andrew-healey commited on
Commit
d196106
·
verified ·
1 Parent(s): 179ff27

Upload folder using huggingface_hub

Browse files
lr2e-4_total_batch_size5120_seq_len256/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr2e-4_total_batch_size5120_seq_len256", "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": 256, "batch_size": null, "total_batch_size": 5120, "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_size5120_seq_len256/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr2e-4_total_batch_size5120_seq_len256/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7488
3
+ 0 val perplexity 126607.6172
4
+ 0 train 11.745071 (lr=6.9930e-07) (hash(x)=12372582)
5
+ 1 train 11.742463 (lr=1.3986e-06) (hash(x)=12683395)
6
+ 2 train 11.779614 (lr=2.0979e-06) (hash(x)=14107219)
7
+ 3 train 11.734921 (lr=2.7972e-06) (hash(x)=12136211)
8
+ 4 train 11.733570 (lr=3.4965e-06) (hash(x)=12407129)
9
+ 5 train 11.707633 (lr=4.1958e-06) (hash(x)=13068552)
10
+ 6 train 11.722017 (lr=4.8951e-06) (hash(x)=17450022)
11
+ 7 train 11.713314 (lr=5.5944e-06) (hash(x)=9915468)
12
+ 8 train 11.712975 (lr=6.2937e-06) (hash(x)=13548512)
13
+ 9 train 11.727014 (lr=6.9930e-06) (hash(x)=12164183)
14
+ 10 train 11.709367 (lr=7.6923e-06) (hash(x)=13803353)
15
+ 11 train 11.679868 (lr=8.3916e-06) (hash(x)=13629277)
16
+ 12 train 11.645482 (lr=9.0909e-06) (hash(x)=12917812)
17
+ 13 train 11.637653 (lr=9.7902e-06) (hash(x)=13413021)
18
+ 14 train 11.599601 (lr=1.0490e-05) (hash(x)=12655054)
19
+ 15 train 11.598496 (lr=1.1189e-05) (hash(x)=12684830)
20
+ 16 train 11.546203 (lr=1.1888e-05) (hash(x)=13439617)
21
+ 17 train 11.568825 (lr=1.2587e-05) (hash(x)=13262550)
22
+ 18 train 11.554014 (lr=1.3287e-05) (hash(x)=11975280)
23
+ 19 train 11.531845 (lr=1.3986e-05) (hash(x)=12003126)
24
+ 20 train 11.490280 (lr=1.4685e-05) (hash(x)=12313617)
25
+ 21 train 11.470226 (lr=1.5385e-05) (hash(x)=14144015)
26
+ 22 train 11.444975 (lr=1.6084e-05) (hash(x)=11178611)
27
+ 23 train 11.380606 (lr=1.6783e-05) (hash(x)=12014715)
28
+ 24 train 11.390724 (lr=1.7483e-05) (hash(x)=15240676)
29
+ 25 train 11.285635 (lr=1.8182e-05) (hash(x)=14912984)
30
+ 26 train 11.271848 (lr=1.8881e-05) (hash(x)=11487406)
31
+ 27 train 11.197708 (lr=1.9580e-05) (hash(x)=9492167)
32
+ 28 train 11.181851 (lr=2.0280e-05) (hash(x)=10765326)
33
+ 29 train 11.181164 (lr=2.0979e-05) (hash(x)=15241928)
34
+ 30 train 11.082594 (lr=2.1678e-05) (hash(x)=12215703)
35
+ 31 train 10.957846 (lr=2.2378e-05) (hash(x)=12595649)
36
+ 32 train 11.008558 (lr=2.3077e-05) (hash(x)=12726494)
37
+ 33 train 10.906214 (lr=2.3776e-05) (hash(x)=15162520)
38
+ 34 train 10.923212 (lr=2.4476e-05) (hash(x)=10640461)
39
+ 35 train 10.911027 (lr=2.5175e-05) (hash(x)=11980631)
40
+ 36 train 10.890951 (lr=2.5874e-05) (hash(x)=10568919)
41
+ 37 train 10.876982 (lr=2.6573e-05) (hash(x)=13172222)
42
+ 38 train 10.761310 (lr=2.7273e-05) (hash(x)=11575367)
43
+ 39 train 10.727571 (lr=2.7972e-05) (hash(x)=11101031)
44
+ 40 train 10.721853 (lr=2.8671e-05) (hash(x)=10304235)
45
+ 41 train 10.662294 (lr=2.9371e-05) (hash(x)=13275746)
46
+ 42 train 10.597552 (lr=3.0070e-05) (hash(x)=14309786)
47
+ 43 train 10.536396 (lr=3.0769e-05) (hash(x)=14316349)
48
+ 44 train 10.530739 (lr=3.1469e-05) (hash(x)=13999350)
49
+ 45 train 10.574337 (lr=3.2168e-05) (hash(x)=14143687)
50
+ 46 train 10.538022 (lr=3.2867e-05) (hash(x)=11365466)
51
+ 47 train 10.497326 (lr=3.3566e-05) (hash(x)=12036401)
52
+ 48 train 10.446421 (lr=3.4266e-05) (hash(x)=11730055)
53
+ 49 train 10.426212 (lr=3.4965e-05) (hash(x)=11143776)
54
+ 50 val loss 10.3841
55
+ 50 val perplexity 32341.5957
56
+ 50 train 10.391515 (lr=3.5664e-05) (hash(x)=12724270)
57
+ 51 train 10.327371 (lr=3.6364e-05) (hash(x)=12337125)
58
+ 52 train 10.329067 (lr=3.7063e-05) (hash(x)=12381632)
59
+ 53 train 10.329236 (lr=3.7762e-05) (hash(x)=10700109)
60
+ 54 train 10.262674 (lr=3.8462e-05) (hash(x)=13197786)
61
+ 55 train 10.232517 (lr=3.9161e-05) (hash(x)=11823810)
62
+ 56 train 10.338673 (lr=3.9860e-05) (hash(x)=12172918)
63
+ 57 train 10.257351 (lr=4.0559e-05) (hash(x)=12070591)
64
+ 58 train 10.236158 (lr=4.1259e-05) (hash(x)=14199166)
65
+ 59 train 10.339070 (lr=4.1958e-05) (hash(x)=13033959)
66
+ 60 train 10.248023 (lr=4.2657e-05) (hash(x)=12802857)
67
+ 61 train 10.193905 (lr=4.3357e-05) (hash(x)=10786770)
68
+ 62 train 10.223665 (lr=4.4056e-05) (hash(x)=11809534)
69
+ 63 train 10.015480 (lr=4.4755e-05) (hash(x)=13549925)
70
+ 64 train 9.997356 (lr=4.5455e-05) (hash(x)=10914323)
71
+ 65 train 10.030985 (lr=4.6154e-05) (hash(x)=14101838)
72
+ 66 train 10.163025 (lr=4.6853e-05) (hash(x)=11235125)
73
+ 67 train 10.157752 (lr=4.7552e-05) (hash(x)=14220701)
74
+ 68 train 9.963398 (lr=4.8252e-05) (hash(x)=12126282)
75
+ 69 train 10.061707 (lr=4.8951e-05) (hash(x)=14571180)
76
+ 70 train 10.011343 (lr=4.9650e-05) (hash(x)=14173731)
77
+ 71 train 10.090284 (lr=5.0350e-05) (hash(x)=13252262)
78
+ 72 train 9.920448 (lr=5.1049e-05) (hash(x)=14118291)
79
+ 73 train 9.922458 (lr=5.1748e-05) (hash(x)=13676730)
80
+ 74 train 9.999580 (lr=5.2448e-05) (hash(x)=13927111)
81
+ 75 train 9.964857 (lr=5.3147e-05) (hash(x)=14638145)
82
+ 76 train 9.900003 (lr=5.3846e-05) (hash(x)=14977819)
83
+ 77 train 9.897355 (lr=5.4545e-05) (hash(x)=14385711)
84
+ 78 train 9.998713 (lr=5.5245e-05) (hash(x)=12219671)
85
+ 79 train 9.980270 (lr=5.5944e-05) (hash(x)=13749054)
86
+ 80 train 9.992213 (lr=5.6643e-05) (hash(x)=11709001)
87
+ 81 train 9.925891 (lr=5.7343e-05) (hash(x)=12469703)
88
+ 82 train 9.888877 (lr=5.8042e-05) (hash(x)=11330478)
89
+ 83 train 9.870050 (lr=5.8741e-05) (hash(x)=12681760)
90
+ 84 train 9.923363 (lr=5.9441e-05) (hash(x)=15215674)
91
+ 85 train 9.870603 (lr=6.0140e-05) (hash(x)=12423697)
92
+ 86 train 9.787323 (lr=6.0839e-05) (hash(x)=11789526)
93
+ 87 train 9.804956 (lr=6.1538e-05) (hash(x)=14461706)
94
+ 88 train 9.836361 (lr=6.2238e-05) (hash(x)=10984941)
95
+ 89 train 9.799957 (lr=6.2937e-05) (hash(x)=13727016)
96
+ 90 train 9.680655 (lr=6.3636e-05) (hash(x)=12862054)
97
+ 91 train 9.648274 (lr=6.4336e-05) (hash(x)=14235218)
98
+ 92 train 9.735443 (lr=6.5035e-05) (hash(x)=13621196)
99
+ 93 train 9.716882 (lr=6.5734e-05) (hash(x)=12501568)
100
+ 94 train 9.874262 (lr=6.6434e-05) (hash(x)=12274822)
101
+ 95 train 9.685357 (lr=6.7133e-05) (hash(x)=11735995)
102
+ 96 train 9.836740 (lr=6.7832e-05) (hash(x)=13050719)
103
+ 97 train 9.859317 (lr=6.8531e-05) (hash(x)=11569945)
104
+ 98 train 9.625572 (lr=6.9231e-05) (hash(x)=11109133)
105
+ 99 train 9.645729 (lr=6.9930e-05) (hash(x)=11641112)
106
+ 100 val loss 9.6865
107
+ 100 val perplexity 16099.0078
108
+ 100 train 9.703830 (lr=7.0629e-05) (hash(x)=11507240)
109
+ 101 train 9.644009 (lr=7.1329e-05) (hash(x)=11536784)
110
+ 102 train 9.664434 (lr=7.2028e-05) (hash(x)=11514312)
111
+ 103 train 9.758986 (lr=7.2727e-05) (hash(x)=12231701)
112
+ 104 train 9.665122 (lr=7.3427e-05) (hash(x)=10619273)
113
+ 105 train 9.624384 (lr=7.4126e-05) (hash(x)=13521301)
114
+ 106 train 9.466059 (lr=7.4825e-05) (hash(x)=10295729)
115
+ 107 train 9.537803 (lr=7.5524e-05) (hash(x)=10175470)
116
+ 108 train 9.589054 (lr=7.6224e-05) (hash(x)=11948399)
117
+ 109 train 9.580799 (lr=7.6923e-05) (hash(x)=13903066)
118
+ 110 train 10.137646 (lr=7.7622e-05) (hash(x)=16265112)
119
+ 111 train 9.576118 (lr=7.8322e-05) (hash(x)=14117945)
120
+ 112 train 9.491982 (lr=7.9021e-05) (hash(x)=12712901)
121
+ 113 train 9.553596 (lr=7.9720e-05) (hash(x)=14851356)
122
+ 114 train 9.624134 (lr=8.0420e-05) (hash(x)=12337690)
123
+ 115 train 9.524712 (lr=8.1119e-05) (hash(x)=13116544)
124
+ 116 train 9.474995 (lr=8.1818e-05) (hash(x)=16794825)
125
+ 117 train 9.552347 (lr=8.2517e-05) (hash(x)=16366718)
126
+ 118 train 9.499178 (lr=8.3217e-05) (hash(x)=13520931)
127
+ 119 train 9.410403 (lr=8.3916e-05) (hash(x)=12678303)
128
+ 120 train 9.375563 (lr=8.4615e-05) (hash(x)=10020764)
129
+ 121 train 9.240090 (lr=8.5315e-05) (hash(x)=11144768)
130
+ 122 train 9.501566 (lr=8.6014e-05) (hash(x)=10378383)
131
+ 123 train 9.326124 (lr=8.6713e-05) (hash(x)=11927462)
132
+ 124 train 9.327136 (lr=8.7413e-05) (hash(x)=13130532)
133
+ 125 train 9.299536 (lr=8.8112e-05) (hash(x)=10418366)
134
+ 126 train 9.422672 (lr=8.8811e-05) (hash(x)=14920480)
135
+ 127 train 9.349115 (lr=8.9510e-05) (hash(x)=12758461)
136
+ 128 train 9.378063 (lr=9.0210e-05) (hash(x)=14632236)
137
+ 129 train 9.379567 (lr=9.0909e-05) (hash(x)=11680802)
138
+ 130 train 9.296677 (lr=9.1608e-05) (hash(x)=12682477)
139
+ 131 train 9.395699 (lr=9.2308e-05) (hash(x)=11946230)
140
+ 132 train 9.376804 (lr=9.3007e-05) (hash(x)=15160610)
141
+ 133 train 9.170813 (lr=9.3706e-05) (hash(x)=11704519)
142
+ 134 train 9.113958 (lr=9.4406e-05) (hash(x)=12746108)
143
+ 135 train 9.153707 (lr=9.5105e-05) (hash(x)=13701311)
144
+ 136 train 9.193964 (lr=9.5804e-05) (hash(x)=15682136)
145
+ 137 train 9.218878 (lr=9.6503e-05) (hash(x)=12751615)
146
+ 138 train 9.223177 (lr=9.7203e-05) (hash(x)=13420600)
147
+ 139 train 9.075116 (lr=9.7902e-05) (hash(x)=11198382)
148
+ 140 train 9.186128 (lr=9.8601e-05) (hash(x)=11556406)
149
+ 141 train 9.254230 (lr=9.9301e-05) (hash(x)=16075171)
150
+ 142 train 9.418408 (lr=1.0000e-04) (hash(x)=16911864)
151
+ 143 train 9.115257 (lr=1.0070e-04) (hash(x)=14346433)
152
+ 144 train 9.148060 (lr=1.0140e-04) (hash(x)=12105279)
153
+ 145 train 9.078602 (lr=1.0210e-04) (hash(x)=12888806)
154
+ 146 train 9.096539 (lr=1.0280e-04) (hash(x)=11676047)
155
+ 147 train 9.325201 (lr=1.0350e-04) (hash(x)=17855984)
156
+ 148 train 9.178595 (lr=1.0420e-04) (hash(x)=14922202)
157
+ 149 train 8.970037 (lr=1.0490e-04) (hash(x)=9616381)
158
+ 150 val loss 8.9505
159
+ 150 val perplexity 7711.4341
160
+ 150 train 8.991547 (lr=1.0559e-04) (hash(x)=11978542)
161
+ 151 train 10.083858 (lr=1.0629e-04) (hash(x)=18910052)
162
+ 152 train 9.488463 (lr=1.0699e-04) (hash(x)=13545339)
163
+ 153 train 8.861468 (lr=1.0769e-04) (hash(x)=12769596)
164
+ 154 train 8.874098 (lr=1.0839e-04) (hash(x)=12235730)
165
+ 155 train 9.159559 (lr=1.0909e-04) (hash(x)=11885403)
166
+ 156 train 8.904396 (lr=1.0979e-04) (hash(x)=12677961)
167
+ 157 train 8.923874 (lr=1.1049e-04) (hash(x)=12476514)
168
+ 158 train 8.972180 (lr=1.1119e-04) (hash(x)=13656193)
169
+ 159 train 8.920395 (lr=1.1189e-04) (hash(x)=12508350)
170
+ 160 train 8.724266 (lr=1.1259e-04) (hash(x)=11204172)
171
+ 161 train 8.763577 (lr=1.1329e-04) (hash(x)=10745637)
172
+ 162 train 8.736413 (lr=1.1399e-04) (hash(x)=11844793)
173
+ 163 train 8.977192 (lr=1.1469e-04) (hash(x)=10862804)
174
+ 164 train 8.969051 (lr=1.1538e-04) (hash(x)=12061659)
175
+ 165 train 8.800344 (lr=1.1608e-04) (hash(x)=15075084)
176
+ 166 train 8.745974 (lr=1.1678e-04) (hash(x)=12508408)
177
+ 167 train 8.685780 (lr=1.1748e-04) (hash(x)=14043988)
178
+ 168 train 8.401215 (lr=1.1818e-04) (hash(x)=10401493)
179
+ 169 train 8.541430 (lr=1.1888e-04) (hash(x)=12667924)
180
+ 170 train 8.471537 (lr=1.1958e-04) (hash(x)=12778664)
181
+ 171 train 8.598648 (lr=1.2028e-04) (hash(x)=11916406)
182
+ 172 train 8.579514 (lr=1.2098e-04) (hash(x)=11516862)
183
+ 173 train 8.490269 (lr=1.2168e-04) (hash(x)=13406195)
184
+ 174 train 8.641919 (lr=1.2238e-04) (hash(x)=12116183)
185
+ 175 train 8.440344 (lr=1.2308e-04) (hash(x)=11968227)
186
+ 176 train 8.646670 (lr=1.2378e-04) (hash(x)=12134811)
187
+ 177 train 8.454968 (lr=1.2448e-04) (hash(x)=11927490)
188
+ 178 train 8.456906 (lr=1.2517e-04) (hash(x)=12684747)
189
+ 179 train 8.510316 (lr=1.2587e-04) (hash(x)=13290191)
190
+ 180 train 8.491484 (lr=1.2657e-04) (hash(x)=11789931)
191
+ 181 train 8.262606 (lr=1.2727e-04) (hash(x)=11771253)
192
+ 182 train 8.375028 (lr=1.2797e-04) (hash(x)=10693812)
193
+ 183 train 8.483016 (lr=1.2867e-04) (hash(x)=15670776)
194
+ 184 train 8.467882 (lr=1.2937e-04) (hash(x)=13586931)
195
+ 185 train 8.222755 (lr=1.3007e-04) (hash(x)=11306133)
196
+ 186 train 8.378218 (lr=1.3077e-04) (hash(x)=12636637)
197
+ 187 train 8.600243 (lr=1.3147e-04) (hash(x)=14358384)
198
+ 188 train 8.283781 (lr=1.3217e-04) (hash(x)=11557997)
199
+ 189 train 8.161809 (lr=1.3287e-04) (hash(x)=10267428)
200
+ 190 train 8.404573 (lr=1.3357e-04) (hash(x)=11861629)
201
+ 191 train 8.236061 (lr=1.3427e-04) (hash(x)=11289341)
202
+ 192 train 8.250552 (lr=1.3497e-04) (hash(x)=11666650)
203
+ 193 train 8.227356 (lr=1.3566e-04) (hash(x)=14062293)
204
+ 194 train 8.328234 (lr=1.3636e-04) (hash(x)=10785993)
205
+ 195 train 8.513978 (lr=1.3706e-04) (hash(x)=14736057)
206
+ 196 train 8.243157 (lr=1.3776e-04) (hash(x)=12054745)
207
+ 197 train 8.803791 (lr=1.3846e-04) (hash(x)=16036112)
208
+ 198 train 8.674994 (lr=1.3916e-04) (hash(x)=16263589)
209
+ 199 train 8.264668 (lr=1.3986e-04) (hash(x)=11482477)
210
+ 200 val loss 8.2186
211
+ 200 val perplexity 3709.1936
212
+ 200 train 8.331128 (lr=1.4056e-04) (hash(x)=13356719)
213
+ 201 train 8.360603 (lr=1.4126e-04) (hash(x)=12742648)
214
+ 202 train 8.178341 (lr=1.4196e-04) (hash(x)=11963702)
215
+ 203 train 8.321255 (lr=1.4266e-04) (hash(x)=15429973)
216
+ 204 train 8.185876 (lr=1.4336e-04) (hash(x)=13257554)
217
+ 205 train 8.299450 (lr=1.4406e-04) (hash(x)=13972190)
218
+ 206 train 8.356649 (lr=1.4476e-04) (hash(x)=13430094)
219
+ 207 train 8.065244 (lr=1.4545e-04) (hash(x)=10345057)
220
+ 208 train 8.241135 (lr=1.4615e-04) (hash(x)=14689084)
221
+ 209 train 8.048694 (lr=1.4685e-04) (hash(x)=11609727)
222
+ 210 train 8.132227 (lr=1.4755e-04) (hash(x)=13860968)
223
+ 211 train 7.983911 (lr=1.4825e-04) (hash(x)=10816510)
224
+ 212 train 7.805987 (lr=1.4895e-04) (hash(x)=10255921)
225
+ 213 train 8.265846 (lr=1.4965e-04) (hash(x)=11801604)
226
+ 214 train 8.024021 (lr=1.5035e-04) (hash(x)=11442918)
227
+ 215 train 8.037505 (lr=1.5105e-04) (hash(x)=11157124)
228
+ 216 train 8.154212 (lr=1.5175e-04) (hash(x)=9862579)
229
+ 217 train 7.988428 (lr=1.5245e-04) (hash(x)=11666699)
230
+ 218 train 8.146107 (lr=1.5315e-04) (hash(x)=15933079)
231
+ 219 train 8.085238 (lr=1.5385e-04) (hash(x)=12349897)
232
+ 220 train 8.070837 (lr=1.5455e-04) (hash(x)=11150337)
233
+ 221 train 8.121499 (lr=1.5524e-04) (hash(x)=11648826)
234
+ 222 train 8.110094 (lr=1.5594e-04) (hash(x)=11198064)
235
+ 223 train 7.869467 (lr=1.5664e-04) (hash(x)=10750452)
236
+ 224 train 7.932137 (lr=1.5734e-04) (hash(x)=11819859)
237
+ 225 train 8.041350 (lr=1.5804e-04) (hash(x)=12114386)
238
+ 226 train 8.131905 (lr=1.5874e-04) (hash(x)=13023365)
239
+ 227 train 8.070331 (lr=1.5944e-04) (hash(x)=11859950)
240
+ 228 train 7.916239 (lr=1.6014e-04) (hash(x)=12542485)
241
+ 229 train 7.945375 (lr=1.6084e-04) (hash(x)=12837162)
242
+ 230 train 7.973840 (lr=1.6154e-04) (hash(x)=12185415)
243
+ 231 train 7.845544 (lr=1.6224e-04) (hash(x)=8921601)
244
+ 232 train 7.823643 (lr=1.6294e-04) (hash(x)=11572366)
245
+ 233 train 7.399642 (lr=1.6364e-04) (hash(x)=9125058)
246
+ 234 train 7.759228 (lr=1.6434e-04) (hash(x)=12007582)
247
+ 235 train 7.883780 (lr=1.6503e-04) (hash(x)=11333049)
248
+ 236 train 7.812930 (lr=1.6573e-04) (hash(x)=10626272)
249
+ 237 train 8.009889 (lr=1.6643e-04) (hash(x)=13807539)
250
+ 238 train 7.948827 (lr=1.6713e-04) (hash(x)=11429976)
251
+ 239 train 7.732471 (lr=1.6783e-04) (hash(x)=12165500)
252
+ 240 train 8.154915 (lr=1.6853e-04) (hash(x)=15491314)
253
+ 241 train 8.031611 (lr=1.6923e-04) (hash(x)=14343620)
254
+ 242 train 7.963512 (lr=1.6993e-04) (hash(x)=12358970)
255
+ 243 train 8.021643 (lr=1.7063e-04) (hash(x)=13153948)
256
+ 244 train 8.157185 (lr=1.7133e-04) (hash(x)=12935023)
257
+ 245 train 7.930007 (lr=1.7203e-04) (hash(x)=14497853)
258
+ 246 train 7.874344 (lr=1.7273e-04) (hash(x)=12545808)
259
+ 247 train 8.218092 (lr=1.7343e-04) (hash(x)=13569316)
260
+ 248 train 7.929280 (lr=1.7413e-04) (hash(x)=13033944)
261
+ 249 train 8.128469 (lr=1.7483e-04) (hash(x)=14212041)
262
+ 250 val loss 7.8500
263
+ 250 val perplexity 2565.6509
264
+ 250 train 8.190201 (lr=1.7552e-04) (hash(x)=10769873)
265
+ 251 train 7.948317 (lr=1.7622e-04) (hash(x)=11456278)
266
+ 252 train 7.892446 (lr=1.7692e-04) (hash(x)=13141720)
267
+ 253 train 7.824191 (lr=1.7762e-04) (hash(x)=13563812)
268
+ 254 train 7.896578 (lr=1.7832e-04) (hash(x)=12645610)
269
+ 255 train 7.865828 (lr=1.7902e-04) (hash(x)=14043713)
270
+ 256 train 8.064789 (lr=1.7972e-04) (hash(x)=14545052)
271
+ 257 train 7.805581 (lr=1.8042e-04) (hash(x)=13968786)
272
+ 258 train 8.033873 (lr=1.8112e-04) (hash(x)=11573299)
273
+ 259 train 7.807196 (lr=1.8182e-04) (hash(x)=11896948)
274
+ 260 train 8.091600 (lr=1.8252e-04) (hash(x)=11602278)
275
+ 261 train 7.977814 (lr=1.8322e-04) (hash(x)=13298324)
276
+ 262 train 8.219219 (lr=1.8392e-04) (hash(x)=13232135)
277
+ 263 train 7.755114 (lr=1.8462e-04) (hash(x)=11969650)
278
+ 264 train 7.849348 (lr=1.8531e-04) (hash(x)=12202825)
279
+ 265 train 7.831344 (lr=1.8601e-04) (hash(x)=12021926)
280
+ 266 train 7.743421 (lr=1.8671e-04) (hash(x)=11486982)
281
+ 267 train 7.743017 (lr=1.8741e-04) (hash(x)=14387351)
282
+ 268 train 7.976931 (lr=1.8811e-04) (hash(x)=14588570)
283
+ 269 train 7.600177 (lr=1.8881e-04) (hash(x)=13939547)
284
+ 270 train 7.712019 (lr=1.8951e-04) (hash(x)=13336938)
285
+ 271 train 7.631325 (lr=1.9021e-04) (hash(x)=13680255)
286
+ 272 train 7.828246 (lr=1.9091e-04) (hash(x)=11276607)
287
+ 273 train 7.859665 (lr=1.9161e-04) (hash(x)=10982564)
288
+ 274 train 8.077077 (lr=1.9231e-04) (hash(x)=13821541)
289
+ 275 train 8.178282 (lr=1.9301e-04) (hash(x)=11875534)
290
+ 276 train 7.903212 (lr=1.9371e-04) (hash(x)=14296280)
291
+ 277 train 8.149975 (lr=1.9441e-04) (hash(x)=13390260)
292
+ 278 train 8.121318 (lr=1.9510e-04) (hash(x)=13142039)
293
+ 279 train 7.871567 (lr=1.9580e-04) (hash(x)=13871675)
294
+ 280 train 7.856265 (lr=1.9650e-04) (hash(x)=13297982)
295
+ 281 train 7.576210 (lr=1.9720e-04) (hash(x)=10728444)
296
+ 282 train 7.767710 (lr=1.9790e-04) (hash(x)=11539550)
297
+ 283 train 7.980420 (lr=1.9860e-04) (hash(x)=13389922)
298
+ 284 train 7.970659 (lr=1.9930e-04) (hash(x)=13390153)
299
+ 285 train 8.135550 (lr=2.0000e-04) (hash(x)=13458742)
300
+ 286 train 7.620024 (lr=2.0000e-04) (hash(x)=10387541)
301
+ 287 train 7.634505 (lr=2.0000e-04) (hash(x)=12263431)
302
+ 288 train 7.617607 (lr=2.0000e-04) (hash(x)=10405877)
303
+ 289 train 7.730730 (lr=1.9999e-04) (hash(x)=12377887)
304
+ 290 train 7.521780 (lr=1.9999e-04) (hash(x)=10511081)
305
+ 291 train 7.839685 (lr=1.9998e-04) (hash(x)=10141365)
306
+ 292 train 7.896434 (lr=1.9997e-04) (hash(x)=13935127)
307
+ 293 train 8.016750 (lr=1.9996e-04) (hash(x)=13965689)
308
+ 294 train 7.658750 (lr=1.9994e-04) (hash(x)=13301076)
309
+ 295 train 7.606771 (lr=1.9993e-04) (hash(x)=12414778)
310
+ 296 train 7.824572 (lr=1.9991e-04) (hash(x)=12497839)
311
+ 297 train 7.494140 (lr=1.9989e-04) (hash(x)=9727144)
312
+ 298 train 7.627941 (lr=1.9987e-04) (hash(x)=11972438)
313
+ 299 train 7.872811 (lr=1.9985e-04) (hash(x)=11738134)
314
+ 300 val loss 7.7282
315
+ 300 val perplexity 2271.5205
316
+ 300 train 8.634287 (lr=1.9983e-04) (hash(x)=15826242)
317
+ 301 train 8.360047 (lr=1.9980e-04) (hash(x)=14924380)
318
+ 302 train 8.059079 (lr=1.9978e-04) (hash(x)=10605039)
319
+ 303 train 7.352702 (lr=1.9975e-04) (hash(x)=14443665)
320
+ 304 train 8.004745 (lr=1.9972e-04) (hash(x)=14431398)
321
+ 305 train 7.804030 (lr=1.9969e-04) (hash(x)=12873827)
322
+ 306 train 7.873482 (lr=1.9965e-04) (hash(x)=11307508)
323
+ 307 train 7.864730 (lr=1.9962e-04) (hash(x)=13435160)
324
+ 308 train 8.038187 (lr=1.9958e-04) (hash(x)=13345013)
325
+ 309 train 8.086296 (lr=1.9954e-04) (hash(x)=13024409)
326
+ 310 train 8.245039 (lr=1.9950e-04) (hash(x)=14339476)
327
+ 311 train 7.999329 (lr=1.9946e-04) (hash(x)=13100449)
328
+ 312 train 7.720766 (lr=1.9941e-04) (hash(x)=10843771)
329
+ 313 train 7.954684 (lr=1.9937e-04) (hash(x)=13304906)
330
+ 314 train 7.787072 (lr=1.9932e-04) (hash(x)=11167598)
331
+ 315 train 7.760123 (lr=1.9927e-04) (hash(x)=11690244)
332
+ 316 train 7.746840 (lr=1.9922e-04) (hash(x)=11802493)
333
+ 317 train 7.955032 (lr=1.9916e-04) (hash(x)=12027838)
334
+ 318 train 7.565988 (lr=1.9911e-04) (hash(x)=12643274)
335
+ 319 train 7.800028 (lr=1.9905e-04) (hash(x)=13953073)
336
+ 320 train 7.793595 (lr=1.9899e-04) (hash(x)=12882377)
337
+ 321 train 7.841989 (lr=1.9893e-04) (hash(x)=12066345)
338
+ 322 train 8.082265 (lr=1.9887e-04) (hash(x)=14822016)
339
+ 323 train 7.995302 (lr=1.9881e-04) (hash(x)=12657159)
340
+ 324 train 7.718143 (lr=1.9874e-04) (hash(x)=11740673)
341
+ 325 train 8.037291 (lr=1.9868e-04) (hash(x)=11335971)
342
+ 326 train 7.806524 (lr=1.9861e-04) (hash(x)=11550890)
343
+ 327 train 7.436589 (lr=1.9854e-04) (hash(x)=10956041)
344
+ 328 train 7.479196 (lr=1.9847e-04) (hash(x)=9941344)
345
+ 329 train 7.386048 (lr=1.9839e-04) (hash(x)=9766608)
346
+ 330 train 7.454962 (lr=1.9832e-04) (hash(x)=11007490)
347
+ 331 train 7.298974 (lr=1.9824e-04) (hash(x)=11060230)
348
+ 332 train 8.090384 (lr=1.9816e-04) (hash(x)=15459248)
349
+ 333 train 7.888937 (lr=1.9808e-04) (hash(x)=11396268)
350
+ 334 train 7.703300 (lr=1.9800e-04) (hash(x)=11127560)
351
+ 335 train 7.754978 (lr=1.9792e-04) (hash(x)=12840879)
352
+ 336 train 7.589744 (lr=1.9783e-04) (hash(x)=11354437)
353
+ 337 train 7.623497 (lr=1.9774e-04) (hash(x)=10987186)
354
+ 338 train 7.503930 (lr=1.9765e-04) (hash(x)=8972293)
355
+ 339 train 7.418931 (lr=1.9756e-04) (hash(x)=8916599)
356
+ 340 train 7.589338 (lr=1.9747e-04) (hash(x)=10487177)
357
+ 341 train 8.049152 (lr=1.9738e-04) (hash(x)=13235433)
358
+ 342 train 7.780016 (lr=1.9728e-04) (hash(x)=12916423)
359
+ 343 train 7.852749 (lr=1.9718e-04) (hash(x)=14428050)
360
+ 344 train 7.842262 (lr=1.9709e-04) (hash(x)=13204551)
361
+ 345 train 7.803088 (lr=1.9698e-04) (hash(x)=12631329)
362
+ 346 train 7.814667 (lr=1.9688e-04) (hash(x)=13809442)
363
+ 347 train 7.740370 (lr=1.9678e-04) (hash(x)=12787108)
364
+ 348 train 7.754910 (lr=1.9667e-04) (hash(x)=11912209)
365
+ 349 train 7.860953 (lr=1.9656e-04) (hash(x)=14499940)
366
+ 350 val loss 7.7193
367
+ 350 val perplexity 2251.4192
368
+ 350 train 7.978496 (lr=1.9646e-04) (hash(x)=13911589)
369
+ 351 train 8.050618 (lr=1.9634e-04) (hash(x)=14104576)
370
+ 352 train 7.914159 (lr=1.9623e-04) (hash(x)=12414592)
371
+ 353 train 7.954380 (lr=1.9612e-04) (hash(x)=13943168)
372
+ 354 train 7.844378 (lr=1.9600e-04) (hash(x)=12770473)
373
+ 355 train 7.729273 (lr=1.9588e-04) (hash(x)=12623743)
374
+ 356 train 7.839377 (lr=1.9576e-04) (hash(x)=12441115)
375
+ 357 train 7.871043 (lr=1.9564e-04) (hash(x)=12689741)
376
+ 358 train 7.779885 (lr=1.9552e-04) (hash(x)=12703188)
377
+ 359 train 7.805457 (lr=1.9540e-04) (hash(x)=14453314)
378
+ 360 train 7.669273 (lr=1.9527e-04) (hash(x)=12038031)
379
+ 361 train 7.726473 (lr=1.9514e-04) (hash(x)=10824391)
380
+ 362 train 7.740187 (lr=1.9501e-04) (hash(x)=11527801)
381
+ 363 train 7.400313 (lr=1.9488e-04) (hash(x)=10470424)
382
+ 364 train 7.301344 (lr=1.9475e-04) (hash(x)=12820231)
383
+ 365 train 8.751458 (lr=1.9462e-04) (hash(x)=15186277)
384
+ 366 train 7.885677 (lr=1.9448e-04) (hash(x)=13477578)
385
+ 367 train 7.874398 (lr=1.9434e-04) (hash(x)=12918172)
386
+ 368 train 7.579510 (lr=1.9421e-04) (hash(x)=12480955)
387
+ 369 train 7.950743 (lr=1.9406e-04) (hash(x)=11936107)
388
+ 370 train 7.590088 (lr=1.9392e-04) (hash(x)=9742856)
389
+ 371 train 7.897593 (lr=1.9378e-04) (hash(x)=13259459)
390
+ 372 train 7.712980 (lr=1.9363e-04) (hash(x)=11608070)
391
+ 373 train 7.665611 (lr=1.9349e-04) (hash(x)=12918385)
392
+ 374 train 7.796329 (lr=1.9334e-04) (hash(x)=12761761)
393
+ 375 train 7.740853 (lr=1.9319e-04) (hash(x)=12368572)
394
+ 376 train 7.802665 (lr=1.9304e-04) (hash(x)=11587019)
395
+ 377 train 8.045978 (lr=1.9288e-04) (hash(x)=13455578)
396
+ 378 train 7.792904 (lr=1.9273e-04) (hash(x)=12043438)
397
+ 379 train 7.730219 (lr=1.9257e-04) (hash(x)=13505144)
398
+ 380 train 7.705760 (lr=1.9241e-04) (hash(x)=12626641)
399
+ 381 train 7.922672 (lr=1.9225e-04) (hash(x)=13370946)
400
+ 382 train 7.450847 (lr=1.9209e-04) (hash(x)=9854862)
401
+ 383 train 7.778313 (lr=1.9193e-04) (hash(x)=13340443)
402
+ 384 train 7.726304 (lr=1.9176e-04) (hash(x)=12641299)
403
+ 385 train 7.676025 (lr=1.9160e-04) (hash(x)=12806141)
404
+ 386 train 8.008121 (lr=1.9143e-04) (hash(x)=12354021)
405
+ 387 train 7.608544 (lr=1.9126e-04) (hash(x)=11469900)
406
+ 388 train 7.259594 (lr=1.9109e-04) (hash(x)=10099706)
407
+ 389 train 7.648619 (lr=1.9091e-04) (hash(x)=11591533)
408
+ 390 train 7.666184 (lr=1.9074e-04) (hash(x)=11174793)
409
+ 391 train 7.459403 (lr=1.9056e-04) (hash(x)=10742101)
410
+ 392 train 7.673098 (lr=1.9039e-04) (hash(x)=13136821)
411
+ 393 train 7.615827 (lr=1.9021e-04) (hash(x)=11781196)
412
+ 394 train 7.449391 (lr=1.9003e-04) (hash(x)=11847817)
413
+ 395 train 7.756083 (lr=1.8985e-04) (hash(x)=13261517)
414
+ 396 train 7.585944 (lr=1.8966e-04) (hash(x)=11622233)
415
+ 397 train 7.881237 (lr=1.8948e-04) (hash(x)=13592858)
416
+ 398 train 7.973800 (lr=1.8929e-04) (hash(x)=15017875)
417
+ 399 train 7.893511 (lr=1.8910e-04) (hash(x)=13025742)
418
+ 400 val loss 7.7060
419
+ 400 val perplexity 2221.6523
420
+ 400 train 7.704328 (lr=1.8891e-04) (hash(x)=11718604)
421
+ 401 train 7.721273 (lr=1.8872e-04) (hash(x)=12152419)
422
+ 402 train 7.802438 (lr=1.8853e-04) (hash(x)=14069856)
423
+ 403 train 8.247646 (lr=1.8834e-04) (hash(x)=15100390)
424
+ 404 train 7.920923 (lr=1.8814e-04) (hash(x)=12718574)
425
+ 405 train 7.554693 (lr=1.8794e-04) (hash(x)=10708197)
426
+ 406 train 7.698997 (lr=1.8774e-04) (hash(x)=10819941)
427
+ 407 train 7.784525 (lr=1.8754e-04) (hash(x)=14351068)
428
+ 408 train 7.895465 (lr=1.8734e-04) (hash(x)=14377468)
429
+ 409 train 7.962371 (lr=1.8714e-04) (hash(x)=12619936)
430
+ 410 train 7.454318 (lr=1.8693e-04) (hash(x)=7908802)
431
+ 411 train 7.643429 (lr=1.8673e-04) (hash(x)=13445545)
432
+ 412 train 7.917080 (lr=1.8652e-04) (hash(x)=14062361)
433
+ 413 train 7.631527 (lr=1.8631e-04) (hash(x)=11375930)
434
+ 414 train 7.680193 (lr=1.8610e-04) (hash(x)=10519208)
435
+ 415 train 8.157230 (lr=1.8589e-04) (hash(x)=13155663)
436
+ 416 train 8.048059 (lr=1.8567e-04) (hash(x)=15774406)
437
+ 417 train 7.685439 (lr=1.8546e-04) (hash(x)=13404244)
438
+ 418 train 7.679968 (lr=1.8524e-04) (hash(x)=9995421)
439
+ 419 train 7.702252 (lr=1.8502e-04) (hash(x)=13309898)
440
+ 420 train 7.913348 (lr=1.8480e-04) (hash(x)=13892965)
441
+ 421 train 7.790037 (lr=1.8458e-04) (hash(x)=12356526)
442
+ 422 train 7.733968 (lr=1.8436e-04) (hash(x)=12133609)
443
+ 423 train 7.497521 (lr=1.8414e-04) (hash(x)=9157946)
444
+ 424 train 7.509605 (lr=1.8391e-04) (hash(x)=11770021)
445
+ 425 train 7.706327 (lr=1.8369e-04) (hash(x)=11837703)
446
+ 426 train 7.633355 (lr=1.8346e-04) (hash(x)=9337790)
447
+ 427 train 7.680883 (lr=1.8323e-04) (hash(x)=12222186)
448
+ 428 train 7.581763 (lr=1.8300e-04) (hash(x)=11408945)
449
+ 429 train 7.695054 (lr=1.8276e-04) (hash(x)=12588150)
450
+ 430 train 7.588254 (lr=1.8253e-04) (hash(x)=12035009)
451
+ 431 train 7.720014 (lr=1.8230e-04) (hash(x)=13872294)
452
+ 432 train 7.754536 (lr=1.8206e-04) (hash(x)=13115811)
453
+ 433 train 7.506521 (lr=1.8182e-04) (hash(x)=10621792)
454
+ 434 train 7.717063 (lr=1.8158e-04) (hash(x)=11829316)
455
+ 435 train 7.545071 (lr=1.8134e-04) (hash(x)=11824400)
456
+ 436 train 7.739806 (lr=1.8110e-04) (hash(x)=13258944)
457
+ 437 train 7.872005 (lr=1.8086e-04) (hash(x)=12495032)
458
+ 438 train 7.738230 (lr=1.8061e-04) (hash(x)=12623570)
459
+ 439 train 7.813132 (lr=1.8036e-04) (hash(x)=12773612)
460
+ 440 train 7.862187 (lr=1.8012e-04) (hash(x)=11183399)
461
+ 441 train 7.588416 (lr=1.7987e-04) (hash(x)=10808441)
462
+ 442 train 7.877497 (lr=1.7962e-04) (hash(x)=13523831)
463
+ 443 train 7.786539 (lr=1.7937e-04) (hash(x)=12794078)
464
+ 444 train 7.707970 (lr=1.7911e-04) (hash(x)=12098496)
465
+ 445 train 7.743603 (lr=1.7886e-04) (hash(x)=11328897)
466
+ 446 train 7.851244 (lr=1.7860e-04) (hash(x)=12482357)
467
+ 447 train 7.699032 (lr=1.7835e-04) (hash(x)=12235713)
468
+ 448 train 7.452949 (lr=1.7809e-04) (hash(x)=12612778)
469
+ 449 train 7.732646 (lr=1.7783e-04) (hash(x)=11512364)
470
+ 450 val loss 7.6713
471
+ 450 val perplexity 2145.8027
472
+ 450 train 7.785788 (lr=1.7757e-04) (hash(x)=13110980)
473
+ 451 train 7.376571 (lr=1.7731e-04) (hash(x)=10270693)
474
+ 452 train 7.389279 (lr=1.7704e-04) (hash(x)=10333232)
475
+ 453 train 7.546477 (lr=1.7678e-04) (hash(x)=12380104)
476
+ 454 train 7.426376 (lr=1.7651e-04) (hash(x)=12106102)
477
+ 455 train 7.636286 (lr=1.7624e-04) (hash(x)=13686441)
478
+ 456 train 8.261515 (lr=1.7597e-04) (hash(x)=11802878)
479
+ 457 train 7.950126 (lr=1.7570e-04) (hash(x)=14151557)
480
+ 458 train 7.973166 (lr=1.7543e-04) (hash(x)=13785241)
481
+ 459 train 7.790209 (lr=1.7516e-04) (hash(x)=13567687)
482
+ 460 train 7.814677 (lr=1.7489e-04) (hash(x)=12731340)
483
+ 461 train 7.681212 (lr=1.7461e-04) (hash(x)=12718282)
484
+ 462 train 7.800984 (lr=1.7434e-04) (hash(x)=13733829)
485
+ 463 train 7.518016 (lr=1.7406e-04) (hash(x)=13214962)
486
+ 464 train 7.693501 (lr=1.7378e-04) (hash(x)=11298890)
487
+ 465 train 7.574960 (lr=1.7350e-04) (hash(x)=10855282)
488
+ 466 train 8.053904 (lr=1.7322e-04) (hash(x)=14082841)
489
+ 467 train 7.806999 (lr=1.7294e-04) (hash(x)=12404167)
490
+ 468 train 8.220016 (lr=1.7265e-04) (hash(x)=15826094)
491
+ 469 train 7.880946 (lr=1.7237e-04) (hash(x)=13017256)
492
+ 470 train 7.646083 (lr=1.7208e-04) (hash(x)=13446769)
493
+ 471 train 7.691029 (lr=1.7179e-04) (hash(x)=12602681)
494
+ 472 train 7.512251 (lr=1.7151e-04) (hash(x)=10655764)
495
+ 473 train 7.375983 (lr=1.7122e-04) (hash(x)=11916856)
496
+ 474 train 8.006416 (lr=1.7092e-04) (hash(x)=13726124)
497
+ 475 train 7.682047 (lr=1.7063e-04) (hash(x)=13850551)
498
+ 476 train 7.523401 (lr=1.7034e-04) (hash(x)=12614910)
499
+ 477 train 7.552313 (lr=1.7004e-04) (hash(x)=11382479)
500
+ 478 train 7.639317 (lr=1.6975e-04) (hash(x)=13387304)
501
+ 479 train 7.739483 (lr=1.6945e-04) (hash(x)=13829219)
502
+ 480 train 7.740978 (lr=1.6915e-04) (hash(x)=14735902)
503
+ 481 train 8.014008 (lr=1.6886e-04) (hash(x)=13540459)
504
+ 482 train 7.717762 (lr=1.6856e-04) (hash(x)=12591112)
505
+ 483 train 7.812782 (lr=1.6825e-04) (hash(x)=15485762)
506
+ 484 train 7.711108 (lr=1.6795e-04) (hash(x)=12950474)
507
+ 485 train 7.671042 (lr=1.6765e-04) (hash(x)=12268519)
508
+ 486 train 7.171349 (lr=1.6734e-04) (hash(x)=8011105)
509
+ 487 train 7.178797 (lr=1.6704e-04) (hash(x)=9335220)
510
+ 488 train 7.342561 (lr=1.6673e-04) (hash(x)=10948116)
511
+ 489 train 7.895513 (lr=1.6642e-04) (hash(x)=13359270)
512
+ 490 train 7.753210 (lr=1.6611e-04) (hash(x)=12487329)
513
+ 491 train 7.701798 (lr=1.6580e-04) (hash(x)=13162637)
514
+ 492 train 7.867634 (lr=1.6549e-04) (hash(x)=12860332)
515
+ 493 train 7.717323 (lr=1.6518e-04) (hash(x)=13305408)
516
+ 494 train 7.678101 (lr=1.6487e-04) (hash(x)=11796684)
517
+ 495 train 7.777458 (lr=1.6455e-04) (hash(x)=15713183)
518
+ 496 train 7.684218 (lr=1.6424e-04) (hash(x)=11904281)
519
+ 497 train 7.661920 (lr=1.6392e-04) (hash(x)=13536419)
520
+ 498 train 7.522082 (lr=1.6360e-04) (hash(x)=11620971)
521
+ 499 train 7.828085 (lr=1.6328e-04) (hash(x)=13326615)
522
+ 500 val loss 7.6389
523
+ 500 val perplexity 2077.4609
524
+ 500 train 7.591279 (lr=1.6296e-04) (hash(x)=13585589)
525
+ 501 train 7.788630 (lr=1.6264e-04) (hash(x)=12112694)
526
+ 502 train 7.763701 (lr=1.6232e-04) (hash(x)=11524363)
527
+ 503 train 7.580710 (lr=1.6200e-04) (hash(x)=9917297)
528
+ 504 train 7.612589 (lr=1.6168e-04) (hash(x)=13392970)
529
+ 505 train 7.945841 (lr=1.6135e-04) (hash(x)=12656639)
530
+ 506 train 7.176032 (lr=1.6103e-04) (hash(x)=8029433)
531
+ 507 train 6.858540 (lr=1.6070e-04) (hash(x)=6829764)
532
+ 508 train 7.220149 (lr=1.6037e-04) (hash(x)=12312323)
533
+ 509 train 7.777761 (lr=1.6004e-04) (hash(x)=12964320)
534
+ 510 train 7.452763 (lr=1.5971e-04) (hash(x)=12460686)
535
+ 511 train 7.748807 (lr=1.5938e-04) (hash(x)=12688412)
536
+ 512 train 7.541424 (lr=1.5905e-04) (hash(x)=11453733)
537
+ 513 train 8.049465 (lr=1.5872e-04) (hash(x)=14906439)
538
+ 514 train 7.320649 (lr=1.5838e-04) (hash(x)=10903202)
539
+ 515 train 7.360019 (lr=1.5805e-04) (hash(x)=13039934)
540
+ 516 train 7.726252 (lr=1.5772e-04) (hash(x)=11526002)
541
+ 517 train 7.700616 (lr=1.5738e-04) (hash(x)=12020357)
542
+ 518 train 8.000734 (lr=1.5704e-04) (hash(x)=17910848)
543
+ 519 train 7.602094 (lr=1.5670e-04) (hash(x)=12108991)
544
+ 520 train 7.810327 (lr=1.5636e-04) (hash(x)=12549643)
545
+ 521 train 7.842011 (lr=1.5602e-04) (hash(x)=12328329)
546
+ 522 train 7.716413 (lr=1.5568e-04) (hash(x)=12084164)
547
+ 523 train 7.935218 (lr=1.5534e-04) (hash(x)=14740147)
548
+ 524 train 7.618327 (lr=1.5500e-04) (hash(x)=10853071)
549
+ 525 train 7.664051 (lr=1.5466e-04) (hash(x)=11559413)
550
+ 526 train 7.780610 (lr=1.5431e-04) (hash(x)=14655032)
551
+ 527 train 7.660082 (lr=1.5397e-04) (hash(x)=12152216)
552
+ 528 train 7.786831 (lr=1.5362e-04) (hash(x)=14655442)
553
+ 529 train 7.557544 (lr=1.5327e-04) (hash(x)=12708922)
554
+ 530 train 7.762427 (lr=1.5293e-04) (hash(x)=12464208)
555
+ 531 train 7.845828 (lr=1.5258e-04) (hash(x)=14469749)
556
+ 532 train 7.935490 (lr=1.5223e-04) (hash(x)=12899435)
557
+ 533 train 7.691104 (lr=1.5188e-04) (hash(x)=12459368)
558
+ 534 train 7.820300 (lr=1.5153e-04) (hash(x)=11414287)
559
+ 535 train 7.699397 (lr=1.5118e-04) (hash(x)=11421483)
560
+ 536 train 7.563045 (lr=1.5082e-04) (hash(x)=12885164)
561
+ 537 train 7.545448 (lr=1.5047e-04) (hash(x)=12939881)
562
+ 538 train 7.686024 (lr=1.5012e-04) (hash(x)=11989344)
563
+ 539 train 7.817342 (lr=1.4976e-04) (hash(x)=12362234)
564
+ 540 train 7.676189 (lr=1.4941e-04) (hash(x)=11521741)
565
+ 541 train 7.579996 (lr=1.4905e-04) (hash(x)=12322333)
566
+ 542 train 7.339329 (lr=1.4869e-04) (hash(x)=10728195)
567
+ 543 train 7.947592 (lr=1.4833e-04) (hash(x)=13448155)
568
+ 544 train 7.662528 (lr=1.4798e-04) (hash(x)=13495082)
569
+ 545 train 7.705361 (lr=1.4762e-04) (hash(x)=12928941)
570
+ 546 train 7.684577 (lr=1.4726e-04) (hash(x)=11423948)
571
+ 547 train 7.473099 (lr=1.4690e-04) (hash(x)=10971430)
572
+ 548 train 7.479210 (lr=1.4653e-04) (hash(x)=11754720)
573
+ 549 train 7.580957 (lr=1.4617e-04) (hash(x)=13627629)
574
+ 550 val loss 7.5884
575
+ 550 val perplexity 1975.2202
576
+ 550 train 7.518768 (lr=1.4581e-04) (hash(x)=11853600)
577
+ 551 train 7.644204 (lr=1.4545e-04) (hash(x)=11905081)
578
+ 552 train 7.546535 (lr=1.4508e-04) (hash(x)=13718040)
579
+ 553 train 7.668506 (lr=1.4472e-04) (hash(x)=14339184)
580
+ 554 train 7.671719 (lr=1.4435e-04) (hash(x)=14022671)
581
+ 555 train 7.480889 (lr=1.4398e-04) (hash(x)=10917438)
582
+ 556 train 7.771840 (lr=1.4362e-04) (hash(x)=12631504)
583
+ 557 train 7.707402 (lr=1.4325e-04) (hash(x)=13804603)
584
+ 558 train 7.947208 (lr=1.4288e-04) (hash(x)=13445207)
585
+ 559 train 7.655127 (lr=1.4251e-04) (hash(x)=12290857)
586
+ 560 train 7.751757 (lr=1.4214e-04) (hash(x)=12852252)
587
+ 561 train 7.369168 (lr=1.4177e-04) (hash(x)=10211118)
588
+ 562 train 7.803599 (lr=1.4140e-04) (hash(x)=15793523)
589
+ 563 train 7.947701 (lr=1.4103e-04) (hash(x)=13407076)
590
+ 564 train 7.754299 (lr=1.4066e-04) (hash(x)=16676586)
591
+ 565 train 7.664621 (lr=1.4029e-04) (hash(x)=12249945)
592
+ 566 train 7.807219 (lr=1.3991e-04) (hash(x)=11660346)
593
+ 567 train 7.692031 (lr=1.3954e-04) (hash(x)=13234933)
594
+ 568 train 7.551330 (lr=1.3916e-04) (hash(x)=10823532)
595
+ 569 train 7.687667 (lr=1.3879e-04) (hash(x)=15194855)
596
+ 570 train 7.759551 (lr=1.3841e-04) (hash(x)=14843612)
597
+ 571 train 7.760289 (lr=1.3804e-04) (hash(x)=12892366)
598
+ 572 train 7.510780 (lr=1.3766e-04) (hash(x)=11056742)
599
+ 573 train 7.598840 (lr=1.3728e-04) (hash(x)=12609795)
600
+ 574 train 7.584177 (lr=1.3691e-04) (hash(x)=11560140)
601
+ 575 train 7.405404 (lr=1.3653e-04) (hash(x)=9768085)
602
+ 576 train 7.632548 (lr=1.3615e-04) (hash(x)=14009803)
603
+ 577 train 7.448309 (lr=1.3577e-04) (hash(x)=12252104)
604
+ 578 train 7.717127 (lr=1.3539e-04) (hash(x)=14304208)
605
+ 579 train 7.764746 (lr=1.3501e-04) (hash(x)=12692791)
606
+ 580 train 7.736430 (lr=1.3463e-04) (hash(x)=13214828)
607
+ 581 train 7.543068 (lr=1.3425e-04) (hash(x)=11957245)
608
+ 582 train 7.175944 (lr=1.3387e-04) (hash(x)=9605788)
609
+ 583 train 7.226079 (lr=1.3348e-04) (hash(x)=9528974)
610
+ 584 train 7.396194 (lr=1.3310e-04) (hash(x)=11036050)
611
+ 585 train 7.713662 (lr=1.3272e-04) (hash(x)=12231837)
612
+ 586 train 7.564991 (lr=1.3234e-04) (hash(x)=12052445)
613
+ 587 train 7.461040 (lr=1.3195e-04) (hash(x)=10032842)
614
+ 588 train 7.829351 (lr=1.3157e-04) (hash(x)=14951059)
615
+ 589 train 8.534807 (lr=1.3118e-04) (hash(x)=17178097)
616
+ 590 train 8.927412 (lr=1.3080e-04) (hash(x)=18009086)
617
+ 591 train 7.548938 (lr=1.3041e-04) (hash(x)=10720403)
618
+ 592 train 7.791643 (lr=1.3003e-04) (hash(x)=15019211)
619
+ 593 train 7.710717 (lr=1.2964e-04) (hash(x)=13541352)
620
+ 594 train 7.922997 (lr=1.2925e-04) (hash(x)=15227056)
621
+ 595 train 8.167940 (lr=1.2887e-04) (hash(x)=15187742)
622
+ 596 train 8.016665 (lr=1.2848e-04) (hash(x)=12858009)
623
+ 597 train 7.620989 (lr=1.2809e-04) (hash(x)=11971935)
624
+ 598 train 7.719777 (lr=1.2770e-04) (hash(x)=10047324)
625
+ 599 train 7.660329 (lr=1.2732e-04) (hash(x)=12874592)
626
+ 600 val loss 7.5667
627
+ 600 val perplexity 1932.7019
628
+ 600 train 7.612109 (lr=1.2693e-04) (hash(x)=11616298)
629
+ 601 train 7.653146 (lr=1.2654e-04) (hash(x)=12462802)
630
+ 602 train 7.611350 (lr=1.2615e-04) (hash(x)=12864255)
631
+ 603 train 7.768061 (lr=1.2576e-04) (hash(x)=13622545)
632
+ 604 train 7.647554 (lr=1.2537e-04) (hash(x)=14649855)
633
+ 605 train 7.962606 (lr=1.2498e-04) (hash(x)=14255974)
634
+ 606 train 7.860838 (lr=1.2459e-04) (hash(x)=14247081)
635
+ 607 train 7.486697 (lr=1.2420e-04) (hash(x)=11025404)
636
+ 608 train 7.843795 (lr=1.2381e-04) (hash(x)=14204393)
637
+ 609 train 7.702729 (lr=1.2341e-04) (hash(x)=11239636)
638
+ 610 train 7.718433 (lr=1.2302e-04) (hash(x)=12606825)
639
+ 611 train 7.778630 (lr=1.2263e-04) (hash(x)=15190123)
640
+ 612 train 7.846795 (lr=1.2224e-04) (hash(x)=15138106)
641
+ 613 train 7.863890 (lr=1.2185e-04) (hash(x)=11464046)
642
+ 614 train 7.745975 (lr=1.2145e-04) (hash(x)=11150867)
643
+ 615 train 7.732019 (lr=1.2106e-04) (hash(x)=12567779)
644
+ 616 train 7.614575 (lr=1.2067e-04) (hash(x)=11611202)
645
+ 617 train 7.723141 (lr=1.2027e-04) (hash(x)=16046446)
646
+ 618 train 7.555303 (lr=1.1988e-04) (hash(x)=11255030)
647
+ 619 train 7.532802 (lr=1.1949e-04) (hash(x)=12439519)
648
+ 620 train 7.526976 (lr=1.1909e-04) (hash(x)=13534296)
649
+ 621 train 7.328829 (lr=1.1870e-04) (hash(x)=11917658)
650
+ 622 train 7.679214 (lr=1.1830e-04) (hash(x)=11009063)
651
+ 623 train 7.575866 (lr=1.1791e-04) (hash(x)=11826091)
652
+ 624 train 8.134135 (lr=1.1752e-04) (hash(x)=14037695)
653
+ 625 train 7.643112 (lr=1.1712e-04) (hash(x)=10807703)
654
+ 626 train 7.492172 (lr=1.1673e-04) (hash(x)=10870729)
655
+ 627 train 7.597502 (lr=1.1633e-04) (hash(x)=14723866)
656
+ 628 train 7.731320 (lr=1.1594e-04) (hash(x)=13363055)
657
+ 629 train 7.913122 (lr=1.1554e-04) (hash(x)=11315230)
658
+ 630 train 8.675401 (lr=1.1515e-04) (hash(x)=11232721)
659
+ 631 train 8.099237 (lr=1.1475e-04) (hash(x)=11259652)
660
+ 632 train 8.211958 (lr=1.1435e-04) (hash(x)=12186888)
661
+ 633 train 8.463395 (lr=1.1396e-04) (hash(x)=11702582)
662
+ 634 train 7.994914 (lr=1.1356e-04) (hash(x)=12503047)
663
+ 635 train 7.664832 (lr=1.1317e-04) (hash(x)=11135951)
664
+ 636 train 7.695299 (lr=1.1277e-04) (hash(x)=11955498)
665
+ 637 train 7.798506 (lr=1.1238e-04) (hash(x)=11570885)
666
+ 638 train 7.590965 (lr=1.1198e-04) (hash(x)=12155844)
667
+ 639 train 7.606598 (lr=1.1158e-04) (hash(x)=14642293)
668
+ 640 train 7.385404 (lr=1.1119e-04) (hash(x)=13839875)
669
+ 641 train 7.415222 (lr=1.1079e-04) (hash(x)=13037070)
670
+ 642 train 7.634869 (lr=1.1040e-04) (hash(x)=14520991)
671
+ 643 train 7.489083 (lr=1.1000e-04) (hash(x)=12485426)
672
+ 644 train 7.467604 (lr=1.0960e-04) (hash(x)=13014970)
673
+ 645 train 7.764840 (lr=1.0921e-04) (hash(x)=15142358)
674
+ 646 train 7.402581 (lr=1.0881e-04) (hash(x)=10027296)
675
+ 647 train 7.455642 (lr=1.0842e-04) (hash(x)=12923132)
676
+ 648 train 7.696591 (lr=1.0802e-04) (hash(x)=12354083)
677
+ 649 train 7.849119 (lr=1.0762e-04) (hash(x)=13055942)
678
+ 650 val loss 7.5418
679
+ 650 val perplexity 1885.1727
680
+ 650 train 7.628264 (lr=1.0723e-04) (hash(x)=13291036)
681
+ 651 train 7.671692 (lr=1.0683e-04) (hash(x)=13559280)
682
+ 652 train 7.621847 (lr=1.0644e-04) (hash(x)=14086368)
683
+ 653 train 7.662685 (lr=1.0604e-04) (hash(x)=15313622)
684
+ 654 train 7.653256 (lr=1.0565e-04) (hash(x)=13812518)
685
+ 655 train 7.513059 (lr=1.0525e-04) (hash(x)=12169964)
686
+ 656 train 7.268090 (lr=1.0485e-04) (hash(x)=11896906)
687
+ 657 train 7.215203 (lr=1.0446e-04) (hash(x)=10902629)
688
+ 658 train 7.008800 (lr=1.0406e-04) (hash(x)=9987921)
689
+ 659 train 7.415229 (lr=1.0367e-04) (hash(x)=12000182)
690
+ 660 train 7.178227 (lr=1.0327e-04) (hash(x)=8940484)
691
+ 661 train 7.503571 (lr=1.0288e-04) (hash(x)=11345877)
692
+ 662 train 7.634153 (lr=1.0248e-04) (hash(x)=14081920)
693
+ 663 train 7.393093 (lr=1.0209e-04) (hash(x)=10948640)
694
+ 664 train 7.221130 (lr=1.0170e-04) (hash(x)=10980535)
695
+ 665 train 7.554907 (lr=1.0130e-04) (hash(x)=13037309)
696
+ 666 train 7.190723 (lr=1.0091e-04) (hash(x)=9375142)
697
+ 667 train 7.223372 (lr=1.0051e-04) (hash(x)=9666433)
698
+ 668 train 7.496276 (lr=1.0012e-04) (hash(x)=11623444)
699
+ 669 train 7.482530 (lr=9.9726e-05) (hash(x)=14049895)
700
+ 670 train 7.668473 (lr=9.9333e-05) (hash(x)=12689579)
701
+ 671 train 7.697872 (lr=9.8940e-05) (hash(x)=14103013)
702
+ 672 train 7.838601 (lr=9.8547e-05) (hash(x)=11809717)
703
+ 673 train 7.637832 (lr=9.8154e-05) (hash(x)=14058870)
704
+ 674 train 7.662465 (lr=9.7762e-05) (hash(x)=13738049)
705
+ 675 train 7.673971 (lr=9.7370e-05) (hash(x)=13553506)
706
+ 676 train 7.644360 (lr=9.6978e-05) (hash(x)=13526463)
707
+ 677 train 7.483742 (lr=9.6586e-05) (hash(x)=12902075)
708
+ 678 train 7.727485 (lr=9.6195e-05) (hash(x)=12466258)
709
+ 679 train 7.196919 (lr=9.5804e-05) (hash(x)=11135042)
710
+ 680 train 7.990532 (lr=9.5413e-05) (hash(x)=14920026)
711
+ 681 train 7.656621 (lr=9.5022e-05) (hash(x)=13650023)
712
+ 682 train 7.472920 (lr=9.4632e-05) (hash(x)=12991042)
713
+ 683 train 7.611134 (lr=9.4242e-05) (hash(x)=14930403)
714
+ 684 train 7.643292 (lr=9.3852e-05) (hash(x)=15386592)
715
+ 685 train 7.760888 (lr=9.3463e-05) (hash(x)=15485935)
716
+ 686 train 8.381590 (lr=9.3073e-05) (hash(x)=15487294)
717
+ 687 train 7.611321 (lr=9.2685e-05) (hash(x)=12404200)
718
+ 688 train 7.420182 (lr=9.2296e-05) (hash(x)=12070779)
719
+ 689 train 7.482160 (lr=9.1908e-05) (hash(x)=12392953)
720
+ 690 train 7.624224 (lr=9.1520e-05) (hash(x)=12331573)
721
+ 691 train 7.481403 (lr=9.1133e-05) (hash(x)=12280314)
722
+ 692 train 7.395922 (lr=9.0746e-05) (hash(x)=11383559)
723
+ 693 train 7.424727 (lr=9.0359e-05) (hash(x)=10900559)
724
+ 694 train 7.572106 (lr=8.9973e-05) (hash(x)=13353547)
725
+ 695 train 7.486903 (lr=8.9587e-05) (hash(x)=11534055)
726
+ 696 train 7.558372 (lr=8.9202e-05) (hash(x)=14797493)
727
+ 697 train 7.489567 (lr=8.8817e-05) (hash(x)=11997160)
728
+ 698 train 7.700190 (lr=8.8432e-05) (hash(x)=13151054)
729
+ 699 train 7.642406 (lr=8.8048e-05) (hash(x)=11308531)
730
+ 700 val loss 7.4913
731
+ 700 val perplexity 1792.3800
732
+ 700 train 7.567808 (lr=8.7664e-05) (hash(x)=13121368)
733
+ 701 train 7.508757 (lr=8.7281e-05) (hash(x)=12759751)
734
+ 702 train 7.569080 (lr=8.6898e-05) (hash(x)=13535835)
735
+ 703 train 7.496762 (lr=8.6515e-05) (hash(x)=10914085)
736
+ 704 train 7.499471 (lr=8.6133e-05) (hash(x)=10617131)
737
+ 705 train 7.839389 (lr=8.5751e-05) (hash(x)=12776027)
738
+ 706 train 7.572231 (lr=8.5370e-05) (hash(x)=11255028)
739
+ 707 train 7.390608 (lr=8.4990e-05) (hash(x)=12357751)
740
+ 708 train 7.507988 (lr=8.4610e-05) (hash(x)=12772363)
741
+ 709 train 7.404257 (lr=8.4230e-05) (hash(x)=11040798)
742
+ 710 train 7.775645 (lr=8.3851e-05) (hash(x)=13528424)
743
+ 711 train 7.807820 (lr=8.3472e-05) (hash(x)=14570713)
744
+ 712 train 7.792303 (lr=8.3094e-05) (hash(x)=13617082)
745
+ 713 train 7.631615 (lr=8.2716e-05) (hash(x)=11492564)
746
+ 714 train 7.482950 (lr=8.2339e-05) (hash(x)=13806329)
747
+ 715 train 7.569128 (lr=8.1963e-05) (hash(x)=12212797)
748
+ 716 train 7.668755 (lr=8.1587e-05) (hash(x)=12715617)
749
+ 717 train 7.488924 (lr=8.1211e-05) (hash(x)=13496879)
750
+ 718 train 7.715611 (lr=8.0836e-05) (hash(x)=12774919)
751
+ 719 train 7.460248 (lr=8.0462e-05) (hash(x)=11680329)
752
+ 720 train 7.567637 (lr=8.0088e-05) (hash(x)=12001582)
753
+ 721 train 7.328339 (lr=7.9715e-05) (hash(x)=11948575)
754
+ 722 train 7.516519 (lr=7.9342e-05) (hash(x)=11246553)
755
+ 723 train 7.449579 (lr=7.8970e-05) (hash(x)=13116772)
756
+ 724 train 7.617133 (lr=7.8599e-05) (hash(x)=13475222)
757
+ 725 train 8.067623 (lr=7.8228e-05) (hash(x)=15527273)
758
+ 726 train 7.168973 (lr=7.7858e-05) (hash(x)=10143298)
759
+ 727 train 7.292366 (lr=7.7488e-05) (hash(x)=12437096)
760
+ 728 train 7.715438 (lr=7.7119e-05) (hash(x)=13259177)
761
+ 729 train 7.555347 (lr=7.6751e-05) (hash(x)=12518287)
762
+ 730 train 7.753469 (lr=7.6383e-05) (hash(x)=12379439)
763
+ 731 train 7.225689 (lr=7.6016e-05) (hash(x)=11220338)
764
+ 732 train 7.348877 (lr=7.5650e-05) (hash(x)=12710818)
765
+ 733 train 7.313824 (lr=7.5284e-05) (hash(x)=12034192)
766
+ 734 train 7.899250 (lr=7.4919e-05) (hash(x)=17970060)
767
+ 735 train 7.802718 (lr=7.4555e-05) (hash(x)=12654843)
768
+ 736 train 7.477330 (lr=7.4191e-05) (hash(x)=13094012)
769
+ 737 train 7.782743 (lr=7.3828e-05) (hash(x)=16116174)
770
+ 738 train 7.619280 (lr=7.3466e-05) (hash(x)=14740205)
771
+ 739 train 7.581219 (lr=7.3105e-05) (hash(x)=13505076)
772
+ 740 train 7.416448 (lr=7.2744e-05) (hash(x)=13104341)
773
+ 741 train 7.650752 (lr=7.2384e-05) (hash(x)=11820607)
774
+ 742 train 8.034712 (lr=7.2024e-05) (hash(x)=16336439)
775
+ 743 train 7.274889 (lr=7.1666e-05) (hash(x)=10606850)
776
+ 744 train 7.405272 (lr=7.1308e-05) (hash(x)=12134587)
777
+ 745 train 7.589751 (lr=7.0950e-05) (hash(x)=13759635)
778
+ 746 train 7.501089 (lr=7.0594e-05) (hash(x)=12898464)
779
+ 747 train 7.551058 (lr=7.0238e-05) (hash(x)=15357227)
780
+ 748 train 7.586432 (lr=6.9884e-05) (hash(x)=13808568)
781
+ 749 train 7.439357 (lr=6.9529e-05) (hash(x)=11597062)
782
+ 750 val loss 7.4628
783
+ 750 val perplexity 1742.0032
784
+ 750 train 7.591985 (lr=6.9176e-05) (hash(x)=11255261)
785
+ 751 train 7.397420 (lr=6.8824e-05) (hash(x)=11281537)
786
+ 752 train 7.083098 (lr=6.8472e-05) (hash(x)=9979447)
787
+ 753 train 7.234200 (lr=6.8121e-05) (hash(x)=13294828)
788
+ 754 train 8.015644 (lr=6.7771e-05) (hash(x)=15954206)
789
+ 755 train 8.042194 (lr=6.7422e-05) (hash(x)=13253249)
790
+ 756 train 7.422839 (lr=6.7073e-05) (hash(x)=10309283)
791
+ 757 train 7.862302 (lr=6.6725e-05) (hash(x)=16101250)
792
+ 758 train 7.793038 (lr=6.6379e-05) (hash(x)=13933994)
793
+ 759 train 7.859650 (lr=6.6033e-05) (hash(x)=12363856)
794
+ 760 train 7.504375 (lr=6.5688e-05) (hash(x)=11923641)
795
+ 761 train 7.589263 (lr=6.5343e-05) (hash(x)=12206540)
796
+ 762 train 7.782245 (lr=6.5000e-05) (hash(x)=14025490)
797
+ 763 train 7.651450 (lr=6.4657e-05) (hash(x)=13580623)
798
+ 764 train 7.662845 (lr=6.4316e-05) (hash(x)=13885651)
799
+ 765 train 7.724654 (lr=6.3975e-05) (hash(x)=12509602)
800
+ 766 train 7.485699 (lr=6.3635e-05) (hash(x)=12966645)
801
+ 767 train 8.091629 (lr=6.3296e-05) (hash(x)=14001265)
802
+ 768 train 7.627005 (lr=6.2958e-05) (hash(x)=13743669)
803
+ 769 train 7.357603 (lr=6.2621e-05) (hash(x)=14290431)
804
+ 770 train 7.468858 (lr=6.2285e-05) (hash(x)=11211355)
805
+ 771 train 7.488520 (lr=6.1950e-05) (hash(x)=13045278)
806
+ 772 train 7.558237 (lr=6.1615e-05) (hash(x)=15088674)
807
+ 773 train 7.575174 (lr=6.1282e-05) (hash(x)=13371383)
808
+ 774 train 7.701767 (lr=6.0949e-05) (hash(x)=12504111)
809
+ 775 train 7.399252 (lr=6.0618e-05) (hash(x)=10444420)
810
+ 776 train 7.169929 (lr=6.0287e-05) (hash(x)=10607988)
811
+ 777 train 7.522538 (lr=5.9958e-05) (hash(x)=12732892)
812
+ 778 train 7.576379 (lr=5.9629e-05) (hash(x)=12650059)
813
+ 779 train 7.709904 (lr=5.9301e-05) (hash(x)=11832964)
814
+ 780 train 7.632542 (lr=5.8974e-05) (hash(x)=14574038)
815
+ 781 train 7.629853 (lr=5.8649e-05) (hash(x)=11477958)
816
+ 782 train 7.415951 (lr=5.8324e-05) (hash(x)=12979425)
817
+ 783 train 7.498928 (lr=5.8000e-05) (hash(x)=12249732)
818
+ 784 train 7.636428 (lr=5.7678e-05) (hash(x)=14867563)
819
+ 785 train 7.540870 (lr=5.7356e-05) (hash(x)=12132933)
820
+ 786 train 7.445901 (lr=5.7035e-05) (hash(x)=13535569)
821
+ 787 train 7.547365 (lr=5.6716e-05) (hash(x)=12307821)
822
+ 788 train 7.086687 (lr=5.6397e-05) (hash(x)=11278357)
823
+ 789 train 7.066176 (lr=5.6079e-05) (hash(x)=11266264)
824
+ 790 train 7.300313 (lr=5.5763e-05) (hash(x)=12955233)
825
+ 791 train 7.589671 (lr=5.5447e-05) (hash(x)=13803785)
826
+ 792 train 7.459930 (lr=5.5133e-05) (hash(x)=14264824)
827
+ 793 train 7.613677 (lr=5.4820e-05) (hash(x)=13256773)
828
+ 794 train 7.413723 (lr=5.4507e-05) (hash(x)=12650537)
829
+ 795 train 7.382134 (lr=5.4196e-05) (hash(x)=12493316)
830
+ 796 train 7.484559 (lr=5.3886e-05) (hash(x)=11836158)
831
+ 797 train 7.360142 (lr=5.3577e-05) (hash(x)=12991346)
832
+ 798 train 7.928028 (lr=5.3269e-05) (hash(x)=13756621)
833
+ 799 train 7.133775 (lr=5.2962e-05) (hash(x)=11689978)
834
+ 800 val loss 7.4363
835
+ 800 val perplexity 1696.4954
836
+ 800 train 7.469406 (lr=5.2656e-05) (hash(x)=10365069)
837
+ 801 train 7.405856 (lr=5.2352e-05) (hash(x)=12056273)
838
+ 802 train 7.568849 (lr=5.2048e-05) (hash(x)=14348732)
839
+ 803 train 7.568042 (lr=5.1746e-05) (hash(x)=11906763)
840
+ 804 train 7.577954 (lr=5.1444e-05) (hash(x)=12437638)
841
+ 805 train 7.527610 (lr=5.1144e-05) (hash(x)=12621559)
842
+ 806 train 7.309839 (lr=5.0845e-05) (hash(x)=11302699)
843
+ 807 train 7.695019 (lr=5.0547e-05) (hash(x)=13865869)
844
+ 808 train 7.495895 (lr=5.0251e-05) (hash(x)=12380967)
845
+ 809 train 7.628712 (lr=4.9955e-05) (hash(x)=11425874)
846
+ 810 train 7.455778 (lr=4.9661e-05) (hash(x)=12723475)
847
+ 811 train 8.334493 (lr=4.9367e-05) (hash(x)=16513375)
848
+ 812 train 7.646509 (lr=4.9075e-05) (hash(x)=13869844)
849
+ 813 train 7.829279 (lr=4.8784e-05) (hash(x)=14492920)
850
+ 814 train 7.660693 (lr=4.8495e-05) (hash(x)=13793114)
851
+ 815 train 7.476605 (lr=4.8206e-05) (hash(x)=13030618)
852
+ 816 train 7.605338 (lr=4.7919e-05) (hash(x)=14792501)
853
+ 817 train 7.410016 (lr=4.7633e-05) (hash(x)=12816692)
854
+ 818 train 7.592035 (lr=4.7348e-05) (hash(x)=12569165)
855
+ 819 train 7.740117 (lr=4.7064e-05) (hash(x)=13701949)
856
+ 820 train 7.475492 (lr=4.6782e-05) (hash(x)=12255108)
857
+ 821 train 7.452014 (lr=4.6501e-05) (hash(x)=12904618)
858
+ 822 train 7.476277 (lr=4.6221e-05) (hash(x)=12258538)
859
+ 823 train 7.730518 (lr=4.5942e-05) (hash(x)=14299833)
860
+ 824 train 7.352813 (lr=4.5664e-05) (hash(x)=10822163)
861
+ 825 train 7.499843 (lr=4.5388e-05) (hash(x)=12586511)
862
+ 826 train 7.438979 (lr=4.5113e-05) (hash(x)=13939370)
863
+ 827 train 7.476046 (lr=4.4839e-05) (hash(x)=11818854)
864
+ 828 train 7.520422 (lr=4.4567e-05) (hash(x)=13752960)
865
+ 829 train 7.513935 (lr=4.4295e-05) (hash(x)=13063115)
866
+ 830 train 7.449214 (lr=4.4025e-05) (hash(x)=12817199)
867
+ 831 train 7.360292 (lr=4.3757e-05) (hash(x)=12908875)
868
+ 832 train 7.687476 (lr=4.3489e-05) (hash(x)=13871627)
869
+ 833 train 7.562818 (lr=4.3223e-05) (hash(x)=12627550)
870
+ 834 train 7.390893 (lr=4.2958e-05) (hash(x)=11338055)
871
+ 835 train 7.410386 (lr=4.2695e-05) (hash(x)=12634656)
872
+ 836 train 7.422427 (lr=4.2432e-05) (hash(x)=12416256)
873
+ 837 train 7.476379 (lr=4.2171e-05) (hash(x)=13466941)
874
+ 838 train 7.624943 (lr=4.1912e-05) (hash(x)=14648030)
875
+ 839 train 7.849475 (lr=4.1654e-05) (hash(x)=13859304)
876
+ 840 train 7.730541 (lr=4.1397e-05) (hash(x)=13483846)
877
+ 841 train 7.651511 (lr=4.1141e-05) (hash(x)=9637450)
878
+ 842 train 7.394114 (lr=4.0887e-05) (hash(x)=11672159)
879
+ 843 train 7.636987 (lr=4.0634e-05) (hash(x)=14940204)
880
+ 844 train 7.459264 (lr=4.0382e-05) (hash(x)=14028640)
881
+ 845 train 7.526485 (lr=4.0132e-05) (hash(x)=13686717)
882
+ 846 train 7.478282 (lr=3.9883e-05) (hash(x)=11037190)
883
+ 847 train 7.676574 (lr=3.9635e-05) (hash(x)=14396964)
884
+ 848 train 7.104583 (lr=3.9389e-05) (hash(x)=10766097)
885
+ 849 train 7.315720 (lr=3.9144e-05) (hash(x)=12949082)
886
+ 850 val loss 7.4190
887
+ 850 val perplexity 1667.3314
888
+ 850 train 7.304620 (lr=3.8901e-05) (hash(x)=13743434)
889
+ 851 train 7.510214 (lr=3.8659e-05) (hash(x)=13577749)
890
+ 852 train 7.531872 (lr=3.8418e-05) (hash(x)=12752830)
891
+ 853 train 7.682255 (lr=3.8178e-05) (hash(x)=13031728)
892
+ 854 train 7.756947 (lr=3.7941e-05) (hash(x)=13630938)
893
+ 855 train 7.512046 (lr=3.7704e-05) (hash(x)=11269801)
894
+ 856 train 7.399909 (lr=3.7469e-05) (hash(x)=12724302)
895
+ 857 train 7.253384 (lr=3.7235e-05) (hash(x)=11106556)
896
+ 858 train 7.186930 (lr=3.7003e-05) (hash(x)=12499441)
897
+ 859 train 7.329587 (lr=3.6772e-05) (hash(x)=11222906)
898
+ 860 train 7.453094 (lr=3.6542e-05) (hash(x)=10586708)
899
+ 861 train 7.632004 (lr=3.6314e-05) (hash(x)=14052269)
900
+ 862 train 7.465525 (lr=3.6088e-05) (hash(x)=13384690)
901
+ 863 train 7.290132 (lr=3.5862e-05) (hash(x)=14486071)
902
+ 864 train 7.486968 (lr=3.5639e-05) (hash(x)=11976037)
903
+ 865 train 7.497848 (lr=3.5416e-05) (hash(x)=13026662)
904
+ 866 train 7.500287 (lr=3.5195e-05) (hash(x)=13443462)
905
+ 867 train 7.514188 (lr=3.4976e-05) (hash(x)=11065744)
906
+ 868 train 7.536957 (lr=3.4758e-05) (hash(x)=12404324)
907
+ 869 train 7.533463 (lr=3.4541e-05) (hash(x)=13318947)
908
+ 870 train 7.739923 (lr=3.4326e-05) (hash(x)=14278749)
909
+ 871 train 7.404979 (lr=3.4113e-05) (hash(x)=11997361)
910
+ 872 train 7.784295 (lr=3.3900e-05) (hash(x)=16591908)
911
+ 873 train 7.432770 (lr=3.3690e-05) (hash(x)=12437178)
912
+ 874 train 7.613464 (lr=3.3480e-05) (hash(x)=11441868)
913
+ 875 train 7.403180 (lr=3.3273e-05) (hash(x)=12705537)
914
+ 876 train 7.524323 (lr=3.3066e-05) (hash(x)=14200097)
915
+ 877 train 7.616915 (lr=3.2862e-05) (hash(x)=15790088)
916
+ 878 train 7.402162 (lr=3.2658e-05) (hash(x)=11804370)
917
+ 879 train 7.278285 (lr=3.2457e-05) (hash(x)=11010336)
918
+ 880 train 7.692947 (lr=3.2256e-05) (hash(x)=10935364)
919
+ 881 train 7.491070 (lr=3.2058e-05) (hash(x)=11401216)
920
+ 882 train 7.541553 (lr=3.1860e-05) (hash(x)=12837899)
921
+ 883 train 7.474320 (lr=3.1665e-05) (hash(x)=15106564)
922
+ 884 train 7.514504 (lr=3.1471e-05) (hash(x)=12856376)
923
+ 885 train 7.450235 (lr=3.1278e-05) (hash(x)=13019227)
924
+ 886 train 7.593466 (lr=3.1087e-05) (hash(x)=13294195)
925
+ 887 train 7.412531 (lr=3.0897e-05) (hash(x)=10864248)
926
+ 888 train 7.234757 (lr=3.0709e-05) (hash(x)=10779619)
927
+ 889 train 7.557247 (lr=3.0522e-05) (hash(x)=12107995)
928
+ 890 train 7.223117 (lr=3.0337e-05) (hash(x)=11879879)
929
+ 891 train 7.410762 (lr=3.0154e-05) (hash(x)=13284954)
930
+ 892 train 7.691232 (lr=2.9972e-05) (hash(x)=12895131)
931
+ 893 train 7.447979 (lr=2.9791e-05) (hash(x)=12430135)
932
+ 894 train 7.538193 (lr=2.9613e-05) (hash(x)=12661748)
933
+ 895 train 7.441061 (lr=2.9435e-05) (hash(x)=11906151)
934
+ 896 train 7.440144 (lr=2.9260e-05) (hash(x)=11476689)
935
+ 897 train 7.231361 (lr=2.9085e-05) (hash(x)=11697970)
936
+ 898 train 7.335961 (lr=2.8913e-05) (hash(x)=10948364)
937
+ 899 train 7.173584 (lr=2.8742e-05) (hash(x)=8629295)
938
+ 900 val loss 7.4015
939
+ 900 val perplexity 1638.3633
940
+ 900 train 7.401530 (lr=2.8572e-05) (hash(x)=12316587)
941
+ 901 train 7.530117 (lr=2.8404e-05) (hash(x)=13977261)
942
+ 902 train 7.495749 (lr=2.8238e-05) (hash(x)=12479583)
943
+ 903 train 7.521832 (lr=2.8073e-05) (hash(x)=15166897)
944
+ 904 train 7.609874 (lr=2.7910e-05) (hash(x)=12472550)
945
+ 905 train 7.659587 (lr=2.7749e-05) (hash(x)=13444474)
946
+ 906 train 7.687065 (lr=2.7589e-05) (hash(x)=13849646)
947
+ 907 train 7.654961 (lr=2.7430e-05) (hash(x)=13108020)
948
+ 908 train 7.425397 (lr=2.7274e-05) (hash(x)=14061108)
949
+ 909 train 7.518952 (lr=2.7119e-05) (hash(x)=12559086)
950
+ 910 train 7.340898 (lr=2.6965e-05) (hash(x)=11562580)
951
+ 911 train 7.489078 (lr=2.6813e-05) (hash(x)=12338963)
952
+ 912 train 7.041017 (lr=2.6663e-05) (hash(x)=10604668)
953
+ 913 train 7.441576 (lr=2.6514e-05) (hash(x)=10411127)
954
+ 914 train 7.468070 (lr=2.6367e-05) (hash(x)=12560107)
955
+ 915 train 7.753387 (lr=2.6221e-05) (hash(x)=17176736)
956
+ 916 train 7.361160 (lr=2.6077e-05) (hash(x)=13034244)
957
+ 917 train 7.398532 (lr=2.5935e-05) (hash(x)=13586859)
958
+ 918 train 7.429888 (lr=2.5795e-05) (hash(x)=14825083)
959
+ 919 train 7.626723 (lr=2.5656e-05) (hash(x)=13235312)
960
+ 920 train 7.405343 (lr=2.5518e-05) (hash(x)=12555636)
961
+ 921 train 7.334251 (lr=2.5383e-05) (hash(x)=10862031)
962
+ 922 train 7.437604 (lr=2.5249e-05) (hash(x)=12129400)
963
+ 923 train 7.589257 (lr=2.5116e-05) (hash(x)=12498316)
964
+ 924 train 7.547704 (lr=2.4985e-05) (hash(x)=13128811)
965
+ 925 train 7.367939 (lr=2.4856e-05) (hash(x)=15109534)
966
+ 926 train 7.190659 (lr=2.4729e-05) (hash(x)=10731002)
967
+ 927 train 7.333186 (lr=2.4603e-05) (hash(x)=11031343)
968
+ 928 train 7.293286 (lr=2.4479e-05) (hash(x)=11552376)
969
+ 929 train 7.447405 (lr=2.4356e-05) (hash(x)=13030101)
970
+ 930 train 7.206554 (lr=2.4235e-05) (hash(x)=13250524)
971
+ 931 train 7.508364 (lr=2.4116e-05) (hash(x)=12787203)
972
+ 932 train 7.371579 (lr=2.3998e-05) (hash(x)=12056192)
973
+ 933 train 7.471001 (lr=2.3883e-05) (hash(x)=13192715)
974
+ 934 train 7.354817 (lr=2.3768e-05) (hash(x)=13092101)
975
+ 935 train 7.572218 (lr=2.3656e-05) (hash(x)=14597757)
976
+ 936 train 7.614328 (lr=2.3545e-05) (hash(x)=13880794)
977
+ 937 train 7.305606 (lr=2.3436e-05) (hash(x)=11548718)
978
+ 938 train 7.541772 (lr=2.3328e-05) (hash(x)=11354422)
979
+ 939 train 7.427123 (lr=2.3222e-05) (hash(x)=12017158)
980
+ 940 train 7.605818 (lr=2.3118e-05) (hash(x)=15839756)
981
+ 941 train 7.334120 (lr=2.3016e-05) (hash(x)=10164479)
982
+ 942 train 7.330738 (lr=2.2915e-05) (hash(x)=13779450)
983
+ 943 train 7.304831 (lr=2.2816e-05) (hash(x)=13571036)
984
+ 944 train 7.278996 (lr=2.2718e-05) (hash(x)=12178618)
985
+ 945 train 7.489436 (lr=2.2623e-05) (hash(x)=11609529)
986
+ 946 train 7.473196 (lr=2.2528e-05) (hash(x)=14297351)
987
+ 947 train 7.439594 (lr=2.2436e-05) (hash(x)=12289875)
988
+ 948 train 7.707874 (lr=2.2345e-05) (hash(x)=13540167)
989
+ 949 train 7.358591 (lr=2.2256e-05) (hash(x)=13086437)
990
+ 950 val loss 7.3895
991
+ 950 val perplexity 1618.8821
992
+ 950 train 7.653452 (lr=2.2169e-05) (hash(x)=15073523)
993
+ 951 train 7.785023 (lr=2.2084e-05) (hash(x)=13225100)
994
+ 952 train 7.300971 (lr=2.2000e-05) (hash(x)=11311171)
995
+ 953 train 7.241302 (lr=2.1918e-05) (hash(x)=12863127)
996
+ 954 train 7.666636 (lr=2.1837e-05) (hash(x)=15288759)
997
+ 955 train 7.188340 (lr=2.1758e-05) (hash(x)=11162887)
998
+ 956 train 7.547403 (lr=2.1681e-05) (hash(x)=9933157)
999
+ 957 train 7.549244 (lr=2.1606e-05) (hash(x)=12364743)
1000
+ 958 train 7.257112 (lr=2.1532e-05) (hash(x)=10723631)
1001
+ 959 train 7.354317 (lr=2.1461e-05) (hash(x)=12281916)
1002
+ 960 train 7.351807 (lr=2.1390e-05) (hash(x)=12131510)
1003
+ 961 train 7.413711 (lr=2.1322e-05) (hash(x)=12249214)
1004
+ 962 train 7.532945 (lr=2.1255e-05) (hash(x)=12301339)
1005
+ 963 train 7.372581 (lr=2.1190e-05) (hash(x)=12348477)
1006
+ 964 train 7.432715 (lr=2.1127e-05) (hash(x)=13002952)
1007
+ 965 train 7.555671 (lr=2.1065e-05) (hash(x)=12468719)
1008
+ 966 train 7.507208 (lr=2.1005e-05) (hash(x)=11222266)
1009
+ 967 train 7.459550 (lr=2.0947e-05) (hash(x)=13634084)
1010
+ 968 train 7.473801 (lr=2.0891e-05) (hash(x)=14045062)
1011
+ 969 train 7.393870 (lr=2.0836e-05) (hash(x)=11740722)
1012
+ 970 train 7.726511 (lr=2.0783e-05) (hash(x)=14816355)
1013
+ 971 train 7.393685 (lr=2.0732e-05) (hash(x)=11378506)
1014
+ 972 train 7.169847 (lr=2.0682e-05) (hash(x)=10458922)
1015
+ 973 train 7.448081 (lr=2.0634e-05) (hash(x)=13417634)
1016
+ 974 train 7.498043 (lr=2.0588e-05) (hash(x)=13825872)
1017
+ 975 train 7.469544 (lr=2.0544e-05) (hash(x)=11801649)
1018
+ 976 train 7.673329 (lr=2.0501e-05) (hash(x)=11686944)
1019
+ 977 train 7.367469 (lr=2.0460e-05) (hash(x)=10917325)
1020
+ 978 train 7.251653 (lr=2.0421e-05) (hash(x)=11207331)
1021
+ 979 train 7.276929 (lr=2.0384e-05) (hash(x)=10552788)
1022
+ 980 train 7.071461 (lr=2.0348e-05) (hash(x)=8947573)
1023
+ 981 train 6.964752 (lr=2.0314e-05) (hash(x)=9421716)
1024
+ 982 train 7.294150 (lr=2.0282e-05) (hash(x)=13067279)
1025
+ 983 train 7.547392 (lr=2.0252e-05) (hash(x)=11746289)
1026
+ 984 train 7.571406 (lr=2.0223e-05) (hash(x)=12788451)
1027
+ 985 train 7.415691 (lr=2.0196e-05) (hash(x)=9589864)
1028
+ 986 train 7.361495 (lr=2.0171e-05) (hash(x)=12600294)
1029
+ 987 train 7.396644 (lr=2.0147e-05) (hash(x)=13826293)
1030
+ 988 train 7.514108 (lr=2.0125e-05) (hash(x)=14679370)
1031
+ 989 train 7.643241 (lr=2.0105e-05) (hash(x)=12720603)
1032
+ 990 train 7.417387 (lr=2.0087e-05) (hash(x)=12387624)
1033
+ 991 train 7.376140 (lr=2.0071e-05) (hash(x)=10783313)
1034
+ 992 train 7.477269 (lr=2.0056e-05) (hash(x)=11773378)
1035
+ 993 train 7.478534 (lr=2.0043e-05) (hash(x)=13077539)
1036
+ 994 train 7.454071 (lr=2.0031e-05) (hash(x)=12541500)
1037
+ 995 train 7.569772 (lr=2.0022e-05) (hash(x)=12913270)
1038
+ 996 train 7.586823 (lr=2.0014e-05) (hash(x)=11426867)
1039
+ 997 train 7.327667 (lr=2.0008e-05) (hash(x)=11228160)
1040
+ 998 train 7.154844 (lr=2.0003e-05) (hash(x)=11408061)
1041
+ 999 val loss 7.3832
1042
+ 999 val perplexity 1608.6898
1043
+ 999 train 7.523584 (lr=2.0001e-05) (hash(x)=14212467)
lr2e-4_total_batch_size5120_seq_len256/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d0b9e63b1377e4e7451fadd8fe33fccfacf628813107ecf5fcf8fe9ddd2f3c3
3
+ size 38587970
lr2e-4_total_batch_size5120_seq_len256/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81f08ff4461e13303e1ebb0080fd8cd61e383f8be9b5355a48ed1d49f747a767
3
+ size 70895430