andrew-healey commited on
Commit
35e024d
·
verified ·
1 Parent(s): e209ba9

Upload folder using huggingface_hub

Browse files
lr2e-4_total_batch_size10240_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr2e-4_total_batch_size10240_seq_len512", "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", "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": 512, "batch_size": null, "total_batch_size": 10240, "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": 0.0002, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr2e-4_total_batch_size10240_seq_len512/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b9fd0bfa27ab267155e5ae2bb7e7ccf3640df9cfec6890472ce93f6d886a52e
3
+ size 964
lr2e-4_total_batch_size10240_seq_len512/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7800
3
+ 0 val perplexity 130608.6406
4
+ 0 train 11.778345 (lr=6.9930e-07) (hash(x)=26375038)
5
+ 1 train 11.759315 (lr=1.3986e-06) (hash(x)=27234506)
6
+ 2 train 11.729763 (lr=2.0979e-06) (hash(x)=23702020)
7
+ 3 train 11.769226 (lr=2.7972e-06) (hash(x)=20782690)
8
+ 4 train 11.783836 (lr=3.4965e-06) (hash(x)=23094976)
9
+ 5 train 11.779711 (lr=4.1958e-06) (hash(x)=26532095)
10
+ 6 train 11.747478 (lr=4.8951e-06) (hash(x)=27151649)
11
+ 7 train 11.733135 (lr=5.5944e-06) (hash(x)=23890908)
12
+ 8 train 11.740841 (lr=6.2937e-06) (hash(x)=25588236)
13
+ 9 train 11.727456 (lr=6.9930e-06) (hash(x)=24011372)
14
+ 10 train 11.708902 (lr=7.6923e-06) (hash(x)=28375581)
15
+ 11 train 11.717477 (lr=8.3916e-06) (hash(x)=24611628)
16
+ 12 train 11.692226 (lr=9.0909e-06) (hash(x)=30298407)
17
+ 13 train 11.674370 (lr=9.7902e-06) (hash(x)=19245352)
18
+ 14 train 11.652133 (lr=1.0490e-05) (hash(x)=28936608)
19
+ 15 train 11.625401 (lr=1.1189e-05) (hash(x)=25767553)
20
+ 16 train 11.622272 (lr=1.1888e-05) (hash(x)=31093473)
21
+ 17 train 11.618754 (lr=1.2587e-05) (hash(x)=24809873)
22
+ 18 train 11.606547 (lr=1.3287e-05) (hash(x)=27852919)
23
+ 19 train 11.579774 (lr=1.3986e-05) (hash(x)=22512166)
24
+ 20 train 11.566683 (lr=1.4685e-05) (hash(x)=27620338)
25
+ 21 train 11.535441 (lr=1.5385e-05) (hash(x)=28092148)
26
+ 22 train 11.529295 (lr=1.6084e-05) (hash(x)=27938767)
27
+ 23 train 11.510841 (lr=1.6783e-05) (hash(x)=24732833)
28
+ 24 train 11.469143 (lr=1.7483e-05) (hash(x)=23200230)
29
+ 25 train 11.453707 (lr=1.8182e-05) (hash(x)=22694718)
30
+ 26 train 11.435548 (lr=1.8881e-05) (hash(x)=23125151)
31
+ 27 train 11.416300 (lr=1.9580e-05) (hash(x)=25129410)
32
+ 28 train 11.369334 (lr=2.0280e-05) (hash(x)=23059154)
33
+ 29 train 11.368062 (lr=2.0979e-05) (hash(x)=27858570)
34
+ 30 train 11.304667 (lr=2.1678e-05) (hash(x)=22402617)
35
+ 31 train 11.236545 (lr=2.2378e-05) (hash(x)=26582391)
36
+ 32 train 11.199144 (lr=2.3077e-05) (hash(x)=26075451)
37
+ 33 train 11.190761 (lr=2.3776e-05) (hash(x)=27279806)
38
+ 34 train 11.127916 (lr=2.4476e-05) (hash(x)=26188136)
39
+ 35 train 11.106805 (lr=2.5175e-05) (hash(x)=26472336)
40
+ 36 train 11.060472 (lr=2.5874e-05) (hash(x)=26376212)
41
+ 37 train 11.039268 (lr=2.6573e-05) (hash(x)=28301589)
42
+ 38 train 10.961640 (lr=2.7273e-05) (hash(x)=28035221)
43
+ 39 train 10.933867 (lr=2.7972e-05) (hash(x)=27654289)
44
+ 40 train 10.937832 (lr=2.8671e-05) (hash(x)=21560904)
45
+ 41 train 10.797760 (lr=2.9371e-05) (hash(x)=24995715)
46
+ 42 train 10.793970 (lr=3.0070e-05) (hash(x)=23792508)
47
+ 43 train 10.764965 (lr=3.0769e-05) (hash(x)=26559876)
48
+ 44 train 10.741049 (lr=3.1469e-05) (hash(x)=25758852)
49
+ 45 train 10.604124 (lr=3.2168e-05) (hash(x)=28364895)
50
+ 46 train 10.622024 (lr=3.2867e-05) (hash(x)=25269299)
51
+ 47 train 10.677426 (lr=3.3566e-05) (hash(x)=21770329)
52
+ 48 train 10.758147 (lr=3.4266e-05) (hash(x)=24785397)
53
+ 49 train 10.574460 (lr=3.4965e-05) (hash(x)=24628606)
54
+ 50 val loss 10.5606
55
+ 50 val perplexity 38586.1211
56
+ 50 train 10.588209 (lr=3.5664e-05) (hash(x)=23181910)
57
+ 51 train 10.510170 (lr=3.6364e-05) (hash(x)=22723459)
58
+ 52 train 10.479529 (lr=3.7063e-05) (hash(x)=25506632)
59
+ 53 train 10.367662 (lr=3.7762e-05) (hash(x)=22897919)
60
+ 54 train 10.452068 (lr=3.8462e-05) (hash(x)=26546719)
61
+ 55 train 10.672297 (lr=3.9161e-05) (hash(x)=30338342)
62
+ 56 train 10.427877 (lr=3.9860e-05) (hash(x)=29175888)
63
+ 57 train 10.482994 (lr=4.0559e-05) (hash(x)=27063280)
64
+ 58 train 10.425466 (lr=4.1259e-05) (hash(x)=32915097)
65
+ 59 train 10.383970 (lr=4.1958e-05) (hash(x)=25678059)
66
+ 60 train 10.319860 (lr=4.2657e-05) (hash(x)=20083773)
67
+ 61 train 10.309414 (lr=4.3357e-05) (hash(x)=21853028)
68
+ 62 train 10.301855 (lr=4.4056e-05) (hash(x)=21808483)
69
+ 63 train 10.366588 (lr=4.4755e-05) (hash(x)=26109335)
70
+ 64 train 10.330067 (lr=4.5455e-05) (hash(x)=24916754)
71
+ 65 train 10.409205 (lr=4.6154e-05) (hash(x)=25334848)
72
+ 66 train 10.393257 (lr=4.6853e-05) (hash(x)=24917705)
73
+ 67 train 10.196725 (lr=4.7552e-05) (hash(x)=25044885)
74
+ 68 train 10.391193 (lr=4.8252e-05) (hash(x)=28747022)
75
+ 69 train 10.257429 (lr=4.8951e-05) (hash(x)=23678349)
76
+ 70 train 10.384089 (lr=4.9650e-05) (hash(x)=28529813)
77
+ 71 train 10.334441 (lr=5.0350e-05) (hash(x)=28870690)
78
+ 72 train 10.290247 (lr=5.1049e-05) (hash(x)=25044834)
79
+ 73 train 10.391140 (lr=5.1748e-05) (hash(x)=32982615)
80
+ 74 train 10.272231 (lr=5.2448e-05) (hash(x)=20970960)
81
+ 75 train 10.419471 (lr=5.3147e-05) (hash(x)=35279941)
82
+ 76 train 10.300147 (lr=5.3846e-05) (hash(x)=25529472)
83
+ 77 train 10.264900 (lr=5.4545e-05) (hash(x)=26400041)
84
+ 78 train 10.167356 (lr=5.5245e-05) (hash(x)=24656138)
85
+ 79 train 10.164246 (lr=5.5944e-05) (hash(x)=25015923)
86
+ 80 train 10.101791 (lr=5.6643e-05) (hash(x)=22924885)
87
+ 81 train 10.167073 (lr=5.7343e-05) (hash(x)=24853586)
88
+ 82 train 10.201824 (lr=5.8042e-05) (hash(x)=27004780)
89
+ 83 train 10.103807 (lr=5.8741e-05) (hash(x)=26740855)
90
+ 84 train 9.977927 (lr=5.9441e-05) (hash(x)=23950114)
91
+ 85 train 10.067707 (lr=6.0140e-05) (hash(x)=24584116)
92
+ 86 train 10.111980 (lr=6.0839e-05) (hash(x)=25718516)
93
+ 87 train 10.062920 (lr=6.1538e-05) (hash(x)=22262151)
94
+ 88 train 10.114091 (lr=6.2238e-05) (hash(x)=23025303)
95
+ 89 train 9.967682 (lr=6.2937e-05) (hash(x)=26627860)
96
+ 90 train 9.957551 (lr=6.3636e-05) (hash(x)=23700532)
97
+ 91 train 9.935550 (lr=6.4336e-05) (hash(x)=26621834)
98
+ 92 train 9.969267 (lr=6.5035e-05) (hash(x)=21821465)
99
+ 93 train 10.038383 (lr=6.5734e-05) (hash(x)=26930630)
100
+ 94 train 9.966433 (lr=6.6434e-05) (hash(x)=19557946)
101
+ 95 train 9.981304 (lr=6.7133e-05) (hash(x)=23234741)
102
+ 96 train 10.005640 (lr=6.7832e-05) (hash(x)=26713563)
103
+ 97 train 9.981532 (lr=6.8531e-05) (hash(x)=28102443)
104
+ 98 train 10.091217 (lr=6.9231e-05) (hash(x)=31689122)
105
+ 99 train 10.067239 (lr=6.9930e-05) (hash(x)=25470563)
106
+ 100 val loss 9.9518
107
+ 100 val perplexity 20990.0586
108
+ 100 train 10.031240 (lr=7.0629e-05) (hash(x)=23757479)
109
+ 101 train 9.884790 (lr=7.1329e-05) (hash(x)=28282861)
110
+ 102 train 9.948372 (lr=7.2028e-05) (hash(x)=26731964)
111
+ 103 train 9.988102 (lr=7.2727e-05) (hash(x)=21799102)
112
+ 104 train 9.878646 (lr=7.3427e-05) (hash(x)=24672077)
113
+ 105 train 9.863609 (lr=7.4126e-05) (hash(x)=23471769)
114
+ 106 train 9.871107 (lr=7.4825e-05) (hash(x)=24058931)
115
+ 107 train 9.844282 (lr=7.5524e-05) (hash(x)=23275479)
116
+ 108 train 9.926214 (lr=7.6224e-05) (hash(x)=24364396)
117
+ 109 train 9.890710 (lr=7.6923e-05) (hash(x)=24473581)
118
+ 110 train 9.880389 (lr=7.7622e-05) (hash(x)=23774644)
119
+ 111 train 9.887297 (lr=7.8322e-05) (hash(x)=22091862)
120
+ 112 train 9.676316 (lr=7.9021e-05) (hash(x)=25482303)
121
+ 113 train 9.838821 (lr=7.9720e-05) (hash(x)=26018202)
122
+ 114 train 9.755852 (lr=8.0420e-05) (hash(x)=24560096)
123
+ 115 train 9.754541 (lr=8.1119e-05) (hash(x)=17430373)
124
+ 116 train 9.584257 (lr=8.1818e-05) (hash(x)=20098681)
125
+ 117 train 9.644231 (lr=8.2517e-05) (hash(x)=24078083)
126
+ 118 train 9.732801 (lr=8.3217e-05) (hash(x)=26595592)
127
+ 119 train 9.667046 (lr=8.3916e-05) (hash(x)=22954861)
128
+ 120 train 9.746193 (lr=8.4615e-05) (hash(x)=29294271)
129
+ 121 train 9.742992 (lr=8.5315e-05) (hash(x)=28695016)
130
+ 122 train 9.685379 (lr=8.6014e-05) (hash(x)=27100115)
131
+ 123 train 9.728571 (lr=8.6713e-05) (hash(x)=26671799)
132
+ 124 train 9.741045 (lr=8.7413e-05) (hash(x)=28137394)
133
+ 125 train 9.703804 (lr=8.8112e-05) (hash(x)=23166092)
134
+ 126 train 9.544738 (lr=8.8811e-05) (hash(x)=25885986)
135
+ 127 train 9.575171 (lr=8.9510e-05) (hash(x)=26924723)
136
+ 128 train 9.642015 (lr=9.0210e-05) (hash(x)=26140590)
137
+ 129 train 9.641081 (lr=9.0909e-05) (hash(x)=25138659)
138
+ 130 train 9.660898 (lr=9.1608e-05) (hash(x)=27224685)
139
+ 131 train 9.610392 (lr=9.2308e-05) (hash(x)=23557495)
140
+ 132 train 9.536975 (lr=9.3007e-05) (hash(x)=23928957)
141
+ 133 train 9.509787 (lr=9.3706e-05) (hash(x)=25288123)
142
+ 134 train 9.451111 (lr=9.4406e-05) (hash(x)=23246294)
143
+ 135 train 9.260262 (lr=9.5105e-05) (hash(x)=27804380)
144
+ 136 train 9.492875 (lr=9.5804e-05) (hash(x)=23089140)
145
+ 137 train 9.637875 (lr=9.6503e-05) (hash(x)=26993263)
146
+ 138 train 9.617380 (lr=9.7203e-05) (hash(x)=27397203)
147
+ 139 train 9.539831 (lr=9.7902e-05) (hash(x)=26654908)
148
+ 140 train 9.439618 (lr=9.8601e-05) (hash(x)=21965022)
149
+ 141 train 9.428035 (lr=9.9301e-05) (hash(x)=27347722)
150
+ 142 train 9.505228 (lr=1.0000e-04) (hash(x)=28314127)
151
+ 143 train 9.464972 (lr=1.0070e-04) (hash(x)=23627518)
152
+ 144 train 9.391310 (lr=1.0140e-04) (hash(x)=25024764)
153
+ 145 train 9.354172 (lr=1.0210e-04) (hash(x)=21768671)
154
+ 146 train 9.415707 (lr=1.0280e-04) (hash(x)=25076667)
155
+ 147 train 9.331602 (lr=1.0350e-04) (hash(x)=21889990)
156
+ 148 train 9.271196 (lr=1.0420e-04) (hash(x)=20932794)
157
+ 149 train 9.304211 (lr=1.0490e-04) (hash(x)=23665838)
158
+ 150 val loss 9.3274
159
+ 150 val perplexity 11242.0811
160
+ 150 train 9.724253 (lr=1.0559e-04) (hash(x)=30223582)
161
+ 151 train 9.270841 (lr=1.0629e-04) (hash(x)=22528001)
162
+ 152 train 9.355865 (lr=1.0699e-04) (hash(x)=25181641)
163
+ 153 train 9.279172 (lr=1.0769e-04) (hash(x)=22487328)
164
+ 154 train 9.372306 (lr=1.0839e-04) (hash(x)=26544630)
165
+ 155 train 9.403337 (lr=1.0909e-04) (hash(x)=29248942)
166
+ 156 train 9.292528 (lr=1.0979e-04) (hash(x)=25052066)
167
+ 157 train 9.180784 (lr=1.1049e-04) (hash(x)=23543273)
168
+ 158 train 9.253603 (lr=1.1119e-04) (hash(x)=27451288)
169
+ 159 train 9.142655 (lr=1.1189e-04) (hash(x)=25533417)
170
+ 160 train 9.195129 (lr=1.1259e-04) (hash(x)=24432899)
171
+ 161 train 9.133533 (lr=1.1329e-04) (hash(x)=25552036)
172
+ 162 train 9.127208 (lr=1.1399e-04) (hash(x)=28149782)
173
+ 163 train 9.066425 (lr=1.1469e-04) (hash(x)=20612533)
174
+ 164 train 9.015961 (lr=1.1538e-04) (hash(x)=19774173)
175
+ 165 train 9.014953 (lr=1.1608e-04) (hash(x)=20216795)
176
+ 166 train 9.217413 (lr=1.1678e-04) (hash(x)=25896435)
177
+ 167 train 9.121740 (lr=1.1748e-04) (hash(x)=24677740)
178
+ 168 train 9.098120 (lr=1.1818e-04) (hash(x)=21835643)
179
+ 169 train 9.002363 (lr=1.1888e-04) (hash(x)=16722715)
180
+ 170 train 9.111904 (lr=1.1958e-04) (hash(x)=26303975)
181
+ 171 train 9.021294 (lr=1.2028e-04) (hash(x)=25347203)
182
+ 172 train 9.016046 (lr=1.2098e-04) (hash(x)=25973417)
183
+ 173 train 8.986620 (lr=1.2168e-04) (hash(x)=24272489)
184
+ 174 train 8.912743 (lr=1.2238e-04) (hash(x)=29260846)
185
+ 175 train 9.017519 (lr=1.2308e-04) (hash(x)=28922363)
186
+ 176 train 8.937632 (lr=1.2378e-04) (hash(x)=26322572)
187
+ 177 train 8.904175 (lr=1.2448e-04) (hash(x)=25510798)
188
+ 178 train 8.990856 (lr=1.2517e-04) (hash(x)=24138948)
189
+ 179 train 8.908133 (lr=1.2587e-04) (hash(x)=25631397)
190
+ 180 train 8.818033 (lr=1.2657e-04) (hash(x)=24001903)
191
+ 181 train 8.769192 (lr=1.2727e-04) (hash(x)=21333676)
192
+ 182 train 8.972531 (lr=1.2797e-04) (hash(x)=30770484)
193
+ 183 train 8.887758 (lr=1.2867e-04) (hash(x)=26237983)
194
+ 184 train 8.788082 (lr=1.2937e-04) (hash(x)=26205744)
195
+ 185 train 8.852192 (lr=1.3007e-04) (hash(x)=25976502)
196
+ 186 train 8.708207 (lr=1.3077e-04) (hash(x)=24080636)
197
+ 187 train 8.893688 (lr=1.3147e-04) (hash(x)=24742645)
198
+ 188 train 8.865075 (lr=1.3217e-04) (hash(x)=28677257)
199
+ 189 train 8.745569 (lr=1.3287e-04) (hash(x)=27086333)
200
+ 190 train 8.747423 (lr=1.3357e-04) (hash(x)=27855233)
201
+ 191 train 8.731753 (lr=1.3427e-04) (hash(x)=26157660)
202
+ 192 train 8.731244 (lr=1.3497e-04) (hash(x)=25046062)
203
+ 193 train 8.739289 (lr=1.3566e-04) (hash(x)=23616370)
204
+ 194 train 8.510015 (lr=1.3636e-04) (hash(x)=24302232)
205
+ 195 train 8.551735 (lr=1.3706e-04) (hash(x)=22582169)
206
+ 196 train 8.601795 (lr=1.3776e-04) (hash(x)=22773833)
207
+ 197 train 8.521738 (lr=1.3846e-04) (hash(x)=26233189)
208
+ 198 train 8.566988 (lr=1.3916e-04) (hash(x)=26431507)
209
+ 199 train 8.716501 (lr=1.3986e-04) (hash(x)=26431960)
210
+ 200 val loss 8.6367
211
+ 200 val perplexity 5634.7241
212
+ 200 train 8.620134 (lr=1.4056e-04) (hash(x)=25112360)
213
+ 201 train 8.779966 (lr=1.4126e-04) (hash(x)=30707498)
214
+ 202 train 8.522272 (lr=1.4196e-04) (hash(x)=22586447)
215
+ 203 train 8.623441 (lr=1.4266e-04) (hash(x)=25716176)
216
+ 204 train 8.698410 (lr=1.4336e-04) (hash(x)=25118971)
217
+ 205 train 8.425785 (lr=1.4406e-04) (hash(x)=26256420)
218
+ 206 train 8.530362 (lr=1.4476e-04) (hash(x)=22633318)
219
+ 207 train 8.583665 (lr=1.4545e-04) (hash(x)=28019788)
220
+ 208 train 8.515127 (lr=1.4615e-04) (hash(x)=27396089)
221
+ 209 train 8.502008 (lr=1.4685e-04) (hash(x)=23510110)
222
+ 210 train 8.579361 (lr=1.4755e-04) (hash(x)=23646926)
223
+ 211 train 8.414207 (lr=1.4825e-04) (hash(x)=20382963)
224
+ 212 train 8.390449 (lr=1.4895e-04) (hash(x)=24304768)
225
+ 213 train 8.329317 (lr=1.4965e-04) (hash(x)=25339466)
226
+ 214 train 8.374365 (lr=1.5035e-04) (hash(x)=22088696)
227
+ 215 train 8.416453 (lr=1.5105e-04) (hash(x)=24489647)
228
+ 216 train 8.313549 (lr=1.5175e-04) (hash(x)=22940357)
229
+ 217 train 8.352634 (lr=1.5245e-04) (hash(x)=22927699)
230
+ 218 train 8.381731 (lr=1.5315e-04) (hash(x)=27631132)
231
+ 219 train 8.372436 (lr=1.5385e-04) (hash(x)=25128502)
232
+ 220 train 8.395638 (lr=1.5455e-04) (hash(x)=23576982)
233
+ 221 train 8.400121 (lr=1.5524e-04) (hash(x)=25808969)
234
+ 222 train 8.325457 (lr=1.5594e-04) (hash(x)=23429962)
235
+ 223 train 8.289244 (lr=1.5664e-04) (hash(x)=25231390)
236
+ 224 train 8.187084 (lr=1.5734e-04) (hash(x)=25395441)
237
+ 225 train 8.195137 (lr=1.5804e-04) (hash(x)=21154388)
238
+ 226 train 8.096304 (lr=1.5874e-04) (hash(x)=24278611)
239
+ 227 train 8.175392 (lr=1.5944e-04) (hash(x)=22708977)
240
+ 228 train 8.553381 (lr=1.6014e-04) (hash(x)=28228490)
241
+ 229 train 8.362590 (lr=1.6084e-04) (hash(x)=27258353)
242
+ 230 train 8.223016 (lr=1.6154e-04) (hash(x)=23252199)
243
+ 231 train 8.154102 (lr=1.6224e-04) (hash(x)=24364920)
244
+ 232 train 8.181886 (lr=1.6294e-04) (hash(x)=23283905)
245
+ 233 train 8.283387 (lr=1.6364e-04) (hash(x)=27243972)
246
+ 234 train 8.396345 (lr=1.6434e-04) (hash(x)=28113043)
247
+ 235 train 8.117346 (lr=1.6503e-04) (hash(x)=24932925)
248
+ 236 train 8.002536 (lr=1.6573e-04) (hash(x)=19001259)
249
+ 237 train 8.263828 (lr=1.6643e-04) (hash(x)=26371091)
250
+ 238 train 8.012959 (lr=1.6713e-04) (hash(x)=24258817)
251
+ 239 train 8.122945 (lr=1.6783e-04) (hash(x)=26913684)
252
+ 240 train 8.216684 (lr=1.6853e-04) (hash(x)=27753043)
253
+ 241 train 8.106688 (lr=1.6923e-04) (hash(x)=29239611)
254
+ 242 train 8.081544 (lr=1.6993e-04) (hash(x)=22984557)
255
+ 243 train 7.647861 (lr=1.7063e-04) (hash(x)=18017792)
256
+ 244 train 7.991449 (lr=1.7133e-04) (hash(x)=23374526)
257
+ 245 train 8.033855 (lr=1.7203e-04) (hash(x)=27574089)
258
+ 246 train 8.094481 (lr=1.7273e-04) (hash(x)=26154906)
259
+ 247 train 8.060309 (lr=1.7343e-04) (hash(x)=28613882)
260
+ 248 train 8.001915 (lr=1.7413e-04) (hash(x)=25272182)
261
+ 249 train 7.997422 (lr=1.7483e-04) (hash(x)=26981914)
262
+ 250 val loss 7.9935
263
+ 250 val perplexity 2961.6565
264
+ 250 train 8.003801 (lr=1.7552e-04) (hash(x)=24232348)
265
+ 251 train 8.017166 (lr=1.7622e-04) (hash(x)=22652243)
266
+ 252 train 8.078834 (lr=1.7692e-04) (hash(x)=25411991)
267
+ 253 train 7.450638 (lr=1.7762e-04) (hash(x)=15446025)
268
+ 254 train 7.831789 (lr=1.7832e-04) (hash(x)=27846176)
269
+ 255 train 7.901950 (lr=1.7902e-04) (hash(x)=27194521)
270
+ 256 train 8.023551 (lr=1.7972e-04) (hash(x)=29430001)
271
+ 257 train 7.663667 (lr=1.8042e-04) (hash(x)=25264518)
272
+ 258 train 7.971980 (lr=1.8112e-04) (hash(x)=25568956)
273
+ 259 train 8.036889 (lr=1.8182e-04) (hash(x)=24936836)
274
+ 260 train 8.012579 (lr=1.8252e-04) (hash(x)=27452099)
275
+ 261 train 8.018111 (lr=1.8322e-04) (hash(x)=28197282)
276
+ 262 train 7.905655 (lr=1.8392e-04) (hash(x)=22931889)
277
+ 263 train 7.951476 (lr=1.8462e-04) (hash(x)=24765578)
278
+ 264 train 7.884571 (lr=1.8531e-04) (hash(x)=25078649)
279
+ 265 train 7.997386 (lr=1.8601e-04) (hash(x)=27448185)
280
+ 266 train 8.005283 (lr=1.8671e-04) (hash(x)=24804856)
281
+ 267 train 7.941479 (lr=1.8741e-04) (hash(x)=23107416)
282
+ 268 train 7.771015 (lr=1.8811e-04) (hash(x)=24960796)
283
+ 269 train 7.960511 (lr=1.8881e-04) (hash(x)=26755138)
284
+ 270 train 7.845014 (lr=1.8951e-04) (hash(x)=24542526)
285
+ 271 train 7.836392 (lr=1.9021e-04) (hash(x)=29314255)
286
+ 272 train 7.890931 (lr=1.9091e-04) (hash(x)=27347755)
287
+ 273 train 7.805033 (lr=1.9161e-04) (hash(x)=22112669)
288
+ 274 train 7.758472 (lr=1.9231e-04) (hash(x)=25161929)
289
+ 275 train 7.828492 (lr=1.9301e-04) (hash(x)=24013079)
290
+ 276 train 7.811082 (lr=1.9371e-04) (hash(x)=26768629)
291
+ 277 train 7.746100 (lr=1.9441e-04) (hash(x)=22537194)
292
+ 278 train 7.867784 (lr=1.9510e-04) (hash(x)=24974360)
293
+ 279 train 7.946060 (lr=1.9580e-04) (hash(x)=25322001)
294
+ 280 train 7.754342 (lr=1.9650e-04) (hash(x)=20562247)
295
+ 281 train 8.045461 (lr=1.9720e-04) (hash(x)=27028126)
296
+ 282 train 7.908051 (lr=1.9790e-04) (hash(x)=25666355)
297
+ 283 train 7.913308 (lr=1.9860e-04) (hash(x)=26690440)
298
+ 284 train 7.823193 (lr=1.9930e-04) (hash(x)=27348418)
299
+ 285 train 7.921791 (lr=2.0000e-04) (hash(x)=26967331)
300
+ 286 train 7.743981 (lr=2.0000e-04) (hash(x)=24765121)
301
+ 287 train 7.701179 (lr=2.0000e-04) (hash(x)=22598512)
302
+ 288 train 7.724514 (lr=2.0000e-04) (hash(x)=23725598)
303
+ 289 train 7.922475 (lr=1.9999e-04) (hash(x)=24697985)
304
+ 290 train 7.827572 (lr=1.9999e-04) (hash(x)=25201962)
305
+ 291 train 7.436179 (lr=1.9998e-04) (hash(x)=18473911)
306
+ 292 train 7.736965 (lr=1.9997e-04) (hash(x)=25982840)
307
+ 293 train 7.702813 (lr=1.9996e-04) (hash(x)=23172124)
308
+ 294 train 8.220150 (lr=1.9994e-04) (hash(x)=34646114)
309
+ 295 train 8.239010 (lr=1.9993e-04) (hash(x)=22322442)
310
+ 296 train 7.900182 (lr=1.9991e-04) (hash(x)=26211794)
311
+ 297 train 8.199044 (lr=1.9989e-04) (hash(x)=29659121)
312
+ 298 train 7.924465 (lr=1.9987e-04) (hash(x)=23538306)
313
+ 299 train 7.847977 (lr=1.9985e-04) (hash(x)=26324153)
314
+ 300 val loss 7.7870
315
+ 300 val perplexity 2409.0305
316
+ 300 train 7.792036 (lr=1.9983e-04) (hash(x)=24910403)
317
+ 301 train 7.865362 (lr=1.9980e-04) (hash(x)=26939970)
318
+ 302 train 7.930648 (lr=1.9978e-04) (hash(x)=27515446)
319
+ 303 train 7.823668 (lr=1.9975e-04) (hash(x)=23446058)
320
+ 304 train 7.916668 (lr=1.9972e-04) (hash(x)=25308253)
321
+ 305 train 7.910151 (lr=1.9969e-04) (hash(x)=29981801)
322
+ 306 train 8.006103 (lr=1.9965e-04) (hash(x)=23470413)
323
+ 307 train 7.908872 (lr=1.9962e-04) (hash(x)=24843741)
324
+ 308 train 7.871950 (lr=1.9958e-04) (hash(x)=27176996)
325
+ 309 train 7.734477 (lr=1.9954e-04) (hash(x)=22855363)
326
+ 310 train 7.628004 (lr=1.9950e-04) (hash(x)=22960957)
327
+ 311 train 7.794220 (lr=1.9946e-04) (hash(x)=23197102)
328
+ 312 train 7.991750 (lr=1.9941e-04) (hash(x)=26132276)
329
+ 313 train 7.717097 (lr=1.9937e-04) (hash(x)=25704919)
330
+ 314 train 7.940597 (lr=1.9932e-04) (hash(x)=22280814)
331
+ 315 train 8.409871 (lr=1.9927e-04) (hash(x)=23054940)
332
+ 316 train 8.361383 (lr=1.9922e-04) (hash(x)=24712416)
333
+ 317 train 7.965640 (lr=1.9916e-04) (hash(x)=24909904)
334
+ 318 train 7.902899 (lr=1.9911e-04) (hash(x)=24321591)
335
+ 319 train 7.763721 (lr=1.9905e-04) (hash(x)=27140757)
336
+ 320 train 7.630103 (lr=1.9899e-04) (hash(x)=26563770)
337
+ 321 train 7.795620 (lr=1.9893e-04) (hash(x)=25993110)
338
+ 322 train 7.809607 (lr=1.9887e-04) (hash(x)=28520222)
339
+ 323 train 7.632773 (lr=1.9881e-04) (hash(x)=25149419)
340
+ 324 train 7.909283 (lr=1.9874e-04) (hash(x)=26112813)
341
+ 325 train 7.821724 (lr=1.9868e-04) (hash(x)=27623643)
342
+ 326 train 7.840419 (lr=1.9861e-04) (hash(x)=29199854)
343
+ 327 train 7.796492 (lr=1.9854e-04) (hash(x)=24727764)
344
+ 328 train 7.401352 (lr=1.9847e-04) (hash(x)=22667179)
345
+ 329 train 7.403568 (lr=1.9839e-04) (hash(x)=23155773)
346
+ 330 train 7.546809 (lr=1.9832e-04) (hash(x)=23453788)
347
+ 331 train 7.680786 (lr=1.9824e-04) (hash(x)=21935931)
348
+ 332 train 7.576919 (lr=1.9816e-04) (hash(x)=24481079)
349
+ 333 train 7.446402 (lr=1.9808e-04) (hash(x)=20054917)
350
+ 334 train 7.699393 (lr=1.9800e-04) (hash(x)=25907741)
351
+ 335 train 7.853298 (lr=1.9792e-04) (hash(x)=27606073)
352
+ 336 train 7.883813 (lr=1.9783e-04) (hash(x)=25202001)
353
+ 337 train 7.844582 (lr=1.9774e-04) (hash(x)=26534487)
354
+ 338 train 7.753564 (lr=1.9765e-04) (hash(x)=24095850)
355
+ 339 train 7.655380 (lr=1.9756e-04) (hash(x)=22450341)
356
+ 340 train 7.980665 (lr=1.9747e-04) (hash(x)=27993763)
357
+ 341 train 7.758741 (lr=1.9738e-04) (hash(x)=25013073)
358
+ 342 train 7.916817 (lr=1.9728e-04) (hash(x)=30934371)
359
+ 343 train 8.134861 (lr=1.9718e-04) (hash(x)=25945859)
360
+ 344 train 7.680793 (lr=1.9709e-04) (hash(x)=25218689)
361
+ 345 train 7.783450 (lr=1.9698e-04) (hash(x)=24418091)
362
+ 346 train 7.662191 (lr=1.9688e-04) (hash(x)=24615466)
363
+ 347 train 7.746573 (lr=1.9678e-04) (hash(x)=21613707)
364
+ 348 train 7.755913 (lr=1.9667e-04) (hash(x)=24348175)
365
+ 349 train 7.847481 (lr=1.9656e-04) (hash(x)=26386157)
366
+ 350 val loss 7.7424
367
+ 350 val perplexity 2303.9036
368
+ 350 train 7.737929 (lr=1.9646e-04) (hash(x)=25771047)
369
+ 351 train 7.740759 (lr=1.9634e-04) (hash(x)=24396519)
370
+ 352 train 7.863147 (lr=1.9623e-04) (hash(x)=25142399)
371
+ 353 train 7.675673 (lr=1.9612e-04) (hash(x)=23265507)
372
+ 354 train 7.660965 (lr=1.9600e-04) (hash(x)=22514858)
373
+ 355 train 7.968797 (lr=1.9588e-04) (hash(x)=29138828)
374
+ 356 train 7.883157 (lr=1.9576e-04) (hash(x)=20531210)
375
+ 357 train 7.724294 (lr=1.9564e-04) (hash(x)=24265353)
376
+ 358 train 7.787123 (lr=1.9552e-04) (hash(x)=24999726)
377
+ 359 train 7.806787 (lr=1.9540e-04) (hash(x)=24685515)
378
+ 360 train 7.688496 (lr=1.9527e-04) (hash(x)=25071701)
379
+ 361 train 7.728350 (lr=1.9514e-04) (hash(x)=24876269)
380
+ 362 train 8.001037 (lr=1.9501e-04) (hash(x)=31733180)
381
+ 363 train 7.546807 (lr=1.9488e-04) (hash(x)=22825749)
382
+ 364 train 7.836477 (lr=1.9475e-04) (hash(x)=26393943)
383
+ 365 train 7.728835 (lr=1.9462e-04) (hash(x)=21239652)
384
+ 366 train 7.573585 (lr=1.9448e-04) (hash(x)=24943881)
385
+ 367 train 8.008204 (lr=1.9434e-04) (hash(x)=28767869)
386
+ 368 train 7.863286 (lr=1.9421e-04) (hash(x)=29341482)
387
+ 369 train 7.804516 (lr=1.9406e-04) (hash(x)=25614301)
388
+ 370 train 7.766779 (lr=1.9392e-04) (hash(x)=26030058)
389
+ 371 train 7.857593 (lr=1.9378e-04) (hash(x)=21159060)
390
+ 372 train 7.709328 (lr=1.9363e-04) (hash(x)=24629937)
391
+ 373 train 7.735011 (lr=1.9349e-04) (hash(x)=26751788)
392
+ 374 train 7.738707 (lr=1.9334e-04) (hash(x)=26012353)
393
+ 375 train 7.705181 (lr=1.9319e-04) (hash(x)=25045397)
394
+ 376 train 7.389698 (lr=1.9304e-04) (hash(x)=23751143)
395
+ 377 train 8.216215 (lr=1.9288e-04) (hash(x)=29989709)
396
+ 378 train 7.847696 (lr=1.9273e-04) (hash(x)=31017246)
397
+ 379 train 7.996924 (lr=1.9257e-04) (hash(x)=25560928)
398
+ 380 train 7.775724 (lr=1.9241e-04) (hash(x)=27352253)
399
+ 381 train 7.942854 (lr=1.9225e-04) (hash(x)=26062687)
400
+ 382 train 7.916823 (lr=1.9209e-04) (hash(x)=27614522)
401
+ 383 train 7.963489 (lr=1.9193e-04) (hash(x)=28959222)
402
+ 384 train 7.720595 (lr=1.9176e-04) (hash(x)=25122598)
403
+ 385 train 7.704392 (lr=1.9160e-04) (hash(x)=25434884)
404
+ 386 train 7.821337 (lr=1.9143e-04) (hash(x)=27824438)
405
+ 387 train 7.758404 (lr=1.9126e-04) (hash(x)=21248405)
406
+ 388 train 7.610556 (lr=1.9109e-04) (hash(x)=26482588)
407
+ 389 train 7.843441 (lr=1.9091e-04) (hash(x)=25044192)
408
+ 390 train 7.841958 (lr=1.9074e-04) (hash(x)=28645524)
409
+ 391 train 7.700953 (lr=1.9056e-04) (hash(x)=25830182)
410
+ 392 train 7.777417 (lr=1.9039e-04) (hash(x)=25585137)
411
+ 393 train 7.741518 (lr=1.9021e-04) (hash(x)=26621419)
412
+ 394 train 7.357493 (lr=1.9003e-04) (hash(x)=22165286)
413
+ 395 train 7.669366 (lr=1.8985e-04) (hash(x)=25173113)
414
+ 396 train 7.740449 (lr=1.8966e-04) (hash(x)=27267091)
415
+ 397 train 7.620950 (lr=1.8948e-04) (hash(x)=24400133)
416
+ 398 train 7.622626 (lr=1.8929e-04) (hash(x)=23103223)
417
+ 399 train 7.758910 (lr=1.8910e-04) (hash(x)=23486277)
418
+ 400 val loss 7.6930
419
+ 400 val perplexity 2192.9568
420
+ 400 train 7.631185 (lr=1.8891e-04) (hash(x)=23421286)
421
+ 401 train 7.759375 (lr=1.8872e-04) (hash(x)=25978130)
422
+ 402 train 7.762667 (lr=1.8853e-04) (hash(x)=25769432)
423
+ 403 train 7.752173 (lr=1.8834e-04) (hash(x)=28916006)
424
+ 404 train 7.782666 (lr=1.8814e-04) (hash(x)=24226056)
425
+ 405 train 8.045784 (lr=1.8794e-04) (hash(x)=33195100)
426
+ 406 train 7.943968 (lr=1.8774e-04) (hash(x)=27730410)
427
+ 407 train 7.838285 (lr=1.8754e-04) (hash(x)=25556929)
428
+ 408 train 7.708522 (lr=1.8734e-04) (hash(x)=25991247)
429
+ 409 train 7.911963 (lr=1.8714e-04) (hash(x)=29536986)
430
+ 410 train 7.707332 (lr=1.8693e-04) (hash(x)=26125216)
431
+ 411 train 7.819774 (lr=1.8673e-04) (hash(x)=29648798)
432
+ 412 train 7.652878 (lr=1.8652e-04) (hash(x)=23195388)
433
+ 413 train 7.711874 (lr=1.8631e-04) (hash(x)=23124767)
434
+ 414 train 7.767929 (lr=1.8610e-04) (hash(x)=25713275)
435
+ 415 train 7.648935 (lr=1.8589e-04) (hash(x)=24976217)
436
+ 416 train 7.844446 (lr=1.8567e-04) (hash(x)=27500801)
437
+ 417 train 7.631258 (lr=1.8546e-04) (hash(x)=23632825)
438
+ 418 train 7.719457 (lr=1.8524e-04) (hash(x)=24456276)
439
+ 419 train 7.949103 (lr=1.8502e-04) (hash(x)=31019606)
440
+ 420 train 7.898537 (lr=1.8480e-04) (hash(x)=25027904)
441
+ 421 train 7.769646 (lr=1.8458e-04) (hash(x)=28236580)
442
+ 422 train 7.700110 (lr=1.8436e-04) (hash(x)=25386473)
443
+ 423 train 7.807137 (lr=1.8414e-04) (hash(x)=28269421)
444
+ 424 train 7.464246 (lr=1.8391e-04) (hash(x)=24308447)
445
+ 425 train 7.647800 (lr=1.8369e-04) (hash(x)=25563279)
446
+ 426 train 7.816898 (lr=1.8346e-04) (hash(x)=26152637)
447
+ 427 train 7.834219 (lr=1.8323e-04) (hash(x)=24865358)
448
+ 428 train 7.586648 (lr=1.8300e-04) (hash(x)=22230964)
449
+ 429 train 7.506748 (lr=1.8276e-04) (hash(x)=22155546)
450
+ 430 train 7.779869 (lr=1.8253e-04) (hash(x)=29142319)
451
+ 431 train 7.614188 (lr=1.8230e-04) (hash(x)=26984272)
452
+ 432 train 7.716298 (lr=1.8206e-04) (hash(x)=27077032)
453
+ 433 train 7.717042 (lr=1.8182e-04) (hash(x)=23193673)
454
+ 434 train 7.784412 (lr=1.8158e-04) (hash(x)=27112558)
455
+ 435 train 7.794037 (lr=1.8134e-04) (hash(x)=24544116)
456
+ 436 train 7.829594 (lr=1.8110e-04) (hash(x)=25890184)
457
+ 437 train 7.671092 (lr=1.8086e-04) (hash(x)=24547533)
458
+ 438 train 7.831473 (lr=1.8061e-04) (hash(x)=27467688)
459
+ 439 train 7.605223 (lr=1.8036e-04) (hash(x)=22059850)
460
+ 440 train 7.818040 (lr=1.8012e-04) (hash(x)=23893130)
461
+ 441 train 7.754534 (lr=1.7987e-04) (hash(x)=25994573)
462
+ 442 train 7.738853 (lr=1.7962e-04) (hash(x)=26577783)
463
+ 443 train 7.732792 (lr=1.7937e-04) (hash(x)=23926632)
464
+ 444 train 7.633517 (lr=1.7911e-04) (hash(x)=23574207)
465
+ 445 train 7.552030 (lr=1.7886e-04) (hash(x)=27111369)
466
+ 446 train 7.782650 (lr=1.7860e-04) (hash(x)=23979820)
467
+ 447 train 7.722751 (lr=1.7835e-04) (hash(x)=27025333)
468
+ 448 train 7.568152 (lr=1.7809e-04) (hash(x)=22846386)
469
+ 449 train 7.494260 (lr=1.7783e-04) (hash(x)=16908068)
470
+ 450 val loss 7.6705
471
+ 450 val perplexity 2144.2561
472
+ 450 train 7.676370 (lr=1.7757e-04) (hash(x)=25664727)
473
+ 451 train 7.737933 (lr=1.7731e-04) (hash(x)=29120407)
474
+ 452 train 7.778521 (lr=1.7704e-04) (hash(x)=25564213)
475
+ 453 train 7.856479 (lr=1.7678e-04) (hash(x)=27092710)
476
+ 454 train 7.672310 (lr=1.7651e-04) (hash(x)=28533197)
477
+ 455 train 7.617917 (lr=1.7624e-04) (hash(x)=23827393)
478
+ 456 train 7.539447 (lr=1.7597e-04) (hash(x)=24154233)
479
+ 457 train 7.854918 (lr=1.7570e-04) (hash(x)=32812727)
480
+ 458 train 7.646098 (lr=1.7543e-04) (hash(x)=26305435)
481
+ 459 train 7.775898 (lr=1.7516e-04) (hash(x)=27230027)
482
+ 460 train 7.627555 (lr=1.7489e-04) (hash(x)=23532437)
483
+ 461 train 7.730556 (lr=1.7461e-04) (hash(x)=25703381)
484
+ 462 train 7.715212 (lr=1.7434e-04) (hash(x)=26961429)
485
+ 463 train 7.524785 (lr=1.7406e-04) (hash(x)=24968260)
486
+ 464 train 7.630342 (lr=1.7378e-04) (hash(x)=24854265)
487
+ 465 train 7.583740 (lr=1.7350e-04) (hash(x)=25676468)
488
+ 466 train 7.717508 (lr=1.7322e-04) (hash(x)=25503865)
489
+ 467 train 7.706821 (lr=1.7294e-04) (hash(x)=27544803)
490
+ 468 train 7.675137 (lr=1.7265e-04) (hash(x)=24658683)
491
+ 469 train 7.685749 (lr=1.7237e-04) (hash(x)=24331407)
492
+ 470 train 7.708381 (lr=1.7208e-04) (hash(x)=21892556)
493
+ 471 train 7.591566 (lr=1.7179e-04) (hash(x)=26540663)
494
+ 472 train 7.612132 (lr=1.7151e-04) (hash(x)=26819462)
495
+ 473 train 7.717683 (lr=1.7122e-04) (hash(x)=25532657)
496
+ 474 train 7.795162 (lr=1.7092e-04) (hash(x)=26515570)
497
+ 475 train 7.950225 (lr=1.7063e-04) (hash(x)=25856625)
498
+ 476 train 7.543755 (lr=1.7034e-04) (hash(x)=25260471)
499
+ 477 train 7.676700 (lr=1.7004e-04) (hash(x)=23437426)
500
+ 478 train 7.768860 (lr=1.6975e-04) (hash(x)=25961833)
501
+ 479 train 7.570470 (lr=1.6945e-04) (hash(x)=23164356)
502
+ 480 train 7.631579 (lr=1.6915e-04) (hash(x)=25052665)
503
+ 481 train 7.712637 (lr=1.6886e-04) (hash(x)=25957896)
504
+ 482 train 7.741110 (lr=1.6856e-04) (hash(x)=24723263)
505
+ 483 train 7.713758 (lr=1.6825e-04) (hash(x)=25646282)
506
+ 484 train 7.704785 (lr=1.6795e-04) (hash(x)=25851993)
507
+ 485 train 7.804077 (lr=1.6765e-04) (hash(x)=21513584)
508
+ 486 train 7.525423 (lr=1.6734e-04) (hash(x)=26017176)
509
+ 487 train 7.743313 (lr=1.6704e-04) (hash(x)=23843233)
510
+ 488 train 7.725643 (lr=1.6673e-04) (hash(x)=21476847)
511
+ 489 train 7.523561 (lr=1.6642e-04) (hash(x)=23226697)
512
+ 490 train 7.328732 (lr=1.6611e-04) (hash(x)=19772969)
513
+ 491 train 7.664299 (lr=1.6580e-04) (hash(x)=24506028)
514
+ 492 train 7.649084 (lr=1.6549e-04) (hash(x)=21077417)
515
+ 493 train 7.602457 (lr=1.6518e-04) (hash(x)=26024321)
516
+ 494 train 7.806077 (lr=1.6487e-04) (hash(x)=27786174)
517
+ 495 train 7.636961 (lr=1.6455e-04) (hash(x)=22781277)
518
+ 496 train 7.692656 (lr=1.6424e-04) (hash(x)=24459895)
519
+ 497 train 7.725582 (lr=1.6392e-04) (hash(x)=24451843)
520
+ 498 train 7.695851 (lr=1.6360e-04) (hash(x)=21116390)
521
+ 499 train 7.634015 (lr=1.6328e-04) (hash(x)=24948650)
522
+ 500 val loss 7.6655
523
+ 500 val perplexity 2133.4058
524
+ 500 train 7.646587 (lr=1.6296e-04) (hash(x)=25617781)
525
+ 501 train 7.554218 (lr=1.6264e-04) (hash(x)=25559534)
526
+ 502 train 7.684109 (lr=1.6232e-04) (hash(x)=25546274)
527
+ 503 train 7.702458 (lr=1.6200e-04) (hash(x)=26373991)
528
+ 504 train 7.630882 (lr=1.6168e-04) (hash(x)=26410662)
529
+ 505 train 7.893291 (lr=1.6135e-04) (hash(x)=27756673)
530
+ 506 train 7.770826 (lr=1.6103e-04) (hash(x)=23661686)
531
+ 507 train 7.651261 (lr=1.6070e-04) (hash(x)=23402396)
532
+ 508 train 7.766200 (lr=1.6037e-04) (hash(x)=26547918)
533
+ 509 train 7.734052 (lr=1.6004e-04) (hash(x)=28250354)
534
+ 510 train 7.734498 (lr=1.5971e-04) (hash(x)=25598861)
535
+ 511 train 7.558579 (lr=1.5938e-04) (hash(x)=21562133)
536
+ 512 train 7.832348 (lr=1.5905e-04) (hash(x)=25312391)
537
+ 513 train 7.833619 (lr=1.5872e-04) (hash(x)=27253861)
538
+ 514 train 7.677152 (lr=1.5838e-04) (hash(x)=26795921)
539
+ 515 train 7.721768 (lr=1.5805e-04) (hash(x)=25009210)
540
+ 516 train 7.619933 (lr=1.5772e-04) (hash(x)=26254538)
541
+ 517 train 7.547478 (lr=1.5738e-04) (hash(x)=23052577)
542
+ 518 train 7.867107 (lr=1.5704e-04) (hash(x)=23667224)
543
+ 519 train 7.660439 (lr=1.5670e-04) (hash(x)=27161811)
544
+ 520 train 7.661022 (lr=1.5636e-04) (hash(x)=26313522)
545
+ 521 train 7.698356 (lr=1.5602e-04) (hash(x)=27302459)
546
+ 522 train 7.841386 (lr=1.5568e-04) (hash(x)=26837963)
547
+ 523 train 7.715165 (lr=1.5534e-04) (hash(x)=29515100)
548
+ 524 train 7.677564 (lr=1.5500e-04) (hash(x)=26674478)
549
+ 525 train 7.616675 (lr=1.5466e-04) (hash(x)=25318001)
550
+ 526 train 7.747953 (lr=1.5431e-04) (hash(x)=25174043)
551
+ 527 train 7.594449 (lr=1.5397e-04) (hash(x)=25305929)
552
+ 528 train 7.762452 (lr=1.5362e-04) (hash(x)=26175477)
553
+ 529 train 7.554554 (lr=1.5327e-04) (hash(x)=24339043)
554
+ 530 train 7.834504 (lr=1.5293e-04) (hash(x)=25412772)
555
+ 531 train 7.628323 (lr=1.5258e-04) (hash(x)=26808543)
556
+ 532 train 7.482993 (lr=1.5223e-04) (hash(x)=24738650)
557
+ 533 train 7.504852 (lr=1.5188e-04) (hash(x)=25763991)
558
+ 534 train 7.809273 (lr=1.5153e-04) (hash(x)=28182190)
559
+ 535 train 7.618079 (lr=1.5118e-04) (hash(x)=25669509)
560
+ 536 train 7.757321 (lr=1.5082e-04) (hash(x)=25616522)
561
+ 537 train 7.735270 (lr=1.5047e-04) (hash(x)=27676869)
562
+ 538 train 7.667931 (lr=1.5012e-04) (hash(x)=26316170)
563
+ 539 train 7.697899 (lr=1.4976e-04) (hash(x)=28435805)
564
+ 540 train 7.645321 (lr=1.4941e-04) (hash(x)=26869022)
565
+ 541 train 7.539723 (lr=1.4905e-04) (hash(x)=23455211)
566
+ 542 train 7.717746 (lr=1.4869e-04) (hash(x)=29321187)
567
+ 543 train 7.811454 (lr=1.4833e-04) (hash(x)=26482758)
568
+ 544 train 7.558270 (lr=1.4798e-04) (hash(x)=26046639)
569
+ 545 train 7.671374 (lr=1.4762e-04) (hash(x)=27068280)
570
+ 546 train 7.584450 (lr=1.4726e-04) (hash(x)=26782091)
571
+ 547 train 7.696431 (lr=1.4690e-04) (hash(x)=24929929)
572
+ 548 train 7.629688 (lr=1.4653e-04) (hash(x)=25950541)
573
+ 549 train 7.552819 (lr=1.4617e-04) (hash(x)=25373676)
574
+ 550 val loss 7.6131
575
+ 550 val perplexity 2024.5040
576
+ 550 train 7.581905 (lr=1.4581e-04) (hash(x)=27283187)
577
+ 551 train 7.560523 (lr=1.4545e-04) (hash(x)=25043037)
578
+ 552 train 7.660954 (lr=1.4508e-04) (hash(x)=27027534)
579
+ 553 train 7.600185 (lr=1.4472e-04) (hash(x)=27906976)
580
+ 554 train 7.640665 (lr=1.4435e-04) (hash(x)=23889709)
581
+ 555 train 8.411392 (lr=1.4398e-04) (hash(x)=29151257)
582
+ 556 train 7.762849 (lr=1.4362e-04) (hash(x)=21307974)
583
+ 557 train 7.701726 (lr=1.4325e-04) (hash(x)=26405613)
584
+ 558 train 7.496874 (lr=1.4288e-04) (hash(x)=24233887)
585
+ 559 train 7.432425 (lr=1.4251e-04) (hash(x)=26400365)
586
+ 560 train 7.683167 (lr=1.4214e-04) (hash(x)=22293114)
587
+ 561 train 7.497227 (lr=1.4177e-04) (hash(x)=23971905)
588
+ 562 train 7.466388 (lr=1.4140e-04) (hash(x)=20076502)
589
+ 563 train 7.595924 (lr=1.4103e-04) (hash(x)=24098812)
590
+ 564 train 7.661902 (lr=1.4066e-04) (hash(x)=27610410)
591
+ 565 train 7.704368 (lr=1.4029e-04) (hash(x)=29402976)
592
+ 566 train 7.342626 (lr=1.3991e-04) (hash(x)=19032564)
593
+ 567 train 7.601684 (lr=1.3954e-04) (hash(x)=20967666)
594
+ 568 train 7.469843 (lr=1.3916e-04) (hash(x)=22666342)
595
+ 569 train 7.661534 (lr=1.3879e-04) (hash(x)=24287610)
596
+ 570 train 7.856849 (lr=1.3841e-04) (hash(x)=28637634)
597
+ 571 train 8.029533 (lr=1.3804e-04) (hash(x)=28667963)
598
+ 572 train 7.755369 (lr=1.3766e-04) (hash(x)=23621685)
599
+ 573 train 7.518147 (lr=1.3728e-04) (hash(x)=25974316)
600
+ 574 train 7.561283 (lr=1.3691e-04) (hash(x)=24487710)
601
+ 575 train 7.886398 (lr=1.3653e-04) (hash(x)=27083417)
602
+ 576 train 7.361611 (lr=1.3615e-04) (hash(x)=22850633)
603
+ 577 train 7.767897 (lr=1.3577e-04) (hash(x)=28605205)
604
+ 578 train 7.538476 (lr=1.3539e-04) (hash(x)=24786468)
605
+ 579 train 7.495500 (lr=1.3501e-04) (hash(x)=23649001)
606
+ 580 train 7.642161 (lr=1.3463e-04) (hash(x)=26473994)
607
+ 581 train 7.404466 (lr=1.3425e-04) (hash(x)=25921700)
608
+ 582 train 7.428670 (lr=1.3387e-04) (hash(x)=24117626)
609
+ 583 train 7.856445 (lr=1.3348e-04) (hash(x)=31338300)
610
+ 584 train 7.493081 (lr=1.3310e-04) (hash(x)=21337692)
611
+ 585 train 7.616754 (lr=1.3272e-04) (hash(x)=23862328)
612
+ 586 train 7.660124 (lr=1.3234e-04) (hash(x)=24380098)
613
+ 587 train 7.532677 (lr=1.3195e-04) (hash(x)=30603174)
614
+ 588 train 7.859857 (lr=1.3157e-04) (hash(x)=30882548)
615
+ 589 train 7.409451 (lr=1.3118e-04) (hash(x)=23603806)
616
+ 590 train 7.610026 (lr=1.3080e-04) (hash(x)=24382442)
617
+ 591 train 7.501811 (lr=1.3041e-04) (hash(x)=24830965)
618
+ 592 train 7.615319 (lr=1.3003e-04) (hash(x)=25547480)
619
+ 593 train 7.685455 (lr=1.2964e-04) (hash(x)=32488729)
620
+ 594 train 7.951099 (lr=1.2925e-04) (hash(x)=31331643)
621
+ 595 train 7.691801 (lr=1.2887e-04) (hash(x)=27269893)
622
+ 596 train 7.533004 (lr=1.2848e-04) (hash(x)=23278725)
623
+ 597 train 7.639126 (lr=1.2809e-04) (hash(x)=25215077)
624
+ 598 train 7.567766 (lr=1.2770e-04) (hash(x)=22442404)
625
+ 599 train 7.496792 (lr=1.2732e-04) (hash(x)=24825041)
626
+ 600 val loss 7.5729
627
+ 600 val perplexity 1944.8180
628
+ 600 train 7.672813 (lr=1.2693e-04) (hash(x)=25590276)
629
+ 601 train 7.566641 (lr=1.2654e-04) (hash(x)=25893804)
630
+ 602 train 7.456523 (lr=1.2615e-04) (hash(x)=22222468)
631
+ 603 train 7.427452 (lr=1.2576e-04) (hash(x)=22724871)
632
+ 604 train 7.476730 (lr=1.2537e-04) (hash(x)=25140744)
633
+ 605 train 7.565583 (lr=1.2498e-04) (hash(x)=27317554)
634
+ 606 train 7.546112 (lr=1.2459e-04) (hash(x)=26692343)
635
+ 607 train 7.340284 (lr=1.2420e-04) (hash(x)=24683717)
636
+ 608 train 7.482062 (lr=1.2381e-04) (hash(x)=26849303)
637
+ 609 train 7.568760 (lr=1.2341e-04) (hash(x)=21584083)
638
+ 610 train 7.677362 (lr=1.2302e-04) (hash(x)=26059939)
639
+ 611 train 7.706423 (lr=1.2263e-04) (hash(x)=26190337)
640
+ 612 train 7.650302 (lr=1.2224e-04) (hash(x)=23663918)
641
+ 613 train 7.587960 (lr=1.2185e-04) (hash(x)=22249019)
642
+ 614 train 7.623057 (lr=1.2145e-04) (hash(x)=25779849)
643
+ 615 train 7.579261 (lr=1.2106e-04) (hash(x)=23086289)
644
+ 616 train 7.682725 (lr=1.2067e-04) (hash(x)=27523941)
645
+ 617 train 7.848848 (lr=1.2027e-04) (hash(x)=28047044)
646
+ 618 train 7.583896 (lr=1.1988e-04) (hash(x)=25125498)
647
+ 619 train 7.444304 (lr=1.1949e-04) (hash(x)=23593875)
648
+ 620 train 7.512729 (lr=1.1909e-04) (hash(x)=25117733)
649
+ 621 train 7.525025 (lr=1.1870e-04) (hash(x)=26059735)
650
+ 622 train 7.468286 (lr=1.1830e-04) (hash(x)=23028679)
651
+ 623 train 7.363839 (lr=1.1791e-04) (hash(x)=22414190)
652
+ 624 train 7.515203 (lr=1.1752e-04) (hash(x)=22223137)
653
+ 625 train 7.738564 (lr=1.1712e-04) (hash(x)=24204591)
654
+ 626 train 7.390532 (lr=1.1673e-04) (hash(x)=22922475)
655
+ 627 train 7.530886 (lr=1.1633e-04) (hash(x)=26758114)
656
+ 628 train 7.458814 (lr=1.1594e-04) (hash(x)=25358064)
657
+ 629 train 7.588549 (lr=1.1554e-04) (hash(x)=20084233)
658
+ 630 train 7.561078 (lr=1.1515e-04) (hash(x)=26395104)
659
+ 631 train 7.368420 (lr=1.1475e-04) (hash(x)=21563184)
660
+ 632 train 7.589888 (lr=1.1435e-04) (hash(x)=25207694)
661
+ 633 train 7.453976 (lr=1.1396e-04) (hash(x)=23861489)
662
+ 634 train 7.659530 (lr=1.1356e-04) (hash(x)=23471007)
663
+ 635 train 7.401705 (lr=1.1317e-04) (hash(x)=19218470)
664
+ 636 train 7.531822 (lr=1.1277e-04) (hash(x)=22944035)
665
+ 637 train 7.588962 (lr=1.1238e-04) (hash(x)=28469562)
666
+ 638 train 7.793529 (lr=1.1198e-04) (hash(x)=26452814)
667
+ 639 train 7.663237 (lr=1.1158e-04) (hash(x)=25588469)
668
+ 640 train 7.503589 (lr=1.1119e-04) (hash(x)=24788298)
669
+ 641 train 7.676475 (lr=1.1079e-04) (hash(x)=25236367)
670
+ 642 train 7.426974 (lr=1.1040e-04) (hash(x)=23069067)
671
+ 643 train 7.826178 (lr=1.1000e-04) (hash(x)=30433767)
672
+ 644 train 8.037190 (lr=1.0960e-04) (hash(x)=29268881)
673
+ 645 train 7.660329 (lr=1.0921e-04) (hash(x)=26302626)
674
+ 646 train 7.818522 (lr=1.0881e-04) (hash(x)=24749682)
675
+ 647 train 7.356236 (lr=1.0842e-04) (hash(x)=17851621)
676
+ 648 train 7.522370 (lr=1.0802e-04) (hash(x)=25620741)
677
+ 649 train 8.050497 (lr=1.0762e-04) (hash(x)=27028191)
678
+ 650 val loss 7.5577
679
+ 650 val perplexity 1915.3685
680
+ 650 train 7.700192 (lr=1.0723e-04) (hash(x)=27590299)
681
+ 651 train 7.763016 (lr=1.0683e-04) (hash(x)=25183690)
682
+ 652 train 7.667663 (lr=1.0644e-04) (hash(x)=31114252)
683
+ 653 train 7.787441 (lr=1.0604e-04) (hash(x)=26182243)
684
+ 654 train 7.909230 (lr=1.0565e-04) (hash(x)=30419908)
685
+ 655 train 7.784121 (lr=1.0525e-04) (hash(x)=27040642)
686
+ 656 train 7.578979 (lr=1.0485e-04) (hash(x)=26240761)
687
+ 657 train 7.594895 (lr=1.0446e-04) (hash(x)=24489607)
688
+ 658 train 7.719632 (lr=1.0406e-04) (hash(x)=25012872)
689
+ 659 train 7.266679 (lr=1.0367e-04) (hash(x)=22183303)
690
+ 660 train 7.587895 (lr=1.0327e-04) (hash(x)=27682633)
691
+ 661 train 7.682282 (lr=1.0288e-04) (hash(x)=28844646)
692
+ 662 train 8.284440 (lr=1.0248e-04) (hash(x)=31082070)
693
+ 663 train 7.910801 (lr=1.0209e-04) (hash(x)=27537063)
694
+ 664 train 7.580163 (lr=1.0170e-04) (hash(x)=26125988)
695
+ 665 train 7.677661 (lr=1.0130e-04) (hash(x)=25446686)
696
+ 666 train 7.719374 (lr=1.0091e-04) (hash(x)=19819857)
697
+ 667 train 7.348328 (lr=1.0051e-04) (hash(x)=19492832)
698
+ 668 train 7.527069 (lr=1.0012e-04) (hash(x)=23741214)
699
+ 669 train 7.500827 (lr=9.9726e-05) (hash(x)=25484958)
700
+ 670 train 7.475147 (lr=9.9333e-05) (hash(x)=25525370)
701
+ 671 train 7.640712 (lr=9.8940e-05) (hash(x)=25951629)
702
+ 672 train 7.919829 (lr=9.8547e-05) (hash(x)=28557663)
703
+ 673 train 7.322699 (lr=9.8154e-05) (hash(x)=23462798)
704
+ 674 train 7.326319 (lr=9.7762e-05) (hash(x)=27320280)
705
+ 675 train 7.653201 (lr=9.7370e-05) (hash(x)=22408682)
706
+ 676 train 7.401831 (lr=9.6978e-05) (hash(x)=22230799)
707
+ 677 train 7.825242 (lr=9.6586e-05) (hash(x)=31317970)
708
+ 678 train 7.380300 (lr=9.6195e-05) (hash(x)=24020983)
709
+ 679 train 7.449243 (lr=9.5804e-05) (hash(x)=27633457)
710
+ 680 train 7.351268 (lr=9.5413e-05) (hash(x)=23766987)
711
+ 681 train 7.520663 (lr=9.5022e-05) (hash(x)=23729283)
712
+ 682 train 7.687955 (lr=9.4632e-05) (hash(x)=25975834)
713
+ 683 train 7.570558 (lr=9.4242e-05) (hash(x)=27635080)
714
+ 684 train 7.528741 (lr=9.3852e-05) (hash(x)=27181156)
715
+ 685 train 7.584823 (lr=9.3463e-05) (hash(x)=25918780)
716
+ 686 train 7.554749 (lr=9.3073e-05) (hash(x)=22370417)
717
+ 687 train 7.388176 (lr=9.2685e-05) (hash(x)=22287596)
718
+ 688 train 7.405398 (lr=9.2296e-05) (hash(x)=21584419)
719
+ 689 train 7.436688 (lr=9.1908e-05) (hash(x)=22694623)
720
+ 690 train 7.583045 (lr=9.1520e-05) (hash(x)=30820846)
721
+ 691 train 8.120363 (lr=9.1133e-05) (hash(x)=35895440)
722
+ 692 train 7.623921 (lr=9.0746e-05) (hash(x)=27058657)
723
+ 693 train 7.505916 (lr=9.0359e-05) (hash(x)=24298886)
724
+ 694 train 7.541343 (lr=8.9973e-05) (hash(x)=27303061)
725
+ 695 train 7.580599 (lr=8.9587e-05) (hash(x)=25888847)
726
+ 696 train 7.453152 (lr=8.9202e-05) (hash(x)=21267641)
727
+ 697 train 7.851780 (lr=8.8817e-05) (hash(x)=30860646)
728
+ 698 train 7.534546 (lr=8.8432e-05) (hash(x)=26881837)
729
+ 699 train 7.583902 (lr=8.8048e-05) (hash(x)=26518800)
730
+ 700 val loss 7.5055
731
+ 700 val perplexity 1817.9297
732
+ 700 train 7.569979 (lr=8.7664e-05) (hash(x)=25863277)
733
+ 701 train 7.374316 (lr=8.7281e-05) (hash(x)=25385523)
734
+ 702 train 7.434670 (lr=8.6898e-05) (hash(x)=23362519)
735
+ 703 train 7.904713 (lr=8.6515e-05) (hash(x)=37519283)
736
+ 704 train 7.932143 (lr=8.6133e-05) (hash(x)=23392584)
737
+ 705 train 8.028492 (lr=8.5751e-05) (hash(x)=20564397)
738
+ 706 train 7.641935 (lr=8.5370e-05) (hash(x)=27065499)
739
+ 707 train 7.535081 (lr=8.4990e-05) (hash(x)=25141945)
740
+ 708 train 7.691893 (lr=8.4610e-05) (hash(x)=25466598)
741
+ 709 train 7.501493 (lr=8.4230e-05) (hash(x)=22717866)
742
+ 710 train 8.098660 (lr=8.3851e-05) (hash(x)=29648992)
743
+ 711 train 7.813345 (lr=8.3472e-05) (hash(x)=28527939)
744
+ 712 train 7.407803 (lr=8.3094e-05) (hash(x)=25134784)
745
+ 713 train 7.923172 (lr=8.2716e-05) (hash(x)=32757059)
746
+ 714 train 7.428549 (lr=8.2339e-05) (hash(x)=24145729)
747
+ 715 train 7.401217 (lr=8.1963e-05) (hash(x)=21942471)
748
+ 716 train 7.526911 (lr=8.1587e-05) (hash(x)=25325444)
749
+ 717 train 7.891680 (lr=8.1211e-05) (hash(x)=24247339)
750
+ 718 train 7.481150 (lr=8.0836e-05) (hash(x)=24142989)
751
+ 719 train 7.457084 (lr=8.0462e-05) (hash(x)=24955630)
752
+ 720 train 7.569934 (lr=8.0088e-05) (hash(x)=25491335)
753
+ 721 train 7.657807 (lr=7.9715e-05) (hash(x)=21368780)
754
+ 722 train 7.566232 (lr=7.9342e-05) (hash(x)=27146278)
755
+ 723 train 7.454957 (lr=7.8970e-05) (hash(x)=25541230)
756
+ 724 train 7.488031 (lr=7.8599e-05) (hash(x)=25375355)
757
+ 725 train 7.340462 (lr=7.8228e-05) (hash(x)=23098806)
758
+ 726 train 7.352211 (lr=7.7858e-05) (hash(x)=22061174)
759
+ 727 train 7.370058 (lr=7.7488e-05) (hash(x)=23501481)
760
+ 728 train 7.549216 (lr=7.7119e-05) (hash(x)=24656430)
761
+ 729 train 7.274225 (lr=7.6751e-05) (hash(x)=23802681)
762
+ 730 train 7.347686 (lr=7.6383e-05) (hash(x)=23496118)
763
+ 731 train 7.529576 (lr=7.6016e-05) (hash(x)=22754988)
764
+ 732 train 7.273293 (lr=7.5650e-05) (hash(x)=21461650)
765
+ 733 train 7.313805 (lr=7.5284e-05) (hash(x)=22183009)
766
+ 734 train 7.463912 (lr=7.4919e-05) (hash(x)=22300616)
767
+ 735 train 7.241005 (lr=7.4555e-05) (hash(x)=23475070)
768
+ 736 train 7.510870 (lr=7.4191e-05) (hash(x)=24508407)
769
+ 737 train 7.546553 (lr=7.3828e-05) (hash(x)=23359061)
770
+ 738 train 7.450968 (lr=7.3466e-05) (hash(x)=22588667)
771
+ 739 train 7.520521 (lr=7.3105e-05) (hash(x)=23609959)
772
+ 740 train 7.510914 (lr=7.2744e-05) (hash(x)=25394852)
773
+ 741 train 7.326530 (lr=7.2384e-05) (hash(x)=22408449)
774
+ 742 train 7.494502 (lr=7.2024e-05) (hash(x)=26947398)
775
+ 743 train 7.221612 (lr=7.1666e-05) (hash(x)=23441847)
776
+ 744 train 7.346112 (lr=7.1308e-05) (hash(x)=23101350)
777
+ 745 train 7.342832 (lr=7.0950e-05) (hash(x)=23676365)
778
+ 746 train 7.540934 (lr=7.0594e-05) (hash(x)=26488811)
779
+ 747 train 7.597052 (lr=7.0238e-05) (hash(x)=23096292)
780
+ 748 train 7.621435 (lr=6.9884e-05) (hash(x)=24512380)
781
+ 749 train 7.503527 (lr=6.9529e-05) (hash(x)=27706294)
782
+ 750 val loss 7.4691
783
+ 750 val perplexity 1753.0149
784
+ 750 train 7.429562 (lr=6.9176e-05) (hash(x)=21892472)
785
+ 751 train 7.394504 (lr=6.8824e-05) (hash(x)=26171093)
786
+ 752 train 7.431930 (lr=6.8472e-05) (hash(x)=23191101)
787
+ 753 train 7.835644 (lr=6.8121e-05) (hash(x)=34946941)
788
+ 754 train 7.627355 (lr=6.7771e-05) (hash(x)=25569942)
789
+ 755 train 7.220369 (lr=6.7422e-05) (hash(x)=20060838)
790
+ 756 train 7.341573 (lr=6.7073e-05) (hash(x)=24581020)
791
+ 757 train 7.301620 (lr=6.6725e-05) (hash(x)=21591702)
792
+ 758 train 7.335374 (lr=6.6379e-05) (hash(x)=21870012)
793
+ 759 train 7.426104 (lr=6.6033e-05) (hash(x)=25451055)
794
+ 760 train 7.209653 (lr=6.5688e-05) (hash(x)=18989695)
795
+ 761 train 6.853334 (lr=6.5343e-05) (hash(x)=18691455)
796
+ 762 train 7.231118 (lr=6.5000e-05) (hash(x)=22406503)
797
+ 763 train 7.473215 (lr=6.4657e-05) (hash(x)=24934792)
798
+ 764 train 7.359167 (lr=6.4316e-05) (hash(x)=24249677)
799
+ 765 train 7.217328 (lr=6.3975e-05) (hash(x)=23864361)
800
+ 766 train 7.236013 (lr=6.3635e-05) (hash(x)=21549554)
801
+ 767 train 7.436182 (lr=6.3296e-05) (hash(x)=25367597)
802
+ 768 train 7.607349 (lr=6.2958e-05) (hash(x)=24409290)
803
+ 769 train 7.377076 (lr=6.2621e-05) (hash(x)=27016702)
804
+ 770 train 7.563516 (lr=6.2285e-05) (hash(x)=25520453)
805
+ 771 train 7.400713 (lr=6.1950e-05) (hash(x)=25022148)
806
+ 772 train 7.527383 (lr=6.1615e-05) (hash(x)=25010319)
807
+ 773 train 7.418715 (lr=6.1282e-05) (hash(x)=24459012)
808
+ 774 train 7.473160 (lr=6.0949e-05) (hash(x)=26528670)
809
+ 775 train 7.459906 (lr=6.0618e-05) (hash(x)=25431365)
810
+ 776 train 7.337970 (lr=6.0287e-05) (hash(x)=24229522)
811
+ 777 train 7.513100 (lr=5.9958e-05) (hash(x)=25958785)
812
+ 778 train 7.404982 (lr=5.9629e-05) (hash(x)=22882640)
813
+ 779 train 7.803152 (lr=5.9301e-05) (hash(x)=29307788)
814
+ 780 train 7.313700 (lr=5.8974e-05) (hash(x)=23167825)
815
+ 781 train 7.382802 (lr=5.8649e-05) (hash(x)=24965435)
816
+ 782 train 7.241666 (lr=5.8324e-05) (hash(x)=23772949)
817
+ 783 train 7.502848 (lr=5.8000e-05) (hash(x)=25651929)
818
+ 784 train 7.385237 (lr=5.7678e-05) (hash(x)=23146724)
819
+ 785 train 7.434648 (lr=5.7356e-05) (hash(x)=28451031)
820
+ 786 train 7.484230 (lr=5.7035e-05) (hash(x)=23796771)
821
+ 787 train 7.510779 (lr=5.6716e-05) (hash(x)=26312387)
822
+ 788 train 7.368457 (lr=5.6397e-05) (hash(x)=21257377)
823
+ 789 train 7.531878 (lr=5.6079e-05) (hash(x)=25326650)
824
+ 790 train 7.412582 (lr=5.5763e-05) (hash(x)=24605603)
825
+ 791 train 7.341255 (lr=5.5447e-05) (hash(x)=23804913)
826
+ 792 train 7.557047 (lr=5.5133e-05) (hash(x)=24322926)
827
+ 793 train 7.534083 (lr=5.4820e-05) (hash(x)=25537529)
828
+ 794 train 7.248249 (lr=5.4507e-05) (hash(x)=25707197)
829
+ 795 train 7.461568 (lr=5.4196e-05) (hash(x)=26115519)
830
+ 796 train 7.380995 (lr=5.3886e-05) (hash(x)=24850654)
831
+ 797 train 7.330876 (lr=5.3577e-05) (hash(x)=25494804)
832
+ 798 train 7.355565 (lr=5.3269e-05) (hash(x)=21972022)
833
+ 799 train 7.690948 (lr=5.2962e-05) (hash(x)=23210747)
834
+ 800 val loss 7.4382
835
+ 800 val perplexity 1699.6434
836
+ 800 train 7.510743 (lr=5.2656e-05) (hash(x)=17064773)
837
+ 801 train 7.478874 (lr=5.2352e-05) (hash(x)=15717069)
838
+ 802 train 7.563032 (lr=5.2048e-05) (hash(x)=16189636)
839
+ 803 train 7.331135 (lr=5.1746e-05) (hash(x)=18463712)
840
+ 804 train 7.492607 (lr=5.1444e-05) (hash(x)=19970072)
841
+ 805 train 7.540052 (lr=5.1144e-05) (hash(x)=20705573)
842
+ 806 train 7.332351 (lr=5.0845e-05) (hash(x)=22913147)
843
+ 807 train 7.424669 (lr=5.0547e-05) (hash(x)=23245699)
844
+ 808 train 7.448200 (lr=5.0251e-05) (hash(x)=27121904)
845
+ 809 train 7.354832 (lr=4.9955e-05) (hash(x)=23444521)
846
+ 810 train 7.264400 (lr=4.9661e-05) (hash(x)=20917937)
847
+ 811 train 7.356593 (lr=4.9367e-05) (hash(x)=21770300)
848
+ 812 train 8.514841 (lr=4.9075e-05) (hash(x)=24642519)
849
+ 813 train 7.587816 (lr=4.8784e-05) (hash(x)=23334569)
850
+ 814 train 7.160762 (lr=4.8495e-05) (hash(x)=21796200)
851
+ 815 train 7.215055 (lr=4.8206e-05) (hash(x)=25387532)
852
+ 816 train 7.144467 (lr=4.7919e-05) (hash(x)=22203733)
853
+ 817 train 7.757704 (lr=4.7633e-05) (hash(x)=29466743)
854
+ 818 train 7.424380 (lr=4.7348e-05) (hash(x)=21756102)
855
+ 819 train 7.475869 (lr=4.7064e-05) (hash(x)=27718703)
856
+ 820 train 7.412989 (lr=4.6782e-05) (hash(x)=22978681)
857
+ 821 train 7.458809 (lr=4.6501e-05) (hash(x)=17631973)
858
+ 822 train 7.538844 (lr=4.6221e-05) (hash(x)=27947891)
859
+ 823 train 7.430303 (lr=4.5942e-05) (hash(x)=27661052)
860
+ 824 train 7.211120 (lr=4.5664e-05) (hash(x)=23061452)
861
+ 825 train 7.332330 (lr=4.5388e-05) (hash(x)=22244096)
862
+ 826 train 7.356231 (lr=4.5113e-05) (hash(x)=25389713)
863
+ 827 train 7.458523 (lr=4.4839e-05) (hash(x)=26538928)
864
+ 828 train 7.545984 (lr=4.4567e-05) (hash(x)=25685569)
865
+ 829 train 7.241985 (lr=4.4295e-05) (hash(x)=26256802)
866
+ 830 train 7.578298 (lr=4.4025e-05) (hash(x)=24412323)
867
+ 831 train 7.371192 (lr=4.3757e-05) (hash(x)=21772654)
868
+ 832 train 7.324841 (lr=4.3489e-05) (hash(x)=25456825)
869
+ 833 train 7.715641 (lr=4.3223e-05) (hash(x)=25026277)
870
+ 834 train 7.573750 (lr=4.2958e-05) (hash(x)=26855439)
871
+ 835 train 7.583344 (lr=4.2695e-05) (hash(x)=26263254)
872
+ 836 train 7.587561 (lr=4.2432e-05) (hash(x)=21917396)
873
+ 837 train 7.392462 (lr=4.2171e-05) (hash(x)=24726314)
874
+ 838 train 7.278290 (lr=4.1912e-05) (hash(x)=24375100)
875
+ 839 train 7.349993 (lr=4.1654e-05) (hash(x)=23552356)
876
+ 840 train 7.440544 (lr=4.1397e-05) (hash(x)=25343112)
877
+ 841 train 7.385789 (lr=4.1141e-05) (hash(x)=26699819)
878
+ 842 train 7.736491 (lr=4.0887e-05) (hash(x)=25218858)
879
+ 843 train 7.931557 (lr=4.0634e-05) (hash(x)=22707141)
880
+ 844 train 7.724117 (lr=4.0382e-05) (hash(x)=29168269)
881
+ 845 train 7.786210 (lr=4.0132e-05) (hash(x)=25676149)
882
+ 846 train 7.354553 (lr=3.9883e-05) (hash(x)=24324401)
883
+ 847 train 7.722795 (lr=3.9635e-05) (hash(x)=34234481)
884
+ 848 train 7.407337 (lr=3.9389e-05) (hash(x)=24511386)
885
+ 849 train 7.405665 (lr=3.9144e-05) (hash(x)=25401624)
886
+ 850 val loss 7.4425
887
+ 850 val perplexity 1706.9906
888
+ 850 train 7.403595 (lr=3.8901e-05) (hash(x)=22340685)
889
+ 851 train 7.743905 (lr=3.8659e-05) (hash(x)=35008670)
890
+ 852 train 7.746662 (lr=3.8418e-05) (hash(x)=29662286)
891
+ 853 train 7.642934 (lr=3.8178e-05) (hash(x)=25682280)
892
+ 854 train 7.588220 (lr=3.7941e-05) (hash(x)=27686498)
893
+ 855 train 7.408626 (lr=3.7704e-05) (hash(x)=25891543)
894
+ 856 train 7.339035 (lr=3.7469e-05) (hash(x)=22538881)
895
+ 857 train 7.405196 (lr=3.7235e-05) (hash(x)=25962407)
896
+ 858 train 7.347361 (lr=3.7003e-05) (hash(x)=24819073)
897
+ 859 train 7.218128 (lr=3.6772e-05) (hash(x)=21971224)
898
+ 860 train 7.173095 (lr=3.6542e-05) (hash(x)=20951317)
899
+ 861 train 7.661697 (lr=3.6314e-05) (hash(x)=30838142)
900
+ 862 train 7.478192 (lr=3.6088e-05) (hash(x)=23102419)
901
+ 863 train 7.469544 (lr=3.5862e-05) (hash(x)=27312186)
902
+ 864 train 7.949055 (lr=3.5639e-05) (hash(x)=30920030)
903
+ 865 train 7.303967 (lr=3.5416e-05) (hash(x)=23657081)
904
+ 866 train 7.826602 (lr=3.5195e-05) (hash(x)=35351214)
905
+ 867 train 7.244772 (lr=3.4976e-05) (hash(x)=22163749)
906
+ 868 train 7.126449 (lr=3.4758e-05) (hash(x)=24629160)
907
+ 869 train 7.288036 (lr=3.4541e-05) (hash(x)=24829818)
908
+ 870 train 7.632834 (lr=3.4326e-05) (hash(x)=25424550)
909
+ 871 train 7.366570 (lr=3.4113e-05) (hash(x)=25942888)
910
+ 872 train 7.522963 (lr=3.3900e-05) (hash(x)=26418501)
911
+ 873 train 7.395418 (lr=3.3690e-05) (hash(x)=23785671)
912
+ 874 train 7.550786 (lr=3.3480e-05) (hash(x)=25097859)
913
+ 875 train 7.343193 (lr=3.3273e-05) (hash(x)=25493916)
914
+ 876 train 7.545170 (lr=3.3066e-05) (hash(x)=19819210)
915
+ 877 train 7.669271 (lr=3.2862e-05) (hash(x)=28618702)
916
+ 878 train 7.849481 (lr=3.2658e-05) (hash(x)=29707907)
917
+ 879 train 7.510124 (lr=3.2457e-05) (hash(x)=24289951)
918
+ 880 train 7.354381 (lr=3.2256e-05) (hash(x)=25147026)
919
+ 881 train 7.475085 (lr=3.2058e-05) (hash(x)=26331486)
920
+ 882 train 7.436450 (lr=3.1860e-05) (hash(x)=25885126)
921
+ 883 train 7.393227 (lr=3.1665e-05) (hash(x)=23648671)
922
+ 884 train 7.195853 (lr=3.1471e-05) (hash(x)=20252079)
923
+ 885 train 7.505919 (lr=3.1278e-05) (hash(x)=27292797)
924
+ 886 train 6.972956 (lr=3.1087e-05) (hash(x)=18317379)
925
+ 887 train 7.351745 (lr=3.0897e-05) (hash(x)=24992055)
926
+ 888 train 7.815773 (lr=3.0709e-05) (hash(x)=24734221)
927
+ 889 train 7.608184 (lr=3.0522e-05) (hash(x)=24398330)
928
+ 890 train 7.340979 (lr=3.0337e-05) (hash(x)=23573984)
929
+ 891 train 7.473004 (lr=3.0154e-05) (hash(x)=26005549)
930
+ 892 train 7.244836 (lr=2.9972e-05) (hash(x)=22583878)
931
+ 893 train 7.564567 (lr=2.9791e-05) (hash(x)=30259737)
932
+ 894 train 7.478389 (lr=2.9613e-05) (hash(x)=25196470)
933
+ 895 train 7.295092 (lr=2.9435e-05) (hash(x)=21046554)
934
+ 896 train 7.492802 (lr=2.9260e-05) (hash(x)=25454371)
935
+ 897 train 7.276515 (lr=2.9085e-05) (hash(x)=18773877)
936
+ 898 train 7.293960 (lr=2.8913e-05) (hash(x)=24645168)
937
+ 899 train 7.540881 (lr=2.8742e-05) (hash(x)=26321539)
938
+ 900 val loss 7.4073
939
+ 900 val perplexity 1647.9661
940
+ 900 train 7.455224 (lr=2.8572e-05) (hash(x)=25722713)
941
+ 901 train 7.324742 (lr=2.8404e-05) (hash(x)=22850804)
942
+ 902 train 7.465915 (lr=2.8238e-05) (hash(x)=22503524)
943
+ 903 train 7.462577 (lr=2.8073e-05) (hash(x)=27807872)
944
+ 904 train 7.492571 (lr=2.7910e-05) (hash(x)=25845590)
945
+ 905 train 7.349280 (lr=2.7749e-05) (hash(x)=25545137)
946
+ 906 train 7.535013 (lr=2.7589e-05) (hash(x)=25471231)
947
+ 907 train 7.262427 (lr=2.7430e-05) (hash(x)=21555364)
948
+ 908 train 7.434108 (lr=2.7274e-05) (hash(x)=24474988)
949
+ 909 train 7.453172 (lr=2.7119e-05) (hash(x)=27243798)
950
+ 910 train 7.336480 (lr=2.6965e-05) (hash(x)=24419085)
951
+ 911 train 7.343315 (lr=2.6813e-05) (hash(x)=26325324)
952
+ 912 train 7.298230 (lr=2.6663e-05) (hash(x)=24822111)
953
+ 913 train 7.300466 (lr=2.6514e-05) (hash(x)=26228728)
954
+ 914 train 7.522482 (lr=2.6367e-05) (hash(x)=24707042)
955
+ 915 train 7.594767 (lr=2.6221e-05) (hash(x)=24956294)
956
+ 916 train 7.157427 (lr=2.6077e-05) (hash(x)=23424737)
957
+ 917 train 7.575006 (lr=2.5935e-05) (hash(x)=25592289)
958
+ 918 train 7.352469 (lr=2.5795e-05) (hash(x)=26826109)
959
+ 919 train 7.393952 (lr=2.5656e-05) (hash(x)=24429389)
960
+ 920 train 7.336396 (lr=2.5518e-05) (hash(x)=24747421)
961
+ 921 train 7.434125 (lr=2.5383e-05) (hash(x)=28733708)
962
+ 922 train 7.524363 (lr=2.5249e-05) (hash(x)=29704803)
963
+ 923 train 7.702175 (lr=2.5116e-05) (hash(x)=27929846)
964
+ 924 train 7.399565 (lr=2.4985e-05) (hash(x)=18961171)
965
+ 925 train 7.390099 (lr=2.4856e-05) (hash(x)=22417476)
966
+ 926 train 7.445345 (lr=2.4729e-05) (hash(x)=26364120)
967
+ 927 train 7.539669 (lr=2.4603e-05) (hash(x)=25282897)
968
+ 928 train 7.282486 (lr=2.4479e-05) (hash(x)=22082503)
969
+ 929 train 7.265842 (lr=2.4356e-05) (hash(x)=22278062)
970
+ 930 train 7.371174 (lr=2.4235e-05) (hash(x)=24271468)
971
+ 931 train 7.244277 (lr=2.4116e-05) (hash(x)=26802508)
972
+ 932 train 8.012707 (lr=2.3998e-05) (hash(x)=33468379)
973
+ 933 train 7.101473 (lr=2.3883e-05) (hash(x)=22094832)
974
+ 934 train 7.345801 (lr=2.3768e-05) (hash(x)=26759850)
975
+ 935 train 7.553965 (lr=2.3656e-05) (hash(x)=28390416)
976
+ 936 train 7.693401 (lr=2.3545e-05) (hash(x)=22727956)
977
+ 937 train 7.626191 (lr=2.3436e-05) (hash(x)=26379331)
978
+ 938 train 7.787500 (lr=2.3328e-05) (hash(x)=25048158)
979
+ 939 train 7.651772 (lr=2.3222e-05) (hash(x)=26292451)
980
+ 940 train 7.494242 (lr=2.3118e-05) (hash(x)=23336031)
981
+ 941 train 7.659499 (lr=2.3016e-05) (hash(x)=26301866)
982
+ 942 train 7.732497 (lr=2.2915e-05) (hash(x)=26687040)
983
+ 943 train 7.280413 (lr=2.2816e-05) (hash(x)=22915344)
984
+ 944 train 7.332717 (lr=2.2718e-05) (hash(x)=19831900)
985
+ 945 train 7.446020 (lr=2.2623e-05) (hash(x)=25448318)
986
+ 946 train 7.384642 (lr=2.2528e-05) (hash(x)=24993116)
987
+ 947 train 7.523602 (lr=2.2436e-05) (hash(x)=24744383)
988
+ 948 train 7.311930 (lr=2.2345e-05) (hash(x)=23731849)
989
+ 949 train 7.270403 (lr=2.2256e-05) (hash(x)=22835201)
990
+ 950 val loss 7.4007
991
+ 950 val perplexity 1637.0513
992
+ 950 train 7.242966 (lr=2.2169e-05) (hash(x)=24652361)
993
+ 951 train 7.521279 (lr=2.2084e-05) (hash(x)=26284678)
994
+ 952 train 7.396550 (lr=2.2000e-05) (hash(x)=27522881)
995
+ 953 train 7.392224 (lr=2.1918e-05) (hash(x)=24745936)
996
+ 954 train 7.211158 (lr=2.1837e-05) (hash(x)=21973969)
997
+ 955 train 7.213524 (lr=2.1758e-05) (hash(x)=21654269)
998
+ 956 train 7.204653 (lr=2.1681e-05) (hash(x)=21708739)
999
+ 957 train 7.634518 (lr=2.1606e-05) (hash(x)=32387999)
1000
+ 958 train 8.142120 (lr=2.1532e-05) (hash(x)=32662682)
1001
+ 959 train 8.143919 (lr=2.1461e-05) (hash(x)=29466268)
1002
+ 960 train 8.067085 (lr=2.1390e-05) (hash(x)=28372973)
1003
+ 961 train 8.103285 (lr=2.1322e-05) (hash(x)=30736727)
1004
+ 962 train 8.054703 (lr=2.1255e-05) (hash(x)=29935562)
1005
+ 963 train 8.062071 (lr=2.1190e-05) (hash(x)=32616762)
1006
+ 964 train 8.016862 (lr=2.1127e-05) (hash(x)=32861403)
1007
+ 965 train 8.018845 (lr=2.1065e-05) (hash(x)=29358578)
1008
+ 966 train 8.067577 (lr=2.1005e-05) (hash(x)=31218966)
1009
+ 967 train 8.150707 (lr=2.0947e-05) (hash(x)=32201138)
1010
+ 968 train 8.020109 (lr=2.0891e-05) (hash(x)=30545133)
1011
+ 969 train 8.218145 (lr=2.0836e-05) (hash(x)=30041621)
1012
+ 970 train 8.039783 (lr=2.0783e-05) (hash(x)=30371023)
1013
+ 971 train 8.181568 (lr=2.0732e-05) (hash(x)=33683468)
1014
+ 972 train 8.084312 (lr=2.0682e-05) (hash(x)=29809830)
1015
+ 973 train 7.541002 (lr=2.0634e-05) (hash(x)=26837755)
1016
+ 974 train 7.524750 (lr=2.0588e-05) (hash(x)=26084089)
1017
+ 975 train 7.653708 (lr=2.0544e-05) (hash(x)=27498125)
1018
+ 976 train 7.891948 (lr=2.0501e-05) (hash(x)=34478395)
1019
+ 977 train 7.532537 (lr=2.0460e-05) (hash(x)=22372507)
1020
+ 978 train 7.460429 (lr=2.0421e-05) (hash(x)=25638879)
1021
+ 979 train 7.298199 (lr=2.0384e-05) (hash(x)=25164318)
1022
+ 980 train 7.447876 (lr=2.0348e-05) (hash(x)=27271900)
1023
+ 981 train 7.375486 (lr=2.0314e-05) (hash(x)=26562947)
1024
+ 982 train 7.401752 (lr=2.0282e-05) (hash(x)=24917192)
1025
+ 983 train 7.302921 (lr=2.0252e-05) (hash(x)=25690221)
1026
+ 984 train 7.645075 (lr=2.0223e-05) (hash(x)=24807841)
1027
+ 985 train 7.300786 (lr=2.0196e-05) (hash(x)=21947305)
1028
+ 986 train 7.302711 (lr=2.0171e-05) (hash(x)=22195089)
1029
+ 987 train 7.434680 (lr=2.0147e-05) (hash(x)=29031550)
1030
+ 988 train 7.548289 (lr=2.0125e-05) (hash(x)=29395823)
1031
+ 989 train 7.490853 (lr=2.0105e-05) (hash(x)=26551589)
1032
+ 990 train 7.448102 (lr=2.0087e-05) (hash(x)=23689487)
1033
+ 991 train 7.502644 (lr=2.0071e-05) (hash(x)=24094093)
1034
+ 992 train 7.209767 (lr=2.0056e-05) (hash(x)=24106453)
1035
+ 993 train 7.307530 (lr=2.0043e-05) (hash(x)=23639076)
1036
+ 994 train 7.442201 (lr=2.0031e-05) (hash(x)=25107314)
1037
+ 995 train 10.471497 (lr=2.0022e-05) (hash(x)=61455960)
1038
+ 996 train 11.432472 (lr=2.0014e-05) (hash(x)=63899424)
1039
+ 997 train 11.386467 (lr=2.0008e-05) (hash(x)=65636572)
1040
+ 998 train 11.762794 (lr=2.0003e-05) (hash(x)=71578682)
1041
+ 999 val loss 7.4048
1042
+ 999 val perplexity 1643.8732
1043
+ 999 train 10.736455 (lr=2.0001e-05) (hash(x)=61104710)
lr2e-4_total_batch_size10240_seq_len512/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d9c8f6f88bbd5e48770cb953fa6a3fc058acd653b1b8eb9c8c5d3e9ae9f22e9
3
+ size 48156226
lr2e-4_total_batch_size10240_seq_len512/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e823f2f4b0af83d612fb7f97a0d5bf57f625d61e7ee3629476d3ab65d9964a98
3
+ size 71157574