andrew-healey commited on
Commit
055a26c
·
verified ·
1 Parent(s): 7d751b3

Upload folder using huggingface_hub

Browse files
lr3e-4_total_batch_size20480_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr3e-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_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 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": 0.0003, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr3e-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
lr3e-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=1.0490e-06) (hash(x)=5113283)
5
+ 1 train 11.676708 (lr=2.0979e-06) (hash(x)=5534142)
6
+ 2 train 11.701111 (lr=3.1469e-06) (hash(x)=6763231)
7
+ 3 train 11.675781 (lr=4.1958e-06) (hash(x)=6328012)
8
+ 4 train 11.666820 (lr=5.2448e-06) (hash(x)=7052161)
9
+ 5 train 11.659793 (lr=6.2937e-06) (hash(x)=5556098)
10
+ 6 train 11.629516 (lr=7.3427e-06) (hash(x)=4696943)
11
+ 7 train 11.586353 (lr=8.3916e-06) (hash(x)=6219138)
12
+ 8 train 11.554424 (lr=9.4406e-06) (hash(x)=6049878)
13
+ 9 train 11.537850 (lr=1.0490e-05) (hash(x)=5491972)
14
+ 10 train 11.525458 (lr=1.1538e-05) (hash(x)=8075458)
15
+ 11 train 11.455849 (lr=1.2587e-05) (hash(x)=5917741)
16
+ 12 train 11.431676 (lr=1.3636e-05) (hash(x)=5750403)
17
+ 13 train 11.387834 (lr=1.4685e-05) (hash(x)=6112458)
18
+ 14 train 11.352572 (lr=1.5734e-05) (hash(x)=5806490)
19
+ 15 train 11.282389 (lr=1.6783e-05) (hash(x)=6568379)
20
+ 16 train 11.215393 (lr=1.7832e-05) (hash(x)=7212403)
21
+ 17 train 11.162113 (lr=1.8881e-05) (hash(x)=7386082)
22
+ 18 train 11.111923 (lr=1.9930e-05) (hash(x)=6359743)
23
+ 19 train 11.061787 (lr=2.0979e-05) (hash(x)=6151674)
24
+ 20 train 11.034362 (lr=2.2028e-05) (hash(x)=7168704)
25
+ 21 train 10.964106 (lr=2.3077e-05) (hash(x)=7859060)
26
+ 22 train 10.882087 (lr=2.4126e-05) (hash(x)=6976239)
27
+ 23 train 10.858698 (lr=2.5175e-05) (hash(x)=6391506)
28
+ 24 train 10.811526 (lr=2.6224e-05) (hash(x)=5245999)
29
+ 25 train 10.752240 (lr=2.7273e-05) (hash(x)=5985675)
30
+ 26 train 10.654097 (lr=2.8322e-05) (hash(x)=4355733)
31
+ 27 train 10.757939 (lr=2.9371e-05) (hash(x)=6984772)
32
+ 28 train 10.629652 (lr=3.0420e-05) (hash(x)=6244096)
33
+ 29 train 10.619232 (lr=3.1469e-05) (hash(x)=6603717)
34
+ 30 train 10.511764 (lr=3.2517e-05) (hash(x)=5669522)
35
+ 31 train 10.487042 (lr=3.3566e-05) (hash(x)=6190579)
36
+ 32 train 10.534955 (lr=3.4615e-05) (hash(x)=4912983)
37
+ 33 train 10.417913 (lr=3.5664e-05) (hash(x)=7276303)
38
+ 34 train 10.463884 (lr=3.6713e-05) (hash(x)=5479947)
39
+ 35 train 10.541850 (lr=3.7762e-05) (hash(x)=6630378)
40
+ 36 train 10.477726 (lr=3.8811e-05) (hash(x)=8097010)
41
+ 37 train 10.524167 (lr=3.9860e-05) (hash(x)=9989744)
42
+ 38 train 10.404040 (lr=4.0909e-05) (hash(x)=6694932)
43
+ 39 train 10.342054 (lr=4.1958e-05) (hash(x)=6317894)
44
+ 40 train 10.266819 (lr=4.3007e-05) (hash(x)=5046327)
45
+ 41 train 10.338000 (lr=4.4056e-05) (hash(x)=7947703)
46
+ 42 train 10.220985 (lr=4.5105e-05) (hash(x)=5916068)
47
+ 43 train 10.306259 (lr=4.6154e-05) (hash(x)=5777498)
48
+ 44 train 10.244427 (lr=4.7203e-05) (hash(x)=5753092)
49
+ 45 train 10.239865 (lr=4.8252e-05) (hash(x)=7723226)
50
+ 46 train 10.233941 (lr=4.9301e-05) (hash(x)=7007016)
51
+ 47 train 10.164026 (lr=5.0350e-05) (hash(x)=6491229)
52
+ 48 train 10.216297 (lr=5.1399e-05) (hash(x)=6390752)
53
+ 49 train 10.332674 (lr=5.2448e-05) (hash(x)=6019389)
54
+ 50 val loss 10.2210
55
+ 50 val perplexity 27474.1973
56
+ 50 train 10.207131 (lr=5.3497e-05) (hash(x)=7662067)
57
+ 51 train 10.215717 (lr=5.4545e-05) (hash(x)=4648609)
58
+ 52 train 10.160510 (lr=5.5594e-05) (hash(x)=4725966)
59
+ 53 train 10.132288 (lr=5.6643e-05) (hash(x)=5558355)
60
+ 54 train 10.174389 (lr=5.7692e-05) (hash(x)=6524933)
61
+ 55 train 10.131278 (lr=5.8741e-05) (hash(x)=6369326)
62
+ 56 train 10.102547 (lr=5.9790e-05) (hash(x)=6022625)
63
+ 57 train 10.055982 (lr=6.0839e-05) (hash(x)=4906853)
64
+ 58 train 9.905285 (lr=6.1888e-05) (hash(x)=5589269)
65
+ 59 train 10.027572 (lr=6.2937e-05) (hash(x)=6437376)
66
+ 60 train 10.062085 (lr=6.3986e-05) (hash(x)=7304892)
67
+ 61 train 10.016556 (lr=6.5035e-05) (hash(x)=7084093)
68
+ 62 train 10.036841 (lr=6.6084e-05) (hash(x)=5908178)
69
+ 63 train 9.849921 (lr=6.7133e-05) (hash(x)=7048804)
70
+ 64 train 9.929809 (lr=6.8182e-05) (hash(x)=6085549)
71
+ 65 train 9.932182 (lr=6.9231e-05) (hash(x)=5752594)
72
+ 66 train 9.758349 (lr=7.0280e-05) (hash(x)=7662769)
73
+ 67 train 9.627929 (lr=7.1329e-05) (hash(x)=6393520)
74
+ 68 train 9.875592 (lr=7.2378e-05) (hash(x)=5942867)
75
+ 69 train 9.865605 (lr=7.3427e-05) (hash(x)=6793550)
76
+ 70 train 9.733513 (lr=7.4476e-05) (hash(x)=7105976)
77
+ 71 train 9.765862 (lr=7.5524e-05) (hash(x)=6581348)
78
+ 72 train 9.716233 (lr=7.6573e-05) (hash(x)=5555598)
79
+ 73 train 9.645407 (lr=7.7622e-05) (hash(x)=6235837)
80
+ 74 train 9.594195 (lr=7.8671e-05) (hash(x)=5591770)
81
+ 75 train 9.782149 (lr=7.9720e-05) (hash(x)=7083021)
82
+ 76 train 9.627707 (lr=8.0769e-05) (hash(x)=7281348)
83
+ 77 train 9.650533 (lr=8.1818e-05) (hash(x)=6155821)
84
+ 78 train 9.516935 (lr=8.2867e-05) (hash(x)=5736241)
85
+ 79 train 9.447936 (lr=8.3916e-05) (hash(x)=5657123)
86
+ 80 train 9.325729 (lr=8.4965e-05) (hash(x)=6452095)
87
+ 81 train 9.352224 (lr=8.6014e-05) (hash(x)=5312267)
88
+ 82 train 9.337975 (lr=8.7063e-05) (hash(x)=5434493)
89
+ 83 train 9.412885 (lr=8.8112e-05) (hash(x)=6830813)
90
+ 84 train 9.311926 (lr=8.9161e-05) (hash(x)=4752771)
91
+ 85 train 9.317992 (lr=9.0210e-05) (hash(x)=6424933)
92
+ 86 train 9.231092 (lr=9.1259e-05) (hash(x)=5786517)
93
+ 87 train 9.228902 (lr=9.2308e-05) (hash(x)=6680023)
94
+ 88 train 9.124507 (lr=9.3357e-05) (hash(x)=6474729)
95
+ 89 train 9.173040 (lr=9.4406e-05) (hash(x)=7523279)
96
+ 90 train 8.999133 (lr=9.5455e-05) (hash(x)=5508768)
97
+ 91 train 9.072744 (lr=9.6503e-05) (hash(x)=6635925)
98
+ 92 train 9.032050 (lr=9.7552e-05) (hash(x)=6663801)
99
+ 93 train 8.996291 (lr=9.8601e-05) (hash(x)=5869441)
100
+ 94 train 8.985525 (lr=9.9650e-05) (hash(x)=6698878)
101
+ 95 train 8.938889 (lr=1.0070e-04) (hash(x)=6349004)
102
+ 96 train 8.915824 (lr=1.0175e-04) (hash(x)=4986137)
103
+ 97 train 8.705990 (lr=1.0280e-04) (hash(x)=6008454)
104
+ 98 train 8.736839 (lr=1.0385e-04) (hash(x)=6550770)
105
+ 99 train 8.823233 (lr=1.0490e-04) (hash(x)=7682741)
106
+ 100 val loss 8.7937
107
+ 100 val perplexity 6592.7949
108
+ 100 train 8.849685 (lr=1.0594e-04) (hash(x)=7892158)
109
+ 101 train 8.745862 (lr=1.0699e-04) (hash(x)=7794028)
110
+ 102 train 8.734698 (lr=1.0804e-04) (hash(x)=7093523)
111
+ 103 train 8.696432 (lr=1.0909e-04) (hash(x)=6166062)
112
+ 104 train 8.634153 (lr=1.1014e-04) (hash(x)=6645781)
113
+ 105 train 8.631236 (lr=1.1119e-04) (hash(x)=4811192)
114
+ 106 train 8.527868 (lr=1.1224e-04) (hash(x)=5970866)
115
+ 107 train 8.566043 (lr=1.1329e-04) (hash(x)=7822680)
116
+ 108 train 8.471962 (lr=1.1434e-04) (hash(x)=6317317)
117
+ 109 train 8.506813 (lr=1.1538e-04) (hash(x)=6431409)
118
+ 110 train 8.519050 (lr=1.1643e-04) (hash(x)=6653337)
119
+ 111 train 8.434496 (lr=1.1748e-04) (hash(x)=5323032)
120
+ 112 train 8.370646 (lr=1.1853e-04) (hash(x)=5125339)
121
+ 113 train 8.351119 (lr=1.1958e-04) (hash(x)=7363286)
122
+ 114 train 8.539948 (lr=1.2063e-04) (hash(x)=7670050)
123
+ 115 train 8.361479 (lr=1.2168e-04) (hash(x)=5616075)
124
+ 116 train 8.390615 (lr=1.2273e-04) (hash(x)=6593764)
125
+ 117 train 8.417207 (lr=1.2378e-04) (hash(x)=5672608)
126
+ 118 train 8.325613 (lr=1.2483e-04) (hash(x)=6828653)
127
+ 119 train 8.275153 (lr=1.2587e-04) (hash(x)=6839867)
128
+ 120 train 8.342884 (lr=1.2692e-04) (hash(x)=8426992)
129
+ 121 train 8.093615 (lr=1.2797e-04) (hash(x)=4916928)
130
+ 122 train 8.249154 (lr=1.2902e-04) (hash(x)=6484050)
131
+ 123 train 8.298876 (lr=1.3007e-04) (hash(x)=8353379)
132
+ 124 train 8.190125 (lr=1.3112e-04) (hash(x)=6552510)
133
+ 125 train 8.198230 (lr=1.3217e-04) (hash(x)=4838871)
134
+ 126 train 7.971184 (lr=1.3322e-04) (hash(x)=3021697)
135
+ 127 train 8.057299 (lr=1.3427e-04) (hash(x)=6795665)
136
+ 128 train 8.020698 (lr=1.3531e-04) (hash(x)=5486935)
137
+ 129 train 8.192800 (lr=1.3636e-04) (hash(x)=6678038)
138
+ 130 train 8.187574 (lr=1.3741e-04) (hash(x)=6657714)
139
+ 131 train 8.138894 (lr=1.3846e-04) (hash(x)=6112215)
140
+ 132 train 8.114148 (lr=1.3951e-04) (hash(x)=8353143)
141
+ 133 train 8.111702 (lr=1.4056e-04) (hash(x)=5915361)
142
+ 134 train 8.003416 (lr=1.4161e-04) (hash(x)=6254885)
143
+ 135 train 8.043524 (lr=1.4266e-04) (hash(x)=5771588)
144
+ 136 train 8.000712 (lr=1.4371e-04) (hash(x)=5561507)
145
+ 137 train 7.969571 (lr=1.4476e-04) (hash(x)=5688829)
146
+ 138 train 7.905279 (lr=1.4580e-04) (hash(x)=5371951)
147
+ 139 train 8.022487 (lr=1.4685e-04) (hash(x)=5443305)
148
+ 140 train 8.070007 (lr=1.4790e-04) (hash(x)=5016757)
149
+ 141 train 8.042087 (lr=1.4895e-04) (hash(x)=6901933)
150
+ 142 train 7.990163 (lr=1.5000e-04) (hash(x)=6019085)
151
+ 143 train 7.879982 (lr=1.5105e-04) (hash(x)=4294425)
152
+ 144 train 7.935308 (lr=1.5210e-04) (hash(x)=6095229)
153
+ 145 train 7.785444 (lr=1.5315e-04) (hash(x)=4363941)
154
+ 146 train 7.870748 (lr=1.5420e-04) (hash(x)=4693798)
155
+ 147 train 8.296747 (lr=1.5524e-04) (hash(x)=5486335)
156
+ 148 train 8.121498 (lr=1.5629e-04) (hash(x)=7907450)
157
+ 149 train 7.992430 (lr=1.5734e-04) (hash(x)=7295165)
158
+ 150 val loss 7.8979
159
+ 150 val perplexity 2691.5112
160
+ 150 train 7.912203 (lr=1.5839e-04) (hash(x)=6814026)
161
+ 151 train 7.993955 (lr=1.5944e-04) (hash(x)=6267424)
162
+ 152 train 7.985445 (lr=1.6049e-04) (hash(x)=8086437)
163
+ 153 train 8.001595 (lr=1.6154e-04) (hash(x)=5270452)
164
+ 154 train 7.871445 (lr=1.6259e-04) (hash(x)=6420820)
165
+ 155 train 7.778224 (lr=1.6364e-04) (hash(x)=6305297)
166
+ 156 train 7.925109 (lr=1.6469e-04) (hash(x)=7889849)
167
+ 157 train 8.225861 (lr=1.6573e-04) (hash(x)=6040077)
168
+ 158 train 8.231306 (lr=1.6678e-04) (hash(x)=5652497)
169
+ 159 train 7.815346 (lr=1.6783e-04) (hash(x)=7130251)
170
+ 160 train 7.704233 (lr=1.6888e-04) (hash(x)=6610177)
171
+ 161 train 7.763867 (lr=1.6993e-04) (hash(x)=6508350)
172
+ 162 train 7.912722 (lr=1.7098e-04) (hash(x)=6994983)
173
+ 163 train 7.852048 (lr=1.7203e-04) (hash(x)=5859576)
174
+ 164 train 7.418035 (lr=1.7308e-04) (hash(x)=5487065)
175
+ 165 train 7.663095 (lr=1.7413e-04) (hash(x)=4892065)
176
+ 166 train 7.556420 (lr=1.7517e-04) (hash(x)=4933674)
177
+ 167 train 7.815187 (lr=1.7622e-04) (hash(x)=7317289)
178
+ 168 train 7.889915 (lr=1.7727e-04) (hash(x)=6811522)
179
+ 169 train 7.721096 (lr=1.7832e-04) (hash(x)=5510218)
180
+ 170 train 7.878978 (lr=1.7937e-04) (hash(x)=7304235)
181
+ 171 train 8.009256 (lr=1.8042e-04) (hash(x)=6752265)
182
+ 172 train 7.722233 (lr=1.8147e-04) (hash(x)=5994476)
183
+ 173 train 7.691530 (lr=1.8252e-04) (hash(x)=6530867)
184
+ 174 train 7.787774 (lr=1.8357e-04) (hash(x)=5746260)
185
+ 175 train 7.746581 (lr=1.8462e-04) (hash(x)=5790078)
186
+ 176 train 7.753730 (lr=1.8566e-04) (hash(x)=6430135)
187
+ 177 train 7.808633 (lr=1.8671e-04) (hash(x)=7614023)
188
+ 178 train 7.795862 (lr=1.8776e-04) (hash(x)=5244861)
189
+ 179 train 7.783302 (lr=1.8881e-04) (hash(x)=5245146)
190
+ 180 train 7.680959 (lr=1.8986e-04) (hash(x)=6415565)
191
+ 181 train 7.732317 (lr=1.9091e-04) (hash(x)=5786972)
192
+ 182 train 7.758905 (lr=1.9196e-04) (hash(x)=6131703)
193
+ 183 train 7.771471 (lr=1.9301e-04) (hash(x)=6346459)
194
+ 184 train 7.810459 (lr=1.9406e-04) (hash(x)=6504131)
195
+ 185 train 7.791246 (lr=1.9510e-04) (hash(x)=5209234)
196
+ 186 train 7.701930 (lr=1.9615e-04) (hash(x)=6163272)
197
+ 187 train 7.693328 (lr=1.9720e-04) (hash(x)=5676768)
198
+ 188 train 7.784185 (lr=1.9825e-04) (hash(x)=6090561)
199
+ 189 train 7.906965 (lr=1.9930e-04) (hash(x)=5863092)
200
+ 190 train 7.827556 (lr=2.0035e-04) (hash(x)=7201312)
201
+ 191 train 7.895401 (lr=2.0140e-04) (hash(x)=7019489)
202
+ 192 train 7.665209 (lr=2.0245e-04) (hash(x)=6225608)
203
+ 193 train 7.753797 (lr=2.0350e-04) (hash(x)=5483019)
204
+ 194 train 7.701184 (lr=2.0455e-04) (hash(x)=5903258)
205
+ 195 train 7.737371 (lr=2.0559e-04) (hash(x)=5968716)
206
+ 196 train 7.723792 (lr=2.0664e-04) (hash(x)=5778017)
207
+ 197 train 7.482740 (lr=2.0769e-04) (hash(x)=7780194)
208
+ 198 train 7.662518 (lr=2.0874e-04) (hash(x)=6140998)
209
+ 199 train 7.652005 (lr=2.0979e-04) (hash(x)=3784321)
210
+ 200 val loss 7.6755
211
+ 200 val perplexity 2155.0044
212
+ 200 train 7.671607 (lr=2.1084e-04) (hash(x)=5860821)
213
+ 201 train 7.723384 (lr=2.1189e-04) (hash(x)=7774109)
214
+ 202 train 7.859271 (lr=2.1294e-04) (hash(x)=7997664)
215
+ 203 train 7.848633 (lr=2.1399e-04) (hash(x)=6576749)
216
+ 204 train 7.770655 (lr=2.1503e-04) (hash(x)=7266655)
217
+ 205 train 7.737950 (lr=2.1608e-04) (hash(x)=7072404)
218
+ 206 train 7.651393 (lr=2.1713e-04) (hash(x)=6626459)
219
+ 207 train 7.667358 (lr=2.1818e-04) (hash(x)=6556025)
220
+ 208 train 7.702430 (lr=2.1923e-04) (hash(x)=5942129)
221
+ 209 train 7.788536 (lr=2.2028e-04) (hash(x)=7325027)
222
+ 210 train 7.782286 (lr=2.2133e-04) (hash(x)=7322467)
223
+ 211 train 7.729346 (lr=2.2238e-04) (hash(x)=6373412)
224
+ 212 train 7.521658 (lr=2.2343e-04) (hash(x)=7332255)
225
+ 213 train 7.785574 (lr=2.2448e-04) (hash(x)=5262868)
226
+ 214 train 7.520272 (lr=2.2552e-04) (hash(x)=5688247)
227
+ 215 train 7.683411 (lr=2.2657e-04) (hash(x)=8055563)
228
+ 216 train 7.685807 (lr=2.2762e-04) (hash(x)=5352405)
229
+ 217 train 7.741713 (lr=2.2867e-04) (hash(x)=5932790)
230
+ 218 train 7.706045 (lr=2.2972e-04) (hash(x)=6249312)
231
+ 219 train 7.655143 (lr=2.3077e-04) (hash(x)=5680154)
232
+ 220 train 7.747571 (lr=2.3182e-04) (hash(x)=8156280)
233
+ 221 train 7.698575 (lr=2.3287e-04) (hash(x)=5914217)
234
+ 222 train 7.556319 (lr=2.3392e-04) (hash(x)=7120215)
235
+ 223 train 7.719099 (lr=2.3497e-04) (hash(x)=6424128)
236
+ 224 train 7.480410 (lr=2.3601e-04) (hash(x)=4390027)
237
+ 225 train 7.682307 (lr=2.3706e-04) (hash(x)=7804089)
238
+ 226 train 7.777087 (lr=2.3811e-04) (hash(x)=7130267)
239
+ 227 train 7.594497 (lr=2.3916e-04) (hash(x)=6490149)
240
+ 228 train 7.627467 (lr=2.4021e-04) (hash(x)=9276378)
241
+ 229 train 7.655799 (lr=2.4126e-04) (hash(x)=6283540)
242
+ 230 train 7.650188 (lr=2.4231e-04) (hash(x)=5781680)
243
+ 231 train 7.608116 (lr=2.4336e-04) (hash(x)=5728337)
244
+ 232 train 7.589152 (lr=2.4441e-04) (hash(x)=6278202)
245
+ 233 train 7.693795 (lr=2.4545e-04) (hash(x)=7400272)
246
+ 234 train 7.658174 (lr=2.4650e-04) (hash(x)=5795128)
247
+ 235 train 7.626445 (lr=2.4755e-04) (hash(x)=6473432)
248
+ 236 train 7.644034 (lr=2.4860e-04) (hash(x)=5201502)
249
+ 237 train 7.837049 (lr=2.4965e-04) (hash(x)=6922646)
250
+ 238 train 7.576366 (lr=2.5070e-04) (hash(x)=5484641)
251
+ 239 train 7.641743 (lr=2.5175e-04) (hash(x)=7277141)
252
+ 240 train 7.646794 (lr=2.5280e-04) (hash(x)=5604614)
253
+ 241 train 7.706406 (lr=2.5385e-04) (hash(x)=7443124)
254
+ 242 train 7.717721 (lr=2.5490e-04) (hash(x)=5601968)
255
+ 243 train 7.611050 (lr=2.5594e-04) (hash(x)=5632735)
256
+ 244 train 7.588712 (lr=2.5699e-04) (hash(x)=5594810)
257
+ 245 train 7.457670 (lr=2.5804e-04) (hash(x)=5881015)
258
+ 246 train 7.599965 (lr=2.5909e-04) (hash(x)=6865361)
259
+ 247 train 7.698141 (lr=2.6014e-04) (hash(x)=5809411)
260
+ 248 train 7.689106 (lr=2.6119e-04) (hash(x)=6293025)
261
+ 249 train 7.615714 (lr=2.6224e-04) (hash(x)=7232964)
262
+ 250 val loss 7.6412
263
+ 250 val perplexity 2082.2432
264
+ 250 train 7.568135 (lr=2.6329e-04) (hash(x)=5646619)
265
+ 251 train 7.694153 (lr=2.6434e-04) (hash(x)=6171829)
266
+ 252 train 7.737913 (lr=2.6538e-04) (hash(x)=7787552)
267
+ 253 train 7.703186 (lr=2.6643e-04) (hash(x)=5884456)
268
+ 254 train 7.724929 (lr=2.6748e-04) (hash(x)=6752162)
269
+ 255 train 7.620440 (lr=2.6853e-04) (hash(x)=5077031)
270
+ 256 train 7.799653 (lr=2.6958e-04) (hash(x)=6883481)
271
+ 257 train 7.692284 (lr=2.7063e-04) (hash(x)=6248895)
272
+ 258 train 7.562183 (lr=2.7168e-04) (hash(x)=6636615)
273
+ 259 train 7.749110 (lr=2.7273e-04) (hash(x)=6918004)
274
+ 260 train 7.660610 (lr=2.7378e-04) (hash(x)=7302874)
275
+ 261 train 7.780438 (lr=2.7483e-04) (hash(x)=6482466)
276
+ 262 train 7.622699 (lr=2.7587e-04) (hash(x)=7298317)
277
+ 263 train 7.654732 (lr=2.7692e-04) (hash(x)=6588573)
278
+ 264 train 7.656054 (lr=2.7797e-04) (hash(x)=6229630)
279
+ 265 train 7.719543 (lr=2.7902e-04) (hash(x)=6136939)
280
+ 266 train 7.480398 (lr=2.8007e-04) (hash(x)=6394645)
281
+ 267 train 7.703724 (lr=2.8112e-04) (hash(x)=7040157)
282
+ 268 train 7.755861 (lr=2.8217e-04) (hash(x)=7041858)
283
+ 269 train 7.674430 (lr=2.8322e-04) (hash(x)=6455595)
284
+ 270 train 7.596068 (lr=2.8427e-04) (hash(x)=6480443)
285
+ 271 train 7.752872 (lr=2.8531e-04) (hash(x)=4122983)
286
+ 272 train 7.589568 (lr=2.8636e-04) (hash(x)=7672190)
287
+ 273 train 7.625854 (lr=2.8741e-04) (hash(x)=4724052)
288
+ 274 train 7.574381 (lr=2.8846e-04) (hash(x)=6654267)
289
+ 275 train 7.564238 (lr=2.8951e-04) (hash(x)=6820797)
290
+ 276 train 7.649192 (lr=2.9056e-04) (hash(x)=6458662)
291
+ 277 train 8.027090 (lr=2.9161e-04) (hash(x)=7477594)
292
+ 278 train 7.702743 (lr=2.9266e-04) (hash(x)=5876574)
293
+ 279 train 7.423265 (lr=2.9371e-04) (hash(x)=6146305)
294
+ 280 train 7.606751 (lr=2.9476e-04) (hash(x)=6162080)
295
+ 281 train 7.544919 (lr=2.9580e-04) (hash(x)=6814831)
296
+ 282 train 7.678919 (lr=2.9685e-04) (hash(x)=7138295)
297
+ 283 train 7.459593 (lr=2.9790e-04) (hash(x)=5548298)
298
+ 284 train 7.562262 (lr=2.9895e-04) (hash(x)=6265078)
299
+ 285 train 7.934306 (lr=3.0000e-04) (hash(x)=7609416)
300
+ 286 train 7.632717 (lr=3.0000e-04) (hash(x)=6909367)
301
+ 287 train 7.729055 (lr=3.0000e-04) (hash(x)=7574342)
302
+ 288 train 7.548546 (lr=2.9999e-04) (hash(x)=8859748)
303
+ 289 train 7.500642 (lr=2.9999e-04) (hash(x)=5906937)
304
+ 290 train 7.521149 (lr=2.9998e-04) (hash(x)=7243024)
305
+ 291 train 7.640011 (lr=2.9997e-04) (hash(x)=8013471)
306
+ 292 train 7.550640 (lr=2.9995e-04) (hash(x)=5659461)
307
+ 293 train 7.586187 (lr=2.9994e-04) (hash(x)=6146213)
308
+ 294 train 7.642905 (lr=2.9992e-04) (hash(x)=6602917)
309
+ 295 train 7.559807 (lr=2.9989e-04) (hash(x)=6709902)
310
+ 296 train 7.672787 (lr=2.9987e-04) (hash(x)=8154341)
311
+ 297 train 7.826360 (lr=2.9984e-04) (hash(x)=6402571)
312
+ 298 train 7.566641 (lr=2.9981e-04) (hash(x)=5605466)
313
+ 299 train 7.530876 (lr=2.9978e-04) (hash(x)=7683614)
314
+ 300 val loss 7.5818
315
+ 300 val perplexity 1962.0988
316
+ 300 train 7.628918 (lr=2.9974e-04) (hash(x)=5421572)
317
+ 301 train 7.460176 (lr=2.9971e-04) (hash(x)=5453026)
318
+ 302 train 7.540658 (lr=2.9967e-04) (hash(x)=6437746)
319
+ 303 train 7.433335 (lr=2.9962e-04) (hash(x)=6260017)
320
+ 304 train 7.545810 (lr=2.9958e-04) (hash(x)=6119053)
321
+ 305 train 7.710012 (lr=2.9953e-04) (hash(x)=6591466)
322
+ 306 train 7.608360 (lr=2.9948e-04) (hash(x)=5652701)
323
+ 307 train 7.588357 (lr=2.9942e-04) (hash(x)=5501472)
324
+ 308 train 7.785321 (lr=2.9937e-04) (hash(x)=6160695)
325
+ 309 train 7.524523 (lr=2.9931e-04) (hash(x)=5871698)
326
+ 310 train 7.499079 (lr=2.9925e-04) (hash(x)=7301184)
327
+ 311 train 7.430365 (lr=2.9918e-04) (hash(x)=4282052)
328
+ 312 train 7.635791 (lr=2.9912e-04) (hash(x)=5679330)
329
+ 313 train 7.482482 (lr=2.9905e-04) (hash(x)=7824660)
330
+ 314 train 7.521547 (lr=2.9898e-04) (hash(x)=4857318)
331
+ 315 train 7.457191 (lr=2.9890e-04) (hash(x)=5954950)
332
+ 316 train 7.550009 (lr=2.9883e-04) (hash(x)=5758178)
333
+ 317 train 7.549527 (lr=2.9875e-04) (hash(x)=5484656)
334
+ 318 train 7.567085 (lr=2.9866e-04) (hash(x)=6818913)
335
+ 319 train 7.758262 (lr=2.9858e-04) (hash(x)=7031328)
336
+ 320 train 7.603953 (lr=2.9849e-04) (hash(x)=6837621)
337
+ 321 train 7.631477 (lr=2.9840e-04) (hash(x)=9157069)
338
+ 322 train 7.881153 (lr=2.9831e-04) (hash(x)=6509536)
339
+ 323 train 7.555462 (lr=2.9821e-04) (hash(x)=5996883)
340
+ 324 train 7.789314 (lr=2.9812e-04) (hash(x)=6850868)
341
+ 325 train 7.738125 (lr=2.9802e-04) (hash(x)=7645447)
342
+ 326 train 7.747761 (lr=2.9791e-04) (hash(x)=6717515)
343
+ 327 train 7.845452 (lr=2.9781e-04) (hash(x)=6336164)
344
+ 328 train 7.582549 (lr=2.9770e-04) (hash(x)=6897445)
345
+ 329 train 7.490012 (lr=2.9759e-04) (hash(x)=5224240)
346
+ 330 train 7.632115 (lr=2.9748e-04) (hash(x)=6612945)
347
+ 331 train 8.103322 (lr=2.9736e-04) (hash(x)=5447801)
348
+ 332 train 7.624482 (lr=2.9724e-04) (hash(x)=7308579)
349
+ 333 train 7.488214 (lr=2.9712e-04) (hash(x)=4155598)
350
+ 334 train 7.518149 (lr=2.9700e-04) (hash(x)=6106773)
351
+ 335 train 7.554475 (lr=2.9687e-04) (hash(x)=7482744)
352
+ 336 train 7.631572 (lr=2.9675e-04) (hash(x)=6310687)
353
+ 337 train 7.508594 (lr=2.9662e-04) (hash(x)=4733414)
354
+ 338 train 7.628943 (lr=2.9648e-04) (hash(x)=6979734)
355
+ 339 train 7.416513 (lr=2.9635e-04) (hash(x)=7806284)
356
+ 340 train 7.435606 (lr=2.9621e-04) (hash(x)=5080536)
357
+ 341 train 7.655305 (lr=2.9607e-04) (hash(x)=7236478)
358
+ 342 train 7.584878 (lr=2.9592e-04) (hash(x)=6619318)
359
+ 343 train 7.485748 (lr=2.9578e-04) (hash(x)=4656255)
360
+ 344 train 7.433942 (lr=2.9563e-04) (hash(x)=5546888)
361
+ 345 train 7.862831 (lr=2.9548e-04) (hash(x)=9298325)
362
+ 346 train 7.570029 (lr=2.9532e-04) (hash(x)=5994232)
363
+ 347 train 7.554264 (lr=2.9517e-04) (hash(x)=6287524)
364
+ 348 train 7.654413 (lr=2.9501e-04) (hash(x)=6200049)
365
+ 349 train 7.571229 (lr=2.9485e-04) (hash(x)=7895970)
366
+ 350 val loss 7.5234
367
+ 350 val perplexity 1850.8209
368
+ 350 train 7.495976 (lr=2.9468e-04) (hash(x)=6051847)
369
+ 351 train 7.645227 (lr=2.9452e-04) (hash(x)=8847992)
370
+ 352 train 7.987313 (lr=2.9435e-04) (hash(x)=5123883)
371
+ 353 train 7.602204 (lr=2.9418e-04) (hash(x)=6338998)
372
+ 354 train 7.579670 (lr=2.9400e-04) (hash(x)=5765866)
373
+ 355 train 7.940150 (lr=2.9383e-04) (hash(x)=7104011)
374
+ 356 train 7.679754 (lr=2.9365e-04) (hash(x)=7438267)
375
+ 357 train 7.416760 (lr=2.9347e-04) (hash(x)=5573214)
376
+ 358 train 7.705954 (lr=2.9328e-04) (hash(x)=6897283)
377
+ 359 train 7.478729 (lr=2.9310e-04) (hash(x)=6882897)
378
+ 360 train 7.612303 (lr=2.9291e-04) (hash(x)=5343615)
379
+ 361 train 7.500973 (lr=2.9272e-04) (hash(x)=5854242)
380
+ 362 train 7.416682 (lr=2.9252e-04) (hash(x)=6317324)
381
+ 363 train 7.352943 (lr=2.9233e-04) (hash(x)=6972494)
382
+ 364 train 7.426230 (lr=2.9213e-04) (hash(x)=5661167)
383
+ 365 train 7.435387 (lr=2.9193e-04) (hash(x)=6796805)
384
+ 366 train 7.289375 (lr=2.9172e-04) (hash(x)=5817209)
385
+ 367 train 7.349096 (lr=2.9152e-04) (hash(x)=6139120)
386
+ 368 train 7.532022 (lr=2.9131e-04) (hash(x)=6121296)
387
+ 369 train 7.492011 (lr=2.9110e-04) (hash(x)=6486796)
388
+ 370 train 7.413053 (lr=2.9088e-04) (hash(x)=5514645)
389
+ 371 train 7.347468 (lr=2.9067e-04) (hash(x)=5578797)
390
+ 372 train 7.349841 (lr=2.9045e-04) (hash(x)=7651738)
391
+ 373 train 7.564703 (lr=2.9023e-04) (hash(x)=6895326)
392
+ 374 train 7.556191 (lr=2.9001e-04) (hash(x)=5490256)
393
+ 375 train 7.428305 (lr=2.8978e-04) (hash(x)=7783160)
394
+ 376 train 7.635077 (lr=2.8955e-04) (hash(x)=5608424)
395
+ 377 train 7.412133 (lr=2.8932e-04) (hash(x)=4867625)
396
+ 378 train 7.313915 (lr=2.8909e-04) (hash(x)=4655472)
397
+ 379 train 7.371718 (lr=2.8885e-04) (hash(x)=6631123)
398
+ 380 train 7.036976 (lr=2.8862e-04) (hash(x)=5751990)
399
+ 381 train 7.337749 (lr=2.8838e-04) (hash(x)=6710965)
400
+ 382 train 7.304642 (lr=2.8813e-04) (hash(x)=5293800)
401
+ 383 train 7.340374 (lr=2.8789e-04) (hash(x)=6294657)
402
+ 384 train 7.477897 (lr=2.8764e-04) (hash(x)=6999538)
403
+ 385 train 7.495056 (lr=2.8739e-04) (hash(x)=5443606)
404
+ 386 train 7.461398 (lr=2.8714e-04) (hash(x)=5650401)
405
+ 387 train 7.479991 (lr=2.8689e-04) (hash(x)=7035770)
406
+ 388 train 7.416465 (lr=2.8663e-04) (hash(x)=6678271)
407
+ 389 train 7.619193 (lr=2.8637e-04) (hash(x)=4899541)
408
+ 390 train 7.344674 (lr=2.8611e-04) (hash(x)=5359260)
409
+ 391 train 7.342348 (lr=2.8585e-04) (hash(x)=6227120)
410
+ 392 train 7.404020 (lr=2.8558e-04) (hash(x)=6643916)
411
+ 393 train 7.489176 (lr=2.8531e-04) (hash(x)=5981648)
412
+ 394 train 7.424512 (lr=2.8504e-04) (hash(x)=6467540)
413
+ 395 train 7.364933 (lr=2.8477e-04) (hash(x)=5558993)
414
+ 396 train 7.519913 (lr=2.8449e-04) (hash(x)=6698959)
415
+ 397 train 7.336070 (lr=2.8422e-04) (hash(x)=7015671)
416
+ 398 train 7.334653 (lr=2.8394e-04) (hash(x)=6800890)
417
+ 399 train 7.496864 (lr=2.8365e-04) (hash(x)=5931234)
418
+ 400 val loss 7.4354
419
+ 400 val perplexity 1694.9639
420
+ 400 train 7.507167 (lr=2.8337e-04) (hash(x)=3995070)
421
+ 401 train 7.432189 (lr=2.8308e-04) (hash(x)=4582159)
422
+ 402 train 7.492960 (lr=2.8279e-04) (hash(x)=7081906)
423
+ 403 train 7.392166 (lr=2.8250e-04) (hash(x)=5029472)
424
+ 404 train 7.399755 (lr=2.8221e-04) (hash(x)=3946232)
425
+ 405 train 7.293565 (lr=2.8191e-04) (hash(x)=5434102)
426
+ 406 train 8.070885 (lr=2.8162e-04) (hash(x)=5340646)
427
+ 407 train 7.181291 (lr=2.8131e-04) (hash(x)=6810638)
428
+ 408 train 7.453653 (lr=2.8101e-04) (hash(x)=6768026)
429
+ 409 train 7.432594 (lr=2.8071e-04) (hash(x)=6521562)
430
+ 410 train 7.405164 (lr=2.8040e-04) (hash(x)=4429434)
431
+ 411 train 7.469281 (lr=2.8009e-04) (hash(x)=7226932)
432
+ 412 train 7.261937 (lr=2.7978e-04) (hash(x)=5065932)
433
+ 413 train 7.414805 (lr=2.7947e-04) (hash(x)=6957849)
434
+ 414 train 7.404327 (lr=2.7915e-04) (hash(x)=6431457)
435
+ 415 train 7.450206 (lr=2.7883e-04) (hash(x)=5334281)
436
+ 416 train 7.492687 (lr=2.7851e-04) (hash(x)=5471747)
437
+ 417 train 7.578084 (lr=2.7819e-04) (hash(x)=5887983)
438
+ 418 train 7.484718 (lr=2.7786e-04) (hash(x)=5075909)
439
+ 419 train 7.282494 (lr=2.7754e-04) (hash(x)=6393576)
440
+ 420 train 7.392065 (lr=2.7721e-04) (hash(x)=6608979)
441
+ 421 train 7.817040 (lr=2.7688e-04) (hash(x)=5057959)
442
+ 422 train 7.755002 (lr=2.7654e-04) (hash(x)=6414858)
443
+ 423 train 7.512033 (lr=2.7621e-04) (hash(x)=8595271)
444
+ 424 train 7.366114 (lr=2.7587e-04) (hash(x)=6500949)
445
+ 425 train 7.531248 (lr=2.7553e-04) (hash(x)=8395167)
446
+ 426 train 7.645133 (lr=2.7519e-04) (hash(x)=6288420)
447
+ 427 train 7.437893 (lr=2.7484e-04) (hash(x)=7634417)
448
+ 428 train 7.342216 (lr=2.7450e-04) (hash(x)=6795561)
449
+ 429 train 7.249804 (lr=2.7415e-04) (hash(x)=5257771)
450
+ 430 train 7.413533 (lr=2.7380e-04) (hash(x)=7506860)
451
+ 431 train 7.453298 (lr=2.7344e-04) (hash(x)=7674238)
452
+ 432 train 7.578542 (lr=2.7309e-04) (hash(x)=5353794)
453
+ 433 train 7.479118 (lr=2.7273e-04) (hash(x)=5300555)
454
+ 434 train 7.190542 (lr=2.7237e-04) (hash(x)=5564419)
455
+ 435 train 7.457651 (lr=2.7201e-04) (hash(x)=6066139)
456
+ 436 train 7.416438 (lr=2.7165e-04) (hash(x)=6863709)
457
+ 437 train 7.407330 (lr=2.7128e-04) (hash(x)=6688212)
458
+ 438 train 7.567044 (lr=2.7092e-04) (hash(x)=7095090)
459
+ 439 train 7.649981 (lr=2.7055e-04) (hash(x)=6642404)
460
+ 440 train 7.373684 (lr=2.7018e-04) (hash(x)=6539227)
461
+ 441 train 7.359828 (lr=2.6980e-04) (hash(x)=4961570)
462
+ 442 train 7.302871 (lr=2.6943e-04) (hash(x)=6606391)
463
+ 443 train 7.098207 (lr=2.6905e-04) (hash(x)=5105190)
464
+ 444 train 7.695505 (lr=2.6867e-04) (hash(x)=5695997)
465
+ 445 train 7.371474 (lr=2.6829e-04) (hash(x)=7155991)
466
+ 446 train 7.365673 (lr=2.6791e-04) (hash(x)=7249727)
467
+ 447 train 7.337553 (lr=2.6752e-04) (hash(x)=5132972)
468
+ 448 train 7.315939 (lr=2.6713e-04) (hash(x)=5381802)
469
+ 449 train 7.363783 (lr=2.6674e-04) (hash(x)=6269946)
470
+ 450 val loss 7.3617
471
+ 450 val perplexity 1574.4968
472
+ 450 train 7.324046 (lr=2.6635e-04) (hash(x)=4976762)
473
+ 451 train 7.407941 (lr=2.6596e-04) (hash(x)=5396133)
474
+ 452 train 7.366235 (lr=2.6556e-04) (hash(x)=6173473)
475
+ 453 train 7.338311 (lr=2.6517e-04) (hash(x)=6394083)
476
+ 454 train 7.400823 (lr=2.6477e-04) (hash(x)=6436030)
477
+ 455 train 7.264842 (lr=2.6437e-04) (hash(x)=5237804)
478
+ 456 train 7.236218 (lr=2.6396e-04) (hash(x)=6540397)
479
+ 457 train 7.494608 (lr=2.6356e-04) (hash(x)=5216638)
480
+ 458 train 7.290399 (lr=2.6315e-04) (hash(x)=5750537)
481
+ 459 train 7.304202 (lr=2.6274e-04) (hash(x)=6735565)
482
+ 460 train 7.309003 (lr=2.6233e-04) (hash(x)=7297488)
483
+ 461 train 7.513011 (lr=2.6192e-04) (hash(x)=5983575)
484
+ 462 train 7.319656 (lr=2.6150e-04) (hash(x)=4808543)
485
+ 463 train 7.437807 (lr=2.6109e-04) (hash(x)=5831619)
486
+ 464 train 7.203455 (lr=2.6067e-04) (hash(x)=6280052)
487
+ 465 train 7.231053 (lr=2.6025e-04) (hash(x)=9109083)
488
+ 466 train 7.455640 (lr=2.5983e-04) (hash(x)=5848604)
489
+ 467 train 7.395865 (lr=2.5940e-04) (hash(x)=6912967)
490
+ 468 train 7.594830 (lr=2.5898e-04) (hash(x)=6615365)
491
+ 469 train 7.645655 (lr=2.5855e-04) (hash(x)=5965674)
492
+ 470 train 7.483243 (lr=2.5812e-04) (hash(x)=6879965)
493
+ 471 train 7.415557 (lr=2.5769e-04) (hash(x)=5822533)
494
+ 472 train 7.302964 (lr=2.5726e-04) (hash(x)=5794462)
495
+ 473 train 7.375607 (lr=2.5682e-04) (hash(x)=5268216)
496
+ 474 train 7.196211 (lr=2.5639e-04) (hash(x)=7058396)
497
+ 475 train 7.300541 (lr=2.5595e-04) (hash(x)=6531900)
498
+ 476 train 7.322757 (lr=2.5551e-04) (hash(x)=5878912)
499
+ 477 train 7.123130 (lr=2.5507e-04) (hash(x)=5426499)
500
+ 478 train 7.342466 (lr=2.5462e-04) (hash(x)=8107004)
501
+ 479 train 8.086945 (lr=2.5418e-04) (hash(x)=7455372)
502
+ 480 train 8.017781 (lr=2.5373e-04) (hash(x)=7727959)
503
+ 481 train 7.947183 (lr=2.5328e-04) (hash(x)=8319119)
504
+ 482 train 7.871157 (lr=2.5283e-04) (hash(x)=7581656)
505
+ 483 train 7.946103 (lr=2.5238e-04) (hash(x)=7592128)
506
+ 484 train 7.910131 (lr=2.5193e-04) (hash(x)=10506633)
507
+ 485 train 7.861582 (lr=2.5147e-04) (hash(x)=8704170)
508
+ 486 train 7.645012 (lr=2.5102e-04) (hash(x)=7461138)
509
+ 487 train 7.525881 (lr=2.5056e-04) (hash(x)=7165037)
510
+ 488 train 7.603084 (lr=2.5010e-04) (hash(x)=4976998)
511
+ 489 train 7.300906 (lr=2.4963e-04) (hash(x)=7678000)
512
+ 490 train 7.346937 (lr=2.4917e-04) (hash(x)=6299409)
513
+ 491 train 7.299150 (lr=2.4871e-04) (hash(x)=7414788)
514
+ 492 train 7.416858 (lr=2.4824e-04) (hash(x)=5063061)
515
+ 493 train 7.330403 (lr=2.4777e-04) (hash(x)=7326216)
516
+ 494 train 7.479515 (lr=2.4730e-04) (hash(x)=7492012)
517
+ 495 train 7.435187 (lr=2.4683e-04) (hash(x)=5857162)
518
+ 496 train 7.226203 (lr=2.4636e-04) (hash(x)=6169001)
519
+ 497 train 9.268858 (lr=2.4588e-04) (hash(x)=14715669)
520
+ 498 train 12.240268 (lr=2.4540e-04) (hash(x)=18256706)
521
+ 499 train 12.077112 (lr=2.4493e-04) (hash(x)=16950560)
522
+ 500 val loss 7.3516
523
+ 500 val perplexity 1558.6382
524
+ 500 train 8.128902 (lr=2.4445e-04) (hash(x)=6113563)
525
+ 501 train 6.989974 (lr=2.4397e-04) (hash(x)=6159125)
526
+ 502 train 7.395797 (lr=2.4348e-04) (hash(x)=5482303)
527
+ 503 train 7.432923 (lr=2.4300e-04) (hash(x)=5908442)
528
+ 504 train 7.303898 (lr=2.4251e-04) (hash(x)=4985045)
529
+ 505 train 7.208208 (lr=2.4203e-04) (hash(x)=6668887)
530
+ 506 train 7.285321 (lr=2.4154e-04) (hash(x)=6123910)
531
+ 507 train 7.364649 (lr=2.4105e-04) (hash(x)=7662975)
532
+ 508 train 7.308952 (lr=2.4056e-04) (hash(x)=7191898)
533
+ 509 train 7.257383 (lr=2.4006e-04) (hash(x)=6669539)
534
+ 510 train 7.339392 (lr=2.3957e-04) (hash(x)=6557481)
535
+ 511 train 7.225530 (lr=2.3907e-04) (hash(x)=5015998)
536
+ 512 train 7.287534 (lr=2.3858e-04) (hash(x)=4964540)
537
+ 513 train 7.099957 (lr=2.3808e-04) (hash(x)=5458386)
538
+ 514 train 7.497488 (lr=2.3758e-04) (hash(x)=6077616)
539
+ 515 train 7.228012 (lr=2.3708e-04) (hash(x)=4849847)
540
+ 516 train 7.290946 (lr=2.3657e-04) (hash(x)=5431422)
541
+ 517 train 7.480175 (lr=2.3607e-04) (hash(x)=5125562)
542
+ 518 train 7.397559 (lr=2.3556e-04) (hash(x)=5587715)
543
+ 519 train 7.438240 (lr=2.3506e-04) (hash(x)=6012047)
544
+ 520 train 7.292498 (lr=2.3455e-04) (hash(x)=5668436)
545
+ 521 train 7.303680 (lr=2.3404e-04) (hash(x)=6079364)
546
+ 522 train 7.433142 (lr=2.3353e-04) (hash(x)=5200504)
547
+ 523 train 7.167505 (lr=2.3301e-04) (hash(x)=6030395)
548
+ 524 train 7.412891 (lr=2.3250e-04) (hash(x)=6855319)
549
+ 525 train 7.187482 (lr=2.3198e-04) (hash(x)=6781762)
550
+ 526 train 7.327936 (lr=2.3147e-04) (hash(x)=6717939)
551
+ 527 train 7.325736 (lr=2.3095e-04) (hash(x)=5238325)
552
+ 528 train 7.225799 (lr=2.3043e-04) (hash(x)=5808069)
553
+ 529 train 7.572337 (lr=2.2991e-04) (hash(x)=8611144)
554
+ 530 train 7.227248 (lr=2.2939e-04) (hash(x)=6028738)
555
+ 531 train 7.261206 (lr=2.2887e-04) (hash(x)=6901605)
556
+ 532 train 7.188833 (lr=2.2834e-04) (hash(x)=6723307)
557
+ 533 train 7.325653 (lr=2.2782e-04) (hash(x)=5499337)
558
+ 534 train 7.233948 (lr=2.2729e-04) (hash(x)=6586372)
559
+ 535 train 7.370315 (lr=2.2676e-04) (hash(x)=7134874)
560
+ 536 train 7.252157 (lr=2.2624e-04) (hash(x)=8415623)
561
+ 537 train 7.281092 (lr=2.2571e-04) (hash(x)=6149038)
562
+ 538 train 7.325523 (lr=2.2517e-04) (hash(x)=6078325)
563
+ 539 train 7.305715 (lr=2.2464e-04) (hash(x)=6609760)
564
+ 540 train 7.189839 (lr=2.2411e-04) (hash(x)=6185253)
565
+ 541 train 7.542022 (lr=2.2357e-04) (hash(x)=6652627)
566
+ 542 train 7.378432 (lr=2.2304e-04) (hash(x)=7758536)
567
+ 543 train 7.324357 (lr=2.2250e-04) (hash(x)=8618579)
568
+ 544 train 7.313974 (lr=2.2196e-04) (hash(x)=6710226)
569
+ 545 train 7.147356 (lr=2.2142e-04) (hash(x)=5896892)
570
+ 546 train 7.178617 (lr=2.2088e-04) (hash(x)=6332995)
571
+ 547 train 7.619903 (lr=2.2034e-04) (hash(x)=8945854)
572
+ 548 train 7.155745 (lr=2.1980e-04) (hash(x)=4329129)
573
+ 549 train 7.212925 (lr=2.1926e-04) (hash(x)=6775103)
574
+ 550 val loss 7.2475
575
+ 550 val perplexity 1404.6465
576
+ 550 train 7.341378 (lr=2.1871e-04) (hash(x)=4465735)
577
+ 551 train 7.187572 (lr=2.1817e-04) (hash(x)=6200024)
578
+ 552 train 7.155844 (lr=2.1762e-04) (hash(x)=5520748)
579
+ 553 train 7.144910 (lr=2.1707e-04) (hash(x)=5238139)
580
+ 554 train 7.433198 (lr=2.1653e-04) (hash(x)=7639817)
581
+ 555 train 7.204960 (lr=2.1598e-04) (hash(x)=5780948)
582
+ 556 train 7.324251 (lr=2.1542e-04) (hash(x)=6733444)
583
+ 557 train 7.262815 (lr=2.1487e-04) (hash(x)=6168407)
584
+ 558 train 7.286927 (lr=2.1432e-04) (hash(x)=6680197)
585
+ 559 train 7.145280 (lr=2.1377e-04) (hash(x)=3775881)
586
+ 560 train 7.001796 (lr=2.1321e-04) (hash(x)=5784096)
587
+ 561 train 7.483177 (lr=2.1266e-04) (hash(x)=6848709)
588
+ 562 train 7.577888 (lr=2.1210e-04) (hash(x)=7449019)
589
+ 563 train 7.628167 (lr=2.1154e-04) (hash(x)=6478113)
590
+ 564 train 7.252421 (lr=2.1099e-04) (hash(x)=5925744)
591
+ 565 train 7.296356 (lr=2.1043e-04) (hash(x)=6483914)
592
+ 566 train 7.164549 (lr=2.0987e-04) (hash(x)=6010847)
593
+ 567 train 7.279039 (lr=2.0931e-04) (hash(x)=5927795)
594
+ 568 train 7.289657 (lr=2.0875e-04) (hash(x)=5896332)
595
+ 569 train 7.230428 (lr=2.0818e-04) (hash(x)=8438033)
596
+ 570 train 7.351763 (lr=2.0762e-04) (hash(x)=6081660)
597
+ 571 train 7.240448 (lr=2.0706e-04) (hash(x)=5684788)
598
+ 572 train 7.130642 (lr=2.0649e-04) (hash(x)=6272055)
599
+ 573 train 7.084376 (lr=2.0593e-04) (hash(x)=5710182)
600
+ 574 train 7.211885 (lr=2.0536e-04) (hash(x)=6219616)
601
+ 575 train 7.176457 (lr=2.0479e-04) (hash(x)=6160063)
602
+ 576 train 7.205252 (lr=2.0422e-04) (hash(x)=5710809)
603
+ 577 train 6.952169 (lr=2.0366e-04) (hash(x)=5955100)
604
+ 578 train 6.968685 (lr=2.0309e-04) (hash(x)=5356088)
605
+ 579 train 7.005916 (lr=2.0252e-04) (hash(x)=6672309)
606
+ 580 train 7.219397 (lr=2.0194e-04) (hash(x)=5841684)
607
+ 581 train 7.580019 (lr=2.0137e-04) (hash(x)=9001564)
608
+ 582 train 8.151119 (lr=2.0080e-04) (hash(x)=6897178)
609
+ 583 train 7.315060 (lr=2.0023e-04) (hash(x)=5454654)
610
+ 584 train 7.071740 (lr=1.9965e-04) (hash(x)=6391454)
611
+ 585 train 7.186828 (lr=1.9908e-04) (hash(x)=5174062)
612
+ 586 train 7.315221 (lr=1.9850e-04) (hash(x)=5197924)
613
+ 587 train 7.254595 (lr=1.9793e-04) (hash(x)=6199203)
614
+ 588 train 7.069732 (lr=1.9735e-04) (hash(x)=5786498)
615
+ 589 train 7.222054 (lr=1.9677e-04) (hash(x)=8352747)
616
+ 590 train 7.510259 (lr=1.9620e-04) (hash(x)=4829700)
617
+ 591 train 6.993373 (lr=1.9562e-04) (hash(x)=4547106)
618
+ 592 train 7.312548 (lr=1.9504e-04) (hash(x)=7686831)
619
+ 593 train 7.175517 (lr=1.9446e-04) (hash(x)=6433099)
620
+ 594 train 7.025823 (lr=1.9388e-04) (hash(x)=4826460)
621
+ 595 train 7.127717 (lr=1.9330e-04) (hash(x)=6424428)
622
+ 596 train 7.020804 (lr=1.9272e-04) (hash(x)=6073605)
623
+ 597 train 7.370596 (lr=1.9214e-04) (hash(x)=5975712)
624
+ 598 train 6.970050 (lr=1.9156e-04) (hash(x)=7030029)
625
+ 599 train 7.296854 (lr=1.9097e-04) (hash(x)=5458532)
626
+ 600 val loss 7.2139
627
+ 600 val perplexity 1358.1952
628
+ 600 train 7.113532 (lr=1.9039e-04) (hash(x)=7823919)
629
+ 601 train 7.584157 (lr=1.8981e-04) (hash(x)=7002774)
630
+ 602 train 7.373654 (lr=1.8922e-04) (hash(x)=7056399)
631
+ 603 train 7.591238 (lr=1.8864e-04) (hash(x)=8206173)
632
+ 604 train 7.436531 (lr=1.8805e-04) (hash(x)=8034846)
633
+ 605 train 7.742330 (lr=1.8747e-04) (hash(x)=7784971)
634
+ 606 train 7.885799 (lr=1.8688e-04) (hash(x)=7118076)
635
+ 607 train 7.150384 (lr=1.8629e-04) (hash(x)=6001843)
636
+ 608 train 7.233588 (lr=1.8571e-04) (hash(x)=5582437)
637
+ 609 train 7.366101 (lr=1.8512e-04) (hash(x)=5698642)
638
+ 610 train 7.208152 (lr=1.8453e-04) (hash(x)=6615301)
639
+ 611 train 7.201260 (lr=1.8395e-04) (hash(x)=6600751)
640
+ 612 train 7.128580 (lr=1.8336e-04) (hash(x)=6880955)
641
+ 613 train 7.173840 (lr=1.8277e-04) (hash(x)=5804610)
642
+ 614 train 7.100860 (lr=1.8218e-04) (hash(x)=6263347)
643
+ 615 train 7.048603 (lr=1.8159e-04) (hash(x)=5162652)
644
+ 616 train 7.115860 (lr=1.8100e-04) (hash(x)=6055710)
645
+ 617 train 7.079264 (lr=1.8041e-04) (hash(x)=5297006)
646
+ 618 train 7.107976 (lr=1.7982e-04) (hash(x)=5923648)
647
+ 619 train 7.110251 (lr=1.7923e-04) (hash(x)=6728323)
648
+ 620 train 7.127951 (lr=1.7864e-04) (hash(x)=6660239)
649
+ 621 train 7.094420 (lr=1.7805e-04) (hash(x)=3964882)
650
+ 622 train 7.150954 (lr=1.7746e-04) (hash(x)=5868942)
651
+ 623 train 7.065094 (lr=1.7686e-04) (hash(x)=6735993)
652
+ 624 train 7.429172 (lr=1.7627e-04) (hash(x)=5926190)
653
+ 625 train 7.075055 (lr=1.7568e-04) (hash(x)=5150316)
654
+ 626 train 7.099816 (lr=1.7509e-04) (hash(x)=7571467)
655
+ 627 train 7.014606 (lr=1.7450e-04) (hash(x)=5220448)
656
+ 628 train 7.017559 (lr=1.7390e-04) (hash(x)=6790397)
657
+ 629 train 7.507487 (lr=1.7331e-04) (hash(x)=6971289)
658
+ 630 train 7.456183 (lr=1.7272e-04) (hash(x)=7123359)
659
+ 631 train 7.259305 (lr=1.7212e-04) (hash(x)=5012069)
660
+ 632 train 7.207749 (lr=1.7153e-04) (hash(x)=6183574)
661
+ 633 train 7.101263 (lr=1.7094e-04) (hash(x)=6576136)
662
+ 634 train 7.125478 (lr=1.7034e-04) (hash(x)=6157346)
663
+ 635 train 7.104994 (lr=1.6975e-04) (hash(x)=6077647)
664
+ 636 train 7.358856 (lr=1.6916e-04) (hash(x)=8369488)
665
+ 637 train 7.150980 (lr=1.6856e-04) (hash(x)=6032381)
666
+ 638 train 7.062288 (lr=1.6797e-04) (hash(x)=5924756)
667
+ 639 train 7.108144 (lr=1.6738e-04) (hash(x)=8410716)
668
+ 640 train 7.093020 (lr=1.6678e-04) (hash(x)=5803440)
669
+ 641 train 7.180191 (lr=1.6619e-04) (hash(x)=6212516)
670
+ 642 train 7.269322 (lr=1.6559e-04) (hash(x)=7165945)
671
+ 643 train 6.970079 (lr=1.6500e-04) (hash(x)=5298545)
672
+ 644 train 7.252603 (lr=1.6441e-04) (hash(x)=6281739)
673
+ 645 train 7.181458 (lr=1.6381e-04) (hash(x)=5779073)
674
+ 646 train 7.066733 (lr=1.6322e-04) (hash(x)=5090107)
675
+ 647 train 7.125894 (lr=1.6262e-04) (hash(x)=6340947)
676
+ 648 train 6.864412 (lr=1.6203e-04) (hash(x)=4741418)
677
+ 649 train 6.780537 (lr=1.6144e-04) (hash(x)=6530441)
678
+ 650 val loss 7.1833
679
+ 650 val perplexity 1317.2255
680
+ 650 train 7.356423 (lr=1.6084e-04) (hash(x)=6396041)
681
+ 651 train 7.016239 (lr=1.6025e-04) (hash(x)=7818007)
682
+ 652 train 7.023707 (lr=1.5966e-04) (hash(x)=6478457)
683
+ 653 train 7.085379 (lr=1.5906e-04) (hash(x)=6277435)
684
+ 654 train 7.056863 (lr=1.5847e-04) (hash(x)=6042369)
685
+ 655 train 7.080772 (lr=1.5788e-04) (hash(x)=6226320)
686
+ 656 train 7.109776 (lr=1.5728e-04) (hash(x)=5826883)
687
+ 657 train 7.208757 (lr=1.5669e-04) (hash(x)=5463751)
688
+ 658 train 7.126280 (lr=1.5610e-04) (hash(x)=6433080)
689
+ 659 train 6.968693 (lr=1.5550e-04) (hash(x)=6579329)
690
+ 660 train 7.001870 (lr=1.5491e-04) (hash(x)=4724703)
691
+ 661 train 7.347678 (lr=1.5432e-04) (hash(x)=6660540)
692
+ 662 train 7.155168 (lr=1.5373e-04) (hash(x)=5918756)
693
+ 663 train 7.086496 (lr=1.5314e-04) (hash(x)=6491272)
694
+ 664 train 7.009777 (lr=1.5254e-04) (hash(x)=5452863)
695
+ 665 train 7.205496 (lr=1.5195e-04) (hash(x)=6766947)
696
+ 666 train 7.583355 (lr=1.5136e-04) (hash(x)=7459504)
697
+ 667 train 7.211386 (lr=1.5077e-04) (hash(x)=3721218)
698
+ 668 train 7.410998 (lr=1.5018e-04) (hash(x)=7929571)
699
+ 669 train 7.061044 (lr=1.4959e-04) (hash(x)=5229898)
700
+ 670 train 7.109720 (lr=1.4900e-04) (hash(x)=6083090)
701
+ 671 train 6.962116 (lr=1.4841e-04) (hash(x)=6054428)
702
+ 672 train 7.128041 (lr=1.4782e-04) (hash(x)=6072563)
703
+ 673 train 7.216414 (lr=1.4723e-04) (hash(x)=6290464)
704
+ 674 train 7.338971 (lr=1.4664e-04) (hash(x)=6314742)
705
+ 675 train 6.990757 (lr=1.4605e-04) (hash(x)=6904498)
706
+ 676 train 7.152540 (lr=1.4547e-04) (hash(x)=5378107)
707
+ 677 train 7.267296 (lr=1.4488e-04) (hash(x)=7765805)
708
+ 678 train 7.088050 (lr=1.4429e-04) (hash(x)=5871280)
709
+ 679 train 7.234810 (lr=1.4371e-04) (hash(x)=6543764)
710
+ 680 train 7.049421 (lr=1.4312e-04) (hash(x)=7369665)
711
+ 681 train 7.182322 (lr=1.4253e-04) (hash(x)=7204955)
712
+ 682 train 7.126574 (lr=1.4195e-04) (hash(x)=5856543)
713
+ 683 train 7.001845 (lr=1.4136e-04) (hash(x)=4423711)
714
+ 684 train 6.982417 (lr=1.4078e-04) (hash(x)=5818153)
715
+ 685 train 7.078627 (lr=1.4019e-04) (hash(x)=5343477)
716
+ 686 train 7.281490 (lr=1.3961e-04) (hash(x)=6581702)
717
+ 687 train 7.111911 (lr=1.3903e-04) (hash(x)=6595856)
718
+ 688 train 7.037158 (lr=1.3844e-04) (hash(x)=5325378)
719
+ 689 train 6.976363 (lr=1.3786e-04) (hash(x)=7229015)
720
+ 690 train 7.132633 (lr=1.3728e-04) (hash(x)=5918186)
721
+ 691 train 6.997124 (lr=1.3670e-04) (hash(x)=5916034)
722
+ 692 train 7.028913 (lr=1.3612e-04) (hash(x)=5280722)
723
+ 693 train 7.399898 (lr=1.3554e-04) (hash(x)=6939521)
724
+ 694 train 7.154862 (lr=1.3496e-04) (hash(x)=5172566)
725
+ 695 train 7.014957 (lr=1.3438e-04) (hash(x)=5492185)
726
+ 696 train 6.883283 (lr=1.3380e-04) (hash(x)=5779011)
727
+ 697 train 7.163246 (lr=1.3323e-04) (hash(x)=6898006)
728
+ 698 train 6.989983 (lr=1.3265e-04) (hash(x)=5779217)
729
+ 699 train 6.862276 (lr=1.3207e-04) (hash(x)=5462008)
730
+ 700 val loss 7.1559
731
+ 700 val perplexity 1281.7029
732
+ 700 train 7.013890 (lr=1.3150e-04) (hash(x)=6243843)
733
+ 701 train 7.063276 (lr=1.3092e-04) (hash(x)=7821902)
734
+ 702 train 7.140563 (lr=1.3035e-04) (hash(x)=5222501)
735
+ 703 train 7.081514 (lr=1.2977e-04) (hash(x)=6608108)
736
+ 704 train 7.203692 (lr=1.2920e-04) (hash(x)=7602879)
737
+ 705 train 7.120367 (lr=1.2863e-04) (hash(x)=6295394)
738
+ 706 train 7.073847 (lr=1.2806e-04) (hash(x)=6579907)
739
+ 707 train 6.963603 (lr=1.2748e-04) (hash(x)=6803449)
740
+ 708 train 7.034765 (lr=1.2691e-04) (hash(x)=5413733)
741
+ 709 train 6.997452 (lr=1.2634e-04) (hash(x)=7444134)
742
+ 710 train 7.158776 (lr=1.2578e-04) (hash(x)=7028652)
743
+ 711 train 6.956982 (lr=1.2521e-04) (hash(x)=5361923)
744
+ 712 train 7.058182 (lr=1.2464e-04) (hash(x)=5478293)
745
+ 713 train 7.067133 (lr=1.2407e-04) (hash(x)=7803039)
746
+ 714 train 7.253763 (lr=1.2351e-04) (hash(x)=6469017)
747
+ 715 train 7.142198 (lr=1.2294e-04) (hash(x)=7042578)
748
+ 716 train 7.190039 (lr=1.2238e-04) (hash(x)=6927957)
749
+ 717 train 7.000715 (lr=1.2182e-04) (hash(x)=6647007)
750
+ 718 train 6.964838 (lr=1.2125e-04) (hash(x)=6230888)
751
+ 719 train 7.253665 (lr=1.2069e-04) (hash(x)=6303658)
752
+ 720 train 7.289773 (lr=1.2013e-04) (hash(x)=6769880)
753
+ 721 train 7.118761 (lr=1.1957e-04) (hash(x)=6158927)
754
+ 722 train 7.097446 (lr=1.1901e-04) (hash(x)=5991631)
755
+ 723 train 6.957667 (lr=1.1846e-04) (hash(x)=6419656)
756
+ 724 train 6.951476 (lr=1.1790e-04) (hash(x)=7739209)
757
+ 725 train 6.918235 (lr=1.1734e-04) (hash(x)=7891271)
758
+ 726 train 7.288694 (lr=1.1679e-04) (hash(x)=7171433)
759
+ 727 train 7.059498 (lr=1.1623e-04) (hash(x)=7099888)
760
+ 728 train 7.055088 (lr=1.1568e-04) (hash(x)=6103730)
761
+ 729 train 7.151111 (lr=1.1513e-04) (hash(x)=6352424)
762
+ 730 train 7.070194 (lr=1.1458e-04) (hash(x)=4893671)
763
+ 731 train 7.028497 (lr=1.1402e-04) (hash(x)=5863145)
764
+ 732 train 7.105975 (lr=1.1347e-04) (hash(x)=5502233)
765
+ 733 train 7.043506 (lr=1.1293e-04) (hash(x)=6844597)
766
+ 734 train 7.047463 (lr=1.1238e-04) (hash(x)=6744526)
767
+ 735 train 7.198891 (lr=1.1183e-04) (hash(x)=6009861)
768
+ 736 train 7.050950 (lr=1.1129e-04) (hash(x)=6152358)
769
+ 737 train 6.983796 (lr=1.1074e-04) (hash(x)=5374957)
770
+ 738 train 7.015686 (lr=1.1020e-04) (hash(x)=6495309)
771
+ 739 train 7.159252 (lr=1.0966e-04) (hash(x)=3923225)
772
+ 740 train 6.932333 (lr=1.0912e-04) (hash(x)=6255243)
773
+ 741 train 7.086520 (lr=1.0858e-04) (hash(x)=5848862)
774
+ 742 train 7.109909 (lr=1.0804e-04) (hash(x)=4535000)
775
+ 743 train 6.965602 (lr=1.0750e-04) (hash(x)=4440634)
776
+ 744 train 6.994028 (lr=1.0696e-04) (hash(x)=5222077)
777
+ 745 train 7.018528 (lr=1.0643e-04) (hash(x)=4788998)
778
+ 746 train 7.019298 (lr=1.0589e-04) (hash(x)=6108350)
779
+ 747 train 6.942775 (lr=1.0536e-04) (hash(x)=7378165)
780
+ 748 train 7.009796 (lr=1.0483e-04) (hash(x)=5475646)
781
+ 749 train 6.998092 (lr=1.0429e-04) (hash(x)=4771243)
782
+ 750 val loss 7.1336
783
+ 750 val perplexity 1253.3616
784
+ 750 train 6.916419 (lr=1.0376e-04) (hash(x)=6668088)
785
+ 751 train 7.119673 (lr=1.0324e-04) (hash(x)=5716089)
786
+ 752 train 7.083768 (lr=1.0271e-04) (hash(x)=6991946)
787
+ 753 train 6.996638 (lr=1.0218e-04) (hash(x)=7991009)
788
+ 754 train 7.081497 (lr=1.0166e-04) (hash(x)=6037018)
789
+ 755 train 7.033475 (lr=1.0113e-04) (hash(x)=5348607)
790
+ 756 train 7.196137 (lr=1.0061e-04) (hash(x)=5190526)
791
+ 757 train 7.071329 (lr=1.0009e-04) (hash(x)=6036146)
792
+ 758 train 6.962649 (lr=9.9568e-05) (hash(x)=5329783)
793
+ 759 train 7.065001 (lr=9.9049e-05) (hash(x)=6658718)
794
+ 760 train 6.959289 (lr=9.8531e-05) (hash(x)=4820939)
795
+ 761 train 6.968325 (lr=9.8015e-05) (hash(x)=6176253)
796
+ 762 train 6.914993 (lr=9.7500e-05) (hash(x)=5316750)
797
+ 763 train 7.020503 (lr=9.6986e-05) (hash(x)=4542895)
798
+ 764 train 6.790451 (lr=9.6474e-05) (hash(x)=5395673)
799
+ 765 train 6.895455 (lr=9.5963e-05) (hash(x)=5549073)
800
+ 766 train 7.168859 (lr=9.5453e-05) (hash(x)=5233723)
801
+ 767 train 7.232811 (lr=9.4944e-05) (hash(x)=5342783)
802
+ 768 train 7.132057 (lr=9.4437e-05) (hash(x)=5552448)
803
+ 769 train 7.291832 (lr=9.3932e-05) (hash(x)=6549546)
804
+ 770 train 6.929078 (lr=9.3427e-05) (hash(x)=3816493)
805
+ 771 train 6.974104 (lr=9.2924e-05) (hash(x)=6170599)
806
+ 772 train 7.482815 (lr=9.2423e-05) (hash(x)=6708759)
807
+ 773 train 7.105374 (lr=9.1923e-05) (hash(x)=8483134)
808
+ 774 train 7.691579 (lr=9.1424e-05) (hash(x)=5583732)
809
+ 775 train 7.103418 (lr=9.0927e-05) (hash(x)=4865036)
810
+ 776 train 6.977034 (lr=9.0431e-05) (hash(x)=7234336)
811
+ 777 train 7.055741 (lr=8.9936e-05) (hash(x)=4433890)
812
+ 778 train 7.375794 (lr=8.9443e-05) (hash(x)=7484636)
813
+ 779 train 7.793103 (lr=8.8952e-05) (hash(x)=6738597)
814
+ 780 train 7.050267 (lr=8.8462e-05) (hash(x)=6752421)
815
+ 781 train 7.253078 (lr=8.7973e-05) (hash(x)=5459620)
816
+ 782 train 7.148220 (lr=8.7486e-05) (hash(x)=6718131)
817
+ 783 train 7.123672 (lr=8.7001e-05) (hash(x)=4604074)
818
+ 784 train 7.005103 (lr=8.6516e-05) (hash(x)=5946124)
819
+ 785 train 7.286037 (lr=8.6034e-05) (hash(x)=8045501)
820
+ 786 train 7.371063 (lr=8.5553e-05) (hash(x)=5712725)
821
+ 787 train 7.108195 (lr=8.5073e-05) (hash(x)=6176464)
822
+ 788 train 7.148711 (lr=8.4596e-05) (hash(x)=6276702)
823
+ 789 train 6.951337 (lr=8.4119e-05) (hash(x)=6127317)
824
+ 790 train 7.103370 (lr=8.3644e-05) (hash(x)=7221374)
825
+ 791 train 7.074277 (lr=8.3171e-05) (hash(x)=5571462)
826
+ 792 train 7.104302 (lr=8.2699e-05) (hash(x)=7518021)
827
+ 793 train 7.083576 (lr=8.2229e-05) (hash(x)=6640755)
828
+ 794 train 7.149507 (lr=8.1761e-05) (hash(x)=6870405)
829
+ 795 train 7.151824 (lr=8.1294e-05) (hash(x)=6147295)
830
+ 796 train 7.017288 (lr=8.0829e-05) (hash(x)=6782583)
831
+ 797 train 7.157561 (lr=8.0365e-05) (hash(x)=6120461)
832
+ 798 train 7.093185 (lr=7.9903e-05) (hash(x)=5893870)
833
+ 799 train 7.023798 (lr=7.9443e-05) (hash(x)=5123928)
834
+ 800 val loss 7.1139
835
+ 800 val perplexity 1228.9559
836
+ 800 train 7.098544 (lr=7.8984e-05) (hash(x)=7037251)
837
+ 801 train 7.080308 (lr=7.8527e-05) (hash(x)=5428813)
838
+ 802 train 7.016583 (lr=7.8072e-05) (hash(x)=5550387)
839
+ 803 train 7.083931 (lr=7.7618e-05) (hash(x)=6448586)
840
+ 804 train 7.228072 (lr=7.7167e-05) (hash(x)=5916934)
841
+ 805 train 7.203781 (lr=7.6716e-05) (hash(x)=8116987)
842
+ 806 train 7.179995 (lr=7.6268e-05) (hash(x)=6210899)
843
+ 807 train 7.299696 (lr=7.5821e-05) (hash(x)=5974342)
844
+ 808 train 7.275043 (lr=7.5376e-05) (hash(x)=7236199)
845
+ 809 train 7.205301 (lr=7.4933e-05) (hash(x)=5261336)
846
+ 810 train 6.991825 (lr=7.4491e-05) (hash(x)=6310720)
847
+ 811 train 7.071819 (lr=7.4051e-05) (hash(x)=5307056)
848
+ 812 train 6.893535 (lr=7.3613e-05) (hash(x)=7356633)
849
+ 813 train 7.015359 (lr=7.3177e-05) (hash(x)=5865119)
850
+ 814 train 6.909586 (lr=7.2742e-05) (hash(x)=5979205)
851
+ 815 train 7.088969 (lr=7.2309e-05) (hash(x)=6672380)
852
+ 816 train 7.068211 (lr=7.1878e-05) (hash(x)=5565030)
853
+ 817 train 7.038513 (lr=7.1449e-05) (hash(x)=6886810)
854
+ 818 train 7.096722 (lr=7.1022e-05) (hash(x)=5926478)
855
+ 819 train 7.320323 (lr=7.0596e-05) (hash(x)=6564071)
856
+ 820 train 6.925412 (lr=7.0173e-05) (hash(x)=5024280)
857
+ 821 train 6.918086 (lr=6.9751e-05) (hash(x)=7705089)
858
+ 822 train 6.894872 (lr=6.9331e-05) (hash(x)=4037371)
859
+ 823 train 7.012532 (lr=6.8913e-05) (hash(x)=5773073)
860
+ 824 train 7.024636 (lr=6.8496e-05) (hash(x)=7158139)
861
+ 825 train 6.872837 (lr=6.8082e-05) (hash(x)=5905976)
862
+ 826 train 7.320297 (lr=6.7669e-05) (hash(x)=13897369)
863
+ 827 train 8.481186 (lr=6.7259e-05) (hash(x)=6936653)
864
+ 828 train 7.058599 (lr=6.6850e-05) (hash(x)=6592246)
865
+ 829 train 7.036511 (lr=6.6443e-05) (hash(x)=6592773)
866
+ 830 train 7.088200 (lr=6.6038e-05) (hash(x)=5461178)
867
+ 831 train 6.859075 (lr=6.5635e-05) (hash(x)=6796551)
868
+ 832 train 7.061387 (lr=6.5234e-05) (hash(x)=4721691)
869
+ 833 train 6.989285 (lr=6.4835e-05) (hash(x)=6151589)
870
+ 834 train 7.140241 (lr=6.4437e-05) (hash(x)=7008995)
871
+ 835 train 7.488571 (lr=6.4042e-05) (hash(x)=7045164)
872
+ 836 train 7.057011 (lr=6.3649e-05) (hash(x)=6614221)
873
+ 837 train 7.090332 (lr=6.3257e-05) (hash(x)=6650506)
874
+ 838 train 6.909191 (lr=6.2868e-05) (hash(x)=6363762)
875
+ 839 train 7.083407 (lr=6.2480e-05) (hash(x)=7839748)
876
+ 840 train 7.006419 (lr=6.2095e-05) (hash(x)=5124077)
877
+ 841 train 7.444353 (lr=6.1711e-05) (hash(x)=6028671)
878
+ 842 train 6.950215 (lr=6.1330e-05) (hash(x)=5427175)
879
+ 843 train 7.279261 (lr=6.0950e-05) (hash(x)=6118408)
880
+ 844 train 7.030025 (lr=6.0573e-05) (hash(x)=6022552)
881
+ 845 train 7.061387 (lr=6.0198e-05) (hash(x)=5757526)
882
+ 846 train 7.530303 (lr=5.9824e-05) (hash(x)=6476549)
883
+ 847 train 6.826754 (lr=5.9453e-05) (hash(x)=6030172)
884
+ 848 train 7.233544 (lr=5.9083e-05) (hash(x)=5917229)
885
+ 849 train 7.285985 (lr=5.8716e-05) (hash(x)=6719882)
886
+ 850 val loss 7.1075
887
+ 850 val perplexity 1221.0654
888
+ 850 train 7.159894 (lr=5.8351e-05) (hash(x)=5441498)
889
+ 851 train 7.011475 (lr=5.7988e-05) (hash(x)=5028050)
890
+ 852 train 7.074201 (lr=5.7627e-05) (hash(x)=8487491)
891
+ 853 train 7.055421 (lr=5.7268e-05) (hash(x)=5682910)
892
+ 854 train 7.239692 (lr=5.6911e-05) (hash(x)=6781362)
893
+ 855 train 7.075415 (lr=5.6556e-05) (hash(x)=5747891)
894
+ 856 train 7.093535 (lr=5.6203e-05) (hash(x)=7627661)
895
+ 857 train 6.758383 (lr=5.5853e-05) (hash(x)=5734208)
896
+ 858 train 6.931219 (lr=5.5504e-05) (hash(x)=6162015)
897
+ 859 train 7.143523 (lr=5.5158e-05) (hash(x)=7024381)
898
+ 860 train 7.368166 (lr=5.4813e-05) (hash(x)=5198498)
899
+ 861 train 7.350578 (lr=5.4471e-05) (hash(x)=7814361)
900
+ 862 train 7.210496 (lr=5.4131e-05) (hash(x)=7394995)
901
+ 863 train 7.083054 (lr=5.3793e-05) (hash(x)=5765893)
902
+ 864 train 6.895495 (lr=5.3458e-05) (hash(x)=4990435)
903
+ 865 train 7.040892 (lr=5.3124e-05) (hash(x)=5572048)
904
+ 866 train 7.185047 (lr=5.2793e-05) (hash(x)=7093293)
905
+ 867 train 7.055923 (lr=5.2464e-05) (hash(x)=5610058)
906
+ 868 train 6.968985 (lr=5.2137e-05) (hash(x)=6771468)
907
+ 869 train 7.006217 (lr=5.1812e-05) (hash(x)=4459483)
908
+ 870 train 6.889050 (lr=5.1489e-05) (hash(x)=4926546)
909
+ 871 train 6.984321 (lr=5.1169e-05) (hash(x)=6758411)
910
+ 872 train 6.988075 (lr=5.0851e-05) (hash(x)=7394277)
911
+ 873 train 7.100116 (lr=5.0534e-05) (hash(x)=6701381)
912
+ 874 train 7.457492 (lr=5.0221e-05) (hash(x)=8120966)
913
+ 875 train 6.954372 (lr=4.9909e-05) (hash(x)=5416400)
914
+ 876 train 6.964521 (lr=4.9600e-05) (hash(x)=6292778)
915
+ 877 train 6.893290 (lr=4.9293e-05) (hash(x)=5919569)
916
+ 878 train 7.227458 (lr=4.8988e-05) (hash(x)=7561740)
917
+ 879 train 6.977690 (lr=4.8685e-05) (hash(x)=5282870)
918
+ 880 train 7.005644 (lr=4.8385e-05) (hash(x)=5473238)
919
+ 881 train 6.863694 (lr=4.8087e-05) (hash(x)=8072181)
920
+ 882 train 6.889382 (lr=4.7791e-05) (hash(x)=6757322)
921
+ 883 train 6.853914 (lr=4.7497e-05) (hash(x)=7996770)
922
+ 884 train 6.935385 (lr=4.7206e-05) (hash(x)=3181021)
923
+ 885 train 6.983203 (lr=4.6917e-05) (hash(x)=5093519)
924
+ 886 train 6.985656 (lr=4.6630e-05) (hash(x)=5391352)
925
+ 887 train 7.008960 (lr=4.6346e-05) (hash(x)=6270917)
926
+ 888 train 7.003362 (lr=4.6063e-05) (hash(x)=5818195)
927
+ 889 train 6.794293 (lr=4.5783e-05) (hash(x)=5852780)
928
+ 890 train 6.716489 (lr=4.5506e-05) (hash(x)=5777928)
929
+ 891 train 6.874502 (lr=4.5231e-05) (hash(x)=5327728)
930
+ 892 train 6.886909 (lr=4.4958e-05) (hash(x)=7651281)
931
+ 893 train 7.122107 (lr=4.4687e-05) (hash(x)=6260765)
932
+ 894 train 6.974381 (lr=4.4419e-05) (hash(x)=6265102)
933
+ 895 train 7.168931 (lr=4.4153e-05) (hash(x)=6212894)
934
+ 896 train 6.929608 (lr=4.3889e-05) (hash(x)=6901967)
935
+ 897 train 7.013860 (lr=4.3628e-05) (hash(x)=6724010)
936
+ 898 train 6.994364 (lr=4.3369e-05) (hash(x)=6299033)
937
+ 899 train 7.164696 (lr=4.3113e-05) (hash(x)=6059934)
938
+ 900 val loss 7.0987
939
+ 900 val perplexity 1210.4152
940
+ 900 train 6.866005 (lr=4.2858e-05) (hash(x)=6029696)
941
+ 901 train 6.771080 (lr=4.2607e-05) (hash(x)=5587382)
942
+ 902 train 6.993401 (lr=4.2357e-05) (hash(x)=5890076)
943
+ 903 train 6.900857 (lr=4.2110e-05) (hash(x)=7274108)
944
+ 904 train 7.014778 (lr=4.1865e-05) (hash(x)=7975528)
945
+ 905 train 7.065679 (lr=4.1623e-05) (hash(x)=6085396)
946
+ 906 train 6.939631 (lr=4.1383e-05) (hash(x)=6724496)
947
+ 907 train 6.947989 (lr=4.1146e-05) (hash(x)=4784666)
948
+ 908 train 6.936305 (lr=4.0910e-05) (hash(x)=7193845)
949
+ 909 train 7.110578 (lr=4.0678e-05) (hash(x)=6288638)
950
+ 910 train 7.079478 (lr=4.0447e-05) (hash(x)=5685285)
951
+ 911 train 7.062114 (lr=4.0220e-05) (hash(x)=7714288)
952
+ 912 train 7.160192 (lr=3.9994e-05) (hash(x)=5983679)
953
+ 913 train 7.086443 (lr=3.9771e-05) (hash(x)=5534748)
954
+ 914 train 7.026258 (lr=3.9550e-05) (hash(x)=6545471)
955
+ 915 train 7.018825 (lr=3.9332e-05) (hash(x)=6361270)
956
+ 916 train 7.007600 (lr=3.9116e-05) (hash(x)=6257744)
957
+ 917 train 7.016111 (lr=3.8903e-05) (hash(x)=4682956)
958
+ 918 train 6.915054 (lr=3.8692e-05) (hash(x)=4521756)
959
+ 919 train 7.064827 (lr=3.8484e-05) (hash(x)=7568935)
960
+ 920 train 6.982286 (lr=3.8277e-05) (hash(x)=7969353)
961
+ 921 train 7.176071 (lr=3.8074e-05) (hash(x)=7973311)
962
+ 922 train 6.797400 (lr=3.7873e-05) (hash(x)=4843937)
963
+ 923 train 6.982713 (lr=3.7674e-05) (hash(x)=6987352)
964
+ 924 train 7.153902 (lr=3.7478e-05) (hash(x)=7179418)
965
+ 925 train 6.755489 (lr=3.7284e-05) (hash(x)=5413904)
966
+ 926 train 7.205327 (lr=3.7093e-05) (hash(x)=6028105)
967
+ 927 train 7.112931 (lr=3.6904e-05) (hash(x)=6732924)
968
+ 928 train 6.974981 (lr=3.6718e-05) (hash(x)=4547651)
969
+ 929 train 7.139859 (lr=3.6534e-05) (hash(x)=6219754)
970
+ 930 train 6.898739 (lr=3.6353e-05) (hash(x)=7465186)
971
+ 931 train 7.050307 (lr=3.6174e-05) (hash(x)=6839070)
972
+ 932 train 6.920786 (lr=3.5998e-05) (hash(x)=5390324)
973
+ 933 train 6.899885 (lr=3.5824e-05) (hash(x)=6090120)
974
+ 934 train 7.566038 (lr=3.5653e-05) (hash(x)=7123181)
975
+ 935 train 7.222597 (lr=3.5484e-05) (hash(x)=7115176)
976
+ 936 train 7.202020 (lr=3.5317e-05) (hash(x)=7016990)
977
+ 937 train 6.988410 (lr=3.5154e-05) (hash(x)=4989299)
978
+ 938 train 6.994831 (lr=3.4992e-05) (hash(x)=6783311)
979
+ 939 train 6.971570 (lr=3.4833e-05) (hash(x)=7497590)
980
+ 940 train 7.215337 (lr=3.4677e-05) (hash(x)=6767514)
981
+ 941 train 6.955915 (lr=3.4523e-05) (hash(x)=3977557)
982
+ 942 train 7.104504 (lr=3.4372e-05) (hash(x)=4806492)
983
+ 943 train 6.988573 (lr=3.4224e-05) (hash(x)=5524165)
984
+ 944 train 7.041864 (lr=3.4077e-05) (hash(x)=4930752)
985
+ 945 train 7.007635 (lr=3.3934e-05) (hash(x)=5124936)
986
+ 946 train 6.897454 (lr=3.3793e-05) (hash(x)=6877351)
987
+ 947 train 7.186145 (lr=3.3654e-05) (hash(x)=5522119)
988
+ 948 train 6.987307 (lr=3.3518e-05) (hash(x)=6183512)
989
+ 949 train 6.911496 (lr=3.3385e-05) (hash(x)=5030865)
990
+ 950 val loss 7.0873
991
+ 950 val perplexity 1196.7218
992
+ 950 train 6.995289 (lr=3.3254e-05) (hash(x)=6619812)
993
+ 951 train 7.037693 (lr=3.3125e-05) (hash(x)=6411783)
994
+ 952 train 6.983159 (lr=3.3000e-05) (hash(x)=6748982)
995
+ 953 train 6.901580 (lr=3.2876e-05) (hash(x)=6581937)
996
+ 954 train 7.031114 (lr=3.2756e-05) (hash(x)=7881022)
997
+ 955 train 6.971148 (lr=3.2638e-05) (hash(x)=5116910)
998
+ 956 train 6.769077 (lr=3.2522e-05) (hash(x)=4880833)
999
+ 957 train 6.799681 (lr=3.2409e-05) (hash(x)=7353459)
1000
+ 958 train 6.878766 (lr=3.2299e-05) (hash(x)=4951302)
1001
+ 959 train 6.764055 (lr=3.2191e-05) (hash(x)=6344535)
1002
+ 960 train 7.426929 (lr=3.2085e-05) (hash(x)=7411763)
1003
+ 961 train 6.972095 (lr=3.1983e-05) (hash(x)=5840169)
1004
+ 962 train 7.080235 (lr=3.1883e-05) (hash(x)=5784788)
1005
+ 963 train 6.881763 (lr=3.1785e-05) (hash(x)=6754290)
1006
+ 964 train 7.177385 (lr=3.1690e-05) (hash(x)=6369963)
1007
+ 965 train 7.269762 (lr=3.1598e-05) (hash(x)=6484573)
1008
+ 966 train 6.816872 (lr=3.1508e-05) (hash(x)=6516338)
1009
+ 967 train 6.849596 (lr=3.1421e-05) (hash(x)=5260125)
1010
+ 968 train 6.992398 (lr=3.1336e-05) (hash(x)=5781624)
1011
+ 969 train 6.760798 (lr=3.1254e-05) (hash(x)=4881179)
1012
+ 970 train 6.919067 (lr=3.1174e-05) (hash(x)=6707821)
1013
+ 971 train 7.061013 (lr=3.1098e-05) (hash(x)=5001634)
1014
+ 972 train 6.917983 (lr=3.1023e-05) (hash(x)=4402949)
1015
+ 973 train 6.939075 (lr=3.0952e-05) (hash(x)=7424810)
1016
+ 974 train 6.821562 (lr=3.0882e-05) (hash(x)=4814803)
1017
+ 975 train 6.723653 (lr=3.0816e-05) (hash(x)=5176431)
1018
+ 976 train 6.927711 (lr=3.0752e-05) (hash(x)=5769393)
1019
+ 977 train 6.855418 (lr=3.0691e-05) (hash(x)=5421624)
1020
+ 978 train 7.100616 (lr=3.0632e-05) (hash(x)=5999773)
1021
+ 979 train 6.799164 (lr=3.0576e-05) (hash(x)=5674439)
1022
+ 980 train 6.891325 (lr=3.0522e-05) (hash(x)=6345351)
1023
+ 981 train 6.823787 (lr=3.0471e-05) (hash(x)=5819157)
1024
+ 982 train 7.023499 (lr=3.0423e-05) (hash(x)=11159355)
1025
+ 983 train 6.992321 (lr=3.0377e-05) (hash(x)=6506856)
1026
+ 984 train 6.912742 (lr=3.0334e-05) (hash(x)=6331023)
1027
+ 985 train 6.825050 (lr=3.0294e-05) (hash(x)=6119124)
1028
+ 986 train 7.028724 (lr=3.0256e-05) (hash(x)=8142528)
1029
+ 987 train 7.216858 (lr=3.0221e-05) (hash(x)=6241669)
1030
+ 988 train 6.827493 (lr=3.0188e-05) (hash(x)=3852991)
1031
+ 989 train 6.532195 (lr=3.0158e-05) (hash(x)=4992756)
1032
+ 990 train 7.306437 (lr=3.0131e-05) (hash(x)=7097145)
1033
+ 991 train 7.063352 (lr=3.0106e-05) (hash(x)=7277927)
1034
+ 992 train 7.173432 (lr=3.0084e-05) (hash(x)=5661400)
1035
+ 993 train 7.063525 (lr=3.0064e-05) (hash(x)=6622097)
1036
+ 994 train 7.200085 (lr=3.0047e-05) (hash(x)=5016804)
1037
+ 995 train 7.278684 (lr=3.0033e-05) (hash(x)=7026582)
1038
+ 996 train 7.044793 (lr=3.0021e-05) (hash(x)=5639136)
1039
+ 997 train 7.163411 (lr=3.0012e-05) (hash(x)=5840987)
1040
+ 998 train 6.831763 (lr=3.0005e-05) (hash(x)=5265410)
1041
+ 999 val loss 7.0862
1042
+ 999 val perplexity 1195.3549
1043
+ 999 train 6.828047 (lr=3.0001e-05) (hash(x)=4351074)
lr3e-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:e9e1fd2272f56a5c196f6daf9d0631bbe509a663a2e9cdaa0f4a0fcc94df1fde
3
+ size 36163010
lr3e-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 sha256:946e72ff4f8166797819e62fa194d04bc4a29571c5e8c725d1000e9ea14f42f1
3
+ size 70764358