andrew-healey commited on
Commit
e35a9db
·
verified ·
1 Parent(s): f30fbaf

Upload folder using huggingface_hub

Browse files
lr4e-4_total_batch_size20480_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr4e-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_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 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": 0.0004, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr4e-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
lr4e-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=1.3986e-06) (hash(x)=27234506)
5
+ 1 train 11.747758 (lr=2.7972e-06) (hash(x)=20782690)
6
+ 2 train 11.782544 (lr=4.1958e-06) (hash(x)=26532095)
7
+ 3 train 11.766343 (lr=5.5944e-06) (hash(x)=23890908)
8
+ 4 train 11.764227 (lr=6.9930e-06) (hash(x)=24011372)
9
+ 5 train 11.748763 (lr=8.3916e-06) (hash(x)=24611628)
10
+ 6 train 11.704254 (lr=9.7902e-06) (hash(x)=19245352)
11
+ 7 train 11.685399 (lr=1.1189e-05) (hash(x)=25767553)
12
+ 8 train 11.671986 (lr=1.2587e-05) (hash(x)=24809873)
13
+ 9 train 11.643412 (lr=1.3986e-05) (hash(x)=22512166)
14
+ 10 train 11.643576 (lr=1.5385e-05) (hash(x)=28092148)
15
+ 11 train 11.613974 (lr=1.6783e-05) (hash(x)=24732833)
16
+ 12 train 11.580401 (lr=1.8182e-05) (hash(x)=22694718)
17
+ 13 train 11.565632 (lr=1.9580e-05) (hash(x)=25129410)
18
+ 14 train 11.526360 (lr=2.0979e-05) (hash(x)=27858570)
19
+ 15 train 11.479332 (lr=2.2378e-05) (hash(x)=26582391)
20
+ 16 train 11.424595 (lr=2.3776e-05) (hash(x)=27279806)
21
+ 17 train 11.395790 (lr=2.5175e-05) (hash(x)=26472336)
22
+ 18 train 11.367981 (lr=2.6573e-05) (hash(x)=28301589)
23
+ 19 train 11.309223 (lr=2.7972e-05) (hash(x)=27654289)
24
+ 20 train 11.269543 (lr=2.9371e-05) (hash(x)=24995715)
25
+ 21 train 11.219220 (lr=3.0769e-05) (hash(x)=26559876)
26
+ 22 train 11.146831 (lr=3.2168e-05) (hash(x)=28364895)
27
+ 23 train 11.104678 (lr=3.3566e-05) (hash(x)=21770329)
28
+ 24 train 11.098833 (lr=3.4965e-05) (hash(x)=24628606)
29
+ 25 train 10.978676 (lr=3.6364e-05) (hash(x)=22723459)
30
+ 26 train 10.862663 (lr=3.7762e-05) (hash(x)=22897919)
31
+ 27 train 10.928136 (lr=3.9161e-05) (hash(x)=30338342)
32
+ 28 train 10.838531 (lr=4.0559e-05) (hash(x)=27063280)
33
+ 29 train 10.761502 (lr=4.1958e-05) (hash(x)=25678059)
34
+ 30 train 10.664656 (lr=4.3357e-05) (hash(x)=21853028)
35
+ 31 train 10.648884 (lr=4.4755e-05) (hash(x)=26109335)
36
+ 32 train 10.658296 (lr=4.6154e-05) (hash(x)=25334848)
37
+ 33 train 10.569092 (lr=4.7552e-05) (hash(x)=25044885)
38
+ 34 train 10.579533 (lr=4.8951e-05) (hash(x)=23678349)
39
+ 35 train 10.607537 (lr=5.0350e-05) (hash(x)=28870690)
40
+ 36 train 10.589782 (lr=5.1748e-05) (hash(x)=32982615)
41
+ 37 train 10.571798 (lr=5.3147e-05) (hash(x)=35279941)
42
+ 38 train 10.516623 (lr=5.4545e-05) (hash(x)=26400041)
43
+ 39 train 10.410390 (lr=5.5944e-05) (hash(x)=25015923)
44
+ 40 train 10.384533 (lr=5.7343e-05) (hash(x)=24853586)
45
+ 41 train 10.407358 (lr=5.8741e-05) (hash(x)=26740855)
46
+ 42 train 10.291468 (lr=6.0140e-05) (hash(x)=24584116)
47
+ 43 train 10.346181 (lr=6.1538e-05) (hash(x)=22262151)
48
+ 44 train 10.312892 (lr=6.2937e-05) (hash(x)=26627860)
49
+ 45 train 10.236773 (lr=6.4336e-05) (hash(x)=26621834)
50
+ 46 train 10.294384 (lr=6.5734e-05) (hash(x)=26930630)
51
+ 47 train 10.276333 (lr=6.7133e-05) (hash(x)=23234741)
52
+ 48 train 10.302801 (lr=6.8531e-05) (hash(x)=28102443)
53
+ 49 train 10.397999 (lr=6.9930e-05) (hash(x)=25470563)
54
+ 50 val loss 10.2835
55
+ 50 val perplexity 29246.5488
56
+ 50 train 10.286386 (lr=7.1329e-05) (hash(x)=28282861)
57
+ 51 train 10.313009 (lr=7.2727e-05) (hash(x)=21799102)
58
+ 52 train 10.224268 (lr=7.4126e-05) (hash(x)=23471769)
59
+ 53 train 10.227499 (lr=7.5524e-05) (hash(x)=23275479)
60
+ 54 train 10.280604 (lr=7.6923e-05) (hash(x)=24473581)
61
+ 55 train 10.263956 (lr=7.8322e-05) (hash(x)=22091862)
62
+ 56 train 10.149853 (lr=7.9720e-05) (hash(x)=26018202)
63
+ 57 train 10.147608 (lr=8.1119e-05) (hash(x)=17430373)
64
+ 58 train 10.030077 (lr=8.2517e-05) (hash(x)=24078083)
65
+ 59 train 10.124784 (lr=8.3916e-05) (hash(x)=22954861)
66
+ 60 train 10.189632 (lr=8.5315e-05) (hash(x)=28695016)
67
+ 61 train 10.131332 (lr=8.6713e-05) (hash(x)=26671799)
68
+ 62 train 10.158911 (lr=8.8112e-05) (hash(x)=23166092)
69
+ 63 train 10.013499 (lr=8.9510e-05) (hash(x)=26924723)
70
+ 64 train 10.090370 (lr=9.0909e-05) (hash(x)=25138659)
71
+ 65 train 10.093479 (lr=9.2308e-05) (hash(x)=23557495)
72
+ 66 train 10.000923 (lr=9.3706e-05) (hash(x)=25288123)
73
+ 67 train 9.819649 (lr=9.5105e-05) (hash(x)=27804380)
74
+ 68 train 10.040302 (lr=9.6503e-05) (hash(x)=26993263)
75
+ 69 train 10.062110 (lr=9.7902e-05) (hash(x)=26654908)
76
+ 70 train 9.925527 (lr=9.9301e-05) (hash(x)=27347722)
77
+ 71 train 9.996675 (lr=1.0070e-04) (hash(x)=23627518)
78
+ 72 train 9.905220 (lr=1.0210e-04) (hash(x)=21768671)
79
+ 73 train 9.902600 (lr=1.0350e-04) (hash(x)=21889990)
80
+ 74 train 9.813425 (lr=1.0490e-04) (hash(x)=23665838)
81
+ 75 train 10.022393 (lr=1.0629e-04) (hash(x)=22528001)
82
+ 76 train 9.866025 (lr=1.0769e-04) (hash(x)=22487328)
83
+ 77 train 9.917924 (lr=1.0909e-04) (hash(x)=29248942)
84
+ 78 train 9.795621 (lr=1.1049e-04) (hash(x)=23543273)
85
+ 79 train 9.772318 (lr=1.1189e-04) (hash(x)=25533417)
86
+ 80 train 9.725587 (lr=1.1329e-04) (hash(x)=25552036)
87
+ 81 train 9.666281 (lr=1.1469e-04) (hash(x)=20612533)
88
+ 82 train 9.612759 (lr=1.1608e-04) (hash(x)=20216795)
89
+ 83 train 9.740935 (lr=1.1748e-04) (hash(x)=24677740)
90
+ 84 train 9.645903 (lr=1.1888e-04) (hash(x)=16722715)
91
+ 85 train 9.665986 (lr=1.2028e-04) (hash(x)=25347203)
92
+ 86 train 9.609003 (lr=1.2168e-04) (hash(x)=24272489)
93
+ 87 train 9.543739 (lr=1.2308e-04) (hash(x)=28922363)
94
+ 88 train 9.509347 (lr=1.2448e-04) (hash(x)=25510798)
95
+ 89 train 9.539851 (lr=1.2587e-04) (hash(x)=25631397)
96
+ 90 train 9.418899 (lr=1.2727e-04) (hash(x)=21333676)
97
+ 91 train 9.523022 (lr=1.2867e-04) (hash(x)=26237983)
98
+ 92 train 9.450417 (lr=1.3007e-04) (hash(x)=25976502)
99
+ 93 train 9.445161 (lr=1.3147e-04) (hash(x)=24742645)
100
+ 94 train 9.452113 (lr=1.3287e-04) (hash(x)=27086333)
101
+ 95 train 9.403681 (lr=1.3427e-04) (hash(x)=26157660)
102
+ 96 train 9.387855 (lr=1.3566e-04) (hash(x)=23616370)
103
+ 97 train 9.161496 (lr=1.3706e-04) (hash(x)=22582169)
104
+ 98 train 9.214252 (lr=1.3846e-04) (hash(x)=26233189)
105
+ 99 train 9.274230 (lr=1.3986e-04) (hash(x)=26431960)
106
+ 100 val loss 9.2678
107
+ 100 val perplexity 10591.6914
108
+ 100 train 9.329738 (lr=1.4126e-04) (hash(x)=30707498)
109
+ 101 train 9.225431 (lr=1.4266e-04) (hash(x)=25716176)
110
+ 102 train 9.212953 (lr=1.4406e-04) (hash(x)=26256420)
111
+ 103 train 9.202058 (lr=1.4545e-04) (hash(x)=28019788)
112
+ 104 train 9.151735 (lr=1.4685e-04) (hash(x)=23510110)
113
+ 105 train 9.145987 (lr=1.4825e-04) (hash(x)=20382963)
114
+ 106 train 9.020278 (lr=1.4965e-04) (hash(x)=25339466)
115
+ 107 train 9.086815 (lr=1.5105e-04) (hash(x)=24489647)
116
+ 108 train 9.025688 (lr=1.5245e-04) (hash(x)=22927699)
117
+ 109 train 9.007959 (lr=1.5385e-04) (hash(x)=25128502)
118
+ 110 train 9.033803 (lr=1.5524e-04) (hash(x)=25808969)
119
+ 111 train 8.936629 (lr=1.5664e-04) (hash(x)=25231390)
120
+ 112 train 8.839972 (lr=1.5804e-04) (hash(x)=21154388)
121
+ 113 train 8.794649 (lr=1.5944e-04) (hash(x)=22708977)
122
+ 114 train 9.046741 (lr=1.6084e-04) (hash(x)=27258353)
123
+ 115 train 8.834043 (lr=1.6224e-04) (hash(x)=24364920)
124
+ 116 train 8.850090 (lr=1.6364e-04) (hash(x)=27243972)
125
+ 117 train 8.883260 (lr=1.6503e-04) (hash(x)=24932925)
126
+ 118 train 8.820827 (lr=1.6643e-04) (hash(x)=26371091)
127
+ 119 train 8.706933 (lr=1.6783e-04) (hash(x)=26913684)
128
+ 120 train 8.764583 (lr=1.6923e-04) (hash(x)=29239611)
129
+ 121 train 8.553930 (lr=1.7063e-04) (hash(x)=18017792)
130
+ 122 train 8.657349 (lr=1.7203e-04) (hash(x)=27574089)
131
+ 123 train 8.670589 (lr=1.7343e-04) (hash(x)=28613882)
132
+ 124 train 8.627396 (lr=1.7483e-04) (hash(x)=26981914)
133
+ 125 train 8.635017 (lr=1.7622e-04) (hash(x)=22652243)
134
+ 126 train 8.501051 (lr=1.7762e-04) (hash(x)=15446025)
135
+ 127 train 8.483841 (lr=1.7902e-04) (hash(x)=27194521)
136
+ 128 train 8.417273 (lr=1.8042e-04) (hash(x)=25264518)
137
+ 129 train 8.573944 (lr=1.8182e-04) (hash(x)=24936836)
138
+ 130 train 8.549899 (lr=1.8322e-04) (hash(x)=28197282)
139
+ 131 train 8.500765 (lr=1.8462e-04) (hash(x)=24765578)
140
+ 132 train 8.494844 (lr=1.8601e-04) (hash(x)=27448185)
141
+ 133 train 8.527194 (lr=1.8741e-04) (hash(x)=23107416)
142
+ 134 train 8.411929 (lr=1.8881e-04) (hash(x)=26755138)
143
+ 135 train 8.388552 (lr=1.9021e-04) (hash(x)=29314255)
144
+ 136 train 8.368528 (lr=1.9161e-04) (hash(x)=22112669)
145
+ 137 train 8.341428 (lr=1.9301e-04) (hash(x)=24013079)
146
+ 138 train 8.299202 (lr=1.9441e-04) (hash(x)=22537194)
147
+ 139 train 8.331915 (lr=1.9580e-04) (hash(x)=25322001)
148
+ 140 train 8.405149 (lr=1.9720e-04) (hash(x)=27028126)
149
+ 141 train 8.383682 (lr=1.9860e-04) (hash(x)=26690440)
150
+ 142 train 8.324874 (lr=2.0000e-04) (hash(x)=26967331)
151
+ 143 train 8.256083 (lr=2.0140e-04) (hash(x)=22598512)
152
+ 144 train 8.263647 (lr=2.0280e-04) (hash(x)=24697985)
153
+ 145 train 8.125901 (lr=2.0420e-04) (hash(x)=18473911)
154
+ 146 train 8.190416 (lr=2.0559e-04) (hash(x)=23172124)
155
+ 147 train 8.542480 (lr=2.0699e-04) (hash(x)=22322442)
156
+ 148 train 8.411720 (lr=2.0839e-04) (hash(x)=29659121)
157
+ 149 train 8.253712 (lr=2.0979e-04) (hash(x)=26324153)
158
+ 150 val loss 8.1874
159
+ 150 val perplexity 3595.3342
160
+ 150 train 8.204757 (lr=2.1119e-04) (hash(x)=26939970)
161
+ 151 train 8.212992 (lr=2.1259e-04) (hash(x)=23446058)
162
+ 152 train 8.250001 (lr=2.1399e-04) (hash(x)=29981801)
163
+ 153 train 8.289589 (lr=2.1538e-04) (hash(x)=24843741)
164
+ 154 train 8.175900 (lr=2.1678e-04) (hash(x)=22855363)
165
+ 155 train 8.072334 (lr=2.1818e-04) (hash(x)=23197102)
166
+ 156 train 8.196078 (lr=2.1958e-04) (hash(x)=25704919)
167
+ 157 train 8.449030 (lr=2.2098e-04) (hash(x)=23054940)
168
+ 158 train 8.448878 (lr=2.2238e-04) (hash(x)=24909904)
169
+ 159 train 8.040490 (lr=2.2378e-04) (hash(x)=27140757)
170
+ 160 train 7.978746 (lr=2.2517e-04) (hash(x)=25993110)
171
+ 161 train 7.988008 (lr=2.2657e-04) (hash(x)=25149419)
172
+ 162 train 8.142767 (lr=2.2797e-04) (hash(x)=27623643)
173
+ 163 train 8.080926 (lr=2.2937e-04) (hash(x)=24727764)
174
+ 164 train 7.677405 (lr=2.3077e-04) (hash(x)=23155773)
175
+ 165 train 7.914826 (lr=2.3217e-04) (hash(x)=21935931)
176
+ 166 train 7.806148 (lr=2.3357e-04) (hash(x)=20054917)
177
+ 167 train 8.031744 (lr=2.3497e-04) (hash(x)=27606073)
178
+ 168 train 8.084712 (lr=2.3636e-04) (hash(x)=26534487)
179
+ 169 train 7.926521 (lr=2.3776e-04) (hash(x)=22450341)
180
+ 170 train 8.066320 (lr=2.3916e-04) (hash(x)=25013073)
181
+ 171 train 8.197744 (lr=2.4056e-04) (hash(x)=25945859)
182
+ 172 train 7.882259 (lr=2.4196e-04) (hash(x)=24418091)
183
+ 173 train 7.898903 (lr=2.4336e-04) (hash(x)=21613707)
184
+ 174 train 7.969786 (lr=2.4476e-04) (hash(x)=26386157)
185
+ 175 train 7.921897 (lr=2.4615e-04) (hash(x)=24396519)
186
+ 176 train 7.942742 (lr=2.4755e-04) (hash(x)=23265507)
187
+ 177 train 7.991883 (lr=2.4895e-04) (hash(x)=29138828)
188
+ 178 train 7.980258 (lr=2.5035e-04) (hash(x)=24265353)
189
+ 179 train 7.972139 (lr=2.5175e-04) (hash(x)=24685515)
190
+ 180 train 7.855184 (lr=2.5315e-04) (hash(x)=24876269)
191
+ 181 train 7.882731 (lr=2.5455e-04) (hash(x)=22825749)
192
+ 182 train 7.911975 (lr=2.5594e-04) (hash(x)=21239652)
193
+ 183 train 7.908867 (lr=2.5734e-04) (hash(x)=28767869)
194
+ 184 train 7.940979 (lr=2.5874e-04) (hash(x)=25614301)
195
+ 185 train 7.922630 (lr=2.6014e-04) (hash(x)=21159060)
196
+ 186 train 7.854911 (lr=2.6154e-04) (hash(x)=26751788)
197
+ 187 train 7.854751 (lr=2.6294e-04) (hash(x)=25045397)
198
+ 188 train 7.940743 (lr=2.6434e-04) (hash(x)=29989709)
199
+ 189 train 8.043810 (lr=2.6573e-04) (hash(x)=25560928)
200
+ 190 train 7.964288 (lr=2.6713e-04) (hash(x)=26062687)
201
+ 191 train 8.024805 (lr=2.6853e-04) (hash(x)=28959222)
202
+ 192 train 7.805037 (lr=2.6993e-04) (hash(x)=25434884)
203
+ 193 train 7.880256 (lr=2.7133e-04) (hash(x)=21248405)
204
+ 194 train 7.824596 (lr=2.7273e-04) (hash(x)=25044192)
205
+ 195 train 7.861133 (lr=2.7413e-04) (hash(x)=25830182)
206
+ 196 train 7.855579 (lr=2.7552e-04) (hash(x)=26621419)
207
+ 197 train 7.626330 (lr=2.7692e-04) (hash(x)=25173113)
208
+ 198 train 7.775539 (lr=2.7832e-04) (hash(x)=24400133)
209
+ 199 train 7.767868 (lr=2.7972e-04) (hash(x)=23486277)
210
+ 200 val loss 7.7680
211
+ 200 val perplexity 2363.7957
212
+ 200 train 7.767083 (lr=2.8112e-04) (hash(x)=25978130)
213
+ 201 train 7.822398 (lr=2.8252e-04) (hash(x)=28916006)
214
+ 202 train 7.968693 (lr=2.8392e-04) (hash(x)=33195100)
215
+ 203 train 7.941406 (lr=2.8531e-04) (hash(x)=25556929)
216
+ 204 train 7.850338 (lr=2.8671e-04) (hash(x)=29536986)
217
+ 205 train 7.841481 (lr=2.8811e-04) (hash(x)=29648798)
218
+ 206 train 7.746860 (lr=2.8951e-04) (hash(x)=23124767)
219
+ 207 train 7.742850 (lr=2.9091e-04) (hash(x)=24976217)
220
+ 208 train 7.795084 (lr=2.9231e-04) (hash(x)=23632825)
221
+ 209 train 7.879206 (lr=2.9371e-04) (hash(x)=31019606)
222
+ 210 train 7.868719 (lr=2.9510e-04) (hash(x)=28236580)
223
+ 211 train 7.795888 (lr=2.9650e-04) (hash(x)=28269421)
224
+ 212 train 7.589768 (lr=2.9790e-04) (hash(x)=25563279)
225
+ 213 train 7.843002 (lr=2.9930e-04) (hash(x)=24865358)
226
+ 214 train 7.587805 (lr=3.0070e-04) (hash(x)=22155546)
227
+ 215 train 7.723507 (lr=3.0210e-04) (hash(x)=26984272)
228
+ 216 train 7.743106 (lr=3.0350e-04) (hash(x)=23193673)
229
+ 217 train 7.812387 (lr=3.0490e-04) (hash(x)=24544116)
230
+ 218 train 7.749648 (lr=3.0629e-04) (hash(x)=24547533)
231
+ 219 train 7.725641 (lr=3.0769e-04) (hash(x)=22059850)
232
+ 220 train 7.794691 (lr=3.0909e-04) (hash(x)=25994573)
233
+ 221 train 7.750957 (lr=3.1049e-04) (hash(x)=23926632)
234
+ 222 train 7.617449 (lr=3.1189e-04) (hash(x)=27111369)
235
+ 223 train 7.772119 (lr=3.1329e-04) (hash(x)=27025333)
236
+ 224 train 7.574084 (lr=3.1469e-04) (hash(x)=16908068)
237
+ 225 train 7.741133 (lr=3.1608e-04) (hash(x)=29120407)
238
+ 226 train 7.820401 (lr=3.1748e-04) (hash(x)=27092710)
239
+ 227 train 7.627608 (lr=3.1888e-04) (hash(x)=23827393)
240
+ 228 train 7.676267 (lr=3.2028e-04) (hash(x)=32812727)
241
+ 229 train 7.705334 (lr=3.2168e-04) (hash(x)=27230027)
242
+ 230 train 7.702010 (lr=3.2308e-04) (hash(x)=25703381)
243
+ 231 train 7.649456 (lr=3.2448e-04) (hash(x)=24968260)
244
+ 232 train 7.628579 (lr=3.2587e-04) (hash(x)=25676468)
245
+ 233 train 7.719288 (lr=3.2727e-04) (hash(x)=27544803)
246
+ 234 train 7.692404 (lr=3.2867e-04) (hash(x)=24331407)
247
+ 235 train 7.666219 (lr=3.3007e-04) (hash(x)=26540663)
248
+ 236 train 7.694482 (lr=3.3147e-04) (hash(x)=25532657)
249
+ 237 train 7.884754 (lr=3.3287e-04) (hash(x)=25856625)
250
+ 238 train 7.638366 (lr=3.3427e-04) (hash(x)=23437426)
251
+ 239 train 7.691709 (lr=3.3566e-04) (hash(x)=23164356)
252
+ 240 train 7.693223 (lr=3.3706e-04) (hash(x)=25957896)
253
+ 241 train 7.755221 (lr=3.3846e-04) (hash(x)=25646282)
254
+ 242 train 7.756562 (lr=3.3986e-04) (hash(x)=21513584)
255
+ 243 train 7.648728 (lr=3.4126e-04) (hash(x)=23843233)
256
+ 244 train 7.638216 (lr=3.4266e-04) (hash(x)=23226697)
257
+ 245 train 7.498558 (lr=3.4406e-04) (hash(x)=24506028)
258
+ 246 train 7.625026 (lr=3.4545e-04) (hash(x)=26024321)
259
+ 247 train 7.738388 (lr=3.4685e-04) (hash(x)=22781277)
260
+ 248 train 7.713893 (lr=3.4825e-04) (hash(x)=24451843)
261
+ 249 train 7.643312 (lr=3.4965e-04) (hash(x)=24948650)
262
+ 250 val loss 7.6682
263
+ 250 val perplexity 2139.1968
264
+ 250 train 7.598648 (lr=3.5105e-04) (hash(x)=25559534)
265
+ 251 train 7.708572 (lr=3.5245e-04) (hash(x)=26373991)
266
+ 252 train 7.763048 (lr=3.5385e-04) (hash(x)=27756673)
267
+ 253 train 7.720646 (lr=3.5524e-04) (hash(x)=23402396)
268
+ 254 train 7.749294 (lr=3.5664e-04) (hash(x)=28250354)
269
+ 255 train 7.660928 (lr=3.5804e-04) (hash(x)=21562133)
270
+ 256 train 7.826489 (lr=3.5944e-04) (hash(x)=27253861)
271
+ 257 train 7.725481 (lr=3.6084e-04) (hash(x)=25009210)
272
+ 258 train 7.610504 (lr=3.6224e-04) (hash(x)=23052577)
273
+ 259 train 7.790030 (lr=3.6364e-04) (hash(x)=27161811)
274
+ 260 train 7.701631 (lr=3.6503e-04) (hash(x)=27302459)
275
+ 261 train 7.786455 (lr=3.6643e-04) (hash(x)=29515100)
276
+ 262 train 7.647666 (lr=3.6783e-04) (hash(x)=25318001)
277
+ 263 train 7.684634 (lr=3.6923e-04) (hash(x)=25305929)
278
+ 264 train 7.681130 (lr=3.7063e-04) (hash(x)=24339043)
279
+ 265 train 7.734348 (lr=3.7203e-04) (hash(x)=26808543)
280
+ 266 train 7.513172 (lr=3.7343e-04) (hash(x)=25763991)
281
+ 267 train 7.729214 (lr=3.7483e-04) (hash(x)=25669509)
282
+ 268 train 7.758537 (lr=3.7622e-04) (hash(x)=27676869)
283
+ 269 train 7.695546 (lr=3.7762e-04) (hash(x)=28435805)
284
+ 270 train 7.619259 (lr=3.7902e-04) (hash(x)=23455211)
285
+ 271 train 7.798424 (lr=3.8042e-04) (hash(x)=26482758)
286
+ 272 train 7.615776 (lr=3.8182e-04) (hash(x)=27068280)
287
+ 273 train 7.667317 (lr=3.8322e-04) (hash(x)=24929929)
288
+ 274 train 7.633095 (lr=3.8462e-04) (hash(x)=25373676)
289
+ 275 train 7.602971 (lr=3.8601e-04) (hash(x)=25043037)
290
+ 276 train 7.666723 (lr=3.8741e-04) (hash(x)=27906976)
291
+ 277 train 8.023095 (lr=3.8881e-04) (hash(x)=29151257)
292
+ 278 train 7.734791 (lr=3.9021e-04) (hash(x)=26405613)
293
+ 279 train 7.493926 (lr=3.9161e-04) (hash(x)=26400365)
294
+ 280 train 7.626794 (lr=3.9301e-04) (hash(x)=23971905)
295
+ 281 train 7.563231 (lr=3.9441e-04) (hash(x)=24098812)
296
+ 282 train 7.707642 (lr=3.9580e-04) (hash(x)=29402976)
297
+ 283 train 7.494372 (lr=3.9720e-04) (hash(x)=20967666)
298
+ 284 train 7.587188 (lr=3.9860e-04) (hash(x)=24287610)
299
+ 285 train 7.937821 (lr=4.0000e-04) (hash(x)=28667963)
300
+ 286 train 7.676493 (lr=4.0000e-04) (hash(x)=25974316)
301
+ 287 train 7.789766 (lr=4.0000e-04) (hash(x)=27083417)
302
+ 288 train 7.616703 (lr=3.9999e-04) (hash(x)=28605205)
303
+ 289 train 7.566914 (lr=3.9998e-04) (hash(x)=23649001)
304
+ 290 train 7.571834 (lr=3.9997e-04) (hash(x)=25921700)
305
+ 291 train 7.685496 (lr=3.9996e-04) (hash(x)=31338300)
306
+ 292 train 7.596811 (lr=3.9994e-04) (hash(x)=23862328)
307
+ 293 train 7.622660 (lr=3.9991e-04) (hash(x)=30603174)
308
+ 294 train 7.684536 (lr=3.9989e-04) (hash(x)=23603806)
309
+ 295 train 7.603960 (lr=3.9986e-04) (hash(x)=24830965)
310
+ 296 train 7.694208 (lr=3.9983e-04) (hash(x)=32488729)
311
+ 297 train 7.845937 (lr=3.9979e-04) (hash(x)=27269893)
312
+ 298 train 7.612996 (lr=3.9975e-04) (hash(x)=25215077)
313
+ 299 train 7.571717 (lr=3.9971e-04) (hash(x)=24825041)
314
+ 300 val loss 7.6199
315
+ 300 val perplexity 2038.4336
316
+ 300 train 7.659795 (lr=3.9966e-04) (hash(x)=25893804)
317
+ 301 train 7.504346 (lr=3.9961e-04) (hash(x)=22724871)
318
+ 302 train 7.589506 (lr=3.9955e-04) (hash(x)=27317554)
319
+ 303 train 7.480072 (lr=3.9950e-04) (hash(x)=24683717)
320
+ 304 train 7.577101 (lr=3.9944e-04) (hash(x)=21584083)
321
+ 305 train 7.731191 (lr=3.9937e-04) (hash(x)=26190337)
322
+ 306 train 7.637913 (lr=3.9930e-04) (hash(x)=22249019)
323
+ 307 train 7.634756 (lr=3.9923e-04) (hash(x)=23086289)
324
+ 308 train 7.804695 (lr=3.9916e-04) (hash(x)=28047044)
325
+ 309 train 7.576525 (lr=3.9908e-04) (hash(x)=23593875)
326
+ 310 train 7.548244 (lr=3.9900e-04) (hash(x)=26059735)
327
+ 311 train 7.474754 (lr=3.9891e-04) (hash(x)=22414190)
328
+ 312 train 7.665969 (lr=3.9882e-04) (hash(x)=24204591)
329
+ 313 train 7.523152 (lr=3.9873e-04) (hash(x)=26758114)
330
+ 314 train 7.560038 (lr=3.9864e-04) (hash(x)=20084233)
331
+ 315 train 7.491889 (lr=3.9854e-04) (hash(x)=21563184)
332
+ 316 train 7.577713 (lr=3.9843e-04) (hash(x)=23861489)
333
+ 317 train 7.607283 (lr=3.9833e-04) (hash(x)=19218470)
334
+ 318 train 7.617937 (lr=3.9822e-04) (hash(x)=28469562)
335
+ 319 train 7.771876 (lr=3.9811e-04) (hash(x)=25588469)
336
+ 320 train 7.652490 (lr=3.9799e-04) (hash(x)=25236367)
337
+ 321 train 7.676331 (lr=3.9787e-04) (hash(x)=30433767)
338
+ 322 train 7.892921 (lr=3.9775e-04) (hash(x)=26302626)
339
+ 323 train 7.619950 (lr=3.9762e-04) (hash(x)=17851621)
340
+ 324 train 7.833998 (lr=3.9749e-04) (hash(x)=27028191)
341
+ 325 train 7.771285 (lr=3.9736e-04) (hash(x)=25183690)
342
+ 326 train 7.770085 (lr=3.9722e-04) (hash(x)=26182243)
343
+ 327 train 7.874911 (lr=3.9708e-04) (hash(x)=27040642)
344
+ 328 train 7.617208 (lr=3.9694e-04) (hash(x)=24489607)
345
+ 329 train 7.567570 (lr=3.9679e-04) (hash(x)=22183303)
346
+ 330 train 7.702882 (lr=3.9664e-04) (hash(x)=28844646)
347
+ 331 train 8.117968 (lr=3.9648e-04) (hash(x)=27537063)
348
+ 332 train 7.707066 (lr=3.9633e-04) (hash(x)=25446686)
349
+ 333 train 7.563775 (lr=3.9616e-04) (hash(x)=19492832)
350
+ 334 train 7.562403 (lr=3.9600e-04) (hash(x)=25484958)
351
+ 335 train 7.629847 (lr=3.9583e-04) (hash(x)=25951629)
352
+ 336 train 7.703831 (lr=3.9566e-04) (hash(x)=23462798)
353
+ 337 train 7.563368 (lr=3.9549e-04) (hash(x)=22408682)
354
+ 338 train 7.684236 (lr=3.9531e-04) (hash(x)=31317970)
355
+ 339 train 7.521949 (lr=3.9513e-04) (hash(x)=27633457)
356
+ 340 train 7.516596 (lr=3.9494e-04) (hash(x)=23729283)
357
+ 341 train 7.709991 (lr=3.9475e-04) (hash(x)=27635080)
358
+ 342 train 7.622772 (lr=3.9456e-04) (hash(x)=25918780)
359
+ 343 train 7.563438 (lr=3.9437e-04) (hash(x)=22287596)
360
+ 344 train 7.470110 (lr=3.9417e-04) (hash(x)=22694623)
361
+ 345 train 7.906068 (lr=3.9397e-04) (hash(x)=35895440)
362
+ 346 train 7.614980 (lr=3.9376e-04) (hash(x)=24298886)
363
+ 347 train 7.622756 (lr=3.9356e-04) (hash(x)=25888847)
364
+ 348 train 7.727657 (lr=3.9334e-04) (hash(x)=30860646)
365
+ 349 train 7.631621 (lr=3.9313e-04) (hash(x)=26518800)
366
+ 350 val loss 7.5700
367
+ 350 val perplexity 1939.1517
368
+ 350 train 7.555439 (lr=3.9291e-04) (hash(x)=25385523)
369
+ 351 train 7.690179 (lr=3.9269e-04) (hash(x)=37519283)
370
+ 352 train 8.018302 (lr=3.9246e-04) (hash(x)=20564397)
371
+ 353 train 7.657287 (lr=3.9223e-04) (hash(x)=25141945)
372
+ 354 train 7.643577 (lr=3.9200e-04) (hash(x)=22717866)
373
+ 355 train 7.974419 (lr=3.9177e-04) (hash(x)=28527939)
374
+ 356 train 7.744243 (lr=3.9153e-04) (hash(x)=32757059)
375
+ 357 train 7.497519 (lr=3.9129e-04) (hash(x)=21942471)
376
+ 358 train 7.764858 (lr=3.9104e-04) (hash(x)=24247339)
377
+ 359 train 7.534533 (lr=3.9079e-04) (hash(x)=24955630)
378
+ 360 train 7.661250 (lr=3.9054e-04) (hash(x)=21368780)
379
+ 361 train 7.556566 (lr=3.9029e-04) (hash(x)=25541230)
380
+ 362 train 7.453659 (lr=3.9003e-04) (hash(x)=23098806)
381
+ 363 train 7.440163 (lr=3.8977e-04) (hash(x)=23501481)
382
+ 364 train 7.494871 (lr=3.8950e-04) (hash(x)=23802681)
383
+ 365 train 7.505114 (lr=3.8923e-04) (hash(x)=22754988)
384
+ 366 train 7.349278 (lr=3.8896e-04) (hash(x)=22183009)
385
+ 367 train 7.403375 (lr=3.8869e-04) (hash(x)=23475070)
386
+ 368 train 7.576678 (lr=3.8841e-04) (hash(x)=23359061)
387
+ 369 train 7.530566 (lr=3.8813e-04) (hash(x)=23609959)
388
+ 370 train 7.468296 (lr=3.8785e-04) (hash(x)=22408449)
389
+ 371 train 7.399762 (lr=3.8756e-04) (hash(x)=23441847)
390
+ 372 train 7.394564 (lr=3.8727e-04) (hash(x)=23676365)
391
+ 373 train 7.607330 (lr=3.8697e-04) (hash(x)=23096292)
392
+ 374 train 7.599124 (lr=3.8667e-04) (hash(x)=27706294)
393
+ 375 train 7.489771 (lr=3.8637e-04) (hash(x)=26171093)
394
+ 376 train 7.675107 (lr=3.8607e-04) (hash(x)=34946941)
395
+ 377 train 7.465498 (lr=3.8576e-04) (hash(x)=20060838)
396
+ 378 train 7.364142 (lr=3.8545e-04) (hash(x)=21591702)
397
+ 379 train 7.424608 (lr=3.8514e-04) (hash(x)=25451055)
398
+ 380 train 7.091950 (lr=3.8482e-04) (hash(x)=18691455)
399
+ 381 train 7.397336 (lr=3.8450e-04) (hash(x)=24934792)
400
+ 382 train 7.353457 (lr=3.8418e-04) (hash(x)=23864361)
401
+ 383 train 7.402672 (lr=3.8385e-04) (hash(x)=25367597)
402
+ 384 train 7.554334 (lr=3.8352e-04) (hash(x)=27016702)
403
+ 385 train 7.535930 (lr=3.8319e-04) (hash(x)=25022148)
404
+ 386 train 7.542823 (lr=3.8286e-04) (hash(x)=24459012)
405
+ 387 train 7.566740 (lr=3.8252e-04) (hash(x)=25431365)
406
+ 388 train 7.489683 (lr=3.8217e-04) (hash(x)=25958785)
407
+ 389 train 7.687195 (lr=3.8183e-04) (hash(x)=29307788)
408
+ 390 train 7.409885 (lr=3.8148e-04) (hash(x)=24965435)
409
+ 391 train 7.425715 (lr=3.8113e-04) (hash(x)=25651929)
410
+ 392 train 7.451488 (lr=3.8077e-04) (hash(x)=28451031)
411
+ 393 train 7.541465 (lr=3.8042e-04) (hash(x)=26312387)
412
+ 394 train 7.504503 (lr=3.8006e-04) (hash(x)=25326650)
413
+ 395 train 7.430471 (lr=3.7969e-04) (hash(x)=23804913)
414
+ 396 train 7.603853 (lr=3.7933e-04) (hash(x)=25537529)
415
+ 397 train 7.408931 (lr=3.7896e-04) (hash(x)=26115519)
416
+ 398 train 7.432256 (lr=3.7858e-04) (hash(x)=25494804)
417
+ 399 train 7.581734 (lr=3.7821e-04) (hash(x)=23210747)
418
+ 400 val loss 7.5042
419
+ 400 val perplexity 1815.7291
420
+ 400 train 7.558111 (lr=3.7783e-04) (hash(x)=15717069)
421
+ 401 train 7.472945 (lr=3.7744e-04) (hash(x)=18463712)
422
+ 402 train 7.477234 (lr=3.7706e-04) (hash(x)=20705573)
423
+ 403 train 7.464781 (lr=3.7667e-04) (hash(x)=23245699)
424
+ 404 train 7.483142 (lr=3.7628e-04) (hash(x)=23444521)
425
+ 405 train 7.382476 (lr=3.7588e-04) (hash(x)=21770300)
426
+ 406 train 8.122724 (lr=3.7549e-04) (hash(x)=23334569)
427
+ 407 train 7.316421 (lr=3.7509e-04) (hash(x)=25387532)
428
+ 408 train 7.573186 (lr=3.7468e-04) (hash(x)=29466743)
429
+ 409 train 7.606859 (lr=3.7428e-04) (hash(x)=27718703)
430
+ 410 train 7.572973 (lr=3.7387e-04) (hash(x)=17631973)
431
+ 411 train 7.586421 (lr=3.7345e-04) (hash(x)=27661052)
432
+ 412 train 7.351886 (lr=3.7304e-04) (hash(x)=22244096)
433
+ 413 train 7.481600 (lr=3.7262e-04) (hash(x)=26538928)
434
+ 414 train 7.459654 (lr=3.7220e-04) (hash(x)=26256802)
435
+ 415 train 7.556507 (lr=3.7177e-04) (hash(x)=21772654)
436
+ 416 train 7.579297 (lr=3.7135e-04) (hash(x)=25026277)
437
+ 417 train 7.646060 (lr=3.7092e-04) (hash(x)=26263254)
438
+ 418 train 7.549091 (lr=3.7048e-04) (hash(x)=24726314)
439
+ 419 train 7.374911 (lr=3.7005e-04) (hash(x)=23552356)
440
+ 420 train 7.474566 (lr=3.6961e-04) (hash(x)=26699819)
441
+ 421 train 7.910850 (lr=3.6917e-04) (hash(x)=22707141)
442
+ 422 train 7.824143 (lr=3.6872e-04) (hash(x)=25676149)
443
+ 423 train 7.602867 (lr=3.6828e-04) (hash(x)=34234481)
444
+ 424 train 7.458756 (lr=3.6782e-04) (hash(x)=25401624)
445
+ 425 train 7.612796 (lr=3.6737e-04) (hash(x)=35008670)
446
+ 426 train 7.722455 (lr=3.6692e-04) (hash(x)=25682280)
447
+ 427 train 7.553664 (lr=3.6646e-04) (hash(x)=25891543)
448
+ 428 train 7.454477 (lr=3.6599e-04) (hash(x)=25962407)
449
+ 429 train 7.387010 (lr=3.6553e-04) (hash(x)=21971224)
450
+ 430 train 7.535604 (lr=3.6506e-04) (hash(x)=30838142)
451
+ 431 train 7.541897 (lr=3.6459e-04) (hash(x)=27312186)
452
+ 432 train 7.628037 (lr=3.6412e-04) (hash(x)=23657081)
453
+ 433 train 7.547190 (lr=3.6364e-04) (hash(x)=22163749)
454
+ 434 train 7.291088 (lr=3.6316e-04) (hash(x)=24829818)
455
+ 435 train 7.512647 (lr=3.6268e-04) (hash(x)=25942888)
456
+ 436 train 7.490554 (lr=3.6220e-04) (hash(x)=23785671)
457
+ 437 train 7.493221 (lr=3.6171e-04) (hash(x)=25493916)
458
+ 438 train 7.653247 (lr=3.6122e-04) (hash(x)=28618702)
459
+ 439 train 7.733469 (lr=3.6073e-04) (hash(x)=24289951)
460
+ 440 train 7.480082 (lr=3.6023e-04) (hash(x)=26331486)
461
+ 441 train 7.458815 (lr=3.5974e-04) (hash(x)=23648671)
462
+ 442 train 7.417950 (lr=3.5924e-04) (hash(x)=27292797)
463
+ 443 train 7.263269 (lr=3.5873e-04) (hash(x)=24992055)
464
+ 444 train 7.760581 (lr=3.5823e-04) (hash(x)=24398330)
465
+ 445 train 7.458492 (lr=3.5772e-04) (hash(x)=26005549)
466
+ 446 train 7.438668 (lr=3.5721e-04) (hash(x)=30259737)
467
+ 447 train 7.421289 (lr=3.5669e-04) (hash(x)=21046554)
468
+ 448 train 7.404891 (lr=3.5618e-04) (hash(x)=18773877)
469
+ 449 train 7.455591 (lr=3.5566e-04) (hash(x)=26321539)
470
+ 450 val loss 7.4464
471
+ 450 val perplexity 1713.7198
472
+ 450 train 7.419096 (lr=3.5514e-04) (hash(x)=22850804)
473
+ 451 train 7.513982 (lr=3.5461e-04) (hash(x)=27807872)
474
+ 452 train 7.468696 (lr=3.5408e-04) (hash(x)=25545137)
475
+ 453 train 7.451216 (lr=3.5355e-04) (hash(x)=21555364)
476
+ 454 train 7.511461 (lr=3.5302e-04) (hash(x)=27243798)
477
+ 455 train 7.367654 (lr=3.5249e-04) (hash(x)=26325324)
478
+ 456 train 7.356295 (lr=3.5195e-04) (hash(x)=26228728)
479
+ 457 train 7.580873 (lr=3.5141e-04) (hash(x)=24956294)
480
+ 458 train 7.428790 (lr=3.5087e-04) (hash(x)=25592289)
481
+ 459 train 7.415205 (lr=3.5032e-04) (hash(x)=24429389)
482
+ 460 train 7.443428 (lr=3.4977e-04) (hash(x)=28733708)
483
+ 461 train 7.600663 (lr=3.4922e-04) (hash(x)=27929846)
484
+ 462 train 7.451354 (lr=3.4867e-04) (hash(x)=22417476)
485
+ 463 train 7.540589 (lr=3.4812e-04) (hash(x)=25282897)
486
+ 464 train 7.340667 (lr=3.4756e-04) (hash(x)=22278062)
487
+ 465 train 7.338343 (lr=3.4700e-04) (hash(x)=26802508)
488
+ 466 train 7.540537 (lr=3.4644e-04) (hash(x)=22094832)
489
+ 467 train 7.515501 (lr=3.4587e-04) (hash(x)=28390416)
490
+ 468 train 7.712140 (lr=3.4530e-04) (hash(x)=26379331)
491
+ 469 train 7.747327 (lr=3.4473e-04) (hash(x)=26292451)
492
+ 470 train 7.611784 (lr=3.4416e-04) (hash(x)=26301866)
493
+ 471 train 7.572252 (lr=3.4359e-04) (hash(x)=22915344)
494
+ 472 train 7.437881 (lr=3.4301e-04) (hash(x)=25448318)
495
+ 473 train 7.528530 (lr=3.4243e-04) (hash(x)=24744383)
496
+ 474 train 7.363411 (lr=3.4185e-04) (hash(x)=22835201)
497
+ 475 train 7.451021 (lr=3.4127e-04) (hash(x)=26284678)
498
+ 476 train 7.472738 (lr=3.4068e-04) (hash(x)=24745936)
499
+ 477 train 7.307781 (lr=3.4009e-04) (hash(x)=21654269)
500
+ 478 train 7.494930 (lr=3.3950e-04) (hash(x)=32387999)
501
+ 479 train 8.169182 (lr=3.3891e-04) (hash(x)=29466268)
502
+ 480 train 8.116629 (lr=3.3831e-04) (hash(x)=30736727)
503
+ 481 train 8.064593 (lr=3.3771e-04) (hash(x)=32616762)
504
+ 482 train 8.025914 (lr=3.3711e-04) (hash(x)=29358578)
505
+ 483 train 8.085496 (lr=3.3651e-04) (hash(x)=32201138)
506
+ 484 train 8.087114 (lr=3.3590e-04) (hash(x)=30041621)
507
+ 485 train 8.053341 (lr=3.3530e-04) (hash(x)=33683468)
508
+ 486 train 7.830552 (lr=3.3469e-04) (hash(x)=26837755)
509
+ 487 train 7.695760 (lr=3.3408e-04) (hash(x)=27498125)
510
+ 488 train 7.839445 (lr=3.3346e-04) (hash(x)=22372507)
511
+ 489 train 7.572035 (lr=3.3285e-04) (hash(x)=25164318)
512
+ 490 train 7.580844 (lr=3.3223e-04) (hash(x)=26562947)
513
+ 491 train 7.516102 (lr=3.3161e-04) (hash(x)=25690221)
514
+ 492 train 7.547541 (lr=3.3099e-04) (hash(x)=21947305)
515
+ 493 train 7.474365 (lr=3.3036e-04) (hash(x)=29031550)
516
+ 494 train 7.611632 (lr=3.2973e-04) (hash(x)=26551589)
517
+ 495 train 7.563972 (lr=3.2911e-04) (hash(x)=24094093)
518
+ 496 train 7.359215 (lr=3.2847e-04) (hash(x)=23639076)
519
+ 497 train 9.138015 (lr=3.2784e-04) (hash(x)=61455960)
520
+ 498 train 11.740412 (lr=3.2721e-04) (hash(x)=65636572)
521
+ 499 train 11.538445 (lr=3.2657e-04) (hash(x)=61104710)
522
+ 500 val loss 7.5033
523
+ 500 val perplexity 1814.0122
524
+ 500 train 8.112293 (lr=3.2593e-04) (hash(x)=24123821)
525
+ 501 train 7.179349 (lr=3.2529e-04) (hash(x)=20881126)
526
+ 502 train 7.577507 (lr=3.2464e-04) (hash(x)=23209969)
527
+ 503 train 7.663974 (lr=3.2400e-04) (hash(x)=26920243)
528
+ 504 train 7.622949 (lr=3.2335e-04) (hash(x)=26127756)
529
+ 505 train 7.546395 (lr=3.2270e-04) (hash(x)=21970343)
530
+ 506 train 7.562242 (lr=3.2205e-04) (hash(x)=24097897)
531
+ 507 train 7.609089 (lr=3.2140e-04) (hash(x)=31676352)
532
+ 508 train 7.499722 (lr=3.2074e-04) (hash(x)=23908053)
533
+ 509 train 7.416518 (lr=3.2008e-04) (hash(x)=25600562)
534
+ 510 train 7.472045 (lr=3.1943e-04) (hash(x)=23549168)
535
+ 511 train 7.381939 (lr=3.1876e-04) (hash(x)=22022924)
536
+ 512 train 7.440255 (lr=3.1810e-04) (hash(x)=23294980)
537
+ 513 train 7.340988 (lr=3.1744e-04) (hash(x)=22777085)
538
+ 514 train 7.687416 (lr=3.1677e-04) (hash(x)=26271115)
539
+ 515 train 7.426669 (lr=3.1610e-04) (hash(x)=21727744)
540
+ 516 train 7.489923 (lr=3.1543e-04) (hash(x)=22729318)
541
+ 517 train 7.688191 (lr=3.1476e-04) (hash(x)=20701998)
542
+ 518 train 7.606686 (lr=3.1408e-04) (hash(x)=27928398)
543
+ 519 train 7.666313 (lr=3.1341e-04) (hash(x)=20863982)
544
+ 520 train 7.544641 (lr=3.1273e-04) (hash(x)=22827054)
545
+ 521 train 7.541405 (lr=3.1205e-04) (hash(x)=25084926)
546
+ 522 train 7.669883 (lr=3.1137e-04) (hash(x)=25540961)
547
+ 523 train 7.401334 (lr=3.1069e-04) (hash(x)=24412897)
548
+ 524 train 7.612433 (lr=3.1000e-04) (hash(x)=26269869)
549
+ 525 train 7.359899 (lr=3.0931e-04) (hash(x)=25434913)
550
+ 526 train 7.499692 (lr=3.0862e-04) (hash(x)=25157653)
551
+ 527 train 7.486452 (lr=3.0793e-04) (hash(x)=26302544)
552
+ 528 train 7.421946 (lr=3.0724e-04) (hash(x)=22010703)
553
+ 529 train 7.716221 (lr=3.0655e-04) (hash(x)=34976197)
554
+ 530 train 7.419102 (lr=3.0585e-04) (hash(x)=28248301)
555
+ 531 train 7.447789 (lr=3.0516e-04) (hash(x)=23870805)
556
+ 532 train 7.402340 (lr=3.0446e-04) (hash(x)=24457252)
557
+ 533 train 7.488316 (lr=3.0376e-04) (hash(x)=25041103)
558
+ 534 train 7.403051 (lr=3.0306e-04) (hash(x)=25998487)
559
+ 535 train 7.545467 (lr=3.0235e-04) (hash(x)=26041745)
560
+ 536 train 7.419512 (lr=3.0165e-04) (hash(x)=27249810)
561
+ 537 train 7.463527 (lr=3.0094e-04) (hash(x)=24068619)
562
+ 538 train 7.515015 (lr=3.0023e-04) (hash(x)=26359528)
563
+ 539 train 7.468028 (lr=2.9952e-04) (hash(x)=25441262)
564
+ 540 train 7.352800 (lr=2.9881e-04) (hash(x)=24413659)
565
+ 541 train 7.754736 (lr=2.9810e-04) (hash(x)=29181867)
566
+ 542 train 7.527767 (lr=2.9738e-04) (hash(x)=31551575)
567
+ 543 train 7.534967 (lr=2.9667e-04) (hash(x)=27093757)
568
+ 544 train 7.506951 (lr=2.9595e-04) (hash(x)=23624298)
569
+ 545 train 7.352910 (lr=2.9523e-04) (hash(x)=25150275)
570
+ 546 train 7.365006 (lr=2.9451e-04) (hash(x)=26239886)
571
+ 547 train 7.721173 (lr=2.9379e-04) (hash(x)=33457102)
572
+ 548 train 7.349096 (lr=2.9307e-04) (hash(x)=17499738)
573
+ 549 train 7.389928 (lr=2.9234e-04) (hash(x)=25755322)
574
+ 550 val loss 7.4194
575
+ 550 val perplexity 1667.9501
576
+ 550 train 7.495744 (lr=2.9162e-04) (hash(x)=23722795)
577
+ 551 train 7.370512 (lr=2.9089e-04) (hash(x)=23630794)
578
+ 552 train 7.332369 (lr=2.9016e-04) (hash(x)=23374168)
579
+ 553 train 7.337629 (lr=2.8943e-04) (hash(x)=25167987)
580
+ 554 train 7.597768 (lr=2.8870e-04) (hash(x)=30941010)
581
+ 555 train 7.359838 (lr=2.8797e-04) (hash(x)=24043224)
582
+ 556 train 7.492447 (lr=2.8723e-04) (hash(x)=29736627)
583
+ 557 train 7.438973 (lr=2.8650e-04) (hash(x)=24665411)
584
+ 558 train 7.482018 (lr=2.8576e-04) (hash(x)=27916043)
585
+ 559 train 7.406168 (lr=2.8502e-04) (hash(x)=17321657)
586
+ 560 train 7.233833 (lr=2.8428e-04) (hash(x)=23829833)
587
+ 561 train 7.648836 (lr=2.8354e-04) (hash(x)=27932591)
588
+ 562 train 7.705634 (lr=2.8280e-04) (hash(x)=28206079)
589
+ 563 train 7.790341 (lr=2.8206e-04) (hash(x)=29447954)
590
+ 564 train 7.437534 (lr=2.8132e-04) (hash(x)=25935364)
591
+ 565 train 7.495313 (lr=2.8057e-04) (hash(x)=24621816)
592
+ 566 train 7.381628 (lr=2.7982e-04) (hash(x)=24070405)
593
+ 567 train 7.475368 (lr=2.7908e-04) (hash(x)=24682294)
594
+ 568 train 7.477361 (lr=2.7833e-04) (hash(x)=24050209)
595
+ 569 train 7.433279 (lr=2.7758e-04) (hash(x)=28057095)
596
+ 570 train 7.545300 (lr=2.7683e-04) (hash(x)=22523232)
597
+ 571 train 7.441713 (lr=2.7607e-04) (hash(x)=24547536)
598
+ 572 train 7.338270 (lr=2.7532e-04) (hash(x)=24978712)
599
+ 573 train 7.304700 (lr=2.7457e-04) (hash(x)=18987545)
600
+ 574 train 7.409098 (lr=2.7381e-04) (hash(x)=24092784)
601
+ 575 train 7.365407 (lr=2.7306e-04) (hash(x)=26745063)
602
+ 576 train 7.381090 (lr=2.7230e-04) (hash(x)=22539112)
603
+ 577 train 7.151959 (lr=2.7154e-04) (hash(x)=21447406)
604
+ 578 train 7.151064 (lr=2.7078e-04) (hash(x)=21082139)
605
+ 579 train 7.204741 (lr=2.7002e-04) (hash(x)=24650655)
606
+ 580 train 7.409270 (lr=2.6926e-04) (hash(x)=22130721)
607
+ 581 train 7.728584 (lr=2.6850e-04) (hash(x)=26519976)
608
+ 582 train 8.259889 (lr=2.6773e-04) (hash(x)=25582440)
609
+ 583 train 7.532710 (lr=2.6697e-04) (hash(x)=20106514)
610
+ 584 train 7.337157 (lr=2.6620e-04) (hash(x)=25695510)
611
+ 585 train 7.419406 (lr=2.6544e-04) (hash(x)=20117808)
612
+ 586 train 7.562501 (lr=2.6467e-04) (hash(x)=24501900)
613
+ 587 train 7.501615 (lr=2.6390e-04) (hash(x)=25510334)
614
+ 588 train 7.330684 (lr=2.6314e-04) (hash(x)=22099158)
615
+ 589 train 7.440312 (lr=2.6237e-04) (hash(x)=25749120)
616
+ 590 train 7.712525 (lr=2.6160e-04) (hash(x)=21720691)
617
+ 591 train 7.254004 (lr=2.6083e-04) (hash(x)=24357241)
618
+ 592 train 7.529492 (lr=2.6005e-04) (hash(x)=28911272)
619
+ 593 train 7.378532 (lr=2.5928e-04) (hash(x)=25584945)
620
+ 594 train 7.264764 (lr=2.5851e-04) (hash(x)=21847287)
621
+ 595 train 7.356828 (lr=2.5773e-04) (hash(x)=25198897)
622
+ 596 train 7.262772 (lr=2.5696e-04) (hash(x)=22436833)
623
+ 597 train 7.600657 (lr=2.5618e-04) (hash(x)=23475891)
624
+ 598 train 7.265197 (lr=2.5541e-04) (hash(x)=25083992)
625
+ 599 train 7.517988 (lr=2.5463e-04) (hash(x)=23644804)
626
+ 600 val loss 7.4405
627
+ 600 val perplexity 1703.5885
628
+ 600 train 7.394555 (lr=2.5385e-04) (hash(x)=31377168)
629
+ 601 train 7.775324 (lr=2.5307e-04) (hash(x)=25629584)
630
+ 602 train 7.579798 (lr=2.5230e-04) (hash(x)=29285135)
631
+ 603 train 7.792214 (lr=2.5152e-04) (hash(x)=29797280)
632
+ 604 train 7.599888 (lr=2.5074e-04) (hash(x)=33122326)
633
+ 605 train 7.885257 (lr=2.4996e-04) (hash(x)=30026394)
634
+ 606 train 8.123857 (lr=2.4917e-04) (hash(x)=34353164)
635
+ 607 train 7.357494 (lr=2.4839e-04) (hash(x)=26594177)
636
+ 608 train 7.461129 (lr=2.4761e-04) (hash(x)=25554738)
637
+ 609 train 7.545344 (lr=2.4683e-04) (hash(x)=23766256)
638
+ 610 train 7.414563 (lr=2.4604e-04) (hash(x)=26853444)
639
+ 611 train 7.422283 (lr=2.4526e-04) (hash(x)=24250150)
640
+ 612 train 7.331684 (lr=2.4448e-04) (hash(x)=26289588)
641
+ 613 train 7.393815 (lr=2.4369e-04) (hash(x)=26108817)
642
+ 614 train 7.318213 (lr=2.4291e-04) (hash(x)=23814987)
643
+ 615 train 7.303033 (lr=2.4212e-04) (hash(x)=18582688)
644
+ 616 train 7.352200 (lr=2.4133e-04) (hash(x)=25558792)
645
+ 617 train 7.295042 (lr=2.4055e-04) (hash(x)=23272705)
646
+ 618 train 7.320442 (lr=2.3976e-04) (hash(x)=24709241)
647
+ 619 train 7.328864 (lr=2.3897e-04) (hash(x)=24793412)
648
+ 620 train 7.346180 (lr=2.3818e-04) (hash(x)=27256707)
649
+ 621 train 7.321583 (lr=2.3740e-04) (hash(x)=19280989)
650
+ 622 train 7.367604 (lr=2.3661e-04) (hash(x)=26508642)
651
+ 623 train 7.293493 (lr=2.3582e-04) (hash(x)=22751150)
652
+ 624 train 7.593561 (lr=2.3503e-04) (hash(x)=23880160)
653
+ 625 train 7.296295 (lr=2.3424e-04) (hash(x)=23537942)
654
+ 626 train 7.304734 (lr=2.3345e-04) (hash(x)=26975313)
655
+ 627 train 7.226646 (lr=2.3266e-04) (hash(x)=23748375)
656
+ 628 train 7.247867 (lr=2.3187e-04) (hash(x)=25541754)
657
+ 629 train 7.664094 (lr=2.3108e-04) (hash(x)=27413825)
658
+ 630 train 7.624827 (lr=2.3029e-04) (hash(x)=27075951)
659
+ 631 train 7.435192 (lr=2.2950e-04) (hash(x)=21682445)
660
+ 632 train 7.410468 (lr=2.2871e-04) (hash(x)=23361504)
661
+ 633 train 7.307510 (lr=2.2792e-04) (hash(x)=22040184)
662
+ 634 train 7.330922 (lr=2.2713e-04) (hash(x)=23727731)
663
+ 635 train 7.324286 (lr=2.2633e-04) (hash(x)=24463229)
664
+ 636 train 7.544067 (lr=2.2554e-04) (hash(x)=32663792)
665
+ 637 train 7.360484 (lr=2.2475e-04) (hash(x)=24091954)
666
+ 638 train 7.308915 (lr=2.2396e-04) (hash(x)=23599013)
667
+ 639 train 7.312502 (lr=2.2317e-04) (hash(x)=26670219)
668
+ 640 train 7.295678 (lr=2.2238e-04) (hash(x)=24380893)
669
+ 641 train 7.382475 (lr=2.2158e-04) (hash(x)=25558986)
670
+ 642 train 7.438945 (lr=2.2079e-04) (hash(x)=25729492)
671
+ 643 train 7.203730 (lr=2.2000e-04) (hash(x)=19218943)
672
+ 644 train 7.428310 (lr=2.1921e-04) (hash(x)=23952206)
673
+ 645 train 7.384051 (lr=2.1842e-04) (hash(x)=24551402)
674
+ 646 train 7.266860 (lr=2.1762e-04) (hash(x)=24007642)
675
+ 647 train 7.320735 (lr=2.1683e-04) (hash(x)=25229281)
676
+ 648 train 7.105819 (lr=2.1604e-04) (hash(x)=18214283)
677
+ 649 train 7.020990 (lr=2.1525e-04) (hash(x)=21251127)
678
+ 650 val loss 7.3726
679
+ 650 val perplexity 1591.7930
680
+ 650 train 7.541228 (lr=2.1446e-04) (hash(x)=25097619)
681
+ 651 train 7.220702 (lr=2.1367e-04) (hash(x)=25800857)
682
+ 652 train 7.223462 (lr=2.1287e-04) (hash(x)=22918428)
683
+ 653 train 7.296169 (lr=2.1208e-04) (hash(x)=25836189)
684
+ 654 train 7.258660 (lr=2.1129e-04) (hash(x)=24504567)
685
+ 655 train 7.296273 (lr=2.1050e-04) (hash(x)=25999255)
686
+ 656 train 7.306149 (lr=2.0971e-04) (hash(x)=27308968)
687
+ 657 train 7.398460 (lr=2.0892e-04) (hash(x)=20149394)
688
+ 658 train 7.358266 (lr=2.0813e-04) (hash(x)=24498556)
689
+ 659 train 7.203475 (lr=2.0734e-04) (hash(x)=21999338)
690
+ 660 train 7.237967 (lr=2.0655e-04) (hash(x)=23280568)
691
+ 661 train 7.552985 (lr=2.0576e-04) (hash(x)=27611302)
692
+ 662 train 7.385753 (lr=2.0497e-04) (hash(x)=23982308)
693
+ 663 train 7.315900 (lr=2.0418e-04) (hash(x)=22778356)
694
+ 664 train 7.209702 (lr=2.0339e-04) (hash(x)=23467046)
695
+ 665 train 7.401742 (lr=2.0260e-04) (hash(x)=24195959)
696
+ 666 train 7.746605 (lr=2.0182e-04) (hash(x)=32710438)
697
+ 667 train 7.445144 (lr=2.0103e-04) (hash(x)=19127465)
698
+ 668 train 7.570690 (lr=2.0024e-04) (hash(x)=28962580)
699
+ 669 train 7.299295 (lr=1.9945e-04) (hash(x)=22426274)
700
+ 670 train 7.328897 (lr=1.9867e-04) (hash(x)=25606512)
701
+ 671 train 7.206809 (lr=1.9788e-04) (hash(x)=23785205)
702
+ 672 train 7.343933 (lr=1.9709e-04) (hash(x)=24779414)
703
+ 673 train 7.462252 (lr=1.9631e-04) (hash(x)=28505676)
704
+ 674 train 7.529500 (lr=1.9552e-04) (hash(x)=29156288)
705
+ 675 train 7.234024 (lr=1.9474e-04) (hash(x)=27706477)
706
+ 676 train 7.357965 (lr=1.9396e-04) (hash(x)=23711800)
707
+ 677 train 7.471089 (lr=1.9317e-04) (hash(x)=28684175)
708
+ 678 train 7.313143 (lr=1.9239e-04) (hash(x)=25225466)
709
+ 679 train 7.454827 (lr=1.9161e-04) (hash(x)=28003600)
710
+ 680 train 7.327992 (lr=1.9083e-04) (hash(x)=27989890)
711
+ 681 train 7.414724 (lr=1.9004e-04) (hash(x)=27586845)
712
+ 682 train 7.372186 (lr=1.8926e-04) (hash(x)=22373927)
713
+ 683 train 7.236530 (lr=1.8848e-04) (hash(x)=21176405)
714
+ 684 train 7.217636 (lr=1.8770e-04) (hash(x)=23806052)
715
+ 685 train 7.302362 (lr=1.8693e-04) (hash(x)=25888433)
716
+ 686 train 7.500119 (lr=1.8615e-04) (hash(x)=27797727)
717
+ 687 train 7.331993 (lr=1.8537e-04) (hash(x)=27916982)
718
+ 688 train 7.280727 (lr=1.8459e-04) (hash(x)=24621793)
719
+ 689 train 7.194295 (lr=1.8382e-04) (hash(x)=24334708)
720
+ 690 train 7.353946 (lr=1.8304e-04) (hash(x)=24419143)
721
+ 691 train 7.244755 (lr=1.8227e-04) (hash(x)=25617709)
722
+ 692 train 7.254500 (lr=1.8149e-04) (hash(x)=24556034)
723
+ 693 train 7.598697 (lr=1.8072e-04) (hash(x)=25152362)
724
+ 694 train 7.369184 (lr=1.7995e-04) (hash(x)=24326376)
725
+ 695 train 7.230223 (lr=1.7917e-04) (hash(x)=24276512)
726
+ 696 train 7.125437 (lr=1.7840e-04) (hash(x)=24344695)
727
+ 697 train 7.373136 (lr=1.7763e-04) (hash(x)=26776133)
728
+ 698 train 7.217302 (lr=1.7686e-04) (hash(x)=21847282)
729
+ 699 train 7.100968 (lr=1.7610e-04) (hash(x)=20254159)
730
+ 700 val loss 7.3721
731
+ 700 val perplexity 1591.0151
732
+ 700 train 7.227040 (lr=1.7533e-04) (hash(x)=25044762)
733
+ 701 train 7.294444 (lr=1.7456e-04) (hash(x)=27873855)
734
+ 702 train 7.359463 (lr=1.7380e-04) (hash(x)=22575615)
735
+ 703 train 7.308659 (lr=1.7303e-04) (hash(x)=25113614)
736
+ 704 train 7.427131 (lr=1.7227e-04) (hash(x)=28062905)
737
+ 705 train 7.341754 (lr=1.7150e-04) (hash(x)=26192886)
738
+ 706 train 7.313987 (lr=1.7074e-04) (hash(x)=25770338)
739
+ 707 train 7.205117 (lr=1.6998e-04) (hash(x)=24614912)
740
+ 708 train 7.282972 (lr=1.6922e-04) (hash(x)=26934071)
741
+ 709 train 7.227664 (lr=1.6846e-04) (hash(x)=25397093)
742
+ 710 train 7.376760 (lr=1.6770e-04) (hash(x)=25925963)
743
+ 711 train 7.184783 (lr=1.6694e-04) (hash(x)=19811802)
744
+ 712 train 7.305120 (lr=1.6619e-04) (hash(x)=23248423)
745
+ 713 train 7.290839 (lr=1.6543e-04) (hash(x)=25579655)
746
+ 714 train 7.457393 (lr=1.6468e-04) (hash(x)=25537603)
747
+ 715 train 7.360604 (lr=1.6393e-04) (hash(x)=26096514)
748
+ 716 train 7.401059 (lr=1.6317e-04) (hash(x)=24132888)
749
+ 717 train 7.225995 (lr=1.6242e-04) (hash(x)=22589633)
750
+ 718 train 7.198757 (lr=1.6167e-04) (hash(x)=27781566)
751
+ 719 train 7.454805 (lr=1.6092e-04) (hash(x)=26757147)
752
+ 720 train 7.474433 (lr=1.6018e-04) (hash(x)=26649864)
753
+ 721 train 7.336629 (lr=1.5943e-04) (hash(x)=24963730)
754
+ 722 train 7.340871 (lr=1.5868e-04) (hash(x)=25249959)
755
+ 723 train 7.212744 (lr=1.5794e-04) (hash(x)=27703124)
756
+ 724 train 7.209770 (lr=1.5720e-04) (hash(x)=26459082)
757
+ 725 train 7.149263 (lr=1.5646e-04) (hash(x)=24781713)
758
+ 726 train 7.484432 (lr=1.5572e-04) (hash(x)=26597693)
759
+ 727 train 7.273416 (lr=1.5498e-04) (hash(x)=24720476)
760
+ 728 train 7.282552 (lr=1.5424e-04) (hash(x)=23691295)
761
+ 729 train 7.395479 (lr=1.5350e-04) (hash(x)=23481301)
762
+ 730 train 7.292824 (lr=1.5277e-04) (hash(x)=23295676)
763
+ 731 train 7.244826 (lr=1.5203e-04) (hash(x)=25641033)
764
+ 732 train 7.342420 (lr=1.5130e-04) (hash(x)=22441379)
765
+ 733 train 7.270470 (lr=1.5057e-04) (hash(x)=27369437)
766
+ 734 train 7.282380 (lr=1.4984e-04) (hash(x)=24482272)
767
+ 735 train 7.425600 (lr=1.4911e-04) (hash(x)=24760900)
768
+ 736 train 7.299913 (lr=1.4838e-04) (hash(x)=23919156)
769
+ 737 train 7.223637 (lr=1.4766e-04) (hash(x)=21692546)
770
+ 738 train 7.243975 (lr=1.4693e-04) (hash(x)=27692087)
771
+ 739 train 7.372296 (lr=1.4621e-04) (hash(x)=22720391)
772
+ 740 train 7.160585 (lr=1.4549e-04) (hash(x)=26146560)
773
+ 741 train 7.315294 (lr=1.4477e-04) (hash(x)=24376979)
774
+ 742 train 7.328500 (lr=1.4405e-04) (hash(x)=22917712)
775
+ 743 train 7.206594 (lr=1.4333e-04) (hash(x)=22244509)
776
+ 744 train 7.208919 (lr=1.4262e-04) (hash(x)=25243385)
777
+ 745 train 7.243530 (lr=1.4190e-04) (hash(x)=21831960)
778
+ 746 train 7.247589 (lr=1.4119e-04) (hash(x)=24842373)
779
+ 747 train 7.177629 (lr=1.4048e-04) (hash(x)=24921535)
780
+ 748 train 7.245176 (lr=1.3977e-04) (hash(x)=24344615)
781
+ 749 train 7.245094 (lr=1.3906e-04) (hash(x)=25381251)
782
+ 750 val loss 7.3543
783
+ 750 val perplexity 1562.8788
784
+ 750 train 7.155465 (lr=1.3835e-04) (hash(x)=22887303)
785
+ 751 train 7.339561 (lr=1.3765e-04) (hash(x)=25687352)
786
+ 752 train 7.299342 (lr=1.3694e-04) (hash(x)=27791412)
787
+ 753 train 7.216559 (lr=1.3624e-04) (hash(x)=25221654)
788
+ 754 train 7.309443 (lr=1.3554e-04) (hash(x)=23243731)
789
+ 755 train 7.288528 (lr=1.3484e-04) (hash(x)=24469863)
790
+ 756 train 7.429299 (lr=1.3415e-04) (hash(x)=22484936)
791
+ 757 train 7.288990 (lr=1.3345e-04) (hash(x)=25291413)
792
+ 758 train 7.205394 (lr=1.3276e-04) (hash(x)=23952225)
793
+ 759 train 7.288749 (lr=1.3207e-04) (hash(x)=25744274)
794
+ 760 train 7.211941 (lr=1.3138e-04) (hash(x)=22616739)
795
+ 761 train 7.197942 (lr=1.3069e-04) (hash(x)=24246963)
796
+ 762 train 7.166834 (lr=1.3000e-04) (hash(x)=19407094)
797
+ 763 train 7.250676 (lr=1.2931e-04) (hash(x)=19724058)
798
+ 764 train 7.086380 (lr=1.2863e-04) (hash(x)=18836491)
799
+ 765 train 7.143649 (lr=1.2795e-04) (hash(x)=20705649)
800
+ 766 train 7.386777 (lr=1.2727e-04) (hash(x)=21325875)
801
+ 767 train 7.426955 (lr=1.2659e-04) (hash(x)=24938685)
802
+ 768 train 7.367415 (lr=1.2592e-04) (hash(x)=25152403)
803
+ 769 train 7.504940 (lr=1.2524e-04) (hash(x)=21707000)
804
+ 770 train 7.191629 (lr=1.2457e-04) (hash(x)=17952018)
805
+ 771 train 7.235609 (lr=1.2390e-04) (hash(x)=24948980)
806
+ 772 train 7.685849 (lr=1.2323e-04) (hash(x)=28012110)
807
+ 773 train 7.330867 (lr=1.2256e-04) (hash(x)=30773958)
808
+ 774 train 7.894384 (lr=1.2190e-04) (hash(x)=29481068)
809
+ 775 train 7.325729 (lr=1.2124e-04) (hash(x)=26769046)
810
+ 776 train 7.251319 (lr=1.2057e-04) (hash(x)=26755048)
811
+ 777 train 7.292826 (lr=1.1992e-04) (hash(x)=18849656)
812
+ 778 train 7.571896 (lr=1.1926e-04) (hash(x)=25449800)
813
+ 779 train 7.969449 (lr=1.1860e-04) (hash(x)=24624950)
814
+ 780 train 7.284887 (lr=1.1795e-04) (hash(x)=26174069)
815
+ 781 train 7.474746 (lr=1.1730e-04) (hash(x)=27299605)
816
+ 782 train 7.377332 (lr=1.1665e-04) (hash(x)=24777273)
817
+ 783 train 7.371468 (lr=1.1600e-04) (hash(x)=24113445)
818
+ 784 train 7.241295 (lr=1.1536e-04) (hash(x)=23970384)
819
+ 785 train 7.505572 (lr=1.1471e-04) (hash(x)=32644465)
820
+ 786 train 7.607574 (lr=1.1407e-04) (hash(x)=29951548)
821
+ 787 train 7.340805 (lr=1.1343e-04) (hash(x)=25829219)
822
+ 788 train 7.379669 (lr=1.1279e-04) (hash(x)=20556094)
823
+ 789 train 7.208964 (lr=1.1216e-04) (hash(x)=25194550)
824
+ 790 train 7.340843 (lr=1.1153e-04) (hash(x)=28482186)
825
+ 791 train 7.323513 (lr=1.1089e-04) (hash(x)=25941804)
826
+ 792 train 7.338962 (lr=1.1027e-04) (hash(x)=27049208)
827
+ 793 train 7.328945 (lr=1.0964e-04) (hash(x)=24075727)
828
+ 794 train 7.395763 (lr=1.0901e-04) (hash(x)=25082806)
829
+ 795 train 7.366691 (lr=1.0839e-04) (hash(x)=23104539)
830
+ 796 train 7.251210 (lr=1.0777e-04) (hash(x)=25351113)
831
+ 797 train 7.378623 (lr=1.0715e-04) (hash(x)=23018983)
832
+ 798 train 7.305014 (lr=1.0654e-04) (hash(x)=25302106)
833
+ 799 train 7.235627 (lr=1.0592e-04) (hash(x)=24422929)
834
+ 800 val loss 7.3372
835
+ 800 val perplexity 1536.3641
836
+ 800 train 7.303794 (lr=1.0531e-04) (hash(x)=29052117)
837
+ 801 train 7.307367 (lr=1.0470e-04) (hash(x)=23389696)
838
+ 802 train 7.247597 (lr=1.0410e-04) (hash(x)=23962503)
839
+ 803 train 7.320859 (lr=1.0349e-04) (hash(x)=29102969)
840
+ 804 train 7.459579 (lr=1.0289e-04) (hash(x)=24948166)
841
+ 805 train 7.406887 (lr=1.0229e-04) (hash(x)=28359859)
842
+ 806 train 7.394737 (lr=1.0169e-04) (hash(x)=22360298)
843
+ 807 train 7.504811 (lr=1.0109e-04) (hash(x)=24630955)
844
+ 808 train 7.500166 (lr=1.0050e-04) (hash(x)=26897402)
845
+ 809 train 7.437614 (lr=9.9910e-05) (hash(x)=22969113)
846
+ 810 train 7.272452 (lr=9.9321e-05) (hash(x)=25205781)
847
+ 811 train 7.284291 (lr=9.8735e-05) (hash(x)=27004415)
848
+ 812 train 7.151403 (lr=9.8151e-05) (hash(x)=22974875)
849
+ 813 train 7.262775 (lr=9.7569e-05) (hash(x)=22265063)
850
+ 814 train 7.167220 (lr=9.6990e-05) (hash(x)=23312114)
851
+ 815 train 7.301248 (lr=9.6413e-05) (hash(x)=26905990)
852
+ 816 train 7.306141 (lr=9.5838e-05) (hash(x)=27089280)
853
+ 817 train 7.282311 (lr=9.5266e-05) (hash(x)=24047679)
854
+ 818 train 7.332480 (lr=9.4696e-05) (hash(x)=25034684)
855
+ 819 train 7.546966 (lr=9.4129e-05) (hash(x)=26905582)
856
+ 820 train 7.185664 (lr=9.3564e-05) (hash(x)=21406950)
857
+ 821 train 7.180654 (lr=9.3001e-05) (hash(x)=29018970)
858
+ 822 train 7.153368 (lr=9.2441e-05) (hash(x)=20853530)
859
+ 823 train 7.251377 (lr=9.1884e-05) (hash(x)=24503315)
860
+ 824 train 7.266328 (lr=9.1328e-05) (hash(x)=23080074)
861
+ 825 train 7.116228 (lr=9.0776e-05) (hash(x)=22607537)
862
+ 826 train 7.530142 (lr=9.0226e-05) (hash(x)=32179773)
863
+ 827 train 8.526848 (lr=8.9678e-05) (hash(x)=26690225)
864
+ 828 train 7.304084 (lr=8.9133e-05) (hash(x)=22109609)
865
+ 829 train 7.273099 (lr=8.8591e-05) (hash(x)=24226255)
866
+ 830 train 7.332911 (lr=8.8051e-05) (hash(x)=23013573)
867
+ 831 train 7.131171 (lr=8.7513e-05) (hash(x)=23622702)
868
+ 832 train 7.312505 (lr=8.6978e-05) (hash(x)=21517429)
869
+ 833 train 7.245609 (lr=8.6446e-05) (hash(x)=23910425)
870
+ 834 train 7.373300 (lr=8.5916e-05) (hash(x)=24407484)
871
+ 835 train 7.680523 (lr=8.5389e-05) (hash(x)=27213318)
872
+ 836 train 7.298686 (lr=8.4865e-05) (hash(x)=25524191)
873
+ 837 train 7.327335 (lr=8.4343e-05) (hash(x)=26701577)
874
+ 838 train 7.154793 (lr=8.3824e-05) (hash(x)=22091266)
875
+ 839 train 7.302507 (lr=8.3307e-05) (hash(x)=26216794)
876
+ 840 train 7.213208 (lr=8.2793e-05) (hash(x)=20817340)
877
+ 841 train 7.665943 (lr=8.2282e-05) (hash(x)=27579623)
878
+ 842 train 7.160152 (lr=8.1773e-05) (hash(x)=21532187)
879
+ 843 train 7.475921 (lr=8.1267e-05) (hash(x)=26086198)
880
+ 844 train 7.262809 (lr=8.0764e-05) (hash(x)=23911729)
881
+ 845 train 7.274337 (lr=8.0263e-05) (hash(x)=24863441)
882
+ 846 train 7.732491 (lr=7.9765e-05) (hash(x)=28856978)
883
+ 847 train 7.083306 (lr=7.9270e-05) (hash(x)=25236385)
884
+ 848 train 7.459886 (lr=7.8778e-05) (hash(x)=24278687)
885
+ 849 train 7.535857 (lr=7.8288e-05) (hash(x)=25330803)
886
+ 850 val loss 7.3276
887
+ 850 val perplexity 1521.7587
888
+ 850 train 7.386761 (lr=7.7801e-05) (hash(x)=21568545)
889
+ 851 train 7.264637 (lr=7.7317e-05) (hash(x)=22508532)
890
+ 852 train 7.291889 (lr=7.6836e-05) (hash(x)=24850662)
891
+ 853 train 7.304751 (lr=7.6357e-05) (hash(x)=21799261)
892
+ 854 train 7.498398 (lr=7.5881e-05) (hash(x)=24884891)
893
+ 855 train 7.290056 (lr=7.5408e-05) (hash(x)=26311168)
894
+ 856 train 7.342678 (lr=7.4938e-05) (hash(x)=27549895)
895
+ 857 train 7.011919 (lr=7.4470e-05) (hash(x)=20977430)
896
+ 858 train 7.166585 (lr=7.4005e-05) (hash(x)=23448855)
897
+ 859 train 7.363522 (lr=7.3544e-05) (hash(x)=26941617)
898
+ 860 train 7.564673 (lr=7.3085e-05) (hash(x)=26953192)
899
+ 861 train 7.532113 (lr=7.2628e-05) (hash(x)=30008957)
900
+ 862 train 7.428288 (lr=7.2175e-05) (hash(x)=27069893)
901
+ 863 train 7.320478 (lr=7.1725e-05) (hash(x)=27125962)
902
+ 864 train 7.144945 (lr=7.1277e-05) (hash(x)=23341415)
903
+ 865 train 7.272380 (lr=7.0832e-05) (hash(x)=25888358)
904
+ 866 train 7.424111 (lr=7.0390e-05) (hash(x)=26761645)
905
+ 867 train 7.311540 (lr=6.9952e-05) (hash(x)=21725719)
906
+ 868 train 7.222312 (lr=6.9516e-05) (hash(x)=25746493)
907
+ 869 train 7.240128 (lr=6.9082e-05) (hash(x)=21669704)
908
+ 870 train 7.155473 (lr=6.8652e-05) (hash(x)=21167493)
909
+ 871 train 7.238923 (lr=6.8225e-05) (hash(x)=23166993)
910
+ 872 train 7.214608 (lr=6.7801e-05) (hash(x)=23947208)
911
+ 873 train 7.329420 (lr=6.7379e-05) (hash(x)=26235974)
912
+ 874 train 7.702572 (lr=6.6961e-05) (hash(x)=29822604)
913
+ 875 train 7.210525 (lr=6.6545e-05) (hash(x)=25426430)
914
+ 876 train 7.200982 (lr=6.6133e-05) (hash(x)=26043735)
915
+ 877 train 7.143534 (lr=6.5723e-05) (hash(x)=22966314)
916
+ 878 train 7.447079 (lr=6.5317e-05) (hash(x)=27944619)
917
+ 879 train 7.199976 (lr=6.4913e-05) (hash(x)=23448505)
918
+ 880 train 7.223325 (lr=6.4513e-05) (hash(x)=24205942)
919
+ 881 train 7.129537 (lr=6.4115e-05) (hash(x)=26330693)
920
+ 882 train 7.141481 (lr=6.3721e-05) (hash(x)=24844466)
921
+ 883 train 7.108101 (lr=6.3329e-05) (hash(x)=29550596)
922
+ 884 train 7.190403 (lr=6.2941e-05) (hash(x)=20506540)
923
+ 885 train 7.221418 (lr=6.2556e-05) (hash(x)=22124892)
924
+ 886 train 7.207169 (lr=6.2173e-05) (hash(x)=24867036)
925
+ 887 train 7.240899 (lr=6.1794e-05) (hash(x)=25607640)
926
+ 888 train 7.213571 (lr=6.1418e-05) (hash(x)=26176930)
927
+ 889 train 7.018155 (lr=6.1045e-05) (hash(x)=23795894)
928
+ 890 train 6.935205 (lr=6.0675e-05) (hash(x)=23756298)
929
+ 891 train 7.108239 (lr=6.0308e-05) (hash(x)=24165449)
930
+ 892 train 7.144867 (lr=5.9944e-05) (hash(x)=26643739)
931
+ 893 train 7.334396 (lr=5.9583e-05) (hash(x)=24697756)
932
+ 894 train 7.234879 (lr=5.9225e-05) (hash(x)=24431101)
933
+ 895 train 7.396071 (lr=5.8871e-05) (hash(x)=27218899)
934
+ 896 train 7.174284 (lr=5.8519e-05) (hash(x)=29462219)
935
+ 897 train 7.242801 (lr=5.8171e-05) (hash(x)=24748522)
936
+ 898 train 7.228515 (lr=5.7826e-05) (hash(x)=24843486)
937
+ 899 train 7.402197 (lr=5.7484e-05) (hash(x)=24965317)
938
+ 900 val loss 7.3338
939
+ 900 val perplexity 1531.2627
940
+ 900 train 7.132629 (lr=5.7145e-05) (hash(x)=26584754)
941
+ 901 train 7.064291 (lr=5.6809e-05) (hash(x)=22031210)
942
+ 902 train 7.244216 (lr=5.6476e-05) (hash(x)=24860582)
943
+ 903 train 7.172890 (lr=5.6147e-05) (hash(x)=25877990)
944
+ 904 train 7.240592 (lr=5.5821e-05) (hash(x)=24970646)
945
+ 905 train 7.285748 (lr=5.5497e-05) (hash(x)=24382381)
946
+ 906 train 7.178356 (lr=5.5178e-05) (hash(x)=25221746)
947
+ 907 train 7.166002 (lr=5.4861e-05) (hash(x)=24551999)
948
+ 908 train 7.175597 (lr=5.4547e-05) (hash(x)=28151378)
949
+ 909 train 7.330429 (lr=5.4237e-05) (hash(x)=25765516)
950
+ 910 train 7.303267 (lr=5.3930e-05) (hash(x)=25630696)
951
+ 911 train 7.283819 (lr=5.3626e-05) (hash(x)=25204247)
952
+ 912 train 7.369516 (lr=5.3325e-05) (hash(x)=26326447)
953
+ 913 train 7.315760 (lr=5.3028e-05) (hash(x)=26321813)
954
+ 914 train 7.256454 (lr=5.2734e-05) (hash(x)=25400396)
955
+ 915 train 7.260317 (lr=5.2443e-05) (hash(x)=25528323)
956
+ 916 train 7.255816 (lr=5.2155e-05) (hash(x)=24035353)
957
+ 917 train 7.263180 (lr=5.1871e-05) (hash(x)=25114729)
958
+ 918 train 7.155625 (lr=5.1589e-05) (hash(x)=21634962)
959
+ 919 train 7.313724 (lr=5.1311e-05) (hash(x)=31341964)
960
+ 920 train 7.211611 (lr=5.1037e-05) (hash(x)=29172813)
961
+ 921 train 7.393250 (lr=5.0765e-05) (hash(x)=26326024)
962
+ 922 train 7.105816 (lr=5.0497e-05) (hash(x)=20140071)
963
+ 923 train 7.251973 (lr=5.0232e-05) (hash(x)=27273311)
964
+ 924 train 7.351120 (lr=4.9971e-05) (hash(x)=31276487)
965
+ 925 train 7.045094 (lr=4.9712e-05) (hash(x)=17819313)
966
+ 926 train 7.412124 (lr=4.9457e-05) (hash(x)=25843618)
967
+ 927 train 7.346514 (lr=4.9206e-05) (hash(x)=23769243)
968
+ 928 train 7.224030 (lr=4.8957e-05) (hash(x)=28731298)
969
+ 929 train 7.384650 (lr=4.8712e-05) (hash(x)=23788738)
970
+ 930 train 7.170901 (lr=4.8470e-05) (hash(x)=26599031)
971
+ 931 train 7.302535 (lr=4.8232e-05) (hash(x)=24818744)
972
+ 932 train 7.179393 (lr=4.7997e-05) (hash(x)=23053871)
973
+ 933 train 7.163075 (lr=4.7765e-05) (hash(x)=23472795)
974
+ 934 train 7.773467 (lr=4.7537e-05) (hash(x)=29748225)
975
+ 935 train 7.460744 (lr=4.7312e-05) (hash(x)=24991832)
976
+ 936 train 7.441347 (lr=4.7090e-05) (hash(x)=23681545)
977
+ 937 train 7.222974 (lr=4.6871e-05) (hash(x)=22424066)
978
+ 938 train 7.241458 (lr=4.6656e-05) (hash(x)=25262748)
979
+ 939 train 7.214849 (lr=4.6445e-05) (hash(x)=26474878)
980
+ 940 train 7.432466 (lr=4.6236e-05) (hash(x)=23930593)
981
+ 941 train 7.215434 (lr=4.6031e-05) (hash(x)=22704349)
982
+ 942 train 7.304269 (lr=4.5830e-05) (hash(x)=24801048)
983
+ 943 train 7.233383 (lr=4.5631e-05) (hash(x)=22293673)
984
+ 944 train 7.292034 (lr=4.5437e-05) (hash(x)=23311934)
985
+ 945 train 7.260554 (lr=4.5245e-05) (hash(x)=21785487)
986
+ 946 train 7.152358 (lr=4.5057e-05) (hash(x)=22981456)
987
+ 947 train 7.439977 (lr=4.4872e-05) (hash(x)=25251063)
988
+ 948 train 7.237205 (lr=4.4691e-05) (hash(x)=25796422)
989
+ 949 train 7.162023 (lr=4.4513e-05) (hash(x)=22722211)
990
+ 950 val loss 7.3227
991
+ 950 val perplexity 1514.2300
992
+ 950 train 7.247365 (lr=4.4338e-05) (hash(x)=24296755)
993
+ 951 train 7.260300 (lr=4.4167e-05) (hash(x)=24522269)
994
+ 952 train 7.241822 (lr=4.4000e-05) (hash(x)=25617184)
995
+ 953 train 7.155941 (lr=4.3835e-05) (hash(x)=23329475)
996
+ 954 train 7.255208 (lr=4.3674e-05) (hash(x)=30652062)
997
+ 955 train 7.204598 (lr=4.3517e-05) (hash(x)=24735165)
998
+ 956 train 7.011867 (lr=4.3363e-05) (hash(x)=23795686)
999
+ 957 train 7.054582 (lr=4.3212e-05) (hash(x)=27002892)
1000
+ 958 train 7.131990 (lr=4.3065e-05) (hash(x)=21189101)
1001
+ 959 train 7.044100 (lr=4.2921e-05) (hash(x)=23499349)
1002
+ 960 train 7.656157 (lr=4.2781e-05) (hash(x)=26807896)
1003
+ 961 train 7.217012 (lr=4.2644e-05) (hash(x)=22461589)
1004
+ 962 train 7.297737 (lr=4.2510e-05) (hash(x)=26904123)
1005
+ 963 train 7.126736 (lr=4.2380e-05) (hash(x)=26415040)
1006
+ 964 train 7.397498 (lr=4.2253e-05) (hash(x)=24463181)
1007
+ 965 train 7.469461 (lr=4.2130e-05) (hash(x)=26349465)
1008
+ 966 train 7.072570 (lr=4.2010e-05) (hash(x)=21205988)
1009
+ 967 train 7.110769 (lr=4.1894e-05) (hash(x)=21596677)
1010
+ 968 train 7.238688 (lr=4.1781e-05) (hash(x)=23575666)
1011
+ 969 train 7.048828 (lr=4.1672e-05) (hash(x)=24117012)
1012
+ 970 train 7.145145 (lr=4.1566e-05) (hash(x)=26013014)
1013
+ 971 train 7.291302 (lr=4.1463e-05) (hash(x)=24105761)
1014
+ 972 train 7.153425 (lr=4.1364e-05) (hash(x)=20369133)
1015
+ 973 train 7.183095 (lr=4.1269e-05) (hash(x)=25725049)
1016
+ 974 train 7.106108 (lr=4.1177e-05) (hash(x)=19808118)
1017
+ 975 train 6.998499 (lr=4.1088e-05) (hash(x)=22892776)
1018
+ 976 train 7.200702 (lr=4.1003e-05) (hash(x)=23736780)
1019
+ 977 train 7.138891 (lr=4.0921e-05) (hash(x)=21630383)
1020
+ 978 train 7.343306 (lr=4.0843e-05) (hash(x)=25397467)
1021
+ 979 train 7.037979 (lr=4.0768e-05) (hash(x)=24727094)
1022
+ 980 train 7.137217 (lr=4.0697e-05) (hash(x)=23951182)
1023
+ 981 train 7.079955 (lr=4.0629e-05) (hash(x)=22835000)
1024
+ 982 train 7.293881 (lr=4.0564e-05) (hash(x)=28767849)
1025
+ 983 train 7.260446 (lr=4.0503e-05) (hash(x)=22839049)
1026
+ 984 train 7.162802 (lr=4.0446e-05) (hash(x)=19930900)
1027
+ 985 train 7.070572 (lr=4.0392e-05) (hash(x)=20654136)
1028
+ 986 train 7.274242 (lr=4.0341e-05) (hash(x)=28495621)
1029
+ 987 train 7.435250 (lr=4.0294e-05) (hash(x)=26776000)
1030
+ 988 train 7.076031 (lr=4.0251e-05) (hash(x)=21584429)
1031
+ 989 train 6.860105 (lr=4.0211e-05) (hash(x)=22560658)
1032
+ 990 train 7.521848 (lr=4.0174e-05) (hash(x)=25276565)
1033
+ 991 train 7.294932 (lr=4.0141e-05) (hash(x)=25995933)
1034
+ 992 train 7.382233 (lr=4.0112e-05) (hash(x)=24163910)
1035
+ 993 train 7.304466 (lr=4.0085e-05) (hash(x)=24635394)
1036
+ 994 train 7.438890 (lr=4.0063e-05) (hash(x)=22556381)
1037
+ 995 train 7.504729 (lr=4.0044e-05) (hash(x)=27537228)
1038
+ 996 train 7.278789 (lr=4.0028e-05) (hash(x)=20883633)
1039
+ 997 train 7.431945 (lr=4.0016e-05) (hash(x)=26496730)
1040
+ 998 train 7.062856 (lr=4.0007e-05) (hash(x)=20342150)
1041
+ 999 val loss 7.3168
1042
+ 999 val perplexity 1505.4510
1043
+ 999 train 7.064411 (lr=4.0002e-05) (hash(x)=16533310)
lr4e-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 sha256:8d343788148276411233ba3444d558d035df740fa2a8bd74432bfcf05f395286
3
+ size 48156226
lr4e-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 sha256:3134c7b3eff4d7ce918606e9d48e6ec64c7005890f9eb681656a2df704d12dc4
3
+ size 71157574