andrew-healey commited on
Commit
2aa01dc
·
verified ·
1 Parent(s): df887a3

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size20480_seq_len256/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size20480_seq_len256", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 256, "batch_size": null, "total_batch_size": 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_len256/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_len256/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7488
3
+ 0 val perplexity 126607.6172
4
+ 0 train 11.750498 (lr=3.4965e-07) (hash(x)=12136211)
5
+ 1 train 11.757016 (lr=6.9930e-07) (hash(x)=9915468)
6
+ 2 train 11.768753 (lr=1.0490e-06) (hash(x)=13629277)
7
+ 3 train 11.756383 (lr=1.3986e-06) (hash(x)=12684830)
8
+ 4 train 11.742002 (lr=1.7483e-06) (hash(x)=12003126)
9
+ 5 train 11.749893 (lr=2.0979e-06) (hash(x)=12014715)
10
+ 6 train 11.716125 (lr=2.4476e-06) (hash(x)=9492167)
11
+ 7 train 11.717700 (lr=2.7972e-06) (hash(x)=12595649)
12
+ 8 train 11.694748 (lr=3.1469e-06) (hash(x)=11980631)
13
+ 9 train 11.734070 (lr=3.4965e-06) (hash(x)=11101031)
14
+ 10 train 11.697481 (lr=3.8462e-06) (hash(x)=14316349)
15
+ 11 train 11.697551 (lr=4.1958e-06) (hash(x)=12036401)
16
+ 12 train 11.679378 (lr=4.5455e-06) (hash(x)=12337125)
17
+ 13 train 11.686071 (lr=4.8951e-06) (hash(x)=11823810)
18
+ 14 train 11.646006 (lr=5.2448e-06) (hash(x)=13033959)
19
+ 15 train 11.662171 (lr=5.5944e-06) (hash(x)=13549925)
20
+ 16 train 11.618073 (lr=5.9441e-06) (hash(x)=14220701)
21
+ 17 train 11.609591 (lr=6.2937e-06) (hash(x)=13252262)
22
+ 18 train 11.613081 (lr=6.6434e-06) (hash(x)=14638145)
23
+ 19 train 11.573792 (lr=6.9930e-06) (hash(x)=13749054)
24
+ 20 train 11.577884 (lr=7.3427e-06) (hash(x)=12681760)
25
+ 21 train 11.546396 (lr=7.6923e-06) (hash(x)=14461706)
26
+ 22 train 11.507826 (lr=8.0420e-06) (hash(x)=14235218)
27
+ 23 train 11.487844 (lr=8.3916e-06) (hash(x)=11735995)
28
+ 24 train 11.463827 (lr=8.7413e-06) (hash(x)=11641112)
29
+ 25 train 11.470420 (lr=9.0909e-06) (hash(x)=12231701)
30
+ 26 train 11.429932 (lr=9.4406e-06) (hash(x)=10175470)
31
+ 27 train 11.437763 (lr=9.7902e-06) (hash(x)=14117945)
32
+ 28 train 11.379354 (lr=1.0140e-05) (hash(x)=13116544)
33
+ 29 train 11.322759 (lr=1.0490e-05) (hash(x)=12678303)
34
+ 30 train 11.322576 (lr=1.0839e-05) (hash(x)=11927462)
35
+ 31 train 11.295232 (lr=1.1189e-05) (hash(x)=12758461)
36
+ 32 train 11.303883 (lr=1.1538e-05) (hash(x)=11946230)
37
+ 33 train 11.218170 (lr=1.1888e-05) (hash(x)=13701311)
38
+ 34 train 11.212880 (lr=1.2238e-05) (hash(x)=11198382)
39
+ 35 train 11.163309 (lr=1.2587e-05) (hash(x)=14346433)
40
+ 36 train 11.165541 (lr=1.2937e-05) (hash(x)=17855984)
41
+ 37 train 11.192873 (lr=1.3287e-05) (hash(x)=18910052)
42
+ 38 train 11.118095 (lr=1.3636e-05) (hash(x)=11885403)
43
+ 39 train 11.079157 (lr=1.3986e-05) (hash(x)=12508350)
44
+ 40 train 11.027081 (lr=1.4336e-05) (hash(x)=10862804)
45
+ 41 train 11.000253 (lr=1.4685e-05) (hash(x)=14043988)
46
+ 42 train 10.883808 (lr=1.5035e-05) (hash(x)=11916406)
47
+ 43 train 10.900768 (lr=1.5385e-05) (hash(x)=11968227)
48
+ 44 train 10.865610 (lr=1.5734e-05) (hash(x)=13290191)
49
+ 45 train 10.853550 (lr=1.6084e-05) (hash(x)=15670776)
50
+ 46 train 10.769351 (lr=1.6434e-05) (hash(x)=14358384)
51
+ 47 train 10.781350 (lr=1.6783e-05) (hash(x)=11289341)
52
+ 48 train 10.719890 (lr=1.7133e-05) (hash(x)=14736057)
53
+ 49 train 10.794022 (lr=1.7483e-05) (hash(x)=11482477)
54
+ 50 val loss 10.6863
55
+ 50 val perplexity 43754.0039
56
+ 50 train 10.677802 (lr=1.7832e-05) (hash(x)=15429973)
57
+ 51 train 10.694228 (lr=1.8182e-05) (hash(x)=10345057)
58
+ 52 train 10.630721 (lr=1.8531e-05) (hash(x)=10816510)
59
+ 53 train 10.576385 (lr=1.8881e-05) (hash(x)=11157124)
60
+ 54 train 10.564247 (lr=1.9231e-05) (hash(x)=12349897)
61
+ 55 train 10.568815 (lr=1.9580e-05) (hash(x)=10750452)
62
+ 56 train 10.523222 (lr=1.9930e-05) (hash(x)=11859950)
63
+ 57 train 10.510275 (lr=2.0280e-05) (hash(x)=8921601)
64
+ 58 train 10.397503 (lr=2.0629e-05) (hash(x)=11333049)
65
+ 59 train 10.453532 (lr=2.0979e-05) (hash(x)=12165500)
66
+ 60 train 10.490846 (lr=2.1329e-05) (hash(x)=13153948)
67
+ 61 train 10.439623 (lr=2.1678e-05) (hash(x)=13569316)
68
+ 62 train 10.480467 (lr=2.2028e-05) (hash(x)=11456278)
69
+ 63 train 10.371814 (lr=2.2378e-05) (hash(x)=14043713)
70
+ 64 train 10.431572 (lr=2.2727e-05) (hash(x)=11896948)
71
+ 65 train 10.407022 (lr=2.3077e-05) (hash(x)=11969650)
72
+ 66 train 10.323514 (lr=2.3427e-05) (hash(x)=14387351)
73
+ 67 train 10.205634 (lr=2.3776e-05) (hash(x)=13680255)
74
+ 68 train 10.399764 (lr=2.4126e-05) (hash(x)=11875534)
75
+ 69 train 10.414874 (lr=2.4476e-05) (hash(x)=13871675)
76
+ 70 train 10.279976 (lr=2.4825e-05) (hash(x)=13389922)
77
+ 71 train 10.338611 (lr=2.5175e-05) (hash(x)=12263431)
78
+ 72 train 10.379599 (lr=2.5524e-05) (hash(x)=10141365)
79
+ 73 train 10.320871 (lr=2.5874e-05) (hash(x)=12414778)
80
+ 74 train 10.271598 (lr=2.6224e-05) (hash(x)=11738134)
81
+ 75 train 10.411397 (lr=2.6573e-05) (hash(x)=14443665)
82
+ 76 train 10.304598 (lr=2.6923e-05) (hash(x)=13435160)
83
+ 77 train 10.357785 (lr=2.7273e-05) (hash(x)=13100449)
84
+ 78 train 10.254242 (lr=2.7622e-05) (hash(x)=11690244)
85
+ 79 train 10.278501 (lr=2.7972e-05) (hash(x)=13953073)
86
+ 80 train 10.165719 (lr=2.8322e-05) (hash(x)=12657159)
87
+ 81 train 10.251500 (lr=2.8671e-05) (hash(x)=10956041)
88
+ 82 train 10.290764 (lr=2.9021e-05) (hash(x)=11060230)
89
+ 83 train 10.305162 (lr=2.9371e-05) (hash(x)=12840879)
90
+ 84 train 10.295480 (lr=2.9720e-05) (hash(x)=8916599)
91
+ 85 train 10.233208 (lr=3.0070e-05) (hash(x)=14428050)
92
+ 86 train 10.207709 (lr=3.0420e-05) (hash(x)=12787108)
93
+ 87 train 10.231522 (lr=3.0769e-05) (hash(x)=14104576)
94
+ 88 train 10.094816 (lr=3.1119e-05) (hash(x)=12623743)
95
+ 89 train 10.303168 (lr=3.1469e-05) (hash(x)=14453314)
96
+ 90 train 10.123506 (lr=3.1818e-05) (hash(x)=10470424)
97
+ 91 train 10.193630 (lr=3.2168e-05) (hash(x)=12918172)
98
+ 92 train 10.165788 (lr=3.2517e-05) (hash(x)=13259459)
99
+ 93 train 10.206167 (lr=3.2867e-05) (hash(x)=12368572)
100
+ 94 train 10.166409 (lr=3.3217e-05) (hash(x)=13505144)
101
+ 95 train 10.214725 (lr=3.3566e-05) (hash(x)=13340443)
102
+ 96 train 10.223375 (lr=3.3916e-05) (hash(x)=11469900)
103
+ 97 train 10.120539 (lr=3.4266e-05) (hash(x)=10742101)
104
+ 98 train 10.117004 (lr=3.4615e-05) (hash(x)=13261517)
105
+ 99 train 10.190212 (lr=3.4965e-05) (hash(x)=13025742)
106
+ 100 val loss 10.1434
107
+ 100 val perplexity 25422.5684
108
+ 100 train 10.224937 (lr=3.5315e-05) (hash(x)=15100390)
109
+ 101 train 10.161534 (lr=3.5664e-05) (hash(x)=14351068)
110
+ 102 train 10.151890 (lr=3.6014e-05) (hash(x)=13445545)
111
+ 103 train 10.128515 (lr=3.6364e-05) (hash(x)=13155663)
112
+ 104 train 10.123762 (lr=3.6713e-05) (hash(x)=13309898)
113
+ 105 train 10.118294 (lr=3.7063e-05) (hash(x)=9157946)
114
+ 106 train 10.063579 (lr=3.7413e-05) (hash(x)=12222186)
115
+ 107 train 10.127416 (lr=3.7762e-05) (hash(x)=13872294)
116
+ 108 train 10.051880 (lr=3.8112e-05) (hash(x)=11824400)
117
+ 109 train 10.039202 (lr=3.8462e-05) (hash(x)=12773612)
118
+ 110 train 10.087670 (lr=3.8811e-05) (hash(x)=12794078)
119
+ 111 train 10.033762 (lr=3.9161e-05) (hash(x)=12235713)
120
+ 112 train 10.058347 (lr=3.9510e-05) (hash(x)=10270693)
121
+ 113 train 10.012417 (lr=3.9860e-05) (hash(x)=13686441)
122
+ 114 train 10.134960 (lr=4.0210e-05) (hash(x)=13567687)
123
+ 115 train 10.042330 (lr=4.0559e-05) (hash(x)=13214962)
124
+ 116 train 10.030949 (lr=4.0909e-05) (hash(x)=12404167)
125
+ 117 train 9.970650 (lr=4.1259e-05) (hash(x)=12602681)
126
+ 118 train 10.053631 (lr=4.1608e-05) (hash(x)=13850551)
127
+ 119 train 9.953950 (lr=4.1958e-05) (hash(x)=13829219)
128
+ 120 train 10.021299 (lr=4.2308e-05) (hash(x)=15485762)
129
+ 121 train 10.011765 (lr=4.2657e-05) (hash(x)=9335220)
130
+ 122 train 9.977592 (lr=4.3007e-05) (hash(x)=13162637)
131
+ 123 train 9.941795 (lr=4.3357e-05) (hash(x)=15713183)
132
+ 124 train 9.938991 (lr=4.3706e-05) (hash(x)=13326615)
133
+ 125 train 9.960690 (lr=4.4056e-05) (hash(x)=9917297)
134
+ 126 train 9.882629 (lr=4.4406e-05) (hash(x)=6829764)
135
+ 127 train 9.874559 (lr=4.4755e-05) (hash(x)=12688412)
136
+ 128 train 9.772630 (lr=4.5105e-05) (hash(x)=13039934)
137
+ 129 train 9.915971 (lr=4.5455e-05) (hash(x)=12108991)
138
+ 130 train 9.905309 (lr=4.5804e-05) (hash(x)=14740147)
139
+ 131 train 9.873522 (lr=4.6154e-05) (hash(x)=12152216)
140
+ 132 train 9.845002 (lr=4.6503e-05) (hash(x)=14469749)
141
+ 133 train 9.848911 (lr=4.6853e-05) (hash(x)=11421483)
142
+ 134 train 9.805295 (lr=4.7203e-05) (hash(x)=12362234)
143
+ 135 train 9.806279 (lr=4.7552e-05) (hash(x)=13448155)
144
+ 136 train 9.795876 (lr=4.7902e-05) (hash(x)=10971430)
145
+ 137 train 9.797690 (lr=4.8252e-05) (hash(x)=11905081)
146
+ 138 train 9.727487 (lr=4.8601e-05) (hash(x)=10917438)
147
+ 139 train 9.699389 (lr=4.8951e-05) (hash(x)=12290857)
148
+ 140 train 9.874486 (lr=4.9301e-05) (hash(x)=13407076)
149
+ 141 train 9.812356 (lr=4.9650e-05) (hash(x)=13234933)
150
+ 142 train 9.742979 (lr=5.0000e-05) (hash(x)=12892366)
151
+ 143 train 9.810158 (lr=5.0350e-05) (hash(x)=9768085)
152
+ 144 train 9.698193 (lr=5.0699e-05) (hash(x)=12692791)
153
+ 145 train 9.653577 (lr=5.1049e-05) (hash(x)=9528974)
154
+ 146 train 9.713407 (lr=5.1399e-05) (hash(x)=10032842)
155
+ 147 train 9.761690 (lr=5.1748e-05) (hash(x)=10720403)
156
+ 148 train 9.627960 (lr=5.2098e-05) (hash(x)=15187742)
157
+ 149 train 9.680120 (lr=5.2448e-05) (hash(x)=12874592)
158
+ 150 val loss 9.6605
159
+ 150 val perplexity 15686.1572
160
+ 150 train 9.696652 (lr=5.2797e-05) (hash(x)=13622545)
161
+ 151 train 9.634906 (lr=5.3147e-05) (hash(x)=11025404)
162
+ 152 train 9.672260 (lr=5.3497e-05) (hash(x)=15190123)
163
+ 153 train 9.671119 (lr=5.3846e-05) (hash(x)=12567779)
164
+ 154 train 9.672889 (lr=5.4196e-05) (hash(x)=12439519)
165
+ 155 train 9.604033 (lr=5.4545e-05) (hash(x)=11826091)
166
+ 156 train 9.651455 (lr=5.4895e-05) (hash(x)=14723866)
167
+ 157 train 9.752838 (lr=5.5245e-05) (hash(x)=11259652)
168
+ 158 train 9.771477 (lr=5.5594e-05) (hash(x)=11135951)
169
+ 159 train 9.481782 (lr=5.5944e-05) (hash(x)=14642293)
170
+ 160 train 9.446519 (lr=5.6294e-05) (hash(x)=12485426)
171
+ 161 train 9.491729 (lr=5.6643e-05) (hash(x)=12923132)
172
+ 162 train 9.530642 (lr=5.6993e-05) (hash(x)=13559280)
173
+ 163 train 9.497276 (lr=5.7343e-05) (hash(x)=12169964)
174
+ 164 train 9.285131 (lr=5.7692e-05) (hash(x)=12000182)
175
+ 165 train 9.521669 (lr=5.8042e-05) (hash(x)=10948640)
176
+ 166 train 9.449587 (lr=5.8392e-05) (hash(x)=9666433)
177
+ 167 train 9.423132 (lr=5.8741e-05) (hash(x)=14103013)
178
+ 168 train 9.523787 (lr=5.9091e-05) (hash(x)=13553506)
179
+ 169 train 9.423008 (lr=5.9441e-05) (hash(x)=11135042)
180
+ 170 train 9.441401 (lr=5.9790e-05) (hash(x)=14930403)
181
+ 171 train 9.412402 (lr=6.0140e-05) (hash(x)=12404200)
182
+ 172 train 9.378063 (lr=6.0490e-05) (hash(x)=12280314)
183
+ 173 train 9.385564 (lr=6.0839e-05) (hash(x)=11534055)
184
+ 174 train 9.339800 (lr=6.1189e-05) (hash(x)=11308531)
185
+ 175 train 9.364421 (lr=6.1538e-05) (hash(x)=10914085)
186
+ 176 train 9.347036 (lr=6.1888e-05) (hash(x)=12357751)
187
+ 177 train 9.360201 (lr=6.2238e-05) (hash(x)=14570713)
188
+ 178 train 9.389318 (lr=6.2587e-05) (hash(x)=12212797)
189
+ 179 train 9.357871 (lr=6.2937e-05) (hash(x)=11680329)
190
+ 180 train 9.301850 (lr=6.3287e-05) (hash(x)=13116772)
191
+ 181 train 9.264181 (lr=6.3636e-05) (hash(x)=12437096)
192
+ 182 train 9.286768 (lr=6.3986e-05) (hash(x)=11220338)
193
+ 183 train 9.228655 (lr=6.4336e-05) (hash(x)=12654843)
194
+ 184 train 9.250509 (lr=6.4685e-05) (hash(x)=13505076)
195
+ 185 train 9.260365 (lr=6.5035e-05) (hash(x)=10606850)
196
+ 186 train 9.182314 (lr=6.5385e-05) (hash(x)=15357227)
197
+ 187 train 9.205271 (lr=6.5734e-05) (hash(x)=11281537)
198
+ 188 train 9.234916 (lr=6.6084e-05) (hash(x)=13253249)
199
+ 189 train 9.285473 (lr=6.6434e-05) (hash(x)=12363856)
200
+ 190 train 9.281722 (lr=6.6783e-05) (hash(x)=13580623)
201
+ 191 train 9.197352 (lr=6.7133e-05) (hash(x)=14001265)
202
+ 192 train 9.081429 (lr=6.7483e-05) (hash(x)=13045278)
203
+ 193 train 9.199512 (lr=6.7832e-05) (hash(x)=10444420)
204
+ 194 train 9.126186 (lr=6.8182e-05) (hash(x)=11832964)
205
+ 195 train 9.170285 (lr=6.8531e-05) (hash(x)=12249732)
206
+ 196 train 9.126087 (lr=6.8881e-05) (hash(x)=12307821)
207
+ 197 train 8.990306 (lr=6.9231e-05) (hash(x)=13803785)
208
+ 198 train 9.076365 (lr=6.9580e-05) (hash(x)=12493316)
209
+ 199 train 9.075480 (lr=6.9930e-05) (hash(x)=11689978)
210
+ 200 val loss 9.0601
211
+ 200 val perplexity 8604.8027
212
+ 200 train 9.040018 (lr=7.0280e-05) (hash(x)=11906763)
213
+ 201 train 9.064518 (lr=7.0629e-05) (hash(x)=13865869)
214
+ 202 train 9.154387 (lr=7.0979e-05) (hash(x)=16513375)
215
+ 203 train 9.171433 (lr=7.1329e-05) (hash(x)=13030618)
216
+ 204 train 9.026261 (lr=7.1678e-05) (hash(x)=13701949)
217
+ 205 train 9.028043 (lr=7.2028e-05) (hash(x)=14299833)
218
+ 206 train 9.047230 (lr=7.2378e-05) (hash(x)=11818854)
219
+ 207 train 8.955934 (lr=7.2727e-05) (hash(x)=12908875)
220
+ 208 train 8.990664 (lr=7.3077e-05) (hash(x)=12634656)
221
+ 209 train 9.021785 (lr=7.3427e-05) (hash(x)=13859304)
222
+ 210 train 9.051846 (lr=7.3776e-05) (hash(x)=14940204)
223
+ 211 train 8.952854 (lr=7.4126e-05) (hash(x)=14396964)
224
+ 212 train 8.843376 (lr=7.4476e-05) (hash(x)=13577749)
225
+ 213 train 8.971315 (lr=7.4825e-05) (hash(x)=11269801)
226
+ 214 train 8.875980 (lr=7.5175e-05) (hash(x)=11222906)
227
+ 215 train 8.813647 (lr=7.5524e-05) (hash(x)=14486071)
228
+ 216 train 8.900749 (lr=7.5874e-05) (hash(x)=11065744)
229
+ 217 train 8.914962 (lr=7.6224e-05) (hash(x)=11997361)
230
+ 218 train 8.893088 (lr=7.6573e-05) (hash(x)=12705537)
231
+ 219 train 8.914719 (lr=7.6923e-05) (hash(x)=11010336)
232
+ 220 train 8.908119 (lr=7.7273e-05) (hash(x)=15106564)
233
+ 221 train 8.846165 (lr=7.7622e-05) (hash(x)=10864248)
234
+ 222 train 8.753251 (lr=7.7972e-05) (hash(x)=13284954)
235
+ 223 train 8.880918 (lr=7.8322e-05) (hash(x)=11906151)
236
+ 224 train 8.849221 (lr=7.8671e-05) (hash(x)=8629295)
237
+ 225 train 8.778005 (lr=7.9021e-05) (hash(x)=15166897)
238
+ 226 train 8.841352 (lr=7.9371e-05) (hash(x)=13108020)
239
+ 227 train 8.721204 (lr=7.9720e-05) (hash(x)=12338963)
240
+ 228 train 8.681835 (lr=8.0070e-05) (hash(x)=17176736)
241
+ 229 train 8.783074 (lr=8.0420e-05) (hash(x)=13235312)
242
+ 230 train 8.749295 (lr=8.0769e-05) (hash(x)=12498316)
243
+ 231 train 8.764608 (lr=8.1119e-05) (hash(x)=11031343)
244
+ 232 train 8.721267 (lr=8.1469e-05) (hash(x)=12787203)
245
+ 233 train 8.764108 (lr=8.1818e-05) (hash(x)=14597757)
246
+ 234 train 8.739358 (lr=8.2168e-05) (hash(x)=12017158)
247
+ 235 train 8.706694 (lr=8.2517e-05) (hash(x)=13571036)
248
+ 236 train 8.723328 (lr=8.2867e-05) (hash(x)=12289875)
249
+ 237 train 8.750812 (lr=8.3217e-05) (hash(x)=13225100)
250
+ 238 train 8.658267 (lr=8.3566e-05) (hash(x)=11162887)
251
+ 239 train 8.733744 (lr=8.3916e-05) (hash(x)=12281916)
252
+ 240 train 8.642264 (lr=8.4266e-05) (hash(x)=12348477)
253
+ 241 train 8.659497 (lr=8.4615e-05) (hash(x)=13634084)
254
+ 242 train 8.687561 (lr=8.4965e-05) (hash(x)=11378506)
255
+ 243 train 8.625514 (lr=8.5315e-05) (hash(x)=11801649)
256
+ 244 train 8.622183 (lr=8.5664e-05) (hash(x)=10552788)
257
+ 245 train 8.543049 (lr=8.6014e-05) (hash(x)=11746289)
258
+ 246 train 8.617659 (lr=8.6364e-05) (hash(x)=13826293)
259
+ 247 train 8.651560 (lr=8.6713e-05) (hash(x)=10783313)
260
+ 248 train 8.624478 (lr=8.7063e-05) (hash(x)=12913270)
261
+ 249 train 8.622437 (lr=8.7413e-05) (hash(x)=14212467)
262
+ 250 val loss 8.5740
263
+ 250 val perplexity 5292.0361
264
+ 250 train 8.579944 (lr=8.7762e-05) (hash(x)=11495162)
265
+ 251 train 8.613096 (lr=8.8112e-05) (hash(x)=12994231)
266
+ 252 train 8.655664 (lr=8.8462e-05) (hash(x)=15319532)
267
+ 253 train 8.599000 (lr=8.8811e-05) (hash(x)=11485137)
268
+ 254 train 8.593289 (lr=8.9161e-05) (hash(x)=14010329)
269
+ 255 train 8.580050 (lr=8.9510e-05) (hash(x)=10470487)
270
+ 256 train 8.662729 (lr=8.9860e-05) (hash(x)=14077502)
271
+ 257 train 8.560527 (lr=9.0210e-05) (hash(x)=12205695)
272
+ 258 train 8.457286 (lr=9.0559e-05) (hash(x)=12096952)
273
+ 259 train 8.538946 (lr=9.0909e-05) (hash(x)=13985657)
274
+ 260 train 8.421590 (lr=9.1259e-05) (hash(x)=14868589)
275
+ 261 train 8.542755 (lr=9.1608e-05) (hash(x)=12944090)
276
+ 262 train 8.445544 (lr=9.1958e-05) (hash(x)=14652164)
277
+ 263 train 8.453466 (lr=9.2308e-05) (hash(x)=12831661)
278
+ 264 train 8.468572 (lr=9.2657e-05) (hash(x)=13472998)
279
+ 265 train 8.497800 (lr=9.3007e-05) (hash(x)=12806095)
280
+ 266 train 8.370776 (lr=9.3357e-05) (hash(x)=13410818)
281
+ 267 train 8.463455 (lr=9.3706e-05) (hash(x)=14617414)
282
+ 268 train 8.506482 (lr=9.4056e-05) (hash(x)=15066658)
283
+ 269 train 8.391692 (lr=9.4406e-05) (hash(x)=13221393)
284
+ 270 train 8.374786 (lr=9.4755e-05) (hash(x)=11528563)
285
+ 271 train 8.507977 (lr=9.5105e-05) (hash(x)=11041023)
286
+ 272 train 8.358087 (lr=9.5455e-05) (hash(x)=14885399)
287
+ 273 train 8.380039 (lr=9.5804e-05) (hash(x)=11362709)
288
+ 274 train 8.412096 (lr=9.6154e-05) (hash(x)=13083412)
289
+ 275 train 8.389572 (lr=9.6503e-05) (hash(x)=13186073)
290
+ 276 train 8.424591 (lr=9.6853e-05) (hash(x)=13988224)
291
+ 277 train 8.588092 (lr=9.7203e-05) (hash(x)=13274177)
292
+ 278 train 8.422585 (lr=9.7552e-05) (hash(x)=12363265)
293
+ 279 train 8.235192 (lr=9.7902e-05) (hash(x)=13507868)
294
+ 280 train 8.290188 (lr=9.8252e-05) (hash(x)=12531978)
295
+ 281 train 8.326450 (lr=9.8601e-05) (hash(x)=12512404)
296
+ 282 train 8.342612 (lr=9.8951e-05) (hash(x)=13439040)
297
+ 283 train 8.272526 (lr=9.9301e-05) (hash(x)=11221721)
298
+ 284 train 8.283500 (lr=9.9650e-05) (hash(x)=13854047)
299
+ 285 train 8.468013 (lr=1.0000e-04) (hash(x)=15372354)
300
+ 286 train 8.307048 (lr=1.0000e-04) (hash(x)=13070460)
301
+ 287 train 8.348429 (lr=1.0000e-04) (hash(x)=14689778)
302
+ 288 train 8.231705 (lr=9.9998e-05) (hash(x)=15364407)
303
+ 289 train 8.290404 (lr=9.9996e-05) (hash(x)=13156509)
304
+ 290 train 8.198296 (lr=9.9993e-05) (hash(x)=13358322)
305
+ 291 train 8.280965 (lr=9.9989e-05) (hash(x)=15734207)
306
+ 292 train 8.193215 (lr=9.9984e-05) (hash(x)=12334129)
307
+ 293 train 8.216010 (lr=9.9979e-05) (hash(x)=13758140)
308
+ 294 train 8.217403 (lr=9.9972e-05) (hash(x)=11809021)
309
+ 295 train 8.194597 (lr=9.9965e-05) (hash(x)=12925955)
310
+ 296 train 8.245657 (lr=9.9956e-05) (hash(x)=15708918)
311
+ 297 train 8.318339 (lr=9.9947e-05) (hash(x)=12750849)
312
+ 298 train 8.169288 (lr=9.9937e-05) (hash(x)=12269439)
313
+ 299 train 8.141202 (lr=9.9926e-05) (hash(x)=14327624)
314
+ 300 val loss 8.1373
315
+ 300 val perplexity 3419.7456
316
+ 300 train 8.182106 (lr=9.9915e-05) (hash(x)=12664301)
317
+ 301 train 8.086181 (lr=9.9902e-05) (hash(x)=11819313)
318
+ 302 train 8.087584 (lr=9.9889e-05) (hash(x)=13250863)
319
+ 303 train 8.058979 (lr=9.9874e-05) (hash(x)=12559422)
320
+ 304 train 8.087760 (lr=9.9859e-05) (hash(x)=12183244)
321
+ 305 train 8.251235 (lr=9.9843e-05) (hash(x)=12816359)
322
+ 306 train 8.161573 (lr=9.9826e-05) (hash(x)=11408330)
323
+ 307 train 8.159526 (lr=9.9808e-05) (hash(x)=10162896)
324
+ 308 train 8.270160 (lr=9.9789e-05) (hash(x)=12904775)
325
+ 309 train 8.045219 (lr=9.9770e-05) (hash(x)=12041552)
326
+ 310 train 8.060559 (lr=9.9749e-05) (hash(x)=14176315)
327
+ 311 train 8.036819 (lr=9.9728e-05) (hash(x)=10256290)
328
+ 312 train 8.094313 (lr=9.9706e-05) (hash(x)=12287950)
329
+ 313 train 8.056343 (lr=9.9683e-05) (hash(x)=14312559)
330
+ 314 train 8.119779 (lr=9.9659e-05) (hash(x)=10078167)
331
+ 315 train 8.021681 (lr=9.9634e-05) (hash(x)=11443472)
332
+ 316 train 8.046126 (lr=9.9609e-05) (hash(x)=11630498)
333
+ 317 train 8.039862 (lr=9.9582e-05) (hash(x)=10480150)
334
+ 318 train 8.109157 (lr=9.9555e-05) (hash(x)=13365961)
335
+ 319 train 8.165606 (lr=9.9526e-05) (hash(x)=13927238)
336
+ 320 train 8.067185 (lr=9.9497e-05) (hash(x)=12661142)
337
+ 321 train 8.073347 (lr=9.9467e-05) (hash(x)=16448258)
338
+ 322 train 8.216824 (lr=9.9437e-05) (hash(x)=12689745)
339
+ 323 train 7.996585 (lr=9.9405e-05) (hash(x)=10701487)
340
+ 324 train 8.162426 (lr=9.9372e-05) (hash(x)=13938367)
341
+ 325 train 8.090876 (lr=9.9339e-05) (hash(x)=13034568)
342
+ 326 train 8.151132 (lr=9.9305e-05) (hash(x)=13398831)
343
+ 327 train 8.196690 (lr=9.9270e-05) (hash(x)=14575913)
344
+ 328 train 7.980875 (lr=9.9234e-05) (hash(x)=12121790)
345
+ 329 train 7.935151 (lr=9.9197e-05) (hash(x)=10875163)
346
+ 330 train 8.045634 (lr=9.9159e-05) (hash(x)=13782256)
347
+ 331 train 8.398884 (lr=9.9121e-05) (hash(x)=13744329)
348
+ 332 train 8.046028 (lr=9.9081e-05) (hash(x)=14443216)
349
+ 333 train 7.939469 (lr=9.9041e-05) (hash(x)=8860388)
350
+ 334 train 7.929966 (lr=9.9000e-05) (hash(x)=13546558)
351
+ 335 train 7.962160 (lr=9.8958e-05) (hash(x)=14097210)
352
+ 336 train 8.032495 (lr=9.8915e-05) (hash(x)=12154683)
353
+ 337 train 7.954828 (lr=9.8872e-05) (hash(x)=11088127)
354
+ 338 train 8.017599 (lr=9.8827e-05) (hash(x)=15853795)
355
+ 339 train 7.890866 (lr=9.8782e-05) (hash(x)=14363085)
356
+ 340 train 7.856545 (lr=9.8736e-05) (hash(x)=12144036)
357
+ 341 train 8.026372 (lr=9.8689e-05) (hash(x)=14133983)
358
+ 342 train 7.954722 (lr=9.8641e-05) (hash(x)=13632329)
359
+ 343 train 7.912153 (lr=9.8592e-05) (hash(x)=9539498)
360
+ 344 train 7.851894 (lr=9.8543e-05) (hash(x)=11883263)
361
+ 345 train 8.173747 (lr=9.8492e-05) (hash(x)=17793549)
362
+ 346 train 7.947198 (lr=9.8441e-05) (hash(x)=11871877)
363
+ 347 train 7.948163 (lr=9.8389e-05) (hash(x)=14163458)
364
+ 348 train 7.979691 (lr=9.8336e-05) (hash(x)=15049180)
365
+ 349 train 7.920407 (lr=9.8282e-05) (hash(x)=14609724)
366
+ 350 val loss 7.8429
367
+ 350 val perplexity 2547.6252
368
+ 350 train 7.886538 (lr=9.8228e-05) (hash(x)=11157433)
369
+ 351 train 8.044968 (lr=9.8172e-05) (hash(x)=18154448)
370
+ 352 train 8.300301 (lr=9.8116e-05) (hash(x)=10541575)
371
+ 353 train 7.939246 (lr=9.8059e-05) (hash(x)=11995482)
372
+ 354 train 7.912446 (lr=9.8001e-05) (hash(x)=11153006)
373
+ 355 train 8.201874 (lr=9.7942e-05) (hash(x)=13713689)
374
+ 356 train 8.039267 (lr=9.7882e-05) (hash(x)=15113554)
375
+ 357 train 7.809542 (lr=9.7822e-05) (hash(x)=11555030)
376
+ 358 train 8.024803 (lr=9.7761e-05) (hash(x)=13374040)
377
+ 359 train 7.849205 (lr=9.7699e-05) (hash(x)=14115772)
378
+ 360 train 7.909541 (lr=9.7636e-05) (hash(x)=9720467)
379
+ 361 train 7.839819 (lr=9.7572e-05) (hash(x)=12279208)
380
+ 362 train 7.761903 (lr=9.7507e-05) (hash(x)=11796701)
381
+ 363 train 7.743669 (lr=9.7442e-05) (hash(x)=12021395)
382
+ 364 train 7.794352 (lr=9.7376e-05) (hash(x)=10845152)
383
+ 365 train 7.830493 (lr=9.7309e-05) (hash(x)=12617648)
384
+ 366 train 7.693230 (lr=9.7241e-05) (hash(x)=11035304)
385
+ 367 train 7.711281 (lr=9.7172e-05) (hash(x)=12798942)
386
+ 368 train 7.827740 (lr=9.7103e-05) (hash(x)=11344910)
387
+ 369 train 7.802253 (lr=9.7032e-05) (hash(x)=12366544)
388
+ 370 train 7.772810 (lr=9.6961e-05) (hash(x)=11321027)
389
+ 371 train 7.693099 (lr=9.6889e-05) (hash(x)=10886073)
390
+ 372 train 7.695246 (lr=9.6817e-05) (hash(x)=12334480)
391
+ 373 train 7.874717 (lr=9.6743e-05) (hash(x)=12555076)
392
+ 374 train 7.888361 (lr=9.6669e-05) (hash(x)=13525646)
393
+ 375 train 7.781948 (lr=9.6593e-05) (hash(x)=14777879)
394
+ 376 train 7.955957 (lr=9.6518e-05) (hash(x)=12138318)
395
+ 377 train 7.733130 (lr=9.6441e-05) (hash(x)=9927010)
396
+ 378 train 7.635750 (lr=9.6363e-05) (hash(x)=10265465)
397
+ 379 train 7.734950 (lr=9.6285e-05) (hash(x)=13073054)
398
+ 380 train 7.410352 (lr=9.6206e-05) (hash(x)=10517302)
399
+ 381 train 7.712189 (lr=9.6126e-05) (hash(x)=14967464)
400
+ 382 train 7.640289 (lr=9.6045e-05) (hash(x)=12190853)
401
+ 383 train 7.687412 (lr=9.5963e-05) (hash(x)=13463198)
402
+ 384 train 7.814234 (lr=9.5881e-05) (hash(x)=13635928)
403
+ 385 train 7.783516 (lr=9.5798e-05) (hash(x)=12209984)
404
+ 386 train 7.811677 (lr=9.5714e-05) (hash(x)=11946913)
405
+ 387 train 7.764989 (lr=9.5629e-05) (hash(x)=11983929)
406
+ 388 train 7.714751 (lr=9.5544e-05) (hash(x)=13044287)
407
+ 389 train 7.914819 (lr=9.5457e-05) (hash(x)=13502211)
408
+ 390 train 7.669989 (lr=9.5370e-05) (hash(x)=11732764)
409
+ 391 train 7.679683 (lr=9.5282e-05) (hash(x)=12563030)
410
+ 392 train 7.731166 (lr=9.5194e-05) (hash(x)=13370486)
411
+ 393 train 7.786655 (lr=9.5104e-05) (hash(x)=12774663)
412
+ 394 train 7.751382 (lr=9.5014e-05) (hash(x)=12581170)
413
+ 395 train 7.679021 (lr=9.4923e-05) (hash(x)=12170905)
414
+ 396 train 7.861018 (lr=9.4831e-05) (hash(x)=13035845)
415
+ 397 train 7.703469 (lr=9.4739e-05) (hash(x)=13348784)
416
+ 398 train 7.675406 (lr=9.4646e-05) (hash(x)=14533849)
417
+ 399 train 7.803824 (lr=9.4551e-05) (hash(x)=11525433)
418
+ 400 val loss 7.6920
419
+ 400 val perplexity 2190.6616
420
+ 400 train 7.735698 (lr=9.4457e-05) (hash(x)=7779807)
421
+ 401 train 7.675568 (lr=9.4361e-05) (hash(x)=9800296)
422
+ 402 train 7.737108 (lr=9.4265e-05) (hash(x)=11761671)
423
+ 403 train 7.697623 (lr=9.4168e-05) (hash(x)=10168268)
424
+ 404 train 7.693434 (lr=9.4070e-05) (hash(x)=10272263)
425
+ 405 train 7.633826 (lr=9.3971e-05) (hash(x)=10276843)
426
+ 406 train 8.305415 (lr=9.3872e-05) (hash(x)=11364751)
427
+ 407 train 7.506081 (lr=9.3772e-05) (hash(x)=12497404)
428
+ 408 train 7.697733 (lr=9.3671e-05) (hash(x)=14966009)
429
+ 409 train 7.745057 (lr=9.3569e-05) (hash(x)=12932339)
430
+ 410 train 7.755435 (lr=9.3467e-05) (hash(x)=8563011)
431
+ 411 train 7.771670 (lr=9.3364e-05) (hash(x)=14203584)
432
+ 412 train 7.592304 (lr=9.3260e-05) (hash(x)=10373543)
433
+ 413 train 7.701405 (lr=9.3155e-05) (hash(x)=13921321)
434
+ 414 train 7.688443 (lr=9.3050e-05) (hash(x)=12997326)
435
+ 415 train 7.782935 (lr=9.2944e-05) (hash(x)=10946961)
436
+ 416 train 7.762925 (lr=9.2837e-05) (hash(x)=11317673)
437
+ 417 train 7.837090 (lr=9.2729e-05) (hash(x)=12168723)
438
+ 418 train 7.778306 (lr=9.2621e-05) (hash(x)=11165585)
439
+ 419 train 7.595994 (lr=9.2512e-05) (hash(x)=10942254)
440
+ 420 train 7.687043 (lr=9.2402e-05) (hash(x)=12927776)
441
+ 421 train 8.019161 (lr=9.2292e-05) (hash(x)=10371632)
442
+ 422 train 8.024521 (lr=9.2181e-05) (hash(x)=12057844)
443
+ 423 train 7.788523 (lr=9.2069e-05) (hash(x)=16703203)
444
+ 424 train 7.665470 (lr=9.1956e-05) (hash(x)=12424204)
445
+ 425 train 7.818272 (lr=9.1843e-05) (hash(x)=17013185)
446
+ 426 train 7.915863 (lr=9.1729e-05) (hash(x)=11881739)
447
+ 427 train 7.772286 (lr=9.1614e-05) (hash(x)=15000294)
448
+ 428 train 7.650777 (lr=9.1499e-05) (hash(x)=13008387)
449
+ 429 train 7.567846 (lr=9.1382e-05) (hash(x)=10834161)
450
+ 430 train 7.683112 (lr=9.1266e-05) (hash(x)=15743796)
451
+ 431 train 7.737546 (lr=9.1148e-05) (hash(x)=14862449)
452
+ 432 train 7.843703 (lr=9.1030e-05) (hash(x)=11257626)
453
+ 433 train 7.784854 (lr=9.0911e-05) (hash(x)=10386532)
454
+ 434 train 7.478770 (lr=9.0791e-05) (hash(x)=11765736)
455
+ 435 train 7.726058 (lr=9.0671e-05) (hash(x)=13806874)
456
+ 436 train 7.695745 (lr=9.0550e-05) (hash(x)=12702206)
457
+ 437 train 7.721119 (lr=9.0428e-05) (hash(x)=12733894)
458
+ 438 train 7.841836 (lr=9.0306e-05) (hash(x)=13860937)
459
+ 439 train 7.913755 (lr=9.0182e-05) (hash(x)=12259111)
460
+ 440 train 7.660017 (lr=9.0059e-05) (hash(x)=14710730)
461
+ 441 train 7.661409 (lr=8.9934e-05) (hash(x)=10003332)
462
+ 442 train 7.606153 (lr=8.9809e-05) (hash(x)=13675590)
463
+ 443 train 7.434155 (lr=8.9683e-05) (hash(x)=11837867)
464
+ 444 train 7.932405 (lr=8.9557e-05) (hash(x)=11296484)
465
+ 445 train 7.679742 (lr=8.9430e-05) (hash(x)=13840537)
466
+ 446 train 7.653571 (lr=8.9302e-05) (hash(x)=14920737)
467
+ 447 train 7.647339 (lr=8.9173e-05) (hash(x)=12159808)
468
+ 448 train 7.643148 (lr=8.9044e-05) (hash(x)=9163004)
469
+ 449 train 7.668732 (lr=8.8914e-05) (hash(x)=13797226)
470
+ 450 val loss 7.6102
471
+ 450 val perplexity 2018.7356
472
+ 450 train 7.635316 (lr=8.8784e-05) (hash(x)=10204611)
473
+ 451 train 7.701510 (lr=8.8653e-05) (hash(x)=12372219)
474
+ 452 train 7.668253 (lr=8.8521e-05) (hash(x)=12434908)
475
+ 453 train 7.664319 (lr=8.8388e-05) (hash(x)=11371435)
476
+ 454 train 7.698636 (lr=8.8255e-05) (hash(x)=13523826)
477
+ 455 train 7.571047 (lr=8.8122e-05) (hash(x)=10910611)
478
+ 456 train 7.558871 (lr=8.7987e-05) (hash(x)=12905087)
479
+ 457 train 7.791741 (lr=8.7852e-05) (hash(x)=10985712)
480
+ 458 train 7.641401 (lr=8.7717e-05) (hash(x)=13245256)
481
+ 459 train 7.608690 (lr=8.7580e-05) (hash(x)=12509276)
482
+ 460 train 7.630426 (lr=8.7444e-05) (hash(x)=13538047)
483
+ 461 train 7.825688 (lr=8.7306e-05) (hash(x)=12953787)
484
+ 462 train 7.643093 (lr=8.7168e-05) (hash(x)=9468319)
485
+ 463 train 7.723547 (lr=8.7029e-05) (hash(x)=12720900)
486
+ 464 train 7.526983 (lr=8.6890e-05) (hash(x)=12030112)
487
+ 465 train 7.541805 (lr=8.6750e-05) (hash(x)=16075757)
488
+ 466 train 7.786083 (lr=8.6609e-05) (hash(x)=11540359)
489
+ 467 train 7.697885 (lr=8.6468e-05) (hash(x)=14048369)
490
+ 468 train 7.855339 (lr=8.6326e-05) (hash(x)=12961504)
491
+ 469 train 7.900305 (lr=8.6184e-05) (hash(x)=12602604)
492
+ 470 train 7.778429 (lr=8.6041e-05) (hash(x)=14173792)
493
+ 471 train 7.762057 (lr=8.5897e-05) (hash(x)=11552959)
494
+ 472 train 7.629869 (lr=8.5753e-05) (hash(x)=11986216)
495
+ 473 train 7.719839 (lr=8.5608e-05) (hash(x)=9890481)
496
+ 474 train 7.526625 (lr=8.5462e-05) (hash(x)=13052540)
497
+ 475 train 7.621621 (lr=8.5316e-05) (hash(x)=13423367)
498
+ 476 train 7.647476 (lr=8.5170e-05) (hash(x)=12494434)
499
+ 477 train 7.474033 (lr=8.5022e-05) (hash(x)=11555634)
500
+ 478 train 7.682240 (lr=8.4875e-05) (hash(x)=17108448)
501
+ 479 train 8.329215 (lr=8.4726e-05) (hash(x)=14914789)
502
+ 480 train 8.289883 (lr=8.4577e-05) (hash(x)=16237319)
503
+ 481 train 8.268783 (lr=8.4428e-05) (hash(x)=16763637)
504
+ 482 train 8.239808 (lr=8.4278e-05) (hash(x)=15222896)
505
+ 483 train 8.299661 (lr=8.4127e-05) (hash(x)=15783017)
506
+ 484 train 8.327317 (lr=8.3976e-05) (hash(x)=17102921)
507
+ 485 train 8.288534 (lr=8.3824e-05) (hash(x)=16354494)
508
+ 486 train 7.983100 (lr=8.3672e-05) (hash(x)=14615340)
509
+ 487 train 7.749866 (lr=8.3519e-05) (hash(x)=13512761)
510
+ 488 train 7.889485 (lr=8.3366e-05) (hash(x)=10750782)
511
+ 489 train 7.586683 (lr=8.3212e-05) (hash(x)=13751537)
512
+ 490 train 7.651933 (lr=8.3057e-05) (hash(x)=12343566)
513
+ 491 train 7.612168 (lr=8.2902e-05) (hash(x)=13830852)
514
+ 492 train 7.680100 (lr=8.2746e-05) (hash(x)=11170832)
515
+ 493 train 7.644613 (lr=8.2590e-05) (hash(x)=14841133)
516
+ 494 train 7.760303 (lr=8.2434e-05) (hash(x)=13683107)
517
+ 495 train 7.721925 (lr=8.2276e-05) (hash(x)=11832803)
518
+ 496 train 7.528362 (lr=8.2119e-05) (hash(x)=11362245)
519
+ 497 train 9.295002 (lr=8.1960e-05) (hash(x)=27572323)
520
+ 498 train 11.952278 (lr=8.1801e-05) (hash(x)=36479466)
521
+ 499 train 11.901650 (lr=8.1642e-05) (hash(x)=31819669)
522
+ 500 val loss 7.6306
523
+ 500 val perplexity 2060.1973
524
+ 500 train 8.378710 (lr=8.1482e-05) (hash(x)=12529262)
525
+ 501 train 7.308687 (lr=8.1322e-05) (hash(x)=11499251)
526
+ 502 train 7.698678 (lr=8.1161e-05) (hash(x)=12171109)
527
+ 503 train 7.738740 (lr=8.1000e-05) (hash(x)=13332674)
528
+ 504 train 7.609941 (lr=8.0838e-05) (hash(x)=11145516)
529
+ 505 train 7.549782 (lr=8.0676e-05) (hash(x)=11728990)
530
+ 506 train 7.623853 (lr=8.0513e-05) (hash(x)=12776190)
531
+ 507 train 7.731164 (lr=8.0349e-05) (hash(x)=15028799)
532
+ 508 train 7.675215 (lr=8.0186e-05) (hash(x)=13064591)
533
+ 509 train 7.623477 (lr=8.0021e-05) (hash(x)=13966115)
534
+ 510 train 7.689294 (lr=7.9856e-05) (hash(x)=11063257)
535
+ 511 train 7.598791 (lr=7.9691e-05) (hash(x)=10783848)
536
+ 512 train 7.649891 (lr=7.9525e-05) (hash(x)=11993069)
537
+ 513 train 7.520018 (lr=7.9359e-05) (hash(x)=11179737)
538
+ 514 train 7.822469 (lr=7.9192e-05) (hash(x)=13338266)
539
+ 515 train 7.617080 (lr=7.9025e-05) (hash(x)=10874609)
540
+ 516 train 7.640716 (lr=7.8858e-05) (hash(x)=11042965)
541
+ 517 train 7.797112 (lr=7.8689e-05) (hash(x)=9234657)
542
+ 518 train 7.727233 (lr=7.8521e-05) (hash(x)=13941557)
543
+ 519 train 7.776536 (lr=7.8352e-05) (hash(x)=12212655)
544
+ 520 train 7.641531 (lr=7.8182e-05) (hash(x)=10706283)
545
+ 521 train 7.656964 (lr=7.8012e-05) (hash(x)=12238338)
546
+ 522 train 7.774545 (lr=7.7842e-05) (hash(x)=11734682)
547
+ 523 train 7.529180 (lr=7.7671e-05) (hash(x)=11079662)
548
+ 524 train 7.750029 (lr=7.7500e-05) (hash(x)=13741369)
549
+ 525 train 7.542380 (lr=7.7328e-05) (hash(x)=13092665)
550
+ 526 train 7.674526 (lr=7.7156e-05) (hash(x)=12440906)
551
+ 527 train 7.694581 (lr=7.6984e-05) (hash(x)=11087308)
552
+ 528 train 7.581886 (lr=7.6811e-05) (hash(x)=11255508)
553
+ 529 train 7.869357 (lr=7.6637e-05) (hash(x)=17694381)
554
+ 530 train 7.586347 (lr=7.6463e-05) (hash(x)=11885779)
555
+ 531 train 7.619039 (lr=7.6289e-05) (hash(x)=13305061)
556
+ 532 train 7.578593 (lr=7.6115e-05) (hash(x)=11091420)
557
+ 533 train 7.672290 (lr=7.5940e-05) (hash(x)=11290716)
558
+ 534 train 7.602916 (lr=7.5764e-05) (hash(x)=13555230)
559
+ 535 train 7.709069 (lr=7.5588e-05) (hash(x)=13664986)
560
+ 536 train 7.615330 (lr=7.5412e-05) (hash(x)=14424424)
561
+ 537 train 7.634376 (lr=7.5235e-05) (hash(x)=12421790)
562
+ 538 train 7.691159 (lr=7.5058e-05) (hash(x)=11997110)
563
+ 539 train 7.638782 (lr=7.4881e-05) (hash(x)=12060664)
564
+ 540 train 7.553851 (lr=7.4703e-05) (hash(x)=12273644)
565
+ 541 train 7.996058 (lr=7.4525e-05) (hash(x)=13653671)
566
+ 542 train 7.735058 (lr=7.4346e-05) (hash(x)=16256309)
567
+ 543 train 7.701868 (lr=7.4167e-05) (hash(x)=15746941)
568
+ 544 train 7.677125 (lr=7.3988e-05) (hash(x)=13358884)
569
+ 545 train 7.526313 (lr=7.3808e-05) (hash(x)=12335295)
570
+ 546 train 7.543874 (lr=7.3628e-05) (hash(x)=13798560)
571
+ 547 train 7.962378 (lr=7.3448e-05) (hash(x)=17767151)
572
+ 548 train 7.520325 (lr=7.3267e-05) (hash(x)=8024709)
573
+ 549 train 7.579728 (lr=7.3086e-05) (hash(x)=13419880)
574
+ 550 val loss 7.5644
575
+ 550 val perplexity 1928.3872
576
+ 550 train 7.695450 (lr=7.2904e-05) (hash(x)=10643641)
577
+ 551 train 7.567838 (lr=7.2723e-05) (hash(x)=13277101)
578
+ 552 train 7.534144 (lr=7.2540e-05) (hash(x)=10131573)
579
+ 553 train 7.534633 (lr=7.2358e-05) (hash(x)=10535318)
580
+ 554 train 7.795949 (lr=7.2175e-05) (hash(x)=15487556)
581
+ 555 train 7.546603 (lr=7.1992e-05) (hash(x)=10935661)
582
+ 556 train 7.683620 (lr=7.1808e-05) (hash(x)=13602491)
583
+ 557 train 7.627586 (lr=7.1624e-05) (hash(x)=12907485)
584
+ 558 train 7.656606 (lr=7.1440e-05) (hash(x)=13358809)
585
+ 559 train 7.575094 (lr=7.1256e-05) (hash(x)=8154583)
586
+ 560 train 7.399242 (lr=7.1071e-05) (hash(x)=12377370)
587
+ 561 train 7.844400 (lr=7.0886e-05) (hash(x)=13402549)
588
+ 562 train 7.938786 (lr=7.0701e-05) (hash(x)=14795263)
589
+ 563 train 8.035863 (lr=7.0515e-05) (hash(x)=12543473)
590
+ 564 train 7.627172 (lr=7.0329e-05) (hash(x)=12711338)
591
+ 565 train 7.650981 (lr=7.0143e-05) (hash(x)=12144679)
592
+ 566 train 7.567720 (lr=6.9956e-05) (hash(x)=11447196)
593
+ 567 train 7.640002 (lr=6.9769e-05) (hash(x)=12101143)
594
+ 568 train 7.652726 (lr=6.9582e-05) (hash(x)=11815259)
595
+ 569 train 7.611240 (lr=6.9394e-05) (hash(x)=15908582)
596
+ 570 train 7.720849 (lr=6.9207e-05) (hash(x)=12254236)
597
+ 571 train 7.635460 (lr=6.9019e-05) (hash(x)=11655051)
598
+ 572 train 7.530869 (lr=6.8830e-05) (hash(x)=12008613)
599
+ 573 train 7.482622 (lr=6.8642e-05) (hash(x)=9481417)
600
+ 574 train 7.601671 (lr=6.8453e-05) (hash(x)=11426076)
601
+ 575 train 7.561141 (lr=6.8264e-05) (hash(x)=12561418)
602
+ 576 train 7.574310 (lr=6.8075e-05) (hash(x)=10629851)
603
+ 577 train 7.342249 (lr=6.7885e-05) (hash(x)=11036935)
604
+ 578 train 7.361825 (lr=6.7695e-05) (hash(x)=10602181)
605
+ 579 train 7.408870 (lr=6.7505e-05) (hash(x)=13778264)
606
+ 580 train 7.621270 (lr=6.7315e-05) (hash(x)=11172916)
607
+ 581 train 7.972743 (lr=6.7124e-05) (hash(x)=15215237)
608
+ 582 train 8.581721 (lr=6.6933e-05) (hash(x)=13372032)
609
+ 583 train 7.708009 (lr=6.6742e-05) (hash(x)=9877852)
610
+ 584 train 7.464104 (lr=6.6551e-05) (hash(x)=12921882)
611
+ 585 train 7.545200 (lr=6.6360e-05) (hash(x)=10232210)
612
+ 586 train 7.632014 (lr=6.6168e-05) (hash(x)=10525131)
613
+ 587 train 7.620470 (lr=6.5976e-05) (hash(x)=11968670)
614
+ 588 train 7.448349 (lr=6.5784e-05) (hash(x)=9852899)
615
+ 589 train 7.615911 (lr=6.5592e-05) (hash(x)=14272928)
616
+ 590 train 7.934957 (lr=6.5399e-05) (hash(x)=10305352)
617
+ 591 train 7.414033 (lr=6.5206e-05) (hash(x)=9256992)
618
+ 592 train 7.752252 (lr=6.5013e-05) (hash(x)=16001552)
619
+ 593 train 7.579786 (lr=6.4820e-05) (hash(x)=12593445)
620
+ 594 train 7.455137 (lr=6.4627e-05) (hash(x)=9538867)
621
+ 595 train 7.529832 (lr=6.4433e-05) (hash(x)=13208924)
622
+ 596 train 7.422639 (lr=6.4240e-05) (hash(x)=10608861)
623
+ 597 train 7.735832 (lr=6.4046e-05) (hash(x)=11578007)
624
+ 598 train 7.382231 (lr=6.3852e-05) (hash(x)=12664043)
625
+ 599 train 7.688111 (lr=6.3658e-05) (hash(x)=9762793)
626
+ 600 val loss 7.5513
627
+ 600 val perplexity 1903.3035
628
+ 600 train 7.519069 (lr=6.3463e-05) (hash(x)=17170548)
629
+ 601 train 7.936024 (lr=6.3269e-05) (hash(x)=12966781)
630
+ 602 train 7.767891 (lr=6.3074e-05) (hash(x)=15271720)
631
+ 603 train 7.949713 (lr=6.2879e-05) (hash(x)=15724292)
632
+ 604 train 7.803950 (lr=6.2684e-05) (hash(x)=16544013)
633
+ 605 train 8.121229 (lr=6.2489e-05) (hash(x)=15100923)
634
+ 606 train 8.392206 (lr=6.2294e-05) (hash(x)=16541600)
635
+ 607 train 7.531761 (lr=6.2098e-05) (hash(x)=13336322)
636
+ 608 train 7.624591 (lr=6.1903e-05) (hash(x)=11803722)
637
+ 609 train 7.724319 (lr=6.1707e-05) (hash(x)=10985744)
638
+ 610 train 7.588097 (lr=6.1511e-05) (hash(x)=11966467)
639
+ 611 train 7.600229 (lr=6.1315e-05) (hash(x)=12715509)
640
+ 612 train 7.508860 (lr=6.1119e-05) (hash(x)=13097367)
641
+ 613 train 7.583557 (lr=6.0923e-05) (hash(x)=11842498)
642
+ 614 train 7.518621 (lr=6.0726e-05) (hash(x)=12344103)
643
+ 615 train 7.487933 (lr=6.0530e-05) (hash(x)=9614210)
644
+ 616 train 7.522042 (lr=6.0333e-05) (hash(x)=11519808)
645
+ 617 train 7.488589 (lr=6.0137e-05) (hash(x)=11918721)
646
+ 618 train 7.500927 (lr=5.9940e-05) (hash(x)=12291116)
647
+ 619 train 7.519720 (lr=5.9743e-05) (hash(x)=12988502)
648
+ 620 train 7.531489 (lr=5.9546e-05) (hash(x)=14604879)
649
+ 621 train 7.499812 (lr=5.9349e-05) (hash(x)=9377388)
650
+ 622 train 7.546618 (lr=5.9152e-05) (hash(x)=11749413)
651
+ 623 train 7.474212 (lr=5.8955e-05) (hash(x)=12043981)
652
+ 624 train 7.771014 (lr=5.8758e-05) (hash(x)=12318014)
653
+ 625 train 7.479993 (lr=5.8560e-05) (hash(x)=12250543)
654
+ 626 train 7.503004 (lr=5.8363e-05) (hash(x)=13442493)
655
+ 627 train 7.430254 (lr=5.8165e-05) (hash(x)=11337762)
656
+ 628 train 7.413014 (lr=5.7968e-05) (hash(x)=13605096)
657
+ 629 train 7.870495 (lr=5.7770e-05) (hash(x)=13608297)
658
+ 630 train 7.822742 (lr=5.7573e-05) (hash(x)=14138512)
659
+ 631 train 7.646680 (lr=5.7375e-05) (hash(x)=10432200)
660
+ 632 train 7.592970 (lr=5.7177e-05) (hash(x)=11412290)
661
+ 633 train 7.497305 (lr=5.6979e-05) (hash(x)=12645115)
662
+ 634 train 7.527453 (lr=5.6782e-05) (hash(x)=12619839)
663
+ 635 train 7.520021 (lr=5.6584e-05) (hash(x)=12948458)
664
+ 636 train 7.754144 (lr=5.6386e-05) (hash(x)=17921494)
665
+ 637 train 7.564725 (lr=5.6188e-05) (hash(x)=12187458)
666
+ 638 train 7.470393 (lr=5.5990e-05) (hash(x)=12361500)
667
+ 639 train 7.518408 (lr=5.5792e-05) (hash(x)=15117606)
668
+ 640 train 7.496388 (lr=5.5594e-05) (hash(x)=12410779)
669
+ 641 train 7.578866 (lr=5.5396e-05) (hash(x)=12491246)
670
+ 642 train 7.657121 (lr=5.5198e-05) (hash(x)=12842140)
671
+ 643 train 7.370091 (lr=5.5000e-05) (hash(x)=9760285)
672
+ 644 train 7.606173 (lr=5.4802e-05) (hash(x)=12266640)
673
+ 645 train 7.587322 (lr=5.4604e-05) (hash(x)=12041167)
674
+ 646 train 7.467998 (lr=5.4406e-05) (hash(x)=12215148)
675
+ 647 train 7.525206 (lr=5.4208e-05) (hash(x)=11748314)
676
+ 648 train 7.280804 (lr=5.4010e-05) (hash(x)=8971899)
677
+ 649 train 7.217208 (lr=5.3812e-05) (hash(x)=11735134)
678
+ 650 val loss 7.5350
679
+ 650 val perplexity 1872.3959
680
+ 650 train 7.764126 (lr=5.3614e-05) (hash(x)=12720532)
681
+ 651 train 7.433120 (lr=5.3416e-05) (hash(x)=14573695)
682
+ 652 train 7.425504 (lr=5.3218e-05) (hash(x)=13490255)
683
+ 653 train 7.502119 (lr=5.3021e-05) (hash(x)=14306208)
684
+ 654 train 7.465779 (lr=5.2823e-05) (hash(x)=12834551)
685
+ 655 train 7.484664 (lr=5.2625e-05) (hash(x)=13862746)
686
+ 656 train 7.494234 (lr=5.2427e-05) (hash(x)=12633018)
687
+ 657 train 7.585762 (lr=5.2230e-05) (hash(x)=10815095)
688
+ 658 train 7.535668 (lr=5.2032e-05) (hash(x)=11964935)
689
+ 659 train 7.396994 (lr=5.1835e-05) (hash(x)=13686163)
690
+ 660 train 7.432317 (lr=5.1637e-05) (hash(x)=9989100)
691
+ 661 train 7.731620 (lr=5.1440e-05) (hash(x)=14377992)
692
+ 662 train 7.541930 (lr=5.1242e-05) (hash(x)=12180716)
693
+ 663 train 7.505529 (lr=5.1045e-05) (hash(x)=12421087)
694
+ 664 train 7.435418 (lr=5.0848e-05) (hash(x)=12172890)
695
+ 665 train 7.614666 (lr=5.0651e-05) (hash(x)=13085792)
696
+ 666 train 7.960114 (lr=5.0454e-05) (hash(x)=15341031)
697
+ 667 train 7.628849 (lr=5.0257e-05) (hash(x)=7228805)
698
+ 668 train 7.805502 (lr=5.0060e-05) (hash(x)=15931243)
699
+ 669 train 7.506854 (lr=4.9863e-05) (hash(x)=10015517)
700
+ 670 train 7.519019 (lr=4.9667e-05) (hash(x)=13147412)
701
+ 671 train 7.385489 (lr=4.9470e-05) (hash(x)=11774322)
702
+ 672 train 7.547682 (lr=4.9274e-05) (hash(x)=12661028)
703
+ 673 train 7.658445 (lr=4.9077e-05) (hash(x)=12674150)
704
+ 674 train 7.752096 (lr=4.8881e-05) (hash(x)=13961527)
705
+ 675 train 7.421803 (lr=4.8685e-05) (hash(x)=13116742)
706
+ 676 train 7.548195 (lr=4.8489e-05) (hash(x)=10925124)
707
+ 677 train 7.658323 (lr=4.8293e-05) (hash(x)=15946962)
708
+ 678 train 7.505723 (lr=4.8097e-05) (hash(x)=13627762)
709
+ 679 train 7.649233 (lr=4.7902e-05) (hash(x)=13171632)
710
+ 680 train 7.521770 (lr=4.7706e-05) (hash(x)=14365188)
711
+ 681 train 7.596652 (lr=4.7511e-05) (hash(x)=14334587)
712
+ 682 train 7.570400 (lr=4.7316e-05) (hash(x)=11123238)
713
+ 683 train 7.446250 (lr=4.7121e-05) (hash(x)=8960004)
714
+ 684 train 7.414915 (lr=4.6926e-05) (hash(x)=11757028)
715
+ 685 train 7.504332 (lr=4.6731e-05) (hash(x)=11642701)
716
+ 686 train 7.667151 (lr=4.6537e-05) (hash(x)=13542346)
717
+ 687 train 7.527892 (lr=4.6342e-05) (hash(x)=12969713)
718
+ 688 train 7.479717 (lr=4.6148e-05) (hash(x)=11014325)
719
+ 689 train 7.381589 (lr=4.5954e-05) (hash(x)=13050602)
720
+ 690 train 7.564822 (lr=4.5760e-05) (hash(x)=12485794)
721
+ 691 train 7.423027 (lr=4.5567e-05) (hash(x)=13929053)
722
+ 692 train 7.439317 (lr=4.5373e-05) (hash(x)=10868985)
723
+ 693 train 7.855669 (lr=4.5180e-05) (hash(x)=13096472)
724
+ 694 train 7.566772 (lr=4.4987e-05) (hash(x)=11312571)
725
+ 695 train 7.436089 (lr=4.4794e-05) (hash(x)=10986878)
726
+ 696 train 7.290952 (lr=4.4601e-05) (hash(x)=11221656)
727
+ 697 train 7.573928 (lr=4.4408e-05) (hash(x)=13231569)
728
+ 698 train 7.407230 (lr=4.4216e-05) (hash(x)=11134013)
729
+ 699 train 7.264856 (lr=4.4024e-05) (hash(x)=10360614)
730
+ 700 val loss 7.5205
731
+ 700 val perplexity 1845.4144
732
+ 700 train 7.408978 (lr=4.3832e-05) (hash(x)=12322428)
733
+ 701 train 7.488769 (lr=4.3640e-05) (hash(x)=14738108)
734
+ 702 train 7.543711 (lr=4.3449e-05) (hash(x)=11219270)
735
+ 703 train 7.504326 (lr=4.3258e-05) (hash(x)=13341647)
736
+ 704 train 7.632721 (lr=4.3067e-05) (hash(x)=13094087)
737
+ 705 train 7.539975 (lr=4.2876e-05) (hash(x)=12687712)
738
+ 706 train 7.512207 (lr=4.2685e-05) (hash(x)=13427553)
739
+ 707 train 7.410624 (lr=4.2495e-05) (hash(x)=12645920)
740
+ 708 train 7.472807 (lr=4.2305e-05) (hash(x)=12655070)
741
+ 709 train 7.433945 (lr=4.2115e-05) (hash(x)=13050918)
742
+ 710 train 7.585203 (lr=4.1925e-05) (hash(x)=13632267)
743
+ 711 train 7.391480 (lr=4.1736e-05) (hash(x)=10203871)
744
+ 712 train 7.493989 (lr=4.1547e-05) (hash(x)=11671729)
745
+ 713 train 7.473155 (lr=4.1358e-05) (hash(x)=12904927)
746
+ 714 train 7.673846 (lr=4.1170e-05) (hash(x)=13065951)
747
+ 715 train 7.559055 (lr=4.0981e-05) (hash(x)=14048300)
748
+ 716 train 7.601460 (lr=4.0793e-05) (hash(x)=12793660)
749
+ 717 train 7.423735 (lr=4.0606e-05) (hash(x)=11802275)
750
+ 718 train 7.406051 (lr=4.0418e-05) (hash(x)=14977049)
751
+ 719 train 7.657815 (lr=4.0231e-05) (hash(x)=13853279)
752
+ 720 train 7.699594 (lr=4.0044e-05) (hash(x)=12732461)
753
+ 721 train 7.548422 (lr=3.9857e-05) (hash(x)=12126692)
754
+ 722 train 7.501635 (lr=3.9671e-05) (hash(x)=13298673)
755
+ 723 train 7.385019 (lr=3.9485e-05) (hash(x)=14104463)
756
+ 724 train 7.369937 (lr=3.9299e-05) (hash(x)=15190284)
757
+ 725 train 7.341875 (lr=3.9114e-05) (hash(x)=13536583)
758
+ 726 train 7.677804 (lr=3.8929e-05) (hash(x)=14186475)
759
+ 727 train 7.473360 (lr=3.8744e-05) (hash(x)=13197896)
760
+ 728 train 7.487577 (lr=3.8560e-05) (hash(x)=11826535)
761
+ 729 train 7.598826 (lr=3.8376e-05) (hash(x)=13408658)
762
+ 730 train 7.503981 (lr=3.8192e-05) (hash(x)=9801100)
763
+ 731 train 7.438087 (lr=3.8008e-05) (hash(x)=11875111)
764
+ 732 train 7.525421 (lr=3.7825e-05) (hash(x)=11266409)
765
+ 733 train 7.460711 (lr=3.7642e-05) (hash(x)=15057683)
766
+ 734 train 7.476129 (lr=3.7460e-05) (hash(x)=12601025)
767
+ 735 train 7.615173 (lr=3.7277e-05) (hash(x)=11993762)
768
+ 736 train 7.494661 (lr=3.7096e-05) (hash(x)=12029315)
769
+ 737 train 7.398742 (lr=3.6914e-05) (hash(x)=10677495)
770
+ 738 train 7.446495 (lr=3.6733e-05) (hash(x)=14172570)
771
+ 739 train 7.574086 (lr=3.6552e-05) (hash(x)=10105877)
772
+ 740 train 7.369134 (lr=3.6372e-05) (hash(x)=13360078)
773
+ 741 train 7.502909 (lr=3.6192e-05) (hash(x)=14131566)
774
+ 742 train 7.530882 (lr=3.6012e-05) (hash(x)=10615972)
775
+ 743 train 7.407176 (lr=3.5833e-05) (hash(x)=9447587)
776
+ 744 train 7.408407 (lr=3.5654e-05) (hash(x)=10502681)
777
+ 745 train 7.423903 (lr=3.5475e-05) (hash(x)=10069430)
778
+ 746 train 7.429781 (lr=3.5297e-05) (hash(x)=12646864)
779
+ 747 train 7.366311 (lr=3.5119e-05) (hash(x)=14464065)
780
+ 748 train 7.440869 (lr=3.4942e-05) (hash(x)=11567941)
781
+ 749 train 7.447198 (lr=3.4765e-05) (hash(x)=11265273)
782
+ 750 val loss 7.5065
783
+ 750 val perplexity 1819.8342
784
+ 750 train 7.328534 (lr=3.4588e-05) (hash(x)=12165514)
785
+ 751 train 7.535354 (lr=3.4412e-05) (hash(x)=12235258)
786
+ 752 train 7.510880 (lr=3.4236e-05) (hash(x)=14053891)
787
+ 753 train 7.401894 (lr=3.4060e-05) (hash(x)=13640309)
788
+ 754 train 7.515865 (lr=3.3885e-05) (hash(x)=11437544)
789
+ 755 train 7.461086 (lr=3.3711e-05) (hash(x)=12814938)
790
+ 756 train 7.624629 (lr=3.3537e-05) (hash(x)=10855665)
791
+ 757 train 7.511748 (lr=3.3363e-05) (hash(x)=12343839)
792
+ 758 train 7.398901 (lr=3.3189e-05) (hash(x)=11548795)
793
+ 759 train 7.487174 (lr=3.3016e-05) (hash(x)=13007077)
794
+ 760 train 7.376635 (lr=3.2844e-05) (hash(x)=10289500)
795
+ 761 train 7.384637 (lr=3.2672e-05) (hash(x)=11835624)
796
+ 762 train 7.371810 (lr=3.2500e-05) (hash(x)=9785564)
797
+ 763 train 7.446455 (lr=3.2329e-05) (hash(x)=9419131)
798
+ 764 train 7.227596 (lr=3.2158e-05) (hash(x)=10301561)
799
+ 765 train 7.329410 (lr=3.1988e-05) (hash(x)=11209923)
800
+ 766 train 7.595720 (lr=3.1818e-05) (hash(x)=9643187)
801
+ 767 train 7.634659 (lr=3.1648e-05) (hash(x)=11977301)
802
+ 768 train 7.573479 (lr=3.1479e-05) (hash(x)=11990833)
803
+ 769 train 7.723416 (lr=3.1311e-05) (hash(x)=11326187)
804
+ 770 train 7.351125 (lr=3.1142e-05) (hash(x)=8063652)
805
+ 771 train 7.422317 (lr=3.0975e-05) (hash(x)=11772591)
806
+ 772 train 7.944698 (lr=3.0808e-05) (hash(x)=15314197)
807
+ 773 train 7.532793 (lr=3.0641e-05) (hash(x)=15539329)
808
+ 774 train 8.127290 (lr=3.0475e-05) (hash(x)=12872419)
809
+ 775 train 7.524062 (lr=3.0309e-05) (hash(x)=10882741)
810
+ 776 train 7.408575 (lr=3.0144e-05) (hash(x)=13856261)
811
+ 777 train 7.428920 (lr=2.9979e-05) (hash(x)=9341637)
812
+ 778 train 7.755001 (lr=2.9814e-05) (hash(x)=13061302)
813
+ 779 train 8.286707 (lr=2.9651e-05) (hash(x)=13174638)
814
+ 780 train 7.459581 (lr=2.9487e-05) (hash(x)=12334241)
815
+ 781 train 7.650098 (lr=2.9324e-05) (hash(x)=12490208)
816
+ 782 train 7.579385 (lr=2.9162e-05) (hash(x)=13425008)
817
+ 783 train 7.562628 (lr=2.9000e-05) (hash(x)=9018988)
818
+ 784 train 7.436006 (lr=2.8839e-05) (hash(x)=11365481)
819
+ 785 train 7.691954 (lr=2.8678e-05) (hash(x)=15752665)
820
+ 786 train 7.823944 (lr=2.8518e-05) (hash(x)=13894180)
821
+ 787 train 7.542935 (lr=2.8358e-05) (hash(x)=12406522)
822
+ 788 train 7.594835 (lr=2.8199e-05) (hash(x)=10427599)
823
+ 789 train 7.398746 (lr=2.8040e-05) (hash(x)=12323425)
824
+ 790 train 7.545414 (lr=2.7881e-05) (hash(x)=15411237)
825
+ 791 train 7.509255 (lr=2.7724e-05) (hash(x)=11486400)
826
+ 792 train 7.519699 (lr=2.7566e-05) (hash(x)=13984672)
827
+ 793 train 7.526208 (lr=2.7410e-05) (hash(x)=13731035)
828
+ 794 train 7.611662 (lr=2.7254e-05) (hash(x)=13318340)
829
+ 795 train 7.596618 (lr=2.7098e-05) (hash(x)=11434626)
830
+ 796 train 7.450301 (lr=2.6943e-05) (hash(x)=12972990)
831
+ 797 train 7.577713 (lr=2.6788e-05) (hash(x)=12174800)
832
+ 798 train 7.523933 (lr=2.6634e-05) (hash(x)=13177865)
833
+ 799 train 7.442208 (lr=2.6481e-05) (hash(x)=11810699)
834
+ 800 val loss 7.4940
835
+ 800 val perplexity 1797.1906
836
+ 800 train 7.523477 (lr=2.6328e-05) (hash(x)=16373080)
837
+ 801 train 7.508332 (lr=2.6176e-05) (hash(x)=10805897)
838
+ 802 train 7.434451 (lr=2.6024e-05) (hash(x)=11454278)
839
+ 803 train 7.529962 (lr=2.5873e-05) (hash(x)=15232839)
840
+ 804 train 7.668612 (lr=2.5722e-05) (hash(x)=13201125)
841
+ 805 train 7.599872 (lr=2.5572e-05) (hash(x)=15206031)
842
+ 806 train 7.594121 (lr=2.5423e-05) (hash(x)=12237999)
843
+ 807 train 7.730491 (lr=2.5274e-05) (hash(x)=12121527)
844
+ 808 train 7.706835 (lr=2.5125e-05) (hash(x)=14596430)
845
+ 809 train 7.649389 (lr=2.4978e-05) (hash(x)=12401285)
846
+ 810 train 7.453089 (lr=2.4830e-05) (hash(x)=13022136)
847
+ 811 train 7.498416 (lr=2.4684e-05) (hash(x)=12535938)
848
+ 812 train 7.328242 (lr=2.4538e-05) (hash(x)=14279137)
849
+ 813 train 7.431974 (lr=2.4392e-05) (hash(x)=11242362)
850
+ 814 train 7.343607 (lr=2.4247e-05) (hash(x)=11818848)
851
+ 815 train 7.511794 (lr=2.4103e-05) (hash(x)=13461204)
852
+ 816 train 7.505830 (lr=2.3959e-05) (hash(x)=13668560)
853
+ 817 train 7.452927 (lr=2.3816e-05) (hash(x)=13281798)
854
+ 818 train 7.520917 (lr=2.3674e-05) (hash(x)=11830904)
855
+ 819 train 7.722205 (lr=2.3532e-05) (hash(x)=13139637)
856
+ 820 train 7.371131 (lr=2.3391e-05) (hash(x)=11203468)
857
+ 821 train 7.357336 (lr=2.3250e-05) (hash(x)=13854352)
858
+ 822 train 7.340005 (lr=2.3110e-05) (hash(x)=8780793)
859
+ 823 train 7.442296 (lr=2.2971e-05) (hash(x)=12479097)
860
+ 824 train 7.463956 (lr=2.2832e-05) (hash(x)=12998035)
861
+ 825 train 7.318312 (lr=2.2694e-05) (hash(x)=12741634)
862
+ 826 train 7.762304 (lr=2.2556e-05) (hash(x)=21787605)
863
+ 827 train 8.981282 (lr=2.2420e-05) (hash(x)=13309670)
864
+ 828 train 7.502049 (lr=2.2283e-05) (hash(x)=11263630)
865
+ 829 train 7.452115 (lr=2.2148e-05) (hash(x)=11812959)
866
+ 830 train 7.515888 (lr=2.2013e-05) (hash(x)=10212403)
867
+ 831 train 7.299543 (lr=2.1878e-05) (hash(x)=13759175)
868
+ 832 train 7.490246 (lr=2.1745e-05) (hash(x)=10802273)
869
+ 833 train 7.446640 (lr=2.1612e-05) (hash(x)=12390919)
870
+ 834 train 7.578722 (lr=2.1479e-05) (hash(x)=13862166)
871
+ 835 train 7.902967 (lr=2.1347e-05) (hash(x)=14417261)
872
+ 836 train 7.489747 (lr=2.1216e-05) (hash(x)=12063172)
873
+ 837 train 7.507782 (lr=2.1086e-05) (hash(x)=13350903)
874
+ 838 train 7.349777 (lr=2.0956e-05) (hash(x)=12052329)
875
+ 839 train 7.491135 (lr=2.0827e-05) (hash(x)=13854240)
876
+ 840 train 7.411604 (lr=2.0698e-05) (hash(x)=10211543)
877
+ 841 train 7.925263 (lr=2.0570e-05) (hash(x)=13510491)
878
+ 842 train 7.341300 (lr=2.0443e-05) (hash(x)=10850957)
879
+ 843 train 7.669669 (lr=2.0317e-05) (hash(x)=12890682)
880
+ 844 train 7.467903 (lr=2.0191e-05) (hash(x)=11487746)
881
+ 845 train 7.460831 (lr=2.0066e-05) (hash(x)=12185495)
882
+ 846 train 7.948375 (lr=1.9941e-05) (hash(x)=12270540)
883
+ 847 train 7.280085 (lr=1.9818e-05) (hash(x)=12600677)
884
+ 848 train 7.630289 (lr=1.9694e-05) (hash(x)=12216705)
885
+ 849 train 7.686987 (lr=1.9572e-05) (hash(x)=13041967)
886
+ 850 val loss 7.4837
887
+ 850 val perplexity 1778.7335
888
+ 850 train 7.561824 (lr=1.9450e-05) (hash(x)=11046862)
889
+ 851 train 7.455768 (lr=1.9329e-05) (hash(x)=10171528)
890
+ 852 train 7.510191 (lr=1.9209e-05) (hash(x)=14053180)
891
+ 853 train 7.483024 (lr=1.9089e-05) (hash(x)=10640767)
892
+ 854 train 7.695465 (lr=1.8970e-05) (hash(x)=11657858)
893
+ 855 train 7.506225 (lr=1.8852e-05) (hash(x)=13458171)
894
+ 856 train 7.535892 (lr=1.8734e-05) (hash(x)=13938760)
895
+ 857 train 7.196707 (lr=1.8618e-05) (hash(x)=11457792)
896
+ 858 train 7.355162 (lr=1.8501e-05) (hash(x)=11314545)
897
+ 859 train 7.568974 (lr=1.8386e-05) (hash(x)=13471569)
898
+ 860 train 7.769083 (lr=1.8271e-05) (hash(x)=13344126)
899
+ 861 train 7.739677 (lr=1.8157e-05) (hash(x)=15321414)
900
+ 862 train 7.625020 (lr=1.8044e-05) (hash(x)=13913131)
901
+ 863 train 7.523155 (lr=1.7931e-05) (hash(x)=11813286)
902
+ 864 train 7.313213 (lr=1.7819e-05) (hash(x)=11185187)
903
+ 865 train 7.444724 (lr=1.7708e-05) (hash(x)=12763755)
904
+ 866 train 7.633044 (lr=1.7598e-05) (hash(x)=13912574)
905
+ 867 train 7.507348 (lr=1.7488e-05) (hash(x)=10811811)
906
+ 868 train 7.412333 (lr=1.7379e-05) (hash(x)=12611213)
907
+ 869 train 7.417366 (lr=1.7271e-05) (hash(x)=9411122)
908
+ 870 train 7.335207 (lr=1.7163e-05) (hash(x)=10049759)
909
+ 871 train 7.441987 (lr=1.7056e-05) (hash(x)=12462956)
910
+ 872 train 7.409105 (lr=1.6950e-05) (hash(x)=14589867)
911
+ 873 train 7.532036 (lr=1.6845e-05) (hash(x)=12764015)
912
+ 874 train 7.924270 (lr=1.6740e-05) (hash(x)=16357780)
913
+ 875 train 7.408873 (lr=1.6636e-05) (hash(x)=13306710)
914
+ 876 train 7.403541 (lr=1.6533e-05) (hash(x)=12575134)
915
+ 877 train 7.330746 (lr=1.6431e-05) (hash(x)=12243473)
916
+ 878 train 7.646022 (lr=1.6329e-05) (hash(x)=13985094)
917
+ 879 train 7.401456 (lr=1.6228e-05) (hash(x)=10772649)
918
+ 880 train 7.430405 (lr=1.6128e-05) (hash(x)=11206603)
919
+ 881 train 7.298787 (lr=1.6029e-05) (hash(x)=15248203)
920
+ 882 train 7.330988 (lr=1.5930e-05) (hash(x)=12324342)
921
+ 883 train 7.297102 (lr=1.5832e-05) (hash(x)=14886975)
922
+ 884 train 7.375770 (lr=1.5735e-05) (hash(x)=8601995)
923
+ 885 train 7.403505 (lr=1.5639e-05) (hash(x)=10539754)
924
+ 886 train 7.413294 (lr=1.5543e-05) (hash(x)=12118155)
925
+ 887 train 7.442027 (lr=1.5449e-05) (hash(x)=12534460)
926
+ 888 train 7.436259 (lr=1.5354e-05) (hash(x)=12258537)
927
+ 889 train 7.200191 (lr=1.5261e-05) (hash(x)=11540156)
928
+ 890 train 7.123234 (lr=1.5169e-05) (hash(x)=12248185)
929
+ 891 train 7.324625 (lr=1.5077e-05) (hash(x)=11701258)
930
+ 892 train 7.346245 (lr=1.4986e-05) (hash(x)=14024542)
931
+ 893 train 7.553559 (lr=1.4896e-05) (hash(x)=13548749)
932
+ 894 train 7.412991 (lr=1.4806e-05) (hash(x)=12754768)
933
+ 895 train 7.638433 (lr=1.4718e-05) (hash(x)=13546394)
934
+ 896 train 7.371797 (lr=1.4630e-05) (hash(x)=12812375)
935
+ 897 train 7.444770 (lr=1.4543e-05) (hash(x)=13046479)
936
+ 898 train 7.424682 (lr=1.4456e-05) (hash(x)=13611076)
937
+ 899 train 7.591202 (lr=1.4371e-05) (hash(x)=12655917)
938
+ 900 val loss 7.4754
939
+ 900 val perplexity 1764.0618
940
+ 900 train 7.302180 (lr=1.4286e-05) (hash(x)=12645297)
941
+ 901 train 7.198061 (lr=1.4202e-05) (hash(x)=10602039)
942
+ 902 train 7.439192 (lr=1.4119e-05) (hash(x)=12170417)
943
+ 903 train 7.358510 (lr=1.4037e-05) (hash(x)=12289584)
944
+ 904 train 7.470107 (lr=1.3955e-05) (hash(x)=14024020)
945
+ 905 train 7.490527 (lr=1.3874e-05) (hash(x)=12105301)
946
+ 906 train 7.376677 (lr=1.3794e-05) (hash(x)=12240359)
947
+ 907 train 7.377388 (lr=1.3715e-05) (hash(x)=11593067)
948
+ 908 train 7.380885 (lr=1.3637e-05) (hash(x)=14711937)
949
+ 909 train 7.565232 (lr=1.3559e-05) (hash(x)=13061011)
950
+ 910 train 7.500811 (lr=1.3482e-05) (hash(x)=12716767)
951
+ 911 train 7.488361 (lr=1.3407e-05) (hash(x)=13707506)
952
+ 912 train 7.578167 (lr=1.3331e-05) (hash(x)=13105180)
953
+ 913 train 7.525228 (lr=1.3257e-05) (hash(x)=11745344)
954
+ 914 train 7.452977 (lr=1.3183e-05) (hash(x)=13991242)
955
+ 915 train 7.476472 (lr=1.3111e-05) (hash(x)=13382803)
956
+ 916 train 7.457714 (lr=1.3039e-05) (hash(x)=12091242)
957
+ 917 train 7.472045 (lr=1.2968e-05) (hash(x)=10951259)
958
+ 918 train 7.357666 (lr=1.2897e-05) (hash(x)=10199997)
959
+ 919 train 7.509995 (lr=1.2828e-05) (hash(x)=15784201)
960
+ 920 train 7.417946 (lr=1.2759e-05) (hash(x)=15638908)
961
+ 921 train 7.605275 (lr=1.2691e-05) (hash(x)=13413178)
962
+ 922 train 7.251301 (lr=1.2624e-05) (hash(x)=9216051)
963
+ 923 train 7.441329 (lr=1.2558e-05) (hash(x)=14569416)
964
+ 924 train 7.553414 (lr=1.2493e-05) (hash(x)=15382648)
965
+ 925 train 7.198500 (lr=1.2428e-05) (hash(x)=10369396)
966
+ 926 train 7.599597 (lr=1.2364e-05) (hash(x)=12649047)
967
+ 927 train 7.542346 (lr=1.2301e-05) (hash(x)=12612709)
968
+ 928 train 7.430590 (lr=1.2239e-05) (hash(x)=11845067)
969
+ 929 train 7.599106 (lr=1.2178e-05) (hash(x)=12196831)
970
+ 930 train 7.370264 (lr=1.2118e-05) (hash(x)=14433763)
971
+ 931 train 7.496485 (lr=1.2058e-05) (hash(x)=13603725)
972
+ 932 train 7.378687 (lr=1.1999e-05) (hash(x)=10399699)
973
+ 933 train 7.341935 (lr=1.1941e-05) (hash(x)=12426620)
974
+ 934 train 7.989953 (lr=1.1884e-05) (hash(x)=14426514)
975
+ 935 train 7.671075 (lr=1.1828e-05) (hash(x)=13034137)
976
+ 936 train 7.617053 (lr=1.1772e-05) (hash(x)=13040891)
977
+ 937 train 7.420403 (lr=1.1718e-05) (hash(x)=10501727)
978
+ 938 train 7.428925 (lr=1.1664e-05) (hash(x)=12854212)
979
+ 939 train 7.405536 (lr=1.1611e-05) (hash(x)=13569897)
980
+ 940 train 7.632082 (lr=1.1559e-05) (hash(x)=12303031)
981
+ 941 train 7.408529 (lr=1.1508e-05) (hash(x)=10398118)
982
+ 942 train 7.516870 (lr=1.1457e-05) (hash(x)=11150447)
983
+ 943 train 7.415589 (lr=1.1408e-05) (hash(x)=11379182)
984
+ 944 train 7.493505 (lr=1.1359e-05) (hash(x)=10606523)
985
+ 945 train 7.484025 (lr=1.1311e-05) (hash(x)=9972768)
986
+ 946 train 7.342205 (lr=1.1264e-05) (hash(x)=13421904)
987
+ 947 train 7.611030 (lr=1.1218e-05) (hash(x)=10772043)
988
+ 948 train 7.421130 (lr=1.1173e-05) (hash(x)=13125366)
989
+ 949 train 7.347414 (lr=1.1128e-05) (hash(x)=11453247)
990
+ 950 val loss 7.4705
991
+ 950 val perplexity 1755.4825
992
+ 950 train 7.446085 (lr=1.1085e-05) (hash(x)=11647755)
993
+ 951 train 7.457398 (lr=1.1042e-05) (hash(x)=13194257)
994
+ 952 train 7.441273 (lr=1.1000e-05) (hash(x)=13526683)
995
+ 953 train 7.330533 (lr=1.0959e-05) (hash(x)=13336861)
996
+ 954 train 7.475265 (lr=1.0919e-05) (hash(x)=15214866)
997
+ 955 train 7.407276 (lr=1.0879e-05) (hash(x)=11410706)
998
+ 956 train 7.200205 (lr=1.0841e-05) (hash(x)=11295935)
999
+ 957 train 7.262398 (lr=1.0803e-05) (hash(x)=14280695)
1000
+ 958 train 7.347809 (lr=1.0766e-05) (hash(x)=11275747)
1001
+ 959 train 7.235084 (lr=1.0730e-05) (hash(x)=12234355)
1002
+ 960 train 7.880729 (lr=1.0695e-05) (hash(x)=13772206)
1003
+ 961 train 7.417648 (lr=1.0661e-05) (hash(x)=11419225)
1004
+ 962 train 7.529213 (lr=1.0628e-05) (hash(x)=14411675)
1005
+ 963 train 7.310859 (lr=1.0595e-05) (hash(x)=13227893)
1006
+ 964 train 7.633803 (lr=1.0563e-05) (hash(x)=12339321)
1007
+ 965 train 7.682081 (lr=1.0533e-05) (hash(x)=11841994)
1008
+ 966 train 7.261723 (lr=1.0503e-05) (hash(x)=10965980)
1009
+ 967 train 7.308535 (lr=1.0474e-05) (hash(x)=11006832)
1010
+ 968 train 7.433401 (lr=1.0445e-05) (hash(x)=11051445)
1011
+ 969 train 7.216193 (lr=1.0418e-05) (hash(x)=11812481)
1012
+ 970 train 7.364482 (lr=1.0391e-05) (hash(x)=13683659)
1013
+ 971 train 7.509920 (lr=1.0366e-05) (hash(x)=10942327)
1014
+ 972 train 7.338023 (lr=1.0341e-05) (hash(x)=9109755)
1015
+ 973 train 7.382717 (lr=1.0317e-05) (hash(x)=13048572)
1016
+ 974 train 7.270310 (lr=1.0294e-05) (hash(x)=9536519)
1017
+ 975 train 7.163351 (lr=1.0272e-05) (hash(x)=11307988)
1018
+ 976 train 7.378426 (lr=1.0251e-05) (hash(x)=11999078)
1019
+ 977 train 7.302054 (lr=1.0230e-05) (hash(x)=10678624)
1020
+ 978 train 7.535415 (lr=1.0211e-05) (hash(x)=12141167)
1021
+ 979 train 7.276427 (lr=1.0192e-05) (hash(x)=11990825)
1022
+ 980 train 7.335519 (lr=1.0174e-05) (hash(x)=12744639)
1023
+ 981 train 7.266788 (lr=1.0157e-05) (hash(x)=11321271)
1024
+ 982 train 7.485752 (lr=1.0141e-05) (hash(x)=18407440)
1025
+ 983 train 7.464877 (lr=1.0126e-05) (hash(x)=11230858)
1026
+ 984 train 7.356620 (lr=1.0111e-05) (hash(x)=11111338)
1027
+ 985 train 7.269953 (lr=1.0098e-05) (hash(x)=11886513)
1028
+ 986 train 7.496714 (lr=1.0085e-05) (hash(x)=15696837)
1029
+ 987 train 7.631768 (lr=1.0074e-05) (hash(x)=13052292)
1030
+ 988 train 7.287791 (lr=1.0063e-05) (hash(x)=10686162)
1031
+ 989 train 7.006116 (lr=1.0053e-05) (hash(x)=10598277)
1032
+ 990 train 7.724509 (lr=1.0044e-05) (hash(x)=13240516)
1033
+ 991 train 7.509223 (lr=1.0035e-05) (hash(x)=13337837)
1034
+ 992 train 7.589272 (lr=1.0028e-05) (hash(x)=12010843)
1035
+ 993 train 7.515950 (lr=1.0021e-05) (hash(x)=12914018)
1036
+ 994 train 7.628997 (lr=1.0016e-05) (hash(x)=10815854)
1037
+ 995 train 7.707650 (lr=1.0011e-05) (hash(x)=13648351)
1038
+ 996 train 7.479869 (lr=1.0007e-05) (hash(x)=10424425)
1039
+ 997 train 7.622347 (lr=1.0004e-05) (hash(x)=11819798)
1040
+ 998 train 7.260644 (lr=1.0002e-05) (hash(x)=9746910)
1041
+ 999 val loss 7.4648
1042
+ 999 val perplexity 1745.5603
1043
+ 999 train 7.243228 (lr=1.0000e-05) (hash(x)=8133784)
lr1e-4_total_batch_size20480_seq_len256/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d03ce4dcf540f4bc84332b26db48dd4bd3d1853bbb9b91ca302e0c136413f23f
3
+ size 38587970
lr1e-4_total_batch_size20480_seq_len256/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1f0d8cb2f08241858fdc5f0fe43f43f11eefdf736171db133cd7044e6a2fa1d
3
+ size 70895430