7.124108 (lr=1.0017e-03) (hash(x)=6710965)
201
+ 191 train 7.262102 (lr=1.0070e-03) (hash(x)=6294657)
202
+ 192 train 7.470639 (lr=1.0122e-03) (hash(x)=5443606)
203
+ 193 train 7.456852 (lr=1.0175e-03) (hash(x)=7035770)
204
+ 194 train 7.488454 (lr=1.0227e-03) (hash(x)=4899541)
205
+ 195 train 7.341967 (lr=1.0280e-03) (hash(x)=6227120)
206
+ 196 train 7.414204 (lr=1.0332e-03) (hash(x)=5981648)
207
+ 197 train 7.368175 (lr=1.0385e-03) (hash(x)=5558993)
208
+ 198 train 7.408672 (lr=1.0437e-03) (hash(x)=7015671)
209
+ 199 train 7.408576 (lr=1.0490e-03) (hash(x)=5931234)
210
+ 200 val loss 7.4130
211
+ 200 val perplexity 1657.3967
212
+ 200 train 7.425208 (lr=1.0542e-03) (hash(x)=4582159)
213
+ 201 train 7.408429 (lr=1.0594e-03) (hash(x)=5029472)
214
+ 202 train 7.329446 (lr=1.0647e-03) (hash(x)=5434102)
215
+ 203 train 7.584406 (lr=1.0699e-03) (hash(x)=6810638)
216
+ 204 train 7.388549 (lr=1.0752e-03) (hash(x)=6521562)
217
+ 205 train 7.435050 (lr=1.0804e-03) (hash(x)=7226932)
218
+ 206 train 7.311177 (lr=1.0857e-03) (hash(x)=6957849)
219
+ 207 train 7.429311 (lr=1.0909e-03) (hash(x)=5334281)
220
+ 208 train 7.537774 (lr=1.0962e-03) (hash(x)=5887983)
221
+ 209 train 7.363080 (lr=1.1014e-03) (hash(x)=6393576)
222
+ 210 train 7.581752 (lr=1.1066e-03) (hash(x)=5057959)
223
+ 211 train 7.630373 (lr=1.1119e-03) (hash(x)=8595271)
224
+ 212 train 7.473483 (lr=1.1171e-03) (hash(x)=8395167)
225
+ 213 train 7.603127 (lr=1.1224e-03) (hash(x)=7634417)
226
+ 214 train 7.319522 (lr=1.1276e-03) (hash(x)=5257771)
227
+ 215 train 7.465933 (lr=1.1329e-03) (hash(x)=7674238)
228
+ 216 train 7.583696 (lr=1.1381e-03) (hash(x)=5300555)
229
+ 217 train 7.366572 (lr=1.1434e-03) (hash(x)=6066139)
230
+ 218 train 7.463611 (lr=1.1486e-03) (hash(x)=6688212)
231
+ 219 train 7.635181 (lr=1.1538e-03) (hash(x)=6642404)
232
+ 220 train 7.366910 (lr=1.1591e-03) (hash(x)=4961570)
233
+ 221 train 7.259629 (lr=1.1643e-03) (hash(x)=5105190)
234
+ 222 train 7.587597 (lr=1.1696e-03) (hash(x)=7155991)
235
+ 223 train 7.404483 (lr=1.1748e-03) (hash(x)=5132972)
236
+ 224 train 7.391991 (lr=1.1801e-03) (hash(x)=6269946)
237
+ 225 train 7.434386 (lr=1.1853e-03) (hash(x)=5396133)
238
+ 226 train 7.442022 (lr=1.1906e-03) (hash(x)=6394083)
239
+ 227 train 7.391151 (lr=1.1958e-03) (hash(x)=5237804)
240
+ 228 train 7.418585 (lr=1.2010e-03) (hash(x)=5216638)
241
+ 229 train 7.341461 (lr=1.2063e-03) (hash(x)=6735565)
242
+ 230 train 7.447753 (lr=1.2115e-03) (hash(x)=5983575)
243
+ 231 train 7.428067 (lr=1.2168e-03) (hash(x)=5831619)
244
+ 232 train 7.255893 (lr=1.2220e-03) (hash(x)=9109083)
245
+ 233 train 7.474394 (lr=1.2273e-03) (hash(x)=6912967)
246
+ 234 train 7.626405 (lr=1.2325e-03) (hash(x)=5965674)
247
+ 235 train 7.476367 (lr=1.2378e-03) (hash(x)=5822533)
248
+ 236 train 7.401931 (lr=1.2430e-03) (hash(x)=5268216)
249
+ 237 train 7.328717 (lr=1.2483e-03) (hash(x)=6531900)
250
+ 238 train 7.318523 (lr=1.2535e-03) (hash(x)=5426499)
251
+ 239 train 7.750913 (lr=1.2587e-03) (hash(x)=7455372)
252
+ 240 train 8.042259 (lr=1.2640e-03) (hash(x)=8319119)
253
+ 241 train 7.988848 (lr=1.2692e-03) (hash(x)=7592128)
254
+ 242 train 7.956560 (lr=1.2745e-03) (hash(x)=8704170)
255
+ 243 train 7.661814 (lr=1.2797e-03) (hash(x)=7165037)
256
+ 244 train 7.633451 (lr=1.2850e-03) (hash(x)=7678000)
257
+ 245 train 7.539085 (lr=1.2902e-03) (hash(x)=7414788)
258
+ 246 train 7.497088 (lr=1.2955e-03) (hash(x)=7326216)
259
+ 247 train 7.597153 (lr=1.3007e-03) (hash(x)=5857162)
260
+ 248 train 8.303952 (lr=1.3059e-03) (hash(x)=14715669)
261
+ 249 train 11.717411 (lr=1.3112e-03) (hash(x)=16950560)
262
+ 250 val loss 7.5656
263
+ 250 val perplexity 1930.5465
264
+ 250 train 7.663959 (lr=1.3164e-03) (hash(x)=6159125)
265
+ 251 train 7.602240 (lr=1.3217e-03) (hash(x)=5908442)
266
+ 252 train 7.496540 (lr=1.3269e-03) (hash(x)=6668887)
267
+ 253 train 7.597426 (lr=1.3322e-03) (hash(x)=7662975)
268
+ 254 train 7.623758 (lr=1.3374e-03) (hash(x)=6669539)
269
+ 255 train 7.569995 (lr=1.3427e-03) (hash(x)=5015998)
270
+ 256 train 7.496862 (lr=1.3479e-03) (hash(x)=5458386)
271
+ 257 train 7.602964 (lr=1.3531e-03) (hash(x)=4849847)
272
+ 258 train 7.596358 (lr=1.3584e-03) (hash(x)=5125562)
273
+ 259 train 7.625378 (lr=1.3636e-03) (hash(x)=6012047)
274
+ 260 train 7.513048 (lr=1.3689e-03) (hash(x)=6079364)
275
+ 261 train 7.517653 (lr=1.3741e-03) (hash(x)=6030395)
276
+ 262 train 7.520674 (lr=1.3794e-03) (hash(x)=6781762)
277
+ 263 train 7.527929 (lr=1.3846e-03) (hash(x)=5238325)
278
+ 264 train 7.611979 (lr=1.3899e-03) (hash(x)=8611144)
279
+ 265 train 7.472679 (lr=1.3951e-03) (hash(x)=6901605)
280
+ 266 train 7.507823 (lr=1.4003e-03) (hash(x)=5499337)
281
+ 267 train 7.537289 (lr=1.4056e-03) (hash(x)=7134874)
282
+ 268 train 7.512462 (lr=1.4108e-03) (hash(x)=6149038)
283
+ 269 train 7.559241 (lr=1.4161e-03) (hash(x)=6609760)
284
+ 270 train 7.588879 (lr=1.4213e-03) (hash(x)=6652627)
285
+ 271 train 7.617541 (lr=1.4266e-03) (hash(x)=8618579)
286
+ 272 train 7.507175 (lr=1.4318e-03) (hash(x)=5896892)
287
+ 273 train 7.658423 (lr=1.4371e-03) (hash(x)=8945854)
288
+ 274 train 7.451437 (lr=1.4423e-03) (hash(x)=6775103)
289
+ 275 train 7.544943 (lr=1.4476e-03) (hash(x)=6200024)
290
+ 276 train 7.460290 (lr=1.4528e-03) (hash(x)=5238139)
291
+ 277 train 7.597555 (lr=1.4580e-03) (hash(x)=5780948)
292
+ 278 train 7.583992 (lr=1.4633e-03) (hash(x)=6168407)
293
+ 279 train 7.568002 (lr=1.4685e-03) (hash(x)=3775881)
294
+ 280 train 7.589237 (lr=1.4738e-03) (hash(x)=6848709)
295
+ 281 train 7.869195 (lr=1.4790e-03) (hash(x)=6478113)
296
+ 282 train 7.572115 (lr=1.4843e-03) (hash(x)=6483914)
297
+ 283 train 7.525204 (lr=1.4895e-03) (hash(x)=5927795)
298
+ 284 train 7.534780 (lr=1.4948e-03) (hash(x)=8438033)
299
+ 285 train 7.571095 (lr=1.5000e-03) (hash(x)=5684788)
300
+ 286 train 7.439157 (lr=1.5000e-03) (hash(x)=5710182)
301
+ 287 train 7.522766 (lr=1.5000e-03) (hash(x)=6160063)
302
+ 288 train 7.427595 (lr=1.5000e-03) (hash(x)=5955100)
303
+ 289 train 7.361252 (lr=1.4999e-03) (hash(x)=6672309)
304
+ 290 train 7.716566 (lr=1.4999e-03) (hash(x)=9001564)
305
+ 291 train 8.046438 (lr=1.4998e-03) (hash(x)=5454654)
306
+ 292 train 7.463018 (lr=1.4998e-03) (hash(x)=5174062)
307
+ 293 train 7.606758 (lr=1.4997e-03) (hash(x)=6199203)
308
+ 294 train 7.532361 (lr=1.4996e-03) (hash(x)=8352747)
309
+ 295 train 7.628793 (lr=1.4995e-03) (hash(x)=4547106)
310
+ 296 train 7.626530 (lr=1.4993e-03) (hash(x)=6433099)
311
+ 297 train 7.485325 (lr=1.4992e-03) (hash(x)=6424428)
312
+ 298 train 7.568564 (lr=1.4991e-03) (hash(x)=5975712)
313
+ 299 train 7.538754 (lr=1.4989e-03) (hash(x)=5458532)
314
+ 300 val loss 7.6289
315
+ 300 val perplexity 2056.7717
316
+ 300 train 7.758170 (lr=1.4987e-03) (hash(x)=7002774)
317
+ 301 train 7.866742 (lr=1.4985e-03) (hash(x)=8206173)
318
+ 302 train 7.927356 (lr=1.4983e-03) (hash(x)=7784971)
319
+ 303 train 7.932564 (lr=1.4981e-03) (hash(x)=6001843)
320
+ 304 train 7.723013 (lr=1.4979e-03) (hash(x)=5698642)
321
+ 305 train 7.665187 (lr=1.4976e-03) (hash(x)=6600751)
322
+ 306 train 7.643217 (lr=1.4974e-03) (hash(x)=5804610)
323
+ 307 train 7.590609 (lr=1.4971e-03) (hash(x)=5162652)
324
+ 308 train 7.563736 (lr=1.4968e-03) (hash(x)=5297006)
325
+ 309 train 7.581231 (lr=1.4965e-03) (hash(x)=6728323)
326
+ 310 train 7.588725 (lr=1.4962e-03) (hash(x)=3964882)
327
+ 311 train 7.602312 (lr=1.4959e-03) (hash(x)=6735993)
328
+ 312 train 7.727785 (lr=1.4956e-03) (hash(x)=5150316)
329
+ 313 train 7.563684 (lr=1.4952e-03) (hash(x)=5220448)
330
+ 314 train 7.719556 (lr=1.4949e-03) (hash(x)=6971289)
331
+ 315 train 7.738544 (lr=1.4945e-03) (hash(x)=5012069)
332
+ 316 train 7.608378 (lr=1.4941e-03) (hash(x)=6576136)
333
+ 317 train 7.623981 (lr=1.4937e-03) (hash(x)=6077647)
334
+ 318 train 7.722754 (lr=1.4933e-03) (hash(x)=6032381)
335
+ 319 train 7.654560 (lr=1.4929e-03) (hash(x)=8410716)
336
+ 320 train 7.659398 (lr=1.4925e-03) (hash(x)=6212516)
337
+ 321 train 7.630398 (lr=1.4920e-03) (hash(x)=5298545)
338
+ 322 train 7.659182 (lr=1.4915e-03) (hash(x)=5779073)
339
+ 323 train 7.533133 (lr=1.4911e-03) (hash(x)=6340947)
340
+ 324 train 7.319706 (lr=1.4906e-03) (hash(x)=6530441)
341
+ 325 train 7.679394 (lr=1.4901e-03) (hash(x)=7818007)
342
+ 326 train 7.517556 (lr=1.4896e-03) (hash(x)=6277435)
343
+ 327 train 7.540113 (lr=1.4890e-03) (hash(x)=6226320)
344
+ 328 train 7.688247 (lr=1.4885e-03) (hash(x)=5463751)
345
+ 329 train 7.672261 (lr=1.4880e-03) (hash(x)=6579329)
346
+ 330 train 7.777030 (lr=1.4874e-03) (hash(x)=6660540)
347
+ 331 train 7.686403 (lr=1.4868e-03) (hash(x)=6491272)
348
+ 332 train 7.664702 (lr=1.4862e-03) (hash(x)=6766947)
349
+ 333 train 7.902317 (lr=1.4856e-03) (hash(x)=3721218)
350
+ 334 train 7.824401 (lr=1.4850e-03) (hash(x)=5229898)
351
+ 335 train 7.659925 (lr=1.4844e-03) (hash(x)=6054428)
352
+ 336 train 7.775100 (lr=1.4837e-03) (hash(x)=6290464)
353
+ 337 train 7.726680 (lr=1.4831e-03) (hash(x)=6904498)
354
+ 338 train 7.740254 (lr=1.4824e-03) (hash(x)=7765805)
355
+ 339 train 7.700435 (lr=1.4817e-03) (hash(x)=6543764)
356
+ 340 train 7.725402 (lr=1.4810e-03) (hash(x)=7204955)
357
+ 341 train 7.607900 (lr=1.4803e-03) (hash(x)=4423711)
358
+ 342 train 7.549972 (lr=1.4796e-03) (hash(x)=5343477)
359
+ 343 train 7.716823 (lr=1.4789e-03) (hash(x)=6595856)
360
+ 344 train 7.578270 (lr=1.4781e-03) (hash(x)=7229015)
361
+ 345 train 7.664701 (lr=1.4774e-03) (hash(x)=5916034)
362
+ 346 train 7.842786 (lr=1.4766e-03) (hash(x)=6939521)
363
+ 347 train 7.716061 (lr=1.4758e-03) (hash(x)=5492185)
364
+ 348 train 7.611262 (lr=1.4750e-03) (hash(x)=6898006)
365
+ 349 train 7.469175 (lr=1.4742e-03) (hash(x)=5462008)
366
+ 350 val loss 7.6374
367
+ 350 val perplexity 2074.3032
368
+ 350 train 7.519867 (lr=1.4734e-03) (hash(x)=7821902)
369
+ 351 train 7.617915 (lr=1.4726e-03) (hash(x)=6608108)
370
+ 352 train 7.660865 (lr=1.4717e-03) (hash(x)=6295394)
371
+ 353 train 7.578392 (lr=1.4709e-03) (hash(x)=6803449)
372
+ 354 train 7.590785 (lr=1.4700e-03) (hash(x)=7444134)
373
+ 355 train 7.613619 (lr=1.4691e-03) (hash(x)=5361923)
374
+ 356 train 7.590117 (lr=1.4682e-03) (hash(x)=7803039)
375
+ 357 train 7.651192 (lr=1.4673e-03) (hash(x)=7042578)
376
+ 358 train 7.535811 (lr=1.4664e-03) (hash(x)=6647007)
377
+ 359 train 7.570312 (lr=1.4655e-03) (hash(x)=6303658)
378
+ 360 train 7.654228 (lr=1.4645e-03) (hash(x)=6158927)
379
+ 361 train 7.496876 (lr=1.4636e-03) (hash(x)=6419656)
380
+ 362 train 7.421102 (lr=1.4626e-03) (hash(x)=7891271)
381
+ 363 train 7.601423 (lr=1.4616e-03) (hash(x)=7099888)
382
+ 364 train 7.600333 (lr=1.4606e-03) (hash(x)=6352424)
383
+ 365 train 7.513255 (lr=1.4596e-03) (hash(x)=5863145)
384
+ 366 train 7.579518 (lr=1.4586e-03) (hash(x)=6844597)
385
+ 367 train 7.624074 (lr=1.4576e-03) (hash(x)=6009861)
386
+ 368 train 7.525167 (lr=1.4565e-03) (hash(x)=5374957)
387
+ 369 train 7.553653 (lr=1.4555e-03) (hash(x)=3923225)
388
+ 370 train 7.495900 (lr=1.4544e-03) (hash(x)=5848862)
389
+ 371 train 7.519052 (lr=1.4533e-03) (hash(x)=4440634)
390
+ 372 train 7.466400 (lr=1.4522e-03) (hash(x)=4788998)
391
+ 373 train 7.445652 (lr=1.4511e-03) (hash(x)=7378165)
392
+ 374 train 7.487044 (lr=1.4500e-03) (hash(x)=4771243)
393
+ 375 train 7.456369 (lr=1.4489e-03) (hash(x)=5716089)
394
+ 376 train 7.462033 (lr=1.4478e-03) (hash(x)=7991009)
395
+ 377 train 7.506260 (lr=1.4466e-03) (hash(x)=5348607)
396
+ 378 train 7.556457 (lr=1.4454e-03) (hash(x)=6036146)
397
+ 379 train 7.449742 (lr=1.4443e-03) (hash(x)=6658718)
398
+ 380 train 7.412056 (lr=1.4431e-03) (hash(x)=6176253)
399
+ 381 train 7.403275 (lr=1.4419e-03) (hash(x)=4542895)
400
+ 382 train 7.267520 (lr=1.4407e-03) (hash(x)=5549073)
401
+ 383 train 7.618551 (lr=1.4394e-03) (hash(x)=5342783)
402
+ 384 train 7.638938 (lr=1.4382e-03) (hash(x)=6549546)
403
+ 385 train 7.408728 (lr=1.4370e-03) (hash(x)=6170599)
404
+ 386 train 7.718097 (lr=1.4357e-03) (hash(x)=8483134)
405
+ 387 train 7.819691 (lr=1.4344e-03) (hash(x)=4865036)
406
+ 388 train 7.448786 (lr=1.4332e-03) (hash(x)=4433890)
407
+ 389 train 8.038866 (lr=1.4319e-03) (hash(x)=6738597)
408
+ 390 train 7.574594 (lr=1.4306e-03) (hash(x)=5459620)
409
+ 391 train 7.572532 (lr=1.4292e-03) (hash(x)=4604074)
410
+ 392 train 7.576383 (lr=1.4279e-03) (hash(x)=8045501)
411
+ 393 train 7.670720 (lr=1.4266e-03) (hash(x)=6176464)
412
+ 394 train 7.511273 (lr=1.4252e-03) (hash(x)=6127317)
413
+ 395 train 7.551556 (lr=1.4238e-03) (hash(x)=5571462)
414
+ 396 train 7.561952 (lr=1.4225e-03) (hash(x)=6640755)
415
+ 397 train 7.618378 (lr=1.4211e-03) (hash(x)=6147295)
416
+ 398 train 7.569329 (lr=1.4197e-03) (hash(x)=6120461)
417
+ 399 train 7.508105 (lr=1.4183e-03) (hash(x)=5123928)
418
+ 400 val loss 7.5784
419
+ 400 val perplexity 1955.4579
420
+ 400 train 7.539470 (lr=1.4169e-03) (hash(x)=5428813)
421
+ 401 train 7.514957 (lr=1.4154e-03) (hash(x)=6448586)
422
+ 402 train 7.640444 (lr=1.4140e-03) (hash(x)=8116987)
423
+ 403 train 7.642405 (lr=1.4125e-03) (hash(x)=5974342)
424
+ 404 train 7.681670 (lr=1.4110e-03) (hash(x)=5261336)
425
+ 405 train 7.483004 (lr=1.4096e-03) (hash(x)=5307056)
426
+ 406 train 7.429078 (lr=1.4081e-03) (hash(x)=5865119)
427
+ 407 train 7.459408 (lr=1.4066e-03) (hash(x)=6672380)
428
+ 408 train 7.496291 (lr=1.4051e-03) (hash(x)=6886810)
429
+ 409 train 7.647774 (lr=1.4035e-03) (hash(x)=6564071)
430
+ 410 train 7.420065 (lr=1.4020e-03) (hash(x)=7705089)
431
+ 411 train 7.426567 (lr=1.4005e-03) (hash(x)=5773073)
432
+ 412 train 7.423654 (lr=1.3989e-03) (hash(x)=5905976)
433
+ 413 train 8.040254 (lr=1.3973e-03) (hash(x)=6936653)
434
+ 414 train 7.516404 (lr=1.3957e-03) (hash(x)=6592773)
435
+ 415 train 7.478351 (lr=1.3942e-03) (hash(x)=6796551)
436
+ 416 train 7.550558 (lr=1.3926e-03) (hash(x)=6151589)
437
+ 417 train 7.779138 (lr=1.3909e-03) (hash(x)=7045164)
438
+ 418 train 7.584764 (lr=1.3893e-03) (hash(x)=6650506)
439
+ 419 train 7.518477 (lr=1.3877e-03) (hash(x)=7839748)
440
+ 420 train 7.753481 (lr=1.3860e-03) (hash(x)=6028671)
441
+ 421 train 7.578267 (lr=1.3844e-03) (hash(x)=6118408)
442
+ 422 train 7.551084 (lr=1.3827e-03) (hash(x)=5757526)
443
+ 423 train 7.665764 (lr=1.3810e-03) (hash(x)=6030172)
444
+ 424 train 7.707296 (lr=1.3793e-03) (hash(x)=6719882)
445
+ 425 train 7.586480 (lr=1.3776e-03) (hash(x)=5028050)
446
+ 426 train 7.567792 (lr=1.3759e-03) (hash(x)=5682910)
447
+ 427 train 7.663215 (lr=1.3742e-03) (hash(x)=5747891)
448
+ 428 train 7.501931 (lr=1.3725e-03) (hash(x)=5734208)
449
+ 429 train 7.558928 (lr=1.3707e-03) (hash(x)=7024381)
450
+ 430 train 7.785365 (lr=1.3690e-03) (hash(x)=7814361)
451
+ 431 train 7.656585 (lr=1.3672e-03) (hash(x)=5765893)
452
+ 432 train 7.515546 (lr=1.3654e-03) (hash(x)=5572048)
453
+ 433 train 7.655213 (lr=1.3637e-03) (hash(x)=5610058)
454
+ 434 train 7.543283 (lr=1.3619e-03) (hash(x)=4459483)
455
+ 435 train 7.503698 (lr=1.3601e-03) (hash(x)=6758411)
456
+ 436 train 7.550706 (lr=1.3582e-03) (hash(x)=6701381)
457
+ 437 train 7.679266 (lr=1.3564e-03) (hash(x)=5416400)
458
+ 438 train 7.466865 (lr=1.3546e-03) (hash(x)=5919569)
459
+ 439 train 7.612395 (lr=1.3527e-03) (hash(x)=5282870)
460
+ 440 train 7.504656 (lr=1.3509e-03) (hash(x)=8072181)
461
+ 441 train 7.450992 (lr=1.3490e-03) (hash(x)=7996770)
462
+ 442 train 7.498672 (lr=1.3471e-03) (hash(x)=5093519)
463
+ 443 train 7.537000 (lr=1.3452e-03) (hash(x)=6270917)
464
+ 444 train 7.393954 (lr=1.3434e-03) (hash(x)=5852780)
465
+ 445 train 7.278063 (lr=1.3414e-03) (hash(x)=5327728)
466
+ 446 train 7.545823 (lr=1.3395e-03) (hash(x)=6260765)
467
+ 447 train 7.618435 (lr=1.3376e-03) (hash(x)=6212894)
468
+ 448 train 7.512340 (lr=1.3357e-03) (hash(x)=6724010)
469
+ 449 train 7.588092 (lr=1.3337e-03) (hash(x)=6059934)
470
+ 450 val loss 7.6117
471
+ 450 val perplexity 2021.6727
472
+ 450 train 7.377148 (lr=1.3318e-03) (hash(x)=5587382)
473
+ 451 train 7.515634 (lr=1.3298e-03) (hash(x)=7274108)
474
+ 452 train 7.556951 (lr=1.3278e-03) (hash(x)=6085396)
475
+ 453 train 7.473550 (lr=1.3258e-03) (hash(x)=4784666)
476
+ 454 train 7.567865 (lr=1.3238e-03) (hash(x)=6288638)
477
+ 455 train 7.596221 (lr=1.3218e-03) (hash(x)=7714288)
478
+ 456 train 7.644174 (lr=1.3198e-03) (hash(x)=5534748)
479
+ 457 train 7.575220 (lr=1.3178e-03) (hash(x)=6361270)
480
+ 458 train 7.578446 (lr=1.3158e-03) (hash(x)=4682956)
481
+ 459 train 7.548480 (lr=1.3137e-03) (hash(x)=7568935)
482
+ 460 train 7.615434 (lr=1.3117e-03) (hash(x)=7973311)
483
+ 461 train 7.541690 (lr=1.3096e-03) (hash(x)=6987352)
484
+ 462 train 7.525329 (lr=1.3075e-03) (hash(x)=5413904)
485
+ 463 train 7.676631 (lr=1.3054e-03) (hash(x)=6732924)
486
+ 464 train 7.656614 (lr=1.3033e-03) (hash(x)=6219754)
487
+ 465 train 7.616294 (lr=1.3012e-03) (hash(x)=6839070)
488
+ 466 train 7.534513 (lr=1.2991e-03) (hash(x)=6090120)
489
+ 467 train 7.892575 (lr=1.2970e-03) (hash(x)=7115176)
490
+ 468 train 7.670271 (lr=1.2949e-03) (hash(x)=4989299)
491
+ 469 train 7.564256 (lr=1.2928e-03) (hash(x)=7497590)
492
+ 470 train 7.652756 (lr=1.2906e-03) (hash(x)=3977557)
493
+ 471 train 7.608488 (lr=1.2885e-03) (hash(x)=5524165)
494
+ 472 train 7.634833 (lr=1.2863e-03) (hash(x)=5124936)
495
+ 473 train 7.653940 (lr=1.2841e-03) (hash(x)=5522119)
496
+ 474 train 7.553419 (lr=1.2819e-03) (hash(x)=5030865)
497
+ 475 train 7.605977 (lr=1.2797e-03) (hash(x)=6411783)
498
+ 476 train 7.573263 (lr=1.2775e-03) (hash(x)=6581937)
499
+ 477 train 7.603786 (lr=1.2753e-03) (hash(x)=5116910)
500
+ 478 train 7.426525 (lr=1.2731e-03) (hash(x)=7353459)
501
+ 479 train 7.439530 (lr=1.2709e-03) (hash(x)=6344535)
502
+ 480 train 7.780142 (lr=1.2687e-03) (hash(x)=5840169)
503
+ 481 train 7.559095 (lr=1.2664e-03) (hash(x)=6754290)
504
+ 482 train 7.778874 (lr=1.2642e-03) (hash(x)=6484573)
505
+ 483 train 7.448756 (lr=1.2619e-03) (hash(x)=5260125)
506
+ 484 train 7.490541 (lr=1.2596e-03) (hash(x)=4881179)
507
+ 485 train 7.564227 (lr=1.2574e-03) (hash(x)=5001634)
508
+ 486 train 7.506314 (lr=1.2551e-03) (hash(x)=7424810)
509
+ 487 train 7.340068 (lr=1.2528e-03) (hash(x)=5176431)
510
+ 488 train 7.474205 (lr=1.2505e-03) (hash(x)=5421624)
511
+ 489 train 7.529961 (lr=1.2482e-03) (hash(x)=5674439)
512
+ 490 train 7.483737 (lr=1.2459e-03) (hash(x)=5819157)
513
+ 491 train 7.599583 (lr=1.2435e-03) (hash(x)=6506856)
514
+ 492 train 7.477900 (lr=1.2412e-03) (hash(x)=6119124)
515
+ 493 train 7.689557 (lr=1.2389e-03) (hash(x)=6241669)
516
+ 494 train 7.341846 (lr=1.2365e-03) (hash(x)=4992756)
517
+ 495 train 7.735241 (lr=1.2341e-03) (hash(x)=7277927)
518
+ 496 train 7.671831 (lr=1.2318e-03) (hash(x)=6622097)
519
+ 497 train 7.763649 (lr=1.2294e-03) (hash(x)=7026582)
520
+ 498 train 7.696640 (lr=1.2270e-03) (hash(x)=5840987)
521
+ 499 train 7.406112 (lr=1.2246e-03) (hash(x)=4351074)
522
+ 500 val loss 7.6584
523
+ 500 val perplexity 2118.3896
524
+ 500 train 7.544330 (lr=1.2222e-03) (hash(x)=7301477)
525
+ 501 train 7.737827 (lr=1.2198e-03) (hash(x)=6264167)
526
+ 502 train 7.493516 (lr=1.2174e-03) (hash(x)=4673363)
527
+ 503 train 7.731082 (lr=1.2150e-03) (hash(x)=6332240)
528
+ 504 train 7.626676 (lr=1.2126e-03) (hash(x)=5093790)
529
+ 505 train 7.754107 (lr=1.2101e-03) (hash(x)=6890713)
530
+ 506 train 7.533003 (lr=1.2077e-03) (hash(x)=6042691)
531
+ 507 train 7.654996 (lr=1.2052e-03) (hash(x)=7080746)
532
+ 508 train 7.648608 (lr=1.2028e-03) (hash(x)=7609207)
533
+ 509 train 7.477878 (lr=1.2003e-03) (hash(x)=4591273)
534
+ 510 train 7.621333 (lr=1.1978e-03) (hash(x)=8183111)
535
+ 511 train 7.887285 (lr=1.1954e-03) (hash(x)=5893095)
536
+ 512 train 7.653993 (lr=1.1929e-03) (hash(x)=5709077)
537
+ 513 train 7.608212 (lr=1.1904e-03) (hash(x)=6218715)
538
+ 514 train 7.495949 (lr=1.1879e-03) (hash(x)=4722188)
539
+ 515 train 7.395737 (lr=1.1854e-03) (hash(x)=6313422)
540
+ 516 train 7.542306 (lr=1.1829e-03) (hash(x)=6350660)
541
+ 517 train 7.621310 (lr=1.1803e-03) (hash(x)=5371913)
542
+ 518 train 7.743192 (lr=1.1778e-03) (hash(x)=6283595)
543
+ 519 train 7.731381 (lr=1.1753e-03) (hash(x)=6179167)
544
+ 520 train 7.613860 (lr=1.1727e-03) (hash(x)=7865194)
545
+ 521 train 7.774142 (lr=1.1702e-03) (hash(x)=4883760)
546
+ 522 train 7.578948 (lr=1.1676e-03) (hash(x)=5769286)
547
+ 523 train 7.574686 (lr=1.1651e-03) (hash(x)=6309912)
548
+ 524 train 7.542689 (lr=1.1625e-03) (hash(x)=7490849)
549
+ 525 train 7.517217 (lr=1.1599e-03) (hash(x)=5631134)
550
+ 526 train 7.533404 (lr=1.1573e-03) (hash(x)=4754560)
551
+ 527 train 7.482868 (lr=1.1548e-03) (hash(x)=5448702)
552
+ 528 train 7.668287 (lr=1.1522e-03) (hash(x)=7168842)
553
+ 529 train 7.663509 (lr=1.1496e-03) (hash(x)=7469195)
554
+ 530 train 7.470301 (lr=1.1470e-03) (hash(x)=5479500)
555
+ 531 train 7.654026 (lr=1.1443e-03) (hash(x)=5634177)
556
+ 532 train 7.554359 (lr=1.1417e-03) (hash(x)=9779982)
557
+ 533 train 7.489776 (lr=1.1391e-03) (hash(x)=3641890)
558
+ 534 train 7.590307 (lr=1.1365e-03) (hash(x)=7994103)
559
+ 535 train 7.574826 (lr=1.1338e-03) (hash(x)=5358547)
560
+ 536 train 7.562074 (lr=1.1312e-03) (hash(x)=6797424)
561
+ 537 train 7.614238 (lr=1.1285e-03) (hash(x)=4876580)
562
+ 538 train 7.417890 (lr=1.1259e-03) (hash(x)=4867629)
563
+ 539 train 7.539329 (lr=1.1232e-03) (hash(x)=6601967)
564
+ 540 train 7.516996 (lr=1.1205e-03) (hash(x)=5480657)
565
+ 541 train 7.585128 (lr=1.1179e-03) (hash(x)=7127346)
566
+ 542 train 7.728262 (lr=1.1152e-03) (hash(x)=7224794)
567
+ 543 train 7.591577 (lr=1.1125e-03) (hash(x)=5967252)
568
+ 544 train 7.681344 (lr=1.1098e-03) (hash(x)=7415262)
569
+ 545 train 7.564496 (lr=1.1071e-03) (hash(x)=6522544)
570
+ 546 train 7.597277 (lr=1.1044e-03) (hash(x)=5181640)
571
+ 547 train 7.512012 (lr=1.1017e-03) (hash(x)=5916880)
572
+ 548 train 7.489431 (lr=1.0990e-03) (hash(x)=5620397)
573
+ 549 train 7.539312 (lr=1.0963e-03) (hash(x)=5994970)
574
+ 550 val loss 7.6094
575
+ 550 val perplexity 2017.1373
576
+ 550 train 7.600243 (lr=1.0936e-03) (hash(x)=6532717)
577
+ 551 train 7.581063 (lr=1.0908e-03) (hash(x)=6262762)
578
+ 552 train 7.433823 (lr=1.0881e-03) (hash(x)=5495326)
579
+ 553 train 7.726925 (lr=1.0854e-03) (hash(x)=5927868)
580
+ 554 train 7.546543 (lr=1.0826e-03) (hash(x)=5485433)
581
+ 555 train 7.501786 (lr=1.0799e-03) (hash(x)=6462425)
582
+ 556 train 7.545203 (lr=1.0771e-03) (hash(x)=6299518)
583
+ 557 train 7.638315 (lr=1.0744e-03) (hash(x)=5950376)
584
+ 558 train 7.767035 (lr=1.0716e-03) (hash(x)=6296440)
585
+ 559 train 7.896042 (lr=1.0688e-03) (hash(x)=6974102)
586
+ 560 train 7.646054 (lr=1.0661e-03) (hash(x)=6228679)
587
+ 561 train 7.412956 (lr=1.0633e-03) (hash(x)=6120319)
588
+ 562 train 7.810788 (lr=1.0605e-03) (hash(x)=6289494)
589
+ 563 train 7.601580 (lr=1.0577e-03) (hash(x)=6877818)
590
+ 564 train 7.563552 (lr=1.0549e-03) (hash(x)=6490976)
591
+ 565 train 7.697146 (lr=1.0521e-03) (hash(x)=6419450)
592
+ 566 train 7.666745 (lr=1.0493e-03) (hash(x)=6840857)
593
+ 567 train 7.601643 (lr=1.0465e-03) (hash(x)=4837686)
594
+ 568 train 7.654056 (lr=1.0437e-03) (hash(x)=5561493)
595
+ 569 train 7.561463 (lr=1.0409e-03) (hash(x)=5831242)
596
+ 570 train 7.667708 (lr=1.0381e-03) (hash(x)=6851511)
597
+ 571 train 7.723593 (lr=1.0353e-03) (hash(x)=6169616)
598
+ 572 train 7.664905 (lr=1.0325e-03) (hash(x)=5357151)
599
+ 573 train 7.502146 (lr=1.0296e-03) (hash(x)=5471028)
600
+ 574 train 7.568137 (lr=1.0268e-03) (hash(x)=5051894)
601
+ 575 train 7.564385 (lr=1.0240e-03) (hash(x)=4930450)
602
+ 576 train 7.688916 (lr=1.0211e-03) (hash(x)=6537527)
603
+ 577 train 7.622617 (lr=1.0183e-03) (hash(x)=6330572)
604
+ 578 train 7.753276 (lr=1.0154e-03) (hash(x)=5315155)
605
+ 579 train 7.615247 (lr=1.0126e-03) (hash(x)=5416079)
606
+ 580 train 7.550607 (lr=1.0097e-03) (hash(x)=6255865)
607
+ 581 train 7.794740 (lr=1.0069e-03) (hash(x)=6723036)
608
+ 582 train 7.540442 (lr=1.0040e-03) (hash(x)=7107190)
609
+ 583 train 7.634871 (lr=1.0011e-03) (hash(x)=5868958)
610
+ 584 train 7.727970 (lr=9.9827e-04) (hash(x)=7385923)
611
+ 585 train 7.542729 (lr=9.9540e-04) (hash(x)=5373053)
612
+ 586 train 7.563944 (lr=9.9252e-04) (hash(x)=5371748)
613
+ 587 train 7.705606 (lr=9.8964e-04) (hash(x)=5783717)
614
+ 588 train 7.481566 (lr=9.8676e-04) (hash(x)=6388848)
615
+ 589 train 7.491884 (lr=9.8387e-04) (hash(x)=4865068)
616
+ 590 train 7.446493 (lr=9.8099e-04) (hash(x)=5844961)
617
+ 591 train 7.537179 (lr=9.7810e-04) (hash(x)=5989273)
618
+ 592 train 7.472892 (lr=9.7520e-04) (hash(x)=5136202)
619
+ 593 train 7.600802 (lr=9.7230e-04) (hash(x)=5811889)
620
+ 594 train 7.563529 (lr=9.6940e-04) (hash(x)=6554415)
621
+ 595 train 7.595642 (lr=9.6650e-04) (hash(x)=5969272)
622
+ 596 train 7.569786 (lr=9.6360e-04) (hash(x)=7121518)
623
+ 597 train 8.023791 (lr=9.6069e-04) (hash(x)=6363768)
624
+ 598 train 7.605362 (lr=9.5778e-04) (hash(x)=7030284)
625
+ 599 train 7.709723 (lr=9.5486e-04) (hash(x)=7243339)
626
+ 600 val loss 7.6365
627
+ 600 val perplexity 2072.4595
628
+ 600 train 7.757580 (lr=9.5195e-04) (hash(x)=7199493)
629
+ 601 train 7.830822 (lr=9.4903e-04) (hash(x)=5819281)
630
+ 602 train 7.558572 (lr=9.4611e-04) (hash(x)=7285909)
631
+ 603 train 7.586792 (lr=9.4319e-04) (hash(x)=6730878)
632
+ 604 train 7.628833 (lr=9.4026e-04) (hash(x)=5200141)
633
+ 605 train 7.710349 (lr=9.3733e-04) (hash(x)=5208573)
634
+ 606 train 7.804736 (lr=9.3440e-04) (hash(x)=5976779)
635
+ 607 train 7.625358 (lr=9.3147e-04) (hash(x)=5921767)
636
+ 608 train 7.683208 (lr=9.2854e-04) (hash(x)=6596032)
637
+ 609 train 7.708938 (lr=9.2560e-04) (hash(x)=5428058)
638
+ 610 train 7.642809 (lr=9.2267e-04) (hash(x)=5748050)
639
+ 611 train 7.651460 (lr=9.1973e-04) (hash(x)=6397421)
640
+ 612 train 7.717648 (lr=9.1678e-04) (hash(x)=5638644)
641
+ 613 train 7.644392 (lr=9.1384e-04) (hash(x)=7317623)
642
+ 614 train 7.686170 (lr=9.1090e-04) (hash(x)=5254932)
643
+ 615 train 7.763377 (lr=9.0795e-04) (hash(x)=4945141)
644
+ 616 train 7.685204 (lr=9.0500e-04) (hash(x)=7093994)
645
+ 617 train 7.626175 (lr=9.0205e-04) (hash(x)=7511894)
646
+ 618 train 7.610992 (lr=8.9910e-04) (hash(x)=6706832)
647
+ 619 train 7.571604 (lr=8.9615e-04) (hash(x)=6175371)
648
+ 620 train 7.636767 (lr=8.9319e-04) (hash(x)=7089897)
649
+ 621 train 7.619788 (lr=8.9024e-04) (hash(x)=6828202)
650
+ 622 train 7.953181 (lr=8.8728e-04) (hash(x)=7095301)
651
+ 623 train 7.639284 (lr=8.8432e-04) (hash(x)=5652191)
652
+ 624 train 7.675911 (lr=8.8136e-04) (hash(x)=7336431)
653
+ 625 train 7.672827 (lr=8.7840e-04) (hash(x)=6426339)
654
+ 626 train 7.782862 (lr=8.7544e-04) (hash(x)=5874982)
655
+ 627 train 7.728220 (lr=8.7248e-04) (hash(x)=6978359)
656
+ 628 train 7.458966 (lr=8.6952e-04) (hash(x)=4438173)
657
+ 629 train 7.562113 (lr=8.6655e-04) (hash(x)=6475356)
658
+ 630 train 7.792675 (lr=8.6359e-04) (hash(x)=5030069)
659
+ 631 train 7.547616 (lr=8.6062e-04) (hash(x)=6507939)
660
+ 632 train 7.571205 (lr=8.5766e-04) (hash(x)=6373406)
661
+ 633 train 7.703252 (lr=8.5469e-04) (hash(x)=6623130)
662
+ 634 train 7.620449 (lr=8.5172e-04) (hash(x)=6276741)
663
+ 635 train 7.534370 (lr=8.4876e-04) (hash(x)=7169874)
664
+ 636 train 7.599548 (lr=8.4579e-04) (hash(x)=4947348)
665
+ 637 train 7.690917 (lr=8.4282e-04) (hash(x)=7407715)
666
+ 638 train 7.809432 (lr=8.3985e-04) (hash(x)=5453341)
667
+ 639 train 7.571879 (lr=8.3688e-04) (hash(x)=6520161)
668
+ 640 train 7.656982 (lr=8.3391e-04) (hash(x)=6011882)
669
+ 641 train 7.622779 (lr=8.3094e-04) (hash(x)=5573372)
670
+ 642 train 7.781800 (lr=8.2797e-04) (hash(x)=5344827)
671
+ 643 train 7.845341 (lr=8.2500e-04) (hash(x)=6127887)
672
+ 644 train 7.622766 (lr=8.2203e-04) (hash(x)=6052304)
673
+ 645 train 7.639464 (lr=8.1906e-04) (hash(x)=7039733)
674
+ 646 train 7.682261 (lr=8.1609e-04) (hash(x)=7746451)
675
+ 647 train 7.870004 (lr=8.1312e-04) (hash(x)=7694373)
676
+ 648 train 7.780158 (lr=8.1015e-04) (hash(x)=6911004)
677
+ 649 train 7.697013 (lr=8.0718e-04) (hash(x)=8055224)
678
+ 650 val loss 7.6305
679
+ 650 val perplexity 2060.0186
680
+ 650 train 7.715835 (lr=8.0421e-04) (hash(x)=7060476)
681
+ 651 train 7.665342 (lr=8.0124e-04) (hash(x)=6002317)
682
+ 652 train 7.699689 (lr=7.9828e-04) (hash(x)=6524390)
683
+ 653 train 7.690838 (lr=7.9531e-04) (hash(x)=8331083)
684
+ 654 train 7.723791 (lr=7.9234e-04) (hash(x)=7084852)
685
+ 655 train 7.647449 (lr=7.8938e-04) (hash(x)=6190174)
686
+ 656 train 7.607740 (lr=7.8641e-04) (hash(x)=5601780)
687
+ 657 train 7.773505 (lr=7.8345e-04) (hash(x)=4764034)
688
+ 658 train 7.651770 (lr=7.8048e-04) (hash(x)=6361606)
689
+ 659 train 7.626073 (lr=7.7752e-04) (hash(x)=5697877)
690
+ 660 train 7.696369 (lr=7.7456e-04) (hash(x)=6249793)
691
+ 661 train 7.733772 (lr=7.7160e-04) (hash(x)=7424709)
692
+ 662 train 7.745449 (lr=7.6864e-04) (hash(x)=8203076)
693
+ 663 train 7.501095 (lr=7.6568e-04) (hash(x)=5290019)
694
+ 664 train 7.548358 (lr=7.6272e-04) (hash(x)=6278399)
695
+ 665 train 7.648700 (lr=7.5976e-04) (hash(x)=5709192)
696
+ 666 train 7.724577 (lr=7.5681e-04) (hash(x)=5660292)
697
+ 667 train 7.691051 (lr=7.5385e-04) (hash(x)=5563697)
698
+ 668 train 7.673583 (lr=7.5090e-04) (hash(x)=6325245)
699
+ 669 train 7.671924 (lr=7.4795e-04) (hash(x)=6259741)
700
+ 670 train 7.795352 (lr=7.4500e-04) (hash(x)=6238355)
701
+ 671 train 7.743815 (lr=7.4205e-04) (hash(x)=7778802)
702
+ 672 train 7.695700 (lr=7.3910e-04) (hash(x)=6283496)
703
+ 673 train 7.756269 (lr=7.3616e-04) (hash(x)=5325614)
704
+ 674 train 7.630458 (lr=7.3322e-04) (hash(x)=5376057)
705
+ 675 train 7.613010 (lr=7.3027e-04) (hash(x)=5203215)
706
+ 676 train 7.741998 (lr=7.2733e-04) (hash(x)=4893030)
707
+ 677 train 7.643628 (lr=7.2440e-04) (hash(x)=7595573)
708
+ 678 train 7.719581 (lr=7.2146e-04) (hash(x)=7188575)
709
+ 679 train 7.737191 (lr=7.1853e-04) (hash(x)=6858739)
710
+ 680 train 7.651198 (lr=7.1560e-04) (hash(x)=7157032)
711
+ 681 train 7.743367 (lr=7.1267e-04) (hash(x)=5871492)
712
+ 682 train 7.611595 (lr=7.0974e-04) (hash(x)=5636927)
713
+ 683 train 7.579608 (lr=7.0681e-04) (hash(x)=6135391)
714
+ 684 train 7.690376 (lr=7.0389e-04) (hash(x)=3351432)
715
+ 685 train 7.646334 (lr=7.0097e-04) (hash(x)=6781160)
716
+ 686 train 7.699903 (lr=6.9805e-04) (hash(x)=5824767)
717
+ 687 train 7.682377 (lr=6.9514e-04) (hash(x)=4512723)
718
+ 688 train 7.628256 (lr=6.9222e-04) (hash(x)=4600884)
719
+ 689 train 7.552753 (lr=6.8931e-04) (hash(x)=5079210)
720
+ 690 train 7.612241 (lr=6.8640e-04) (hash(x)=4671963)
721
+ 691 train 7.477903 (lr=6.8350e-04) (hash(x)=4169508)
722
+ 692 train 7.746994 (lr=6.8060e-04) (hash(x)=5341676)
723
+ 693 train 7.545423 (lr=6.7770e-04) (hash(x)=7593391)
724
+ 694 train 7.662059 (lr=6.7480e-04) (hash(x)=6723106)
725
+ 695 train 7.737453 (lr=6.7190e-04) (hash(x)=5136994)
726
+ 696 train 7.723119 (lr=6.6901e-04) (hash(x)=6571636)
727
+ 697 train 7.724689 (lr=6.6613e-04) (hash(x)=6975234)
728
+ 698 train 7.547842 (lr=6.6324e-04) (hash(x)=4868479)
729
+ 699 train 7.889332 (lr=6.6036e-04) (hash(x)=7086593)
730
+ 700 val loss 7.6620
731
+ 700 val perplexity 2125.9346
732
+ 700 train 7.716095 (lr=6.5748e-04) (hash(x)=6870808)
733
+ 701 train 8.007128 (lr=6.5460e-04) (hash(x)=7186398)
734
+ 702 train 7.777761 (lr=6.5173e-04) (hash(x)=4907570)
735
+ 703 train 7.664205 (lr=6.4886e-04) (hash(x)=5812649)
736
+ 704 train 7.688201 (lr=6.4600e-04) (hash(x)=6577447)
737
+ 705 train 7.659649 (lr=6.4314e-04) (hash(x)=5923453)
738
+ 706 train 7.859843 (lr=6.4028e-04) (hash(x)=7282962)
739
+ 707 train 7.595630 (lr=6.3742e-04) (hash(x)=4767725)
740
+ 708 train 7.616772 (lr=6.3457e-04) (hash(x)=6894202)
741
+ 709 train 8.061256 (lr=6.3172e-04) (hash(x)=4904377)
742
+ 710 train 7.648824 (lr=6.2888e-04) (hash(x)=5810443)
743
+ 711 train 7.732968 (lr=6.2604e-04) (hash(x)=5193944)
744
+ 712 train 7.671754 (lr=6.2320e-04) (hash(x)=5188803)
745
+ 713 train 7.674891 (lr=6.2037e-04) (hash(x)=7235549)
746
+ 714 train 7.839170 (lr=6.1754e-04) (hash(x)=5692576)
747
+ 715 train 7.767762 (lr=6.1472e-04) (hash(x)=5103770)
748
+ 716 train 7.737634 (lr=6.1190e-04) (hash(x)=6084786)
749
+ 717 train 7.645361 (lr=6.0908e-04) (hash(x)=5949126)
750
+ 718 train 7.728585 (lr=6.0627e-04) (hash(x)=6321938)
751
+ 719 train 7.757533 (lr=6.0346e-04) (hash(x)=6227413)
752
+ 720 train 7.806387 (lr=6.0066e-04) (hash(x)=6795372)
753
+ 721 train 7.669972 (lr=5.9786e-04) (hash(x)=6506400)
754
+ 722 train 7.673567 (lr=5.9507e-04) (hash(x)=4595742)
755
+ 723 train 7.816919 (lr=5.9228e-04) (hash(x)=7342185)
756
+ 724 train 7.822548 (lr=5.8949e-04) (hash(x)=5762071)
757
+ 725 train 7.680350 (lr=5.8671e-04) (hash(x)=6337670)
758
+ 726 train 7.724812 (lr=5.8393e-04) (hash(x)=6831170)
759
+ 727 train 7.765401 (lr=5.8116e-04) (hash(x)=7080870)
760
+ 728 train 7.687064 (lr=5.7839e-04) (hash(x)=6449970)
761
+ 729 train 7.838898 (lr=5.7563e-04) (hash(x)=5588514)
762
+ 730 train 7.637728 (lr=5.7288e-04) (hash(x)=5379746)
763
+ 731 train 7.809268 (lr=5.7012e-04) (hash(x)=8049901)
764
+ 732 train 7.863852 (lr=5.6737e-04) (hash(x)=8398209)
765
+ 733 train 7.747255 (lr=5.6463e-04) (hash(x)=6152707)
766
+ 734 train 7.806954 (lr=5.6189e-04) (hash(x)=7682617)
767
+ 735 train 7.718179 (lr=5.5916e-04) (hash(x)=7053352)
768
+ 736 train 7.782071 (lr=5.5643e-04) (hash(x)=8491505)
769
+ 737 train 7.742905 (lr=5.5371e-04) (hash(x)=6102853)
770
+ 738 train 7.815732 (lr=5.5100e-04) (hash(x)=4166591)
771
+ 739 train 7.708291 (lr=5.4828e-04) (hash(x)=7501152)
772
+ 740 train 7.664650 (lr=5.4558e-04) (hash(x)=8231949)
773
+ 741 train 7.705490 (lr=5.4288e-04) (hash(x)=5741690)
774
+ 742 train 7.832082 (lr=5.4018e-04) (hash(x)=5420426)
775
+ 743 train 7.637193 (lr=5.3749e-04) (hash(x)=5808766)
776
+ 744 train 7.609255 (lr=5.3481e-04) (hash(x)=7533322)
777
+ 745 train 7.648741 (lr=5.3213e-04) (hash(x)=6350197)
778
+ 746 train 7.768043 (lr=5.2946e-04) (hash(x)=6207170)
779
+ 747 train 7.739127 (lr=5.2679e-04) (hash(x)=5811552)
780
+ 748 train 7.684457 (lr=5.2413e-04) (hash(x)=4540607)
781
+ 749 train 7.549143 (lr=5.2147e-04) (hash(x)=5599396)
782
+ 750 val loss 7.6554
783
+ 750 val perplexity 2112.1191
784
+ 750 train 7.555338 (lr=5.1882e-04) (hash(x)=6180984)
785
+ 751 train 7.713306 (lr=5.1618e-04) (hash(x)=6519152)
786
+ 752 train 7.677018 (lr=5.1354e-04) (hash(x)=9018585)
787
+ 753 train 7.721793 (lr=5.1091e-04) (hash(x)=5988202)
788
+ 754 train 7.726769 (lr=5.0828e-04) (hash(x)=7243449)
789
+ 755 train 7.570024 (lr=5.0566e-04) (hash(x)=4377360)
790
+ 756 train 7.651197 (lr=5.0305e-04) (hash(x)=5928295)
791
+ 757 train 7.586645 (lr=5.0044e-04) (hash(x)=4996189)
792
+ 758 train 7.624186 (lr=4.9784e-04) (hash(x)=6829435)
793
+ 759 train 7.702364 (lr=4.9525e-04) (hash(x)=6929283)
794
+ 760 train 7.774829 (lr=4.9266e-04) (hash(x)=7733662)
795
+ 761 train 7.742947 (lr=4.9008e-04) (hash(x)=5975918)
796
+ 762 train 7.602759 (lr=4.8750e-04) (hash(x)=5566111)
797
+ 763 train 7.682924 (lr=4.8493e-04) (hash(x)=7407511)
798
+ 764 train 7.612732 (lr=4.8237e-04) (hash(x)=5896879)
799
+ 765 train 7.587609 (lr=4.7981e-04) (hash(x)=5569196)
800
+ 766 train 7.774881 (lr=4.7726e-04) (hash(x)=6235218)
801
+ 767 train 7.711050 (lr=4.7472e-04) (hash(x)=7467630)
802
+ 768 train 7.682931 (lr=4.7219e-04) (hash(x)=5268563)
803
+ 769 train 7.875369 (lr=4.6966e-04) (hash(x)=5915215)
804
+ 770 train 7.693166 (lr=4.6714e-04) (hash(x)=5439067)
805
+ 771 train 7.652749 (lr=4.6462e-04) (hash(x)=6167203)
806
+ 772 train 7.718116 (lr=4.6211e-04) (hash(x)=4940526)
807
+ 773 train 7.757366 (lr=4.5961e-04) (hash(x)=5897107)
808
+ 774 train 7.625262 (lr=4.5712e-04) (hash(x)=6780557)
809
+ 775 train 7.712064 (lr=4.5463e-04) (hash(x)=7566733)
810
+ 776 train 7.841503 (lr=4.5215e-04) (hash(x)=7939982)
811
+ 777 train 7.602079 (lr=4.4968e-04) (hash(x)=6447254)
812
+ 778 train 7.586424 (lr=4.4722e-04) (hash(x)=5525831)
813
+ 779 train 7.536219 (lr=4.4476e-04) (hash(x)=5059170)
814
+ 780 train 7.633175 (lr=4.4231e-04) (hash(x)=7485229)
815
+ 781 train 7.737871 (lr=4.3987e-04) (hash(x)=5955219)
816
+ 782 train 7.601485 (lr=4.3743e-04) (hash(x)=6625876)
817
+ 783 train 7.881254 (lr=4.3500e-04) (hash(x)=7141085)
818
+ 784 train 7.940269 (lr=4.3258e-04) (hash(x)=5777157)
819
+ 785 train 7.548592 (lr=4.3017e-04) (hash(x)=6236468)
820
+ 786 train 7.406718 (lr=4.2776e-04) (hash(x)=5381032)
821
+ 787 train 7.583264 (lr=4.2537e-04) (hash(x)=6733716)
822
+ 788 train 7.821527 (lr=4.2298e-04) (hash(x)=7073122)
823
+ 789 train 7.791065 (lr=4.2060e-04) (hash(x)=7622163)
824
+ 790 train 7.839431 (lr=4.1822e-04) (hash(x)=6866834)
825
+ 791 train 7.786200 (lr=4.1586e-04) (hash(x)=7008739)
826
+ 792 train 7.717621 (lr=4.1350e-04) (hash(x)=7839835)
827
+ 793 train 7.613276 (lr=4.1115e-04) (hash(x)=5313637)
828
+ 794 train 7.751465 (lr=4.0880e-04) (hash(x)=6410359)
829
+ 795 train 7.603687 (lr=4.0647e-04) (hash(x)=5721008)
830
+ 796 train 7.625974 (lr=4.0414e-04) (hash(x)=6137936)
831
+ 797 train 7.491819 (lr=4.0183e-04) (hash(x)=6881406)
832
+ 798 train 7.693223 (lr=3.9952e-04) (hash(x)=6088931)
833
+ 799 train 7.717697 (lr=3.9722e-04) (hash(x)=5689728)
834
+ 800 val loss 7.6370
835
+ 800 val perplexity 2073.5161
836
+ 800 train 7.762728 (lr=3.9492e-04) (hash(x)=5800778)
837
+ 801 train 7.675423 (lr=3.9264e-04) (hash(x)=4419021)
838
+ 802 train 7.624688 (lr=3.9036e-04) (hash(x)=5544294)
839
+ 803 train 7.697114 (lr=3.8809e-04) (hash(x)=5663946)
840
+ 804 train 7.615922 (lr=3.8583e-04) (hash(x)=4573911)
841
+ 805 train 7.557411 (lr=3.8358e-04) (hash(x)=5879465)
842
+ 806 train 7.658273 (lr=3.8134e-04) (hash(x)=5641517)
843
+ 807 train 7.802371 (lr=3.7910e-04) (hash(x)=6390336)
844
+ 808 train 7.684554 (lr=3.7688e-04) (hash(x)=8412819)
845
+ 809 train 7.816953 (lr=3.7466e-04) (hash(x)=7005536)
846
+ 810 train 7.554421 (lr=3.7245e-04) (hash(x)=6630048)
847
+ 811 train 7.682105 (lr=3.7026e-04) (hash(x)=7267424)
848
+ 812 train 7.596612 (lr=3.6806e-04) (hash(x)=6497645)
849
+ 813 train 7.536392 (lr=3.6588e-04) (hash(x)=5498697)
850
+ 814 train 7.629416 (lr=3.6371e-04) (hash(x)=6323524)
851
+ 815 train 7.594470 (lr=3.6155e-04) (hash(x)=6529460)
852
+ 816 train 7.453266 (lr=3.5939e-04) (hash(x)=5498587)
853
+ 817 train 7.627351 (lr=3.5725e-04) (hash(x)=6828893)
854
+ 818 train 7.719358 (lr=3.5511e-04) (hash(x)=6591055)
855
+ 819 train 7.591519 (lr=3.5298e-04) (hash(x)=6739572)
856
+ 820 train 7.602864 (lr=3.5086e-04) (hash(x)=5614231)
857
+ 821 train 7.652092 (lr=3.4875e-04) (hash(x)=5462474)
858
+ 822 train 7.642527 (lr=3.4665e-04) (hash(x)=5906742)
859
+ 823 train 7.728227 (lr=3.4456e-04) (hash(x)=8249268)
860
+ 824 train 7.723638 (lr=3.4248e-04) (hash(x)=4886590)
861
+ 825 train 7.503180 (lr=3.4041e-04) (hash(x)=6736150)
862
+ 826 train 7.615807 (lr=3.3835e-04) (hash(x)=6601014)
863
+ 827 train 7.682602 (lr=3.3629e-04) (hash(x)=8283020)
864
+ 828 train 7.632830 (lr=3.3425e-04) (hash(x)=6526568)
865
+ 829 train 7.712421 (lr=3.3221e-04) (hash(x)=6609259)
866
+ 830 train 7.597031 (lr=3.3019e-04) (hash(x)=6079648)
867
+ 831 train 7.573944 (lr=3.2817e-04) (hash(x)=6607178)
868
+ 832 train 7.656188 (lr=3.2617e-04) (hash(x)=6815798)
869
+ 833 train 7.867294 (lr=3.2417e-04) (hash(x)=7277017)
870
+ 834 train 7.582785 (lr=3.2219e-04) (hash(x)=5326415)
871
+ 835 train 7.635470 (lr=3.2021e-04) (hash(x)=4215944)
872
+ 836 train 7.710712 (lr=3.1824e-04) (hash(x)=4689010)
873
+ 837 train 7.668991 (lr=3.1629e-04) (hash(x)=5690469)
874
+ 838 train 7.570385 (lr=3.1434e-04) (hash(x)=6699212)
875
+ 839 train 7.556406 (lr=3.1240e-04) (hash(x)=5494876)
876
+ 840 train 7.939544 (lr=3.1047e-04) (hash(x)=10756054)
877
+ 841 train 7.602897 (lr=3.0856e-04) (hash(x)=6553219)
878
+ 842 train 7.458749 (lr=3.0665e-04) (hash(x)=7323893)
879
+ 843 train 7.903305 (lr=3.0475e-04) (hash(x)=5481970)
880
+ 844 train 7.717767 (lr=3.0286e-04) (hash(x)=6102855)
881
+ 845 train 7.666049 (lr=3.0099e-04) (hash(x)=6889406)
882
+ 846 train 7.710120 (lr=2.9912e-04) (hash(x)=4468013)
883
+ 847 train 7.524120 (lr=2.9726e-04) (hash(x)=6621321)
884
+ 848 train 7.663353 (lr=2.9542e-04) (hash(x)=6185717)
885
+ 849 train 7.558887 (lr=2.9358e-04) (hash(x)=5594828)
886
+ 850 val loss 7.6217
887
+ 850 val perplexity 2042.0741
888
+ 850 train 7.672707 (lr=2.9175e-04) (hash(x)=6530853)
889
+ 851 train 7.604573 (lr=2.8994e-04) (hash(x)=5688551)
890
+ 852 train 7.708513 (lr=2.8813e-04) (hash(x)=6979027)
891
+ 853 train 7.606856 (lr=2.8634e-04) (hash(x)=6065040)
892
+ 854 train 7.567966 (lr=2.8455e-04) (hash(x)=6836310)
893
+ 855 train 7.619243 (lr=2.8278e-04) (hash(x)=4871742)
894
+ 856 train 7.539971 (lr=2.8102e-04) (hash(x)=7461300)
895
+ 857 train 7.681456 (lr=2.7926e-04) (hash(x)=5073610)
896
+ 858 train 7.592435 (lr=2.7752e-04) (hash(x)=6725326)
897
+ 859 train 7.721680 (lr=2.7579e-04) (hash(x)=6114900)
898
+ 860 train 7.694625 (lr=2.7407e-04) (hash(x)=5249155)
899
+ 861 train 7.698965 (lr=2.7236e-04) (hash(x)=8179785)
900
+ 862 train 7.611760 (lr=2.7066e-04) (hash(x)=5588627)
901
+ 863 train 7.634285 (lr=2.6897e-04) (hash(x)=7084451)
902
+ 864 train 7.819812 (lr=2.6729e-04) (hash(x)=7478960)
903
+ 865 train 7.685376 (lr=2.6562e-04) (hash(x)=5609947)
904
+ 866 train 7.663353 (lr=2.6396e-04) (hash(x)=5741827)
905
+ 867 train 7.630328 (lr=2.6232e-04) (hash(x)=3359718)
906
+ 868 train 7.542492 (lr=2.6068e-04) (hash(x)=5480813)
907
+ 869 train 7.620015 (lr=2.5906e-04) (hash(x)=4774534)
908
+ 870 train 7.706778 (lr=2.5745e-04) (hash(x)=5895917)
909
+ 871 train 7.568813 (lr=2.5584e-04) (hash(x)=4902539)
910
+ 872 train 7.639246 (lr=2.5425e-04) (hash(x)=5734618)
911
+ 873 train 7.568552 (lr=2.5267e-04) (hash(x)=7086920)
912
+ 874 train 7.811159 (lr=2.5110e-04) (hash(x)=5338668)
913
+ 875 train 7.994176 (lr=2.4955e-04) (hash(x)=5238889)
914
+ 876 train 7.556047 (lr=2.4800e-04) (hash(x)=6244468)
915
+ 877 train 7.592893 (lr=2.4646e-04) (hash(x)=6715860)
916
+ 878 train 7.712711 (lr=2.4494e-04) (hash(x)=6216047)
917
+ 879 train 7.670996 (lr=2.4343e-04) (hash(x)=6565285)
918
+ 880 train 7.625661 (lr=2.4192e-04) (hash(x)=5257335)
919
+ 881 train 7.623128 (lr=2.4043e-04) (hash(x)=6474071)
920
+ 882 train 7.681149 (lr=2.3895e-04) (hash(x)=5705620)
921
+ 883 train 7.579249 (lr=2.3749e-04) (hash(x)=7193697)
922
+ 884 train 7.657279 (lr=2.3603e-04) (hash(x)=6264039)
923
+ 885 train 7.647198 (lr=2.3458e-04) (hash(x)=6303643)
924
+ 886 train 7.589214 (lr=2.3315e-04) (hash(x)=6120402)
925
+ 887 train 7.603663 (lr=2.3173e-04) (hash(x)=4158764)
926
+ 888 train 7.367469 (lr=2.3032e-04) (hash(x)=6173916)
927
+ 889 train 7.624500 (lr=2.2892e-04) (hash(x)=6211683)
928
+ 890 train 7.661010 (lr=2.2753e-04) (hash(x)=4950548)
929
+ 891 train 7.506677 (lr=2.2615e-04) (hash(x)=4885446)
930
+ 892 train 7.617564 (lr=2.2479e-04) (hash(x)=8117411)
931
+ 893 train 7.648463 (lr=2.2344e-04) (hash(x)=7446434)
932
+ 894 train 7.544267 (lr=2.2209e-04) (hash(x)=5371870)
933
+ 895 train 7.527243 (lr=2.2076e-04) (hash(x)=5022800)
934
+ 896 train 7.518435 (lr=2.1945e-04) (hash(x)=5666758)
935
+ 897 train 7.512265 (lr=2.1814e-04) (hash(x)=5872751)
936
+ 898 train 7.942310 (lr=2.1685e-04) (hash(x)=8428026)
937
+ 899 train 7.767822 (lr=2.1556e-04) (hash(x)=5222699)
938
+ 900 val loss 7.6077
939
+ 900 val perplexity 2013.6699
940
+ 900 train 7.637766 (lr=2.1429e-04) (hash(x)=6048200)
941
+ 901 train 7.494919 (lr=2.1303e-04) (hash(x)=5125556)
942
+ 902 train 7.517148 (lr=2.1179e-04) (hash(x)=6198583)
943
+ 903 train 7.482060 (lr=2.1055e-04) (hash(x)=5112805)
944
+ 904 train 7.569543 (lr=2.0933e-04) (hash(x)=6212803)
945
+ 905 train 7.607596 (lr=2.0812e-04) (hash(x)=5642978)
946
+ 906 train 7.600864 (lr=2.0692e-04) (hash(x)=7307168)
947
+ 907 train 7.532407 (lr=2.0573e-04) (hash(x)=6115590)
948
+ 908 train 7.441804 (lr=2.0455e-04) (hash(x)=7014937)
949
+ 909 train 7.816400 (lr=2.0339e-04) (hash(x)=6374852)
950
+ 910 train 7.530696 (lr=2.0224e-04) (hash(x)=5780051)
951
+ 911 train 7.565453 (lr=2.0110e-04) (hash(x)=4762818)
952
+ 912 train 7.414027 (lr=1.9997e-04) (hash(x)=6585420)
953
+ 913 train 7.730681 (lr=1.9885e-04) (hash(x)=6291590)
954
+ 914 train 7.781445 (lr=1.9775e-04) (hash(x)=5941561)
955
+ 915 train 7.581264 (lr=1.9666e-04) (hash(x)=6824617)
956
+ 916 train 7.681425 (lr=1.9558e-04) (hash(x)=5763341)
957
+ 917 train 7.699814 (lr=1.9451e-04) (hash(x)=6030150)
958
+ 918 train 7.623231 (lr=1.9346e-04) (hash(x)=5325490)
959
+ 919 train 7.624593 (lr=1.9242e-04) (hash(x)=6373236)
960
+ 920 train 7.643997 (lr=1.9139e-04) (hash(x)=6189151)
961
+ 921 train 7.589466 (lr=1.9037e-04) (hash(x)=6645157)
962
+ 922 train 7.601607 (lr=1.8936e-04) (hash(x)=6478394)
963
+ 923 train 7.741062 (lr=1.8837e-04) (hash(x)=5615918)
964
+ 924 train 7.628603 (lr=1.8739e-04) (hash(x)=6376988)
965
+ 925 train 7.664116 (lr=1.8642e-04) (hash(x)=5456350)
966
+ 926 train 7.696933 (lr=1.8547e-04) (hash(x)=5491217)
967
+ 927 train 7.507014 (lr=1.8452e-04) (hash(x)=6871163)
968
+ 928 train 7.524287 (lr=1.8359e-04) (hash(x)=7003708)
969
+ 929 train 7.638659 (lr=1.8267e-04) (hash(x)=7271154)
970
+ 930 train 7.606356 (lr=1.8176e-04) (hash(x)=5378966)
971
+ 931 train 7.582332 (lr=1.8087e-04) (hash(x)=5912082)
972
+ 932 train 7.508318 (lr=1.7999e-04) (hash(x)=5199606)
973
+ 933 train 7.542276 (lr=1.7912e-04) (hash(x)=5195464)
974
+ 934 train 7.490858 (lr=1.7826e-04) (hash(x)=7534052)
975
+ 935 train 7.507756 (lr=1.7742e-04) (hash(x)=6594135)
976
+ 936 train 7.575208 (lr=1.7659e-04) (hash(x)=6782051)
977
+ 937 train 7.710385 (lr=1.7577e-04) (hash(x)=6485577)
978
+ 938 train 7.641093 (lr=1.7496e-04) (hash(x)=6356813)
979
+ 939 train 7.539979 (lr=1.7417e-04) (hash(x)=6903068)
980
+ 940 train 7.610451 (lr=1.7339e-04) (hash(x)=6200862)
981
+ 941 train 7.606176 (lr=1.7262e-04) (hash(x)=5873929)
982
+ 942 train 7.611000 (lr=1.7186e-04) (hash(x)=5624440)
983
+ 943 train 7.584079 (lr=1.7112e-04) (hash(x)=7166972)
984
+ 944 train 7.697545 (lr=1.7039e-04) (hash(x)=6973234)
985
+ 945 train 7.521423 (lr=1.6967e-04) (hash(x)=6626956)
986
+ 946 train 7.661423 (lr=1.6896e-04) (hash(x)=5265592)
987
+ 947 train 7.398880 (lr=1.6827e-04) (hash(x)=5569439)
988
+ 948 train 7.527370 (lr=1.6759e-04) (hash(x)=4981059)
989
+ 949 train 7.536789 (lr=1.6692e-04) (hash(x)=7488974)
990
+ 950 val loss 7.6015
991
+ 950 val perplexity 2001.1573
992
+ 950 train 7.625254 (lr=1.6627e-04) (hash(x)=5160526)
993
+ 951 train 7.545371 (lr=1.6563e-04) (hash(x)=5340877)
994
+ 952 train 7.401459 (lr=1.6500e-04) (hash(x)=6333269)
995
+ 953 train 7.606927 (lr=1.6438e-04) (hash(x)=7205151)
996
+ 954 train 7.655581 (lr=1.6378e-04) (hash(x)=7256545)
997
+ 955 train 7.528460 (lr=1.6319e-04) (hash(x)=6374904)
998
+ 956 train 7.574969 (lr=1.6261e-04) (hash(x)=7124927)
999
+ 957 train 7.541004 (lr=1.6205e-04) (hash(x)=5263132)
1000
+ 958 train 7.719974 (lr=1.6149e-04) (hash(x)=8427208)
1001
+ 959 train 7.533807 (lr=1.6095e-04) (hash(x)=6884232)
1002
+ 960 train 7.523619 (lr=1.6043e-04) (hash(x)=7416977)
1003
+ 961 train 7.632861 (lr=1.5991e-04) (hash(x)=6733050)
1004
+ 962 train 7.545758 (lr=1.5941e-04) (hash(x)=5251404)
1005
+ 963 train 7.669500 (lr=1.5893e-04) (hash(x)=8162997)
1006
+ 964 train 7.800950 (lr=1.5845e-04) (hash(x)=7234382)
1007
+ 965 train 7.714700 (lr=1.5799e-04) (hash(x)=6255275)
1008
+ 966 train 7.664640 (lr=1.5754e-04) (hash(x)=4874742)
1009
+ 967 train 7.626337 (lr=1.5710e-04) (hash(x)=6303037)
1010
+ 968 train 7.547512 (lr=1.5668e-04) (hash(x)=5002696)
1011
+ 969 train 7.583476 (lr=1.5627e-04) (hash(x)=7636268)
1012
+ 970 train 7.564806 (lr=1.5587e-04) (hash(x)=5843590)
1013
+ 971 train 7.633920 (lr=1.5549e-04) (hash(x)=5122289)
1014
+ 972 train 7.529389 (lr=1.5512e-04) (hash(x)=5785819)
1015
+ 973 train 7.596242 (lr=1.5476e-04) (hash(x)=7350343)
1016
+ 974 train 7.417065 (lr=1.5441e-04) (hash(x)=7612943)
1017
+ 975 train 7.677065 (lr=1.5408e-04) (hash(x)=7955067)
1018
+ 976 train 7.582063 (lr=1.5376e-04) (hash(x)=5520886)
1019
+ 977 train 7.468924 (lr=1.5345e-04) (hash(x)=5299200)
1020
+ 978 train 7.503417 (lr=1.5316e-04) (hash(x)=7820550)
1021
+ 979 train 7.467607 (lr=1.5288e-04) (hash(x)=3895312)
1022
+ 980 train 7.590773 (lr=1.5261e-04) (hash(x)=4726896)
1023
+ 981 train 7.375548 (lr=1.5236e-04) (hash(x)=5061167)
1024
+ 982 train 7.556042 (lr=1.5212e-04) (hash(x)=7006804)
1025
+ 983 train 7.565936 (lr=1.5189e-04) (hash(x)=7392600)
1026
+ 984 train 7.766809 (lr=1.5167e-04) (hash(x)=7717172)
1027
+ 985 train 7.893339 (lr=1.5147e-04) (hash(x)=5800587)
1028
+ 986 train 7.572031 (lr=1.5128e-04) (hash(x)=5813784)
1029
+ 987 train 7.556030 (lr=1.5110e-04) (hash(x)=5513267)
1030
+ 988 train 7.536576 (lr=1.5094e-04) (hash(x)=7679326)
1031
+ 989 train 7.618683 (lr=1.5079e-04) (hash(x)=4820278)
1032
+ 990 train 7.631702 (lr=1.5065e-04) (hash(x)=5950845)
1033
+ 991 train 7.539297 (lr=1.5053e-04) (hash(x)=6988067)
1034
+ 992 train 7.470384 (lr=1.5042e-04) (hash(x)=4536923)
1035
+ 993 train 7.618794 (lr=1.5032e-04) (hash(x)=6874134)
1036
+ 994 train 7.847885 (lr=1.5024e-04) (hash(x)=7252013)
1037
+ 995 train 7.498935 (lr=1.5016e-04) (hash(x)=6180446)
1038
+ 996 train 7.731575 (lr=1.5010e-04) (hash(x)=5936769)
1039
+ 997 train 7.653919 (lr=1.5006e-04) (hash(x)=5627313)
1040
+ 998 train 7.604517 (lr=1.5003e-04) (hash(x)=7541743)
1041
+ 999 val loss 7.6014
1042
+ 999 val perplexity 2001.0438
1043
+ 999 train 7.944356 (lr=1.5001e-04) (hash(x)=9812956)
lr1.5e-3_total_batch_size40960_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f072eebdcdd91d21c588828879347a262378a6e48c036ffa78ef6ec124991b4
3
+ size 36163010
lr1.5e-3_total_batch_size40960_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c4eafd4ce740ce46b92a9765043d956bf6f6a947ec47422ce8ae4f850834404
3
+ size 70764358