andrew-healey commited on
Commit
c6c05b8
·
verified ·
1 Parent(s): fa255d2

Upload folder using huggingface_hub

Browse files
lr8e-4_total_batch_size40960_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr8e-4_total_batch_size40960_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": 40960, "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": 0.0008, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr8e-4_total_batch_size40960_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cec289292cf51a5462e2631af59109e64df3eca2d9e62e1fbd1c53a776a0f874
3
+ size 964
lr8e-4_total_batch_size40960_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 11.697288 (lr=2.7972e-06) (hash(x)=5534142)
5
+ 1 train 11.695920 (lr=5.5944e-06) (hash(x)=6328012)
6
+ 2 train 11.667794 (lr=8.3916e-06) (hash(x)=5556098)
7
+ 3 train 11.643607 (lr=1.1189e-05) (hash(x)=6219138)
8
+ 4 train 11.585120 (lr=1.3986e-05) (hash(x)=5491972)
9
+ 5 train 11.533912 (lr=1.6783e-05) (hash(x)=5917741)
10
+ 6 train 11.475008 (lr=1.9580e-05) (hash(x)=6112458)
11
+ 7 train 11.392788 (lr=2.2378e-05) (hash(x)=6568379)
12
+ 8 train 11.291488 (lr=2.5175e-05) (hash(x)=7386082)
13
+ 9 train 11.214754 (lr=2.7972e-05) (hash(x)=6151674)
14
+ 10 train 11.142526 (lr=3.0769e-05) (hash(x)=7859060)
15
+ 11 train 11.032938 (lr=3.3566e-05) (hash(x)=6391506)
16
+ 12 train 10.950451 (lr=3.6364e-05) (hash(x)=5985675)
17
+ 13 train 10.864902 (lr=3.9161e-05) (hash(x)=6984772)
18
+ 14 train 10.791240 (lr=4.1958e-05) (hash(x)=6603717)
19
+ 15 train 10.668466 (lr=4.4755e-05) (hash(x)=6190579)
20
+ 16 train 10.621742 (lr=4.7552e-05) (hash(x)=7276303)
21
+ 17 train 10.631202 (lr=5.0350e-05) (hash(x)=6630378)
22
+ 18 train 10.618442 (lr=5.3147e-05) (hash(x)=9989744)
23
+ 19 train 10.492077 (lr=5.5944e-05) (hash(x)=6317894)
24
+ 20 train 10.412903 (lr=5.8741e-05) (hash(x)=7947703)
25
+ 21 train 10.372401 (lr=6.1538e-05) (hash(x)=5777498)
26
+ 22 train 10.352867 (lr=6.4336e-05) (hash(x)=7723226)
27
+ 23 train 10.312781 (lr=6.7133e-05) (hash(x)=6491229)
28
+ 24 train 10.399549 (lr=6.9930e-05) (hash(x)=6019389)
29
+ 25 train 10.341461 (lr=7.2727e-05) (hash(x)=4648609)
30
+ 26 train 10.291364 (lr=7.5524e-05) (hash(x)=5558355)
31
+ 27 train 10.315466 (lr=7.8322e-05) (hash(x)=6369326)
32
+ 28 train 10.252752 (lr=8.1119e-05) (hash(x)=4906853)
33
+ 29 train 10.158627 (lr=8.3916e-05) (hash(x)=6437376)
34
+ 30 train 10.241501 (lr=8.6713e-05) (hash(x)=7084093)
35
+ 31 train 10.172038 (lr=8.9510e-05) (hash(x)=7048804)
36
+ 32 train 10.173751 (lr=9.2308e-05) (hash(x)=5752594)
37
+ 33 train 9.963683 (lr=9.5105e-05) (hash(x)=6393520)
38
+ 34 train 10.158913 (lr=9.7902e-05) (hash(x)=6793550)
39
+ 35 train 10.068394 (lr=1.0070e-04) (hash(x)=6581348)
40
+ 36 train 10.014685 (lr=1.0350e-04) (hash(x)=6235837)
41
+ 37 train 10.035067 (lr=1.0629e-04) (hash(x)=7083021)
42
+ 38 train 9.994874 (lr=1.0909e-04) (hash(x)=6155821)
43
+ 39 train 9.867517 (lr=1.1189e-04) (hash(x)=5657123)
44
+ 40 train 9.744009 (lr=1.1469e-04) (hash(x)=5312267)
45
+ 41 train 9.797932 (lr=1.1748e-04) (hash(x)=6830813)
46
+ 42 train 9.737255 (lr=1.2028e-04) (hash(x)=6424933)
47
+ 43 train 9.664257 (lr=1.2308e-04) (hash(x)=6680023)
48
+ 44 train 9.590752 (lr=1.2587e-04) (hash(x)=7523279)
49
+ 45 train 9.486513 (lr=1.2867e-04) (hash(x)=6635925)
50
+ 46 train 9.483954 (lr=1.3147e-04) (hash(x)=5869441)
51
+ 47 train 9.436051 (lr=1.3427e-04) (hash(x)=6349004)
52
+ 48 train 9.286825 (lr=1.3706e-04) (hash(x)=6008454)
53
+ 49 train 9.253229 (lr=1.3986e-04) (hash(x)=7682741)
54
+ 50 val loss 9.2472
55
+ 50 val perplexity 10375.5732
56
+ 50 train 9.264073 (lr=1.4266e-04) (hash(x)=7794028)
57
+ 51 train 9.178639 (lr=1.4545e-04) (hash(x)=6166062)
58
+ 52 train 9.089400 (lr=1.4825e-04) (hash(x)=4811192)
59
+ 53 train 8.997089 (lr=1.5105e-04) (hash(x)=7822680)
60
+ 54 train 8.917319 (lr=1.5385e-04) (hash(x)=6431409)
61
+ 55 train 8.890619 (lr=1.5664e-04) (hash(x)=5323032)
62
+ 56 train 8.770432 (lr=1.5944e-04) (hash(x)=7363286)
63
+ 57 train 8.800536 (lr=1.6224e-04) (hash(x)=5616075)
64
+ 58 train 8.748420 (lr=1.6503e-04) (hash(x)=5672608)
65
+ 59 train 8.623936 (lr=1.6783e-04) (hash(x)=6839867)
66
+ 60 train 8.526201 (lr=1.7063e-04) (hash(x)=4916928)
67
+ 61 train 8.547891 (lr=1.7343e-04) (hash(x)=8353379)
68
+ 62 train 8.448561 (lr=1.7622e-04) (hash(x)=4838871)
69
+ 63 train 8.265900 (lr=1.7902e-04) (hash(x)=6795665)
70
+ 64 train 8.313754 (lr=1.8182e-04) (hash(x)=6678038)
71
+ 65 train 8.340294 (lr=1.8462e-04) (hash(x)=6112215)
72
+ 66 train 8.281316 (lr=1.8741e-04) (hash(x)=5915361)
73
+ 67 train 8.162537 (lr=1.9021e-04) (hash(x)=5771588)
74
+ 68 train 8.114316 (lr=1.9301e-04) (hash(x)=5688829)
75
+ 69 train 8.067454 (lr=1.9580e-04) (hash(x)=5443305)
76
+ 70 train 8.153781 (lr=1.9860e-04) (hash(x)=6901933)
77
+ 71 train 8.019279 (lr=2.0140e-04) (hash(x)=4294425)
78
+ 72 train 7.936037 (lr=2.0420e-04) (hash(x)=4363941)
79
+ 73 train 8.121943 (lr=2.0699e-04) (hash(x)=5486335)
80
+ 74 train 8.098098 (lr=2.0979e-04) (hash(x)=7295165)
81
+ 75 train 7.991174 (lr=2.1259e-04) (hash(x)=6267424)
82
+ 76 train 8.019372 (lr=2.1538e-04) (hash(x)=5270452)
83
+ 77 train 7.846412 (lr=2.1818e-04) (hash(x)=6305297)
84
+ 78 train 8.090322 (lr=2.2098e-04) (hash(x)=6040077)
85
+ 79 train 8.031216 (lr=2.2378e-04) (hash(x)=7130251)
86
+ 80 train 7.738321 (lr=2.2657e-04) (hash(x)=6508350)
87
+ 81 train 7.884501 (lr=2.2937e-04) (hash(x)=5859576)
88
+ 82 train 7.538318 (lr=2.3217e-04) (hash(x)=4892065)
89
+ 83 train 7.670144 (lr=2.3497e-04) (hash(x)=7317289)
90
+ 84 train 7.794332 (lr=2.3776e-04) (hash(x)=5510218)
91
+ 85 train 7.953207 (lr=2.4056e-04) (hash(x)=6752265)
92
+ 86 train 7.697105 (lr=2.4336e-04) (hash(x)=6530867)
93
+ 87 train 7.766182 (lr=2.4615e-04) (hash(x)=5790078)
94
+ 88 train 7.772139 (lr=2.4895e-04) (hash(x)=7614023)
95
+ 89 train 7.789983 (lr=2.5175e-04) (hash(x)=5245146)
96
+ 90 train 7.704303 (lr=2.5455e-04) (hash(x)=5786972)
97
+ 91 train 7.764390 (lr=2.5734e-04) (hash(x)=6346459)
98
+ 92 train 7.794429 (lr=2.6014e-04) (hash(x)=5209234)
99
+ 93 train 7.699342 (lr=2.6294e-04) (hash(x)=5676768)
100
+ 94 train 7.847878 (lr=2.6573e-04) (hash(x)=5863092)
101
+ 95 train 7.874211 (lr=2.6853e-04) (hash(x)=7019489)
102
+ 96 train 7.712156 (lr=2.7133e-04) (hash(x)=5483019)
103
+ 97 train 7.720919 (lr=2.7413e-04) (hash(x)=5968716)
104
+ 98 train 7.621480 (lr=2.7692e-04) (hash(x)=7780194)
105
+ 99 train 7.734790 (lr=2.7972e-04) (hash(x)=3784321)
106
+ 100 val loss 7.7036
107
+ 100 val perplexity 2216.2815
108
+ 100 train 7.727891 (lr=2.8252e-04) (hash(x)=7774109)
109
+ 101 train 7.861131 (lr=2.8531e-04) (hash(x)=6576749)
110
+ 102 train 7.763587 (lr=2.8811e-04) (hash(x)=7072404)
111
+ 103 train 7.676697 (lr=2.9091e-04) (hash(x)=6556025)
112
+ 104 train 7.757595 (lr=2.9371e-04) (hash(x)=7325027)
113
+ 105 train 7.762927 (lr=2.9650e-04) (hash(x)=6373412)
114
+ 106 train 7.664719 (lr=2.9930e-04) (hash(x)=5262868)
115
+ 107 train 7.593937 (lr=3.0210e-04) (hash(x)=8055563)
116
+ 108 train 7.727453 (lr=3.0490e-04) (hash(x)=5932790)
117
+ 109 train 7.681774 (lr=3.0769e-04) (hash(x)=5680154)
118
+ 110 train 7.723314 (lr=3.1049e-04) (hash(x)=5914217)
119
+ 111 train 7.640156 (lr=3.1329e-04) (hash(x)=6424128)
120
+ 112 train 7.580154 (lr=3.1608e-04) (hash(x)=7804089)
121
+ 113 train 7.708635 (lr=3.1888e-04) (hash(x)=6490149)
122
+ 114 train 7.644102 (lr=3.2168e-04) (hash(x)=6283540)
123
+ 115 train 7.613694 (lr=3.2448e-04) (hash(x)=5728337)
124
+ 116 train 7.629118 (lr=3.2727e-04) (hash(x)=7400272)
125
+ 117 train 7.640052 (lr=3.3007e-04) (hash(x)=6473432)
126
+ 118 train 7.731514 (lr=3.3287e-04) (hash(x)=6922646)
127
+ 119 train 7.613554 (lr=3.3566e-04) (hash(x)=7277141)
128
+ 120 train 7.670222 (lr=3.3846e-04) (hash(x)=7443124)
129
+ 121 train 7.659196 (lr=3.4126e-04) (hash(x)=5632735)
130
+ 122 train 7.512862 (lr=3.4406e-04) (hash(x)=5881015)
131
+ 123 train 7.652102 (lr=3.4685e-04) (hash(x)=5809411)
132
+ 124 train 7.653436 (lr=3.4965e-04) (hash(x)=7232964)
133
+ 125 train 7.623687 (lr=3.5245e-04) (hash(x)=6171829)
134
+ 126 train 7.704216 (lr=3.5524e-04) (hash(x)=5884456)
135
+ 127 train 7.662256 (lr=3.5804e-04) (hash(x)=5077031)
136
+ 128 train 7.725883 (lr=3.6084e-04) (hash(x)=6248895)
137
+ 129 train 7.644543 (lr=3.6364e-04) (hash(x)=6918004)
138
+ 130 train 7.703830 (lr=3.6643e-04) (hash(x)=6482466)
139
+ 131 train 7.635147 (lr=3.6923e-04) (hash(x)=6588573)
140
+ 132 train 7.676833 (lr=3.7203e-04) (hash(x)=6136939)
141
+ 133 train 7.586064 (lr=3.7483e-04) (hash(x)=7040157)
142
+ 134 train 7.697452 (lr=3.7762e-04) (hash(x)=6455595)
143
+ 135 train 7.667699 (lr=3.8042e-04) (hash(x)=4122983)
144
+ 136 train 7.613140 (lr=3.8322e-04) (hash(x)=4724052)
145
+ 137 train 7.569327 (lr=3.8601e-04) (hash(x)=6820797)
146
+ 138 train 7.809066 (lr=3.8881e-04) (hash(x)=7477594)
147
+ 139 train 7.555345 (lr=3.9161e-04) (hash(x)=6146305)
148
+ 140 train 7.570525 (lr=3.9441e-04) (hash(x)=6814831)
149
+ 141 train 7.573858 (lr=3.9720e-04) (hash(x)=5548298)
150
+ 142 train 7.746276 (lr=4.0000e-04) (hash(x)=7609416)
151
+ 143 train 7.682101 (lr=4.0280e-04) (hash(x)=7574342)
152
+ 144 train 7.530268 (lr=4.0559e-04) (hash(x)=5906937)
153
+ 145 train 7.586014 (lr=4.0839e-04) (hash(x)=8013471)
154
+ 146 train 7.569292 (lr=4.1119e-04) (hash(x)=6146213)
155
+ 147 train 7.596301 (lr=4.1399e-04) (hash(x)=6709902)
156
+ 148 train 7.734204 (lr=4.1678e-04) (hash(x)=6402571)
157
+ 149 train 7.545006 (lr=4.1958e-04) (hash(x)=7683614)
158
+ 150 val loss 7.5814
159
+ 150 val perplexity 1961.4075
160
+ 150 train 7.540520 (lr=4.2238e-04) (hash(x)=5453026)
161
+ 151 train 7.487573 (lr=4.2517e-04) (hash(x)=6260017)
162
+ 152 train 7.619231 (lr=4.2797e-04) (hash(x)=6591466)
163
+ 153 train 7.598324 (lr=4.3077e-04) (hash(x)=5501472)
164
+ 154 train 7.634311 (lr=4.3357e-04) (hash(x)=5871698)
165
+ 155 train 7.469506 (lr=4.3636e-04) (hash(x)=4282052)
166
+ 156 train 7.559871 (lr=4.3916e-04) (hash(x)=7824660)
167
+ 157 train 7.495238 (lr=4.4196e-04) (hash(x)=5954950)
168
+ 158 train 7.561706 (lr=4.4476e-04) (hash(x)=5484656)
169
+ 159 train 7.676786 (lr=4.4755e-04) (hash(x)=7031328)
170
+ 160 train 7.630304 (lr=4.5035e-04) (hash(x)=9157069)
171
+ 161 train 7.714020 (lr=4.5315e-04) (hash(x)=5996883)
172
+ 162 train 7.777853 (lr=4.5594e-04) (hash(x)=7645447)
173
+ 163 train 7.815012 (lr=4.5874e-04) (hash(x)=6336164)
174
+ 164 train 7.543817 (lr=4.6154e-04) (hash(x)=5224240)
175
+ 165 train 7.880888 (lr=4.6434e-04) (hash(x)=5447801)
176
+ 166 train 7.570288 (lr=4.6713e-04) (hash(x)=4155598)
177
+ 167 train 7.554580 (lr=4.6993e-04) (hash(x)=7482744)
178
+ 168 train 7.575298 (lr=4.7273e-04) (hash(x)=4733414)
179
+ 169 train 7.544789 (lr=4.7552e-04) (hash(x)=7806284)
180
+ 170 train 7.557264 (lr=4.7832e-04) (hash(x)=7236478)
181
+ 171 train 7.568693 (lr=4.8112e-04) (hash(x)=4656255)
182
+ 172 train 7.667747 (lr=4.8392e-04) (hash(x)=9298325)
183
+ 173 train 7.589840 (lr=4.8671e-04) (hash(x)=6287524)
184
+ 174 train 7.633582 (lr=4.8951e-04) (hash(x)=7895970)
185
+ 175 train 7.586063 (lr=4.9231e-04) (hash(x)=8847992)
186
+ 176 train 7.805615 (lr=4.9510e-04) (hash(x)=6338998)
187
+ 177 train 7.760677 (lr=4.9790e-04) (hash(x)=7104011)
188
+ 178 train 7.584731 (lr=5.0070e-04) (hash(x)=5573214)
189
+ 179 train 7.609766 (lr=5.0350e-04) (hash(x)=6882897)
190
+ 180 train 7.585401 (lr=5.0629e-04) (hash(x)=5854242)
191
+ 181 train 7.429110 (lr=5.0909e-04) (hash(x)=6972494)
192
+ 182 train 7.465399 (lr=5.1189e-04) (hash(x)=6796805)
193
+ 183 train 7.358451 (lr=5.1469e-04) (hash(x)=6139120)
194
+ 184 train 7.536239 (lr=5.1748e-04) (hash(x)=6486796)
195
+ 185 train 7.410870 (lr=5.2028e-04) (hash(x)=5578797)
196
+ 186 train 7.492826 (lr=5.2308e-04) (hash(x)=6895326)
197
+ 187 train 7.531305 (lr=5.2587e-04) (hash(x)=7783160)
198
+ 188 train 7.542838 (lr=5.2867e-04) (hash(x)=4867625)
199
+ 189 train 7.368277 (lr=5.3147e-04) (hash(x)=6631123)
200
+ 190 train 7.222600 (lr=5.3427e-04) (hash(x)=6710965)
201
+ 191 train 7.360646 (lr=5.3706e-04) (hash(x)=6294657)
202
+ 192 train 7.515742 (lr=5.3986e-04) (hash(x)=5443606)
203
+ 193 train 7.486578 (lr=5.4266e-04) (hash(x)=7035770)
204
+ 194 train 7.527805 (lr=5.4545e-04) (hash(x)=4899541)
205
+ 195 train 7.375118 (lr=5.4825e-04) (hash(x)=6227120)
206
+ 196 train 7.464990 (lr=5.5105e-04) (hash(x)=5981648)
207
+ 197 train 7.429571 (lr=5.5385e-04) (hash(x)=5558993)
208
+ 198 train 7.462946 (lr=5.5664e-04) (hash(x)=7015671)
209
+ 199 train 7.451362 (lr=5.5944e-04) (hash(x)=5931234)
210
+ 200 val loss 7.4482
211
+ 200 val perplexity 1716.7419
212
+ 200 train 7.477434 (lr=5.6224e-04) (hash(x)=4582159)
213
+ 201 train 7.466753 (lr=5.6503e-04) (hash(x)=5029472)
214
+ 202 train 7.372465 (lr=5.6783e-04) (hash(x)=5434102)
215
+ 203 train 7.635233 (lr=5.7063e-04) (hash(x)=6810638)
216
+ 204 train 7.475424 (lr=5.7343e-04) (hash(x)=6521562)
217
+ 205 train 7.479112 (lr=5.7622e-04) (hash(x)=7226932)
218
+ 206 train 7.370938 (lr=5.7902e-04) (hash(x)=6957849)
219
+ 207 train 7.485872 (lr=5.8182e-04) (hash(x)=5334281)
220
+ 208 train 7.583689 (lr=5.8462e-04) (hash(x)=5887983)
221
+ 209 train 7.413815 (lr=5.8741e-04) (hash(x)=6393576)
222
+ 210 train 7.634243 (lr=5.9021e-04) (hash(x)=5057959)
223
+ 211 train 7.671282 (lr=5.9301e-04) (hash(x)=8595271)
224
+ 212 train 7.493760 (lr=5.9580e-04) (hash(x)=8395167)
225
+ 213 train 7.610686 (lr=5.9860e-04) (hash(x)=7634417)
226
+ 214 train 7.347445 (lr=6.0140e-04) (hash(x)=5257771)
227
+ 215 train 7.497778 (lr=6.0420e-04) (hash(x)=7674238)
228
+ 216 train 7.593218 (lr=6.0699e-04) (hash(x)=5300555)
229
+ 217 train 7.352460 (lr=6.0979e-04) (hash(x)=6066139)
230
+ 218 train 7.455792 (lr=6.1259e-04) (hash(x)=6688212)
231
+ 219 train 7.658871 (lr=6.1538e-04) (hash(x)=6642404)
232
+ 220 train 7.414104 (lr=6.1818e-04) (hash(x)=4961570)
233
+ 221 train 7.266774 (lr=6.2098e-04) (hash(x)=5105190)
234
+ 222 train 7.573030 (lr=6.2378e-04) (hash(x)=7155991)
235
+ 223 train 7.407368 (lr=6.2657e-04) (hash(x)=5132972)
236
+ 224 train 7.402404 (lr=6.2937e-04) (hash(x)=6269946)
237
+ 225 train 7.443851 (lr=6.3217e-04) (hash(x)=5396133)
238
+ 226 train 7.465405 (lr=6.3497e-04) (hash(x)=6394083)
239
+ 227 train 7.411294 (lr=6.3776e-04) (hash(x)=5237804)
240
+ 228 train 7.438427 (lr=6.4056e-04) (hash(x)=5216638)
241
+ 229 train 7.384326 (lr=6.4336e-04) (hash(x)=6735565)
242
+ 230 train 7.482626 (lr=6.4615e-04) (hash(x)=5983575)
243
+ 231 train 7.458193 (lr=6.4895e-04) (hash(x)=5831619)
244
+ 232 train 7.276896 (lr=6.5175e-04) (hash(x)=9109083)
245
+ 233 train 7.496134 (lr=6.5455e-04) (hash(x)=6912967)
246
+ 234 train 7.685136 (lr=6.5734e-04) (hash(x)=5965674)
247
+ 235 train 7.544114 (lr=6.6014e-04) (hash(x)=5822533)
248
+ 236 train 7.419838 (lr=6.6294e-04) (hash(x)=5268216)
249
+ 237 train 7.326520 (lr=6.6573e-04) (hash(x)=6531900)
250
+ 238 train 7.314795 (lr=6.6853e-04) (hash(x)=5426499)
251
+ 239 train 7.764974 (lr=6.7133e-04) (hash(x)=7455372)
252
+ 240 train 8.041131 (lr=6.7413e-04) (hash(x)=8319119)
253
+ 241 train 7.968082 (lr=6.7692e-04) (hash(x)=7592128)
254
+ 242 train 7.974820 (lr=6.7972e-04) (hash(x)=8704170)
255
+ 243 train 7.700039 (lr=6.8252e-04) (hash(x)=7165037)
256
+ 244 train 7.567582 (lr=6.8531e-04) (hash(x)=7678000)
257
+ 245 train 7.383462 (lr=6.8811e-04) (hash(x)=7414788)
258
+ 246 train 7.412508 (lr=6.9091e-04) (hash(x)=7326216)
259
+ 247 train 7.532023 (lr=6.9371e-04) (hash(x)=5857162)
260
+ 248 train 8.263190 (lr=6.9650e-04) (hash(x)=14715669)
261
+ 249 train 11.879148 (lr=6.9930e-04) (hash(x)=16950560)
262
+ 250 val loss 7.4056
263
+ 250 val perplexity 1645.1035
264
+ 250 train 7.582276 (lr=7.0210e-04) (hash(x)=6159125)
265
+ 251 train 7.444682 (lr=7.0490e-04) (hash(x)=5908442)
266
+ 252 train 7.292677 (lr=7.0769e-04) (hash(x)=6668887)
267
+ 253 train 7.394005 (lr=7.1049e-04) (hash(x)=7662975)
268
+ 254 train 7.384941 (lr=7.1329e-04) (hash(x)=6669539)
269
+ 255 train 7.375414 (lr=7.1608e-04) (hash(x)=5015998)
270
+ 256 train 7.304709 (lr=7.1888e-04) (hash(x)=5458386)
271
+ 257 train 7.477495 (lr=7.2168e-04) (hash(x)=4849847)
272
+ 258 train 7.491002 (lr=7.2448e-04) (hash(x)=5125562)
273
+ 259 train 7.522124 (lr=7.2727e-04) (hash(x)=6012047)
274
+ 260 train 7.410746 (lr=7.3007e-04) (hash(x)=6079364)
275
+ 261 train 7.409963 (lr=7.3287e-04) (hash(x)=6030395)
276
+ 262 train 7.407788 (lr=7.3566e-04) (hash(x)=6781762)
277
+ 263 train 7.436899 (lr=7.3846e-04) (hash(x)=5238325)
278
+ 264 train 7.500021 (lr=7.4126e-04) (hash(x)=8611144)
279
+ 265 train 7.353480 (lr=7.4406e-04) (hash(x)=6901605)
280
+ 266 train 7.381721 (lr=7.4685e-04) (hash(x)=5499337)
281
+ 267 train 7.412154 (lr=7.4965e-04) (hash(x)=7134874)
282
+ 268 train 7.369044 (lr=7.5245e-04) (hash(x)=6149038)
283
+ 269 train 7.416449 (lr=7.5524e-04) (hash(x)=6609760)
284
+ 270 train 7.472561 (lr=7.5804e-04) (hash(x)=6652627)
285
+ 271 train 7.455261 (lr=7.6084e-04) (hash(x)=8618579)
286
+ 272 train 7.341047 (lr=7.6364e-04) (hash(x)=5896892)
287
+ 273 train 7.498653 (lr=7.6643e-04) (hash(x)=8945854)
288
+ 274 train 7.308757 (lr=7.6923e-04) (hash(x)=6775103)
289
+ 275 train 7.383062 (lr=7.7203e-04) (hash(x)=6200024)
290
+ 276 train 7.282226 (lr=7.7483e-04) (hash(x)=5238139)
291
+ 277 train 7.416149 (lr=7.7762e-04) (hash(x)=5780948)
292
+ 278 train 7.394005 (lr=7.8042e-04) (hash(x)=6168407)
293
+ 279 train 7.357039 (lr=7.8322e-04) (hash(x)=3775881)
294
+ 280 train 7.354521 (lr=7.8601e-04) (hash(x)=6848709)
295
+ 281 train 7.632778 (lr=7.8881e-04) (hash(x)=6478113)
296
+ 282 train 7.403832 (lr=7.9161e-04) (hash(x)=6483914)
297
+ 283 train 7.375517 (lr=7.9441e-04) (hash(x)=5927795)
298
+ 284 train 7.389233 (lr=7.9720e-04) (hash(x)=8438033)
299
+ 285 train 7.433768 (lr=8.0000e-04) (hash(x)=5684788)
300
+ 286 train 7.249988 (lr=8.0000e-04) (hash(x)=5710182)
301
+ 287 train 7.333145 (lr=8.0000e-04) (hash(x)=6160063)
302
+ 288 train 7.211267 (lr=7.9999e-04) (hash(x)=5955100)
303
+ 289 train 7.089900 (lr=7.9997e-04) (hash(x)=6672309)
304
+ 290 train 7.499460 (lr=7.9994e-04) (hash(x)=9001564)
305
+ 291 train 7.865353 (lr=7.9991e-04) (hash(x)=5454654)
306
+ 292 train 7.264449 (lr=7.9987e-04) (hash(x)=5174062)
307
+ 293 train 7.429892 (lr=7.9983e-04) (hash(x)=6199203)
308
+ 294 train 7.312432 (lr=7.9978e-04) (hash(x)=8352747)
309
+ 295 train 7.408596 (lr=7.9972e-04) (hash(x)=4547106)
310
+ 296 train 7.380010 (lr=7.9965e-04) (hash(x)=6433099)
311
+ 297 train 7.241283 (lr=7.9958e-04) (hash(x)=6424428)
312
+ 298 train 7.375280 (lr=7.9950e-04) (hash(x)=5975712)
313
+ 299 train 7.265843 (lr=7.9941e-04) (hash(x)=5458532)
314
+ 300 val loss 7.3456
315
+ 300 val perplexity 1549.3386
316
+ 300 train 7.494257 (lr=7.9932e-04) (hash(x)=7002774)
317
+ 301 train 7.600027 (lr=7.9922e-04) (hash(x)=8206173)
318
+ 302 train 7.690063 (lr=7.9911e-04) (hash(x)=7784971)
319
+ 303 train 7.588319 (lr=7.9899e-04) (hash(x)=6001843)
320
+ 304 train 7.438201 (lr=7.9887e-04) (hash(x)=5698642)
321
+ 305 train 7.353337 (lr=7.9874e-04) (hash(x)=6600751)
322
+ 306 train 7.314136 (lr=7.9861e-04) (hash(x)=5804610)
323
+ 307 train 7.248654 (lr=7.9846e-04) (hash(x)=5162652)
324
+ 308 train 7.271367 (lr=7.9831e-04) (hash(x)=5297006)
325
+ 309 train 7.283448 (lr=7.9816e-04) (hash(x)=6728323)
326
+ 310 train 7.286610 (lr=7.9799e-04) (hash(x)=3964882)
327
+ 311 train 7.267771 (lr=7.9782e-04) (hash(x)=6735993)
328
+ 312 train 7.408009 (lr=7.9765e-04) (hash(x)=5150316)
329
+ 313 train 7.221607 (lr=7.9746e-04) (hash(x)=5220448)
330
+ 314 train 7.436646 (lr=7.9727e-04) (hash(x)=6971289)
331
+ 315 train 7.455480 (lr=7.9707e-04) (hash(x)=5012069)
332
+ 316 train 7.308303 (lr=7.9687e-04) (hash(x)=6576136)
333
+ 317 train 7.315253 (lr=7.9666e-04) (hash(x)=6077647)
334
+ 318 train 7.399382 (lr=7.9644e-04) (hash(x)=6032381)
335
+ 319 train 7.256501 (lr=7.9621e-04) (hash(x)=8410716)
336
+ 320 train 7.286648 (lr=7.9598e-04) (hash(x)=6212516)
337
+ 321 train 7.264132 (lr=7.9574e-04) (hash(x)=5298545)
338
+ 322 train 7.369779 (lr=7.9549e-04) (hash(x)=5779073)
339
+ 323 train 7.224717 (lr=7.9524e-04) (hash(x)=6340947)
340
+ 324 train 6.981560 (lr=7.9498e-04) (hash(x)=6530441)
341
+ 325 train 7.318573 (lr=7.9471e-04) (hash(x)=7818007)
342
+ 326 train 7.197862 (lr=7.9444e-04) (hash(x)=6277435)
343
+ 327 train 7.212821 (lr=7.9416e-04) (hash(x)=6226320)
344
+ 328 train 7.302688 (lr=7.9387e-04) (hash(x)=5463751)
345
+ 329 train 7.206116 (lr=7.9358e-04) (hash(x)=6579329)
346
+ 330 train 7.315201 (lr=7.9327e-04) (hash(x)=6660540)
347
+ 331 train 7.274509 (lr=7.9297e-04) (hash(x)=6491272)
348
+ 332 train 7.235404 (lr=7.9265e-04) (hash(x)=6766947)
349
+ 333 train 7.533473 (lr=7.9233e-04) (hash(x)=3721218)
350
+ 334 train 7.377615 (lr=7.9200e-04) (hash(x)=5229898)
351
+ 335 train 7.168855 (lr=7.9167e-04) (hash(x)=6054428)
352
+ 336 train 7.317795 (lr=7.9132e-04) (hash(x)=6290464)
353
+ 337 train 7.294703 (lr=7.9097e-04) (hash(x)=6904498)
354
+ 338 train 7.349016 (lr=7.9062e-04) (hash(x)=7765805)
355
+ 339 train 7.296751 (lr=7.9026e-04) (hash(x)=6543764)
356
+ 340 train 7.301435 (lr=7.8989e-04) (hash(x)=7204955)
357
+ 341 train 7.214976 (lr=7.8951e-04) (hash(x)=4423711)
358
+ 342 train 7.173480 (lr=7.8913e-04) (hash(x)=5343477)
359
+ 343 train 7.338549 (lr=7.8874e-04) (hash(x)=6595856)
360
+ 344 train 7.152942 (lr=7.8834e-04) (hash(x)=7229015)
361
+ 345 train 7.231004 (lr=7.8794e-04) (hash(x)=5916034)
362
+ 346 train 7.362188 (lr=7.8753e-04) (hash(x)=6939521)
363
+ 347 train 7.228534 (lr=7.8711e-04) (hash(x)=5492185)
364
+ 348 train 7.178028 (lr=7.8669e-04) (hash(x)=6898006)
365
+ 349 train 7.103750 (lr=7.8626e-04) (hash(x)=5462008)
366
+ 350 val loss 7.3251
367
+ 350 val perplexity 1517.9257
368
+ 350 train 7.223419 (lr=7.8582e-04) (hash(x)=7821902)
369
+ 351 train 7.331910 (lr=7.8538e-04) (hash(x)=6608108)
370
+ 352 train 7.367597 (lr=7.8493e-04) (hash(x)=6295394)
371
+ 353 train 7.227644 (lr=7.8447e-04) (hash(x)=6803449)
372
+ 354 train 7.225511 (lr=7.8401e-04) (hash(x)=7444134)
373
+ 355 train 7.274103 (lr=7.8354e-04) (hash(x)=5361923)
374
+ 356 train 7.300616 (lr=7.8306e-04) (hash(x)=7803039)
375
+ 357 train 7.419339 (lr=7.8258e-04) (hash(x)=7042578)
376
+ 358 train 7.289340 (lr=7.8209e-04) (hash(x)=6647007)
377
+ 359 train 7.292564 (lr=7.8159e-04) (hash(x)=6303658)
378
+ 360 train 7.388300 (lr=7.8109e-04) (hash(x)=6158927)
379
+ 361 train 7.265947 (lr=7.8058e-04) (hash(x)=6419656)
380
+ 362 train 7.189330 (lr=7.8006e-04) (hash(x)=7891271)
381
+ 363 train 7.397287 (lr=7.7954e-04) (hash(x)=7099888)
382
+ 364 train 7.351587 (lr=7.7901e-04) (hash(x)=6352424)
383
+ 365 train 7.260668 (lr=7.7847e-04) (hash(x)=5863145)
384
+ 366 train 7.293188 (lr=7.7793e-04) (hash(x)=6844597)
385
+ 367 train 7.330185 (lr=7.7738e-04) (hash(x)=6009861)
386
+ 368 train 7.216932 (lr=7.7682e-04) (hash(x)=5374957)
387
+ 369 train 7.284505 (lr=7.7626e-04) (hash(x)=3923225)
388
+ 370 train 7.224683 (lr=7.7569e-04) (hash(x)=5848862)
389
+ 371 train 7.253258 (lr=7.7511e-04) (hash(x)=4440634)
390
+ 372 train 7.207953 (lr=7.7453e-04) (hash(x)=4788998)
391
+ 373 train 7.184224 (lr=7.7394e-04) (hash(x)=7378165)
392
+ 374 train 7.218213 (lr=7.7335e-04) (hash(x)=4771243)
393
+ 375 train 7.219984 (lr=7.7275e-04) (hash(x)=5716089)
394
+ 376 train 7.251999 (lr=7.7214e-04) (hash(x)=7991009)
395
+ 377 train 7.303088 (lr=7.7153e-04) (hash(x)=5348607)
396
+ 378 train 7.367265 (lr=7.7091e-04) (hash(x)=6036146)
397
+ 379 train 7.239356 (lr=7.7028e-04) (hash(x)=6658718)
398
+ 380 train 7.190146 (lr=7.6964e-04) (hash(x)=6176253)
399
+ 381 train 7.202798 (lr=7.6901e-04) (hash(x)=4542895)
400
+ 382 train 7.104035 (lr=7.6836e-04) (hash(x)=5549073)
401
+ 383 train 7.383494 (lr=7.6771e-04) (hash(x)=5342783)
402
+ 384 train 7.411261 (lr=7.6705e-04) (hash(x)=6549546)
403
+ 385 train 7.196146 (lr=7.6638e-04) (hash(x)=6170599)
404
+ 386 train 7.476060 (lr=7.6571e-04) (hash(x)=8483134)
405
+ 387 train 7.592927 (lr=7.6503e-04) (hash(x)=4865036)
406
+ 388 train 7.245173 (lr=7.6435e-04) (hash(x)=4433890)
407
+ 389 train 7.766350 (lr=7.6366e-04) (hash(x)=6738597)
408
+ 390 train 7.374644 (lr=7.6296e-04) (hash(x)=5459620)
409
+ 391 train 7.373478 (lr=7.6226e-04) (hash(x)=4604074)
410
+ 392 train 7.372497 (lr=7.6155e-04) (hash(x)=8045501)
411
+ 393 train 7.464375 (lr=7.6083e-04) (hash(x)=6176464)
412
+ 394 train 7.288664 (lr=7.6011e-04) (hash(x)=6127317)
413
+ 395 train 7.338772 (lr=7.5938e-04) (hash(x)=5571462)
414
+ 396 train 7.318034 (lr=7.5865e-04) (hash(x)=6640755)
415
+ 397 train 7.366526 (lr=7.5791e-04) (hash(x)=6147295)
416
+ 398 train 7.303747 (lr=7.5716e-04) (hash(x)=6120461)
417
+ 399 train 7.267434 (lr=7.5641e-04) (hash(x)=5123928)
418
+ 400 val loss 7.3113
419
+ 400 val perplexity 1497.1090
420
+ 400 train 7.286520 (lr=7.5565e-04) (hash(x)=5428813)
421
+ 401 train 7.262196 (lr=7.5489e-04) (hash(x)=6448586)
422
+ 402 train 7.420949 (lr=7.5412e-04) (hash(x)=8116987)
423
+ 403 train 7.440536 (lr=7.5334e-04) (hash(x)=5974342)
424
+ 404 train 7.459017 (lr=7.5256e-04) (hash(x)=5261336)
425
+ 405 train 7.281918 (lr=7.5177e-04) (hash(x)=5307056)
426
+ 406 train 7.200621 (lr=7.5097e-04) (hash(x)=5865119)
427
+ 407 train 7.227658 (lr=7.5017e-04) (hash(x)=6672380)
428
+ 408 train 7.278050 (lr=7.4937e-04) (hash(x)=6886810)
429
+ 409 train 7.432333 (lr=7.4855e-04) (hash(x)=6564071)
430
+ 410 train 7.174788 (lr=7.4773e-04) (hash(x)=7705089)
431
+ 411 train 7.192070 (lr=7.4691e-04) (hash(x)=5773073)
432
+ 412 train 7.184642 (lr=7.4608e-04) (hash(x)=5905976)
433
+ 413 train 7.929121 (lr=7.4524e-04) (hash(x)=6936653)
434
+ 414 train 7.288633 (lr=7.4440e-04) (hash(x)=6592773)
435
+ 415 train 7.233203 (lr=7.4355e-04) (hash(x)=6796551)
436
+ 416 train 7.283764 (lr=7.4270e-04) (hash(x)=6151589)
437
+ 417 train 7.527581 (lr=7.4184e-04) (hash(x)=7045164)
438
+ 418 train 7.315282 (lr=7.4097e-04) (hash(x)=6650506)
439
+ 419 train 7.243340 (lr=7.4010e-04) (hash(x)=7839748)
440
+ 420 train 7.461590 (lr=7.3922e-04) (hash(x)=6028671)
441
+ 421 train 7.320601 (lr=7.3834e-04) (hash(x)=6118408)
442
+ 422 train 7.279452 (lr=7.3745e-04) (hash(x)=5757526)
443
+ 423 train 7.420662 (lr=7.3655e-04) (hash(x)=6030172)
444
+ 424 train 7.491510 (lr=7.3565e-04) (hash(x)=6719882)
445
+ 425 train 7.343468 (lr=7.3474e-04) (hash(x)=5028050)
446
+ 426 train 7.351423 (lr=7.3383e-04) (hash(x)=5682910)
447
+ 427 train 7.423174 (lr=7.3291e-04) (hash(x)=5747891)
448
+ 428 train 7.229238 (lr=7.3199e-04) (hash(x)=5734208)
449
+ 429 train 7.283911 (lr=7.3106e-04) (hash(x)=7024381)
450
+ 430 train 7.552984 (lr=7.3012e-04) (hash(x)=7814361)
451
+ 431 train 7.399691 (lr=7.2918e-04) (hash(x)=5765893)
452
+ 432 train 7.258517 (lr=7.2824e-04) (hash(x)=5572048)
453
+ 433 train 7.421177 (lr=7.2729e-04) (hash(x)=5610058)
454
+ 434 train 7.270415 (lr=7.2633e-04) (hash(x)=4459483)
455
+ 435 train 7.235401 (lr=7.2537e-04) (hash(x)=6758411)
456
+ 436 train 7.336760 (lr=7.2440e-04) (hash(x)=6701381)
457
+ 437 train 7.485813 (lr=7.2342e-04) (hash(x)=5416400)
458
+ 438 train 7.236591 (lr=7.2244e-04) (hash(x)=5919569)
459
+ 439 train 7.384229 (lr=7.2146e-04) (hash(x)=5282870)
460
+ 440 train 7.245607 (lr=7.2047e-04) (hash(x)=8072181)
461
+ 441 train 7.198156 (lr=7.1947e-04) (hash(x)=7996770)
462
+ 442 train 7.264748 (lr=7.1847e-04) (hash(x)=5093519)
463
+ 443 train 7.293458 (lr=7.1747e-04) (hash(x)=6270917)
464
+ 444 train 7.175251 (lr=7.1645e-04) (hash(x)=5852780)
465
+ 445 train 7.074408 (lr=7.1544e-04) (hash(x)=5327728)
466
+ 446 train 7.309034 (lr=7.1441e-04) (hash(x)=6260765)
467
+ 447 train 7.374115 (lr=7.1339e-04) (hash(x)=6212894)
468
+ 448 train 7.271809 (lr=7.1235e-04) (hash(x)=6724010)
469
+ 449 train 7.370953 (lr=7.1131e-04) (hash(x)=6059934)
470
+ 450 val loss 7.3727
471
+ 450 val perplexity 1591.9281
472
+ 450 train 7.150157 (lr=7.1027e-04) (hash(x)=5587382)
473
+ 451 train 7.283811 (lr=7.0922e-04) (hash(x)=7274108)
474
+ 452 train 7.322528 (lr=7.0817e-04) (hash(x)=6085396)
475
+ 453 train 7.230937 (lr=7.0711e-04) (hash(x)=4784666)
476
+ 454 train 7.327444 (lr=7.0604e-04) (hash(x)=6288638)
477
+ 455 train 7.347158 (lr=7.0497e-04) (hash(x)=7714288)
478
+ 456 train 7.432976 (lr=7.0390e-04) (hash(x)=5534748)
479
+ 457 train 7.399427 (lr=7.0282e-04) (hash(x)=6361270)
480
+ 458 train 7.439818 (lr=7.0173e-04) (hash(x)=4682956)
481
+ 459 train 7.402696 (lr=7.0064e-04) (hash(x)=7568935)
482
+ 460 train 7.454140 (lr=6.9955e-04) (hash(x)=7973311)
483
+ 461 train 7.347954 (lr=6.9845e-04) (hash(x)=6987352)
484
+ 462 train 7.339832 (lr=6.9734e-04) (hash(x)=5413904)
485
+ 463 train 7.504987 (lr=6.9623e-04) (hash(x)=6732924)
486
+ 464 train 7.444580 (lr=6.9512e-04) (hash(x)=6219754)
487
+ 465 train 7.382334 (lr=6.9400e-04) (hash(x)=6839070)
488
+ 466 train 7.316649 (lr=6.9287e-04) (hash(x)=6090120)
489
+ 467 train 7.725515 (lr=6.9174e-04) (hash(x)=7115176)
490
+ 468 train 7.465732 (lr=6.9061e-04) (hash(x)=4989299)
491
+ 469 train 7.356131 (lr=6.8947e-04) (hash(x)=7497590)
492
+ 470 train 7.430796 (lr=6.8832e-04) (hash(x)=3977557)
493
+ 471 train 7.400104 (lr=6.8717e-04) (hash(x)=5524165)
494
+ 472 train 7.400688 (lr=6.8602e-04) (hash(x)=5124936)
495
+ 473 train 7.430546 (lr=6.8486e-04) (hash(x)=5522119)
496
+ 474 train 7.364988 (lr=6.8370e-04) (hash(x)=5030865)
497
+ 475 train 7.427809 (lr=6.8253e-04) (hash(x)=6411783)
498
+ 476 train 7.370835 (lr=6.8136e-04) (hash(x)=6581937)
499
+ 477 train 7.403676 (lr=6.8018e-04) (hash(x)=5116910)
500
+ 478 train 7.214337 (lr=6.7900e-04) (hash(x)=7353459)
501
+ 479 train 7.254469 (lr=6.7781e-04) (hash(x)=6344535)
502
+ 480 train 7.572206 (lr=6.7662e-04) (hash(x)=5840169)
503
+ 481 train 7.377628 (lr=6.7542e-04) (hash(x)=6754290)
504
+ 482 train 7.576836 (lr=6.7422e-04) (hash(x)=6484573)
505
+ 483 train 7.267538 (lr=6.7302e-04) (hash(x)=5260125)
506
+ 484 train 7.306373 (lr=6.7181e-04) (hash(x)=4881179)
507
+ 485 train 7.380256 (lr=6.7059e-04) (hash(x)=5001634)
508
+ 486 train 7.321090 (lr=6.6937e-04) (hash(x)=7424810)
509
+ 487 train 7.167312 (lr=6.6815e-04) (hash(x)=5176431)
510
+ 488 train 7.294439 (lr=6.6692e-04) (hash(x)=5421624)
511
+ 489 train 7.315212 (lr=6.6569e-04) (hash(x)=5674439)
512
+ 490 train 7.254185 (lr=6.6446e-04) (hash(x)=5819157)
513
+ 491 train 7.436361 (lr=6.6322e-04) (hash(x)=6506856)
514
+ 492 train 7.250735 (lr=6.6197e-04) (hash(x)=6119124)
515
+ 493 train 7.498211 (lr=6.6072e-04) (hash(x)=6241669)
516
+ 494 train 7.105004 (lr=6.5947e-04) (hash(x)=4992756)
517
+ 495 train 7.543459 (lr=6.5821e-04) (hash(x)=7277927)
518
+ 496 train 7.461330 (lr=6.5695e-04) (hash(x)=6622097)
519
+ 497 train 7.564340 (lr=6.5568e-04) (hash(x)=7026582)
520
+ 498 train 7.485524 (lr=6.5441e-04) (hash(x)=5840987)
521
+ 499 train 7.201328 (lr=6.5314e-04) (hash(x)=4351074)
522
+ 500 val loss 7.4345
523
+ 500 val perplexity 1693.4774
524
+ 500 train 7.340623 (lr=6.5186e-04) (hash(x)=7301477)
525
+ 501 train 7.541590 (lr=6.5058e-04) (hash(x)=6264167)
526
+ 502 train 7.303625 (lr=6.4929e-04) (hash(x)=4673363)
527
+ 503 train 7.530320 (lr=6.4800e-04) (hash(x)=6332240)
528
+ 504 train 7.442071 (lr=6.4670e-04) (hash(x)=5093790)
529
+ 505 train 7.594797 (lr=6.4540e-04) (hash(x)=6890713)
530
+ 506 train 7.352987 (lr=6.4410e-04) (hash(x)=6042691)
531
+ 507 train 7.477703 (lr=6.4280e-04) (hash(x)=7080746)
532
+ 508 train 7.501744 (lr=6.4148e-04) (hash(x)=7609207)
533
+ 509 train 7.310081 (lr=6.4017e-04) (hash(x)=4591273)
534
+ 510 train 7.454575 (lr=6.3885e-04) (hash(x)=8183111)
535
+ 511 train 7.718560 (lr=6.3753e-04) (hash(x)=5893095)
536
+ 512 train 7.497441 (lr=6.3620e-04) (hash(x)=5709077)
537
+ 513 train 7.461885 (lr=6.3487e-04) (hash(x)=6218715)
538
+ 514 train 7.344151 (lr=6.3354e-04) (hash(x)=4722188)
539
+ 515 train 7.250729 (lr=6.3220e-04) (hash(x)=6313422)
540
+ 516 train 7.400622 (lr=6.3086e-04) (hash(x)=6350660)
541
+ 517 train 7.446066 (lr=6.2952e-04) (hash(x)=5371913)
542
+ 518 train 7.563063 (lr=6.2817e-04) (hash(x)=6283595)
543
+ 519 train 7.579663 (lr=6.2681e-04) (hash(x)=6179167)
544
+ 520 train 7.450664 (lr=6.2546e-04) (hash(x)=7865194)
545
+ 521 train 7.640111 (lr=6.2410e-04) (hash(x)=4883760)
546
+ 522 train 7.438840 (lr=6.2274e-04) (hash(x)=5769286)
547
+ 523 train 7.445566 (lr=6.2137e-04) (hash(x)=6309912)
548
+ 524 train 7.431752 (lr=6.2000e-04) (hash(x)=7490849)
549
+ 525 train 7.419092 (lr=6.1863e-04) (hash(x)=5631134)
550
+ 526 train 7.431568 (lr=6.1725e-04) (hash(x)=4754560)
551
+ 527 train 7.384674 (lr=6.1587e-04) (hash(x)=5448702)
552
+ 528 train 7.570445 (lr=6.1449e-04) (hash(x)=7168842)
553
+ 529 train 7.561579 (lr=6.1310e-04) (hash(x)=7469195)
554
+ 530 train 7.382328 (lr=6.1171e-04) (hash(x)=5479500)
555
+ 531 train 7.555226 (lr=6.1031e-04) (hash(x)=5634177)
556
+ 532 train 7.403822 (lr=6.0892e-04) (hash(x)=9779982)
557
+ 533 train 7.353205 (lr=6.0752e-04) (hash(x)=3641890)
558
+ 534 train 7.455035 (lr=6.0611e-04) (hash(x)=7994103)
559
+ 535 train 7.453609 (lr=6.0471e-04) (hash(x)=5358547)
560
+ 536 train 7.431534 (lr=6.0330e-04) (hash(x)=6797424)
561
+ 537 train 7.487229 (lr=6.0188e-04) (hash(x)=4876580)
562
+ 538 train 7.331033 (lr=6.0047e-04) (hash(x)=4867629)
563
+ 539 train 7.445199 (lr=5.9905e-04) (hash(x)=6601967)
564
+ 540 train 7.397231 (lr=5.9762e-04) (hash(x)=5480657)
565
+ 541 train 7.452838 (lr=5.9620e-04) (hash(x)=7127346)
566
+ 542 train 7.567822 (lr=5.9477e-04) (hash(x)=7224794)
567
+ 543 train 7.441413 (lr=5.9334e-04) (hash(x)=5967252)
568
+ 544 train 7.551286 (lr=5.9190e-04) (hash(x)=7415262)
569
+ 545 train 7.436929 (lr=5.9047e-04) (hash(x)=6522544)
570
+ 546 train 7.471183 (lr=5.8903e-04) (hash(x)=5181640)
571
+ 547 train 7.404430 (lr=5.8758e-04) (hash(x)=5916880)
572
+ 548 train 7.387734 (lr=5.8614e-04) (hash(x)=5620397)
573
+ 549 train 7.451409 (lr=5.8469e-04) (hash(x)=5994970)
574
+ 550 val loss 7.5364
575
+ 550 val perplexity 1875.0897
576
+ 550 train 7.521690 (lr=5.8323e-04) (hash(x)=6532717)
577
+ 551 train 7.515255 (lr=5.8178e-04) (hash(x)=6262762)
578
+ 552 train 7.379282 (lr=5.8032e-04) (hash(x)=5495326)
579
+ 553 train 7.646127 (lr=5.7886e-04) (hash(x)=5927868)
580
+ 554 train 7.468727 (lr=5.7740e-04) (hash(x)=5485433)
581
+ 555 train 7.425300 (lr=5.7593e-04) (hash(x)=6462425)
582
+ 556 train 7.462692 (lr=5.7447e-04) (hash(x)=6299518)
583
+ 557 train 7.525301 (lr=5.7300e-04) (hash(x)=5950376)
584
+ 558 train 7.657178 (lr=5.7152e-04) (hash(x)=6296440)
585
+ 559 train 7.771070 (lr=5.7005e-04) (hash(x)=6974102)
586
+ 560 train 7.558245 (lr=5.6857e-04) (hash(x)=6228679)
587
+ 561 train 7.341441 (lr=5.6709e-04) (hash(x)=6120319)
588
+ 562 train 7.707808 (lr=5.6560e-04) (hash(x)=6289494)
589
+ 563 train 7.537265 (lr=5.6412e-04) (hash(x)=6877818)
590
+ 564 train 7.493026 (lr=5.6263e-04) (hash(x)=6490976)
591
+ 565 train 7.597544 (lr=5.6114e-04) (hash(x)=6419450)
592
+ 566 train 7.581350 (lr=5.5965e-04) (hash(x)=6840857)
593
+ 567 train 7.517493 (lr=5.5815e-04) (hash(x)=4837686)
594
+ 568 train 7.536008 (lr=5.5666e-04) (hash(x)=5561493)
595
+ 569 train 7.474400 (lr=5.5516e-04) (hash(x)=5831242)
596
+ 570 train 7.561130 (lr=5.5365e-04) (hash(x)=6851511)
597
+ 571 train 7.625948 (lr=5.5215e-04) (hash(x)=6169616)
598
+ 572 train 7.550555 (lr=5.5064e-04) (hash(x)=5357151)
599
+ 573 train 7.421032 (lr=5.4913e-04) (hash(x)=5471028)
600
+ 574 train 7.468271 (lr=5.4762e-04) (hash(x)=5051894)
601
+ 575 train 7.468203 (lr=5.4611e-04) (hash(x)=4930450)
602
+ 576 train 7.563958 (lr=5.4460e-04) (hash(x)=6537527)
603
+ 577 train 7.500878 (lr=5.4308e-04) (hash(x)=6330572)
604
+ 578 train 7.625853 (lr=5.4156e-04) (hash(x)=5315155)
605
+ 579 train 7.492888 (lr=5.4004e-04) (hash(x)=5416079)
606
+ 580 train 7.441052 (lr=5.3852e-04) (hash(x)=6255865)
607
+ 581 train 7.667192 (lr=5.3699e-04) (hash(x)=6723036)
608
+ 582 train 7.450230 (lr=5.3547e-04) (hash(x)=7107190)
609
+ 583 train 7.546109 (lr=5.3394e-04) (hash(x)=5868958)
610
+ 584 train 7.643386 (lr=5.3241e-04) (hash(x)=7385923)
611
+ 585 train 7.460397 (lr=5.3088e-04) (hash(x)=5373053)
612
+ 586 train 7.474876 (lr=5.2934e-04) (hash(x)=5371748)
613
+ 587 train 7.615180 (lr=5.2781e-04) (hash(x)=5783717)
614
+ 588 train 7.403784 (lr=5.2627e-04) (hash(x)=6388848)
615
+ 589 train 7.390203 (lr=5.2473e-04) (hash(x)=4865068)
616
+ 590 train 7.347885 (lr=5.2319e-04) (hash(x)=5844961)
617
+ 591 train 7.448218 (lr=5.2165e-04) (hash(x)=5989273)
618
+ 592 train 7.384857 (lr=5.2011e-04) (hash(x)=5136202)
619
+ 593 train 7.517279 (lr=5.1856e-04) (hash(x)=5811889)
620
+ 594 train 7.469808 (lr=5.1702e-04) (hash(x)=6554415)
621
+ 595 train 7.511178 (lr=5.1547e-04) (hash(x)=5969272)
622
+ 596 train 7.480851 (lr=5.1392e-04) (hash(x)=7121518)
623
+ 597 train 7.936923 (lr=5.1237e-04) (hash(x)=6363768)
624
+ 598 train 7.542945 (lr=5.1082e-04) (hash(x)=7030284)
625
+ 599 train 7.640202 (lr=5.0926e-04) (hash(x)=7243339)
626
+ 600 val loss 7.5766
627
+ 600 val perplexity 1952.0062
628
+ 600 train 7.713429 (lr=5.0771e-04) (hash(x)=7199493)
629
+ 601 train 7.756603 (lr=5.0615e-04) (hash(x)=5819281)
630
+ 602 train 7.500884 (lr=5.0459e-04) (hash(x)=7285909)
631
+ 603 train 7.535044 (lr=5.0303e-04) (hash(x)=6730878)
632
+ 604 train 7.575915 (lr=5.0147e-04) (hash(x)=5200141)
633
+ 605 train 7.664170 (lr=4.9991e-04) (hash(x)=5208573)
634
+ 606 train 7.755271 (lr=4.9835e-04) (hash(x)=5976779)
635
+ 607 train 7.574374 (lr=4.9679e-04) (hash(x)=5921767)
636
+ 608 train 7.631646 (lr=4.9522e-04) (hash(x)=6596032)
637
+ 609 train 7.660711 (lr=4.9366e-04) (hash(x)=5428058)
638
+ 610 train 7.591249 (lr=4.9209e-04) (hash(x)=5748050)
639
+ 611 train 7.598941 (lr=4.9052e-04) (hash(x)=6397421)
640
+ 612 train 7.675171 (lr=4.8895e-04) (hash(x)=5638644)
641
+ 613 train 7.600860 (lr=4.8738e-04) (hash(x)=7317623)
642
+ 614 train 7.638929 (lr=4.8581e-04) (hash(x)=5254932)
643
+ 615 train 7.712118 (lr=4.8424e-04) (hash(x)=4945141)
644
+ 616 train 7.626518 (lr=4.8267e-04) (hash(x)=7093994)
645
+ 617 train 7.581876 (lr=4.8109e-04) (hash(x)=7511894)
646
+ 618 train 7.554394 (lr=4.7952e-04) (hash(x)=6706832)
647
+ 619 train 7.499073 (lr=4.7795e-04) (hash(x)=6175371)
648
+ 620 train 7.570307 (lr=4.7637e-04) (hash(x)=7089897)
649
+ 621 train 7.557417 (lr=4.7479e-04) (hash(x)=6828202)
650
+ 622 train 7.847225 (lr=4.7322e-04) (hash(x)=7095301)
651
+ 623 train 7.600407 (lr=4.7164e-04) (hash(x)=5652191)
652
+ 624 train 7.612131 (lr=4.7006e-04) (hash(x)=7336431)
653
+ 625 train 7.613230 (lr=4.6848e-04) (hash(x)=6426339)
654
+ 626 train 7.718176 (lr=4.6690e-04) (hash(x)=5874982)
655
+ 627 train 7.662855 (lr=4.6532e-04) (hash(x)=6978359)
656
+ 628 train 7.419153 (lr=4.6374e-04) (hash(x)=4438173)
657
+ 629 train 7.501123 (lr=4.6216e-04) (hash(x)=6475356)
658
+ 630 train 7.705752 (lr=4.6058e-04) (hash(x)=5030069)
659
+ 631 train 7.476147 (lr=4.5900e-04) (hash(x)=6507939)
660
+ 632 train 7.504259 (lr=4.5742e-04) (hash(x)=6373406)
661
+ 633 train 7.642270 (lr=4.5583e-04) (hash(x)=6623130)
662
+ 634 train 7.569536 (lr=4.5425e-04) (hash(x)=6276741)
663
+ 635 train 7.462178 (lr=4.5267e-04) (hash(x)=7169874)
664
+ 636 train 7.519847 (lr=4.5109e-04) (hash(x)=4947348)
665
+ 637 train 7.601785 (lr=4.4950e-04) (hash(x)=7407715)
666
+ 638 train 7.690968 (lr=4.4792e-04) (hash(x)=5453341)
667
+ 639 train 7.503454 (lr=4.4634e-04) (hash(x)=6520161)
668
+ 640 train 7.568887 (lr=4.4475e-04) (hash(x)=6011882)
669
+ 641 train 7.524736 (lr=4.4317e-04) (hash(x)=5573372)
670
+ 642 train 7.706033 (lr=4.4158e-04) (hash(x)=5344827)
671
+ 643 train 7.758078 (lr=4.4000e-04) (hash(x)=6127887)
672
+ 644 train 7.533940 (lr=4.3842e-04) (hash(x)=6052304)
673
+ 645 train 7.523339 (lr=4.3683e-04) (hash(x)=7039733)
674
+ 646 train 7.604224 (lr=4.3525e-04) (hash(x)=7746451)
675
+ 647 train 7.773371 (lr=4.3366e-04) (hash(x)=7694373)
676
+ 648 train 7.684050 (lr=4.3208e-04) (hash(x)=6911004)
677
+ 649 train 7.593968 (lr=4.3050e-04) (hash(x)=8055224)
678
+ 650 val loss 7.5403
679
+ 650 val perplexity 1882.3513
680
+ 650 train 7.619267 (lr=4.2891e-04) (hash(x)=7060476)
681
+ 651 train 7.578489 (lr=4.2733e-04) (hash(x)=6002317)
682
+ 652 train 7.588496 (lr=4.2575e-04) (hash(x)=6524390)
683
+ 653 train 7.595119 (lr=4.2417e-04) (hash(x)=8331083)
684
+ 654 train 7.621775 (lr=4.2258e-04) (hash(x)=7084852)
685
+ 655 train 7.548296 (lr=4.2100e-04) (hash(x)=6190174)
686
+ 656 train 7.515909 (lr=4.1942e-04) (hash(x)=5601780)
687
+ 657 train 7.655635 (lr=4.1784e-04) (hash(x)=4764034)
688
+ 658 train 7.549284 (lr=4.1626e-04) (hash(x)=6361606)
689
+ 659 train 7.531516 (lr=4.1468e-04) (hash(x)=5697877)
690
+ 660 train 7.583778 (lr=4.1310e-04) (hash(x)=6249793)
691
+ 661 train 7.630911 (lr=4.1152e-04) (hash(x)=7424709)
692
+ 662 train 7.640082 (lr=4.0994e-04) (hash(x)=8203076)
693
+ 663 train 7.401597 (lr=4.0836e-04) (hash(x)=5290019)
694
+ 664 train 7.453696 (lr=4.0678e-04) (hash(x)=6278399)
695
+ 665 train 7.536967 (lr=4.0521e-04) (hash(x)=5709192)
696
+ 666 train 7.619995 (lr=4.0363e-04) (hash(x)=5660292)
697
+ 667 train 7.582291 (lr=4.0205e-04) (hash(x)=5563697)
698
+ 668 train 7.569189 (lr=4.0048e-04) (hash(x)=6325245)
699
+ 669 train 7.566551 (lr=3.9891e-04) (hash(x)=6259741)
700
+ 670 train 7.687680 (lr=3.9733e-04) (hash(x)=6238355)
701
+ 671 train 7.637065 (lr=3.9576e-04) (hash(x)=7778802)
702
+ 672 train 7.574772 (lr=3.9419e-04) (hash(x)=6283496)
703
+ 673 train 7.647129 (lr=3.9262e-04) (hash(x)=5325614)
704
+ 674 train 7.524629 (lr=3.9105e-04) (hash(x)=5376057)
705
+ 675 train 7.512283 (lr=3.8948e-04) (hash(x)=5203215)
706
+ 676 train 7.588980 (lr=3.8791e-04) (hash(x)=4893030)
707
+ 677 train 7.534074 (lr=3.8634e-04) (hash(x)=7595573)
708
+ 678 train 7.620516 (lr=3.8478e-04) (hash(x)=7188575)
709
+ 679 train 7.625021 (lr=3.8321e-04) (hash(x)=6858739)
710
+ 680 train 7.526333 (lr=3.8165e-04) (hash(x)=7157032)
711
+ 681 train 7.630484 (lr=3.8009e-04) (hash(x)=5871492)
712
+ 682 train 7.502760 (lr=3.7853e-04) (hash(x)=5636927)
713
+ 683 train 7.478837 (lr=3.7697e-04) (hash(x)=6135391)
714
+ 684 train 7.573928 (lr=3.7541e-04) (hash(x)=3351432)
715
+ 685 train 7.546045 (lr=3.7385e-04) (hash(x)=6781160)
716
+ 686 train 7.584134 (lr=3.7229e-04) (hash(x)=5824767)
717
+ 687 train 7.584021 (lr=3.7074e-04) (hash(x)=4512723)
718
+ 688 train 7.529224 (lr=3.6918e-04) (hash(x)=4600884)
719
+ 689 train 7.438166 (lr=3.6763e-04) (hash(x)=5079210)
720
+ 690 train 7.516575 (lr=3.6608e-04) (hash(x)=4671963)
721
+ 691 train 7.379745 (lr=3.6453e-04) (hash(x)=4169508)
722
+ 692 train 7.645580 (lr=3.6298e-04) (hash(x)=5341676)
723
+ 693 train 7.423436 (lr=3.6144e-04) (hash(x)=7593391)
724
+ 694 train 7.550332 (lr=3.5989e-04) (hash(x)=6723106)
725
+ 695 train 7.653798 (lr=3.5835e-04) (hash(x)=5136994)
726
+ 696 train 7.613504 (lr=3.5681e-04) (hash(x)=6571636)
727
+ 697 train 7.619715 (lr=3.5527e-04) (hash(x)=6975234)
728
+ 698 train 7.436259 (lr=3.5373e-04) (hash(x)=4868479)
729
+ 699 train 7.762106 (lr=3.5219e-04) (hash(x)=7086593)
730
+ 700 val loss 7.5435
731
+ 700 val perplexity 1888.4423
732
+ 700 train 7.594830 (lr=3.5066e-04) (hash(x)=6870808)
733
+ 701 train 7.889438 (lr=3.4912e-04) (hash(x)=7186398)
734
+ 702 train 7.665181 (lr=3.4759e-04) (hash(x)=4907570)
735
+ 703 train 7.535140 (lr=3.4606e-04) (hash(x)=5812649)
736
+ 704 train 7.564381 (lr=3.4453e-04) (hash(x)=6577447)
737
+ 705 train 7.514263 (lr=3.4301e-04) (hash(x)=5923453)
738
+ 706 train 7.762703 (lr=3.4148e-04) (hash(x)=7282962)
739
+ 707 train 7.457568 (lr=3.3996e-04) (hash(x)=4767725)
740
+ 708 train 7.482709 (lr=3.3844e-04) (hash(x)=6894202)
741
+ 709 train 7.972928 (lr=3.3692e-04) (hash(x)=4904377)
742
+ 710 train 7.522717 (lr=3.3540e-04) (hash(x)=5810443)
743
+ 711 train 7.601889 (lr=3.3389e-04) (hash(x)=5193944)
744
+ 712 train 7.542549 (lr=3.3238e-04) (hash(x)=5188803)
745
+ 713 train 7.551065 (lr=3.3087e-04) (hash(x)=7235549)
746
+ 714 train 7.701838 (lr=3.2936e-04) (hash(x)=5692576)
747
+ 715 train 7.635889 (lr=3.2785e-04) (hash(x)=5103770)
748
+ 716 train 7.594498 (lr=3.2635e-04) (hash(x)=6084786)
749
+ 717 train 7.519592 (lr=3.2484e-04) (hash(x)=5949126)
750
+ 718 train 7.587924 (lr=3.2334e-04) (hash(x)=6321938)
751
+ 719 train 7.629817 (lr=3.2185e-04) (hash(x)=6227413)
752
+ 720 train 7.670659 (lr=3.2035e-04) (hash(x)=6795372)
753
+ 721 train 7.543125 (lr=3.1886e-04) (hash(x)=6506400)
754
+ 722 train 7.564490 (lr=3.1737e-04) (hash(x)=4595742)
755
+ 723 train 7.668403 (lr=3.1588e-04) (hash(x)=7342185)
756
+ 724 train 7.704750 (lr=3.1440e-04) (hash(x)=5762071)
757
+ 725 train 7.559199 (lr=3.1291e-04) (hash(x)=6337670)
758
+ 726 train 7.600758 (lr=3.1143e-04) (hash(x)=6831170)
759
+ 727 train 7.622196 (lr=3.0995e-04) (hash(x)=7080870)
760
+ 728 train 7.568775 (lr=3.0848e-04) (hash(x)=6449970)
761
+ 729 train 7.719101 (lr=3.0700e-04) (hash(x)=5588514)
762
+ 730 train 7.521279 (lr=3.0553e-04) (hash(x)=5379746)
763
+ 731 train 7.691173 (lr=3.0407e-04) (hash(x)=8049901)
764
+ 732 train 7.725621 (lr=3.0260e-04) (hash(x)=8398209)
765
+ 733 train 7.636964 (lr=3.0114e-04) (hash(x)=6152707)
766
+ 734 train 7.691932 (lr=2.9968e-04) (hash(x)=7682617)
767
+ 735 train 7.608679 (lr=2.9822e-04) (hash(x)=7053352)
768
+ 736 train 7.659449 (lr=2.9677e-04) (hash(x)=8491505)
769
+ 737 train 7.624277 (lr=2.9531e-04) (hash(x)=6102853)
770
+ 738 train 7.706184 (lr=2.9386e-04) (hash(x)=4166591)
771
+ 739 train 7.587080 (lr=2.9242e-04) (hash(x)=7501152)
772
+ 740 train 7.553897 (lr=2.9097e-04) (hash(x)=8231949)
773
+ 741 train 7.608720 (lr=2.8953e-04) (hash(x)=5741690)
774
+ 742 train 7.734529 (lr=2.8810e-04) (hash(x)=5420426)
775
+ 743 train 7.538952 (lr=2.8666e-04) (hash(x)=5808766)
776
+ 744 train 7.522714 (lr=2.8523e-04) (hash(x)=7533322)
777
+ 745 train 7.572011 (lr=2.8380e-04) (hash(x)=6350197)
778
+ 746 train 7.676644 (lr=2.8238e-04) (hash(x)=6207170)
779
+ 747 train 7.656901 (lr=2.8095e-04) (hash(x)=5811552)
780
+ 748 train 7.592969 (lr=2.7953e-04) (hash(x)=4540607)
781
+ 749 train 7.471862 (lr=2.7812e-04) (hash(x)=5599396)
782
+ 750 val loss 7.5669
783
+ 750 val perplexity 1933.0872
784
+ 750 train 7.470884 (lr=2.7670e-04) (hash(x)=6180984)
785
+ 751 train 7.634400 (lr=2.7529e-04) (hash(x)=6519152)
786
+ 752 train 7.590250 (lr=2.7389e-04) (hash(x)=9018585)
787
+ 753 train 7.648584 (lr=2.7248e-04) (hash(x)=5988202)
788
+ 754 train 7.660338 (lr=2.7108e-04) (hash(x)=7243449)
789
+ 755 train 7.485209 (lr=2.6969e-04) (hash(x)=4377360)
790
+ 756 train 7.573912 (lr=2.6829e-04) (hash(x)=5928295)
791
+ 757 train 7.528889 (lr=2.6690e-04) (hash(x)=4996189)
792
+ 758 train 7.556436 (lr=2.6551e-04) (hash(x)=6829435)
793
+ 759 train 7.637048 (lr=2.6413e-04) (hash(x)=6929283)
794
+ 760 train 7.717125 (lr=2.6275e-04) (hash(x)=7733662)
795
+ 761 train 7.672530 (lr=2.6137e-04) (hash(x)=5975918)
796
+ 762 train 7.542972 (lr=2.6000e-04) (hash(x)=5566111)
797
+ 763 train 7.625130 (lr=2.5863e-04) (hash(x)=7407511)
798
+ 764 train 7.557744 (lr=2.5726e-04) (hash(x)=5896879)
799
+ 765 train 7.522753 (lr=2.5590e-04) (hash(x)=5569196)
800
+ 766 train 7.716023 (lr=2.5454e-04) (hash(x)=6235218)
801
+ 767 train 7.649037 (lr=2.5319e-04) (hash(x)=7467630)
802
+ 768 train 7.633235 (lr=2.5183e-04) (hash(x)=5268563)
803
+ 769 train 7.809764 (lr=2.5048e-04) (hash(x)=5915215)
804
+ 770 train 7.636974 (lr=2.4914e-04) (hash(x)=5439067)
805
+ 771 train 7.587815 (lr=2.4780e-04) (hash(x)=6167203)
806
+ 772 train 7.648864 (lr=2.4646e-04) (hash(x)=4940526)
807
+ 773 train 7.692421 (lr=2.4513e-04) (hash(x)=5897107)
808
+ 774 train 7.550602 (lr=2.4380e-04) (hash(x)=6780557)
809
+ 775 train 7.648821 (lr=2.4247e-04) (hash(x)=7566733)
810
+ 776 train 7.772978 (lr=2.4115e-04) (hash(x)=7939982)
811
+ 777 train 7.545461 (lr=2.3983e-04) (hash(x)=6447254)
812
+ 778 train 7.529480 (lr=2.3852e-04) (hash(x)=5525831)
813
+ 779 train 7.470535 (lr=2.3720e-04) (hash(x)=5059170)
814
+ 780 train 7.568174 (lr=2.3590e-04) (hash(x)=7485229)
815
+ 781 train 7.679751 (lr=2.3460e-04) (hash(x)=5955219)
816
+ 782 train 7.548208 (lr=2.3330e-04) (hash(x)=6625876)
817
+ 783 train 7.839372 (lr=2.3200e-04) (hash(x)=7141085)
818
+ 784 train 7.850155 (lr=2.3071e-04) (hash(x)=5777157)
819
+ 785 train 7.506167 (lr=2.2942e-04) (hash(x)=6236468)
820
+ 786 train 7.354688 (lr=2.2814e-04) (hash(x)=5381032)
821
+ 787 train 7.533047 (lr=2.2686e-04) (hash(x)=6733716)
822
+ 788 train 7.781629 (lr=2.2559e-04) (hash(x)=7073122)
823
+ 789 train 7.737010 (lr=2.2432e-04) (hash(x)=7622163)
824
+ 790 train 7.787405 (lr=2.2305e-04) (hash(x)=6866834)
825
+ 791 train 7.729915 (lr=2.2179e-04) (hash(x)=7008739)
826
+ 792 train 7.652678 (lr=2.2053e-04) (hash(x)=7839835)
827
+ 793 train 7.538663 (lr=2.1928e-04) (hash(x)=5313637)
828
+ 794 train 7.699767 (lr=2.1803e-04) (hash(x)=6410359)
829
+ 795 train 7.549256 (lr=2.1678e-04) (hash(x)=5721008)
830
+ 796 train 7.570126 (lr=2.1554e-04) (hash(x)=6137936)
831
+ 797 train 7.436895 (lr=2.1431e-04) (hash(x)=6881406)
832
+ 798 train 7.639191 (lr=2.1308e-04) (hash(x)=6088931)
833
+ 799 train 7.654595 (lr=2.1185e-04) (hash(x)=5689728)
834
+ 800 val loss 7.5694
835
+ 800 val perplexity 1937.9481
836
+ 800 train 7.699054 (lr=2.1063e-04) (hash(x)=5800778)
837
+ 801 train 7.616796 (lr=2.0941e-04) (hash(x)=4419021)
838
+ 802 train 7.570331 (lr=2.0819e-04) (hash(x)=5544294)
839
+ 803 train 7.629237 (lr=2.0698e-04) (hash(x)=5663946)
840
+ 804 train 7.556306 (lr=2.0578e-04) (hash(x)=4573911)
841
+ 805 train 7.495501 (lr=2.0458e-04) (hash(x)=5879465)
842
+ 806 train 7.587636 (lr=2.0338e-04) (hash(x)=5641517)
843
+ 807 train 7.739289 (lr=2.0219e-04) (hash(x)=6390336)
844
+ 808 train 7.626567 (lr=2.0100e-04) (hash(x)=8412819)
845
+ 809 train 7.741934 (lr=1.9982e-04) (hash(x)=7005536)
846
+ 810 train 7.492167 (lr=1.9864e-04) (hash(x)=6630048)
847
+ 811 train 7.613661 (lr=1.9747e-04) (hash(x)=7267424)
848
+ 812 train 7.537196 (lr=1.9630e-04) (hash(x)=6497645)
849
+ 813 train 7.475287 (lr=1.9514e-04) (hash(x)=5498697)
850
+ 814 train 7.564942 (lr=1.9398e-04) (hash(x)=6323524)
851
+ 815 train 7.520209 (lr=1.9283e-04) (hash(x)=6529460)
852
+ 816 train 7.392698 (lr=1.9168e-04) (hash(x)=5498587)
853
+ 817 train 7.558104 (lr=1.9053e-04) (hash(x)=6828893)
854
+ 818 train 7.670970 (lr=1.8939e-04) (hash(x)=6591055)
855
+ 819 train 7.529550 (lr=1.8826e-04) (hash(x)=6739572)
856
+ 820 train 7.538704 (lr=1.8713e-04) (hash(x)=5614231)
857
+ 821 train 7.594312 (lr=1.8600e-04) (hash(x)=5462474)
858
+ 822 train 7.572201 (lr=1.8488e-04) (hash(x)=5906742)
859
+ 823 train 7.656599 (lr=1.8377e-04) (hash(x)=8249268)
860
+ 824 train 7.658707 (lr=1.8266e-04) (hash(x)=4886590)
861
+ 825 train 7.435260 (lr=1.8155e-04) (hash(x)=6736150)
862
+ 826 train 7.547906 (lr=1.8045e-04) (hash(x)=6601014)
863
+ 827 train 7.618305 (lr=1.7936e-04) (hash(x)=8283020)
864
+ 828 train 7.575827 (lr=1.7827e-04) (hash(x)=6526568)
865
+ 829 train 7.655956 (lr=1.7718e-04) (hash(x)=6609259)
866
+ 830 train 7.526250 (lr=1.7610e-04) (hash(x)=6079648)
867
+ 831 train 7.513031 (lr=1.7503e-04) (hash(x)=6607178)
868
+ 832 train 7.590842 (lr=1.7396e-04) (hash(x)=6815798)
869
+ 833 train 7.819019 (lr=1.7289e-04) (hash(x)=7277017)
870
+ 834 train 7.514294 (lr=1.7183e-04) (hash(x)=5326415)
871
+ 835 train 7.571322 (lr=1.7078e-04) (hash(x)=4215944)
872
+ 836 train 7.645205 (lr=1.6973e-04) (hash(x)=4689010)
873
+ 837 train 7.602386 (lr=1.6869e-04) (hash(x)=5690469)
874
+ 838 train 7.492938 (lr=1.6765e-04) (hash(x)=6699212)
875
+ 839 train 7.486357 (lr=1.6661e-04) (hash(x)=5494876)
876
+ 840 train 7.850960 (lr=1.6559e-04) (hash(x)=10756054)
877
+ 841 train 7.528389 (lr=1.6456e-04) (hash(x)=6553219)
878
+ 842 train 7.372905 (lr=1.6355e-04) (hash(x)=7323893)
879
+ 843 train 7.828763 (lr=1.6253e-04) (hash(x)=5481970)
880
+ 844 train 7.646289 (lr=1.6153e-04) (hash(x)=6102855)
881
+ 845 train 7.595944 (lr=1.6053e-04) (hash(x)=6889406)
882
+ 846 train 7.642897 (lr=1.5953e-04) (hash(x)=4468013)
883
+ 847 train 7.456985 (lr=1.5854e-04) (hash(x)=6621321)
884
+ 848 train 7.597471 (lr=1.5756e-04) (hash(x)=6185717)
885
+ 849 train 7.491279 (lr=1.5658e-04) (hash(x)=5594828)
886
+ 850 val loss 7.5465
887
+ 850 val perplexity 1894.0236
888
+ 850 train 7.604753 (lr=1.5560e-04) (hash(x)=6530853)
889
+ 851 train 7.543923 (lr=1.5463e-04) (hash(x)=5688551)
890
+ 852 train 7.648221 (lr=1.5367e-04) (hash(x)=6979027)
891
+ 853 train 7.537507 (lr=1.5271e-04) (hash(x)=6065040)
892
+ 854 train 7.496067 (lr=1.5176e-04) (hash(x)=6836310)
893
+ 855 train 7.564645 (lr=1.5082e-04) (hash(x)=4871742)
894
+ 856 train 7.477651 (lr=1.4988e-04) (hash(x)=7461300)
895
+ 857 train 7.612236 (lr=1.4894e-04) (hash(x)=5073610)
896
+ 858 train 7.540531 (lr=1.4801e-04) (hash(x)=6725326)
897
+ 859 train 7.656981 (lr=1.4709e-04) (hash(x)=6114900)
898
+ 860 train 7.639676 (lr=1.4617e-04) (hash(x)=5249155)
899
+ 861 train 7.636175 (lr=1.4526e-04) (hash(x)=8179785)
900
+ 862 train 7.542575 (lr=1.4435e-04) (hash(x)=5588627)
901
+ 863 train 7.563974 (lr=1.4345e-04) (hash(x)=7084451)
902
+ 864 train 7.760869 (lr=1.4255e-04) (hash(x)=7478960)
903
+ 865 train 7.620910 (lr=1.4166e-04) (hash(x)=5609947)
904
+ 866 train 7.596589 (lr=1.4078e-04) (hash(x)=5741827)
905
+ 867 train 7.565795 (lr=1.3990e-04) (hash(x)=3359718)
906
+ 868 train 7.474165 (lr=1.3903e-04) (hash(x)=5480813)
907
+ 869 train 7.559143 (lr=1.3816e-04) (hash(x)=4774534)
908
+ 870 train 7.646445 (lr=1.3730e-04) (hash(x)=5895917)
909
+ 871 train 7.514635 (lr=1.3645e-04) (hash(x)=4902539)
910
+ 872 train 7.577642 (lr=1.3560e-04) (hash(x)=5734618)
911
+ 873 train 7.517354 (lr=1.3476e-04) (hash(x)=7086920)
912
+ 874 train 7.754157 (lr=1.3392e-04) (hash(x)=5338668)
913
+ 875 train 7.957647 (lr=1.3309e-04) (hash(x)=5238889)
914
+ 876 train 7.488467 (lr=1.3227e-04) (hash(x)=6244468)
915
+ 877 train 7.530331 (lr=1.3145e-04) (hash(x)=6715860)
916
+ 878 train 7.642595 (lr=1.3063e-04) (hash(x)=6216047)
917
+ 879 train 7.621879 (lr=1.2983e-04) (hash(x)=6565285)
918
+ 880 train 7.569536 (lr=1.2903e-04) (hash(x)=5257335)
919
+ 881 train 7.569623 (lr=1.2823e-04) (hash(x)=6474071)
920
+ 882 train 7.612330 (lr=1.2744e-04) (hash(x)=5705620)
921
+ 883 train 7.516779 (lr=1.2666e-04) (hash(x)=7193697)
922
+ 884 train 7.593920 (lr=1.2588e-04) (hash(x)=6264039)
923
+ 885 train 7.585778 (lr=1.2511e-04) (hash(x)=6303643)
924
+ 886 train 7.519769 (lr=1.2435e-04) (hash(x)=6120402)
925
+ 887 train 7.534740 (lr=1.2359e-04) (hash(x)=4158764)
926
+ 888 train 7.299137 (lr=1.2284e-04) (hash(x)=6173916)
927
+ 889 train 7.556824 (lr=1.2209e-04) (hash(x)=6211683)
928
+ 890 train 7.588429 (lr=1.2135e-04) (hash(x)=4950548)
929
+ 891 train 7.433133 (lr=1.2062e-04) (hash(x)=4885446)
930
+ 892 train 7.554171 (lr=1.1989e-04) (hash(x)=8117411)
931
+ 893 train 7.588366 (lr=1.1917e-04) (hash(x)=7446434)
932
+ 894 train 7.481071 (lr=1.1845e-04) (hash(x)=5371870)
933
+ 895 train 7.447979 (lr=1.1774e-04) (hash(x)=5022800)
934
+ 896 train 7.460223 (lr=1.1704e-04) (hash(x)=5666758)
935
+ 897 train 7.444650 (lr=1.1634e-04) (hash(x)=5872751)
936
+ 898 train 7.886844 (lr=1.1565e-04) (hash(x)=8428026)
937
+ 899 train 7.704290 (lr=1.1497e-04) (hash(x)=5222699)
938
+ 900 val loss 7.5360
939
+ 900 val perplexity 1874.3691
940
+ 900 train 7.567203 (lr=1.1429e-04) (hash(x)=6048200)
941
+ 901 train 7.434710 (lr=1.1362e-04) (hash(x)=5125556)
942
+ 902 train 7.446784 (lr=1.1295e-04) (hash(x)=6198583)
943
+ 903 train 7.427861 (lr=1.1229e-04) (hash(x)=5112805)
944
+ 904 train 7.518618 (lr=1.1164e-04) (hash(x)=6212803)
945
+ 905 train 7.551873 (lr=1.1099e-04) (hash(x)=5642978)
946
+ 906 train 7.539564 (lr=1.1036e-04) (hash(x)=7307168)
947
+ 907 train 7.478080 (lr=1.0972e-04) (hash(x)=6115590)
948
+ 908 train 7.395474 (lr=1.0909e-04) (hash(x)=7014937)
949
+ 909 train 7.720736 (lr=1.0847e-04) (hash(x)=6374852)
950
+ 910 train 7.478175 (lr=1.0786e-04) (hash(x)=5780051)
951
+ 911 train 7.515276 (lr=1.0725e-04) (hash(x)=4762818)
952
+ 912 train 7.367507 (lr=1.0665e-04) (hash(x)=6585420)
953
+ 913 train 7.656269 (lr=1.0606e-04) (hash(x)=6291590)
954
+ 914 train 7.710505 (lr=1.0547e-04) (hash(x)=5941561)
955
+ 915 train 7.522068 (lr=1.0489e-04) (hash(x)=6824617)
956
+ 916 train 7.601200 (lr=1.0431e-04) (hash(x)=5763341)
957
+ 917 train 7.625293 (lr=1.0374e-04) (hash(x)=6030150)
958
+ 918 train 7.545836 (lr=1.0318e-04) (hash(x)=5325490)
959
+ 919 train 7.575538 (lr=1.0262e-04) (hash(x)=6373236)
960
+ 920 train 7.575417 (lr=1.0207e-04) (hash(x)=6189151)
961
+ 921 train 7.535617 (lr=1.0153e-04) (hash(x)=6645157)
962
+ 922 train 7.543871 (lr=1.0099e-04) (hash(x)=6478394)
963
+ 923 train 7.666284 (lr=1.0046e-04) (hash(x)=5615918)
964
+ 924 train 7.567281 (lr=9.9941e-05) (hash(x)=6376988)
965
+ 925 train 7.598716 (lr=9.9425e-05) (hash(x)=5456350)
966
+ 926 train 7.626200 (lr=9.8915e-05) (hash(x)=5491217)
967
+ 927 train 7.442505 (lr=9.8411e-05) (hash(x)=6871163)
968
+ 928 train 7.462374 (lr=9.7915e-05) (hash(x)=7003708)
969
+ 929 train 7.563483 (lr=9.7424e-05) (hash(x)=7271154)
970
+ 930 train 7.542067 (lr=9.6941e-05) (hash(x)=5378966)
971
+ 931 train 7.527862 (lr=9.6464e-05) (hash(x)=5912082)
972
+ 932 train 7.435543 (lr=9.5994e-05) (hash(x)=5199606)
973
+ 933 train 7.481464 (lr=9.5530e-05) (hash(x)=5195464)
974
+ 934 train 7.430543 (lr=9.5073e-05) (hash(x)=7534052)
975
+ 935 train 7.438519 (lr=9.4623e-05) (hash(x)=6594135)
976
+ 936 train 7.518635 (lr=9.4180e-05) (hash(x)=6782051)
977
+ 937 train 7.622554 (lr=9.3743e-05) (hash(x)=6485577)
978
+ 938 train 7.572618 (lr=9.3313e-05) (hash(x)=6356813)
979
+ 939 train 7.471385 (lr=9.2889e-05) (hash(x)=6903068)
980
+ 940 train 7.529856 (lr=9.2473e-05) (hash(x)=6200862)
981
+ 941 train 7.544037 (lr=9.2063e-05) (hash(x)=5873929)
982
+ 942 train 7.543595 (lr=9.1659e-05) (hash(x)=5624440)
983
+ 943 train 7.528181 (lr=9.1263e-05) (hash(x)=7166972)
984
+ 944 train 7.639604 (lr=9.0873e-05) (hash(x)=6973234)
985
+ 945 train 7.460605 (lr=9.0490e-05) (hash(x)=6626956)
986
+ 946 train 7.597321 (lr=9.0114e-05) (hash(x)=5265592)
987
+ 947 train 7.341638 (lr=8.9744e-05) (hash(x)=5569439)
988
+ 948 train 7.472672 (lr=8.9382e-05) (hash(x)=4981059)
989
+ 949 train 7.480445 (lr=8.9026e-05) (hash(x)=7488974)
990
+ 950 val loss 7.5422
991
+ 950 val perplexity 1886.0386
992
+ 950 train 7.568463 (lr=8.8677e-05) (hash(x)=5160526)
993
+ 951 train 7.497169 (lr=8.8335e-05) (hash(x)=5340877)
994
+ 952 train 7.351625 (lr=8.7999e-05) (hash(x)=6333269)
995
+ 953 train 7.543328 (lr=8.7670e-05) (hash(x)=7205151)
996
+ 954 train 7.601792 (lr=8.7349e-05) (hash(x)=7256545)
997
+ 955 train 7.469706 (lr=8.7034e-05) (hash(x)=6374904)
998
+ 956 train 7.528821 (lr=8.6725e-05) (hash(x)=7124927)
999
+ 957 train 7.485075 (lr=8.6424e-05) (hash(x)=5263132)
1000
+ 958 train 7.662750 (lr=8.6130e-05) (hash(x)=8427208)
1001
+ 959 train 7.471940 (lr=8.5842e-05) (hash(x)=6884232)
1002
+ 960 train 7.439706 (lr=8.5561e-05) (hash(x)=7416977)
1003
+ 961 train 7.561843 (lr=8.5287e-05) (hash(x)=6733050)
1004
+ 962 train 7.474877 (lr=8.5020e-05) (hash(x)=5251404)
1005
+ 963 train 7.587639 (lr=8.4760e-05) (hash(x)=8162997)
1006
+ 964 train 7.726987 (lr=8.4507e-05) (hash(x)=7234382)
1007
+ 965 train 7.632571 (lr=8.4260e-05) (hash(x)=6255275)
1008
+ 966 train 7.590325 (lr=8.4021e-05) (hash(x)=4874742)
1009
+ 967 train 7.536074 (lr=8.3788e-05) (hash(x)=6303037)
1010
+ 968 train 7.487821 (lr=8.3563e-05) (hash(x)=5002696)
1011
+ 969 train 7.531425 (lr=8.3344e-05) (hash(x)=7636268)
1012
+ 970 train 7.504858 (lr=8.3132e-05) (hash(x)=5843590)
1013
+ 971 train 7.570895 (lr=8.2927e-05) (hash(x)=5122289)
1014
+ 972 train 7.461473 (lr=8.2729e-05) (hash(x)=5785819)
1015
+ 973 train 7.532471 (lr=8.2537e-05) (hash(x)=7350343)
1016
+ 974 train 7.350913 (lr=8.2353e-05) (hash(x)=7612943)
1017
+ 975 train 7.612738 (lr=8.2176e-05) (hash(x)=7955067)
1018
+ 976 train 7.517225 (lr=8.2005e-05) (hash(x)=5520886)
1019
+ 977 train 7.408623 (lr=8.1842e-05) (hash(x)=5299200)
1020
+ 978 train 7.432864 (lr=8.1685e-05) (hash(x)=7820550)
1021
+ 979 train 7.397061 (lr=8.1536e-05) (hash(x)=3895312)
1022
+ 980 train 7.522735 (lr=8.1393e-05) (hash(x)=4726896)
1023
+ 981 train 7.307961 (lr=8.1257e-05) (hash(x)=5061167)
1024
+ 982 train 7.491364 (lr=8.1128e-05) (hash(x)=7006804)
1025
+ 983 train 7.492296 (lr=8.1007e-05) (hash(x)=7392600)
1026
+ 984 train 7.677797 (lr=8.0892e-05) (hash(x)=7717172)
1027
+ 985 train 7.781394 (lr=8.0784e-05) (hash(x)=5800587)
1028
+ 986 train 7.508896 (lr=8.0683e-05) (hash(x)=5813784)
1029
+ 987 train 7.478319 (lr=8.0589e-05) (hash(x)=5513267)
1030
+ 988 train 7.476517 (lr=8.0502e-05) (hash(x)=7679326)
1031
+ 989 train 7.545918 (lr=8.0422e-05) (hash(x)=4820278)
1032
+ 990 train 7.560172 (lr=8.0348e-05) (hash(x)=5950845)
1033
+ 991 train 7.474019 (lr=8.0282e-05) (hash(x)=6988067)
1034
+ 992 train 7.404217 (lr=8.0223e-05) (hash(x)=4536923)
1035
+ 993 train 7.552340 (lr=8.0171e-05) (hash(x)=6874134)
1036
+ 994 train 7.774990 (lr=8.0125e-05) (hash(x)=7252013)
1037
+ 995 train 7.430938 (lr=8.0087e-05) (hash(x)=6180446)
1038
+ 996 train 7.662426 (lr=8.0056e-05) (hash(x)=5936769)
1039
+ 997 train 7.582118 (lr=8.0031e-05) (hash(x)=5627313)
1040
+ 998 train 7.533885 (lr=8.0014e-05) (hash(x)=7541743)
1041
+ 999 val loss 7.5209
1042
+ 999 val perplexity 1846.1749
1043
+ 999 train 7.867359 (lr=8.0003e-05) (hash(x)=9812956)
lr8e-4_total_batch_size40960_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5e1a67360050786129f8b5b4b6051a2f00394ea54da30d422df67da36fc3cb6
3
+ size 36163010
lr8e-4_total_batch_size40960_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a28cb53ef695d20b89fb11f564372a82660f7f1e4f9c40f24f4510b014f7595
3
+ size 70764358