andrew-healey commited on
Commit
f55407c
·
verified ·
1 Parent(s): bedbc00

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size10240_seq_len1024/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size10240_seq_len1024", "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": 1024, "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": 0.0001, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1e-4_total_batch_size10240_seq_len1024/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b9fd0bfa27ab267155e5ae2bb7e7ccf3640df9cfec6890472ce93f6d886a52e
3
+ size 964
lr1e-4_total_batch_size10240_seq_len1024/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7486
3
+ 0 val perplexity 126579.2500
4
+ 0 train 11.752059 (lr=3.4965e-07) (hash(x)=49261872)
5
+ 1 train 11.749063 (lr=6.9930e-07) (hash(x)=57833283)
6
+ 2 train 11.757311 (lr=1.0490e-06) (hash(x)=51469900)
7
+ 3 train 11.707736 (lr=1.3986e-06) (hash(x)=52769534)
8
+ 4 train 11.743955 (lr=1.7483e-06) (hash(x)=48296575)
9
+ 5 train 11.734865 (lr=2.0979e-06) (hash(x)=50373191)
10
+ 6 train 11.760165 (lr=2.4476e-06) (hash(x)=51583947)
11
+ 7 train 11.734324 (lr=2.7972e-06) (hash(x)=48487754)
12
+ 8 train 11.728084 (lr=3.1469e-06) (hash(x)=54502191)
13
+ 9 train 11.745715 (lr=3.4965e-06) (hash(x)=47781406)
14
+ 10 train 11.735178 (lr=3.8462e-06) (hash(x)=53817479)
15
+ 11 train 11.690192 (lr=4.1958e-06) (hash(x)=48658307)
16
+ 12 train 11.727763 (lr=4.5455e-06) (hash(x)=56467437)
17
+ 13 train 11.727351 (lr=4.8951e-06) (hash(x)=42956464)
18
+ 14 train 11.728564 (lr=5.2448e-06) (hash(x)=50465744)
19
+ 15 train 11.690876 (lr=5.5944e-06) (hash(x)=50106566)
20
+ 16 train 11.653931 (lr=5.9441e-06) (hash(x)=57533378)
21
+ 17 train 11.691808 (lr=6.2937e-06) (hash(x)=50260247)
22
+ 18 train 11.684580 (lr=6.6434e-06) (hash(x)=51106171)
23
+ 19 train 11.646943 (lr=6.9930e-06) (hash(x)=45839663)
24
+ 20 train 11.620096 (lr=7.3427e-06) (hash(x)=50479757)
25
+ 21 train 11.627672 (lr=7.6923e-06) (hash(x)=54489985)
26
+ 22 train 11.612410 (lr=8.0420e-06) (hash(x)=52601470)
27
+ 23 train 11.615122 (lr=8.3916e-06) (hash(x)=50770821)
28
+ 24 train 11.591175 (lr=8.7413e-06) (hash(x)=48459756)
29
+ 25 train 11.562224 (lr=9.0909e-06) (hash(x)=49416075)
30
+ 26 train 11.566208 (lr=9.4406e-06) (hash(x)=51191917)
31
+ 27 train 11.570426 (lr=9.7902e-06) (hash(x)=52323135)
32
+ 28 train 11.534632 (lr=1.0140e-05) (hash(x)=47323142)
33
+ 29 train 11.561912 (lr=1.0490e-05) (hash(x)=53922434)
34
+ 30 train 11.529024 (lr=1.0839e-05) (hash(x)=46277237)
35
+ 31 train 11.465873 (lr=1.1189e-05) (hash(x)=50183213)
36
+ 32 train 11.443418 (lr=1.1538e-05) (hash(x)=49300734)
37
+ 33 train 11.468081 (lr=1.1888e-05) (hash(x)=52003225)
38
+ 34 train 11.436772 (lr=1.2238e-05) (hash(x)=52058527)
39
+ 35 train 11.414644 (lr=1.2587e-05) (hash(x)=56845779)
40
+ 36 train 11.389026 (lr=1.2937e-05) (hash(x)=53027784)
41
+ 37 train 11.364678 (lr=1.3287e-05) (hash(x)=55034939)
42
+ 38 train 11.336169 (lr=1.3636e-05) (hash(x)=55634780)
43
+ 39 train 11.359215 (lr=1.3986e-05) (hash(x)=52671072)
44
+ 40 train 11.329524 (lr=1.4336e-05) (hash(x)=46158462)
45
+ 41 train 11.302370 (lr=1.4685e-05) (hash(x)=46979552)
46
+ 42 train 11.304674 (lr=1.5035e-05) (hash(x)=53668921)
47
+ 43 train 11.265863 (lr=1.5385e-05) (hash(x)=52068996)
48
+ 44 train 11.237091 (lr=1.5734e-05) (hash(x)=49328499)
49
+ 45 train 11.119827 (lr=1.6084e-05) (hash(x)=54071193)
50
+ 46 train 11.125580 (lr=1.6434e-05) (hash(x)=50573962)
51
+ 47 train 11.159789 (lr=1.6783e-05) (hash(x)=47640895)
52
+ 48 train 11.170653 (lr=1.7133e-05) (hash(x)=52381297)
53
+ 49 train 11.104982 (lr=1.7483e-05) (hash(x)=46150086)
54
+ 50 val loss 11.0638
55
+ 50 val perplexity 63821.3320
56
+ 50 train 11.078732 (lr=1.7832e-05) (hash(x)=47852060)
57
+ 51 train 11.031740 (lr=1.8182e-05) (hash(x)=45438450)
58
+ 52 train 11.031905 (lr=1.8531e-05) (hash(x)=47030948)
59
+ 53 train 10.928069 (lr=1.8881e-05) (hash(x)=44573591)
60
+ 54 train 10.963228 (lr=1.9231e-05) (hash(x)=49868350)
61
+ 55 train 11.028304 (lr=1.9580e-05) (hash(x)=62300690)
62
+ 56 train 10.920276 (lr=1.9930e-05) (hash(x)=52900359)
63
+ 57 train 10.900271 (lr=2.0280e-05) (hash(x)=50319996)
64
+ 58 train 10.944350 (lr=2.0629e-05) (hash(x)=63972756)
65
+ 59 train 10.805855 (lr=2.0979e-05) (hash(x)=54520776)
66
+ 60 train 10.751052 (lr=2.1329e-05) (hash(x)=41677283)
67
+ 61 train 10.761518 (lr=2.1678e-05) (hash(x)=44855848)
68
+ 62 train 10.700194 (lr=2.2028e-05) (hash(x)=48794108)
69
+ 63 train 10.728315 (lr=2.2378e-05) (hash(x)=54982586)
70
+ 64 train 10.719250 (lr=2.2727e-05) (hash(x)=51251428)
71
+ 65 train 10.736531 (lr=2.3077e-05) (hash(x)=50784472)
72
+ 66 train 10.688812 (lr=2.3427e-05) (hash(x)=52402568)
73
+ 67 train 10.530327 (lr=2.3776e-05) (hash(x)=50188334)
74
+ 68 train 10.676181 (lr=2.4126e-05) (hash(x)=56568050)
75
+ 69 train 10.580331 (lr=2.4476e-05) (hash(x)=50861237)
76
+ 70 train 10.581285 (lr=2.4825e-05) (hash(x)=52123048)
77
+ 71 train 10.588673 (lr=2.5175e-05) (hash(x)=60945351)
78
+ 72 train 10.562842 (lr=2.5524e-05) (hash(x)=52352539)
79
+ 73 train 10.638957 (lr=2.5874e-05) (hash(x)=56694638)
80
+ 74 train 10.524961 (lr=2.6224e-05) (hash(x)=51084620)
81
+ 75 train 10.615081 (lr=2.6573e-05) (hash(x)=58412625)
82
+ 76 train 10.530808 (lr=2.6923e-05) (hash(x)=56756916)
83
+ 77 train 10.535865 (lr=2.7273e-05) (hash(x)=50750450)
84
+ 78 train 10.505095 (lr=2.7622e-05) (hash(x)=49918759)
85
+ 79 train 10.593051 (lr=2.7972e-05) (hash(x)=51819337)
86
+ 80 train 10.430758 (lr=2.8322e-05) (hash(x)=46506057)
87
+ 81 train 10.429204 (lr=2.8671e-05) (hash(x)=48267882)
88
+ 82 train 10.432565 (lr=2.9021e-05) (hash(x)=52004910)
89
+ 83 train 10.384971 (lr=2.9371e-05) (hash(x)=52889428)
90
+ 84 train 10.326497 (lr=2.9720e-05) (hash(x)=44915533)
91
+ 85 train 10.311108 (lr=3.0070e-05) (hash(x)=49536098)
92
+ 86 train 10.385897 (lr=3.0420e-05) (hash(x)=50097275)
93
+ 87 train 10.350885 (lr=3.0769e-05) (hash(x)=50686547)
94
+ 88 train 10.375159 (lr=3.1119e-05) (hash(x)=49463715)
95
+ 89 train 10.346013 (lr=3.1469e-05) (hash(x)=50818630)
96
+ 90 train 10.307083 (lr=3.1818e-05) (hash(x)=47363971)
97
+ 91 train 10.338363 (lr=3.2168e-05) (hash(x)=49697510)
98
+ 92 train 10.275852 (lr=3.2517e-05) (hash(x)=51247734)
99
+ 93 train 10.332953 (lr=3.2867e-05) (hash(x)=53436760)
100
+ 94 train 10.290838 (lr=3.3217e-05) (hash(x)=42551739)
101
+ 95 train 10.313580 (lr=3.3566e-05) (hash(x)=46807632)
102
+ 96 train 10.290268 (lr=3.3916e-05) (hash(x)=52261514)
103
+ 97 train 10.367702 (lr=3.4266e-05) (hash(x)=54016065)
104
+ 98 train 10.411796 (lr=3.4615e-05) (hash(x)=54782473)
105
+ 99 train 10.439738 (lr=3.4965e-05) (hash(x)=57723080)
106
+ 100 val loss 10.3006
107
+ 100 val perplexity 29750.6074
108
+ 100 train 10.379817 (lr=3.5315e-05) (hash(x)=49355093)
109
+ 101 train 10.213956 (lr=3.5664e-05) (hash(x)=53422909)
110
+ 102 train 10.341072 (lr=3.6014e-05) (hash(x)=51486849)
111
+ 103 train 10.354255 (lr=3.6364e-05) (hash(x)=48459663)
112
+ 104 train 10.289363 (lr=3.6713e-05) (hash(x)=51698063)
113
+ 105 train 10.234508 (lr=3.7063e-05) (hash(x)=48794753)
114
+ 106 train 10.249657 (lr=3.7413e-05) (hash(x)=44825422)
115
+ 107 train 10.282411 (lr=3.7762e-05) (hash(x)=47148322)
116
+ 108 train 10.352004 (lr=3.8112e-05) (hash(x)=49279091)
117
+ 109 train 10.259987 (lr=3.8462e-05) (hash(x)=52460055)
118
+ 110 train 10.284278 (lr=3.8811e-05) (hash(x)=46749333)
119
+ 111 train 10.265469 (lr=3.9161e-05) (hash(x)=45337189)
120
+ 112 train 10.169753 (lr=3.9510e-05) (hash(x)=49845142)
121
+ 113 train 10.276927 (lr=3.9860e-05) (hash(x)=50930055)
122
+ 114 train 10.178627 (lr=4.0210e-05) (hash(x)=50684591)
123
+ 115 train 10.255511 (lr=4.0559e-05) (hash(x)=42125704)
124
+ 116 train 10.075107 (lr=4.0909e-05) (hash(x)=41912026)
125
+ 117 train 10.080246 (lr=4.1259e-05) (hash(x)=49174011)
126
+ 118 train 10.195078 (lr=4.1608e-05) (hash(x)=49497097)
127
+ 119 train 10.161194 (lr=4.1958e-05) (hash(x)=50618057)
128
+ 120 train 10.176349 (lr=4.2308e-05) (hash(x)=59453505)
129
+ 121 train 10.244926 (lr=4.2657e-05) (hash(x)=52423338)
130
+ 122 train 10.171401 (lr=4.3007e-05) (hash(x)=53802843)
131
+ 123 train 10.170405 (lr=4.3357e-05) (hash(x)=51754551)
132
+ 124 train 10.257604 (lr=4.3706e-05) (hash(x)=51856340)
133
+ 125 train 10.216125 (lr=4.4056e-05) (hash(x)=47059587)
134
+ 126 train 10.068848 (lr=4.4406e-05) (hash(x)=51793651)
135
+ 127 train 10.096539 (lr=4.4755e-05) (hash(x)=52187435)
136
+ 128 train 10.100333 (lr=4.5105e-05) (hash(x)=54885326)
137
+ 129 train 10.157353 (lr=4.5455e-05) (hash(x)=50919108)
138
+ 130 train 10.180940 (lr=4.5804e-05) (hash(x)=54468731)
139
+ 131 train 10.088813 (lr=4.6154e-05) (hash(x)=51831972)
140
+ 132 train 10.061813 (lr=4.6503e-05) (hash(x)=48609553)
141
+ 133 train 10.084247 (lr=4.6853e-05) (hash(x)=49049513)
142
+ 134 train 10.033011 (lr=4.7203e-05) (hash(x)=51951796)
143
+ 135 train 9.802541 (lr=4.7552e-05) (hash(x)=56443459)
144
+ 136 train 10.081003 (lr=4.7902e-05) (hash(x)=47261375)
145
+ 137 train 10.150689 (lr=4.8252e-05) (hash(x)=54156964)
146
+ 138 train 10.207404 (lr=4.8601e-05) (hash(x)=55621436)
147
+ 139 train 10.106061 (lr=4.8951e-05) (hash(x)=54578790)
148
+ 140 train 10.054338 (lr=4.9301e-05) (hash(x)=46178169)
149
+ 141 train 9.997040 (lr=4.9650e-05) (hash(x)=52813407)
150
+ 142 train 9.997985 (lr=5.0000e-05) (hash(x)=55046177)
151
+ 143 train 10.062449 (lr=5.0350e-05) (hash(x)=45098704)
152
+ 144 train 10.073506 (lr=5.0699e-05) (hash(x)=45895117)
153
+ 145 train 10.027283 (lr=5.1049e-05) (hash(x)=42452493)
154
+ 146 train 10.038409 (lr=5.1399e-05) (hash(x)=50633976)
155
+ 147 train 10.036118 (lr=5.1748e-05) (hash(x)=45655812)
156
+ 148 train 9.971221 (lr=5.2098e-05) (hash(x)=47272687)
157
+ 149 train 10.007463 (lr=5.2448e-05) (hash(x)=45415908)
158
+ 150 val loss 10.0021
159
+ 150 val perplexity 22071.7383
160
+ 150 train 10.312955 (lr=5.2797e-05) (hash(x)=63111643)
161
+ 151 train 9.903798 (lr=5.3147e-05) (hash(x)=49436419)
162
+ 152 train 10.026584 (lr=5.3497e-05) (hash(x)=52633828)
163
+ 153 train 10.069984 (lr=5.3846e-05) (hash(x)=48033921)
164
+ 154 train 10.029150 (lr=5.4196e-05) (hash(x)=54348904)
165
+ 155 train 10.044086 (lr=5.4545e-05) (hash(x)=56987876)
166
+ 156 train 9.982650 (lr=5.4895e-05) (hash(x)=50155518)
167
+ 157 train 9.925389 (lr=5.5245e-05) (hash(x)=48024575)
168
+ 158 train 9.937182 (lr=5.5594e-05) (hash(x)=53059532)
169
+ 159 train 9.906391 (lr=5.5944e-05) (hash(x)=48339908)
170
+ 160 train 9.916343 (lr=5.6294e-05) (hash(x)=48990896)
171
+ 161 train 9.865720 (lr=5.6643e-05) (hash(x)=53135323)
172
+ 162 train 9.843231 (lr=5.6993e-05) (hash(x)=52351650)
173
+ 163 train 9.909871 (lr=5.7343e-05) (hash(x)=46142231)
174
+ 164 train 9.930681 (lr=5.7692e-05) (hash(x)=40473173)
175
+ 165 train 9.866372 (lr=5.8042e-05) (hash(x)=41898441)
176
+ 166 train 9.980217 (lr=5.8392e-05) (hash(x)=53594433)
177
+ 167 train 9.948334 (lr=5.8741e-05) (hash(x)=46263050)
178
+ 168 train 9.938109 (lr=5.9091e-05) (hash(x)=43863543)
179
+ 169 train 9.851551 (lr=5.9441e-05) (hash(x)=37281177)
180
+ 170 train 9.940963 (lr=5.9790e-05) (hash(x)=47276630)
181
+ 171 train 9.868153 (lr=6.0140e-05) (hash(x)=48285373)
182
+ 172 train 9.852020 (lr=6.0490e-05) (hash(x)=53372103)
183
+ 173 train 9.874806 (lr=6.0839e-05) (hash(x)=51190878)
184
+ 174 train 9.826859 (lr=6.1189e-05) (hash(x)=54854560)
185
+ 175 train 9.872436 (lr=6.1538e-05) (hash(x)=56873965)
186
+ 176 train 9.728643 (lr=6.1888e-05) (hash(x)=53533306)
187
+ 177 train 9.810177 (lr=6.2238e-05) (hash(x)=52595463)
188
+ 178 train 9.909182 (lr=6.2587e-05) (hash(x)=49109869)
189
+ 179 train 9.817554 (lr=6.2937e-05) (hash(x)=50421608)
190
+ 180 train 9.755355 (lr=6.3287e-05) (hash(x)=47228528)
191
+ 181 train 9.711909 (lr=6.3636e-05) (hash(x)=45921624)
192
+ 182 train 9.856186 (lr=6.3986e-05) (hash(x)=54444263)
193
+ 183 train 9.802772 (lr=6.4336e-05) (hash(x)=52802882)
194
+ 184 train 9.752012 (lr=6.4685e-05) (hash(x)=49970100)
195
+ 185 train 9.826804 (lr=6.5035e-05) (hash(x)=45185272)
196
+ 186 train 9.734459 (lr=6.5385e-05) (hash(x)=48064569)
197
+ 187 train 9.848223 (lr=6.5734e-05) (hash(x)=49146692)
198
+ 188 train 9.823517 (lr=6.6084e-05) (hash(x)=52836857)
199
+ 189 train 9.642829 (lr=6.6434e-05) (hash(x)=52690444)
200
+ 190 train 9.749191 (lr=6.6783e-05) (hash(x)=53043440)
201
+ 191 train 9.745676 (lr=6.7133e-05) (hash(x)=45627699)
202
+ 192 train 9.776206 (lr=6.7483e-05) (hash(x)=50337632)
203
+ 193 train 9.769264 (lr=6.7832e-05) (hash(x)=50636707)
204
+ 194 train 9.568782 (lr=6.8182e-05) (hash(x)=43415450)
205
+ 195 train 9.569918 (lr=6.8531e-05) (hash(x)=44771118)
206
+ 196 train 9.613774 (lr=6.8881e-05) (hash(x)=47474403)
207
+ 197 train 9.562140 (lr=6.9231e-05) (hash(x)=48109117)
208
+ 198 train 9.657481 (lr=6.9580e-05) (hash(x)=50752974)
209
+ 199 train 9.736411 (lr=6.9930e-05) (hash(x)=55122837)
210
+ 200 val loss 9.6704
211
+ 200 val perplexity 15841.7549
212
+ 200 train 9.599234 (lr=7.0280e-05) (hash(x)=49692660)
213
+ 201 train 9.822425 (lr=7.0629e-05) (hash(x)=58304741)
214
+ 202 train 9.636789 (lr=7.0979e-05) (hash(x)=51071912)
215
+ 203 train 9.662524 (lr=7.1329e-05) (hash(x)=48891446)
216
+ 204 train 9.710877 (lr=7.1678e-05) (hash(x)=51980344)
217
+ 205 train 9.538668 (lr=7.2028e-05) (hash(x)=46085486)
218
+ 206 train 9.630286 (lr=7.2378e-05) (hash(x)=50429471)
219
+ 207 train 9.644734 (lr=7.2727e-05) (hash(x)=50609171)
220
+ 208 train 9.557456 (lr=7.3077e-05) (hash(x)=56366529)
221
+ 209 train 9.669631 (lr=7.3427e-05) (hash(x)=44693623)
222
+ 210 train 9.655536 (lr=7.3776e-05) (hash(x)=52480393)
223
+ 211 train 9.609107 (lr=7.4126e-05) (hash(x)=45080235)
224
+ 212 train 9.569507 (lr=7.4476e-05) (hash(x)=47772363)
225
+ 213 train 9.451037 (lr=7.4825e-05) (hash(x)=46731794)
226
+ 214 train 9.553939 (lr=7.5175e-05) (hash(x)=44181238)
227
+ 215 train 9.598185 (lr=7.5524e-05) (hash(x)=46674118)
228
+ 216 train 9.425181 (lr=7.5874e-05) (hash(x)=49734489)
229
+ 217 train 9.549738 (lr=7.6224e-05) (hash(x)=46647221)
230
+ 218 train 9.442024 (lr=7.6573e-05) (hash(x)=53699708)
231
+ 219 train 9.519777 (lr=7.6923e-05) (hash(x)=51868493)
232
+ 220 train 9.491533 (lr=7.7273e-05) (hash(x)=49234242)
233
+ 221 train 9.532344 (lr=7.7622e-05) (hash(x)=52926855)
234
+ 222 train 9.430117 (lr=7.7972e-05) (hash(x)=48168200)
235
+ 223 train 9.468180 (lr=7.8322e-05) (hash(x)=50306555)
236
+ 224 train 9.396657 (lr=7.8671e-05) (hash(x)=47450495)
237
+ 225 train 9.402873 (lr=7.9021e-05) (hash(x)=47017597)
238
+ 226 train 9.383057 (lr=7.9371e-05) (hash(x)=45879487)
239
+ 227 train 9.400734 (lr=7.9720e-05) (hash(x)=45930697)
240
+ 228 train 9.539317 (lr=8.0070e-05) (hash(x)=51866248)
241
+ 229 train 9.445711 (lr=8.0420e-05) (hash(x)=55896424)
242
+ 230 train 9.455660 (lr=8.0769e-05) (hash(x)=49856927)
243
+ 231 train 9.369067 (lr=8.1119e-05) (hash(x)=50806347)
244
+ 232 train 9.435968 (lr=8.1469e-05) (hash(x)=48907697)
245
+ 233 train 9.407043 (lr=8.1818e-05) (hash(x)=53269239)
246
+ 234 train 9.490478 (lr=8.2168e-05) (hash(x)=58562988)
247
+ 235 train 9.311584 (lr=8.2517e-05) (hash(x)=50115446)
248
+ 236 train 9.370076 (lr=8.2867e-05) (hash(x)=39354357)
249
+ 237 train 9.415224 (lr=8.3217e-05) (hash(x)=53956776)
250
+ 238 train 9.249819 (lr=8.3566e-05) (hash(x)=49150615)
251
+ 239 train 9.315452 (lr=8.3916e-05) (hash(x)=51243947)
252
+ 240 train 9.414477 (lr=8.4266e-05) (hash(x)=54091498)
253
+ 241 train 9.253624 (lr=8.4615e-05) (hash(x)=55362900)
254
+ 242 train 9.318666 (lr=8.4965e-05) (hash(x)=49537560)
255
+ 243 train 9.182846 (lr=8.5315e-05) (hash(x)=34965283)
256
+ 244 train 9.235908 (lr=8.5664e-05) (hash(x)=43293134)
257
+ 245 train 9.234424 (lr=8.6014e-05) (hash(x)=52583594)
258
+ 246 train 9.304277 (lr=8.6364e-05) (hash(x)=50277570)
259
+ 247 train 9.176147 (lr=8.6713e-05) (hash(x)=53806649)
260
+ 248 train 9.210715 (lr=8.7063e-05) (hash(x)=48819401)
261
+ 249 train 9.170201 (lr=8.7413e-05) (hash(x)=51974675)
262
+ 250 val loss 9.1947
263
+ 250 val perplexity 9844.8154
264
+ 250 train 9.217349 (lr=8.7762e-05) (hash(x)=46284281)
265
+ 251 train 9.216488 (lr=8.8112e-05) (hash(x)=45810574)
266
+ 252 train 9.226006 (lr=8.8462e-05) (hash(x)=49217593)
267
+ 253 train 8.942690 (lr=8.8811e-05) (hash(x)=34273240)
268
+ 254 train 9.047447 (lr=8.9161e-05) (hash(x)=48362439)
269
+ 255 train 9.057227 (lr=8.9510e-05) (hash(x)=50536970)
270
+ 256 train 9.111984 (lr=8.9860e-05) (hash(x)=52438285)
271
+ 257 train 8.845273 (lr=9.0210e-05) (hash(x)=47843837)
272
+ 258 train 9.128763 (lr=9.0559e-05) (hash(x)=50928031)
273
+ 259 train 9.163085 (lr=9.0909e-05) (hash(x)=56940944)
274
+ 260 train 9.125155 (lr=9.1259e-05) (hash(x)=54715437)
275
+ 261 train 9.100197 (lr=9.1608e-05) (hash(x)=54162688)
276
+ 262 train 9.061078 (lr=9.1958e-05) (hash(x)=45398098)
277
+ 263 train 9.067197 (lr=9.2308e-05) (hash(x)=51669498)
278
+ 264 train 8.972820 (lr=9.2657e-05) (hash(x)=52890008)
279
+ 265 train 9.082358 (lr=9.3007e-05) (hash(x)=53020601)
280
+ 266 train 9.079623 (lr=9.3357e-05) (hash(x)=50728575)
281
+ 267 train 9.001111 (lr=9.3706e-05) (hash(x)=46315245)
282
+ 268 train 8.919011 (lr=9.4056e-05) (hash(x)=51700327)
283
+ 269 train 9.079071 (lr=9.4406e-05) (hash(x)=51422940)
284
+ 270 train 8.938221 (lr=9.4755e-05) (hash(x)=50079658)
285
+ 271 train 8.924780 (lr=9.5105e-05) (hash(x)=50610610)
286
+ 272 train 8.852625 (lr=9.5455e-05) (hash(x)=53349554)
287
+ 273 train 8.874093 (lr=9.5804e-05) (hash(x)=47220467)
288
+ 274 train 8.855607 (lr=9.6154e-05) (hash(x)=47059896)
289
+ 275 train 8.926280 (lr=9.6503e-05) (hash(x)=51478185)
290
+ 276 train 8.868791 (lr=9.6853e-05) (hash(x)=49910644)
291
+ 277 train 8.736941 (lr=9.7203e-05) (hash(x)=48930577)
292
+ 278 train 8.783195 (lr=9.7552e-05) (hash(x)=49020396)
293
+ 279 train 8.900206 (lr=9.7902e-05) (hash(x)=53701929)
294
+ 280 train 8.834047 (lr=9.8252e-05) (hash(x)=47184278)
295
+ 281 train 8.963490 (lr=9.8601e-05) (hash(x)=54410011)
296
+ 282 train 8.867006 (lr=9.8951e-05) (hash(x)=54549283)
297
+ 283 train 8.840093 (lr=9.9301e-05) (hash(x)=51021250)
298
+ 284 train 8.762712 (lr=9.9650e-05) (hash(x)=50272010)
299
+ 285 train 8.757231 (lr=1.0000e-04) (hash(x)=55817179)
300
+ 286 train 8.750861 (lr=1.0000e-04) (hash(x)=47596588)
301
+ 287 train 8.712130 (lr=1.0000e-04) (hash(x)=46930369)
302
+ 288 train 8.686503 (lr=9.9998e-05) (hash(x)=48874951)
303
+ 289 train 8.728017 (lr=9.9996e-05) (hash(x)=51147542)
304
+ 290 train 8.626772 (lr=9.9993e-05) (hash(x)=52125021)
305
+ 291 train 8.464854 (lr=9.9989e-05) (hash(x)=39405431)
306
+ 292 train 8.644427 (lr=9.9984e-05) (hash(x)=47289107)
307
+ 293 train 8.615517 (lr=9.9979e-05) (hash(x)=48536998)
308
+ 294 train 8.858381 (lr=9.9972e-05) (hash(x)=62523011)
309
+ 295 train 8.839559 (lr=9.9965e-05) (hash(x)=57476018)
310
+ 296 train 8.603618 (lr=9.9956e-05) (hash(x)=54119125)
311
+ 297 train 8.817293 (lr=9.9947e-05) (hash(x)=58950633)
312
+ 298 train 8.615675 (lr=9.9937e-05) (hash(x)=53212705)
313
+ 299 train 8.563722 (lr=9.9926e-05) (hash(x)=48315677)
314
+ 300 val loss 8.5515
315
+ 300 val perplexity 5174.6353
316
+ 300 train 8.566295 (lr=9.9915e-05) (hash(x)=48622485)
317
+ 301 train 8.572163 (lr=9.9902e-05) (hash(x)=53677175)
318
+ 302 train 8.572646 (lr=9.9889e-05) (hash(x)=55167389)
319
+ 303 train 8.577269 (lr=9.9874e-05) (hash(x)=50199187)
320
+ 304 train 8.640165 (lr=9.9859e-05) (hash(x)=52896102)
321
+ 305 train 8.492141 (lr=9.9843e-05) (hash(x)=56596899)
322
+ 306 train 8.540364 (lr=9.9826e-05) (hash(x)=53062758)
323
+ 307 train 8.555486 (lr=9.9808e-05) (hash(x)=49586111)
324
+ 308 train 8.496590 (lr=9.9789e-05) (hash(x)=52369544)
325
+ 309 train 8.484304 (lr=9.9770e-05) (hash(x)=46819915)
326
+ 310 train 8.344594 (lr=9.9749e-05) (hash(x)=49293953)
327
+ 311 train 8.601306 (lr=9.9728e-05) (hash(x)=45949699)
328
+ 312 train 8.603287 (lr=9.9706e-05) (hash(x)=53515595)
329
+ 313 train 8.436904 (lr=9.9683e-05) (hash(x)=45515416)
330
+ 314 train 8.550052 (lr=9.9659e-05) (hash(x)=49455078)
331
+ 315 train 8.902791 (lr=9.9634e-05) (hash(x)=47334388)
332
+ 316 train 8.836808 (lr=9.9609e-05) (hash(x)=48654816)
333
+ 317 train 8.576250 (lr=9.9582e-05) (hash(x)=49474562)
334
+ 318 train 8.512582 (lr=9.9555e-05) (hash(x)=47371125)
335
+ 319 train 8.280191 (lr=9.9526e-05) (hash(x)=53294055)
336
+ 320 train 8.217039 (lr=9.9497e-05) (hash(x)=51679677)
337
+ 321 train 8.350601 (lr=9.9467e-05) (hash(x)=54082362)
338
+ 322 train 8.387424 (lr=9.9437e-05) (hash(x)=53670230)
339
+ 323 train 8.226344 (lr=9.9405e-05) (hash(x)=46499362)
340
+ 324 train 8.453465 (lr=9.9372e-05) (hash(x)=51843454)
341
+ 325 train 8.405087 (lr=9.9339e-05) (hash(x)=53531448)
342
+ 326 train 8.334021 (lr=9.9305e-05) (hash(x)=55684813)
343
+ 327 train 8.355392 (lr=9.9270e-05) (hash(x)=53097392)
344
+ 328 train 8.012247 (lr=9.9234e-05) (hash(x)=45277852)
345
+ 329 train 7.994800 (lr=9.9197e-05) (hash(x)=41633073)
346
+ 330 train 8.149121 (lr=9.9159e-05) (hash(x)=42914820)
347
+ 331 train 8.243296 (lr=9.9121e-05) (hash(x)=46479397)
348
+ 332 train 8.152773 (lr=9.9081e-05) (hash(x)=44391371)
349
+ 333 train 8.088024 (lr=9.9041e-05) (hash(x)=38977328)
350
+ 334 train 8.221487 (lr=9.9000e-05) (hash(x)=50758211)
351
+ 335 train 8.354858 (lr=9.8958e-05) (hash(x)=54479595)
352
+ 336 train 8.432354 (lr=9.8915e-05) (hash(x)=51411646)
353
+ 337 train 8.345534 (lr=9.8872e-05) (hash(x)=52103949)
354
+ 338 train 8.290290 (lr=9.8827e-05) (hash(x)=50550907)
355
+ 339 train 8.189072 (lr=9.8782e-05) (hash(x)=47738093)
356
+ 340 train 8.466667 (lr=9.8736e-05) (hash(x)=56998616)
357
+ 341 train 8.214229 (lr=9.8689e-05) (hash(x)=51395731)
358
+ 342 train 8.312585 (lr=9.8641e-05) (hash(x)=61530180)
359
+ 343 train 8.477888 (lr=9.8592e-05) (hash(x)=59006693)
360
+ 344 train 8.147644 (lr=9.8543e-05) (hash(x)=48594367)
361
+ 345 train 8.186057 (lr=9.8492e-05) (hash(x)=52545488)
362
+ 346 train 8.145518 (lr=9.8441e-05) (hash(x)=47376565)
363
+ 347 train 8.193451 (lr=9.8389e-05) (hash(x)=46312947)
364
+ 348 train 8.179934 (lr=9.8336e-05) (hash(x)=49325729)
365
+ 349 train 8.204646 (lr=9.8282e-05) (hash(x)=51488924)
366
+ 350 val loss 8.1946
367
+ 350 val perplexity 3621.5222
368
+ 350 train 8.189511 (lr=9.8228e-05) (hash(x)=52194507)
369
+ 351 train 8.151454 (lr=9.8172e-05) (hash(x)=51396394)
370
+ 352 train 8.327136 (lr=9.8116e-05) (hash(x)=47730521)
371
+ 353 train 8.080447 (lr=9.8059e-05) (hash(x)=43705721)
372
+ 354 train 8.130113 (lr=9.8001e-05) (hash(x)=47078328)
373
+ 355 train 8.350009 (lr=9.7942e-05) (hash(x)=55830040)
374
+ 356 train 8.318402 (lr=9.7882e-05) (hash(x)=48559738)
375
+ 357 train 8.149508 (lr=9.7822e-05) (hash(x)=49340705)
376
+ 358 train 8.217707 (lr=9.7761e-05) (hash(x)=49635452)
377
+ 359 train 8.237566 (lr=9.7699e-05) (hash(x)=52098425)
378
+ 360 train 8.104035 (lr=9.7636e-05) (hash(x)=48852030)
379
+ 361 train 8.177330 (lr=9.7572e-05) (hash(x)=48643399)
380
+ 362 train 8.330093 (lr=9.7507e-05) (hash(x)=58138953)
381
+ 363 train 7.890167 (lr=9.7442e-05) (hash(x)=44163258)
382
+ 364 train 8.220734 (lr=9.7376e-05) (hash(x)=55032638)
383
+ 365 train 8.114673 (lr=9.7309e-05) (hash(x)=48803235)
384
+ 366 train 7.981875 (lr=9.7241e-05) (hash(x)=48930309)
385
+ 367 train 8.317338 (lr=9.7172e-05) (hash(x)=58459317)
386
+ 368 train 8.213313 (lr=9.7103e-05) (hash(x)=52969670)
387
+ 369 train 8.141819 (lr=9.7032e-05) (hash(x)=55950871)
388
+ 370 train 8.142927 (lr=9.6961e-05) (hash(x)=50190558)
389
+ 371 train 8.202598 (lr=9.6889e-05) (hash(x)=50402996)
390
+ 372 train 8.089105 (lr=9.6817e-05) (hash(x)=48331790)
391
+ 373 train 8.092799 (lr=9.6743e-05) (hash(x)=51861896)
392
+ 374 train 8.085764 (lr=9.6669e-05) (hash(x)=52442780)
393
+ 375 train 8.085851 (lr=9.6593e-05) (hash(x)=47781307)
394
+ 376 train 7.830320 (lr=9.6518e-05) (hash(x)=45305570)
395
+ 377 train 8.512580 (lr=9.6441e-05) (hash(x)=58591982)
396
+ 378 train 8.199300 (lr=9.6363e-05) (hash(x)=53348894)
397
+ 379 train 8.301464 (lr=9.6285e-05) (hash(x)=55269973)
398
+ 380 train 8.130864 (lr=9.6206e-05) (hash(x)=52427717)
399
+ 381 train 8.285108 (lr=9.6126e-05) (hash(x)=54250578)
400
+ 382 train 8.201990 (lr=9.6045e-05) (hash(x)=55042333)
401
+ 383 train 8.247171 (lr=9.5963e-05) (hash(x)=55088766)
402
+ 384 train 8.036704 (lr=9.5881e-05) (hash(x)=51982665)
403
+ 385 train 8.022803 (lr=9.5798e-05) (hash(x)=50679914)
404
+ 386 train 8.110034 (lr=9.5714e-05) (hash(x)=55557228)
405
+ 387 train 8.131469 (lr=9.5629e-05) (hash(x)=48450358)
406
+ 388 train 7.961640 (lr=9.5544e-05) (hash(x)=49288522)
407
+ 389 train 8.153949 (lr=9.5457e-05) (hash(x)=49197883)
408
+ 390 train 8.142331 (lr=9.5370e-05) (hash(x)=54555602)
409
+ 391 train 8.018296 (lr=9.5282e-05) (hash(x)=50198680)
410
+ 392 train 8.097953 (lr=9.5194e-05) (hash(x)=54766944)
411
+ 393 train 8.075603 (lr=9.5104e-05) (hash(x)=51419665)
412
+ 394 train 7.658637 (lr=9.5014e-05) (hash(x)=43612177)
413
+ 395 train 8.002937 (lr=9.4923e-05) (hash(x)=48650332)
414
+ 396 train 8.057268 (lr=9.4831e-05) (hash(x)=53120879)
415
+ 397 train 7.949635 (lr=9.4739e-05) (hash(x)=48143827)
416
+ 398 train 7.954962 (lr=9.4646e-05) (hash(x)=46766862)
417
+ 399 train 8.057114 (lr=9.4551e-05) (hash(x)=52234688)
418
+ 400 val loss 7.9971
419
+ 400 val perplexity 2972.3240
420
+ 400 train 7.951595 (lr=9.4457e-05) (hash(x)=49099804)
421
+ 401 train 8.059956 (lr=9.4361e-05) (hash(x)=52032234)
422
+ 402 train 8.054568 (lr=9.4265e-05) (hash(x)=51775957)
423
+ 403 train 8.040034 (lr=9.4168e-05) (hash(x)=51346801)
424
+ 404 train 8.035719 (lr=9.4070e-05) (hash(x)=49392856)
425
+ 405 train 8.301320 (lr=9.3971e-05) (hash(x)=58253034)
426
+ 406 train 8.202806 (lr=9.3872e-05) (hash(x)=54043298)
427
+ 407 train 8.059733 (lr=9.3772e-05) (hash(x)=52929403)
428
+ 408 train 7.977715 (lr=9.3671e-05) (hash(x)=52901232)
429
+ 409 train 8.140979 (lr=9.3569e-05) (hash(x)=56975127)
430
+ 410 train 7.990594 (lr=9.3467e-05) (hash(x)=50603607)
431
+ 411 train 8.080699 (lr=9.3364e-05) (hash(x)=56070928)
432
+ 412 train 7.955007 (lr=9.3260e-05) (hash(x)=44443158)
433
+ 413 train 7.943355 (lr=9.3155e-05) (hash(x)=48921492)
434
+ 414 train 8.007022 (lr=9.3050e-05) (hash(x)=50946739)
435
+ 415 train 7.885238 (lr=9.2944e-05) (hash(x)=50526384)
436
+ 416 train 8.065828 (lr=9.2837e-05) (hash(x)=56037628)
437
+ 417 train 7.870002 (lr=9.2729e-05) (hash(x)=49178590)
438
+ 418 train 7.947971 (lr=9.2621e-05) (hash(x)=50164285)
439
+ 419 train 8.126010 (lr=9.2512e-05) (hash(x)=60209461)
440
+ 420 train 8.120618 (lr=9.2402e-05) (hash(x)=51355917)
441
+ 421 train 7.991400 (lr=9.2292e-05) (hash(x)=51970769)
442
+ 422 train 7.949481 (lr=9.2181e-05) (hash(x)=51896253)
443
+ 423 train 8.023345 (lr=9.2069e-05) (hash(x)=52322092)
444
+ 424 train 7.702393 (lr=9.1956e-05) (hash(x)=46560171)
445
+ 425 train 7.887349 (lr=9.1843e-05) (hash(x)=49806109)
446
+ 426 train 8.024118 (lr=9.1729e-05) (hash(x)=52507118)
447
+ 427 train 8.060771 (lr=9.1614e-05) (hash(x)=52916383)
448
+ 428 train 7.824626 (lr=9.1499e-05) (hash(x)=46519875)
449
+ 429 train 7.761441 (lr=9.1382e-05) (hash(x)=43459378)
450
+ 430 train 7.984489 (lr=9.1266e-05) (hash(x)=54438747)
451
+ 431 train 7.753470 (lr=9.1148e-05) (hash(x)=52529702)
452
+ 432 train 7.916400 (lr=9.1030e-05) (hash(x)=52506037)
453
+ 433 train 7.960931 (lr=9.0911e-05) (hash(x)=49688097)
454
+ 434 train 7.975182 (lr=9.0791e-05) (hash(x)=52187692)
455
+ 435 train 7.992792 (lr=9.0671e-05) (hash(x)=51980724)
456
+ 436 train 7.967807 (lr=9.0550e-05) (hash(x)=57522870)
457
+ 437 train 7.876134 (lr=9.0428e-05) (hash(x)=47435088)
458
+ 438 train 8.030954 (lr=9.0306e-05) (hash(x)=54021184)
459
+ 439 train 7.778221 (lr=9.0182e-05) (hash(x)=46826784)
460
+ 440 train 7.986296 (lr=9.0059e-05) (hash(x)=46764832)
461
+ 441 train 7.911371 (lr=8.9934e-05) (hash(x)=51120264)
462
+ 442 train 7.893799 (lr=8.9809e-05) (hash(x)=52654128)
463
+ 443 train 7.919742 (lr=8.9683e-05) (hash(x)=51321857)
464
+ 444 train 7.837077 (lr=8.9557e-05) (hash(x)=45311446)
465
+ 445 train 7.747219 (lr=8.9430e-05) (hash(x)=51476600)
466
+ 446 train 7.974338 (lr=8.9302e-05) (hash(x)=51269835)
467
+ 447 train 7.903169 (lr=8.9173e-05) (hash(x)=53475454)
468
+ 448 train 7.767452 (lr=8.9044e-05) (hash(x)=46470991)
469
+ 449 train 7.763066 (lr=8.8914e-05) (hash(x)=39878629)
470
+ 450 val loss 7.8599
471
+ 450 val perplexity 2591.3057
472
+ 450 train 7.851364 (lr=8.8784e-05) (hash(x)=50326173)
473
+ 451 train 7.896492 (lr=8.8653e-05) (hash(x)=54787418)
474
+ 452 train 8.018271 (lr=8.8521e-05) (hash(x)=48949948)
475
+ 453 train 8.042957 (lr=8.8388e-05) (hash(x)=52506608)
476
+ 454 train 7.887018 (lr=8.8255e-05) (hash(x)=54323120)
477
+ 455 train 7.804399 (lr=8.8122e-05) (hash(x)=46810389)
478
+ 456 train 7.668405 (lr=8.7987e-05) (hash(x)=45396873)
479
+ 457 train 7.960999 (lr=8.7852e-05) (hash(x)=57144694)
480
+ 458 train 7.789538 (lr=8.7717e-05) (hash(x)=49878429)
481
+ 459 train 7.910660 (lr=8.7580e-05) (hash(x)=53498382)
482
+ 460 train 7.761792 (lr=8.7444e-05) (hash(x)=47417814)
483
+ 461 train 7.900574 (lr=8.7306e-05) (hash(x)=51280415)
484
+ 462 train 7.866807 (lr=8.7168e-05) (hash(x)=54075295)
485
+ 463 train 7.708292 (lr=8.7029e-05) (hash(x)=46323632)
486
+ 464 train 7.808490 (lr=8.6890e-05) (hash(x)=50211782)
487
+ 465 train 7.771979 (lr=8.6750e-05) (hash(x)=46779238)
488
+ 466 train 7.874959 (lr=8.6609e-05) (hash(x)=48313734)
489
+ 467 train 7.826059 (lr=8.6468e-05) (hash(x)=52398798)
490
+ 468 train 7.816012 (lr=8.6326e-05) (hash(x)=50640616)
491
+ 469 train 7.846576 (lr=8.6184e-05) (hash(x)=48186608)
492
+ 470 train 7.871347 (lr=8.6041e-05) (hash(x)=51158107)
493
+ 471 train 7.749991 (lr=8.5897e-05) (hash(x)=53724068)
494
+ 472 train 7.767381 (lr=8.5753e-05) (hash(x)=52537855)
495
+ 473 train 7.857733 (lr=8.5608e-05) (hash(x)=52960197)
496
+ 474 train 7.925149 (lr=8.5462e-05) (hash(x)=54156942)
497
+ 475 train 8.058942 (lr=8.5316e-05) (hash(x)=52768582)
498
+ 476 train 7.712003 (lr=8.5170e-05) (hash(x)=50479600)
499
+ 477 train 7.811925 (lr=8.5022e-05) (hash(x)=52810796)
500
+ 478 train 7.945230 (lr=8.4875e-05) (hash(x)=49731354)
501
+ 479 train 7.721236 (lr=8.4726e-05) (hash(x)=46747022)
502
+ 480 train 7.748267 (lr=8.4577e-05) (hash(x)=49495779)
503
+ 481 train 7.858725 (lr=8.4428e-05) (hash(x)=53760168)
504
+ 482 train 7.845321 (lr=8.4278e-05) (hash(x)=51460514)
505
+ 483 train 7.846571 (lr=8.4127e-05) (hash(x)=50353293)
506
+ 484 train 7.811156 (lr=8.3976e-05) (hash(x)=53396658)
507
+ 485 train 7.902220 (lr=8.3824e-05) (hash(x)=50573284)
508
+ 486 train 7.683843 (lr=8.3672e-05) (hash(x)=49168443)
509
+ 487 train 7.881670 (lr=8.3519e-05) (hash(x)=50822751)
510
+ 488 train 7.855075 (lr=8.3366e-05) (hash(x)=45669857)
511
+ 489 train 7.678635 (lr=8.3212e-05) (hash(x)=44593201)
512
+ 490 train 7.447061 (lr=8.3057e-05) (hash(x)=39734742)
513
+ 491 train 7.778190 (lr=8.2902e-05) (hash(x)=47616170)
514
+ 492 train 7.759972 (lr=8.2746e-05) (hash(x)=46558148)
515
+ 493 train 7.761379 (lr=8.2590e-05) (hash(x)=49711034)
516
+ 494 train 7.910164 (lr=8.2434e-05) (hash(x)=55210283)
517
+ 495 train 7.790407 (lr=8.2276e-05) (hash(x)=48013779)
518
+ 496 train 7.831361 (lr=8.2119e-05) (hash(x)=50644422)
519
+ 497 train 7.860594 (lr=8.1960e-05) (hash(x)=49696467)
520
+ 498 train 7.851898 (lr=8.1801e-05) (hash(x)=43246287)
521
+ 499 train 7.720505 (lr=8.1642e-05) (hash(x)=45598720)
522
+ 500 val loss 7.7905
523
+ 500 val perplexity 2417.5437
524
+ 500 train 7.776520 (lr=8.1482e-05) (hash(x)=51062334)
525
+ 501 train 7.685298 (lr=8.1322e-05) (hash(x)=49421968)
526
+ 502 train 7.828423 (lr=8.1161e-05) (hash(x)=52123859)
527
+ 503 train 7.816712 (lr=8.1000e-05) (hash(x)=52658193)
528
+ 504 train 7.766056 (lr=8.0838e-05) (hash(x)=51023513)
529
+ 505 train 8.006373 (lr=8.0676e-05) (hash(x)=51581514)
530
+ 506 train 7.915479 (lr=8.0513e-05) (hash(x)=49089133)
531
+ 507 train 7.761624 (lr=8.0349e-05) (hash(x)=48531900)
532
+ 508 train 7.873342 (lr=8.0186e-05) (hash(x)=52693475)
533
+ 509 train 7.816837 (lr=8.0021e-05) (hash(x)=54903424)
534
+ 510 train 7.857077 (lr=7.9856e-05) (hash(x)=51036262)
535
+ 511 train 7.720460 (lr=7.9691e-05) (hash(x)=46134720)
536
+ 512 train 7.958547 (lr=7.9525e-05) (hash(x)=53031267)
537
+ 513 train 7.952726 (lr=7.9359e-05) (hash(x)=53833396)
538
+ 514 train 7.813370 (lr=7.9192e-05) (hash(x)=55247788)
539
+ 515 train 7.835004 (lr=7.9025e-05) (hash(x)=51156418)
540
+ 516 train 7.728214 (lr=7.8858e-05) (hash(x)=51468309)
541
+ 517 train 7.647869 (lr=7.8689e-05) (hash(x)=48618191)
542
+ 518 train 7.962184 (lr=7.8521e-05) (hash(x)=53297837)
543
+ 519 train 7.777779 (lr=7.8352e-05) (hash(x)=52902481)
544
+ 520 train 7.772885 (lr=7.8182e-05) (hash(x)=52699185)
545
+ 521 train 7.809433 (lr=7.8012e-05) (hash(x)=54117145)
546
+ 522 train 7.957872 (lr=7.7842e-05) (hash(x)=53596095)
547
+ 523 train 7.796145 (lr=7.7671e-05) (hash(x)=51604647)
548
+ 524 train 7.770963 (lr=7.7500e-05) (hash(x)=52145920)
549
+ 525 train 7.740326 (lr=7.7328e-05) (hash(x)=50691387)
550
+ 526 train 7.859173 (lr=7.7156e-05) (hash(x)=52429064)
551
+ 527 train 7.718513 (lr=7.6984e-05) (hash(x)=49163526)
552
+ 528 train 7.892435 (lr=7.6811e-05) (hash(x)=53184723)
553
+ 529 train 7.676492 (lr=7.6637e-05) (hash(x)=47039068)
554
+ 530 train 7.914508 (lr=7.6463e-05) (hash(x)=47552672)
555
+ 531 train 7.751335 (lr=7.6289e-05) (hash(x)=53829392)
556
+ 532 train 7.620055 (lr=7.6115e-05) (hash(x)=47800177)
557
+ 533 train 7.633024 (lr=7.5940e-05) (hash(x)=49800706)
558
+ 534 train 7.937207 (lr=7.5764e-05) (hash(x)=55575943)
559
+ 535 train 7.735637 (lr=7.5588e-05) (hash(x)=49028078)
560
+ 536 train 7.896054 (lr=7.5412e-05) (hash(x)=54755546)
561
+ 537 train 7.843447 (lr=7.5235e-05) (hash(x)=53372658)
562
+ 538 train 7.799034 (lr=7.5058e-05) (hash(x)=52268865)
563
+ 539 train 7.791011 (lr=7.4881e-05) (hash(x)=52517672)
564
+ 540 train 7.785433 (lr=7.4703e-05) (hash(x)=50244085)
565
+ 541 train 7.663471 (lr=7.4525e-05) (hash(x)=49248218)
566
+ 542 train 7.841891 (lr=7.4346e-05) (hash(x)=49762688)
567
+ 543 train 7.892095 (lr=7.4167e-05) (hash(x)=59110263)
568
+ 544 train 7.653158 (lr=7.3988e-05) (hash(x)=47478150)
569
+ 545 train 7.773090 (lr=7.3808e-05) (hash(x)=54533121)
570
+ 546 train 7.703787 (lr=7.3628e-05) (hash(x)=49901224)
571
+ 547 train 7.821720 (lr=7.3448e-05) (hash(x)=51195255)
572
+ 548 train 7.761472 (lr=7.3267e-05) (hash(x)=49109169)
573
+ 549 train 7.713020 (lr=7.3086e-05) (hash(x)=47467588)
574
+ 550 val loss 7.7450
575
+ 550 val perplexity 2309.9668
576
+ 550 train 7.725455 (lr=7.2904e-05) (hash(x)=46269857)
577
+ 551 train 7.712228 (lr=7.2723e-05) (hash(x)=50517780)
578
+ 552 train 7.813435 (lr=7.2540e-05) (hash(x)=50183795)
579
+ 553 train 7.763023 (lr=7.2358e-05) (hash(x)=50640606)
580
+ 554 train 7.772913 (lr=7.2175e-05) (hash(x)=51738364)
581
+ 555 train 8.473052 (lr=7.1992e-05) (hash(x)=62341175)
582
+ 556 train 7.849590 (lr=7.1808e-05) (hash(x)=45189486)
583
+ 557 train 7.786495 (lr=7.1624e-05) (hash(x)=51670137)
584
+ 558 train 7.586957 (lr=7.1440e-05) (hash(x)=46152565)
585
+ 559 train 7.569258 (lr=7.1256e-05) (hash(x)=45890874)
586
+ 560 train 7.822829 (lr=7.1071e-05) (hash(x)=50865200)
587
+ 561 train 7.657534 (lr=7.0886e-05) (hash(x)=50817384)
588
+ 562 train 7.622824 (lr=7.0701e-05) (hash(x)=45716461)
589
+ 563 train 7.764955 (lr=7.0515e-05) (hash(x)=49188067)
590
+ 564 train 7.815670 (lr=7.0329e-05) (hash(x)=51104117)
591
+ 565 train 7.813921 (lr=7.0143e-05) (hash(x)=53943162)
592
+ 566 train 7.481145 (lr=6.9956e-05) (hash(x)=42808589)
593
+ 567 train 7.726647 (lr=6.9769e-05) (hash(x)=47888783)
594
+ 568 train 7.606959 (lr=6.9582e-05) (hash(x)=46060882)
595
+ 569 train 7.796121 (lr=6.9394e-05) (hash(x)=47770108)
596
+ 570 train 7.951019 (lr=6.9207e-05) (hash(x)=53150465)
597
+ 571 train 8.140530 (lr=6.9019e-05) (hash(x)=52775090)
598
+ 572 train 7.867815 (lr=6.8830e-05) (hash(x)=49924177)
599
+ 573 train 7.661129 (lr=6.8642e-05) (hash(x)=49971730)
600
+ 574 train 7.700123 (lr=6.8453e-05) (hash(x)=47952698)
601
+ 575 train 7.988646 (lr=6.8264e-05) (hash(x)=56693767)
602
+ 576 train 7.508914 (lr=6.8075e-05) (hash(x)=43516921)
603
+ 577 train 7.890891 (lr=6.7885e-05) (hash(x)=53347077)
604
+ 578 train 7.683504 (lr=6.7695e-05) (hash(x)=47410931)
605
+ 579 train 7.654891 (lr=6.7505e-05) (hash(x)=45014400)
606
+ 580 train 7.772875 (lr=6.7315e-05) (hash(x)=54677976)
607
+ 581 train 7.546669 (lr=6.7124e-05) (hash(x)=45398141)
608
+ 582 train 7.596748 (lr=6.6933e-05) (hash(x)=47181969)
609
+ 583 train 7.962359 (lr=6.6742e-05) (hash(x)=53102856)
610
+ 584 train 7.640815 (lr=6.6551e-05) (hash(x)=48780879)
611
+ 585 train 7.755972 (lr=6.6360e-05) (hash(x)=51707711)
612
+ 586 train 7.800344 (lr=6.6168e-05) (hash(x)=48191112)
613
+ 587 train 7.670135 (lr=6.5976e-05) (hash(x)=52954310)
614
+ 588 train 7.958051 (lr=6.5784e-05) (hash(x)=58807144)
615
+ 589 train 7.560582 (lr=6.5592e-05) (hash(x)=45943270)
616
+ 590 train 7.745490 (lr=6.5399e-05) (hash(x)=49191483)
617
+ 591 train 7.649088 (lr=6.5206e-05) (hash(x)=47965042)
618
+ 592 train 7.769397 (lr=6.5013e-05) (hash(x)=51074739)
619
+ 593 train 7.826696 (lr=6.4820e-05) (hash(x)=56913043)
620
+ 594 train 8.040998 (lr=6.4627e-05) (hash(x)=62500105)
621
+ 595 train 7.833272 (lr=6.4433e-05) (hash(x)=56016526)
622
+ 596 train 7.639240 (lr=6.4240e-05) (hash(x)=46762756)
623
+ 597 train 7.755949 (lr=6.4046e-05) (hash(x)=50655822)
624
+ 598 train 7.718080 (lr=6.3852e-05) (hash(x)=50751670)
625
+ 599 train 7.648071 (lr=6.3658e-05) (hash(x)=45600303)
626
+ 600 val loss 7.7189
627
+ 600 val perplexity 2250.4822
628
+ 600 train 7.785832 (lr=6.3463e-05) (hash(x)=54607172)
629
+ 601 train 7.719733 (lr=6.3269e-05) (hash(x)=48710188)
630
+ 602 train 7.606162 (lr=6.3074e-05) (hash(x)=46206284)
631
+ 603 train 7.592145 (lr=6.2879e-05) (hash(x)=45764012)
632
+ 604 train 7.626038 (lr=6.2684e-05) (hash(x)=50393806)
633
+ 605 train 7.719620 (lr=6.2489e-05) (hash(x)=52034772)
634
+ 606 train 7.679569 (lr=6.2294e-05) (hash(x)=48705579)
635
+ 607 train 7.489473 (lr=6.2098e-05) (hash(x)=44542942)
636
+ 608 train 7.637166 (lr=6.1903e-05) (hash(x)=48781316)
637
+ 609 train 7.724702 (lr=6.1707e-05) (hash(x)=48042083)
638
+ 610 train 7.836951 (lr=6.1511e-05) (hash(x)=49658564)
639
+ 611 train 7.848541 (lr=6.1315e-05) (hash(x)=51672319)
640
+ 612 train 7.768613 (lr=6.1119e-05) (hash(x)=52431673)
641
+ 613 train 7.672872 (lr=6.0923e-05) (hash(x)=43542246)
642
+ 614 train 7.748245 (lr=6.0726e-05) (hash(x)=52666378)
643
+ 615 train 7.713056 (lr=6.0530e-05) (hash(x)=49315253)
644
+ 616 train 7.833696 (lr=6.0333e-05) (hash(x)=50722863)
645
+ 617 train 8.026576 (lr=6.0137e-05) (hash(x)=52341036)
646
+ 618 train 7.740505 (lr=5.9940e-05) (hash(x)=48814169)
647
+ 619 train 7.596574 (lr=5.9743e-05) (hash(x)=45571167)
648
+ 620 train 7.632347 (lr=5.9546e-05) (hash(x)=47776763)
649
+ 621 train 7.675299 (lr=5.9349e-05) (hash(x)=48382543)
650
+ 622 train 7.636768 (lr=5.9152e-05) (hash(x)=45514205)
651
+ 623 train 7.542202 (lr=5.8955e-05) (hash(x)=45320225)
652
+ 624 train 7.676067 (lr=5.8758e-05) (hash(x)=48452761)
653
+ 625 train 7.832915 (lr=5.8560e-05) (hash(x)=49116111)
654
+ 626 train 7.560065 (lr=5.8363e-05) (hash(x)=47302076)
655
+ 627 train 7.735370 (lr=5.8165e-05) (hash(x)=50184363)
656
+ 628 train 7.612583 (lr=5.7968e-05) (hash(x)=49385175)
657
+ 629 train 7.753914 (lr=5.7770e-05) (hash(x)=42655518)
658
+ 630 train 7.719224 (lr=5.7573e-05) (hash(x)=45181685)
659
+ 631 train 7.507550 (lr=5.7375e-05) (hash(x)=45775751)
660
+ 632 train 7.745899 (lr=5.7177e-05) (hash(x)=51697844)
661
+ 633 train 7.629016 (lr=5.6979e-05) (hash(x)=47776033)
662
+ 634 train 7.836730 (lr=5.6782e-05) (hash(x)=54185547)
663
+ 635 train 7.604667 (lr=5.6584e-05) (hash(x)=42462763)
664
+ 636 train 7.692198 (lr=5.6386e-05) (hash(x)=49909171)
665
+ 637 train 7.747829 (lr=5.6188e-05) (hash(x)=50472276)
666
+ 638 train 7.911973 (lr=5.5990e-05) (hash(x)=53342806)
667
+ 639 train 7.808676 (lr=5.5792e-05) (hash(x)=53985957)
668
+ 640 train 7.677028 (lr=5.5594e-05) (hash(x)=49621437)
669
+ 641 train 7.811014 (lr=5.5396e-05) (hash(x)=50215750)
670
+ 642 train 7.594242 (lr=5.5198e-05) (hash(x)=45707324)
671
+ 643 train 7.957663 (lr=5.5000e-05) (hash(x)=55567006)
672
+ 644 train 8.148246 (lr=5.4802e-05) (hash(x)=63587960)
673
+ 645 train 7.777864 (lr=5.4604e-05) (hash(x)=50830962)
674
+ 646 train 7.938086 (lr=5.4406e-05) (hash(x)=50228793)
675
+ 647 train 7.488471 (lr=5.4208e-05) (hash(x)=41963014)
676
+ 648 train 7.669334 (lr=5.4010e-05) (hash(x)=52083811)
677
+ 649 train 8.175434 (lr=5.3812e-05) (hash(x)=55253867)
678
+ 650 val loss 7.7039
679
+ 650 val perplexity 2216.9443
680
+ 650 train 7.821275 (lr=5.3614e-05) (hash(x)=56596815)
681
+ 651 train 7.931311 (lr=5.3416e-05) (hash(x)=53862673)
682
+ 652 train 7.821924 (lr=5.3218e-05) (hash(x)=51032349)
683
+ 653 train 7.935736 (lr=5.3021e-05) (hash(x)=57095498)
684
+ 654 train 8.067757 (lr=5.2823e-05) (hash(x)=57840576)
685
+ 655 train 7.908059 (lr=5.2625e-05) (hash(x)=56042923)
686
+ 656 train 7.704258 (lr=5.2427e-05) (hash(x)=53170061)
687
+ 657 train 7.742249 (lr=5.2230e-05) (hash(x)=51651483)
688
+ 658 train 7.862979 (lr=5.2032e-05) (hash(x)=52052987)
689
+ 659 train 7.469182 (lr=5.1835e-05) (hash(x)=43078023)
690
+ 660 train 7.780211 (lr=5.1637e-05) (hash(x)=53974411)
691
+ 661 train 7.842247 (lr=5.1440e-05) (hash(x)=55335538)
692
+ 662 train 8.362656 (lr=5.1242e-05) (hash(x)=60627374)
693
+ 663 train 8.043871 (lr=5.1045e-05) (hash(x)=55023379)
694
+ 664 train 7.755903 (lr=5.0848e-05) (hash(x)=53081545)
695
+ 665 train 7.859747 (lr=5.0651e-05) (hash(x)=48246811)
696
+ 666 train 7.866852 (lr=5.0454e-05) (hash(x)=48563603)
697
+ 667 train 7.484613 (lr=5.0257e-05) (hash(x)=43011460)
698
+ 668 train 7.682555 (lr=5.0060e-05) (hash(x)=48368934)
699
+ 669 train 7.665777 (lr=4.9863e-05) (hash(x)=51322872)
700
+ 670 train 7.659311 (lr=4.9667e-05) (hash(x)=49196654)
701
+ 671 train 7.817737 (lr=4.9470e-05) (hash(x)=52536929)
702
+ 672 train 8.029195 (lr=4.9274e-05) (hash(x)=57300798)
703
+ 673 train 7.532197 (lr=4.9077e-05) (hash(x)=46105549)
704
+ 674 train 7.548130 (lr=4.8881e-05) (hash(x)=51612608)
705
+ 675 train 7.795131 (lr=4.8685e-05) (hash(x)=49761494)
706
+ 676 train 7.587813 (lr=4.8489e-05) (hash(x)=45375531)
707
+ 677 train 7.983675 (lr=4.8293e-05) (hash(x)=61065657)
708
+ 678 train 7.563208 (lr=4.8097e-05) (hash(x)=44152741)
709
+ 679 train 7.636041 (lr=4.7902e-05) (hash(x)=51631508)
710
+ 680 train 7.513665 (lr=4.7706e-05) (hash(x)=43922234)
711
+ 681 train 7.684584 (lr=4.7511e-05) (hash(x)=49689666)
712
+ 682 train 7.876114 (lr=4.7316e-05) (hash(x)=53751279)
713
+ 683 train 7.757041 (lr=4.7121e-05) (hash(x)=50400339)
714
+ 684 train 7.732320 (lr=4.6926e-05) (hash(x)=51029698)
715
+ 685 train 7.771144 (lr=4.6731e-05) (hash(x)=54240120)
716
+ 686 train 7.763153 (lr=4.6537e-05) (hash(x)=45708714)
717
+ 687 train 7.589183 (lr=4.6342e-05) (hash(x)=46560264)
718
+ 688 train 7.591962 (lr=4.6148e-05) (hash(x)=46841822)
719
+ 689 train 7.632158 (lr=4.5954e-05) (hash(x)=48013446)
720
+ 690 train 7.773297 (lr=4.5760e-05) (hash(x)=54564252)
721
+ 691 train 8.262938 (lr=4.5567e-05) (hash(x)=70534997)
722
+ 692 train 7.798925 (lr=4.5373e-05) (hash(x)=53363496)
723
+ 693 train 7.695941 (lr=4.5180e-05) (hash(x)=49508415)
724
+ 694 train 7.711966 (lr=4.4987e-05) (hash(x)=53503261)
725
+ 695 train 7.765871 (lr=4.4794e-05) (hash(x)=52420475)
726
+ 696 train 7.643469 (lr=4.4601e-05) (hash(x)=50191708)
727
+ 697 train 8.033561 (lr=4.4408e-05) (hash(x)=60279802)
728
+ 698 train 7.711820 (lr=4.4216e-05) (hash(x)=52626271)
729
+ 699 train 7.790769 (lr=4.4024e-05) (hash(x)=53194976)
730
+ 700 val loss 7.6937
731
+ 700 val perplexity 2194.4797
732
+ 700 train 7.805597 (lr=4.3832e-05) (hash(x)=50906515)
733
+ 701 train 7.580327 (lr=4.3640e-05) (hash(x)=49459146)
734
+ 702 train 7.631964 (lr=4.3449e-05) (hash(x)=48320631)
735
+ 703 train 8.062301 (lr=4.3258e-05) (hash(x)=66781899)
736
+ 704 train 8.115291 (lr=4.3067e-05) (hash(x)=57109514)
737
+ 705 train 8.128964 (lr=4.2876e-05) (hash(x)=43442176)
738
+ 706 train 7.828445 (lr=4.2685e-05) (hash(x)=51431734)
739
+ 707 train 7.701361 (lr=4.2495e-05) (hash(x)=51986059)
740
+ 708 train 7.852036 (lr=4.2305e-05) (hash(x)=54279714)
741
+ 709 train 7.653358 (lr=4.2115e-05) (hash(x)=45517898)
742
+ 710 train 8.244445 (lr=4.1925e-05) (hash(x)=57689755)
743
+ 711 train 8.000155 (lr=4.1736e-05) (hash(x)=60275167)
744
+ 712 train 7.635049 (lr=4.1547e-05) (hash(x)=46698776)
745
+ 713 train 8.121480 (lr=4.1358e-05) (hash(x)=61199882)
746
+ 714 train 7.641793 (lr=4.1170e-05) (hash(x)=48256229)
747
+ 715 train 7.614938 (lr=4.0981e-05) (hash(x)=45483557)
748
+ 716 train 7.738168 (lr=4.0793e-05) (hash(x)=50062280)
749
+ 717 train 8.045675 (lr=4.0606e-05) (hash(x)=49436293)
750
+ 718 train 7.717941 (lr=4.0418e-05) (hash(x)=47916352)
751
+ 719 train 7.660004 (lr=4.0231e-05) (hash(x)=49182582)
752
+ 720 train 7.746307 (lr=4.0044e-05) (hash(x)=50054568)
753
+ 721 train 7.864446 (lr=3.9857e-05) (hash(x)=54622543)
754
+ 722 train 7.758401 (lr=3.9671e-05) (hash(x)=53761778)
755
+ 723 train 7.638814 (lr=3.9485e-05) (hash(x)=51446091)
756
+ 724 train 7.691039 (lr=3.9299e-05) (hash(x)=50809582)
757
+ 725 train 7.547813 (lr=3.9114e-05) (hash(x)=45019935)
758
+ 726 train 7.567287 (lr=3.8929e-05) (hash(x)=44061718)
759
+ 727 train 7.590410 (lr=3.8744e-05) (hash(x)=47827767)
760
+ 728 train 7.799716 (lr=3.8560e-05) (hash(x)=51054368)
761
+ 729 train 7.533257 (lr=3.8376e-05) (hash(x)=47297652)
762
+ 730 train 7.590787 (lr=3.8192e-05) (hash(x)=45486271)
763
+ 731 train 7.737047 (lr=3.8008e-05) (hash(x)=46988810)
764
+ 732 train 7.491913 (lr=3.7825e-05) (hash(x)=44578285)
765
+ 733 train 7.539136 (lr=3.7642e-05) (hash(x)=44457482)
766
+ 734 train 7.655244 (lr=3.7460e-05) (hash(x)=47273057)
767
+ 735 train 7.454480 (lr=3.7277e-05) (hash(x)=44918130)
768
+ 736 train 7.691038 (lr=3.7096e-05) (hash(x)=46825217)
769
+ 737 train 7.715501 (lr=3.6914e-05) (hash(x)=47962618)
770
+ 738 train 7.624187 (lr=3.6733e-05) (hash(x)=45539511)
771
+ 739 train 7.734653 (lr=3.6552e-05) (hash(x)=49693485)
772
+ 740 train 7.696205 (lr=3.6372e-05) (hash(x)=49483023)
773
+ 741 train 7.555534 (lr=3.6192e-05) (hash(x)=46911489)
774
+ 742 train 7.703938 (lr=3.6012e-05) (hash(x)=52016723)
775
+ 743 train 7.429154 (lr=3.5833e-05) (hash(x)=40969008)
776
+ 744 train 7.560429 (lr=3.5654e-05) (hash(x)=48694211)
777
+ 745 train 7.567380 (lr=3.5475e-05) (hash(x)=47877745)
778
+ 746 train 7.738072 (lr=3.5297e-05) (hash(x)=53156899)
779
+ 747 train 7.777902 (lr=3.5119e-05) (hash(x)=50344401)
780
+ 748 train 7.770007 (lr=3.4942e-05) (hash(x)=46614361)
781
+ 749 train 7.726585 (lr=3.4765e-05) (hash(x)=53074032)
782
+ 750 val loss 7.6766
783
+ 750 val perplexity 2157.2366
784
+ 750 train 7.647449 (lr=3.4588e-05) (hash(x)=46046498)
785
+ 751 train 7.636253 (lr=3.4412e-05) (hash(x)=49833636)
786
+ 752 train 7.651859 (lr=3.4236e-05) (hash(x)=49165393)
787
+ 753 train 8.025078 (lr=3.4060e-05) (hash(x)=61445802)
788
+ 754 train 7.844114 (lr=3.3885e-05) (hash(x)=51012661)
789
+ 755 train 7.386491 (lr=3.3711e-05) (hash(x)=41326666)
790
+ 756 train 7.493817 (lr=3.3537e-05) (hash(x)=47169271)
791
+ 757 train 7.501117 (lr=3.3363e-05) (hash(x)=43474195)
792
+ 758 train 7.513590 (lr=3.3189e-05) (hash(x)=41683297)
793
+ 759 train 7.659019 (lr=3.3016e-05) (hash(x)=52346553)
794
+ 760 train 7.412308 (lr=3.2844e-05) (hash(x)=44385560)
795
+ 761 train 7.081166 (lr=3.2672e-05) (hash(x)=35732133)
796
+ 762 train 7.485312 (lr=3.2500e-05) (hash(x)=41423670)
797
+ 763 train 7.679850 (lr=3.2329e-05) (hash(x)=47970930)
798
+ 764 train 7.561603 (lr=3.2158e-05) (hash(x)=47092804)
799
+ 765 train 7.461565 (lr=3.1988e-05) (hash(x)=47182916)
800
+ 766 train 7.480068 (lr=3.1818e-05) (hash(x)=41791614)
801
+ 767 train 7.656696 (lr=3.1648e-05) (hash(x)=47009621)
802
+ 768 train 7.816598 (lr=3.1479e-05) (hash(x)=51633434)
803
+ 769 train 7.603348 (lr=3.1311e-05) (hash(x)=52003882)
804
+ 770 train 7.758319 (lr=3.1142e-05) (hash(x)=51340089)
805
+ 771 train 7.582633 (lr=3.0975e-05) (hash(x)=48602644)
806
+ 772 train 7.766180 (lr=3.0808e-05) (hash(x)=50457646)
807
+ 773 train 7.629657 (lr=3.0641e-05) (hash(x)=49970789)
808
+ 774 train 7.701898 (lr=3.0475e-05) (hash(x)=51611455)
809
+ 775 train 7.664772 (lr=3.0309e-05) (hash(x)=52115732)
810
+ 776 train 7.527327 (lr=3.0144e-05) (hash(x)=49032269)
811
+ 777 train 7.715684 (lr=2.9979e-05) (hash(x)=50715961)
812
+ 778 train 7.631848 (lr=2.9814e-05) (hash(x)=49693375)
813
+ 779 train 8.003014 (lr=2.9651e-05) (hash(x)=56527845)
814
+ 780 train 7.534776 (lr=2.9487e-05) (hash(x)=48381855)
815
+ 781 train 7.616484 (lr=2.9324e-05) (hash(x)=49623946)
816
+ 782 train 7.464725 (lr=2.9162e-05) (hash(x)=47393528)
817
+ 783 train 7.707607 (lr=2.9000e-05) (hash(x)=49459868)
818
+ 784 train 7.620192 (lr=2.8839e-05) (hash(x)=48696657)
819
+ 785 train 7.687817 (lr=2.8678e-05) (hash(x)=52550306)
820
+ 786 train 7.697450 (lr=2.8518e-05) (hash(x)=49285282)
821
+ 787 train 7.695724 (lr=2.8358e-05) (hash(x)=48955151)
822
+ 788 train 7.569065 (lr=2.8199e-05) (hash(x)=43743600)
823
+ 789 train 7.741786 (lr=2.8040e-05) (hash(x)=49655015)
824
+ 790 train 7.620924 (lr=2.7881e-05) (hash(x)=51090440)
825
+ 791 train 7.577444 (lr=2.7724e-05) (hash(x)=48473450)
826
+ 792 train 7.795185 (lr=2.7566e-05) (hash(x)=47938317)
827
+ 793 train 7.793788 (lr=2.7410e-05) (hash(x)=45290633)
828
+ 794 train 7.506030 (lr=2.7254e-05) (hash(x)=48542673)
829
+ 795 train 7.695592 (lr=2.7098e-05) (hash(x)=50306722)
830
+ 796 train 7.616338 (lr=2.6943e-05) (hash(x)=50632201)
831
+ 797 train 7.582848 (lr=2.6788e-05) (hash(x)=48795732)
832
+ 798 train 7.610859 (lr=2.6634e-05) (hash(x)=50141432)
833
+ 799 train 7.886966 (lr=2.6481e-05) (hash(x)=50898037)
834
+ 800 val loss 7.6653
835
+ 800 val perplexity 2133.0374
836
+ 800 train 7.681482 (lr=2.6328e-05) (hash(x)=37427531)
837
+ 801 train 7.657777 (lr=2.6176e-05) (hash(x)=32890763)
838
+ 802 train 7.709007 (lr=2.6024e-05) (hash(x)=31307609)
839
+ 803 train 7.523466 (lr=2.5873e-05) (hash(x)=35015807)
840
+ 804 train 7.672347 (lr=2.5722e-05) (hash(x)=37542227)
841
+ 805 train 7.743680 (lr=2.5572e-05) (hash(x)=41169444)
842
+ 806 train 7.584801 (lr=2.5423e-05) (hash(x)=47354793)
843
+ 807 train 7.652699 (lr=2.5274e-05) (hash(x)=46327297)
844
+ 808 train 7.647086 (lr=2.5125e-05) (hash(x)=50124976)
845
+ 809 train 7.538122 (lr=2.4978e-05) (hash(x)=48536826)
846
+ 810 train 7.490990 (lr=2.4830e-05) (hash(x)=43048468)
847
+ 811 train 7.574680 (lr=2.4684e-05) (hash(x)=44297138)
848
+ 812 train 8.610119 (lr=2.4538e-05) (hash(x)=46976056)
849
+ 813 train 7.821051 (lr=2.4392e-05) (hash(x)=48533607)
850
+ 814 train 7.409503 (lr=2.4247e-05) (hash(x)=45581560)
851
+ 815 train 7.445683 (lr=2.4103e-05) (hash(x)=47776613)
852
+ 816 train 7.366590 (lr=2.3959e-05) (hash(x)=39451311)
853
+ 817 train 7.950234 (lr=2.3816e-05) (hash(x)=53854198)
854
+ 818 train 7.668997 (lr=2.3674e-05) (hash(x)=47110254)
855
+ 819 train 7.699428 (lr=2.3532e-05) (hash(x)=51262732)
856
+ 820 train 7.658064 (lr=2.3391e-05) (hash(x)=48675819)
857
+ 821 train 7.712848 (lr=2.3250e-05) (hash(x)=47159040)
858
+ 822 train 7.783525 (lr=2.3110e-05) (hash(x)=52808063)
859
+ 823 train 7.686244 (lr=2.2971e-05) (hash(x)=51480608)
860
+ 824 train 7.454303 (lr=2.2832e-05) (hash(x)=46075661)
861
+ 825 train 7.562265 (lr=2.2694e-05) (hash(x)=44193147)
862
+ 826 train 7.609120 (lr=2.2556e-05) (hash(x)=51387298)
863
+ 827 train 7.710419 (lr=2.2420e-05) (hash(x)=50929448)
864
+ 828 train 7.796356 (lr=2.2283e-05) (hash(x)=54393657)
865
+ 829 train 7.507691 (lr=2.2148e-05) (hash(x)=48420333)
866
+ 830 train 7.806491 (lr=2.2013e-05) (hash(x)=54428349)
867
+ 831 train 7.688338 (lr=2.1878e-05) (hash(x)=46494961)
868
+ 832 train 7.567462 (lr=2.1745e-05) (hash(x)=50441352)
869
+ 833 train 7.891117 (lr=2.1612e-05) (hash(x)=49410253)
870
+ 834 train 7.831018 (lr=2.1479e-05) (hash(x)=54977154)
871
+ 835 train 7.828668 (lr=2.1347e-05) (hash(x)=59634178)
872
+ 836 train 7.856194 (lr=2.1216e-05) (hash(x)=51875554)
873
+ 837 train 7.634380 (lr=2.1086e-05) (hash(x)=48922609)
874
+ 838 train 7.543328 (lr=2.0956e-05) (hash(x)=46512842)
875
+ 839 train 7.565852 (lr=2.0827e-05) (hash(x)=47517395)
876
+ 840 train 7.672166 (lr=2.0698e-05) (hash(x)=51480081)
877
+ 841 train 7.642575 (lr=2.0570e-05) (hash(x)=49531327)
878
+ 842 train 7.936496 (lr=2.0443e-05) (hash(x)=50654362)
879
+ 843 train 8.087562 (lr=2.0317e-05) (hash(x)=48371589)
880
+ 844 train 7.942711 (lr=2.0191e-05) (hash(x)=57169652)
881
+ 845 train 8.035479 (lr=2.0066e-05) (hash(x)=55313671)
882
+ 846 train 7.589671 (lr=1.9941e-05) (hash(x)=45244232)
883
+ 847 train 7.926842 (lr=1.9818e-05) (hash(x)=61097106)
884
+ 848 train 7.639060 (lr=1.9694e-05) (hash(x)=51926581)
885
+ 849 train 7.646481 (lr=1.9572e-05) (hash(x)=52305923)
886
+ 850 val loss 7.6650
887
+ 850 val perplexity 2132.3948
888
+ 850 train 7.628397 (lr=1.9450e-05) (hash(x)=48839174)
889
+ 851 train 7.959109 (lr=1.9329e-05) (hash(x)=61405427)
890
+ 852 train 7.951845 (lr=1.9209e-05) (hash(x)=61444306)
891
+ 853 train 7.837207 (lr=1.9089e-05) (hash(x)=59898062)
892
+ 854 train 7.823995 (lr=1.8970e-05) (hash(x)=54795546)
893
+ 855 train 7.662750 (lr=1.8852e-05) (hash(x)=50139108)
894
+ 856 train 7.584324 (lr=1.8734e-05) (hash(x)=48975659)
895
+ 857 train 7.651093 (lr=1.8618e-05) (hash(x)=47955919)
896
+ 858 train 7.591760 (lr=1.8501e-05) (hash(x)=48949357)
897
+ 859 train 7.480790 (lr=1.8386e-05) (hash(x)=45467872)
898
+ 860 train 7.412431 (lr=1.8271e-05) (hash(x)=42449398)
899
+ 861 train 7.912760 (lr=1.8157e-05) (hash(x)=59686276)
900
+ 862 train 7.725308 (lr=1.8044e-05) (hash(x)=51946811)
901
+ 863 train 7.723258 (lr=1.7931e-05) (hash(x)=49686665)
902
+ 864 train 8.136472 (lr=1.7819e-05) (hash(x)=65282589)
903
+ 865 train 7.541974 (lr=1.7708e-05) (hash(x)=47341207)
904
+ 866 train 8.048573 (lr=1.7598e-05) (hash(x)=63814513)
905
+ 867 train 7.529201 (lr=1.7488e-05) (hash(x)=51019411)
906
+ 868 train 7.398736 (lr=1.7379e-05) (hash(x)=46653745)
907
+ 869 train 7.545042 (lr=1.7271e-05) (hash(x)=50050759)
908
+ 870 train 7.842685 (lr=1.7163e-05) (hash(x)=53878448)
909
+ 871 train 7.618465 (lr=1.7056e-05) (hash(x)=49380702)
910
+ 872 train 7.734842 (lr=1.6950e-05) (hash(x)=50922302)
911
+ 873 train 7.616608 (lr=1.6845e-05) (hash(x)=50963362)
912
+ 874 train 7.792284 (lr=1.6740e-05) (hash(x)=52460631)
913
+ 875 train 7.616323 (lr=1.6636e-05) (hash(x)=50156382)
914
+ 876 train 7.805536 (lr=1.6533e-05) (hash(x)=44475078)
915
+ 877 train 7.881967 (lr=1.6431e-05) (hash(x)=49995857)
916
+ 878 train 8.070649 (lr=1.6329e-05) (hash(x)=57032860)
917
+ 879 train 7.761130 (lr=1.6228e-05) (hash(x)=48775872)
918
+ 880 train 7.594008 (lr=1.6128e-05) (hash(x)=47643742)
919
+ 881 train 7.719578 (lr=1.6029e-05) (hash(x)=53183314)
920
+ 882 train 7.693049 (lr=1.5930e-05) (hash(x)=50268180)
921
+ 883 train 7.628825 (lr=1.5832e-05) (hash(x)=49497141)
922
+ 884 train 7.452913 (lr=1.5735e-05) (hash(x)=43420753)
923
+ 885 train 7.726367 (lr=1.5639e-05) (hash(x)=51207084)
924
+ 886 train 7.242729 (lr=1.5543e-05) (hash(x)=34492530)
925
+ 887 train 7.606430 (lr=1.5449e-05) (hash(x)=49456326)
926
+ 888 train 8.003901 (lr=1.5354e-05) (hash(x)=50766672)
927
+ 889 train 7.872369 (lr=1.5261e-05) (hash(x)=49811760)
928
+ 890 train 7.611205 (lr=1.5169e-05) (hash(x)=48142033)
929
+ 891 train 7.733113 (lr=1.5077e-05) (hash(x)=53343959)
930
+ 892 train 7.514139 (lr=1.4986e-05) (hash(x)=46916209)
931
+ 893 train 7.809450 (lr=1.4896e-05) (hash(x)=54858850)
932
+ 894 train 7.726833 (lr=1.4806e-05) (hash(x)=51700707)
933
+ 895 train 7.555516 (lr=1.4718e-05) (hash(x)=44654425)
934
+ 896 train 7.741243 (lr=1.4630e-05) (hash(x)=52740006)
935
+ 897 train 7.537305 (lr=1.4543e-05) (hash(x)=45260452)
936
+ 898 train 7.575133 (lr=1.4456e-05) (hash(x)=47151929)
937
+ 899 train 7.753841 (lr=1.4371e-05) (hash(x)=51367775)
938
+ 900 val loss 7.6536
939
+ 900 val perplexity 2108.3196
940
+ 900 train 7.665564 (lr=1.4286e-05) (hash(x)=50325207)
941
+ 901 train 7.589430 (lr=1.4202e-05) (hash(x)=46562023)
942
+ 902 train 7.706729 (lr=1.4119e-05) (hash(x)=48332912)
943
+ 903 train 7.709047 (lr=1.4037e-05) (hash(x)=54477325)
944
+ 904 train 7.744818 (lr=1.3955e-05) (hash(x)=49389150)
945
+ 905 train 7.622040 (lr=1.3874e-05) (hash(x)=48571772)
946
+ 906 train 7.794766 (lr=1.3794e-05) (hash(x)=53218072)
947
+ 907 train 7.561443 (lr=1.3715e-05) (hash(x)=45888568)
948
+ 908 train 7.688207 (lr=1.3637e-05) (hash(x)=50792965)
949
+ 909 train 7.720834 (lr=1.3559e-05) (hash(x)=49219751)
950
+ 910 train 7.605235 (lr=1.3482e-05) (hash(x)=49516452)
951
+ 911 train 7.579093 (lr=1.3407e-05) (hash(x)=49384161)
952
+ 912 train 7.551650 (lr=1.3331e-05) (hash(x)=45403134)
953
+ 913 train 7.582731 (lr=1.3257e-05) (hash(x)=50565032)
954
+ 914 train 7.772027 (lr=1.3183e-05) (hash(x)=51152823)
955
+ 915 train 7.826147 (lr=1.3111e-05) (hash(x)=52839038)
956
+ 916 train 7.447977 (lr=1.3039e-05) (hash(x)=44409465)
957
+ 917 train 7.846397 (lr=1.2968e-05) (hash(x)=51799409)
958
+ 918 train 7.598882 (lr=1.2897e-05) (hash(x)=51152758)
959
+ 919 train 7.645781 (lr=1.2828e-05) (hash(x)=49188683)
960
+ 920 train 7.606798 (lr=1.2759e-05) (hash(x)=50284940)
961
+ 921 train 7.674716 (lr=1.2691e-05) (hash(x)=51812773)
962
+ 922 train 7.755337 (lr=1.2624e-05) (hash(x)=53642545)
963
+ 923 train 7.938771 (lr=1.2558e-05) (hash(x)=62547001)
964
+ 924 train 7.671631 (lr=1.2493e-05) (hash(x)=47242049)
965
+ 925 train 7.651288 (lr=1.2428e-05) (hash(x)=49563491)
966
+ 926 train 7.703235 (lr=1.2364e-05) (hash(x)=50947272)
967
+ 927 train 7.788035 (lr=1.2301e-05) (hash(x)=51229664)
968
+ 928 train 7.544437 (lr=1.2239e-05) (hash(x)=47237060)
969
+ 929 train 7.521131 (lr=1.2178e-05) (hash(x)=46252668)
970
+ 930 train 7.597550 (lr=1.2118e-05) (hash(x)=43781107)
971
+ 931 train 7.531198 (lr=1.2058e-05) (hash(x)=48746317)
972
+ 932 train 8.224815 (lr=1.1999e-05) (hash(x)=67646788)
973
+ 933 train 7.406954 (lr=1.1941e-05) (hash(x)=44204094)
974
+ 934 train 7.618474 (lr=1.1884e-05) (hash(x)=50776803)
975
+ 935 train 7.811872 (lr=1.1828e-05) (hash(x)=51961195)
976
+ 936 train 7.929801 (lr=1.1772e-05) (hash(x)=48943726)
977
+ 937 train 7.853168 (lr=1.1718e-05) (hash(x)=47683918)
978
+ 938 train 8.004213 (lr=1.1664e-05) (hash(x)=50300577)
979
+ 939 train 7.882885 (lr=1.1611e-05) (hash(x)=49881442)
980
+ 940 train 7.720823 (lr=1.1559e-05) (hash(x)=46953117)
981
+ 941 train 7.899009 (lr=1.1508e-05) (hash(x)=45500608)
982
+ 942 train 8.026901 (lr=1.1457e-05) (hash(x)=56801743)
983
+ 943 train 7.568245 (lr=1.1408e-05) (hash(x)=48726061)
984
+ 944 train 7.585130 (lr=1.1359e-05) (hash(x)=46878089)
985
+ 945 train 7.702103 (lr=1.1311e-05) (hash(x)=53235305)
986
+ 946 train 7.651208 (lr=1.1264e-05) (hash(x)=50160069)
987
+ 947 train 7.831829 (lr=1.1218e-05) (hash(x)=50492385)
988
+ 948 train 7.567551 (lr=1.1173e-05) (hash(x)=49291969)
989
+ 949 train 7.513515 (lr=1.1128e-05) (hash(x)=46951052)
990
+ 950 val loss 7.6523
991
+ 950 val perplexity 2105.5437
992
+ 950 train 7.531526 (lr=1.1085e-05) (hash(x)=46580257)
993
+ 951 train 7.766010 (lr=1.1042e-05) (hash(x)=53617083)
994
+ 952 train 7.650432 (lr=1.1000e-05) (hash(x)=50964186)
995
+ 953 train 7.652650 (lr=1.0959e-05) (hash(x)=50267127)
996
+ 954 train 7.485906 (lr=1.0919e-05) (hash(x)=43627333)
997
+ 955 train 7.504456 (lr=1.0879e-05) (hash(x)=45972461)
998
+ 956 train 7.498619 (lr=1.0841e-05) (hash(x)=45851643)
999
+ 957 train 7.888245 (lr=1.0803e-05) (hash(x)=55325922)
1000
+ 958 train 8.352912 (lr=1.0766e-05) (hash(x)=61621353)
1001
+ 959 train 8.347492 (lr=1.0730e-05) (hash(x)=60830807)
1002
+ 960 train 8.300445 (lr=1.0695e-05) (hash(x)=59780537)
1003
+ 961 train 8.310517 (lr=1.0661e-05) (hash(x)=59357720)
1004
+ 962 train 8.256862 (lr=1.0628e-05) (hash(x)=60305936)
1005
+ 963 train 8.318660 (lr=1.0595e-05) (hash(x)=63145389)
1006
+ 964 train 8.271323 (lr=1.0563e-05) (hash(x)=62011447)
1007
+ 965 train 8.293187 (lr=1.0533e-05) (hash(x)=60412496)
1008
+ 966 train 8.327641 (lr=1.0503e-05) (hash(x)=62740258)
1009
+ 967 train 8.379521 (lr=1.0474e-05) (hash(x)=64013010)
1010
+ 968 train 8.286026 (lr=1.0445e-05) (hash(x)=59176481)
1011
+ 969 train 8.492831 (lr=1.0418e-05) (hash(x)=62177701)
1012
+ 970 train 8.278195 (lr=1.0391e-05) (hash(x)=60908204)
1013
+ 971 train 8.438822 (lr=1.0366e-05) (hash(x)=64418102)
1014
+ 972 train 8.351231 (lr=1.0341e-05) (hash(x)=60282281)
1015
+ 973 train 7.745143 (lr=1.0317e-05) (hash(x)=52915348)
1016
+ 974 train 7.715331 (lr=1.0294e-05) (hash(x)=49995009)
1017
+ 975 train 7.838100 (lr=1.0272e-05) (hash(x)=54461658)
1018
+ 976 train 8.158129 (lr=1.0251e-05) (hash(x)=63990903)
1019
+ 977 train 7.735990 (lr=1.0230e-05) (hash(x)=47729553)
1020
+ 978 train 7.666760 (lr=1.0211e-05) (hash(x)=50140658)
1021
+ 979 train 7.533794 (lr=1.0192e-05) (hash(x)=49407427)
1022
+ 980 train 7.718092 (lr=1.0174e-05) (hash(x)=51411433)
1023
+ 981 train 7.635004 (lr=1.0157e-05) (hash(x)=51432179)
1024
+ 982 train 7.631053 (lr=1.0141e-05) (hash(x)=51394518)
1025
+ 983 train 7.568357 (lr=1.0126e-05) (hash(x)=48800368)
1026
+ 984 train 7.825583 (lr=1.0111e-05) (hash(x)=51807114)
1027
+ 985 train 7.569704 (lr=1.0098e-05) (hash(x)=45866136)
1028
+ 986 train 7.580410 (lr=1.0085e-05) (hash(x)=45069575)
1029
+ 987 train 7.705280 (lr=1.0074e-05) (hash(x)=54376909)
1030
+ 988 train 7.793777 (lr=1.0063e-05) (hash(x)=59343246)
1031
+ 989 train 7.719244 (lr=1.0053e-05) (hash(x)=53519805)
1032
+ 990 train 7.710960 (lr=1.0044e-05) (hash(x)=49346697)
1033
+ 991 train 7.769209 (lr=1.0035e-05) (hash(x)=49838333)
1034
+ 992 train 7.512253 (lr=1.0028e-05) (hash(x)=45831325)
1035
+ 993 train 7.582121 (lr=1.0021e-05) (hash(x)=48643653)
1036
+ 994 train 7.698822 (lr=1.0016e-05) (hash(x)=54516111)
1037
+ 995 train 10.790346 (lr=1.0011e-05) (hash(x)=111351979)
1038
+ 996 train 11.813456 (lr=1.0007e-05) (hash(x)=127687623)
1039
+ 997 train 11.844747 (lr=1.0004e-05) (hash(x)=132049261)
1040
+ 998 train 12.362039 (lr=1.0002e-05) (hash(x)=141228471)
1041
+ 999 val loss 7.6654
1042
+ 999 val perplexity 2133.1575
1043
+ 999 train 11.282540 (lr=1.0000e-05) (hash(x)=119446440)
lr1e-4_total_batch_size10240_seq_len1024/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4040ebc702753545998cc2933abf3bf10519567bdc0e65eafa9f31feda4fd051
3
+ size 86167106
lr1e-4_total_batch_size10240_seq_len1024/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:796fa6c28ee0a6900bec19d856652a1736383510f5ef01edbbb38a5680cc6de4
3
+ size 71681862