andrew-healey commited on
Commit
df887a3
·
verified ·
1 Parent(s): f55407c

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size20480_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size20480_seq_len128", "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": 128, "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_len128/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_len128/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.6870
3
+ 0 val perplexity 119016.5000
4
+ 0 train 11.700421 (lr=3.4965e-07) (hash(x)=5113283)
5
+ 1 train 11.682228 (lr=6.9930e-07) (hash(x)=5534142)
6
+ 2 train 11.712296 (lr=1.0490e-06) (hash(x)=6763231)
7
+ 3 train 11.697494 (lr=1.3986e-06) (hash(x)=6328012)
8
+ 4 train 11.691818 (lr=1.7483e-06) (hash(x)=7052161)
9
+ 5 train 11.691360 (lr=2.0979e-06) (hash(x)=5556098)
10
+ 6 train 11.683790 (lr=2.4476e-06) (hash(x)=4696943)
11
+ 7 train 11.662252 (lr=2.7972e-06) (hash(x)=6219138)
12
+ 8 train 11.644224 (lr=3.1469e-06) (hash(x)=6049878)
13
+ 9 train 11.646851 (lr=3.4965e-06) (hash(x)=5491972)
14
+ 10 train 11.651977 (lr=3.8462e-06) (hash(x)=8075458)
15
+ 11 train 11.621681 (lr=4.1958e-06) (hash(x)=5917741)
16
+ 12 train 11.614285 (lr=4.5455e-06) (hash(x)=5750403)
17
+ 13 train 11.600780 (lr=4.8951e-06) (hash(x)=6112458)
18
+ 14 train 11.584670 (lr=5.2448e-06) (hash(x)=5806490)
19
+ 15 train 11.559491 (lr=5.5944e-06) (hash(x)=6568379)
20
+ 16 train 11.540793 (lr=5.9441e-06) (hash(x)=7212403)
21
+ 17 train 11.507972 (lr=6.2937e-06) (hash(x)=7386082)
22
+ 18 train 11.492175 (lr=6.6434e-06) (hash(x)=6359743)
23
+ 19 train 11.480099 (lr=6.9930e-06) (hash(x)=6151674)
24
+ 20 train 11.458134 (lr=7.3427e-06) (hash(x)=7168704)
25
+ 21 train 11.462021 (lr=7.6923e-06) (hash(x)=7859060)
26
+ 22 train 11.394769 (lr=8.0420e-06) (hash(x)=6976239)
27
+ 23 train 11.398275 (lr=8.3916e-06) (hash(x)=6391506)
28
+ 24 train 11.355160 (lr=8.7413e-06) (hash(x)=5245999)
29
+ 25 train 11.330033 (lr=9.0909e-06) (hash(x)=5985675)
30
+ 26 train 11.289556 (lr=9.4406e-06) (hash(x)=4355733)
31
+ 27 train 11.292900 (lr=9.7902e-06) (hash(x)=6984772)
32
+ 28 train 11.247354 (lr=1.0140e-05) (hash(x)=6244096)
33
+ 29 train 11.201186 (lr=1.0490e-05) (hash(x)=6603717)
34
+ 30 train 11.138947 (lr=1.0839e-05) (hash(x)=5669522)
35
+ 31 train 11.122600 (lr=1.1189e-05) (hash(x)=6190579)
36
+ 32 train 11.127979 (lr=1.1538e-05) (hash(x)=4912983)
37
+ 33 train 11.026930 (lr=1.1888e-05) (hash(x)=7276303)
38
+ 34 train 11.022292 (lr=1.2238e-05) (hash(x)=5479947)
39
+ 35 train 11.061768 (lr=1.2587e-05) (hash(x)=6630378)
40
+ 36 train 10.985852 (lr=1.2937e-05) (hash(x)=8097010)
41
+ 37 train 11.013044 (lr=1.3287e-05) (hash(x)=9989744)
42
+ 38 train 10.901842 (lr=1.3636e-05) (hash(x)=6694932)
43
+ 39 train 10.844536 (lr=1.3986e-05) (hash(x)=6317894)
44
+ 40 train 10.762856 (lr=1.4336e-05) (hash(x)=5046327)
45
+ 41 train 10.786942 (lr=1.4685e-05) (hash(x)=7947703)
46
+ 42 train 10.681293 (lr=1.5035e-05) (hash(x)=5916068)
47
+ 43 train 10.723731 (lr=1.5385e-05) (hash(x)=5777498)
48
+ 44 train 10.680823 (lr=1.5734e-05) (hash(x)=5753092)
49
+ 45 train 10.655180 (lr=1.6084e-05) (hash(x)=7723226)
50
+ 46 train 10.622224 (lr=1.6434e-05) (hash(x)=7007016)
51
+ 47 train 10.577086 (lr=1.6783e-05) (hash(x)=6491229)
52
+ 48 train 10.579188 (lr=1.7133e-05) (hash(x)=6390752)
53
+ 49 train 10.690311 (lr=1.7483e-05) (hash(x)=6019389)
54
+ 50 val loss 10.5770
55
+ 50 val perplexity 39220.3867
56
+ 50 train 10.562006 (lr=1.7832e-05) (hash(x)=7662067)
57
+ 51 train 10.577480 (lr=1.8182e-05) (hash(x)=4648609)
58
+ 52 train 10.520283 (lr=1.8531e-05) (hash(x)=4725966)
59
+ 53 train 10.488559 (lr=1.8881e-05) (hash(x)=5558355)
60
+ 54 train 10.538498 (lr=1.9231e-05) (hash(x)=6524933)
61
+ 55 train 10.487376 (lr=1.9580e-05) (hash(x)=6369326)
62
+ 56 train 10.465241 (lr=1.9930e-05) (hash(x)=6022625)
63
+ 57 train 10.434535 (lr=2.0280e-05) (hash(x)=4906853)
64
+ 58 train 10.284534 (lr=2.0629e-05) (hash(x)=5589269)
65
+ 59 train 10.405723 (lr=2.0979e-05) (hash(x)=6437376)
66
+ 60 train 10.447899 (lr=2.1329e-05) (hash(x)=7304892)
67
+ 61 train 10.418019 (lr=2.1678e-05) (hash(x)=7084093)
68
+ 62 train 10.444640 (lr=2.2028e-05) (hash(x)=5908178)
69
+ 63 train 10.291209 (lr=2.2378e-05) (hash(x)=7048804)
70
+ 64 train 10.380711 (lr=2.2727e-05) (hash(x)=6085549)
71
+ 65 train 10.384246 (lr=2.3077e-05) (hash(x)=5752594)
72
+ 66 train 10.257653 (lr=2.3427e-05) (hash(x)=7662769)
73
+ 67 train 10.141635 (lr=2.3776e-05) (hash(x)=6393520)
74
+ 68 train 10.390976 (lr=2.4126e-05) (hash(x)=5942867)
75
+ 69 train 10.405921 (lr=2.4476e-05) (hash(x)=6793550)
76
+ 70 train 10.303573 (lr=2.4825e-05) (hash(x)=7105976)
77
+ 71 train 10.369252 (lr=2.5175e-05) (hash(x)=6581348)
78
+ 72 train 10.340770 (lr=2.5524e-05) (hash(x)=5555598)
79
+ 73 train 10.284508 (lr=2.5874e-05) (hash(x)=6235837)
80
+ 74 train 10.256501 (lr=2.6224e-05) (hash(x)=5591770)
81
+ 75 train 10.459804 (lr=2.6573e-05) (hash(x)=7083021)
82
+ 76 train 10.321479 (lr=2.6923e-05) (hash(x)=7281348)
83
+ 77 train 10.358477 (lr=2.7273e-05) (hash(x)=6155821)
84
+ 78 train 10.260435 (lr=2.7622e-05) (hash(x)=5736241)
85
+ 79 train 10.240286 (lr=2.7972e-05) (hash(x)=5657123)
86
+ 80 train 10.121307 (lr=2.8322e-05) (hash(x)=6452095)
87
+ 81 train 10.204829 (lr=2.8671e-05) (hash(x)=5312267)
88
+ 82 train 10.249396 (lr=2.9021e-05) (hash(x)=5434493)
89
+ 83 train 10.286786 (lr=2.9371e-05) (hash(x)=6830813)
90
+ 84 train 10.236574 (lr=2.9720e-05) (hash(x)=4752771)
91
+ 85 train 10.234578 (lr=3.0070e-05) (hash(x)=6424933)
92
+ 86 train 10.185545 (lr=3.0420e-05) (hash(x)=5786517)
93
+ 87 train 10.203669 (lr=3.0769e-05) (hash(x)=6680023)
94
+ 88 train 10.117901 (lr=3.1119e-05) (hash(x)=6474729)
95
+ 89 train 10.194358 (lr=3.1469e-05) (hash(x)=7523279)
96
+ 90 train 10.077053 (lr=3.1818e-05) (hash(x)=5508768)
97
+ 91 train 10.118877 (lr=3.2168e-05) (hash(x)=6635925)
98
+ 92 train 10.147852 (lr=3.2517e-05) (hash(x)=6663801)
99
+ 93 train 10.129238 (lr=3.2867e-05) (hash(x)=5869441)
100
+ 94 train 10.124088 (lr=3.3217e-05) (hash(x)=6698878)
101
+ 95 train 10.127551 (lr=3.3566e-05) (hash(x)=6349004)
102
+ 96 train 10.119530 (lr=3.3916e-05) (hash(x)=4986137)
103
+ 97 train 9.948666 (lr=3.4266e-05) (hash(x)=6008454)
104
+ 98 train 9.990538 (lr=3.4615e-05) (hash(x)=6550770)
105
+ 99 train 10.087302 (lr=3.4965e-05) (hash(x)=7682741)
106
+ 100 val loss 10.0828
107
+ 100 val perplexity 23928.9824
108
+ 100 train 10.112022 (lr=3.5315e-05) (hash(x)=7892158)
109
+ 101 train 10.060993 (lr=3.5664e-05) (hash(x)=7794028)
110
+ 102 train 10.067136 (lr=3.6014e-05) (hash(x)=7093523)
111
+ 103 train 10.035433 (lr=3.6364e-05) (hash(x)=6166062)
112
+ 104 train 9.965083 (lr=3.6713e-05) (hash(x)=6645781)
113
+ 105 train 10.025417 (lr=3.7063e-05) (hash(x)=4811192)
114
+ 106 train 9.922396 (lr=3.7413e-05) (hash(x)=5970866)
115
+ 107 train 9.992115 (lr=3.7762e-05) (hash(x)=7822680)
116
+ 108 train 9.897361 (lr=3.8112e-05) (hash(x)=6317317)
117
+ 109 train 9.911785 (lr=3.8462e-05) (hash(x)=6431409)
118
+ 110 train 9.932535 (lr=3.8811e-05) (hash(x)=6653337)
119
+ 111 train 9.892779 (lr=3.9161e-05) (hash(x)=5323032)
120
+ 112 train 9.857840 (lr=3.9510e-05) (hash(x)=5125339)
121
+ 113 train 9.867743 (lr=3.9860e-05) (hash(x)=7363286)
122
+ 114 train 9.905087 (lr=4.0210e-05) (hash(x)=7670050)
123
+ 115 train 9.821703 (lr=4.0559e-05) (hash(x)=5616075)
124
+ 116 train 9.858139 (lr=4.0909e-05) (hash(x)=6593764)
125
+ 117 train 9.880978 (lr=4.1259e-05) (hash(x)=5672608)
126
+ 118 train 9.832921 (lr=4.1608e-05) (hash(x)=6828653)
127
+ 119 train 9.758484 (lr=4.1958e-05) (hash(x)=6839867)
128
+ 120 train 9.791478 (lr=4.2308e-05) (hash(x)=8426992)
129
+ 121 train 9.679622 (lr=4.2657e-05) (hash(x)=4916928)
130
+ 122 train 9.749115 (lr=4.3007e-05) (hash(x)=6484050)
131
+ 123 train 9.749975 (lr=4.3357e-05) (hash(x)=8353379)
132
+ 124 train 9.674648 (lr=4.3706e-05) (hash(x)=6552510)
133
+ 125 train 9.696353 (lr=4.4056e-05) (hash(x)=4838871)
134
+ 126 train 9.578297 (lr=4.4406e-05) (hash(x)=3021697)
135
+ 127 train 9.563817 (lr=4.4755e-05) (hash(x)=6795665)
136
+ 128 train 9.507409 (lr=4.5105e-05) (hash(x)=5486935)
137
+ 129 train 9.648417 (lr=4.5455e-05) (hash(x)=6678038)
138
+ 130 train 9.611389 (lr=4.5804e-05) (hash(x)=6657714)
139
+ 131 train 9.598157 (lr=4.6154e-05) (hash(x)=6112215)
140
+ 132 train 9.533468 (lr=4.6503e-05) (hash(x)=8353143)
141
+ 133 train 9.537433 (lr=4.6853e-05) (hash(x)=5915361)
142
+ 134 train 9.393261 (lr=4.7203e-05) (hash(x)=6254885)
143
+ 135 train 9.485544 (lr=4.7552e-05) (hash(x)=5771588)
144
+ 136 train 9.395518 (lr=4.7902e-05) (hash(x)=5561507)
145
+ 137 train 9.412107 (lr=4.8252e-05) (hash(x)=5688829)
146
+ 138 train 9.315217 (lr=4.8601e-05) (hash(x)=5371951)
147
+ 139 train 9.321527 (lr=4.8951e-05) (hash(x)=5443305)
148
+ 140 train 9.456476 (lr=4.9301e-05) (hash(x)=5016757)
149
+ 141 train 9.383716 (lr=4.9650e-05) (hash(x)=6901933)
150
+ 142 train 9.295671 (lr=5.0000e-05) (hash(x)=6019085)
151
+ 143 train 9.321017 (lr=5.0350e-05) (hash(x)=4294425)
152
+ 144 train 9.304499 (lr=5.0699e-05) (hash(x)=6095229)
153
+ 145 train 9.140566 (lr=5.1049e-05) (hash(x)=4363941)
154
+ 146 train 9.237564 (lr=5.1399e-05) (hash(x)=4693798)
155
+ 147 train 9.357787 (lr=5.1748e-05) (hash(x)=5486335)
156
+ 148 train 9.215295 (lr=5.2098e-05) (hash(x)=7907450)
157
+ 149 train 9.191633 (lr=5.2448e-05) (hash(x)=7295165)
158
+ 150 val loss 9.1792
159
+ 150 val perplexity 9693.3877
160
+ 150 train 9.162370 (lr=5.2797e-05) (hash(x)=6814026)
161
+ 151 train 9.177898 (lr=5.3147e-05) (hash(x)=6267424)
162
+ 152 train 9.175431 (lr=5.3497e-05) (hash(x)=8086437)
163
+ 153 train 9.099119 (lr=5.3846e-05) (hash(x)=5270452)
164
+ 154 train 9.088470 (lr=5.4196e-05) (hash(x)=6420820)
165
+ 155 train 9.030128 (lr=5.4545e-05) (hash(x)=6305297)
166
+ 156 train 9.045231 (lr=5.4895e-05) (hash(x)=7889849)
167
+ 157 train 9.204832 (lr=5.5245e-05) (hash(x)=6040077)
168
+ 158 train 9.203299 (lr=5.5594e-05) (hash(x)=5652497)
169
+ 159 train 8.849765 (lr=5.5944e-05) (hash(x)=7130251)
170
+ 160 train 8.805563 (lr=5.6294e-05) (hash(x)=6610177)
171
+ 161 train 8.829044 (lr=5.6643e-05) (hash(x)=6508350)
172
+ 162 train 8.948931 (lr=5.6993e-05) (hash(x)=6994983)
173
+ 163 train 8.908875 (lr=5.7343e-05) (hash(x)=5859576)
174
+ 164 train 8.555028 (lr=5.7692e-05) (hash(x)=5487065)
175
+ 165 train 8.783297 (lr=5.8042e-05) (hash(x)=4892065)
176
+ 166 train 8.698184 (lr=5.8392e-05) (hash(x)=4933674)
177
+ 167 train 8.813088 (lr=5.8741e-05) (hash(x)=7317289)
178
+ 168 train 8.852915 (lr=5.9091e-05) (hash(x)=6811522)
179
+ 169 train 8.713274 (lr=5.9441e-05) (hash(x)=5510218)
180
+ 170 train 8.792453 (lr=5.9790e-05) (hash(x)=7304235)
181
+ 171 train 8.811553 (lr=6.0140e-05) (hash(x)=6752265)
182
+ 172 train 8.659269 (lr=6.0490e-05) (hash(x)=5994476)
183
+ 173 train 8.641792 (lr=6.0839e-05) (hash(x)=6530867)
184
+ 174 train 8.640290 (lr=6.1189e-05) (hash(x)=5746260)
185
+ 175 train 8.646400 (lr=6.1538e-05) (hash(x)=5790078)
186
+ 176 train 8.609084 (lr=6.1888e-05) (hash(x)=6430135)
187
+ 177 train 8.634834 (lr=6.2238e-05) (hash(x)=7614023)
188
+ 178 train 8.613640 (lr=6.2587e-05) (hash(x)=5244861)
189
+ 179 train 8.609360 (lr=6.2937e-05) (hash(x)=5245146)
190
+ 180 train 8.525265 (lr=6.3287e-05) (hash(x)=6415565)
191
+ 181 train 8.527716 (lr=6.3636e-05) (hash(x)=5786972)
192
+ 182 train 8.548317 (lr=6.3986e-05) (hash(x)=6131703)
193
+ 183 train 8.467394 (lr=6.4336e-05) (hash(x)=6346459)
194
+ 184 train 8.491040 (lr=6.4685e-05) (hash(x)=6504131)
195
+ 185 train 8.525329 (lr=6.5035e-05) (hash(x)=5209234)
196
+ 186 train 8.421936 (lr=6.5385e-05) (hash(x)=6163272)
197
+ 187 train 8.395388 (lr=6.5734e-05) (hash(x)=5676768)
198
+ 188 train 8.507473 (lr=6.6084e-05) (hash(x)=6090561)
199
+ 189 train 8.557103 (lr=6.6434e-05) (hash(x)=5863092)
200
+ 190 train 8.509046 (lr=6.6783e-05) (hash(x)=7201312)
201
+ 191 train 8.467498 (lr=6.7133e-05) (hash(x)=7019489)
202
+ 192 train 8.307016 (lr=6.7483e-05) (hash(x)=6225608)
203
+ 193 train 8.397382 (lr=6.7832e-05) (hash(x)=5483019)
204
+ 194 train 8.358557 (lr=6.8182e-05) (hash(x)=5903258)
205
+ 195 train 8.371078 (lr=6.8531e-05) (hash(x)=5968716)
206
+ 196 train 8.326005 (lr=6.8881e-05) (hash(x)=5778017)
207
+ 197 train 8.196486 (lr=6.9231e-05) (hash(x)=7780194)
208
+ 198 train 8.302087 (lr=6.9580e-05) (hash(x)=6140998)
209
+ 199 train 8.255290 (lr=6.9930e-05) (hash(x)=3784321)
210
+ 200 val loss 8.2862
211
+ 200 val perplexity 3968.7908
212
+ 200 train 8.244012 (lr=7.0280e-05) (hash(x)=5860821)
213
+ 201 train 8.298063 (lr=7.0629e-05) (hash(x)=7774109)
214
+ 202 train 8.390723 (lr=7.0979e-05) (hash(x)=7997664)
215
+ 203 train 8.368900 (lr=7.1329e-05) (hash(x)=6576749)
216
+ 204 train 8.282912 (lr=7.1678e-05) (hash(x)=7266655)
217
+ 205 train 8.283048 (lr=7.2028e-05) (hash(x)=7072404)
218
+ 206 train 8.219134 (lr=7.2378e-05) (hash(x)=6626459)
219
+ 207 train 8.119039 (lr=7.2727e-05) (hash(x)=6556025)
220
+ 208 train 8.193781 (lr=7.3077e-05) (hash(x)=5942129)
221
+ 209 train 8.293331 (lr=7.3427e-05) (hash(x)=7325027)
222
+ 210 train 8.274752 (lr=7.3776e-05) (hash(x)=7322467)
223
+ 211 train 8.206050 (lr=7.4126e-05) (hash(x)=6373412)
224
+ 212 train 8.041122 (lr=7.4476e-05) (hash(x)=7332255)
225
+ 213 train 8.198691 (lr=7.4825e-05) (hash(x)=5262868)
226
+ 214 train 8.034361 (lr=7.5175e-05) (hash(x)=5688247)
227
+ 215 train 8.066957 (lr=7.5524e-05) (hash(x)=8055563)
228
+ 216 train 8.116670 (lr=7.5874e-05) (hash(x)=5352405)
229
+ 217 train 8.169200 (lr=7.6224e-05) (hash(x)=5932790)
230
+ 218 train 8.105790 (lr=7.6573e-05) (hash(x)=6249312)
231
+ 219 train 8.086445 (lr=7.6923e-05) (hash(x)=5680154)
232
+ 220 train 8.156798 (lr=7.7273e-05) (hash(x)=8156280)
233
+ 221 train 8.114058 (lr=7.7622e-05) (hash(x)=5914217)
234
+ 222 train 7.993467 (lr=7.7972e-05) (hash(x)=7120215)
235
+ 223 train 8.108563 (lr=7.8322e-05) (hash(x)=6424128)
236
+ 224 train 7.960476 (lr=7.8671e-05) (hash(x)=4390027)
237
+ 225 train 8.050363 (lr=7.9021e-05) (hash(x)=7804089)
238
+ 226 train 8.179266 (lr=7.9371e-05) (hash(x)=7130267)
239
+ 227 train 8.017311 (lr=7.9720e-05) (hash(x)=6490149)
240
+ 228 train 7.965081 (lr=8.0070e-05) (hash(x)=9276378)
241
+ 229 train 8.044571 (lr=8.0420e-05) (hash(x)=6283540)
242
+ 230 train 8.003944 (lr=8.0769e-05) (hash(x)=5781680)
243
+ 231 train 7.951007 (lr=8.1119e-05) (hash(x)=5728337)
244
+ 232 train 7.956471 (lr=8.1469e-05) (hash(x)=6278202)
245
+ 233 train 7.983962 (lr=8.1818e-05) (hash(x)=7400272)
246
+ 234 train 7.982558 (lr=8.2168e-05) (hash(x)=5795128)
247
+ 235 train 7.965426 (lr=8.2517e-05) (hash(x)=6473432)
248
+ 236 train 7.957043 (lr=8.2867e-05) (hash(x)=5201502)
249
+ 237 train 8.084440 (lr=8.3217e-05) (hash(x)=6922646)
250
+ 238 train 7.886778 (lr=8.3566e-05) (hash(x)=5484641)
251
+ 239 train 7.960029 (lr=8.3916e-05) (hash(x)=7277141)
252
+ 240 train 7.929717 (lr=8.4266e-05) (hash(x)=5604614)
253
+ 241 train 7.973738 (lr=8.4615e-05) (hash(x)=7443124)
254
+ 242 train 7.981045 (lr=8.4965e-05) (hash(x)=5601968)
255
+ 243 train 7.889190 (lr=8.5315e-05) (hash(x)=5632735)
256
+ 244 train 7.819806 (lr=8.5664e-05) (hash(x)=5594810)
257
+ 245 train 7.687109 (lr=8.6014e-05) (hash(x)=5881015)
258
+ 246 train 7.840022 (lr=8.6364e-05) (hash(x)=6865361)
259
+ 247 train 7.944674 (lr=8.6713e-05) (hash(x)=5809411)
260
+ 248 train 7.931084 (lr=8.7063e-05) (hash(x)=6293025)
261
+ 249 train 7.871254 (lr=8.7413e-05) (hash(x)=7232964)
262
+ 250 val loss 7.8811
263
+ 250 val perplexity 2646.7034
264
+ 250 train 7.827769 (lr=8.7762e-05) (hash(x)=5646619)
265
+ 251 train 7.892281 (lr=8.8112e-05) (hash(x)=6171829)
266
+ 252 train 7.937308 (lr=8.8462e-05) (hash(x)=7787552)
267
+ 253 train 7.915157 (lr=8.8811e-05) (hash(x)=5884456)
268
+ 254 train 7.904029 (lr=8.9161e-05) (hash(x)=6752162)
269
+ 255 train 7.845023 (lr=8.9510e-05) (hash(x)=5077031)
270
+ 256 train 7.959042 (lr=8.9860e-05) (hash(x)=6883481)
271
+ 257 train 7.877989 (lr=9.0210e-05) (hash(x)=6248895)
272
+ 258 train 7.752935 (lr=9.0559e-05) (hash(x)=6636615)
273
+ 259 train 7.897634 (lr=9.0909e-05) (hash(x)=6918004)
274
+ 260 train 7.823270 (lr=9.1259e-05) (hash(x)=7302874)
275
+ 261 train 7.914125 (lr=9.1608e-05) (hash(x)=6482466)
276
+ 262 train 7.773370 (lr=9.1958e-05) (hash(x)=7298317)
277
+ 263 train 7.807083 (lr=9.2308e-05) (hash(x)=6588573)
278
+ 264 train 7.809088 (lr=9.2657e-05) (hash(x)=6229630)
279
+ 265 train 7.866171 (lr=9.3007e-05) (hash(x)=6136939)
280
+ 266 train 7.667040 (lr=9.3357e-05) (hash(x)=6394645)
281
+ 267 train 7.846242 (lr=9.3706e-05) (hash(x)=7040157)
282
+ 268 train 7.871849 (lr=9.4056e-05) (hash(x)=7041858)
283
+ 269 train 7.788403 (lr=9.4406e-05) (hash(x)=6455595)
284
+ 270 train 7.722827 (lr=9.4755e-05) (hash(x)=6480443)
285
+ 271 train 7.896258 (lr=9.5105e-05) (hash(x)=4122983)
286
+ 272 train 7.723101 (lr=9.5455e-05) (hash(x)=7672190)
287
+ 273 train 7.770048 (lr=9.5804e-05) (hash(x)=4724052)
288
+ 274 train 7.725244 (lr=9.6154e-05) (hash(x)=6654267)
289
+ 275 train 7.718994 (lr=9.6503e-05) (hash(x)=6820797)
290
+ 276 train 7.774723 (lr=9.6853e-05) (hash(x)=6458662)
291
+ 277 train 8.084826 (lr=9.7203e-05) (hash(x)=7477594)
292
+ 278 train 7.809586 (lr=9.7552e-05) (hash(x)=5876574)
293
+ 279 train 7.569035 (lr=9.7902e-05) (hash(x)=6146305)
294
+ 280 train 7.734031 (lr=9.8252e-05) (hash(x)=6162080)
295
+ 281 train 7.662701 (lr=9.8601e-05) (hash(x)=6814831)
296
+ 282 train 7.795294 (lr=9.8951e-05) (hash(x)=7138295)
297
+ 283 train 7.600586 (lr=9.9301e-05) (hash(x)=5548298)
298
+ 284 train 7.680980 (lr=9.9650e-05) (hash(x)=6265078)
299
+ 285 train 8.010313 (lr=1.0000e-04) (hash(x)=7609416)
300
+ 286 train 7.733868 (lr=1.0000e-04) (hash(x)=6909367)
301
+ 287 train 7.826205 (lr=1.0000e-04) (hash(x)=7574342)
302
+ 288 train 7.667184 (lr=9.9998e-05) (hash(x)=8859748)
303
+ 289 train 7.628573 (lr=9.9996e-05) (hash(x)=5906937)
304
+ 290 train 7.638769 (lr=9.9993e-05) (hash(x)=7243024)
305
+ 291 train 7.722898 (lr=9.9989e-05) (hash(x)=8013471)
306
+ 292 train 7.656594 (lr=9.9984e-05) (hash(x)=5659461)
307
+ 293 train 7.681818 (lr=9.9979e-05) (hash(x)=6146213)
308
+ 294 train 7.733034 (lr=9.9972e-05) (hash(x)=6602917)
309
+ 295 train 7.662315 (lr=9.9965e-05) (hash(x)=6709902)
310
+ 296 train 7.765413 (lr=9.9956e-05) (hash(x)=8154341)
311
+ 297 train 7.892011 (lr=9.9947e-05) (hash(x)=6402571)
312
+ 298 train 7.668901 (lr=9.9937e-05) (hash(x)=5605466)
313
+ 299 train 7.633203 (lr=9.9926e-05) (hash(x)=7683614)
314
+ 300 val loss 7.6736
315
+ 300 val perplexity 2150.8684
316
+ 300 train 7.707984 (lr=9.9915e-05) (hash(x)=5421572)
317
+ 301 train 7.558808 (lr=9.9902e-05) (hash(x)=5453026)
318
+ 302 train 7.632991 (lr=9.9889e-05) (hash(x)=6437746)
319
+ 303 train 7.533190 (lr=9.9874e-05) (hash(x)=6260017)
320
+ 304 train 7.626817 (lr=9.9859e-05) (hash(x)=6119053)
321
+ 305 train 7.784616 (lr=9.9843e-05) (hash(x)=6591466)
322
+ 306 train 7.680906 (lr=9.9826e-05) (hash(x)=5652701)
323
+ 307 train 7.666694 (lr=9.9808e-05) (hash(x)=5501472)
324
+ 308 train 7.852896 (lr=9.9789e-05) (hash(x)=6160695)
325
+ 309 train 7.602336 (lr=9.9770e-05) (hash(x)=5871698)
326
+ 310 train 7.586343 (lr=9.9749e-05) (hash(x)=7301184)
327
+ 311 train 7.543208 (lr=9.9728e-05) (hash(x)=4282052)
328
+ 312 train 7.706687 (lr=9.9706e-05) (hash(x)=5679330)
329
+ 313 train 7.589229 (lr=9.9683e-05) (hash(x)=7824660)
330
+ 314 train 7.606934 (lr=9.9659e-05) (hash(x)=4857318)
331
+ 315 train 7.541104 (lr=9.9634e-05) (hash(x)=5954950)
332
+ 316 train 7.625545 (lr=9.9609e-05) (hash(x)=5758178)
333
+ 317 train 7.653512 (lr=9.9582e-05) (hash(x)=5484656)
334
+ 318 train 7.658661 (lr=9.9555e-05) (hash(x)=6818913)
335
+ 319 train 7.817195 (lr=9.9526e-05) (hash(x)=7031328)
336
+ 320 train 7.692027 (lr=9.9497e-05) (hash(x)=6837621)
337
+ 321 train 7.704087 (lr=9.9467e-05) (hash(x)=9157069)
338
+ 322 train 7.913099 (lr=9.9437e-05) (hash(x)=6509536)
339
+ 323 train 7.637184 (lr=9.9405e-05) (hash(x)=5996883)
340
+ 324 train 7.861008 (lr=9.9372e-05) (hash(x)=6850868)
341
+ 325 train 7.797667 (lr=9.9339e-05) (hash(x)=7645447)
342
+ 326 train 7.820003 (lr=9.9305e-05) (hash(x)=6717515)
343
+ 327 train 7.932030 (lr=9.9270e-05) (hash(x)=6336164)
344
+ 328 train 7.674425 (lr=9.9234e-05) (hash(x)=6897445)
345
+ 329 train 7.584448 (lr=9.9197e-05) (hash(x)=5224240)
346
+ 330 train 7.726440 (lr=9.9159e-05) (hash(x)=6612945)
347
+ 331 train 8.153300 (lr=9.9121e-05) (hash(x)=5447801)
348
+ 332 train 7.717077 (lr=9.9081e-05) (hash(x)=7308579)
349
+ 333 train 7.589918 (lr=9.9041e-05) (hash(x)=4155598)
350
+ 334 train 7.606115 (lr=9.9000e-05) (hash(x)=6106773)
351
+ 335 train 7.658552 (lr=9.8958e-05) (hash(x)=7482744)
352
+ 336 train 7.709564 (lr=9.8915e-05) (hash(x)=6310687)
353
+ 337 train 7.609346 (lr=9.8872e-05) (hash(x)=4733414)
354
+ 338 train 7.736682 (lr=9.8827e-05) (hash(x)=6979734)
355
+ 339 train 7.535175 (lr=9.8782e-05) (hash(x)=7806284)
356
+ 340 train 7.531532 (lr=9.8736e-05) (hash(x)=5080536)
357
+ 341 train 7.747163 (lr=9.8689e-05) (hash(x)=7236478)
358
+ 342 train 7.684106 (lr=9.8641e-05) (hash(x)=6619318)
359
+ 343 train 7.597320 (lr=9.8592e-05) (hash(x)=4656255)
360
+ 344 train 7.550783 (lr=9.8543e-05) (hash(x)=5546888)
361
+ 345 train 7.968453 (lr=9.8492e-05) (hash(x)=9298325)
362
+ 346 train 7.684812 (lr=9.8441e-05) (hash(x)=5994232)
363
+ 347 train 7.663421 (lr=9.8389e-05) (hash(x)=6287524)
364
+ 348 train 7.755795 (lr=9.8336e-05) (hash(x)=6200049)
365
+ 349 train 7.691012 (lr=9.8282e-05) (hash(x)=7895970)
366
+ 350 val loss 7.6151
367
+ 350 val perplexity 2028.6235
368
+ 350 train 7.610491 (lr=9.8228e-05) (hash(x)=6051847)
369
+ 351 train 7.773247 (lr=9.8172e-05) (hash(x)=8847992)
370
+ 352 train 8.034668 (lr=9.8116e-05) (hash(x)=5123883)
371
+ 353 train 7.690849 (lr=9.8059e-05) (hash(x)=6338998)
372
+ 354 train 7.682508 (lr=9.8001e-05) (hash(x)=5765866)
373
+ 355 train 8.039083 (lr=9.7942e-05) (hash(x)=7104011)
374
+ 356 train 7.796448 (lr=9.7882e-05) (hash(x)=7438267)
375
+ 357 train 7.538000 (lr=9.7822e-05) (hash(x)=5573214)
376
+ 358 train 7.818782 (lr=9.7761e-05) (hash(x)=6897283)
377
+ 359 train 7.611669 (lr=9.7699e-05) (hash(x)=6882897)
378
+ 360 train 7.709434 (lr=9.7636e-05) (hash(x)=5343615)
379
+ 361 train 7.625771 (lr=9.7572e-05) (hash(x)=5854242)
380
+ 362 train 7.541291 (lr=9.7507e-05) (hash(x)=6317324)
381
+ 363 train 7.501956 (lr=9.7442e-05) (hash(x)=6972494)
382
+ 364 train 7.567398 (lr=9.7376e-05) (hash(x)=5661167)
383
+ 365 train 7.570722 (lr=9.7309e-05) (hash(x)=6796805)
384
+ 366 train 7.437843 (lr=9.7241e-05) (hash(x)=5817209)
385
+ 367 train 7.487501 (lr=9.7172e-05) (hash(x)=6139120)
386
+ 368 train 7.648262 (lr=9.7103e-05) (hash(x)=6121296)
387
+ 369 train 7.611656 (lr=9.7032e-05) (hash(x)=6486796)
388
+ 370 train 7.553037 (lr=9.6961e-05) (hash(x)=5514645)
389
+ 371 train 7.488541 (lr=9.6889e-05) (hash(x)=5578797)
390
+ 372 train 7.488960 (lr=9.6817e-05) (hash(x)=7651738)
391
+ 373 train 7.697730 (lr=9.6743e-05) (hash(x)=6895326)
392
+ 374 train 7.676789 (lr=9.6669e-05) (hash(x)=5490256)
393
+ 375 train 7.574334 (lr=9.6593e-05) (hash(x)=7783160)
394
+ 376 train 7.756738 (lr=9.6518e-05) (hash(x)=5608424)
395
+ 377 train 7.547126 (lr=9.6441e-05) (hash(x)=4867625)
396
+ 378 train 7.432821 (lr=9.6363e-05) (hash(x)=4655472)
397
+ 379 train 7.514024 (lr=9.6285e-05) (hash(x)=6631123)
398
+ 380 train 7.177055 (lr=9.6206e-05) (hash(x)=5751990)
399
+ 381 train 7.479978 (lr=9.6126e-05) (hash(x)=6710965)
400
+ 382 train 7.456696 (lr=9.6045e-05) (hash(x)=5293800)
401
+ 383 train 7.486141 (lr=9.5963e-05) (hash(x)=6294657)
402
+ 384 train 7.627875 (lr=9.5881e-05) (hash(x)=6999538)
403
+ 385 train 7.616038 (lr=9.5798e-05) (hash(x)=5443606)
404
+ 386 train 7.590320 (lr=9.5714e-05) (hash(x)=5650401)
405
+ 387 train 7.598116 (lr=9.5629e-05) (hash(x)=7035770)
406
+ 388 train 7.545592 (lr=9.5544e-05) (hash(x)=6678271)
407
+ 389 train 7.736907 (lr=9.5457e-05) (hash(x)=4899541)
408
+ 390 train 7.498661 (lr=9.5370e-05) (hash(x)=5359260)
409
+ 391 train 7.503405 (lr=9.5282e-05) (hash(x)=6227120)
410
+ 392 train 7.582795 (lr=9.5194e-05) (hash(x)=6643916)
411
+ 393 train 7.624684 (lr=9.5104e-05) (hash(x)=5981648)
412
+ 394 train 7.564916 (lr=9.5014e-05) (hash(x)=6467540)
413
+ 395 train 7.528152 (lr=9.4923e-05) (hash(x)=5558993)
414
+ 396 train 7.674169 (lr=9.4831e-05) (hash(x)=6698959)
415
+ 397 train 7.511356 (lr=9.4739e-05) (hash(x)=7015671)
416
+ 398 train 7.507686 (lr=9.4646e-05) (hash(x)=6800890)
417
+ 399 train 7.665562 (lr=9.4551e-05) (hash(x)=5931234)
418
+ 400 val loss 7.5845
419
+ 400 val perplexity 1967.3995
420
+ 400 train 7.562352 (lr=9.4457e-05) (hash(x)=3995070)
421
+ 401 train 7.514524 (lr=9.4361e-05) (hash(x)=4582159)
422
+ 402 train 7.602766 (lr=9.4265e-05) (hash(x)=7081906)
423
+ 403 train 7.553264 (lr=9.4168e-05) (hash(x)=5029472)
424
+ 404 train 7.544349 (lr=9.4070e-05) (hash(x)=3946232)
425
+ 405 train 7.455920 (lr=9.3971e-05) (hash(x)=5434102)
426
+ 406 train 8.172092 (lr=9.3872e-05) (hash(x)=5340646)
427
+ 407 train 7.349815 (lr=9.3772e-05) (hash(x)=6810638)
428
+ 408 train 7.578692 (lr=9.3671e-05) (hash(x)=6768026)
429
+ 409 train 7.594790 (lr=9.3569e-05) (hash(x)=6521562)
430
+ 410 train 7.593122 (lr=9.3467e-05) (hash(x)=4429434)
431
+ 411 train 7.643532 (lr=9.3364e-05) (hash(x)=7226932)
432
+ 412 train 7.429601 (lr=9.3260e-05) (hash(x)=5065932)
433
+ 413 train 7.576879 (lr=9.3155e-05) (hash(x)=6957849)
434
+ 414 train 7.572788 (lr=9.3050e-05) (hash(x)=6431457)
435
+ 415 train 7.629786 (lr=9.2944e-05) (hash(x)=5334281)
436
+ 416 train 7.642087 (lr=9.2837e-05) (hash(x)=5471747)
437
+ 417 train 7.736324 (lr=9.2729e-05) (hash(x)=5887983)
438
+ 418 train 7.657819 (lr=9.2621e-05) (hash(x)=5075909)
439
+ 419 train 7.445121 (lr=9.2512e-05) (hash(x)=6393576)
440
+ 420 train 7.569709 (lr=9.2402e-05) (hash(x)=6608979)
441
+ 421 train 7.941503 (lr=9.2292e-05) (hash(x)=5057959)
442
+ 422 train 7.932811 (lr=9.2181e-05) (hash(x)=6414858)
443
+ 423 train 7.684020 (lr=9.2069e-05) (hash(x)=8595271)
444
+ 424 train 7.558131 (lr=9.1956e-05) (hash(x)=6500949)
445
+ 425 train 7.717246 (lr=9.1843e-05) (hash(x)=8395167)
446
+ 426 train 7.844435 (lr=9.1729e-05) (hash(x)=6288420)
447
+ 427 train 7.643422 (lr=9.1614e-05) (hash(x)=7634417)
448
+ 428 train 7.544155 (lr=9.1499e-05) (hash(x)=6795561)
449
+ 429 train 7.443240 (lr=9.1382e-05) (hash(x)=5257771)
450
+ 430 train 7.587779 (lr=9.1266e-05) (hash(x)=7506860)
451
+ 431 train 7.641050 (lr=9.1148e-05) (hash(x)=7674238)
452
+ 432 train 7.774417 (lr=9.1030e-05) (hash(x)=5353794)
453
+ 433 train 7.685770 (lr=9.0911e-05) (hash(x)=5300555)
454
+ 434 train 7.367756 (lr=9.0791e-05) (hash(x)=5564419)
455
+ 435 train 7.636134 (lr=9.0671e-05) (hash(x)=6066139)
456
+ 436 train 7.592438 (lr=9.0550e-05) (hash(x)=6863709)
457
+ 437 train 7.601659 (lr=9.0428e-05) (hash(x)=6688212)
458
+ 438 train 7.763230 (lr=9.0306e-05) (hash(x)=7095090)
459
+ 439 train 7.821415 (lr=9.0182e-05) (hash(x)=6642404)
460
+ 440 train 7.560585 (lr=9.0059e-05) (hash(x)=6539227)
461
+ 441 train 7.558822 (lr=8.9934e-05) (hash(x)=4961570)
462
+ 442 train 7.493913 (lr=8.9809e-05) (hash(x)=6606391)
463
+ 443 train 7.303205 (lr=8.9683e-05) (hash(x)=5105190)
464
+ 444 train 7.872974 (lr=8.9557e-05) (hash(x)=5695997)
465
+ 445 train 7.579685 (lr=8.9430e-05) (hash(x)=7155991)
466
+ 446 train 7.568652 (lr=8.9302e-05) (hash(x)=7249727)
467
+ 447 train 7.553354 (lr=8.9173e-05) (hash(x)=5132972)
468
+ 448 train 7.542585 (lr=8.9044e-05) (hash(x)=5381802)
469
+ 449 train 7.573223 (lr=8.8914e-05) (hash(x)=6269946)
470
+ 450 val loss 7.5568
471
+ 450 val perplexity 1913.6842
472
+ 450 train 7.545667 (lr=8.8784e-05) (hash(x)=4976762)
473
+ 451 train 7.622306 (lr=8.8653e-05) (hash(x)=5396133)
474
+ 452 train 7.571563 (lr=8.8521e-05) (hash(x)=6173473)
475
+ 453 train 7.566725 (lr=8.8388e-05) (hash(x)=6394083)
476
+ 454 train 7.604074 (lr=8.8255e-05) (hash(x)=6436030)
477
+ 455 train 7.484024 (lr=8.8122e-05) (hash(x)=5237804)
478
+ 456 train 7.447423 (lr=8.7987e-05) (hash(x)=6540397)
479
+ 457 train 7.698504 (lr=8.7852e-05) (hash(x)=5216638)
480
+ 458 train 7.533554 (lr=8.7717e-05) (hash(x)=5750537)
481
+ 459 train 7.520440 (lr=8.7580e-05) (hash(x)=6735565)
482
+ 460 train 7.527417 (lr=8.7444e-05) (hash(x)=7297488)
483
+ 461 train 7.759251 (lr=8.7306e-05) (hash(x)=5983575)
484
+ 462 train 7.546457 (lr=8.7168e-05) (hash(x)=4808543)
485
+ 463 train 7.653461 (lr=8.7029e-05) (hash(x)=5831619)
486
+ 464 train 7.422575 (lr=8.6890e-05) (hash(x)=6280052)
487
+ 465 train 7.443268 (lr=8.6750e-05) (hash(x)=9109083)
488
+ 466 train 7.693081 (lr=8.6609e-05) (hash(x)=5848604)
489
+ 467 train 7.615723 (lr=8.6468e-05) (hash(x)=6912967)
490
+ 468 train 7.786067 (lr=8.6326e-05) (hash(x)=6615365)
491
+ 469 train 7.848358 (lr=8.6184e-05) (hash(x)=5965674)
492
+ 470 train 7.695670 (lr=8.6041e-05) (hash(x)=6879965)
493
+ 471 train 7.670931 (lr=8.5897e-05) (hash(x)=5822533)
494
+ 472 train 7.536402 (lr=8.5753e-05) (hash(x)=5794462)
495
+ 473 train 7.617669 (lr=8.5608e-05) (hash(x)=5268216)
496
+ 474 train 7.425871 (lr=8.5462e-05) (hash(x)=7058396)
497
+ 475 train 7.552730 (lr=8.5316e-05) (hash(x)=6531900)
498
+ 476 train 7.561369 (lr=8.5170e-05) (hash(x)=5878912)
499
+ 477 train 7.371514 (lr=8.5022e-05) (hash(x)=5426499)
500
+ 478 train 7.595307 (lr=8.4875e-05) (hash(x)=8107004)
501
+ 479 train 8.290646 (lr=8.4726e-05) (hash(x)=7455372)
502
+ 480 train 8.245323 (lr=8.4577e-05) (hash(x)=7727959)
503
+ 481 train 8.214881 (lr=8.4428e-05) (hash(x)=8319119)
504
+ 482 train 8.178007 (lr=8.4278e-05) (hash(x)=7581656)
505
+ 483 train 8.239472 (lr=8.4127e-05) (hash(x)=7592128)
506
+ 484 train 8.249104 (lr=8.3976e-05) (hash(x)=10506633)
507
+ 485 train 8.222606 (lr=8.3824e-05) (hash(x)=8704170)
508
+ 486 train 7.901765 (lr=8.3672e-05) (hash(x)=7461138)
509
+ 487 train 7.674077 (lr=8.3519e-05) (hash(x)=7165037)
510
+ 488 train 7.844088 (lr=8.3366e-05) (hash(x)=4976998)
511
+ 489 train 7.524818 (lr=8.3212e-05) (hash(x)=7678000)
512
+ 490 train 7.573645 (lr=8.3057e-05) (hash(x)=6299409)
513
+ 491 train 7.514767 (lr=8.2902e-05) (hash(x)=7414788)
514
+ 492 train 7.604311 (lr=8.2746e-05) (hash(x)=5063061)
515
+ 493 train 7.561200 (lr=8.2590e-05) (hash(x)=7326216)
516
+ 494 train 7.687844 (lr=8.2434e-05) (hash(x)=7492012)
517
+ 495 train 7.625070 (lr=8.2276e-05) (hash(x)=5857162)
518
+ 496 train 7.432327 (lr=8.2119e-05) (hash(x)=6169001)
519
+ 497 train 9.275414 (lr=8.1960e-05) (hash(x)=14715669)
520
+ 498 train 12.021711 (lr=8.1801e-05) (hash(x)=18256706)
521
+ 499 train 11.997477 (lr=8.1642e-05) (hash(x)=16950560)
522
+ 500 val loss 7.5712
523
+ 500 val perplexity 1941.3759
524
+ 500 train 8.319768 (lr=8.1482e-05) (hash(x)=6113563)
525
+ 501 train 7.188338 (lr=8.1322e-05) (hash(x)=6159125)
526
+ 502 train 7.592231 (lr=8.1161e-05) (hash(x)=5482303)
527
+ 503 train 7.627305 (lr=8.1000e-05) (hash(x)=5908442)
528
+ 504 train 7.497091 (lr=8.0838e-05) (hash(x)=4985045)
529
+ 505 train 7.429093 (lr=8.0676e-05) (hash(x)=6668887)
530
+ 506 train 7.526974 (lr=8.0513e-05) (hash(x)=6123910)
531
+ 507 train 7.629738 (lr=8.0349e-05) (hash(x)=7662975)
532
+ 508 train 7.570232 (lr=8.0186e-05) (hash(x)=7191898)
533
+ 509 train 7.508491 (lr=8.0021e-05) (hash(x)=6669539)
534
+ 510 train 7.580479 (lr=7.9856e-05) (hash(x)=6557481)
535
+ 511 train 7.481220 (lr=7.9691e-05) (hash(x)=5015998)
536
+ 512 train 7.529236 (lr=7.9525e-05) (hash(x)=4964540)
537
+ 513 train 7.387224 (lr=7.9359e-05) (hash(x)=5458386)
538
+ 514 train 7.716806 (lr=7.9192e-05) (hash(x)=6077616)
539
+ 515 train 7.508154 (lr=7.9025e-05) (hash(x)=4849847)
540
+ 516 train 7.541708 (lr=7.8858e-05) (hash(x)=5431422)
541
+ 517 train 7.724491 (lr=7.8689e-05) (hash(x)=5125562)
542
+ 518 train 7.633122 (lr=7.8521e-05) (hash(x)=5587715)
543
+ 519 train 7.692536 (lr=7.8352e-05) (hash(x)=6012047)
544
+ 520 train 7.546825 (lr=7.8182e-05) (hash(x)=5668436)
545
+ 521 train 7.569431 (lr=7.8012e-05) (hash(x)=6079364)
546
+ 522 train 7.689155 (lr=7.7842e-05) (hash(x)=5200504)
547
+ 523 train 7.429648 (lr=7.7671e-05) (hash(x)=6030395)
548
+ 524 train 7.664275 (lr=7.7500e-05) (hash(x)=6855319)
549
+ 525 train 7.448567 (lr=7.7328e-05) (hash(x)=6781762)
550
+ 526 train 7.593457 (lr=7.7156e-05) (hash(x)=6717939)
551
+ 527 train 7.601754 (lr=7.6984e-05) (hash(x)=5238325)
552
+ 528 train 7.491256 (lr=7.6811e-05) (hash(x)=5808069)
553
+ 529 train 7.807417 (lr=7.6637e-05) (hash(x)=8611144)
554
+ 530 train 7.504074 (lr=7.6463e-05) (hash(x)=6028738)
555
+ 531 train 7.535662 (lr=7.6289e-05) (hash(x)=6901605)
556
+ 532 train 7.484989 (lr=7.6115e-05) (hash(x)=6723307)
557
+ 533 train 7.589094 (lr=7.5940e-05) (hash(x)=5499337)
558
+ 534 train 7.492435 (lr=7.5764e-05) (hash(x)=6586372)
559
+ 535 train 7.625252 (lr=7.5588e-05) (hash(x)=7134874)
560
+ 536 train 7.526498 (lr=7.5412e-05) (hash(x)=8415623)
561
+ 537 train 7.556602 (lr=7.5235e-05) (hash(x)=6149038)
562
+ 538 train 7.597550 (lr=7.5058e-05) (hash(x)=6078325)
563
+ 539 train 7.561991 (lr=7.4881e-05) (hash(x)=6609760)
564
+ 540 train 7.465807 (lr=7.4703e-05) (hash(x)=6185253)
565
+ 541 train 7.896461 (lr=7.4525e-05) (hash(x)=6652627)
566
+ 542 train 7.633656 (lr=7.4346e-05) (hash(x)=7758536)
567
+ 543 train 7.612063 (lr=7.4167e-05) (hash(x)=8618579)
568
+ 544 train 7.588217 (lr=7.3988e-05) (hash(x)=6710226)
569
+ 545 train 7.434253 (lr=7.3808e-05) (hash(x)=5896892)
570
+ 546 train 7.443069 (lr=7.3628e-05) (hash(x)=6332995)
571
+ 547 train 7.860254 (lr=7.3448e-05) (hash(x)=8945854)
572
+ 548 train 7.432050 (lr=7.3267e-05) (hash(x)=4329129)
573
+ 549 train 7.493251 (lr=7.3086e-05) (hash(x)=6775103)
574
+ 550 val loss 7.5143
575
+ 550 val perplexity 1834.1144
576
+ 550 train 7.607581 (lr=7.2904e-05) (hash(x)=4465735)
577
+ 551 train 7.468003 (lr=7.2723e-05) (hash(x)=6200024)
578
+ 552 train 7.439372 (lr=7.2540e-05) (hash(x)=5520748)
579
+ 553 train 7.416797 (lr=7.2358e-05) (hash(x)=5238139)
580
+ 554 train 7.693781 (lr=7.2175e-05) (hash(x)=7639817)
581
+ 555 train 7.454398 (lr=7.1992e-05) (hash(x)=5780948)
582
+ 556 train 7.585551 (lr=7.1808e-05) (hash(x)=6733444)
583
+ 557 train 7.537384 (lr=7.1624e-05) (hash(x)=6168407)
584
+ 558 train 7.562554 (lr=7.1440e-05) (hash(x)=6680197)
585
+ 559 train 7.462463 (lr=7.1256e-05) (hash(x)=3775881)
586
+ 560 train 7.290215 (lr=7.1071e-05) (hash(x)=5784096)
587
+ 561 train 7.762159 (lr=7.0886e-05) (hash(x)=6848709)
588
+ 562 train 7.847752 (lr=7.0701e-05) (hash(x)=7449019)
589
+ 563 train 7.931695 (lr=7.0515e-05) (hash(x)=6478113)
590
+ 564 train 7.535942 (lr=7.0329e-05) (hash(x)=5925744)
591
+ 565 train 7.570724 (lr=7.0143e-05) (hash(x)=6483914)
592
+ 566 train 7.464733 (lr=6.9956e-05) (hash(x)=6010847)
593
+ 567 train 7.557147 (lr=6.9769e-05) (hash(x)=5927795)
594
+ 568 train 7.569900 (lr=6.9582e-05) (hash(x)=5896332)
595
+ 569 train 7.515122 (lr=6.9394e-05) (hash(x)=8438033)
596
+ 570 train 7.634714 (lr=6.9207e-05) (hash(x)=6081660)
597
+ 571 train 7.540883 (lr=6.9019e-05) (hash(x)=5684788)
598
+ 572 train 7.417416 (lr=6.8830e-05) (hash(x)=6272055)
599
+ 573 train 7.380927 (lr=6.8642e-05) (hash(x)=5710182)
600
+ 574 train 7.500779 (lr=6.8453e-05) (hash(x)=6219616)
601
+ 575 train 7.473444 (lr=6.8264e-05) (hash(x)=6160063)
602
+ 576 train 7.498240 (lr=6.8075e-05) (hash(x)=5710809)
603
+ 577 train 7.250515 (lr=6.7885e-05) (hash(x)=5955100)
604
+ 578 train 7.258733 (lr=6.7695e-05) (hash(x)=5356088)
605
+ 579 train 7.312251 (lr=6.7505e-05) (hash(x)=6672309)
606
+ 580 train 7.530699 (lr=6.7315e-05) (hash(x)=5841684)
607
+ 581 train 7.877331 (lr=6.7124e-05) (hash(x)=9001564)
608
+ 582 train 8.499863 (lr=6.6933e-05) (hash(x)=6897178)
609
+ 583 train 7.613977 (lr=6.6742e-05) (hash(x)=5454654)
610
+ 584 train 7.360872 (lr=6.6551e-05) (hash(x)=6391454)
611
+ 585 train 7.444867 (lr=6.6360e-05) (hash(x)=5174062)
612
+ 586 train 7.548015 (lr=6.6168e-05) (hash(x)=5197924)
613
+ 587 train 7.520078 (lr=6.5976e-05) (hash(x)=6199203)
614
+ 588 train 7.358311 (lr=6.5784e-05) (hash(x)=5786498)
615
+ 589 train 7.511392 (lr=6.5592e-05) (hash(x)=8352747)
616
+ 590 train 7.833665 (lr=6.5399e-05) (hash(x)=4829700)
617
+ 591 train 7.287157 (lr=6.5206e-05) (hash(x)=4547106)
618
+ 592 train 7.622293 (lr=6.5013e-05) (hash(x)=7686831)
619
+ 593 train 7.474770 (lr=6.4820e-05) (hash(x)=6433099)
620
+ 594 train 7.336624 (lr=6.4627e-05) (hash(x)=4826460)
621
+ 595 train 7.424981 (lr=6.4433e-05) (hash(x)=6424428)
622
+ 596 train 7.311295 (lr=6.4240e-05) (hash(x)=6073605)
623
+ 597 train 7.638861 (lr=6.4046e-05) (hash(x)=5975712)
624
+ 598 train 7.268995 (lr=6.3852e-05) (hash(x)=7030029)
625
+ 599 train 7.593262 (lr=6.3658e-05) (hash(x)=5458532)
626
+ 600 val loss 7.4961
627
+ 600 val perplexity 1800.9910
628
+ 600 train 7.420565 (lr=6.3463e-05) (hash(x)=7823919)
629
+ 601 train 7.835981 (lr=6.3269e-05) (hash(x)=7002774)
630
+ 602 train 7.672714 (lr=6.3074e-05) (hash(x)=7056399)
631
+ 603 train 7.867592 (lr=6.2879e-05) (hash(x)=8206173)
632
+ 604 train 7.711513 (lr=6.2684e-05) (hash(x)=8034846)
633
+ 605 train 8.019062 (lr=6.2489e-05) (hash(x)=7784971)
634
+ 606 train 8.257688 (lr=6.2294e-05) (hash(x)=7118076)
635
+ 607 train 7.430291 (lr=6.2098e-05) (hash(x)=6001843)
636
+ 608 train 7.511605 (lr=6.1903e-05) (hash(x)=5582437)
637
+ 609 train 7.628926 (lr=6.1707e-05) (hash(x)=5698642)
638
+ 610 train 7.480402 (lr=6.1511e-05) (hash(x)=6615301)
639
+ 611 train 7.507148 (lr=6.1315e-05) (hash(x)=6600751)
640
+ 612 train 7.410814 (lr=6.1119e-05) (hash(x)=6880955)
641
+ 613 train 7.475226 (lr=6.0923e-05) (hash(x)=5804610)
642
+ 614 train 7.412912 (lr=6.0726e-05) (hash(x)=6263347)
643
+ 615 train 7.371877 (lr=6.0530e-05) (hash(x)=5162652)
644
+ 616 train 7.411459 (lr=6.0333e-05) (hash(x)=6055710)
645
+ 617 train 7.371758 (lr=6.0137e-05) (hash(x)=5297006)
646
+ 618 train 7.380520 (lr=5.9940e-05) (hash(x)=5923648)
647
+ 619 train 7.402765 (lr=5.9743e-05) (hash(x)=6728323)
648
+ 620 train 7.421280 (lr=5.9546e-05) (hash(x)=6660239)
649
+ 621 train 7.393060 (lr=5.9349e-05) (hash(x)=3964882)
650
+ 622 train 7.440411 (lr=5.9152e-05) (hash(x)=5868942)
651
+ 623 train 7.358283 (lr=5.8955e-05) (hash(x)=6735993)
652
+ 624 train 7.682373 (lr=5.8758e-05) (hash(x)=5926190)
653
+ 625 train 7.365679 (lr=5.8560e-05) (hash(x)=5150316)
654
+ 626 train 7.399684 (lr=5.8363e-05) (hash(x)=7571467)
655
+ 627 train 7.320965 (lr=5.8165e-05) (hash(x)=5220448)
656
+ 628 train 7.296511 (lr=5.7968e-05) (hash(x)=6790397)
657
+ 629 train 7.786151 (lr=5.7770e-05) (hash(x)=6971289)
658
+ 630 train 7.730562 (lr=5.7573e-05) (hash(x)=7123359)
659
+ 631 train 7.529173 (lr=5.7375e-05) (hash(x)=5012069)
660
+ 632 train 7.487381 (lr=5.7177e-05) (hash(x)=6183574)
661
+ 633 train 7.387636 (lr=5.6979e-05) (hash(x)=6576136)
662
+ 634 train 7.407414 (lr=5.6782e-05) (hash(x)=6157346)
663
+ 635 train 7.401323 (lr=5.6584e-05) (hash(x)=6077647)
664
+ 636 train 7.640889 (lr=5.6386e-05) (hash(x)=8369488)
665
+ 637 train 7.455949 (lr=5.6188e-05) (hash(x)=6032381)
666
+ 638 train 7.349550 (lr=5.5990e-05) (hash(x)=5924756)
667
+ 639 train 7.404820 (lr=5.5792e-05) (hash(x)=8410716)
668
+ 640 train 7.374012 (lr=5.5594e-05) (hash(x)=5803440)
669
+ 641 train 7.473207 (lr=5.5396e-05) (hash(x)=6212516)
670
+ 642 train 7.536135 (lr=5.5198e-05) (hash(x)=7165945)
671
+ 643 train 7.261624 (lr=5.5000e-05) (hash(x)=5298545)
672
+ 644 train 7.520327 (lr=5.4802e-05) (hash(x)=6281739)
673
+ 645 train 7.464900 (lr=5.4604e-05) (hash(x)=5779073)
674
+ 646 train 7.359646 (lr=5.4406e-05) (hash(x)=5090107)
675
+ 647 train 7.418220 (lr=5.4208e-05) (hash(x)=6340947)
676
+ 648 train 7.161737 (lr=5.4010e-05) (hash(x)=4741418)
677
+ 649 train 7.094602 (lr=5.3812e-05) (hash(x)=6530441)
678
+ 650 val loss 7.4723
679
+ 650 val perplexity 1758.5824
680
+ 650 train 7.642398 (lr=5.3614e-05) (hash(x)=6396041)
681
+ 651 train 7.316564 (lr=5.3416e-05) (hash(x)=7818007)
682
+ 652 train 7.312077 (lr=5.3218e-05) (hash(x)=6478457)
683
+ 653 train 7.380846 (lr=5.3021e-05) (hash(x)=6277435)
684
+ 654 train 7.342876 (lr=5.2823e-05) (hash(x)=6042369)
685
+ 655 train 7.375009 (lr=5.2625e-05) (hash(x)=6226320)
686
+ 656 train 7.380784 (lr=5.2427e-05) (hash(x)=5826883)
687
+ 657 train 7.477278 (lr=5.2230e-05) (hash(x)=5463751)
688
+ 658 train 7.424911 (lr=5.2032e-05) (hash(x)=6433080)
689
+ 659 train 7.261593 (lr=5.1835e-05) (hash(x)=6579329)
690
+ 660 train 7.300441 (lr=5.1637e-05) (hash(x)=4724703)
691
+ 661 train 7.632395 (lr=5.1440e-05) (hash(x)=6660540)
692
+ 662 train 7.433490 (lr=5.1242e-05) (hash(x)=5918756)
693
+ 663 train 7.379390 (lr=5.1045e-05) (hash(x)=6491272)
694
+ 664 train 7.319147 (lr=5.0848e-05) (hash(x)=5452863)
695
+ 665 train 7.490001 (lr=5.0651e-05) (hash(x)=6766947)
696
+ 666 train 7.844019 (lr=5.0454e-05) (hash(x)=7459504)
697
+ 667 train 7.524452 (lr=5.0257e-05) (hash(x)=3721218)
698
+ 668 train 7.676506 (lr=5.0060e-05) (hash(x)=7929571)
699
+ 669 train 7.381314 (lr=4.9863e-05) (hash(x)=5229898)
700
+ 670 train 7.408013 (lr=4.9667e-05) (hash(x)=6083090)
701
+ 671 train 7.262152 (lr=4.9470e-05) (hash(x)=6054428)
702
+ 672 train 7.431007 (lr=4.9274e-05) (hash(x)=6072563)
703
+ 673 train 7.533266 (lr=4.9077e-05) (hash(x)=6290464)
704
+ 674 train 7.625160 (lr=4.8881e-05) (hash(x)=6314742)
705
+ 675 train 7.297334 (lr=4.8685e-05) (hash(x)=6904498)
706
+ 676 train 7.428061 (lr=4.8489e-05) (hash(x)=5378107)
707
+ 677 train 7.555771 (lr=4.8293e-05) (hash(x)=7765805)
708
+ 678 train 7.383022 (lr=4.8097e-05) (hash(x)=5871280)
709
+ 679 train 7.532508 (lr=4.7902e-05) (hash(x)=6543764)
710
+ 680 train 7.369411 (lr=4.7706e-05) (hash(x)=7369665)
711
+ 681 train 7.485236 (lr=4.7511e-05) (hash(x)=7204955)
712
+ 682 train 7.444676 (lr=4.7316e-05) (hash(x)=5856543)
713
+ 683 train 7.315861 (lr=4.7121e-05) (hash(x)=4423711)
714
+ 684 train 7.284998 (lr=4.6926e-05) (hash(x)=5818153)
715
+ 685 train 7.374858 (lr=4.6731e-05) (hash(x)=5343477)
716
+ 686 train 7.569736 (lr=4.6537e-05) (hash(x)=6581702)
717
+ 687 train 7.407937 (lr=4.6342e-05) (hash(x)=6595856)
718
+ 688 train 7.348412 (lr=4.6148e-05) (hash(x)=5325378)
719
+ 689 train 7.258319 (lr=4.5954e-05) (hash(x)=7229015)
720
+ 690 train 7.440554 (lr=4.5760e-05) (hash(x)=5918186)
721
+ 691 train 7.287374 (lr=4.5567e-05) (hash(x)=5916034)
722
+ 692 train 7.325947 (lr=4.5373e-05) (hash(x)=5280722)
723
+ 693 train 7.734975 (lr=4.5180e-05) (hash(x)=6939521)
724
+ 694 train 7.454517 (lr=4.4987e-05) (hash(x)=5172566)
725
+ 695 train 7.316304 (lr=4.4794e-05) (hash(x)=5492185)
726
+ 696 train 7.174698 (lr=4.4601e-05) (hash(x)=5779011)
727
+ 697 train 7.455735 (lr=4.4408e-05) (hash(x)=6898006)
728
+ 698 train 7.275985 (lr=4.4216e-05) (hash(x)=5779217)
729
+ 699 train 7.138043 (lr=4.4024e-05) (hash(x)=5462008)
730
+ 700 val loss 7.4483
731
+ 700 val perplexity 1716.8975
732
+ 700 train 7.277759 (lr=4.3832e-05) (hash(x)=6243843)
733
+ 701 train 7.363749 (lr=4.3640e-05) (hash(x)=7821902)
734
+ 702 train 7.417922 (lr=4.3449e-05) (hash(x)=5222501)
735
+ 703 train 7.373488 (lr=4.3258e-05) (hash(x)=6608108)
736
+ 704 train 7.513809 (lr=4.3067e-05) (hash(x)=7602879)
737
+ 705 train 7.411582 (lr=4.2876e-05) (hash(x)=6295394)
738
+ 706 train 7.377085 (lr=4.2685e-05) (hash(x)=6579907)
739
+ 707 train 7.293796 (lr=4.2495e-05) (hash(x)=6803449)
740
+ 708 train 7.338017 (lr=4.2305e-05) (hash(x)=5413733)
741
+ 709 train 7.295708 (lr=4.2115e-05) (hash(x)=7444134)
742
+ 710 train 7.448079 (lr=4.1925e-05) (hash(x)=7028652)
743
+ 711 train 7.264015 (lr=4.1736e-05) (hash(x)=5361923)
744
+ 712 train 7.357029 (lr=4.1547e-05) (hash(x)=5478293)
745
+ 713 train 7.337646 (lr=4.1358e-05) (hash(x)=7803039)
746
+ 714 train 7.548423 (lr=4.1170e-05) (hash(x)=6469017)
747
+ 715 train 7.433737 (lr=4.0981e-05) (hash(x)=7042578)
748
+ 716 train 7.446400 (lr=4.0793e-05) (hash(x)=6927957)
749
+ 717 train 7.295939 (lr=4.0606e-05) (hash(x)=6647007)
750
+ 718 train 7.269823 (lr=4.0418e-05) (hash(x)=6230888)
751
+ 719 train 7.533604 (lr=4.0231e-05) (hash(x)=6303658)
752
+ 720 train 7.577978 (lr=4.0044e-05) (hash(x)=6769880)
753
+ 721 train 7.423652 (lr=3.9857e-05) (hash(x)=6158927)
754
+ 722 train 7.369067 (lr=3.9671e-05) (hash(x)=5991631)
755
+ 723 train 7.255892 (lr=3.9485e-05) (hash(x)=6419656)
756
+ 724 train 7.239484 (lr=3.9299e-05) (hash(x)=7739209)
757
+ 725 train 7.221469 (lr=3.9114e-05) (hash(x)=7891271)
758
+ 726 train 7.576676 (lr=3.8929e-05) (hash(x)=7171433)
759
+ 727 train 7.333326 (lr=3.8744e-05) (hash(x)=7099888)
760
+ 728 train 7.345532 (lr=3.8560e-05) (hash(x)=6103730)
761
+ 729 train 7.457294 (lr=3.8376e-05) (hash(x)=6352424)
762
+ 730 train 7.363271 (lr=3.8192e-05) (hash(x)=4893671)
763
+ 731 train 7.308297 (lr=3.8008e-05) (hash(x)=5863145)
764
+ 732 train 7.377453 (lr=3.7825e-05) (hash(x)=5502233)
765
+ 733 train 7.318246 (lr=3.7642e-05) (hash(x)=6844597)
766
+ 734 train 7.341473 (lr=3.7460e-05) (hash(x)=6744526)
767
+ 735 train 7.494901 (lr=3.7277e-05) (hash(x)=6009861)
768
+ 736 train 7.364418 (lr=3.7096e-05) (hash(x)=6152358)
769
+ 737 train 7.263221 (lr=3.6914e-05) (hash(x)=5374957)
770
+ 738 train 7.300702 (lr=3.6733e-05) (hash(x)=6495309)
771
+ 739 train 7.431802 (lr=3.6552e-05) (hash(x)=3923225)
772
+ 740 train 7.242917 (lr=3.6372e-05) (hash(x)=6255243)
773
+ 741 train 7.374507 (lr=3.6192e-05) (hash(x)=5848862)
774
+ 742 train 7.386904 (lr=3.6012e-05) (hash(x)=4535000)
775
+ 743 train 7.259915 (lr=3.5833e-05) (hash(x)=4440634)
776
+ 744 train 7.274788 (lr=3.5654e-05) (hash(x)=5222077)
777
+ 745 train 7.297772 (lr=3.5475e-05) (hash(x)=4788998)
778
+ 746 train 7.295279 (lr=3.5297e-05) (hash(x)=6108350)
779
+ 747 train 7.223709 (lr=3.5119e-05) (hash(x)=7378165)
780
+ 748 train 7.308453 (lr=3.4942e-05) (hash(x)=5475646)
781
+ 749 train 7.316569 (lr=3.4765e-05) (hash(x)=4771243)
782
+ 750 val loss 7.4290
783
+ 750 val perplexity 1684.0638
784
+ 750 train 7.209068 (lr=3.4588e-05) (hash(x)=6668088)
785
+ 751 train 7.410710 (lr=3.4412e-05) (hash(x)=5716089)
786
+ 752 train 7.379200 (lr=3.4236e-05) (hash(x)=6991946)
787
+ 753 train 7.276920 (lr=3.4060e-05) (hash(x)=7991009)
788
+ 754 train 7.377810 (lr=3.3885e-05) (hash(x)=6037018)
789
+ 755 train 7.333932 (lr=3.3711e-05) (hash(x)=5348607)
790
+ 756 train 7.488078 (lr=3.3537e-05) (hash(x)=5190526)
791
+ 757 train 7.365550 (lr=3.3363e-05) (hash(x)=6036146)
792
+ 758 train 7.255570 (lr=3.3189e-05) (hash(x)=5329783)
793
+ 759 train 7.352256 (lr=3.3016e-05) (hash(x)=6658718)
794
+ 760 train 7.242823 (lr=3.2844e-05) (hash(x)=4820939)
795
+ 761 train 7.252680 (lr=3.2672e-05) (hash(x)=6176253)
796
+ 762 train 7.222097 (lr=3.2500e-05) (hash(x)=5316750)
797
+ 763 train 7.310517 (lr=3.2329e-05) (hash(x)=4542895)
798
+ 764 train 7.092785 (lr=3.2158e-05) (hash(x)=5395673)
799
+ 765 train 7.196866 (lr=3.1988e-05) (hash(x)=5549073)
800
+ 766 train 7.454140 (lr=3.1818e-05) (hash(x)=5233723)
801
+ 767 train 7.505625 (lr=3.1648e-05) (hash(x)=5342783)
802
+ 768 train 7.439860 (lr=3.1479e-05) (hash(x)=5552448)
803
+ 769 train 7.596329 (lr=3.1311e-05) (hash(x)=6549546)
804
+ 770 train 7.231628 (lr=3.1142e-05) (hash(x)=3816493)
805
+ 771 train 7.292289 (lr=3.0975e-05) (hash(x)=6170599)
806
+ 772 train 7.800466 (lr=3.0808e-05) (hash(x)=6708759)
807
+ 773 train 7.394939 (lr=3.0641e-05) (hash(x)=8483134)
808
+ 774 train 7.991918 (lr=3.0475e-05) (hash(x)=5583732)
809
+ 775 train 7.392884 (lr=3.0309e-05) (hash(x)=4865036)
810
+ 776 train 7.268393 (lr=3.0144e-05) (hash(x)=7234336)
811
+ 777 train 7.306252 (lr=2.9979e-05) (hash(x)=4433890)
812
+ 778 train 7.625133 (lr=2.9814e-05) (hash(x)=7484636)
813
+ 779 train 8.148495 (lr=2.9651e-05) (hash(x)=6738597)
814
+ 780 train 7.326751 (lr=2.9487e-05) (hash(x)=6752421)
815
+ 781 train 7.532525 (lr=2.9324e-05) (hash(x)=5459620)
816
+ 782 train 7.432144 (lr=2.9162e-05) (hash(x)=6718131)
817
+ 783 train 7.418931 (lr=2.9000e-05) (hash(x)=4604074)
818
+ 784 train 7.296824 (lr=2.8839e-05) (hash(x)=5946124)
819
+ 785 train 7.577932 (lr=2.8678e-05) (hash(x)=8045501)
820
+ 786 train 7.705376 (lr=2.8518e-05) (hash(x)=5712725)
821
+ 787 train 7.397311 (lr=2.8358e-05) (hash(x)=6176464)
822
+ 788 train 7.462009 (lr=2.8199e-05) (hash(x)=6276702)
823
+ 789 train 7.255145 (lr=2.8040e-05) (hash(x)=6127317)
824
+ 790 train 7.397742 (lr=2.7881e-05) (hash(x)=7221374)
825
+ 791 train 7.368383 (lr=2.7724e-05) (hash(x)=5571462)
826
+ 792 train 7.401064 (lr=2.7566e-05) (hash(x)=7518021)
827
+ 793 train 7.392969 (lr=2.7410e-05) (hash(x)=6640755)
828
+ 794 train 7.458731 (lr=2.7254e-05) (hash(x)=6870405)
829
+ 795 train 7.475354 (lr=2.7098e-05) (hash(x)=6147295)
830
+ 796 train 7.313713 (lr=2.6943e-05) (hash(x)=6782583)
831
+ 797 train 7.440812 (lr=2.6788e-05) (hash(x)=6120461)
832
+ 798 train 7.395951 (lr=2.6634e-05) (hash(x)=5893870)
833
+ 799 train 7.301673 (lr=2.6481e-05) (hash(x)=5123928)
834
+ 800 val loss 7.4132
835
+ 800 val perplexity 1657.6805
836
+ 800 train 7.384249 (lr=2.6328e-05) (hash(x)=7037251)
837
+ 801 train 7.373734 (lr=2.6176e-05) (hash(x)=5428813)
838
+ 802 train 7.306117 (lr=2.6024e-05) (hash(x)=5550387)
839
+ 803 train 7.390250 (lr=2.5873e-05) (hash(x)=6448586)
840
+ 804 train 7.534525 (lr=2.5722e-05) (hash(x)=5916934)
841
+ 805 train 7.460178 (lr=2.5572e-05) (hash(x)=8116987)
842
+ 806 train 7.465727 (lr=2.5423e-05) (hash(x)=6210899)
843
+ 807 train 7.590629 (lr=2.5274e-05) (hash(x)=5974342)
844
+ 808 train 7.577329 (lr=2.5125e-05) (hash(x)=7236199)
845
+ 809 train 7.513281 (lr=2.4978e-05) (hash(x)=5261336)
846
+ 810 train 7.334577 (lr=2.4830e-05) (hash(x)=6310720)
847
+ 811 train 7.354394 (lr=2.4684e-05) (hash(x)=5307056)
848
+ 812 train 7.198193 (lr=2.4538e-05) (hash(x)=7356633)
849
+ 813 train 7.297832 (lr=2.4392e-05) (hash(x)=5865119)
850
+ 814 train 7.192027 (lr=2.4247e-05) (hash(x)=5979205)
851
+ 815 train 7.370776 (lr=2.4103e-05) (hash(x)=6672380)
852
+ 816 train 7.375759 (lr=2.3959e-05) (hash(x)=5565030)
853
+ 817 train 7.332466 (lr=2.3816e-05) (hash(x)=6886810)
854
+ 818 train 7.392402 (lr=2.3674e-05) (hash(x)=5926478)
855
+ 819 train 7.599762 (lr=2.3532e-05) (hash(x)=6564071)
856
+ 820 train 7.237552 (lr=2.3391e-05) (hash(x)=5024280)
857
+ 821 train 7.213919 (lr=2.3250e-05) (hash(x)=7705089)
858
+ 822 train 7.193397 (lr=2.3110e-05) (hash(x)=4037371)
859
+ 823 train 7.303055 (lr=2.2971e-05) (hash(x)=5773073)
860
+ 824 train 7.321165 (lr=2.2832e-05) (hash(x)=7158139)
861
+ 825 train 7.186443 (lr=2.2694e-05) (hash(x)=5905976)
862
+ 826 train 7.622326 (lr=2.2556e-05) (hash(x)=13897369)
863
+ 827 train 8.801919 (lr=2.2420e-05) (hash(x)=6936653)
864
+ 828 train 7.360253 (lr=2.2283e-05) (hash(x)=6592246)
865
+ 829 train 7.310474 (lr=2.2148e-05) (hash(x)=6592773)
866
+ 830 train 7.395019 (lr=2.2013e-05) (hash(x)=5461178)
867
+ 831 train 7.158001 (lr=2.1878e-05) (hash(x)=6796551)
868
+ 832 train 7.362367 (lr=2.1745e-05) (hash(x)=4721691)
869
+ 833 train 7.307372 (lr=2.1612e-05) (hash(x)=6151589)
870
+ 834 train 7.440302 (lr=2.1479e-05) (hash(x)=7008995)
871
+ 835 train 7.771308 (lr=2.1347e-05) (hash(x)=7045164)
872
+ 836 train 7.356120 (lr=2.1216e-05) (hash(x)=6614221)
873
+ 837 train 7.380269 (lr=2.1086e-05) (hash(x)=6650506)
874
+ 838 train 7.217260 (lr=2.0956e-05) (hash(x)=6363762)
875
+ 839 train 7.373951 (lr=2.0827e-05) (hash(x)=7839748)
876
+ 840 train 7.272520 (lr=2.0698e-05) (hash(x)=5124077)
877
+ 841 train 7.797008 (lr=2.0570e-05) (hash(x)=6028671)
878
+ 842 train 7.226677 (lr=2.0443e-05) (hash(x)=5427175)
879
+ 843 train 7.547626 (lr=2.0317e-05) (hash(x)=6118408)
880
+ 844 train 7.331814 (lr=2.0191e-05) (hash(x)=6022552)
881
+ 845 train 7.316694 (lr=2.0066e-05) (hash(x)=5757526)
882
+ 846 train 7.836789 (lr=1.9941e-05) (hash(x)=6476549)
883
+ 847 train 7.135097 (lr=1.9818e-05) (hash(x)=6030172)
884
+ 848 train 7.519093 (lr=1.9694e-05) (hash(x)=5917229)
885
+ 849 train 7.544286 (lr=1.9572e-05) (hash(x)=6719882)
886
+ 850 val loss 7.4008
887
+ 850 val perplexity 1637.3448
888
+ 850 train 7.427021 (lr=1.9450e-05) (hash(x)=5441498)
889
+ 851 train 7.315691 (lr=1.9329e-05) (hash(x)=5028050)
890
+ 852 train 7.375894 (lr=1.9209e-05) (hash(x)=8487491)
891
+ 853 train 7.340427 (lr=1.9089e-05) (hash(x)=5682910)
892
+ 854 train 7.544554 (lr=1.8970e-05) (hash(x)=6781362)
893
+ 855 train 7.365804 (lr=1.8852e-05) (hash(x)=5747891)
894
+ 856 train 7.398807 (lr=1.8734e-05) (hash(x)=7627661)
895
+ 857 train 7.050218 (lr=1.8618e-05) (hash(x)=5734208)
896
+ 858 train 7.211808 (lr=1.8501e-05) (hash(x)=6162015)
897
+ 859 train 7.413092 (lr=1.8386e-05) (hash(x)=7024381)
898
+ 860 train 7.627446 (lr=1.8271e-05) (hash(x)=5198498)
899
+ 861 train 7.617412 (lr=1.8157e-05) (hash(x)=7814361)
900
+ 862 train 7.489380 (lr=1.8044e-05) (hash(x)=7394995)
901
+ 863 train 7.373221 (lr=1.7931e-05) (hash(x)=5765893)
902
+ 864 train 7.173771 (lr=1.7819e-05) (hash(x)=4990435)
903
+ 865 train 7.316307 (lr=1.7708e-05) (hash(x)=5572048)
904
+ 866 train 7.480632 (lr=1.7598e-05) (hash(x)=7093293)
905
+ 867 train 7.356171 (lr=1.7488e-05) (hash(x)=5610058)
906
+ 868 train 7.269330 (lr=1.7379e-05) (hash(x)=6771468)
907
+ 869 train 7.295402 (lr=1.7271e-05) (hash(x)=4459483)
908
+ 870 train 7.186852 (lr=1.7163e-05) (hash(x)=4926546)
909
+ 871 train 7.290637 (lr=1.7056e-05) (hash(x)=6758411)
910
+ 872 train 7.273262 (lr=1.6950e-05) (hash(x)=7394277)
911
+ 873 train 7.390432 (lr=1.6845e-05) (hash(x)=6701381)
912
+ 874 train 7.797503 (lr=1.6740e-05) (hash(x)=8120966)
913
+ 875 train 7.253914 (lr=1.6636e-05) (hash(x)=5416400)
914
+ 876 train 7.252780 (lr=1.6533e-05) (hash(x)=6292778)
915
+ 877 train 7.188558 (lr=1.6431e-05) (hash(x)=5919569)
916
+ 878 train 7.511118 (lr=1.6329e-05) (hash(x)=7561740)
917
+ 879 train 7.266269 (lr=1.6228e-05) (hash(x)=5282870)
918
+ 880 train 7.292302 (lr=1.6128e-05) (hash(x)=5473238)
919
+ 881 train 7.158679 (lr=1.6029e-05) (hash(x)=8072181)
920
+ 882 train 7.188862 (lr=1.5930e-05) (hash(x)=6757322)
921
+ 883 train 7.155596 (lr=1.5832e-05) (hash(x)=7996770)
922
+ 884 train 7.227199 (lr=1.5735e-05) (hash(x)=3181021)
923
+ 885 train 7.260509 (lr=1.5639e-05) (hash(x)=5093519)
924
+ 886 train 7.263531 (lr=1.5543e-05) (hash(x)=5391352)
925
+ 887 train 7.296053 (lr=1.5449e-05) (hash(x)=6270917)
926
+ 888 train 7.301110 (lr=1.5354e-05) (hash(x)=5818195)
927
+ 889 train 7.031780 (lr=1.5261e-05) (hash(x)=5852780)
928
+ 890 train 6.959300 (lr=1.5169e-05) (hash(x)=5777928)
929
+ 891 train 7.168554 (lr=1.5077e-05) (hash(x)=5327728)
930
+ 892 train 7.187828 (lr=1.4986e-05) (hash(x)=7651281)
931
+ 893 train 7.398753 (lr=1.4896e-05) (hash(x)=6260765)
932
+ 894 train 7.264123 (lr=1.4806e-05) (hash(x)=6265102)
933
+ 895 train 7.488179 (lr=1.4718e-05) (hash(x)=6212894)
934
+ 896 train 7.223478 (lr=1.4630e-05) (hash(x)=6901967)
935
+ 897 train 7.296988 (lr=1.4543e-05) (hash(x)=6724010)
936
+ 898 train 7.288163 (lr=1.4456e-05) (hash(x)=6299033)
937
+ 899 train 7.447861 (lr=1.4371e-05) (hash(x)=6059934)
938
+ 900 val loss 7.3955
939
+ 900 val perplexity 1628.6270
940
+ 900 train 7.157464 (lr=1.4286e-05) (hash(x)=6029696)
941
+ 901 train 7.054863 (lr=1.4202e-05) (hash(x)=5587382)
942
+ 902 train 7.289368 (lr=1.4119e-05) (hash(x)=5890076)
943
+ 903 train 7.214302 (lr=1.4037e-05) (hash(x)=7274108)
944
+ 904 train 7.329493 (lr=1.3955e-05) (hash(x)=7975528)
945
+ 905 train 7.345973 (lr=1.3874e-05) (hash(x)=6085396)
946
+ 906 train 7.229750 (lr=1.3794e-05) (hash(x)=6724496)
947
+ 907 train 7.238358 (lr=1.3715e-05) (hash(x)=4784666)
948
+ 908 train 7.241588 (lr=1.3637e-05) (hash(x)=7193845)
949
+ 909 train 7.426858 (lr=1.3559e-05) (hash(x)=6288638)
950
+ 910 train 7.367076 (lr=1.3482e-05) (hash(x)=5685285)
951
+ 911 train 7.355628 (lr=1.3407e-05) (hash(x)=7714288)
952
+ 912 train 7.440588 (lr=1.3331e-05) (hash(x)=5983679)
953
+ 913 train 7.379293 (lr=1.3257e-05) (hash(x)=5534748)
954
+ 914 train 7.312242 (lr=1.3183e-05) (hash(x)=6545471)
955
+ 915 train 7.330667 (lr=1.3111e-05) (hash(x)=6361270)
956
+ 916 train 7.312091 (lr=1.3039e-05) (hash(x)=6257744)
957
+ 917 train 7.326831 (lr=1.2968e-05) (hash(x)=4682956)
958
+ 918 train 7.212829 (lr=1.2897e-05) (hash(x)=4521756)
959
+ 919 train 7.364055 (lr=1.2828e-05) (hash(x)=7568935)
960
+ 920 train 7.279306 (lr=1.2759e-05) (hash(x)=7969353)
961
+ 921 train 7.456676 (lr=1.2691e-05) (hash(x)=7973311)
962
+ 922 train 7.112971 (lr=1.2624e-05) (hash(x)=4843937)
963
+ 923 train 7.297282 (lr=1.2558e-05) (hash(x)=6987352)
964
+ 924 train 7.425914 (lr=1.2493e-05) (hash(x)=7179418)
965
+ 925 train 7.054530 (lr=1.2428e-05) (hash(x)=5413904)
966
+ 926 train 7.455254 (lr=1.2364e-05) (hash(x)=6028105)
967
+ 927 train 7.401783 (lr=1.2301e-05) (hash(x)=6732924)
968
+ 928 train 7.282667 (lr=1.2239e-05) (hash(x)=4547651)
969
+ 929 train 7.452469 (lr=1.2178e-05) (hash(x)=6219754)
970
+ 930 train 7.212047 (lr=1.2118e-05) (hash(x)=7465186)
971
+ 931 train 7.370359 (lr=1.2058e-05) (hash(x)=6839070)
972
+ 932 train 7.232346 (lr=1.1999e-05) (hash(x)=5390324)
973
+ 933 train 7.205455 (lr=1.1941e-05) (hash(x)=6090120)
974
+ 934 train 7.856067 (lr=1.1884e-05) (hash(x)=7123181)
975
+ 935 train 7.527866 (lr=1.1828e-05) (hash(x)=7115176)
976
+ 936 train 7.494783 (lr=1.1772e-05) (hash(x)=7016990)
977
+ 937 train 7.291622 (lr=1.1718e-05) (hash(x)=4989299)
978
+ 938 train 7.282129 (lr=1.1664e-05) (hash(x)=6783311)
979
+ 939 train 7.260326 (lr=1.1611e-05) (hash(x)=7497590)
980
+ 940 train 7.492116 (lr=1.1559e-05) (hash(x)=6767514)
981
+ 941 train 7.267005 (lr=1.1508e-05) (hash(x)=3977557)
982
+ 942 train 7.387469 (lr=1.1457e-05) (hash(x)=4806492)
983
+ 943 train 7.275816 (lr=1.1408e-05) (hash(x)=5524165)
984
+ 944 train 7.355778 (lr=1.1359e-05) (hash(x)=4930752)
985
+ 945 train 7.340404 (lr=1.1311e-05) (hash(x)=5124936)
986
+ 946 train 7.189746 (lr=1.1264e-05) (hash(x)=6877351)
987
+ 947 train 7.485983 (lr=1.1218e-05) (hash(x)=5522119)
988
+ 948 train 7.272012 (lr=1.1173e-05) (hash(x)=6183512)
989
+ 949 train 7.196634 (lr=1.1128e-05) (hash(x)=5030865)
990
+ 950 val loss 7.3856
991
+ 950 val perplexity 1612.5575
992
+ 950 train 7.288042 (lr=1.1085e-05) (hash(x)=6619812)
993
+ 951 train 7.329567 (lr=1.1042e-05) (hash(x)=6411783)
994
+ 952 train 7.286347 (lr=1.1000e-05) (hash(x)=6748982)
995
+ 953 train 7.186785 (lr=1.0959e-05) (hash(x)=6581937)
996
+ 954 train 7.322198 (lr=1.0919e-05) (hash(x)=7881022)
997
+ 955 train 7.253920 (lr=1.0879e-05) (hash(x)=5116910)
998
+ 956 train 7.060466 (lr=1.0841e-05) (hash(x)=4880833)
999
+ 957 train 7.103237 (lr=1.0803e-05) (hash(x)=7353459)
1000
+ 958 train 7.189532 (lr=1.0766e-05) (hash(x)=4951302)
1001
+ 959 train 7.081435 (lr=1.0730e-05) (hash(x)=6344535)
1002
+ 960 train 7.758891 (lr=1.0695e-05) (hash(x)=7411763)
1003
+ 961 train 7.272539 (lr=1.0661e-05) (hash(x)=5840169)
1004
+ 962 train 7.384135 (lr=1.0628e-05) (hash(x)=5784788)
1005
+ 963 train 7.166114 (lr=1.0595e-05) (hash(x)=6754290)
1006
+ 964 train 7.475963 (lr=1.0563e-05) (hash(x)=6369963)
1007
+ 965 train 7.559347 (lr=1.0533e-05) (hash(x)=6484573)
1008
+ 966 train 7.106179 (lr=1.0503e-05) (hash(x)=6516338)
1009
+ 967 train 7.159538 (lr=1.0474e-05) (hash(x)=5260125)
1010
+ 968 train 7.296071 (lr=1.0445e-05) (hash(x)=5781624)
1011
+ 969 train 7.060678 (lr=1.0418e-05) (hash(x)=4881179)
1012
+ 970 train 7.215987 (lr=1.0391e-05) (hash(x)=6707821)
1013
+ 971 train 7.346059 (lr=1.0366e-05) (hash(x)=5001634)
1014
+ 972 train 7.197507 (lr=1.0341e-05) (hash(x)=4402949)
1015
+ 973 train 7.229362 (lr=1.0317e-05) (hash(x)=7424810)
1016
+ 974 train 7.127823 (lr=1.0294e-05) (hash(x)=4814803)
1017
+ 975 train 7.030563 (lr=1.0272e-05) (hash(x)=5176431)
1018
+ 976 train 7.231709 (lr=1.0251e-05) (hash(x)=5769393)
1019
+ 977 train 7.163228 (lr=1.0230e-05) (hash(x)=5421624)
1020
+ 978 train 7.393646 (lr=1.0211e-05) (hash(x)=5999773)
1021
+ 979 train 7.115360 (lr=1.0192e-05) (hash(x)=5674439)
1022
+ 980 train 7.189202 (lr=1.0174e-05) (hash(x)=6345351)
1023
+ 981 train 7.123557 (lr=1.0157e-05) (hash(x)=5819157)
1024
+ 982 train 7.339509 (lr=1.0141e-05) (hash(x)=11159355)
1025
+ 983 train 7.319966 (lr=1.0126e-05) (hash(x)=6506856)
1026
+ 984 train 7.196962 (lr=1.0111e-05) (hash(x)=6331023)
1027
+ 985 train 7.129354 (lr=1.0098e-05) (hash(x)=6119124)
1028
+ 986 train 7.345874 (lr=1.0085e-05) (hash(x)=8142528)
1029
+ 987 train 7.509965 (lr=1.0074e-05) (hash(x)=6241669)
1030
+ 988 train 7.135898 (lr=1.0063e-05) (hash(x)=3852991)
1031
+ 989 train 6.854803 (lr=1.0053e-05) (hash(x)=4992756)
1032
+ 990 train 7.607345 (lr=1.0044e-05) (hash(x)=7097145)
1033
+ 991 train 7.367468 (lr=1.0035e-05) (hash(x)=7277927)
1034
+ 992 train 7.453647 (lr=1.0028e-05) (hash(x)=5661400)
1035
+ 993 train 7.381728 (lr=1.0021e-05) (hash(x)=6622097)
1036
+ 994 train 7.500656 (lr=1.0016e-05) (hash(x)=5016804)
1037
+ 995 train 7.572945 (lr=1.0011e-05) (hash(x)=7026582)
1038
+ 996 train 7.344687 (lr=1.0007e-05) (hash(x)=5639136)
1039
+ 997 train 7.465575 (lr=1.0004e-05) (hash(x)=5840987)
1040
+ 998 train 7.100306 (lr=1.0002e-05) (hash(x)=5265410)
1041
+ 999 val loss 7.3793
1042
+ 999 val perplexity 1602.3990
1043
+ 999 train 7.094858 (lr=1.0000e-05) (hash(x)=4351074)
lr1e-4_total_batch_size20480_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a4d0f2bbd0fdc72910bcf671ca733b370aa44c5b16e9105cdf7034670e5f0b16
3
+ size 36163010
lr1e-4_total_batch_size20480_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2338771406a7e3d4f86484503bf8908f1cbda5024e49aee5ceb59bb3a57921ca
3
+ size 70764358