andrew-healey commited on
Commit
76bb476
·
verified ·
1 Parent(s): 2aa01dc

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size20480_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size20480_seq_len512", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 512, "batch_size": null, "total_batch_size": 20480, "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_size20480_seq_len512/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8513131d2bf732f8f6d97ff82d32c505b2a2f3f9f2247bac315bb09ce0e1f97
3
+ size 964
lr1e-4_total_batch_size20480_seq_len512/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7800
3
+ 0 val perplexity 130608.6406
4
+ 0 train 11.774407 (lr=3.4965e-07) (hash(x)=27234506)
5
+ 1 train 11.756070 (lr=6.9930e-07) (hash(x)=20782690)
6
+ 2 train 11.788530 (lr=1.0490e-06) (hash(x)=26532095)
7
+ 3 train 11.767059 (lr=1.3986e-06) (hash(x)=23890908)
8
+ 4 train 11.770683 (lr=1.7483e-06) (hash(x)=24011372)
9
+ 5 train 11.763443 (lr=2.0979e-06) (hash(x)=24611628)
10
+ 6 train 11.748604 (lr=2.4476e-06) (hash(x)=19245352)
11
+ 7 train 11.735285 (lr=2.7972e-06) (hash(x)=25767553)
12
+ 8 train 11.743330 (lr=3.1469e-06) (hash(x)=24809873)
13
+ 9 train 11.748739 (lr=3.4965e-06) (hash(x)=22512166)
14
+ 10 train 11.726679 (lr=3.8462e-06) (hash(x)=28092148)
15
+ 11 train 11.724897 (lr=4.1958e-06) (hash(x)=24732833)
16
+ 12 train 11.712705 (lr=4.5455e-06) (hash(x)=22694718)
17
+ 13 train 11.718848 (lr=4.8951e-06) (hash(x)=25129410)
18
+ 14 train 11.696350 (lr=5.2448e-06) (hash(x)=27858570)
19
+ 15 train 11.697028 (lr=5.5944e-06) (hash(x)=26582391)
20
+ 16 train 11.680421 (lr=5.9441e-06) (hash(x)=27279806)
21
+ 17 train 11.666603 (lr=6.2937e-06) (hash(x)=26472336)
22
+ 18 train 11.681172 (lr=6.6434e-06) (hash(x)=28301589)
23
+ 19 train 11.645747 (lr=6.9930e-06) (hash(x)=27654289)
24
+ 20 train 11.640273 (lr=7.3427e-06) (hash(x)=24995715)
25
+ 21 train 11.622377 (lr=7.6923e-06) (hash(x)=26559876)
26
+ 22 train 11.614866 (lr=8.0420e-06) (hash(x)=28364895)
27
+ 23 train 11.610596 (lr=8.3916e-06) (hash(x)=21770329)
28
+ 24 train 11.582096 (lr=8.7413e-06) (hash(x)=24628606)
29
+ 25 train 11.576576 (lr=9.0909e-06) (hash(x)=22723459)
30
+ 26 train 11.537670 (lr=9.4406e-06) (hash(x)=22897919)
31
+ 27 train 11.565722 (lr=9.7902e-06) (hash(x)=30338342)
32
+ 28 train 11.518240 (lr=1.0140e-05) (hash(x)=27063280)
33
+ 29 train 11.489194 (lr=1.0490e-05) (hash(x)=25678059)
34
+ 30 train 11.464165 (lr=1.0839e-05) (hash(x)=21853028)
35
+ 31 train 11.473089 (lr=1.1189e-05) (hash(x)=26109335)
36
+ 32 train 11.471902 (lr=1.1538e-05) (hash(x)=25334848)
37
+ 33 train 11.419803 (lr=1.1888e-05) (hash(x)=25044885)
38
+ 34 train 11.420362 (lr=1.2238e-05) (hash(x)=23678349)
39
+ 35 train 11.401827 (lr=1.2587e-05) (hash(x)=28870690)
40
+ 36 train 11.379728 (lr=1.2937e-05) (hash(x)=32982615)
41
+ 37 train 11.377997 (lr=1.3287e-05) (hash(x)=35279941)
42
+ 38 train 11.340857 (lr=1.3636e-05) (hash(x)=26400041)
43
+ 39 train 11.273310 (lr=1.3986e-05) (hash(x)=25015923)
44
+ 40 train 11.238981 (lr=1.4336e-05) (hash(x)=24853586)
45
+ 41 train 11.229668 (lr=1.4685e-05) (hash(x)=26740855)
46
+ 42 train 11.195984 (lr=1.5035e-05) (hash(x)=24584116)
47
+ 43 train 11.176802 (lr=1.5385e-05) (hash(x)=22262151)
48
+ 44 train 11.120602 (lr=1.5734e-05) (hash(x)=26627860)
49
+ 45 train 11.087852 (lr=1.6084e-05) (hash(x)=26621834)
50
+ 46 train 11.053520 (lr=1.6434e-05) (hash(x)=26930630)
51
+ 47 train 11.019162 (lr=1.6783e-05) (hash(x)=23234741)
52
+ 48 train 10.987256 (lr=1.7133e-05) (hash(x)=28102443)
53
+ 49 train 11.058925 (lr=1.7483e-05) (hash(x)=25470563)
54
+ 50 val loss 10.9568
55
+ 50 val perplexity 57343.2812
56
+ 50 train 10.951854 (lr=1.7832e-05) (hash(x)=28282861)
57
+ 51 train 10.940174 (lr=1.8182e-05) (hash(x)=21799102)
58
+ 52 train 10.872891 (lr=1.8531e-05) (hash(x)=23471769)
59
+ 53 train 10.837331 (lr=1.8881e-05) (hash(x)=23275479)
60
+ 54 train 10.848922 (lr=1.9231e-05) (hash(x)=24473581)
61
+ 55 train 10.831827 (lr=1.9580e-05) (hash(x)=22091862)
62
+ 56 train 10.725104 (lr=1.9930e-05) (hash(x)=26018202)
63
+ 57 train 10.713504 (lr=2.0280e-05) (hash(x)=17430373)
64
+ 58 train 10.597141 (lr=2.0629e-05) (hash(x)=24078083)
65
+ 59 train 10.670362 (lr=2.0979e-05) (hash(x)=22954861)
66
+ 60 train 10.698197 (lr=2.1329e-05) (hash(x)=28695016)
67
+ 61 train 10.634047 (lr=2.1678e-05) (hash(x)=26671799)
68
+ 62 train 10.668319 (lr=2.2028e-05) (hash(x)=23166092)
69
+ 63 train 10.536644 (lr=2.2378e-05) (hash(x)=26924723)
70
+ 64 train 10.597767 (lr=2.2727e-05) (hash(x)=25138659)
71
+ 65 train 10.585502 (lr=2.3077e-05) (hash(x)=23557495)
72
+ 66 train 10.513854 (lr=2.3427e-05) (hash(x)=25288123)
73
+ 67 train 10.345476 (lr=2.3776e-05) (hash(x)=27804380)
74
+ 68 train 10.538805 (lr=2.4126e-05) (hash(x)=26993263)
75
+ 69 train 10.559528 (lr=2.4476e-05) (hash(x)=26654908)
76
+ 70 train 10.442593 (lr=2.4825e-05) (hash(x)=27347722)
77
+ 71 train 10.509675 (lr=2.5175e-05) (hash(x)=23627518)
78
+ 72 train 10.459628 (lr=2.5524e-05) (hash(x)=21768671)
79
+ 73 train 10.445970 (lr=2.5874e-05) (hash(x)=21889990)
80
+ 74 train 10.371089 (lr=2.6224e-05) (hash(x)=23665838)
81
+ 75 train 10.584812 (lr=2.6573e-05) (hash(x)=22528001)
82
+ 76 train 10.457139 (lr=2.6923e-05) (hash(x)=22487328)
83
+ 77 train 10.499977 (lr=2.7273e-05) (hash(x)=29248942)
84
+ 78 train 10.403926 (lr=2.7622e-05) (hash(x)=23543273)
85
+ 79 train 10.396155 (lr=2.7972e-05) (hash(x)=25533417)
86
+ 80 train 10.364779 (lr=2.8322e-05) (hash(x)=25552036)
87
+ 81 train 10.338118 (lr=2.8671e-05) (hash(x)=20612533)
88
+ 82 train 10.301330 (lr=2.9021e-05) (hash(x)=20216795)
89
+ 83 train 10.425492 (lr=2.9371e-05) (hash(x)=24677740)
90
+ 84 train 10.362882 (lr=2.9720e-05) (hash(x)=16722715)
91
+ 85 train 10.390660 (lr=3.0070e-05) (hash(x)=25347203)
92
+ 86 train 10.357518 (lr=3.0420e-05) (hash(x)=24272489)
93
+ 87 train 10.313563 (lr=3.0769e-05) (hash(x)=28922363)
94
+ 88 train 10.296288 (lr=3.1119e-05) (hash(x)=25510798)
95
+ 89 train 10.352019 (lr=3.1469e-05) (hash(x)=25631397)
96
+ 90 train 10.265753 (lr=3.1818e-05) (hash(x)=21333676)
97
+ 91 train 10.362594 (lr=3.2168e-05) (hash(x)=26237983)
98
+ 92 train 10.321539 (lr=3.2517e-05) (hash(x)=25976502)
99
+ 93 train 10.332174 (lr=3.2867e-05) (hash(x)=24742645)
100
+ 94 train 10.357476 (lr=3.3217e-05) (hash(x)=27086333)
101
+ 95 train 10.337040 (lr=3.3566e-05) (hash(x)=26157660)
102
+ 96 train 10.332930 (lr=3.3916e-05) (hash(x)=23616370)
103
+ 97 train 10.138109 (lr=3.4266e-05) (hash(x)=22582169)
104
+ 98 train 10.223929 (lr=3.4615e-05) (hash(x)=26233189)
105
+ 99 train 10.295856 (lr=3.4965e-05) (hash(x)=26431960)
106
+ 100 val loss 10.3068
107
+ 100 val perplexity 29936.4863
108
+ 100 train 10.359409 (lr=3.5315e-05) (hash(x)=30707498)
109
+ 101 train 10.295092 (lr=3.5664e-05) (hash(x)=25716176)
110
+ 102 train 10.295593 (lr=3.6014e-05) (hash(x)=26256420)
111
+ 103 train 10.312172 (lr=3.6364e-05) (hash(x)=28019788)
112
+ 104 train 10.262430 (lr=3.6713e-05) (hash(x)=23510110)
113
+ 105 train 10.305017 (lr=3.7063e-05) (hash(x)=20382963)
114
+ 106 train 10.192451 (lr=3.7413e-05) (hash(x)=25339466)
115
+ 107 train 10.282358 (lr=3.7762e-05) (hash(x)=24489647)
116
+ 108 train 10.245831 (lr=3.8112e-05) (hash(x)=22927699)
117
+ 109 train 10.226080 (lr=3.8462e-05) (hash(x)=25128502)
118
+ 110 train 10.269227 (lr=3.8811e-05) (hash(x)=25808969)
119
+ 111 train 10.207922 (lr=3.9161e-05) (hash(x)=25231390)
120
+ 112 train 10.158957 (lr=3.9510e-05) (hash(x)=21154388)
121
+ 113 train 10.147013 (lr=3.9860e-05) (hash(x)=22708977)
122
+ 114 train 10.322182 (lr=4.0210e-05) (hash(x)=27258353)
123
+ 115 train 10.175892 (lr=4.0559e-05) (hash(x)=24364920)
124
+ 116 train 10.206132 (lr=4.0909e-05) (hash(x)=27243972)
125
+ 117 train 10.235865 (lr=4.1259e-05) (hash(x)=24932925)
126
+ 118 train 10.233398 (lr=4.1608e-05) (hash(x)=26371091)
127
+ 119 train 10.155094 (lr=4.1958e-05) (hash(x)=26913684)
128
+ 120 train 10.180317 (lr=4.2308e-05) (hash(x)=29239611)
129
+ 121 train 10.057978 (lr=4.2657e-05) (hash(x)=18017792)
130
+ 122 train 10.148741 (lr=4.3007e-05) (hash(x)=27574089)
131
+ 123 train 10.158352 (lr=4.3357e-05) (hash(x)=28613882)
132
+ 124 train 10.152716 (lr=4.3706e-05) (hash(x)=26981914)
133
+ 125 train 10.159129 (lr=4.4056e-05) (hash(x)=22652243)
134
+ 126 train 10.111862 (lr=4.4406e-05) (hash(x)=15446025)
135
+ 127 train 10.059793 (lr=4.4755e-05) (hash(x)=27194521)
136
+ 128 train 10.018322 (lr=4.5105e-05) (hash(x)=25264518)
137
+ 129 train 10.153021 (lr=4.5455e-05) (hash(x)=24936836)
138
+ 130 train 10.123075 (lr=4.5804e-05) (hash(x)=28197282)
139
+ 131 train 10.091504 (lr=4.6154e-05) (hash(x)=24765578)
140
+ 132 train 10.096424 (lr=4.6503e-05) (hash(x)=27448185)
141
+ 133 train 10.181149 (lr=4.6853e-05) (hash(x)=23107416)
142
+ 134 train 10.008262 (lr=4.7203e-05) (hash(x)=26755138)
143
+ 135 train 10.023855 (lr=4.7552e-05) (hash(x)=29314255)
144
+ 136 train 9.964910 (lr=4.7902e-05) (hash(x)=22112669)
145
+ 137 train 10.033711 (lr=4.8252e-05) (hash(x)=24013079)
146
+ 138 train 9.977947 (lr=4.8601e-05) (hash(x)=22537194)
147
+ 139 train 9.957504 (lr=4.8951e-05) (hash(x)=25322001)
148
+ 140 train 10.094177 (lr=4.9301e-05) (hash(x)=27028126)
149
+ 141 train 10.046268 (lr=4.9650e-05) (hash(x)=26690440)
150
+ 142 train 9.974261 (lr=5.0000e-05) (hash(x)=26967331)
151
+ 143 train 10.013650 (lr=5.0350e-05) (hash(x)=22598512)
152
+ 144 train 10.017159 (lr=5.0699e-05) (hash(x)=24697985)
153
+ 145 train 9.888652 (lr=5.1049e-05) (hash(x)=18473911)
154
+ 146 train 9.961899 (lr=5.1399e-05) (hash(x)=23172124)
155
+ 147 train 10.053072 (lr=5.1748e-05) (hash(x)=22322442)
156
+ 148 train 10.023013 (lr=5.2098e-05) (hash(x)=29659121)
157
+ 149 train 9.928902 (lr=5.2448e-05) (hash(x)=26324153)
158
+ 150 val loss 9.9480
159
+ 150 val perplexity 20909.6211
160
+ 150 train 9.916352 (lr=5.2797e-05) (hash(x)=26939970)
161
+ 151 train 9.927941 (lr=5.3147e-05) (hash(x)=23446058)
162
+ 152 train 9.953926 (lr=5.3497e-05) (hash(x)=29981801)
163
+ 153 train 9.980227 (lr=5.3846e-05) (hash(x)=24843741)
164
+ 154 train 9.962902 (lr=5.4196e-05) (hash(x)=22855363)
165
+ 155 train 9.858494 (lr=5.4545e-05) (hash(x)=23197102)
166
+ 156 train 9.880893 (lr=5.4895e-05) (hash(x)=25704919)
167
+ 157 train 10.102352 (lr=5.5245e-05) (hash(x)=23054940)
168
+ 158 train 10.063993 (lr=5.5594e-05) (hash(x)=24909904)
169
+ 159 train 9.748272 (lr=5.5944e-05) (hash(x)=27140757)
170
+ 160 train 9.775261 (lr=5.6294e-05) (hash(x)=25993110)
171
+ 161 train 9.759039 (lr=5.6643e-05) (hash(x)=25149419)
172
+ 162 train 9.859231 (lr=5.6993e-05) (hash(x)=27623643)
173
+ 163 train 9.816277 (lr=5.7343e-05) (hash(x)=24727764)
174
+ 164 train 9.583190 (lr=5.7692e-05) (hash(x)=23155773)
175
+ 165 train 9.752445 (lr=5.8042e-05) (hash(x)=21935931)
176
+ 166 train 9.736070 (lr=5.8392e-05) (hash(x)=20054917)
177
+ 167 train 9.784201 (lr=5.8741e-05) (hash(x)=27606073)
178
+ 168 train 9.830671 (lr=5.9091e-05) (hash(x)=26534487)
179
+ 169 train 9.736032 (lr=5.9441e-05) (hash(x)=22450341)
180
+ 170 train 9.770388 (lr=5.9790e-05) (hash(x)=25013073)
181
+ 171 train 9.799050 (lr=6.0140e-05) (hash(x)=25945859)
182
+ 172 train 9.695324 (lr=6.0490e-05) (hash(x)=24418091)
183
+ 173 train 9.687039 (lr=6.0839e-05) (hash(x)=21613707)
184
+ 174 train 9.679859 (lr=6.1189e-05) (hash(x)=26386157)
185
+ 175 train 9.705043 (lr=6.1538e-05) (hash(x)=24396519)
186
+ 176 train 9.651868 (lr=6.1888e-05) (hash(x)=23265507)
187
+ 177 train 9.701689 (lr=6.2238e-05) (hash(x)=29138828)
188
+ 178 train 9.735908 (lr=6.2587e-05) (hash(x)=24265353)
189
+ 179 train 9.729469 (lr=6.2937e-05) (hash(x)=24685515)
190
+ 180 train 9.636333 (lr=6.3287e-05) (hash(x)=24876269)
191
+ 181 train 9.565554 (lr=6.3636e-05) (hash(x)=22825749)
192
+ 182 train 9.626949 (lr=6.3986e-05) (hash(x)=21239652)
193
+ 183 train 9.625591 (lr=6.4336e-05) (hash(x)=28767869)
194
+ 184 train 9.589849 (lr=6.4685e-05) (hash(x)=25614301)
195
+ 185 train 9.590500 (lr=6.5035e-05) (hash(x)=21159060)
196
+ 186 train 9.574510 (lr=6.5385e-05) (hash(x)=26751788)
197
+ 187 train 9.523219 (lr=6.5734e-05) (hash(x)=25045397)
198
+ 188 train 9.614997 (lr=6.6084e-05) (hash(x)=29989709)
199
+ 189 train 9.642633 (lr=6.6434e-05) (hash(x)=25560928)
200
+ 190 train 9.615669 (lr=6.6783e-05) (hash(x)=26062687)
201
+ 191 train 9.531675 (lr=6.7133e-05) (hash(x)=28959222)
202
+ 192 train 9.450694 (lr=6.7483e-05) (hash(x)=25434884)
203
+ 193 train 9.529901 (lr=6.7832e-05) (hash(x)=21248405)
204
+ 194 train 9.460155 (lr=6.8182e-05) (hash(x)=25044192)
205
+ 195 train 9.504607 (lr=6.8531e-05) (hash(x)=25830182)
206
+ 196 train 9.476355 (lr=6.8881e-05) (hash(x)=26621419)
207
+ 197 train 9.332607 (lr=6.9231e-05) (hash(x)=25173113)
208
+ 198 train 9.451179 (lr=6.9580e-05) (hash(x)=24400133)
209
+ 199 train 9.434359 (lr=6.9930e-05) (hash(x)=23486277)
210
+ 200 val loss 9.4299
211
+ 200 val perplexity 12455.8535
212
+ 200 train 9.377373 (lr=7.0280e-05) (hash(x)=25978130)
213
+ 201 train 9.419920 (lr=7.0629e-05) (hash(x)=28916006)
214
+ 202 train 9.488194 (lr=7.0979e-05) (hash(x)=33195100)
215
+ 203 train 9.453103 (lr=7.1329e-05) (hash(x)=25556929)
216
+ 204 train 9.371488 (lr=7.1678e-05) (hash(x)=29536986)
217
+ 205 train 9.426601 (lr=7.2028e-05) (hash(x)=29648798)
218
+ 206 train 9.394245 (lr=7.2378e-05) (hash(x)=23124767)
219
+ 207 train 9.271253 (lr=7.2727e-05) (hash(x)=24976217)
220
+ 208 train 9.312224 (lr=7.3077e-05) (hash(x)=23632825)
221
+ 209 train 9.385268 (lr=7.3427e-05) (hash(x)=31019606)
222
+ 210 train 9.405113 (lr=7.3776e-05) (hash(x)=28236580)
223
+ 211 train 9.278949 (lr=7.4126e-05) (hash(x)=28269421)
224
+ 212 train 9.218432 (lr=7.4476e-05) (hash(x)=25563279)
225
+ 213 train 9.334978 (lr=7.4825e-05) (hash(x)=24865358)
226
+ 214 train 9.217773 (lr=7.5175e-05) (hash(x)=22155546)
227
+ 215 train 9.136641 (lr=7.5524e-05) (hash(x)=26984272)
228
+ 216 train 9.244029 (lr=7.5874e-05) (hash(x)=23193673)
229
+ 217 train 9.283193 (lr=7.6224e-05) (hash(x)=24544116)
230
+ 218 train 9.227083 (lr=7.6573e-05) (hash(x)=24547533)
231
+ 219 train 9.234766 (lr=7.6923e-05) (hash(x)=22059850)
232
+ 220 train 9.214176 (lr=7.7273e-05) (hash(x)=25994573)
233
+ 221 train 9.200909 (lr=7.7622e-05) (hash(x)=23926632)
234
+ 222 train 9.113145 (lr=7.7972e-05) (hash(x)=27111369)
235
+ 223 train 9.193667 (lr=7.8322e-05) (hash(x)=27025333)
236
+ 224 train 9.105745 (lr=7.8671e-05) (hash(x)=16908068)
237
+ 225 train 9.137980 (lr=7.9021e-05) (hash(x)=29120407)
238
+ 226 train 9.251821 (lr=7.9371e-05) (hash(x)=27092710)
239
+ 227 train 9.099687 (lr=7.9720e-05) (hash(x)=23827393)
240
+ 228 train 9.019650 (lr=8.0070e-05) (hash(x)=32812727)
241
+ 229 train 9.110296 (lr=8.0420e-05) (hash(x)=27230027)
242
+ 230 train 9.074903 (lr=8.0769e-05) (hash(x)=25703381)
243
+ 231 train 9.060614 (lr=8.1119e-05) (hash(x)=24968260)
244
+ 232 train 9.063730 (lr=8.1469e-05) (hash(x)=25676468)
245
+ 233 train 9.089333 (lr=8.1818e-05) (hash(x)=27544803)
246
+ 234 train 9.062958 (lr=8.2168e-05) (hash(x)=24331407)
247
+ 235 train 9.016635 (lr=8.2517e-05) (hash(x)=26540663)
248
+ 236 train 9.009319 (lr=8.2867e-05) (hash(x)=25532657)
249
+ 237 train 9.116049 (lr=8.3217e-05) (hash(x)=25856625)
250
+ 238 train 8.958660 (lr=8.3566e-05) (hash(x)=23437426)
251
+ 239 train 9.006875 (lr=8.3916e-05) (hash(x)=23164356)
252
+ 240 train 8.964865 (lr=8.4266e-05) (hash(x)=25957896)
253
+ 241 train 9.035215 (lr=8.4615e-05) (hash(x)=25646282)
254
+ 242 train 9.052797 (lr=8.4965e-05) (hash(x)=21513584)
255
+ 243 train 8.944491 (lr=8.5315e-05) (hash(x)=23843233)
256
+ 244 train 8.916464 (lr=8.5664e-05) (hash(x)=23226697)
257
+ 245 train 8.823786 (lr=8.6014e-05) (hash(x)=24506028)
258
+ 246 train 8.909817 (lr=8.6364e-05) (hash(x)=26024321)
259
+ 247 train 8.896194 (lr=8.6713e-05) (hash(x)=22781277)
260
+ 248 train 8.929066 (lr=8.7063e-05) (hash(x)=24451843)
261
+ 249 train 8.950446 (lr=8.7413e-05) (hash(x)=24948650)
262
+ 250 val loss 8.9133
263
+ 250 val perplexity 7429.9385
264
+ 250 train 8.871754 (lr=8.7762e-05) (hash(x)=25559534)
265
+ 251 train 8.873118 (lr=8.8112e-05) (hash(x)=26373991)
266
+ 252 train 8.928583 (lr=8.8462e-05) (hash(x)=27756673)
267
+ 253 train 8.873676 (lr=8.8811e-05) (hash(x)=23402396)
268
+ 254 train 8.826410 (lr=8.9161e-05) (hash(x)=28250354)
269
+ 255 train 8.858820 (lr=8.9510e-05) (hash(x)=21562133)
270
+ 256 train 8.931221 (lr=8.9860e-05) (hash(x)=27253861)
271
+ 257 train 8.846184 (lr=9.0210e-05) (hash(x)=25009210)
272
+ 258 train 8.697174 (lr=9.0559e-05) (hash(x)=23052577)
273
+ 259 train 8.801602 (lr=9.0909e-05) (hash(x)=27161811)
274
+ 260 train 8.740181 (lr=9.1259e-05) (hash(x)=27302459)
275
+ 261 train 8.759892 (lr=9.1608e-05) (hash(x)=29515100)
276
+ 262 train 8.715187 (lr=9.1958e-05) (hash(x)=25318001)
277
+ 263 train 8.707712 (lr=9.2308e-05) (hash(x)=25305929)
278
+ 264 train 8.712384 (lr=9.2657e-05) (hash(x)=24339043)
279
+ 265 train 8.714159 (lr=9.3007e-05) (hash(x)=26808543)
280
+ 266 train 8.629950 (lr=9.3357e-05) (hash(x)=25763991)
281
+ 267 train 8.675389 (lr=9.3706e-05) (hash(x)=25669509)
282
+ 268 train 8.730092 (lr=9.4056e-05) (hash(x)=27676869)
283
+ 269 train 8.648525 (lr=9.4406e-05) (hash(x)=28435805)
284
+ 270 train 8.617868 (lr=9.4755e-05) (hash(x)=23455211)
285
+ 271 train 8.716974 (lr=9.5105e-05) (hash(x)=26482758)
286
+ 272 train 8.556553 (lr=9.5455e-05) (hash(x)=27068280)
287
+ 273 train 8.599915 (lr=9.5804e-05) (hash(x)=24929929)
288
+ 274 train 8.618320 (lr=9.6154e-05) (hash(x)=25373676)
289
+ 275 train 8.611420 (lr=9.6503e-05) (hash(x)=25043037)
290
+ 276 train 8.597288 (lr=9.6853e-05) (hash(x)=27906976)
291
+ 277 train 8.788007 (lr=9.7203e-05) (hash(x)=29151257)
292
+ 278 train 8.678926 (lr=9.7552e-05) (hash(x)=26405613)
293
+ 279 train 8.449980 (lr=9.7902e-05) (hash(x)=26400365)
294
+ 280 train 8.522198 (lr=9.8252e-05) (hash(x)=23971905)
295
+ 281 train 8.492514 (lr=9.8601e-05) (hash(x)=24098812)
296
+ 282 train 8.570681 (lr=9.8951e-05) (hash(x)=29402976)
297
+ 283 train 8.477932 (lr=9.9301e-05) (hash(x)=20967666)
298
+ 284 train 8.517657 (lr=9.9650e-05) (hash(x)=24287610)
299
+ 285 train 8.680602 (lr=1.0000e-04) (hash(x)=28667963)
300
+ 286 train 8.527136 (lr=1.0000e-04) (hash(x)=25974316)
301
+ 287 train 8.556179 (lr=1.0000e-04) (hash(x)=27083417)
302
+ 288 train 8.461832 (lr=9.9998e-05) (hash(x)=28605205)
303
+ 289 train 8.472486 (lr=9.9996e-05) (hash(x)=23649001)
304
+ 290 train 8.435099 (lr=9.9993e-05) (hash(x)=25921700)
305
+ 291 train 8.478551 (lr=9.9989e-05) (hash(x)=31338300)
306
+ 292 train 8.417875 (lr=9.9984e-05) (hash(x)=23862328)
307
+ 293 train 8.414188 (lr=9.9979e-05) (hash(x)=30603174)
308
+ 294 train 8.424736 (lr=9.9972e-05) (hash(x)=23603806)
309
+ 295 train 8.393325 (lr=9.9965e-05) (hash(x)=24830965)
310
+ 296 train 8.464300 (lr=9.9956e-05) (hash(x)=32488729)
311
+ 297 train 8.498103 (lr=9.9947e-05) (hash(x)=27269893)
312
+ 298 train 8.390118 (lr=9.9937e-05) (hash(x)=25215077)
313
+ 299 train 8.370158 (lr=9.9926e-05) (hash(x)=24825041)
314
+ 300 val loss 8.3697
315
+ 300 val perplexity 4314.2568
316
+ 300 train 8.396864 (lr=9.9915e-05) (hash(x)=25893804)
317
+ 301 train 8.295256 (lr=9.9902e-05) (hash(x)=22724871)
318
+ 302 train 8.312361 (lr=9.9889e-05) (hash(x)=27317554)
319
+ 303 train 8.250538 (lr=9.9874e-05) (hash(x)=24683717)
320
+ 304 train 8.285161 (lr=9.9859e-05) (hash(x)=21584083)
321
+ 305 train 8.447642 (lr=9.9843e-05) (hash(x)=26190337)
322
+ 306 train 8.341165 (lr=9.9826e-05) (hash(x)=22249019)
323
+ 307 train 8.328814 (lr=9.9808e-05) (hash(x)=23086289)
324
+ 308 train 8.484922 (lr=9.9789e-05) (hash(x)=28047044)
325
+ 309 train 8.272476 (lr=9.9770e-05) (hash(x)=23593875)
326
+ 310 train 8.263665 (lr=9.9749e-05) (hash(x)=26059735)
327
+ 311 train 8.244967 (lr=9.9728e-05) (hash(x)=22414190)
328
+ 312 train 8.300624 (lr=9.9706e-05) (hash(x)=24204591)
329
+ 313 train 8.267406 (lr=9.9683e-05) (hash(x)=26758114)
330
+ 314 train 8.272927 (lr=9.9659e-05) (hash(x)=20084233)
331
+ 315 train 8.201039 (lr=9.9634e-05) (hash(x)=21563184)
332
+ 316 train 8.252774 (lr=9.9609e-05) (hash(x)=23861489)
333
+ 317 train 8.295657 (lr=9.9582e-05) (hash(x)=19218470)
334
+ 318 train 8.286336 (lr=9.9555e-05) (hash(x)=28469562)
335
+ 319 train 8.363187 (lr=9.9526e-05) (hash(x)=25588469)
336
+ 320 train 8.263462 (lr=9.9497e-05) (hash(x)=25236367)
337
+ 321 train 8.261034 (lr=9.9467e-05) (hash(x)=30433767)
338
+ 322 train 8.391470 (lr=9.9437e-05) (hash(x)=26302626)
339
+ 323 train 8.183725 (lr=9.9405e-05) (hash(x)=17851621)
340
+ 324 train 8.355635 (lr=9.9372e-05) (hash(x)=27028191)
341
+ 325 train 8.318135 (lr=9.9339e-05) (hash(x)=25183690)
342
+ 326 train 8.356015 (lr=9.9305e-05) (hash(x)=26182243)
343
+ 327 train 8.377423 (lr=9.9270e-05) (hash(x)=27040642)
344
+ 328 train 8.177889 (lr=9.9234e-05) (hash(x)=24489607)
345
+ 329 train 8.170621 (lr=9.9197e-05) (hash(x)=22183303)
346
+ 330 train 8.244492 (lr=9.9159e-05) (hash(x)=28844646)
347
+ 331 train 8.551321 (lr=9.9121e-05) (hash(x)=27537063)
348
+ 332 train 8.228240 (lr=9.9081e-05) (hash(x)=25446686)
349
+ 333 train 8.115274 (lr=9.9041e-05) (hash(x)=19492832)
350
+ 334 train 8.097966 (lr=9.9000e-05) (hash(x)=25484958)
351
+ 335 train 8.160414 (lr=9.8958e-05) (hash(x)=25951629)
352
+ 336 train 8.210470 (lr=9.8915e-05) (hash(x)=23462798)
353
+ 337 train 8.103142 (lr=9.8872e-05) (hash(x)=22408682)
354
+ 338 train 8.212992 (lr=9.8827e-05) (hash(x)=31317970)
355
+ 339 train 8.085554 (lr=9.8782e-05) (hash(x)=27633457)
356
+ 340 train 8.050351 (lr=9.8736e-05) (hash(x)=23729283)
357
+ 341 train 8.220807 (lr=9.8689e-05) (hash(x)=27635080)
358
+ 342 train 8.147395 (lr=9.8641e-05) (hash(x)=25918780)
359
+ 343 train 8.120698 (lr=9.8592e-05) (hash(x)=22287596)
360
+ 344 train 8.057014 (lr=9.8543e-05) (hash(x)=22694623)
361
+ 345 train 8.361918 (lr=9.8492e-05) (hash(x)=35895440)
362
+ 346 train 8.137057 (lr=9.8441e-05) (hash(x)=24298886)
363
+ 347 train 8.125513 (lr=9.8389e-05) (hash(x)=25888847)
364
+ 348 train 8.159514 (lr=9.8336e-05) (hash(x)=30860646)
365
+ 349 train 8.107687 (lr=9.8282e-05) (hash(x)=26518800)
366
+ 350 val loss 8.0751
367
+ 350 val perplexity 3213.5435
368
+ 350 train 8.070807 (lr=9.8228e-05) (hash(x)=25385523)
369
+ 351 train 8.205559 (lr=9.8172e-05) (hash(x)=37519283)
370
+ 352 train 8.408539 (lr=9.8116e-05) (hash(x)=20564397)
371
+ 353 train 8.115188 (lr=9.8059e-05) (hash(x)=25141945)
372
+ 354 train 8.108124 (lr=9.8001e-05) (hash(x)=22717866)
373
+ 355 train 8.433540 (lr=9.7942e-05) (hash(x)=28527939)
374
+ 356 train 8.181572 (lr=9.7882e-05) (hash(x)=32757059)
375
+ 357 train 7.993158 (lr=9.7822e-05) (hash(x)=21942471)
376
+ 358 train 8.201951 (lr=9.7761e-05) (hash(x)=24247339)
377
+ 359 train 8.058681 (lr=9.7699e-05) (hash(x)=24955630)
378
+ 360 train 8.109174 (lr=9.7636e-05) (hash(x)=21368780)
379
+ 361 train 8.036007 (lr=9.7572e-05) (hash(x)=25541230)
380
+ 362 train 7.946487 (lr=9.7507e-05) (hash(x)=23098806)
381
+ 363 train 7.951101 (lr=9.7442e-05) (hash(x)=23501481)
382
+ 364 train 7.979908 (lr=9.7376e-05) (hash(x)=23802681)
383
+ 365 train 8.031181 (lr=9.7309e-05) (hash(x)=22754988)
384
+ 366 train 7.901494 (lr=9.7241e-05) (hash(x)=22183009)
385
+ 367 train 7.886165 (lr=9.7172e-05) (hash(x)=23475070)
386
+ 368 train 8.019608 (lr=9.7103e-05) (hash(x)=23359061)
387
+ 369 train 8.010147 (lr=9.7032e-05) (hash(x)=23609959)
388
+ 370 train 7.975481 (lr=9.6961e-05) (hash(x)=22408449)
389
+ 371 train 7.906122 (lr=9.6889e-05) (hash(x)=23441847)
390
+ 372 train 7.886661 (lr=9.6817e-05) (hash(x)=23676365)
391
+ 373 train 8.045522 (lr=9.6743e-05) (hash(x)=23096292)
392
+ 374 train 8.029465 (lr=9.6669e-05) (hash(x)=27706294)
393
+ 375 train 7.949654 (lr=9.6593e-05) (hash(x)=26171093)
394
+ 376 train 8.112991 (lr=9.6518e-05) (hash(x)=34946941)
395
+ 377 train 7.902636 (lr=9.6441e-05) (hash(x)=20060838)
396
+ 378 train 7.792796 (lr=9.6363e-05) (hash(x)=21591702)
397
+ 379 train 7.909789 (lr=9.6285e-05) (hash(x)=25451055)
398
+ 380 train 7.564116 (lr=9.6206e-05) (hash(x)=18691455)
399
+ 381 train 7.882860 (lr=9.6126e-05) (hash(x)=24934792)
400
+ 382 train 7.808902 (lr=9.6045e-05) (hash(x)=23864361)
401
+ 383 train 7.853098 (lr=9.5963e-05) (hash(x)=25367597)
402
+ 384 train 7.958365 (lr=9.5881e-05) (hash(x)=27016702)
403
+ 385 train 7.940325 (lr=9.5798e-05) (hash(x)=25022148)
404
+ 386 train 7.956110 (lr=9.5714e-05) (hash(x)=24459012)
405
+ 387 train 7.909300 (lr=9.5629e-05) (hash(x)=25431365)
406
+ 388 train 7.871985 (lr=9.5544e-05) (hash(x)=25958785)
407
+ 389 train 8.025242 (lr=9.5457e-05) (hash(x)=29307788)
408
+ 390 train 7.804760 (lr=9.5370e-05) (hash(x)=24965435)
409
+ 391 train 7.853750 (lr=9.5282e-05) (hash(x)=25651929)
410
+ 392 train 7.887435 (lr=9.5194e-05) (hash(x)=28451031)
411
+ 393 train 7.935885 (lr=9.5104e-05) (hash(x)=26312387)
412
+ 394 train 7.905349 (lr=9.5014e-05) (hash(x)=25326650)
413
+ 395 train 7.818499 (lr=9.4923e-05) (hash(x)=23804913)
414
+ 396 train 7.998854 (lr=9.4831e-05) (hash(x)=25537529)
415
+ 397 train 7.809297 (lr=9.4739e-05) (hash(x)=26115519)
416
+ 398 train 7.803030 (lr=9.4646e-05) (hash(x)=25494804)
417
+ 399 train 7.935295 (lr=9.4551e-05) (hash(x)=23210747)
418
+ 400 val loss 7.8653
419
+ 400 val perplexity 2605.3943
420
+ 400 train 7.847210 (lr=9.4457e-05) (hash(x)=15717069)
421
+ 401 train 7.820347 (lr=9.4361e-05) (hash(x)=18463712)
422
+ 402 train 7.862066 (lr=9.4265e-05) (hash(x)=20705573)
423
+ 403 train 7.827834 (lr=9.4168e-05) (hash(x)=23245699)
424
+ 404 train 7.814764 (lr=9.4070e-05) (hash(x)=23444521)
425
+ 405 train 7.770213 (lr=9.3971e-05) (hash(x)=21770300)
426
+ 406 train 8.436256 (lr=9.3872e-05) (hash(x)=23334569)
427
+ 407 train 7.679902 (lr=9.3772e-05) (hash(x)=25387532)
428
+ 408 train 7.844475 (lr=9.3671e-05) (hash(x)=29466743)
429
+ 409 train 7.862890 (lr=9.3569e-05) (hash(x)=27718703)
430
+ 410 train 7.901713 (lr=9.3467e-05) (hash(x)=17631973)
431
+ 411 train 7.923923 (lr=9.3364e-05) (hash(x)=27661052)
432
+ 412 train 7.722741 (lr=9.3260e-05) (hash(x)=22244096)
433
+ 413 train 7.814191 (lr=9.3155e-05) (hash(x)=26538928)
434
+ 414 train 7.815310 (lr=9.3050e-05) (hash(x)=26256802)
435
+ 415 train 7.904099 (lr=9.2944e-05) (hash(x)=21772654)
436
+ 416 train 7.877647 (lr=9.2837e-05) (hash(x)=25026277)
437
+ 417 train 7.960544 (lr=9.2729e-05) (hash(x)=26263254)
438
+ 418 train 7.886663 (lr=9.2621e-05) (hash(x)=24726314)
439
+ 419 train 7.731798 (lr=9.2512e-05) (hash(x)=23552356)
440
+ 420 train 7.812923 (lr=9.2402e-05) (hash(x)=26699819)
441
+ 421 train 8.146421 (lr=9.2292e-05) (hash(x)=22707141)
442
+ 422 train 8.129635 (lr=9.2181e-05) (hash(x)=25676149)
443
+ 423 train 7.900240 (lr=9.2069e-05) (hash(x)=34234481)
444
+ 424 train 7.775521 (lr=9.1956e-05) (hash(x)=25401624)
445
+ 425 train 7.939219 (lr=9.1843e-05) (hash(x)=35008670)
446
+ 426 train 8.017336 (lr=9.1729e-05) (hash(x)=25682280)
447
+ 427 train 7.864930 (lr=9.1614e-05) (hash(x)=25891543)
448
+ 428 train 7.757998 (lr=9.1499e-05) (hash(x)=25962407)
449
+ 429 train 7.685889 (lr=9.1382e-05) (hash(x)=21971224)
450
+ 430 train 7.801029 (lr=9.1266e-05) (hash(x)=30838142)
451
+ 431 train 7.842700 (lr=9.1148e-05) (hash(x)=27312186)
452
+ 432 train 7.938363 (lr=9.1030e-05) (hash(x)=23657081)
453
+ 433 train 7.906349 (lr=9.0911e-05) (hash(x)=22163749)
454
+ 434 train 7.575752 (lr=9.0791e-05) (hash(x)=24829818)
455
+ 435 train 7.832965 (lr=9.0671e-05) (hash(x)=25942888)
456
+ 436 train 7.795085 (lr=9.0550e-05) (hash(x)=23785671)
457
+ 437 train 7.823710 (lr=9.0428e-05) (hash(x)=25493916)
458
+ 438 train 7.944253 (lr=9.0306e-05) (hash(x)=28618702)
459
+ 439 train 7.995876 (lr=9.0182e-05) (hash(x)=24289951)
460
+ 440 train 7.768553 (lr=9.0059e-05) (hash(x)=26331486)
461
+ 441 train 7.767030 (lr=8.9934e-05) (hash(x)=23648671)
462
+ 442 train 7.707991 (lr=8.9809e-05) (hash(x)=27292797)
463
+ 443 train 7.581060 (lr=8.9683e-05) (hash(x)=24992055)
464
+ 444 train 8.011742 (lr=8.9557e-05) (hash(x)=24398330)
465
+ 445 train 7.766413 (lr=8.9430e-05) (hash(x)=26005549)
466
+ 446 train 7.749378 (lr=8.9302e-05) (hash(x)=30259737)
467
+ 447 train 7.746036 (lr=8.9173e-05) (hash(x)=21046554)
468
+ 448 train 7.751832 (lr=8.9044e-05) (hash(x)=18773877)
469
+ 449 train 7.749691 (lr=8.8914e-05) (hash(x)=26321539)
470
+ 450 val loss 7.7555
471
+ 450 val perplexity 2334.2913
472
+ 450 train 7.740405 (lr=8.8784e-05) (hash(x)=22850804)
473
+ 451 train 7.786320 (lr=8.8653e-05) (hash(x)=27807872)
474
+ 452 train 7.773264 (lr=8.8521e-05) (hash(x)=25545137)
475
+ 453 train 7.767291 (lr=8.8388e-05) (hash(x)=21555364)
476
+ 454 train 7.809669 (lr=8.8255e-05) (hash(x)=27243798)
477
+ 455 train 7.690829 (lr=8.8122e-05) (hash(x)=26325324)
478
+ 456 train 7.655442 (lr=8.7987e-05) (hash(x)=26228728)
479
+ 457 train 7.866999 (lr=8.7852e-05) (hash(x)=24956294)
480
+ 458 train 7.734053 (lr=8.7717e-05) (hash(x)=25592289)
481
+ 459 train 7.715889 (lr=8.7580e-05) (hash(x)=24429389)
482
+ 460 train 7.735579 (lr=8.7444e-05) (hash(x)=28733708)
483
+ 461 train 7.919900 (lr=8.7306e-05) (hash(x)=27929846)
484
+ 462 train 7.745178 (lr=8.7168e-05) (hash(x)=22417476)
485
+ 463 train 7.816964 (lr=8.7029e-05) (hash(x)=25282897)
486
+ 464 train 7.635806 (lr=8.6890e-05) (hash(x)=22278062)
487
+ 465 train 7.636587 (lr=8.6750e-05) (hash(x)=26802508)
488
+ 466 train 7.875931 (lr=8.6609e-05) (hash(x)=22094832)
489
+ 467 train 7.772000 (lr=8.6468e-05) (hash(x)=28390416)
490
+ 468 train 7.910661 (lr=8.6326e-05) (hash(x)=26379331)
491
+ 469 train 7.942529 (lr=8.6184e-05) (hash(x)=26292451)
492
+ 470 train 7.838595 (lr=8.6041e-05) (hash(x)=26301866)
493
+ 471 train 7.841110 (lr=8.5897e-05) (hash(x)=22915344)
494
+ 472 train 7.720595 (lr=8.5753e-05) (hash(x)=25448318)
495
+ 473 train 7.794286 (lr=8.5608e-05) (hash(x)=24744383)
496
+ 474 train 7.635128 (lr=8.5462e-05) (hash(x)=22835201)
497
+ 475 train 7.711008 (lr=8.5316e-05) (hash(x)=26284678)
498
+ 476 train 7.724489 (lr=8.5170e-05) (hash(x)=24745936)
499
+ 477 train 7.586637 (lr=8.5022e-05) (hash(x)=21654269)
500
+ 478 train 7.758162 (lr=8.4875e-05) (hash(x)=32387999)
501
+ 479 train 8.394434 (lr=8.4726e-05) (hash(x)=29466268)
502
+ 480 train 8.340124 (lr=8.4577e-05) (hash(x)=30736727)
503
+ 481 train 8.318363 (lr=8.4428e-05) (hash(x)=32616762)
504
+ 482 train 8.294589 (lr=8.4278e-05) (hash(x)=29358578)
505
+ 483 train 8.367889 (lr=8.4127e-05) (hash(x)=32201138)
506
+ 484 train 8.390668 (lr=8.3976e-05) (hash(x)=30041621)
507
+ 485 train 8.355347 (lr=8.3824e-05) (hash(x)=33683468)
508
+ 486 train 8.055648 (lr=8.3672e-05) (hash(x)=26837755)
509
+ 487 train 7.830695 (lr=8.3519e-05) (hash(x)=27498125)
510
+ 488 train 7.984925 (lr=8.3366e-05) (hash(x)=22372507)
511
+ 489 train 7.670789 (lr=8.3212e-05) (hash(x)=25164318)
512
+ 490 train 7.717383 (lr=8.3057e-05) (hash(x)=26562947)
513
+ 491 train 7.675936 (lr=8.2902e-05) (hash(x)=25690221)
514
+ 492 train 7.746218 (lr=8.2746e-05) (hash(x)=21947305)
515
+ 493 train 7.687648 (lr=8.2590e-05) (hash(x)=29031550)
516
+ 494 train 7.815752 (lr=8.2434e-05) (hash(x)=26551589)
517
+ 495 train 7.803751 (lr=8.2276e-05) (hash(x)=24094093)
518
+ 496 train 7.602840 (lr=8.2119e-05) (hash(x)=23639076)
519
+ 497 train 9.217509 (lr=8.1960e-05) (hash(x)=61455960)
520
+ 498 train 11.649544 (lr=8.1801e-05) (hash(x)=65636572)
521
+ 499 train 11.616954 (lr=8.1642e-05) (hash(x)=61104710)
522
+ 500 val loss 7.7316
523
+ 500 val perplexity 2279.2478
524
+ 500 train 8.383205 (lr=8.1482e-05) (hash(x)=24123821)
525
+ 501 train 7.390849 (lr=8.1322e-05) (hash(x)=20881126)
526
+ 502 train 7.773952 (lr=8.1161e-05) (hash(x)=23209969)
527
+ 503 train 7.796063 (lr=8.1000e-05) (hash(x)=26920243)
528
+ 504 train 7.672505 (lr=8.0838e-05) (hash(x)=26127756)
529
+ 505 train 7.614229 (lr=8.0676e-05) (hash(x)=21970343)
530
+ 506 train 7.679968 (lr=8.0513e-05) (hash(x)=24097897)
531
+ 507 train 7.764169 (lr=8.0349e-05) (hash(x)=31676352)
532
+ 508 train 7.719649 (lr=8.0186e-05) (hash(x)=23908053)
533
+ 509 train 7.674220 (lr=8.0021e-05) (hash(x)=25600562)
534
+ 510 train 7.715953 (lr=7.9856e-05) (hash(x)=23549168)
535
+ 511 train 7.649692 (lr=7.9691e-05) (hash(x)=22022924)
536
+ 512 train 7.671283 (lr=7.9525e-05) (hash(x)=23294980)
537
+ 513 train 7.574963 (lr=7.9359e-05) (hash(x)=22777085)
538
+ 514 train 7.854635 (lr=7.9192e-05) (hash(x)=26271115)
539
+ 515 train 7.656049 (lr=7.9025e-05) (hash(x)=21727744)
540
+ 516 train 7.704733 (lr=7.8858e-05) (hash(x)=22729318)
541
+ 517 train 7.865342 (lr=7.8689e-05) (hash(x)=20701998)
542
+ 518 train 7.760269 (lr=7.8521e-05) (hash(x)=27928398)
543
+ 519 train 7.834486 (lr=7.8352e-05) (hash(x)=20863982)
544
+ 520 train 7.692726 (lr=7.8182e-05) (hash(x)=22827054)
545
+ 521 train 7.703193 (lr=7.8012e-05) (hash(x)=25084926)
546
+ 522 train 7.816121 (lr=7.7842e-05) (hash(x)=25540961)
547
+ 523 train 7.595809 (lr=7.7671e-05) (hash(x)=24412897)
548
+ 524 train 7.812642 (lr=7.7500e-05) (hash(x)=26269869)
549
+ 525 train 7.584125 (lr=7.7328e-05) (hash(x)=25434913)
550
+ 526 train 7.735545 (lr=7.7156e-05) (hash(x)=25157653)
551
+ 527 train 7.727058 (lr=7.6984e-05) (hash(x)=26302544)
552
+ 528 train 7.638879 (lr=7.6811e-05) (hash(x)=22010703)
553
+ 529 train 7.932634 (lr=7.6637e-05) (hash(x)=34976197)
554
+ 530 train 7.649604 (lr=7.6463e-05) (hash(x)=28248301)
555
+ 531 train 7.681161 (lr=7.6289e-05) (hash(x)=23870805)
556
+ 532 train 7.649999 (lr=7.6115e-05) (hash(x)=24457252)
557
+ 533 train 7.725609 (lr=7.5940e-05) (hash(x)=25041103)
558
+ 534 train 7.654294 (lr=7.5764e-05) (hash(x)=25998487)
559
+ 535 train 7.781257 (lr=7.5588e-05) (hash(x)=26041745)
560
+ 536 train 7.657908 (lr=7.5412e-05) (hash(x)=27249810)
561
+ 537 train 7.699659 (lr=7.5235e-05) (hash(x)=24068619)
562
+ 538 train 7.731132 (lr=7.5058e-05) (hash(x)=26359528)
563
+ 539 train 7.698642 (lr=7.4881e-05) (hash(x)=25441262)
564
+ 540 train 7.616361 (lr=7.4703e-05) (hash(x)=24413659)
565
+ 541 train 7.989346 (lr=7.4525e-05) (hash(x)=29181867)
566
+ 542 train 7.762399 (lr=7.4346e-05) (hash(x)=31551575)
567
+ 543 train 7.744263 (lr=7.4167e-05) (hash(x)=27093757)
568
+ 544 train 7.722213 (lr=7.3988e-05) (hash(x)=23624298)
569
+ 545 train 7.586331 (lr=7.3808e-05) (hash(x)=25150275)
570
+ 546 train 7.602991 (lr=7.3628e-05) (hash(x)=26239886)
571
+ 547 train 7.969000 (lr=7.3448e-05) (hash(x)=33457102)
572
+ 548 train 7.597427 (lr=7.3267e-05) (hash(x)=17499738)
573
+ 549 train 7.637483 (lr=7.3086e-05) (hash(x)=25755322)
574
+ 550 val loss 7.6587
575
+ 550 val perplexity 2119.0200
576
+ 550 train 7.736249 (lr=7.2904e-05) (hash(x)=23722795)
577
+ 551 train 7.620814 (lr=7.2723e-05) (hash(x)=23630794)
578
+ 552 train 7.584390 (lr=7.2540e-05) (hash(x)=23374168)
579
+ 553 train 7.587133 (lr=7.2358e-05) (hash(x)=25167987)
580
+ 554 train 7.852980 (lr=7.2175e-05) (hash(x)=30941010)
581
+ 555 train 7.594168 (lr=7.1992e-05) (hash(x)=24043224)
582
+ 556 train 7.721970 (lr=7.1808e-05) (hash(x)=29736627)
583
+ 557 train 7.684642 (lr=7.1624e-05) (hash(x)=24665411)
584
+ 558 train 7.699383 (lr=7.1440e-05) (hash(x)=27916043)
585
+ 559 train 7.644013 (lr=7.1256e-05) (hash(x)=17321657)
586
+ 560 train 7.467325 (lr=7.1071e-05) (hash(x)=23829833)
587
+ 561 train 7.871204 (lr=7.0886e-05) (hash(x)=27932591)
588
+ 562 train 7.932806 (lr=7.0701e-05) (hash(x)=28206079)
589
+ 563 train 8.044420 (lr=7.0515e-05) (hash(x)=29447954)
590
+ 564 train 7.688857 (lr=7.0329e-05) (hash(x)=25935364)
591
+ 565 train 7.695575 (lr=7.0143e-05) (hash(x)=24621816)
592
+ 566 train 7.635980 (lr=6.9956e-05) (hash(x)=24070405)
593
+ 567 train 7.703582 (lr=6.9769e-05) (hash(x)=24682294)
594
+ 568 train 7.711880 (lr=6.9582e-05) (hash(x)=24050209)
595
+ 569 train 7.666784 (lr=6.9394e-05) (hash(x)=28057095)
596
+ 570 train 7.769289 (lr=6.9207e-05) (hash(x)=22523232)
597
+ 571 train 7.684893 (lr=6.9019e-05) (hash(x)=24547536)
598
+ 572 train 7.575873 (lr=6.8830e-05) (hash(x)=24978712)
599
+ 573 train 7.532546 (lr=6.8642e-05) (hash(x)=18987545)
600
+ 574 train 7.646169 (lr=6.8453e-05) (hash(x)=24092784)
601
+ 575 train 7.621340 (lr=6.8264e-05) (hash(x)=26745063)
602
+ 576 train 7.637866 (lr=6.8075e-05) (hash(x)=22539112)
603
+ 577 train 7.391376 (lr=6.7885e-05) (hash(x)=21447406)
604
+ 578 train 7.420279 (lr=6.7695e-05) (hash(x)=21082139)
605
+ 579 train 7.470431 (lr=6.7505e-05) (hash(x)=24650655)
606
+ 580 train 7.669219 (lr=6.7315e-05) (hash(x)=22130721)
607
+ 581 train 7.986447 (lr=6.7124e-05) (hash(x)=26519976)
608
+ 582 train 8.554877 (lr=6.6933e-05) (hash(x)=25582440)
609
+ 583 train 7.771469 (lr=6.6742e-05) (hash(x)=20106514)
610
+ 584 train 7.528815 (lr=6.6551e-05) (hash(x)=25695510)
611
+ 585 train 7.568365 (lr=6.6360e-05) (hash(x)=20117808)
612
+ 586 train 7.664600 (lr=6.6168e-05) (hash(x)=24501900)
613
+ 587 train 7.672115 (lr=6.5976e-05) (hash(x)=25510334)
614
+ 588 train 7.528102 (lr=6.5784e-05) (hash(x)=22099158)
615
+ 589 train 7.650597 (lr=6.5592e-05) (hash(x)=25749120)
616
+ 590 train 7.940101 (lr=6.5399e-05) (hash(x)=21720691)
617
+ 591 train 7.485755 (lr=6.5206e-05) (hash(x)=24357241)
618
+ 592 train 7.759824 (lr=6.5013e-05) (hash(x)=28911272)
619
+ 593 train 7.644472 (lr=6.4820e-05) (hash(x)=25584945)
620
+ 594 train 7.536990 (lr=6.4627e-05) (hash(x)=21847287)
621
+ 595 train 7.589716 (lr=6.4433e-05) (hash(x)=25198897)
622
+ 596 train 7.479732 (lr=6.4240e-05) (hash(x)=22436833)
623
+ 597 train 7.768996 (lr=6.4046e-05) (hash(x)=23475891)
624
+ 598 train 7.439800 (lr=6.3852e-05) (hash(x)=25083992)
625
+ 599 train 7.721510 (lr=6.3658e-05) (hash(x)=23644804)
626
+ 600 val loss 7.6346
627
+ 600 val perplexity 2068.6345
628
+ 600 train 7.589725 (lr=6.3463e-05) (hash(x)=31377168)
629
+ 601 train 7.970809 (lr=6.3269e-05) (hash(x)=25629584)
630
+ 602 train 7.803752 (lr=6.3074e-05) (hash(x)=29285135)
631
+ 603 train 7.984015 (lr=6.2879e-05) (hash(x)=29797280)
632
+ 604 train 7.820681 (lr=6.2684e-05) (hash(x)=33122326)
633
+ 605 train 8.106792 (lr=6.2489e-05) (hash(x)=30026394)
634
+ 606 train 8.333110 (lr=6.2294e-05) (hash(x)=34353164)
635
+ 607 train 7.594029 (lr=6.2098e-05) (hash(x)=26594177)
636
+ 608 train 7.678337 (lr=6.1903e-05) (hash(x)=25554738)
637
+ 609 train 7.762151 (lr=6.1707e-05) (hash(x)=23766256)
638
+ 610 train 7.652583 (lr=6.1511e-05) (hash(x)=26853444)
639
+ 611 train 7.673648 (lr=6.1315e-05) (hash(x)=24250150)
640
+ 612 train 7.566403 (lr=6.1119e-05) (hash(x)=26289588)
641
+ 613 train 7.653089 (lr=6.0923e-05) (hash(x)=26108817)
642
+ 614 train 7.575051 (lr=6.0726e-05) (hash(x)=23814987)
643
+ 615 train 7.567137 (lr=6.0530e-05) (hash(x)=18582688)
644
+ 616 train 7.583081 (lr=6.0333e-05) (hash(x)=25558792)
645
+ 617 train 7.555091 (lr=6.0137e-05) (hash(x)=23272705)
646
+ 618 train 7.554976 (lr=5.9940e-05) (hash(x)=24709241)
647
+ 619 train 7.571392 (lr=5.9743e-05) (hash(x)=24793412)
648
+ 620 train 7.570584 (lr=5.9546e-05) (hash(x)=27256707)
649
+ 621 train 7.552554 (lr=5.9349e-05) (hash(x)=19280989)
650
+ 622 train 7.591905 (lr=5.9152e-05) (hash(x)=26508642)
651
+ 623 train 7.534531 (lr=5.8955e-05) (hash(x)=22751150)
652
+ 624 train 7.808561 (lr=5.8758e-05) (hash(x)=23880160)
653
+ 625 train 7.547490 (lr=5.8560e-05) (hash(x)=23537942)
654
+ 626 train 7.556577 (lr=5.8363e-05) (hash(x)=26975313)
655
+ 627 train 7.486595 (lr=5.8165e-05) (hash(x)=23748375)
656
+ 628 train 7.485116 (lr=5.7968e-05) (hash(x)=25541754)
657
+ 629 train 7.871311 (lr=5.7770e-05) (hash(x)=27413825)
658
+ 630 train 7.823086 (lr=5.7573e-05) (hash(x)=27075951)
659
+ 631 train 7.654891 (lr=5.7375e-05) (hash(x)=21682445)
660
+ 632 train 7.632730 (lr=5.7177e-05) (hash(x)=23361504)
661
+ 633 train 7.548299 (lr=5.6979e-05) (hash(x)=22040184)
662
+ 634 train 7.574142 (lr=5.6782e-05) (hash(x)=23727731)
663
+ 635 train 7.576105 (lr=5.6584e-05) (hash(x)=24463229)
664
+ 636 train 7.780320 (lr=5.6386e-05) (hash(x)=32663792)
665
+ 637 train 7.614997 (lr=5.6188e-05) (hash(x)=24091954)
666
+ 638 train 7.550580 (lr=5.5990e-05) (hash(x)=23599013)
667
+ 639 train 7.551748 (lr=5.5792e-05) (hash(x)=26670219)
668
+ 640 train 7.540818 (lr=5.5594e-05) (hash(x)=24380893)
669
+ 641 train 7.628519 (lr=5.5396e-05) (hash(x)=25558986)
670
+ 642 train 7.683051 (lr=5.5198e-05) (hash(x)=25729492)
671
+ 643 train 7.444208 (lr=5.5000e-05) (hash(x)=19218943)
672
+ 644 train 7.666581 (lr=5.4802e-05) (hash(x)=23952206)
673
+ 645 train 7.639890 (lr=5.4604e-05) (hash(x)=24551402)
674
+ 646 train 7.527094 (lr=5.4406e-05) (hash(x)=24007642)
675
+ 647 train 7.568041 (lr=5.4208e-05) (hash(x)=25229281)
676
+ 648 train 7.349470 (lr=5.4010e-05) (hash(x)=18214283)
677
+ 649 train 7.308761 (lr=5.3812e-05) (hash(x)=21251127)
678
+ 650 val loss 7.6231
679
+ 650 val perplexity 2044.8160
680
+ 650 train 7.788250 (lr=5.3614e-05) (hash(x)=25097619)
681
+ 651 train 7.492816 (lr=5.3416e-05) (hash(x)=25800857)
682
+ 652 train 7.500025 (lr=5.3218e-05) (hash(x)=22918428)
683
+ 653 train 7.538114 (lr=5.3021e-05) (hash(x)=25836189)
684
+ 654 train 7.524338 (lr=5.2823e-05) (hash(x)=24504567)
685
+ 655 train 7.552638 (lr=5.2625e-05) (hash(x)=25999255)
686
+ 656 train 7.552148 (lr=5.2427e-05) (hash(x)=27308968)
687
+ 657 train 7.639724 (lr=5.2230e-05) (hash(x)=20149394)
688
+ 658 train 7.607770 (lr=5.2032e-05) (hash(x)=24498556)
689
+ 659 train 7.475309 (lr=5.1835e-05) (hash(x)=21999338)
690
+ 660 train 7.500627 (lr=5.1637e-05) (hash(x)=23280568)
691
+ 661 train 7.773887 (lr=5.1440e-05) (hash(x)=27611302)
692
+ 662 train 7.607110 (lr=5.1242e-05) (hash(x)=23982308)
693
+ 663 train 7.559577 (lr=5.1045e-05) (hash(x)=22778356)
694
+ 664 train 7.483366 (lr=5.0848e-05) (hash(x)=23467046)
695
+ 665 train 7.645515 (lr=5.0651e-05) (hash(x)=24195959)
696
+ 666 train 7.980962 (lr=5.0454e-05) (hash(x)=32710438)
697
+ 667 train 7.679995 (lr=5.0257e-05) (hash(x)=19127465)
698
+ 668 train 7.837014 (lr=5.0060e-05) (hash(x)=28962580)
699
+ 669 train 7.572300 (lr=4.9863e-05) (hash(x)=22426274)
700
+ 670 train 7.577860 (lr=4.9667e-05) (hash(x)=25606512)
701
+ 671 train 7.447354 (lr=4.9470e-05) (hash(x)=23785205)
702
+ 672 train 7.596399 (lr=4.9274e-05) (hash(x)=24779414)
703
+ 673 train 7.717569 (lr=4.9077e-05) (hash(x)=28505676)
704
+ 674 train 7.767404 (lr=4.8881e-05) (hash(x)=29156288)
705
+ 675 train 7.457535 (lr=4.8685e-05) (hash(x)=27706477)
706
+ 676 train 7.595258 (lr=4.8489e-05) (hash(x)=23711800)
707
+ 677 train 7.710373 (lr=4.8293e-05) (hash(x)=28684175)
708
+ 678 train 7.558385 (lr=4.8097e-05) (hash(x)=25225466)
709
+ 679 train 7.700738 (lr=4.7902e-05) (hash(x)=28003600)
710
+ 680 train 7.616917 (lr=4.7706e-05) (hash(x)=27989890)
711
+ 681 train 7.666185 (lr=4.7511e-05) (hash(x)=27586845)
712
+ 682 train 7.630069 (lr=4.7316e-05) (hash(x)=22373927)
713
+ 683 train 7.510782 (lr=4.7121e-05) (hash(x)=21176405)
714
+ 684 train 7.474745 (lr=4.6926e-05) (hash(x)=23806052)
715
+ 685 train 7.555648 (lr=4.6731e-05) (hash(x)=25888433)
716
+ 686 train 7.731767 (lr=4.6537e-05) (hash(x)=27797727)
717
+ 687 train 7.592105 (lr=4.6342e-05) (hash(x)=27916982)
718
+ 688 train 7.536187 (lr=4.6148e-05) (hash(x)=24621793)
719
+ 689 train 7.437708 (lr=4.5954e-05) (hash(x)=24334708)
720
+ 690 train 7.622230 (lr=4.5760e-05) (hash(x)=24419143)
721
+ 691 train 7.506947 (lr=4.5567e-05) (hash(x)=25617709)
722
+ 692 train 7.510048 (lr=4.5373e-05) (hash(x)=24556034)
723
+ 693 train 7.887058 (lr=4.5180e-05) (hash(x)=25152362)
724
+ 694 train 7.610797 (lr=4.4987e-05) (hash(x)=24326376)
725
+ 695 train 7.496637 (lr=4.4794e-05) (hash(x)=24276512)
726
+ 696 train 7.349700 (lr=4.4601e-05) (hash(x)=24344695)
727
+ 697 train 7.606761 (lr=4.4408e-05) (hash(x)=26776133)
728
+ 698 train 7.467590 (lr=4.4216e-05) (hash(x)=21847282)
729
+ 699 train 7.333020 (lr=4.4024e-05) (hash(x)=20254159)
730
+ 700 val loss 7.6080
731
+ 700 val perplexity 2014.1713
732
+ 700 train 7.441834 (lr=4.3832e-05) (hash(x)=25044762)
733
+ 701 train 7.541940 (lr=4.3640e-05) (hash(x)=27873855)
734
+ 702 train 7.582738 (lr=4.3449e-05) (hash(x)=22575615)
735
+ 703 train 7.565621 (lr=4.3258e-05) (hash(x)=25113614)
736
+ 704 train 7.672441 (lr=4.3067e-05) (hash(x)=28062905)
737
+ 705 train 7.590787 (lr=4.2876e-05) (hash(x)=26192886)
738
+ 706 train 7.566239 (lr=4.2685e-05) (hash(x)=25770338)
739
+ 707 train 7.469866 (lr=4.2495e-05) (hash(x)=24614912)
740
+ 708 train 7.530090 (lr=4.2305e-05) (hash(x)=26934071)
741
+ 709 train 7.505883 (lr=4.2115e-05) (hash(x)=25397093)
742
+ 710 train 7.649030 (lr=4.1925e-05) (hash(x)=25925963)
743
+ 711 train 7.458528 (lr=4.1736e-05) (hash(x)=19811802)
744
+ 712 train 7.562682 (lr=4.1547e-05) (hash(x)=23248423)
745
+ 713 train 7.530380 (lr=4.1358e-05) (hash(x)=25579655)
746
+ 714 train 7.705186 (lr=4.1170e-05) (hash(x)=25537603)
747
+ 715 train 7.613539 (lr=4.0981e-05) (hash(x)=26096514)
748
+ 716 train 7.624192 (lr=4.0793e-05) (hash(x)=24132888)
749
+ 717 train 7.495349 (lr=4.0606e-05) (hash(x)=22589633)
750
+ 718 train 7.463028 (lr=4.0418e-05) (hash(x)=27781566)
751
+ 719 train 7.705812 (lr=4.0231e-05) (hash(x)=26757147)
752
+ 720 train 7.745534 (lr=4.0044e-05) (hash(x)=26649864)
753
+ 721 train 7.600248 (lr=3.9857e-05) (hash(x)=24963730)
754
+ 722 train 7.602943 (lr=3.9671e-05) (hash(x)=25249959)
755
+ 723 train 7.458243 (lr=3.9485e-05) (hash(x)=27703124)
756
+ 724 train 7.464323 (lr=3.9299e-05) (hash(x)=26459082)
757
+ 725 train 7.420321 (lr=3.9114e-05) (hash(x)=24781713)
758
+ 726 train 7.725396 (lr=3.8929e-05) (hash(x)=26597693)
759
+ 727 train 7.523052 (lr=3.8744e-05) (hash(x)=24720476)
760
+ 728 train 7.545893 (lr=3.8560e-05) (hash(x)=23691295)
761
+ 729 train 7.655793 (lr=3.8376e-05) (hash(x)=23481301)
762
+ 730 train 7.549343 (lr=3.8192e-05) (hash(x)=23295676)
763
+ 731 train 7.492396 (lr=3.8008e-05) (hash(x)=25641033)
764
+ 732 train 7.598472 (lr=3.7825e-05) (hash(x)=22441379)
765
+ 733 train 7.511992 (lr=3.7642e-05) (hash(x)=27369437)
766
+ 734 train 7.539789 (lr=3.7460e-05) (hash(x)=24482272)
767
+ 735 train 7.659534 (lr=3.7277e-05) (hash(x)=24760900)
768
+ 736 train 7.557656 (lr=3.7096e-05) (hash(x)=23919156)
769
+ 737 train 7.447842 (lr=3.6914e-05) (hash(x)=21692546)
770
+ 738 train 7.501382 (lr=3.6733e-05) (hash(x)=27692087)
771
+ 739 train 7.633173 (lr=3.6552e-05) (hash(x)=22720391)
772
+ 740 train 7.448800 (lr=3.6372e-05) (hash(x)=26146560)
773
+ 741 train 7.578500 (lr=3.6192e-05) (hash(x)=24376979)
774
+ 742 train 7.586529 (lr=3.6012e-05) (hash(x)=22917712)
775
+ 743 train 7.478118 (lr=3.5833e-05) (hash(x)=22244509)
776
+ 744 train 7.480198 (lr=3.5654e-05) (hash(x)=25243385)
777
+ 745 train 7.484527 (lr=3.5475e-05) (hash(x)=21831960)
778
+ 746 train 7.493180 (lr=3.5297e-05) (hash(x)=24842373)
779
+ 747 train 7.422365 (lr=3.5119e-05) (hash(x)=24921535)
780
+ 748 train 7.491825 (lr=3.4942e-05) (hash(x)=24344615)
781
+ 749 train 7.499056 (lr=3.4765e-05) (hash(x)=25381251)
782
+ 750 val loss 7.5999
783
+ 750 val perplexity 1997.9175
784
+ 750 train 7.403535 (lr=3.4588e-05) (hash(x)=22887303)
785
+ 751 train 7.585867 (lr=3.4412e-05) (hash(x)=25687352)
786
+ 752 train 7.564346 (lr=3.4236e-05) (hash(x)=27791412)
787
+ 753 train 7.466172 (lr=3.4060e-05) (hash(x)=25221654)
788
+ 754 train 7.576991 (lr=3.3885e-05) (hash(x)=23243731)
789
+ 755 train 7.536428 (lr=3.3711e-05) (hash(x)=24469863)
790
+ 756 train 7.674136 (lr=3.3537e-05) (hash(x)=22484936)
791
+ 757 train 7.566036 (lr=3.3363e-05) (hash(x)=25291413)
792
+ 758 train 7.467584 (lr=3.3189e-05) (hash(x)=23952225)
793
+ 759 train 7.561107 (lr=3.3016e-05) (hash(x)=25744274)
794
+ 760 train 7.443997 (lr=3.2844e-05) (hash(x)=22616739)
795
+ 761 train 7.439696 (lr=3.2672e-05) (hash(x)=24246963)
796
+ 762 train 7.434772 (lr=3.2500e-05) (hash(x)=19407094)
797
+ 763 train 7.517981 (lr=3.2329e-05) (hash(x)=19724058)
798
+ 764 train 7.362933 (lr=3.2158e-05) (hash(x)=18836491)
799
+ 765 train 7.420434 (lr=3.1988e-05) (hash(x)=20705649)
800
+ 766 train 7.645336 (lr=3.1818e-05) (hash(x)=21325875)
801
+ 767 train 7.666421 (lr=3.1648e-05) (hash(x)=24938685)
802
+ 768 train 7.630509 (lr=3.1479e-05) (hash(x)=25152403)
803
+ 769 train 7.762554 (lr=3.1311e-05) (hash(x)=21707000)
804
+ 770 train 7.450766 (lr=3.1142e-05) (hash(x)=17952018)
805
+ 771 train 7.507354 (lr=3.0975e-05) (hash(x)=24948980)
806
+ 772 train 7.957976 (lr=3.0808e-05) (hash(x)=28012110)
807
+ 773 train 7.581994 (lr=3.0641e-05) (hash(x)=30773958)
808
+ 774 train 8.110618 (lr=3.0475e-05) (hash(x)=29481068)
809
+ 775 train 7.572071 (lr=3.0309e-05) (hash(x)=26769046)
810
+ 776 train 7.512348 (lr=3.0144e-05) (hash(x)=26755048)
811
+ 777 train 7.514757 (lr=2.9979e-05) (hash(x)=18849656)
812
+ 778 train 7.807108 (lr=2.9814e-05) (hash(x)=25449800)
813
+ 779 train 8.299426 (lr=2.9651e-05) (hash(x)=24624950)
814
+ 780 train 7.532690 (lr=2.9487e-05) (hash(x)=26174069)
815
+ 781 train 7.715192 (lr=2.9324e-05) (hash(x)=27299605)
816
+ 782 train 7.624476 (lr=2.9162e-05) (hash(x)=24777273)
817
+ 783 train 7.616799 (lr=2.9000e-05) (hash(x)=24113445)
818
+ 784 train 7.503209 (lr=2.8839e-05) (hash(x)=23970384)
819
+ 785 train 7.747850 (lr=2.8678e-05) (hash(x)=32644465)
820
+ 786 train 7.863757 (lr=2.8518e-05) (hash(x)=29951548)
821
+ 787 train 7.601475 (lr=2.8358e-05) (hash(x)=25829219)
822
+ 788 train 7.635177 (lr=2.8199e-05) (hash(x)=20556094)
823
+ 789 train 7.459713 (lr=2.8040e-05) (hash(x)=25194550)
824
+ 790 train 7.603440 (lr=2.7881e-05) (hash(x)=28482186)
825
+ 791 train 7.570147 (lr=2.7724e-05) (hash(x)=25941804)
826
+ 792 train 7.582454 (lr=2.7566e-05) (hash(x)=27049208)
827
+ 793 train 7.597131 (lr=2.7410e-05) (hash(x)=24075727)
828
+ 794 train 7.660492 (lr=2.7254e-05) (hash(x)=25082806)
829
+ 795 train 7.633698 (lr=2.7098e-05) (hash(x)=23104539)
830
+ 796 train 7.516155 (lr=2.6943e-05) (hash(x)=25351113)
831
+ 797 train 7.622557 (lr=2.6788e-05) (hash(x)=23018983)
832
+ 798 train 7.568786 (lr=2.6634e-05) (hash(x)=25302106)
833
+ 799 train 7.489184 (lr=2.6481e-05) (hash(x)=24422929)
834
+ 800 val loss 7.5857
835
+ 800 val perplexity 1969.8767
836
+ 800 train 7.568878 (lr=2.6328e-05) (hash(x)=29052117)
837
+ 801 train 7.565760 (lr=2.6176e-05) (hash(x)=23389696)
838
+ 802 train 7.495582 (lr=2.6024e-05) (hash(x)=23962503)
839
+ 803 train 7.573020 (lr=2.5873e-05) (hash(x)=29102969)
840
+ 804 train 7.712683 (lr=2.5722e-05) (hash(x)=24948166)
841
+ 805 train 7.638350 (lr=2.5572e-05) (hash(x)=28359859)
842
+ 806 train 7.641867 (lr=2.5423e-05) (hash(x)=22360298)
843
+ 807 train 7.745175 (lr=2.5274e-05) (hash(x)=24630955)
844
+ 808 train 7.747856 (lr=2.5125e-05) (hash(x)=26897402)
845
+ 809 train 7.701249 (lr=2.4978e-05) (hash(x)=22969113)
846
+ 810 train 7.542614 (lr=2.4830e-05) (hash(x)=25205781)
847
+ 811 train 7.546130 (lr=2.4684e-05) (hash(x)=27004415)
848
+ 812 train 7.420169 (lr=2.4538e-05) (hash(x)=22974875)
849
+ 813 train 7.504194 (lr=2.4392e-05) (hash(x)=22265063)
850
+ 814 train 7.422019 (lr=2.4247e-05) (hash(x)=23312114)
851
+ 815 train 7.561425 (lr=2.4103e-05) (hash(x)=26905990)
852
+ 816 train 7.562366 (lr=2.3959e-05) (hash(x)=27089280)
853
+ 817 train 7.511436 (lr=2.3816e-05) (hash(x)=24047679)
854
+ 818 train 7.578308 (lr=2.3674e-05) (hash(x)=25034684)
855
+ 819 train 7.773652 (lr=2.3532e-05) (hash(x)=26905582)
856
+ 820 train 7.433074 (lr=2.3391e-05) (hash(x)=21406950)
857
+ 821 train 7.433749 (lr=2.3250e-05) (hash(x)=29018970)
858
+ 822 train 7.416494 (lr=2.3110e-05) (hash(x)=20853530)
859
+ 823 train 7.501361 (lr=2.2971e-05) (hash(x)=24503315)
860
+ 824 train 7.530054 (lr=2.2832e-05) (hash(x)=23080074)
861
+ 825 train 7.398076 (lr=2.2694e-05) (hash(x)=22607537)
862
+ 826 train 7.808821 (lr=2.2556e-05) (hash(x)=32179773)
863
+ 827 train 8.881022 (lr=2.2420e-05) (hash(x)=26690225)
864
+ 828 train 7.569139 (lr=2.2283e-05) (hash(x)=22109609)
865
+ 829 train 7.520998 (lr=2.2148e-05) (hash(x)=24226255)
866
+ 830 train 7.595139 (lr=2.2013e-05) (hash(x)=23013573)
867
+ 831 train 7.396942 (lr=2.1878e-05) (hash(x)=23622702)
868
+ 832 train 7.586246 (lr=2.1745e-05) (hash(x)=21517429)
869
+ 833 train 7.535779 (lr=2.1612e-05) (hash(x)=23910425)
870
+ 834 train 7.634594 (lr=2.1479e-05) (hash(x)=24407484)
871
+ 835 train 7.919840 (lr=2.1347e-05) (hash(x)=27213318)
872
+ 836 train 7.554382 (lr=2.1216e-05) (hash(x)=25524191)
873
+ 837 train 7.586559 (lr=2.1086e-05) (hash(x)=26701577)
874
+ 838 train 7.432321 (lr=2.0956e-05) (hash(x)=22091266)
875
+ 839 train 7.562690 (lr=2.0827e-05) (hash(x)=26216794)
876
+ 840 train 7.482749 (lr=2.0698e-05) (hash(x)=20817340)
877
+ 841 train 7.944189 (lr=2.0570e-05) (hash(x)=27579623)
878
+ 842 train 7.412446 (lr=2.0443e-05) (hash(x)=21532187)
879
+ 843 train 7.714222 (lr=2.0317e-05) (hash(x)=26086198)
880
+ 844 train 7.532524 (lr=2.0191e-05) (hash(x)=23911729)
881
+ 845 train 7.515644 (lr=2.0066e-05) (hash(x)=24863441)
882
+ 846 train 7.968708 (lr=1.9941e-05) (hash(x)=28856978)
883
+ 847 train 7.363721 (lr=1.9818e-05) (hash(x)=25236385)
884
+ 848 train 7.665146 (lr=1.9694e-05) (hash(x)=24278687)
885
+ 849 train 7.738198 (lr=1.9572e-05) (hash(x)=25330803)
886
+ 850 val loss 7.5848
887
+ 850 val perplexity 1968.1276
888
+ 850 train 7.615569 (lr=1.9450e-05) (hash(x)=21568545)
889
+ 851 train 7.526470 (lr=1.9329e-05) (hash(x)=22508532)
890
+ 852 train 7.561185 (lr=1.9209e-05) (hash(x)=24850662)
891
+ 853 train 7.562243 (lr=1.9089e-05) (hash(x)=21799261)
892
+ 854 train 7.747930 (lr=1.8970e-05) (hash(x)=24884891)
893
+ 855 train 7.555419 (lr=1.8852e-05) (hash(x)=26311168)
894
+ 856 train 7.597106 (lr=1.8734e-05) (hash(x)=27549895)
895
+ 857 train 7.283412 (lr=1.8618e-05) (hash(x)=20977430)
896
+ 858 train 7.413096 (lr=1.8501e-05) (hash(x)=23448855)
897
+ 859 train 7.614607 (lr=1.8386e-05) (hash(x)=26941617)
898
+ 860 train 7.789804 (lr=1.8271e-05) (hash(x)=26953192)
899
+ 861 train 7.770515 (lr=1.8157e-05) (hash(x)=30008957)
900
+ 862 train 7.683902 (lr=1.8044e-05) (hash(x)=27069893)
901
+ 863 train 7.569233 (lr=1.7931e-05) (hash(x)=27125962)
902
+ 864 train 7.398902 (lr=1.7819e-05) (hash(x)=23341415)
903
+ 865 train 7.524310 (lr=1.7708e-05) (hash(x)=25888358)
904
+ 866 train 7.689504 (lr=1.7598e-05) (hash(x)=26761645)
905
+ 867 train 7.572095 (lr=1.7488e-05) (hash(x)=21725719)
906
+ 868 train 7.472510 (lr=1.7379e-05) (hash(x)=25746493)
907
+ 869 train 7.501115 (lr=1.7271e-05) (hash(x)=21669704)
908
+ 870 train 7.416508 (lr=1.7163e-05) (hash(x)=21167493)
909
+ 871 train 7.515857 (lr=1.7056e-05) (hash(x)=23166993)
910
+ 872 train 7.473018 (lr=1.6950e-05) (hash(x)=23947208)
911
+ 873 train 7.600850 (lr=1.6845e-05) (hash(x)=26235974)
912
+ 874 train 7.926110 (lr=1.6740e-05) (hash(x)=29822604)
913
+ 875 train 7.467489 (lr=1.6636e-05) (hash(x)=25426430)
914
+ 876 train 7.470237 (lr=1.6533e-05) (hash(x)=26043735)
915
+ 877 train 7.418916 (lr=1.6431e-05) (hash(x)=22966314)
916
+ 878 train 7.701723 (lr=1.6329e-05) (hash(x)=27944619)
917
+ 879 train 7.468981 (lr=1.6228e-05) (hash(x)=23448505)
918
+ 880 train 7.494925 (lr=1.6128e-05) (hash(x)=24205942)
919
+ 881 train 7.390527 (lr=1.6029e-05) (hash(x)=26330693)
920
+ 882 train 7.409158 (lr=1.5930e-05) (hash(x)=24844466)
921
+ 883 train 7.378872 (lr=1.5832e-05) (hash(x)=29550596)
922
+ 884 train 7.461069 (lr=1.5735e-05) (hash(x)=20506540)
923
+ 885 train 7.481843 (lr=1.5639e-05) (hash(x)=22124892)
924
+ 886 train 7.482552 (lr=1.5543e-05) (hash(x)=24867036)
925
+ 887 train 7.504941 (lr=1.5449e-05) (hash(x)=25607640)
926
+ 888 train 7.499252 (lr=1.5354e-05) (hash(x)=26176930)
927
+ 889 train 7.215381 (lr=1.5261e-05) (hash(x)=23795894)
928
+ 890 train 7.144152 (lr=1.5169e-05) (hash(x)=23756298)
929
+ 891 train 7.376658 (lr=1.5077e-05) (hash(x)=24165449)
930
+ 892 train 7.420582 (lr=1.4986e-05) (hash(x)=26643739)
931
+ 893 train 7.605301 (lr=1.4896e-05) (hash(x)=24697756)
932
+ 894 train 7.499039 (lr=1.4806e-05) (hash(x)=24431101)
933
+ 895 train 7.675683 (lr=1.4718e-05) (hash(x)=27218899)
934
+ 896 train 7.448135 (lr=1.4630e-05) (hash(x)=29462219)
935
+ 897 train 7.516937 (lr=1.4543e-05) (hash(x)=24748522)
936
+ 898 train 7.500336 (lr=1.4456e-05) (hash(x)=24843486)
937
+ 899 train 7.653016 (lr=1.4371e-05) (hash(x)=24965317)
938
+ 900 val loss 7.5805
939
+ 900 val perplexity 1959.5200
940
+ 900 train 7.390455 (lr=1.4286e-05) (hash(x)=26584754)
941
+ 901 train 7.309467 (lr=1.4202e-05) (hash(x)=22031210)
942
+ 902 train 7.507323 (lr=1.4119e-05) (hash(x)=24860582)
943
+ 903 train 7.442982 (lr=1.4037e-05) (hash(x)=25877990)
944
+ 904 train 7.523367 (lr=1.3955e-05) (hash(x)=24970646)
945
+ 905 train 7.550287 (lr=1.3874e-05) (hash(x)=24382381)
946
+ 906 train 7.433103 (lr=1.3794e-05) (hash(x)=25221746)
947
+ 907 train 7.426545 (lr=1.3715e-05) (hash(x)=24551999)
948
+ 908 train 7.455559 (lr=1.3637e-05) (hash(x)=28151378)
949
+ 909 train 7.608099 (lr=1.3559e-05) (hash(x)=25765516)
950
+ 910 train 7.566759 (lr=1.3482e-05) (hash(x)=25630696)
951
+ 911 train 7.546997 (lr=1.3407e-05) (hash(x)=25204247)
952
+ 912 train 7.625235 (lr=1.3331e-05) (hash(x)=26326447)
953
+ 913 train 7.584642 (lr=1.3257e-05) (hash(x)=26321813)
954
+ 914 train 7.520039 (lr=1.3183e-05) (hash(x)=25400396)
955
+ 915 train 7.540230 (lr=1.3111e-05) (hash(x)=25528323)
956
+ 916 train 7.528754 (lr=1.3039e-05) (hash(x)=24035353)
957
+ 917 train 7.545255 (lr=1.2968e-05) (hash(x)=25114729)
958
+ 918 train 7.437885 (lr=1.2897e-05) (hash(x)=21634962)
959
+ 919 train 7.578315 (lr=1.2828e-05) (hash(x)=31341964)
960
+ 920 train 7.485369 (lr=1.2759e-05) (hash(x)=29172813)
961
+ 921 train 7.637546 (lr=1.2691e-05) (hash(x)=26326024)
962
+ 922 train 7.395048 (lr=1.2624e-05) (hash(x)=20140071)
963
+ 923 train 7.539844 (lr=1.2558e-05) (hash(x)=27273311)
964
+ 924 train 7.615642 (lr=1.2493e-05) (hash(x)=31276487)
965
+ 925 train 7.311555 (lr=1.2428e-05) (hash(x)=17819313)
966
+ 926 train 7.652135 (lr=1.2364e-05) (hash(x)=25843618)
967
+ 927 train 7.601595 (lr=1.2301e-05) (hash(x)=23769243)
968
+ 928 train 7.514042 (lr=1.2239e-05) (hash(x)=28731298)
969
+ 929 train 7.657512 (lr=1.2178e-05) (hash(x)=23788738)
970
+ 930 train 7.464962 (lr=1.2118e-05) (hash(x)=26599031)
971
+ 931 train 7.573407 (lr=1.2058e-05) (hash(x)=24818744)
972
+ 932 train 7.474153 (lr=1.1999e-05) (hash(x)=23053871)
973
+ 933 train 7.430338 (lr=1.1941e-05) (hash(x)=23472795)
974
+ 934 train 8.027657 (lr=1.1884e-05) (hash(x)=29748225)
975
+ 935 train 7.720096 (lr=1.1828e-05) (hash(x)=24991832)
976
+ 936 train 7.696406 (lr=1.1772e-05) (hash(x)=23681545)
977
+ 937 train 7.490154 (lr=1.1718e-05) (hash(x)=22424066)
978
+ 938 train 7.499488 (lr=1.1664e-05) (hash(x)=25262748)
979
+ 939 train 7.481741 (lr=1.1611e-05) (hash(x)=26474878)
980
+ 940 train 7.688588 (lr=1.1559e-05) (hash(x)=23930593)
981
+ 941 train 7.485000 (lr=1.1508e-05) (hash(x)=22704349)
982
+ 942 train 7.552230 (lr=1.1457e-05) (hash(x)=24801048)
983
+ 943 train 7.501627 (lr=1.1408e-05) (hash(x)=22293673)
984
+ 944 train 7.567864 (lr=1.1359e-05) (hash(x)=23311934)
985
+ 945 train 7.566318 (lr=1.1311e-05) (hash(x)=21785487)
986
+ 946 train 7.430105 (lr=1.1264e-05) (hash(x)=22981456)
987
+ 947 train 7.690678 (lr=1.1218e-05) (hash(x)=25251063)
988
+ 948 train 7.496724 (lr=1.1173e-05) (hash(x)=25796422)
989
+ 949 train 7.410078 (lr=1.1128e-05) (hash(x)=22722211)
990
+ 950 val loss 7.5736
991
+ 950 val perplexity 1946.0972
992
+ 950 train 7.514771 (lr=1.1085e-05) (hash(x)=24296755)
993
+ 951 train 7.508792 (lr=1.1042e-05) (hash(x)=24522269)
994
+ 952 train 7.515880 (lr=1.1000e-05) (hash(x)=25617184)
995
+ 953 train 7.418437 (lr=1.0959e-05) (hash(x)=23329475)
996
+ 954 train 7.535826 (lr=1.0919e-05) (hash(x)=30652062)
997
+ 955 train 7.465960 (lr=1.0879e-05) (hash(x)=24735165)
998
+ 956 train 7.297657 (lr=1.0841e-05) (hash(x)=23795686)
999
+ 957 train 7.338871 (lr=1.0803e-05) (hash(x)=27002892)
1000
+ 958 train 7.411456 (lr=1.0766e-05) (hash(x)=21189101)
1001
+ 959 train 7.315903 (lr=1.0730e-05) (hash(x)=23499349)
1002
+ 960 train 7.916732 (lr=1.0695e-05) (hash(x)=26807896)
1003
+ 961 train 7.489057 (lr=1.0661e-05) (hash(x)=22461589)
1004
+ 962 train 7.569326 (lr=1.0628e-05) (hash(x)=26904123)
1005
+ 963 train 7.388714 (lr=1.0595e-05) (hash(x)=26415040)
1006
+ 964 train 7.680634 (lr=1.0563e-05) (hash(x)=24463181)
1007
+ 965 train 7.729903 (lr=1.0533e-05) (hash(x)=26349465)
1008
+ 966 train 7.332331 (lr=1.0503e-05) (hash(x)=21205988)
1009
+ 967 train 7.385901 (lr=1.0474e-05) (hash(x)=21596677)
1010
+ 968 train 7.503863 (lr=1.0445e-05) (hash(x)=23575666)
1011
+ 969 train 7.317882 (lr=1.0418e-05) (hash(x)=24117012)
1012
+ 970 train 7.436379 (lr=1.0391e-05) (hash(x)=26013014)
1013
+ 971 train 7.550871 (lr=1.0366e-05) (hash(x)=24105761)
1014
+ 972 train 7.416203 (lr=1.0341e-05) (hash(x)=20369133)
1015
+ 973 train 7.449209 (lr=1.0317e-05) (hash(x)=25725049)
1016
+ 974 train 7.333949 (lr=1.0294e-05) (hash(x)=19808118)
1017
+ 975 train 7.242748 (lr=1.0272e-05) (hash(x)=22892776)
1018
+ 976 train 7.423972 (lr=1.0251e-05) (hash(x)=23736780)
1019
+ 977 train 7.370919 (lr=1.0230e-05) (hash(x)=21630383)
1020
+ 978 train 7.583869 (lr=1.0211e-05) (hash(x)=25397467)
1021
+ 979 train 7.335042 (lr=1.0192e-05) (hash(x)=24727094)
1022
+ 980 train 7.406897 (lr=1.0174e-05) (hash(x)=23951182)
1023
+ 981 train 7.351386 (lr=1.0157e-05) (hash(x)=22835000)
1024
+ 982 train 7.581035 (lr=1.0141e-05) (hash(x)=28767849)
1025
+ 983 train 7.542294 (lr=1.0126e-05) (hash(x)=22839049)
1026
+ 984 train 7.453597 (lr=1.0111e-05) (hash(x)=19930900)
1027
+ 985 train 7.346589 (lr=1.0098e-05) (hash(x)=20654136)
1028
+ 986 train 7.561928 (lr=1.0085e-05) (hash(x)=28495621)
1029
+ 987 train 7.698074 (lr=1.0074e-05) (hash(x)=26776000)
1030
+ 988 train 7.367422 (lr=1.0063e-05) (hash(x)=21584429)
1031
+ 989 train 7.129485 (lr=1.0053e-05) (hash(x)=22560658)
1032
+ 990 train 7.772963 (lr=1.0044e-05) (hash(x)=25276565)
1033
+ 991 train 7.573019 (lr=1.0035e-05) (hash(x)=25995933)
1034
+ 992 train 7.633466 (lr=1.0028e-05) (hash(x)=24163910)
1035
+ 993 train 7.562485 (lr=1.0021e-05) (hash(x)=24635394)
1036
+ 994 train 7.674159 (lr=1.0016e-05) (hash(x)=22556381)
1037
+ 995 train 7.752596 (lr=1.0011e-05) (hash(x)=27537228)
1038
+ 996 train 7.535579 (lr=1.0007e-05) (hash(x)=20883633)
1039
+ 997 train 7.690817 (lr=1.0004e-05) (hash(x)=26496730)
1040
+ 998 train 7.326190 (lr=1.0002e-05) (hash(x)=20342150)
1041
+ 999 val loss 7.5717
1042
+ 999 val perplexity 1942.4972
1043
+ 999 train 7.335561 (lr=1.0000e-05) (hash(x)=16533310)
lr1e-4_total_batch_size20480_seq_len512/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dba5299212bc73baf4641b4128b4609892d223b94a8fc71552ab43d716c9d292
3
+ size 48156226
lr1e-4_total_batch_size20480_seq_len512/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e84cb6f5f282642bd6c6173c535031740aeafdffd921ded9c30c58f9ab5b3aa
3
+ size 71157574