andrew-healey commited on
Commit
f30fbaf
·
verified ·
1 Parent(s): 0199a3f

Upload folder using huggingface_hub

Browse files
lr4e-4_total_batch_size20480_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr4e-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": 0.0004, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr4e-4_total_batch_size20480_seq_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
lr4e-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=1.3986e-06) (hash(x)=5113283)
5
+ 1 train 11.672828 (lr=2.7972e-06) (hash(x)=5534142)
6
+ 2 train 11.699075 (lr=4.1958e-06) (hash(x)=6763231)
7
+ 3 train 11.663286 (lr=5.5944e-06) (hash(x)=6328012)
8
+ 4 train 11.656357 (lr=6.9930e-06) (hash(x)=7052161)
9
+ 5 train 11.640671 (lr=8.3916e-06) (hash(x)=5556098)
10
+ 6 train 11.608338 (lr=9.7902e-06) (hash(x)=4696943)
11
+ 7 train 11.559019 (lr=1.1189e-05) (hash(x)=6219138)
12
+ 8 train 11.508968 (lr=1.2587e-05) (hash(x)=6049878)
13
+ 9 train 11.499625 (lr=1.3986e-05) (hash(x)=5491972)
14
+ 10 train 11.449233 (lr=1.5385e-05) (hash(x)=8075458)
15
+ 11 train 11.378419 (lr=1.6783e-05) (hash(x)=5917741)
16
+ 12 train 11.348023 (lr=1.8182e-05) (hash(x)=5750403)
17
+ 13 train 11.303619 (lr=1.9580e-05) (hash(x)=6112458)
18
+ 14 train 11.241801 (lr=2.0979e-05) (hash(x)=5806490)
19
+ 15 train 11.150006 (lr=2.2378e-05) (hash(x)=6568379)
20
+ 16 train 11.067209 (lr=2.3776e-05) (hash(x)=7212403)
21
+ 17 train 10.995203 (lr=2.5175e-05) (hash(x)=7386082)
22
+ 18 train 10.955493 (lr=2.6573e-05) (hash(x)=6359743)
23
+ 19 train 10.880232 (lr=2.7972e-05) (hash(x)=6151674)
24
+ 20 train 10.850606 (lr=2.9371e-05) (hash(x)=7168704)
25
+ 21 train 10.788665 (lr=3.0769e-05) (hash(x)=7859060)
26
+ 22 train 10.694252 (lr=3.2168e-05) (hash(x)=6976239)
27
+ 23 train 10.683619 (lr=3.3566e-05) (hash(x)=6391506)
28
+ 24 train 10.643327 (lr=3.4965e-05) (hash(x)=5245999)
29
+ 25 train 10.593281 (lr=3.6364e-05) (hash(x)=5985675)
30
+ 26 train 10.498267 (lr=3.7762e-05) (hash(x)=4355733)
31
+ 27 train 10.631542 (lr=3.9161e-05) (hash(x)=6984772)
32
+ 28 train 10.496611 (lr=4.0559e-05) (hash(x)=6244096)
33
+ 29 train 10.495267 (lr=4.1958e-05) (hash(x)=6603717)
34
+ 30 train 10.398873 (lr=4.3357e-05) (hash(x)=5669522)
35
+ 31 train 10.381268 (lr=4.4755e-05) (hash(x)=6190579)
36
+ 32 train 10.434013 (lr=4.6154e-05) (hash(x)=4912983)
37
+ 33 train 10.316533 (lr=4.7552e-05) (hash(x)=7276303)
38
+ 34 train 10.372952 (lr=4.8951e-05) (hash(x)=5479947)
39
+ 35 train 10.459206 (lr=5.0350e-05) (hash(x)=6630378)
40
+ 36 train 10.396424 (lr=5.1748e-05) (hash(x)=8097010)
41
+ 37 train 10.448638 (lr=5.3147e-05) (hash(x)=9989744)
42
+ 38 train 10.327752 (lr=5.4545e-05) (hash(x)=6694932)
43
+ 39 train 10.256584 (lr=5.5944e-05) (hash(x)=6317894)
44
+ 40 train 10.175003 (lr=5.7343e-05) (hash(x)=5046327)
45
+ 41 train 10.250850 (lr=5.8741e-05) (hash(x)=7947703)
46
+ 42 train 10.121830 (lr=6.0140e-05) (hash(x)=5916068)
47
+ 43 train 10.215761 (lr=6.1538e-05) (hash(x)=5777498)
48
+ 44 train 10.148030 (lr=6.2937e-05) (hash(x)=5753092)
49
+ 45 train 10.130989 (lr=6.4336e-05) (hash(x)=7723226)
50
+ 46 train 10.124299 (lr=6.5734e-05) (hash(x)=7007016)
51
+ 47 train 10.047805 (lr=6.7133e-05) (hash(x)=6491229)
52
+ 48 train 10.097555 (lr=6.8531e-05) (hash(x)=6390752)
53
+ 49 train 10.204393 (lr=6.9930e-05) (hash(x)=6019389)
54
+ 50 val loss 10.0811
55
+ 50 val perplexity 23886.7793
56
+ 50 train 10.069889 (lr=7.1329e-05) (hash(x)=7662067)
57
+ 51 train 10.069789 (lr=7.2727e-05) (hash(x)=4648609)
58
+ 52 train 10.007125 (lr=7.4126e-05) (hash(x)=4725966)
59
+ 53 train 9.967831 (lr=7.5524e-05) (hash(x)=5558355)
60
+ 54 train 10.001527 (lr=7.6923e-05) (hash(x)=6524933)
61
+ 55 train 9.953378 (lr=7.8322e-05) (hash(x)=6369326)
62
+ 56 train 9.915675 (lr=7.9720e-05) (hash(x)=6022625)
63
+ 57 train 9.855737 (lr=8.1119e-05) (hash(x)=4906853)
64
+ 58 train 9.688809 (lr=8.2517e-05) (hash(x)=5589269)
65
+ 59 train 9.800064 (lr=8.3916e-05) (hash(x)=6437376)
66
+ 60 train 9.826277 (lr=8.5315e-05) (hash(x)=7304892)
67
+ 61 train 9.777195 (lr=8.6713e-05) (hash(x)=7084093)
68
+ 62 train 9.788608 (lr=8.8112e-05) (hash(x)=5908178)
69
+ 63 train 9.581896 (lr=8.9510e-05) (hash(x)=7048804)
70
+ 64 train 9.655831 (lr=9.0909e-05) (hash(x)=6085549)
71
+ 65 train 9.653721 (lr=9.2308e-05) (hash(x)=5752594)
72
+ 66 train 9.456213 (lr=9.3706e-05) (hash(x)=7662769)
73
+ 67 train 9.309429 (lr=9.5105e-05) (hash(x)=6393520)
74
+ 68 train 9.551364 (lr=9.6503e-05) (hash(x)=5942867)
75
+ 69 train 9.532688 (lr=9.7902e-05) (hash(x)=6793550)
76
+ 70 train 9.377486 (lr=9.9301e-05) (hash(x)=7105976)
77
+ 71 train 9.401495 (lr=1.0070e-04) (hash(x)=6581348)
78
+ 72 train 9.338243 (lr=1.0210e-04) (hash(x)=5555598)
79
+ 73 train 9.265688 (lr=1.0350e-04) (hash(x)=6235837)
80
+ 74 train 9.200271 (lr=1.0490e-04) (hash(x)=5591770)
81
+ 75 train 9.390544 (lr=1.0629e-04) (hash(x)=7083021)
82
+ 76 train 9.214975 (lr=1.0769e-04) (hash(x)=7281348)
83
+ 77 train 9.240290 (lr=1.0909e-04) (hash(x)=6155821)
84
+ 78 train 9.080822 (lr=1.1049e-04) (hash(x)=5736241)
85
+ 79 train 8.993293 (lr=1.1189e-04) (hash(x)=5657123)
86
+ 80 train 8.892729 (lr=1.1329e-04) (hash(x)=6452095)
87
+ 81 train 8.879316 (lr=1.1469e-04) (hash(x)=5312267)
88
+ 82 train 8.814240 (lr=1.1608e-04) (hash(x)=5434493)
89
+ 83 train 8.938129 (lr=1.1748e-04) (hash(x)=6830813)
90
+ 84 train 8.809291 (lr=1.1888e-04) (hash(x)=4752771)
91
+ 85 train 8.839609 (lr=1.2028e-04) (hash(x)=6424933)
92
+ 86 train 8.738819 (lr=1.2168e-04) (hash(x)=5786517)
93
+ 87 train 8.764196 (lr=1.2308e-04) (hash(x)=6680023)
94
+ 88 train 8.645253 (lr=1.2448e-04) (hash(x)=6474729)
95
+ 89 train 8.689062 (lr=1.2587e-04) (hash(x)=7523279)
96
+ 90 train 8.485591 (lr=1.2727e-04) (hash(x)=5508768)
97
+ 91 train 8.615630 (lr=1.2867e-04) (hash(x)=6635925)
98
+ 92 train 8.540660 (lr=1.3007e-04) (hash(x)=6663801)
99
+ 93 train 8.502014 (lr=1.3147e-04) (hash(x)=5869441)
100
+ 94 train 8.503474 (lr=1.3287e-04) (hash(x)=6698878)
101
+ 95 train 8.440034 (lr=1.3427e-04) (hash(x)=6349004)
102
+ 96 train 8.428391 (lr=1.3566e-04) (hash(x)=4986137)
103
+ 97 train 8.213653 (lr=1.3706e-04) (hash(x)=6008454)
104
+ 98 train 8.250460 (lr=1.3846e-04) (hash(x)=6550770)
105
+ 99 train 8.353949 (lr=1.3986e-04) (hash(x)=7682741)
106
+ 100 val loss 8.3184
107
+ 100 val perplexity 4098.7666
108
+ 100 train 8.396103 (lr=1.4126e-04) (hash(x)=7892158)
109
+ 101 train 8.275209 (lr=1.4266e-04) (hash(x)=7794028)
110
+ 102 train 8.269009 (lr=1.4406e-04) (hash(x)=7093523)
111
+ 103 train 8.260232 (lr=1.4545e-04) (hash(x)=6166062)
112
+ 104 train 8.194337 (lr=1.4685e-04) (hash(x)=6645781)
113
+ 105 train 8.170750 (lr=1.4825e-04) (hash(x)=4811192)
114
+ 106 train 8.084870 (lr=1.4965e-04) (hash(x)=5970866)
115
+ 107 train 8.108835 (lr=1.5105e-04) (hash(x)=7822680)
116
+ 108 train 8.034785 (lr=1.5245e-04) (hash(x)=6317317)
117
+ 109 train 8.106247 (lr=1.5385e-04) (hash(x)=6431409)
118
+ 110 train 8.122334 (lr=1.5524e-04) (hash(x)=6653337)
119
+ 111 train 8.042282 (lr=1.5664e-04) (hash(x)=5323032)
120
+ 112 train 7.955510 (lr=1.5804e-04) (hash(x)=5125339)
121
+ 113 train 7.916279 (lr=1.5944e-04) (hash(x)=7363286)
122
+ 114 train 8.207185 (lr=1.6084e-04) (hash(x)=7670050)
123
+ 115 train 7.988898 (lr=1.6224e-04) (hash(x)=5616075)
124
+ 116 train 8.030732 (lr=1.6364e-04) (hash(x)=6593764)
125
+ 117 train 8.082897 (lr=1.6503e-04) (hash(x)=5672608)
126
+ 118 train 7.939550 (lr=1.6643e-04) (hash(x)=6828653)
127
+ 119 train 7.930516 (lr=1.6783e-04) (hash(x)=6839867)
128
+ 120 train 8.023301 (lr=1.6923e-04) (hash(x)=8426992)
129
+ 121 train 7.712131 (lr=1.7063e-04) (hash(x)=4916928)
130
+ 122 train 7.917670 (lr=1.7203e-04) (hash(x)=6484050)
131
+ 123 train 7.983311 (lr=1.7343e-04) (hash(x)=8353379)
132
+ 124 train 7.881751 (lr=1.7483e-04) (hash(x)=6552510)
133
+ 125 train 7.879014 (lr=1.7622e-04) (hash(x)=4838871)
134
+ 126 train 7.617235 (lr=1.7762e-04) (hash(x)=3021697)
135
+ 127 train 7.751260 (lr=1.7902e-04) (hash(x)=6795665)
136
+ 128 train 7.748051 (lr=1.8042e-04) (hash(x)=5486935)
137
+ 129 train 7.928158 (lr=1.8182e-04) (hash(x)=6678038)
138
+ 130 train 7.940475 (lr=1.8322e-04) (hash(x)=6657714)
139
+ 131 train 7.849248 (lr=1.8462e-04) (hash(x)=6112215)
140
+ 132 train 7.868167 (lr=1.8601e-04) (hash(x)=8353143)
141
+ 133 train 7.882381 (lr=1.8741e-04) (hash(x)=5915361)
142
+ 134 train 7.779057 (lr=1.8881e-04) (hash(x)=6254885)
143
+ 135 train 7.807247 (lr=1.9021e-04) (hash(x)=5771588)
144
+ 136 train 7.778200 (lr=1.9161e-04) (hash(x)=5561507)
145
+ 137 train 7.730793 (lr=1.9301e-04) (hash(x)=5688829)
146
+ 138 train 7.696267 (lr=1.9441e-04) (hash(x)=5371951)
147
+ 139 train 7.855422 (lr=1.9580e-04) (hash(x)=5443305)
148
+ 140 train 7.853623 (lr=1.9720e-04) (hash(x)=5016757)
149
+ 141 train 7.847708 (lr=1.9860e-04) (hash(x)=6901933)
150
+ 142 train 7.797873 (lr=2.0000e-04) (hash(x)=6019085)
151
+ 143 train 7.649871 (lr=2.0140e-04) (hash(x)=4294425)
152
+ 144 train 7.747959 (lr=2.0280e-04) (hash(x)=6095229)
153
+ 145 train 7.588701 (lr=2.0420e-04) (hash(x)=4363941)
154
+ 146 train 7.664509 (lr=2.0559e-04) (hash(x)=4693798)
155
+ 147 train 8.173172 (lr=2.0699e-04) (hash(x)=5486335)
156
+ 148 train 7.993349 (lr=2.0839e-04) (hash(x)=7907450)
157
+ 149 train 7.828030 (lr=2.0979e-04) (hash(x)=7295165)
158
+ 150 val loss 7.7270
159
+ 150 val perplexity 2268.7385
160
+ 150 train 7.757653 (lr=2.1119e-04) (hash(x)=6814026)
161
+ 151 train 7.846723 (lr=2.1259e-04) (hash(x)=6267424)
162
+ 152 train 7.828159 (lr=2.1399e-04) (hash(x)=8086437)
163
+ 153 train 7.879587 (lr=2.1538e-04) (hash(x)=5270452)
164
+ 154 train 7.730044 (lr=2.1678e-04) (hash(x)=6420820)
165
+ 155 train 7.628667 (lr=2.1818e-04) (hash(x)=6305297)
166
+ 156 train 7.788452 (lr=2.1958e-04) (hash(x)=7889849)
167
+ 157 train 8.134259 (lr=2.2098e-04) (hash(x)=6040077)
168
+ 158 train 8.125292 (lr=2.2238e-04) (hash(x)=5652497)
169
+ 159 train 7.714570 (lr=2.2378e-04) (hash(x)=7130251)
170
+ 160 train 7.591569 (lr=2.2517e-04) (hash(x)=6610177)
171
+ 161 train 7.643689 (lr=2.2657e-04) (hash(x)=6508350)
172
+ 162 train 7.794116 (lr=2.2797e-04) (hash(x)=6994983)
173
+ 163 train 7.741472 (lr=2.2937e-04) (hash(x)=5859576)
174
+ 164 train 7.313273 (lr=2.3077e-04) (hash(x)=5487065)
175
+ 165 train 7.542570 (lr=2.3217e-04) (hash(x)=4892065)
176
+ 166 train 7.436973 (lr=2.3357e-04) (hash(x)=4933674)
177
+ 167 train 7.707512 (lr=2.3497e-04) (hash(x)=7317289)
178
+ 168 train 7.806946 (lr=2.3636e-04) (hash(x)=6811522)
179
+ 169 train 7.634039 (lr=2.3776e-04) (hash(x)=5510218)
180
+ 170 train 7.789076 (lr=2.3916e-04) (hash(x)=7304235)
181
+ 171 train 7.947345 (lr=2.4056e-04) (hash(x)=6752265)
182
+ 172 train 7.644037 (lr=2.4196e-04) (hash(x)=5994476)
183
+ 173 train 7.613106 (lr=2.4336e-04) (hash(x)=6530867)
184
+ 174 train 7.719955 (lr=2.4476e-04) (hash(x)=5746260)
185
+ 175 train 7.669672 (lr=2.4615e-04) (hash(x)=5790078)
186
+ 176 train 7.685869 (lr=2.4755e-04) (hash(x)=6430135)
187
+ 177 train 7.755925 (lr=2.4895e-04) (hash(x)=7614023)
188
+ 178 train 7.745198 (lr=2.5035e-04) (hash(x)=5244861)
189
+ 179 train 7.718951 (lr=2.5175e-04) (hash(x)=5245146)
190
+ 180 train 7.623196 (lr=2.5315e-04) (hash(x)=6415565)
191
+ 181 train 7.671936 (lr=2.5455e-04) (hash(x)=5786972)
192
+ 182 train 7.711048 (lr=2.5594e-04) (hash(x)=6131703)
193
+ 183 train 7.731692 (lr=2.5734e-04) (hash(x)=6346459)
194
+ 184 train 7.763904 (lr=2.5874e-04) (hash(x)=6504131)
195
+ 185 train 7.757150 (lr=2.6014e-04) (hash(x)=5209234)
196
+ 186 train 7.662239 (lr=2.6154e-04) (hash(x)=6163272)
197
+ 187 train 7.664566 (lr=2.6294e-04) (hash(x)=5676768)
198
+ 188 train 7.727629 (lr=2.6434e-04) (hash(x)=6090561)
199
+ 189 train 7.876177 (lr=2.6573e-04) (hash(x)=5863092)
200
+ 190 train 7.793953 (lr=2.6713e-04) (hash(x)=7201312)
201
+ 191 train 7.870019 (lr=2.6853e-04) (hash(x)=7019489)
202
+ 192 train 7.640486 (lr=2.6993e-04) (hash(x)=6225608)
203
+ 193 train 7.726340 (lr=2.7133e-04) (hash(x)=5483019)
204
+ 194 train 7.662089 (lr=2.7273e-04) (hash(x)=5903258)
205
+ 195 train 7.703291 (lr=2.7413e-04) (hash(x)=5968716)
206
+ 196 train 7.695203 (lr=2.7552e-04) (hash(x)=5778017)
207
+ 197 train 7.426719 (lr=2.7692e-04) (hash(x)=7780194)
208
+ 198 train 7.632934 (lr=2.7832e-04) (hash(x)=6140998)
209
+ 199 train 7.620529 (lr=2.7972e-04) (hash(x)=3784321)
210
+ 200 val loss 7.6541
211
+ 200 val perplexity 2109.2607
212
+ 200 train 7.648793 (lr=2.8112e-04) (hash(x)=5860821)
213
+ 201 train 7.695865 (lr=2.8252e-04) (hash(x)=7774109)
214
+ 202 train 7.862022 (lr=2.8392e-04) (hash(x)=7997664)
215
+ 203 train 7.839151 (lr=2.8531e-04) (hash(x)=6576749)
216
+ 204 train 7.745502 (lr=2.8671e-04) (hash(x)=7266655)
217
+ 205 train 7.705393 (lr=2.8811e-04) (hash(x)=7072404)
218
+ 206 train 7.620603 (lr=2.8951e-04) (hash(x)=6626459)
219
+ 207 train 7.640181 (lr=2.9091e-04) (hash(x)=6556025)
220
+ 208 train 7.670080 (lr=2.9231e-04) (hash(x)=5942129)
221
+ 209 train 7.773512 (lr=2.9371e-04) (hash(x)=7325027)
222
+ 210 train 7.769457 (lr=2.9510e-04) (hash(x)=7322467)
223
+ 211 train 7.707861 (lr=2.9650e-04) (hash(x)=6373412)
224
+ 212 train 7.487411 (lr=2.9790e-04) (hash(x)=7332255)
225
+ 213 train 7.769811 (lr=2.9930e-04) (hash(x)=5262868)
226
+ 214 train 7.481462 (lr=3.0070e-04) (hash(x)=5688247)
227
+ 215 train 7.656564 (lr=3.0210e-04) (hash(x)=8055563)
228
+ 216 train 7.666641 (lr=3.0350e-04) (hash(x)=5352405)
229
+ 217 train 7.719462 (lr=3.0490e-04) (hash(x)=5932790)
230
+ 218 train 7.671588 (lr=3.0629e-04) (hash(x)=6249312)
231
+ 219 train 7.631391 (lr=3.0769e-04) (hash(x)=5680154)
232
+ 220 train 7.730334 (lr=3.0909e-04) (hash(x)=8156280)
233
+ 221 train 7.675781 (lr=3.1049e-04) (hash(x)=5914217)
234
+ 222 train 7.517913 (lr=3.1189e-04) (hash(x)=7120215)
235
+ 223 train 7.699361 (lr=3.1329e-04) (hash(x)=6424128)
236
+ 224 train 7.449519 (lr=3.1469e-04) (hash(x)=4390027)
237
+ 225 train 7.652171 (lr=3.1608e-04) (hash(x)=7804089)
238
+ 226 train 7.747726 (lr=3.1748e-04) (hash(x)=7130267)
239
+ 227 train 7.547658 (lr=3.1888e-04) (hash(x)=6490149)
240
+ 228 train 7.601025 (lr=3.2028e-04) (hash(x)=9276378)
241
+ 229 train 7.637883 (lr=3.2168e-04) (hash(x)=6283540)
242
+ 230 train 7.616992 (lr=3.2308e-04) (hash(x)=5781680)
243
+ 231 train 7.560999 (lr=3.2448e-04) (hash(x)=5728337)
244
+ 232 train 7.546410 (lr=3.2587e-04) (hash(x)=6278202)
245
+ 233 train 7.648108 (lr=3.2727e-04) (hash(x)=7400272)
246
+ 234 train 7.611826 (lr=3.2867e-04) (hash(x)=5795128)
247
+ 235 train 7.589363 (lr=3.3007e-04) (hash(x)=6473432)
248
+ 236 train 7.605611 (lr=3.3147e-04) (hash(x)=5201502)
249
+ 237 train 7.831241 (lr=3.3287e-04) (hash(x)=6922646)
250
+ 238 train 7.539192 (lr=3.3427e-04) (hash(x)=5484641)
251
+ 239 train 7.610104 (lr=3.3566e-04) (hash(x)=7277141)
252
+ 240 train 7.605514 (lr=3.3706e-04) (hash(x)=5604614)
253
+ 241 train 7.660578 (lr=3.3846e-04) (hash(x)=7443124)
254
+ 242 train 7.677866 (lr=3.3986e-04) (hash(x)=5601968)
255
+ 243 train 7.572083 (lr=3.4126e-04) (hash(x)=5632735)
256
+ 244 train 7.565017 (lr=3.4266e-04) (hash(x)=5594810)
257
+ 245 train 7.418117 (lr=3.4406e-04) (hash(x)=5881015)
258
+ 246 train 7.574681 (lr=3.4545e-04) (hash(x)=6865361)
259
+ 247 train 7.673841 (lr=3.4685e-04) (hash(x)=5809411)
260
+ 248 train 7.631116 (lr=3.4825e-04) (hash(x)=6293025)
261
+ 249 train 7.576523 (lr=3.4965e-04) (hash(x)=7232964)
262
+ 250 val loss 7.5903
263
+ 250 val perplexity 1978.9337
264
+ 250 train 7.521509 (lr=3.5105e-04) (hash(x)=5646619)
265
+ 251 train 7.637913 (lr=3.5245e-04) (hash(x)=6171829)
266
+ 252 train 7.682011 (lr=3.5385e-04) (hash(x)=7787552)
267
+ 253 train 7.628629 (lr=3.5524e-04) (hash(x)=5884456)
268
+ 254 train 7.660127 (lr=3.5664e-04) (hash(x)=6752162)
269
+ 255 train 7.570142 (lr=3.5804e-04) (hash(x)=5077031)
270
+ 256 train 7.745372 (lr=3.5944e-04) (hash(x)=6883481)
271
+ 257 train 7.630184 (lr=3.6084e-04) (hash(x)=6248895)
272
+ 258 train 7.495297 (lr=3.6224e-04) (hash(x)=6636615)
273
+ 259 train 7.691655 (lr=3.6364e-04) (hash(x)=6918004)
274
+ 260 train 7.614872 (lr=3.6503e-04) (hash(x)=7302874)
275
+ 261 train 7.718871 (lr=3.6643e-04) (hash(x)=6482466)
276
+ 262 train 7.554568 (lr=3.6783e-04) (hash(x)=7298317)
277
+ 263 train 7.597640 (lr=3.6923e-04) (hash(x)=6588573)
278
+ 264 train 7.590458 (lr=3.7063e-04) (hash(x)=6229630)
279
+ 265 train 7.670510 (lr=3.7203e-04) (hash(x)=6136939)
280
+ 266 train 7.430586 (lr=3.7343e-04) (hash(x)=6394645)
281
+ 267 train 7.649239 (lr=3.7483e-04) (hash(x)=7040157)
282
+ 268 train 7.694551 (lr=3.7622e-04) (hash(x)=7041858)
283
+ 269 train 7.601514 (lr=3.7762e-04) (hash(x)=6455595)
284
+ 270 train 7.527848 (lr=3.7902e-04) (hash(x)=6480443)
285
+ 271 train 7.705554 (lr=3.8042e-04) (hash(x)=4122983)
286
+ 272 train 7.526193 (lr=3.8182e-04) (hash(x)=7672190)
287
+ 273 train 7.558549 (lr=3.8322e-04) (hash(x)=4724052)
288
+ 274 train 7.516334 (lr=3.8462e-04) (hash(x)=6654267)
289
+ 275 train 7.504004 (lr=3.8601e-04) (hash(x)=6820797)
290
+ 276 train 7.572878 (lr=3.8741e-04) (hash(x)=6458662)
291
+ 277 train 7.959230 (lr=3.8881e-04) (hash(x)=7477594)
292
+ 278 train 7.626729 (lr=3.9021e-04) (hash(x)=5876574)
293
+ 279 train 7.350047 (lr=3.9161e-04) (hash(x)=6146305)
294
+ 280 train 7.519811 (lr=3.9301e-04) (hash(x)=6162080)
295
+ 281 train 7.486890 (lr=3.9441e-04) (hash(x)=6814831)
296
+ 282 train 7.618848 (lr=3.9580e-04) (hash(x)=7138295)
297
+ 283 train 7.408764 (lr=3.9720e-04) (hash(x)=5548298)
298
+ 284 train 7.497830 (lr=3.9860e-04) (hash(x)=6265078)
299
+ 285 train 7.869653 (lr=4.0000e-04) (hash(x)=7609416)
300
+ 286 train 7.565840 (lr=4.0000e-04) (hash(x)=6909367)
301
+ 287 train 7.667139 (lr=4.0000e-04) (hash(x)=7574342)
302
+ 288 train 7.482962 (lr=3.9999e-04) (hash(x)=8859748)
303
+ 289 train 7.437656 (lr=3.9998e-04) (hash(x)=5906937)
304
+ 290 train 7.448060 (lr=3.9997e-04) (hash(x)=7243024)
305
+ 291 train 7.568493 (lr=3.9996e-04) (hash(x)=8013471)
306
+ 292 train 7.480510 (lr=3.9994e-04) (hash(x)=5659461)
307
+ 293 train 7.522944 (lr=3.9991e-04) (hash(x)=6146213)
308
+ 294 train 7.575163 (lr=3.9989e-04) (hash(x)=6602917)
309
+ 295 train 7.490837 (lr=3.9986e-04) (hash(x)=6709902)
310
+ 296 train 7.593400 (lr=3.9983e-04) (hash(x)=8154341)
311
+ 297 train 7.746493 (lr=3.9979e-04) (hash(x)=6402571)
312
+ 298 train 7.485673 (lr=3.9975e-04) (hash(x)=5605466)
313
+ 299 train 7.451880 (lr=3.9971e-04) (hash(x)=7683614)
314
+ 300 val loss 7.5142
315
+ 300 val perplexity 1833.8905
316
+ 300 train 7.542973 (lr=3.9966e-04) (hash(x)=5421572)
317
+ 301 train 7.396899 (lr=3.9961e-04) (hash(x)=5453026)
318
+ 302 train 7.457862 (lr=3.9955e-04) (hash(x)=6437746)
319
+ 303 train 7.374797 (lr=3.9950e-04) (hash(x)=6260017)
320
+ 304 train 7.478588 (lr=3.9944e-04) (hash(x)=6119053)
321
+ 305 train 7.627518 (lr=3.9937e-04) (hash(x)=6591466)
322
+ 306 train 7.541867 (lr=3.9930e-04) (hash(x)=5652701)
323
+ 307 train 7.522711 (lr=3.9923e-04) (hash(x)=5501472)
324
+ 308 train 7.696172 (lr=3.9916e-04) (hash(x)=6160695)
325
+ 309 train 7.446716 (lr=3.9908e-04) (hash(x)=5871698)
326
+ 310 train 7.423694 (lr=3.9900e-04) (hash(x)=7301184)
327
+ 311 train 7.355983 (lr=3.9891e-04) (hash(x)=4282052)
328
+ 312 train 7.567953 (lr=3.9882e-04) (hash(x)=5679330)
329
+ 313 train 7.401403 (lr=3.9873e-04) (hash(x)=7824660)
330
+ 314 train 7.453237 (lr=3.9864e-04) (hash(x)=4857318)
331
+ 315 train 7.389629 (lr=3.9854e-04) (hash(x)=5954950)
332
+ 316 train 7.464486 (lr=3.9843e-04) (hash(x)=5758178)
333
+ 317 train 7.470260 (lr=3.9833e-04) (hash(x)=5484656)
334
+ 318 train 7.483916 (lr=3.9822e-04) (hash(x)=6818913)
335
+ 319 train 7.682001 (lr=3.9811e-04) (hash(x)=7031328)
336
+ 320 train 7.523411 (lr=3.9799e-04) (hash(x)=6837621)
337
+ 321 train 7.539083 (lr=3.9787e-04) (hash(x)=9157069)
338
+ 322 train 7.783797 (lr=3.9775e-04) (hash(x)=6509536)
339
+ 323 train 7.488003 (lr=3.9762e-04) (hash(x)=5996883)
340
+ 324 train 7.736810 (lr=3.9749e-04) (hash(x)=6850868)
341
+ 325 train 7.658040 (lr=3.9736e-04) (hash(x)=7645447)
342
+ 326 train 7.682364 (lr=3.9722e-04) (hash(x)=6717515)
343
+ 327 train 7.785436 (lr=3.9708e-04) (hash(x)=6336164)
344
+ 328 train 7.491637 (lr=3.9694e-04) (hash(x)=6897445)
345
+ 329 train 7.396650 (lr=3.9679e-04) (hash(x)=5224240)
346
+ 330 train 7.555936 (lr=3.9664e-04) (hash(x)=6612945)
347
+ 331 train 8.023050 (lr=3.9648e-04) (hash(x)=5447801)
348
+ 332 train 7.544331 (lr=3.9633e-04) (hash(x)=7308579)
349
+ 333 train 7.421063 (lr=3.9616e-04) (hash(x)=4155598)
350
+ 334 train 7.434925 (lr=3.9600e-04) (hash(x)=6106773)
351
+ 335 train 7.486042 (lr=3.9583e-04) (hash(x)=7482744)
352
+ 336 train 7.555331 (lr=3.9566e-04) (hash(x)=6310687)
353
+ 337 train 7.439886 (lr=3.9549e-04) (hash(x)=4733414)
354
+ 338 train 7.542549 (lr=3.9531e-04) (hash(x)=6979734)
355
+ 339 train 7.344533 (lr=3.9513e-04) (hash(x)=7806284)
356
+ 340 train 7.354739 (lr=3.9494e-04) (hash(x)=5080536)
357
+ 341 train 7.557064 (lr=3.9475e-04) (hash(x)=7236478)
358
+ 342 train 7.489819 (lr=3.9456e-04) (hash(x)=6619318)
359
+ 343 train 7.415064 (lr=3.9437e-04) (hash(x)=4656255)
360
+ 344 train 7.342458 (lr=3.9417e-04) (hash(x)=5546888)
361
+ 345 train 7.797825 (lr=3.9397e-04) (hash(x)=9298325)
362
+ 346 train 7.483396 (lr=3.9376e-04) (hash(x)=5994232)
363
+ 347 train 7.464743 (lr=3.9356e-04) (hash(x)=6287524)
364
+ 348 train 7.583871 (lr=3.9334e-04) (hash(x)=6200049)
365
+ 349 train 7.502278 (lr=3.9313e-04) (hash(x)=7895970)
366
+ 350 val loss 7.4495
367
+ 350 val perplexity 1718.9470
368
+ 350 train 7.427850 (lr=3.9291e-04) (hash(x)=6051847)
369
+ 351 train 7.552623 (lr=3.9269e-04) (hash(x)=8847992)
370
+ 352 train 7.923444 (lr=3.9246e-04) (hash(x)=5123883)
371
+ 353 train 7.526021 (lr=3.9223e-04) (hash(x)=6338998)
372
+ 354 train 7.524910 (lr=3.9200e-04) (hash(x)=5765866)
373
+ 355 train 7.864393 (lr=3.9177e-04) (hash(x)=7104011)
374
+ 356 train 7.587577 (lr=3.9153e-04) (hash(x)=7438267)
375
+ 357 train 7.320564 (lr=3.9129e-04) (hash(x)=5573214)
376
+ 358 train 7.632944 (lr=3.9104e-04) (hash(x)=6897283)
377
+ 359 train 7.391943 (lr=3.9079e-04) (hash(x)=6882897)
378
+ 360 train 7.533899 (lr=3.9054e-04) (hash(x)=5343615)
379
+ 361 train 7.436355 (lr=3.9029e-04) (hash(x)=5854242)
380
+ 362 train 7.344234 (lr=3.9003e-04) (hash(x)=6317324)
381
+ 363 train 7.275468 (lr=3.8977e-04) (hash(x)=6972494)
382
+ 364 train 7.358955 (lr=3.8950e-04) (hash(x)=5661167)
383
+ 365 train 7.379583 (lr=3.8923e-04) (hash(x)=6796805)
384
+ 366 train 7.215770 (lr=3.8896e-04) (hash(x)=5817209)
385
+ 367 train 7.265789 (lr=3.8869e-04) (hash(x)=6139120)
386
+ 368 train 7.442384 (lr=3.8841e-04) (hash(x)=6121296)
387
+ 369 train 7.423337 (lr=3.8813e-04) (hash(x)=6486796)
388
+ 370 train 7.342756 (lr=3.8785e-04) (hash(x)=5514645)
389
+ 371 train 7.254919 (lr=3.8756e-04) (hash(x)=5578797)
390
+ 372 train 7.281869 (lr=3.8727e-04) (hash(x)=7651738)
391
+ 373 train 7.471611 (lr=3.8697e-04) (hash(x)=6895326)
392
+ 374 train 7.485798 (lr=3.8667e-04) (hash(x)=5490256)
393
+ 375 train 7.370027 (lr=3.8637e-04) (hash(x)=7783160)
394
+ 376 train 7.544970 (lr=3.8607e-04) (hash(x)=5608424)
395
+ 377 train 7.333013 (lr=3.8576e-04) (hash(x)=4867625)
396
+ 378 train 7.243606 (lr=3.8545e-04) (hash(x)=4655472)
397
+ 379 train 7.309593 (lr=3.8514e-04) (hash(x)=6631123)
398
+ 380 train 6.973507 (lr=3.8482e-04) (hash(x)=5751990)
399
+ 381 train 7.271508 (lr=3.8450e-04) (hash(x)=6710965)
400
+ 382 train 7.243104 (lr=3.8418e-04) (hash(x)=5293800)
401
+ 383 train 7.276038 (lr=3.8385e-04) (hash(x)=6294657)
402
+ 384 train 7.443870 (lr=3.8352e-04) (hash(x)=6999538)
403
+ 385 train 7.417187 (lr=3.8319e-04) (hash(x)=5443606)
404
+ 386 train 7.384171 (lr=3.8286e-04) (hash(x)=5650401)
405
+ 387 train 7.408008 (lr=3.8252e-04) (hash(x)=7035770)
406
+ 388 train 7.338608 (lr=3.8217e-04) (hash(x)=6678271)
407
+ 389 train 7.547683 (lr=3.8183e-04) (hash(x)=4899541)
408
+ 390 train 7.292682 (lr=3.8148e-04) (hash(x)=5359260)
409
+ 391 train 7.294185 (lr=3.8113e-04) (hash(x)=6227120)
410
+ 392 train 7.355081 (lr=3.8077e-04) (hash(x)=6643916)
411
+ 393 train 7.424819 (lr=3.8042e-04) (hash(x)=5981648)
412
+ 394 train 7.377169 (lr=3.8006e-04) (hash(x)=6467540)
413
+ 395 train 7.313494 (lr=3.7969e-04) (hash(x)=5558993)
414
+ 396 train 7.467900 (lr=3.7933e-04) (hash(x)=6698959)
415
+ 397 train 7.292468 (lr=3.7896e-04) (hash(x)=7015671)
416
+ 398 train 7.287083 (lr=3.7858e-04) (hash(x)=6800890)
417
+ 399 train 7.448864 (lr=3.7821e-04) (hash(x)=5931234)
418
+ 400 val loss 7.3784
419
+ 400 val perplexity 1600.9830
420
+ 400 train 7.432587 (lr=3.7783e-04) (hash(x)=3995070)
421
+ 401 train 7.363217 (lr=3.7744e-04) (hash(x)=4582159)
422
+ 402 train 7.439198 (lr=3.7706e-04) (hash(x)=7081906)
423
+ 403 train 7.332005 (lr=3.7667e-04) (hash(x)=5029472)
424
+ 404 train 7.327302 (lr=3.7628e-04) (hash(x)=3946232)
425
+ 405 train 7.241664 (lr=3.7588e-04) (hash(x)=5434102)
426
+ 406 train 8.010228 (lr=3.7549e-04) (hash(x)=5340646)
427
+ 407 train 7.132459 (lr=3.7509e-04) (hash(x)=6810638)
428
+ 408 train 7.397131 (lr=3.7468e-04) (hash(x)=6768026)
429
+ 409 train 7.401068 (lr=3.7428e-04) (hash(x)=6521562)
430
+ 410 train 7.363834 (lr=3.7387e-04) (hash(x)=4429434)
431
+ 411 train 7.427893 (lr=3.7345e-04) (hash(x)=7226932)
432
+ 412 train 7.236605 (lr=3.7304e-04) (hash(x)=5065932)
433
+ 413 train 7.369333 (lr=3.7262e-04) (hash(x)=6957849)
434
+ 414 train 7.363637 (lr=3.7220e-04) (hash(x)=6431457)
435
+ 415 train 7.426032 (lr=3.7177e-04) (hash(x)=5334281)
436
+ 416 train 7.437128 (lr=3.7135e-04) (hash(x)=5471747)
437
+ 417 train 7.526937 (lr=3.7092e-04) (hash(x)=5887983)
438
+ 418 train 7.425808 (lr=3.7048e-04) (hash(x)=5075909)
439
+ 419 train 7.241815 (lr=3.7005e-04) (hash(x)=6393576)
440
+ 420 train 7.371863 (lr=3.6961e-04) (hash(x)=6608979)
441
+ 421 train 7.781045 (lr=3.6917e-04) (hash(x)=5057959)
442
+ 422 train 7.693833 (lr=3.6872e-04) (hash(x)=6414858)
443
+ 423 train 7.466870 (lr=3.6828e-04) (hash(x)=8595271)
444
+ 424 train 7.336835 (lr=3.6782e-04) (hash(x)=6500949)
445
+ 425 train 7.500569 (lr=3.6737e-04) (hash(x)=8395167)
446
+ 426 train 7.617700 (lr=3.6692e-04) (hash(x)=6288420)
447
+ 427 train 7.422984 (lr=3.6646e-04) (hash(x)=7634417)
448
+ 428 train 7.320395 (lr=3.6599e-04) (hash(x)=6795561)
449
+ 429 train 7.226711 (lr=3.6553e-04) (hash(x)=5257771)
450
+ 430 train 7.380797 (lr=3.6506e-04) (hash(x)=7506860)
451
+ 431 train 7.427154 (lr=3.6459e-04) (hash(x)=7674238)
452
+ 432 train 7.529054 (lr=3.6412e-04) (hash(x)=5353794)
453
+ 433 train 7.442874 (lr=3.6364e-04) (hash(x)=5300555)
454
+ 434 train 7.158638 (lr=3.6316e-04) (hash(x)=5564419)
455
+ 435 train 7.421156 (lr=3.6268e-04) (hash(x)=6066139)
456
+ 436 train 7.384982 (lr=3.6220e-04) (hash(x)=6863709)
457
+ 437 train 7.384432 (lr=3.6171e-04) (hash(x)=6688212)
458
+ 438 train 7.548279 (lr=3.6122e-04) (hash(x)=7095090)
459
+ 439 train 7.637533 (lr=3.6073e-04) (hash(x)=6642404)
460
+ 440 train 7.350913 (lr=3.6023e-04) (hash(x)=6539227)
461
+ 441 train 7.343325 (lr=3.5974e-04) (hash(x)=4961570)
462
+ 442 train 7.282474 (lr=3.5924e-04) (hash(x)=6606391)
463
+ 443 train 7.081844 (lr=3.5873e-04) (hash(x)=5105190)
464
+ 444 train 7.669587 (lr=3.5823e-04) (hash(x)=5695997)
465
+ 445 train 7.361680 (lr=3.5772e-04) (hash(x)=7155991)
466
+ 446 train 7.332823 (lr=3.5721e-04) (hash(x)=7249727)
467
+ 447 train 7.320885 (lr=3.5669e-04) (hash(x)=5132972)
468
+ 448 train 7.311207 (lr=3.5618e-04) (hash(x)=5381802)
469
+ 449 train 7.330151 (lr=3.5566e-04) (hash(x)=6269946)
470
+ 450 val loss 7.3349
471
+ 450 val perplexity 1532.9175
472
+ 450 train 7.304922 (lr=3.5514e-04) (hash(x)=4976762)
473
+ 451 train 7.394916 (lr=3.5461e-04) (hash(x)=5396133)
474
+ 452 train 7.351671 (lr=3.5408e-04) (hash(x)=6173473)
475
+ 453 train 7.329982 (lr=3.5355e-04) (hash(x)=6394083)
476
+ 454 train 7.386403 (lr=3.5302e-04) (hash(x)=6436030)
477
+ 455 train 7.248203 (lr=3.5249e-04) (hash(x)=5237804)
478
+ 456 train 7.217991 (lr=3.5195e-04) (hash(x)=6540397)
479
+ 457 train 7.480265 (lr=3.5141e-04) (hash(x)=5216638)
480
+ 458 train 7.296997 (lr=3.5087e-04) (hash(x)=5750537)
481
+ 459 train 7.295883 (lr=3.5032e-04) (hash(x)=6735565)
482
+ 460 train 7.305536 (lr=3.4977e-04) (hash(x)=7297488)
483
+ 461 train 7.497890 (lr=3.4922e-04) (hash(x)=5983575)
484
+ 462 train 7.319165 (lr=3.4867e-04) (hash(x)=4808543)
485
+ 463 train 7.431988 (lr=3.4812e-04) (hash(x)=5831619)
486
+ 464 train 7.193366 (lr=3.4756e-04) (hash(x)=6280052)
487
+ 465 train 7.224925 (lr=3.4700e-04) (hash(x)=9109083)
488
+ 466 train 7.464886 (lr=3.4644e-04) (hash(x)=5848604)
489
+ 467 train 7.403017 (lr=3.4587e-04) (hash(x)=6912967)
490
+ 468 train 7.598088 (lr=3.4530e-04) (hash(x)=6615365)
491
+ 469 train 7.643711 (lr=3.4473e-04) (hash(x)=5965674)
492
+ 470 train 7.469454 (lr=3.4416e-04) (hash(x)=6879965)
493
+ 471 train 7.434535 (lr=3.4359e-04) (hash(x)=5822533)
494
+ 472 train 7.317868 (lr=3.4301e-04) (hash(x)=5794462)
495
+ 473 train 7.386673 (lr=3.4243e-04) (hash(x)=5268216)
496
+ 474 train 7.210080 (lr=3.4185e-04) (hash(x)=7058396)
497
+ 475 train 7.314373 (lr=3.4127e-04) (hash(x)=6531900)
498
+ 476 train 7.340644 (lr=3.4068e-04) (hash(x)=5878912)
499
+ 477 train 7.157888 (lr=3.4009e-04) (hash(x)=5426499)
500
+ 478 train 7.357786 (lr=3.3950e-04) (hash(x)=8107004)
501
+ 479 train 8.084682 (lr=3.3891e-04) (hash(x)=7455372)
502
+ 480 train 8.043467 (lr=3.3831e-04) (hash(x)=7727959)
503
+ 481 train 8.002517 (lr=3.3771e-04) (hash(x)=8319119)
504
+ 482 train 7.923566 (lr=3.3711e-04) (hash(x)=7581656)
505
+ 483 train 7.973314 (lr=3.3651e-04) (hash(x)=7592128)
506
+ 484 train 7.972460 (lr=3.3590e-04) (hash(x)=10506633)
507
+ 485 train 7.926563 (lr=3.3530e-04) (hash(x)=8704170)
508
+ 486 train 7.699511 (lr=3.3469e-04) (hash(x)=7461138)
509
+ 487 train 7.539885 (lr=3.3408e-04) (hash(x)=7165037)
510
+ 488 train 7.610208 (lr=3.3346e-04) (hash(x)=4976998)
511
+ 489 train 7.311361 (lr=3.3285e-04) (hash(x)=7678000)
512
+ 490 train 7.371429 (lr=3.3223e-04) (hash(x)=6299409)
513
+ 491 train 7.297880 (lr=3.3161e-04) (hash(x)=7414788)
514
+ 492 train 7.416718 (lr=3.3099e-04) (hash(x)=5063061)
515
+ 493 train 7.351502 (lr=3.3036e-04) (hash(x)=7326216)
516
+ 494 train 7.483407 (lr=3.2973e-04) (hash(x)=7492012)
517
+ 495 train 7.431917 (lr=3.2911e-04) (hash(x)=5857162)
518
+ 496 train 7.224462 (lr=3.2847e-04) (hash(x)=6169001)
519
+ 497 train 9.223996 (lr=3.2784e-04) (hash(x)=14715669)
520
+ 498 train 12.014521 (lr=3.2721e-04) (hash(x)=18256706)
521
+ 499 train 11.784342 (lr=3.2657e-04) (hash(x)=16950560)
522
+ 500 val loss 7.3506
523
+ 500 val perplexity 1557.1532
524
+ 500 train 8.058232 (lr=3.2593e-04) (hash(x)=6113563)
525
+ 501 train 6.992707 (lr=3.2529e-04) (hash(x)=6159125)
526
+ 502 train 7.386683 (lr=3.2464e-04) (hash(x)=5482303)
527
+ 503 train 7.446765 (lr=3.2400e-04) (hash(x)=5908442)
528
+ 504 train 7.311045 (lr=3.2335e-04) (hash(x)=4985045)
529
+ 505 train 7.238109 (lr=3.2270e-04) (hash(x)=6668887)
530
+ 506 train 7.326504 (lr=3.2205e-04) (hash(x)=6123910)
531
+ 507 train 7.395772 (lr=3.2140e-04) (hash(x)=7662975)
532
+ 508 train 7.346252 (lr=3.2074e-04) (hash(x)=7191898)
533
+ 509 train 7.294250 (lr=3.2008e-04) (hash(x)=6669539)
534
+ 510 train 7.358017 (lr=3.1943e-04) (hash(x)=6557481)
535
+ 511 train 7.255944 (lr=3.1876e-04) (hash(x)=5015998)
536
+ 512 train 7.317005 (lr=3.1810e-04) (hash(x)=4964540)
537
+ 513 train 7.142015 (lr=3.1744e-04) (hash(x)=5458386)
538
+ 514 train 7.541059 (lr=3.1677e-04) (hash(x)=6077616)
539
+ 515 train 7.268070 (lr=3.1610e-04) (hash(x)=4849847)
540
+ 516 train 7.318647 (lr=3.1543e-04) (hash(x)=5431422)
541
+ 517 train 7.528705 (lr=3.1476e-04) (hash(x)=5125562)
542
+ 518 train 7.420020 (lr=3.1408e-04) (hash(x)=5587715)
543
+ 519 train 7.478749 (lr=3.1341e-04) (hash(x)=6012047)
544
+ 520 train 7.329851 (lr=3.1273e-04) (hash(x)=5668436)
545
+ 521 train 7.361660 (lr=3.1205e-04) (hash(x)=6079364)
546
+ 522 train 7.473340 (lr=3.1137e-04) (hash(x)=5200504)
547
+ 523 train 7.210963 (lr=3.1069e-04) (hash(x)=6030395)
548
+ 524 train 7.451573 (lr=3.1000e-04) (hash(x)=6855319)
549
+ 525 train 7.225374 (lr=3.0931e-04) (hash(x)=6781762)
550
+ 526 train 7.378330 (lr=3.0862e-04) (hash(x)=6717939)
551
+ 527 train 7.363630 (lr=3.0793e-04) (hash(x)=5238325)
552
+ 528 train 7.265601 (lr=3.0724e-04) (hash(x)=5808069)
553
+ 529 train 7.628414 (lr=3.0655e-04) (hash(x)=8611144)
554
+ 530 train 7.271699 (lr=3.0585e-04) (hash(x)=6028738)
555
+ 531 train 7.303637 (lr=3.0516e-04) (hash(x)=6901605)
556
+ 532 train 7.242479 (lr=3.0446e-04) (hash(x)=6723307)
557
+ 533 train 7.365186 (lr=3.0376e-04) (hash(x)=5499337)
558
+ 534 train 7.286484 (lr=3.0306e-04) (hash(x)=6586372)
559
+ 535 train 7.408593 (lr=3.0235e-04) (hash(x)=7134874)
560
+ 536 train 7.297486 (lr=3.0165e-04) (hash(x)=8415623)
561
+ 537 train 7.330126 (lr=3.0094e-04) (hash(x)=6149038)
562
+ 538 train 7.363118 (lr=3.0023e-04) (hash(x)=6078325)
563
+ 539 train 7.337748 (lr=2.9952e-04) (hash(x)=6609760)
564
+ 540 train 7.231339 (lr=2.9881e-04) (hash(x)=6185253)
565
+ 541 train 7.614024 (lr=2.9810e-04) (hash(x)=6652627)
566
+ 542 train 7.417749 (lr=2.9738e-04) (hash(x)=7758536)
567
+ 543 train 7.380561 (lr=2.9667e-04) (hash(x)=8618579)
568
+ 544 train 7.360254 (lr=2.9595e-04) (hash(x)=6710226)
569
+ 545 train 7.206186 (lr=2.9523e-04) (hash(x)=5896892)
570
+ 546 train 7.224288 (lr=2.9451e-04) (hash(x)=6332995)
571
+ 547 train 7.649704 (lr=2.9379e-04) (hash(x)=8945854)
572
+ 548 train 7.222267 (lr=2.9307e-04) (hash(x)=4329129)
573
+ 549 train 7.260049 (lr=2.9234e-04) (hash(x)=6775103)
574
+ 550 val loss 7.2928
575
+ 550 val perplexity 1469.7389
576
+ 550 train 7.375446 (lr=2.9162e-04) (hash(x)=4465735)
577
+ 551 train 7.241213 (lr=2.9089e-04) (hash(x)=6200024)
578
+ 552 train 7.211312 (lr=2.9016e-04) (hash(x)=5520748)
579
+ 553 train 7.214302 (lr=2.8943e-04) (hash(x)=5238139)
580
+ 554 train 7.477874 (lr=2.8870e-04) (hash(x)=7639817)
581
+ 555 train 7.244286 (lr=2.8797e-04) (hash(x)=5780948)
582
+ 556 train 7.365036 (lr=2.8723e-04) (hash(x)=6733444)
583
+ 557 train 7.318333 (lr=2.8650e-04) (hash(x)=6168407)
584
+ 558 train 7.351531 (lr=2.8576e-04) (hash(x)=6680197)
585
+ 559 train 7.230827 (lr=2.8502e-04) (hash(x)=3775881)
586
+ 560 train 7.085499 (lr=2.8428e-04) (hash(x)=5784096)
587
+ 561 train 7.513249 (lr=2.8354e-04) (hash(x)=6848709)
588
+ 562 train 7.548810 (lr=2.8280e-04) (hash(x)=7449019)
589
+ 563 train 7.647755 (lr=2.8206e-04) (hash(x)=6478113)
590
+ 564 train 7.310112 (lr=2.8132e-04) (hash(x)=5925744)
591
+ 565 train 7.360231 (lr=2.8057e-04) (hash(x)=6483914)
592
+ 566 train 7.233283 (lr=2.7982e-04) (hash(x)=6010847)
593
+ 567 train 7.331893 (lr=2.7908e-04) (hash(x)=5927795)
594
+ 568 train 7.343916 (lr=2.7833e-04) (hash(x)=5896332)
595
+ 569 train 7.303408 (lr=2.7758e-04) (hash(x)=8438033)
596
+ 570 train 7.404140 (lr=2.7683e-04) (hash(x)=6081660)
597
+ 571 train 7.299306 (lr=2.7607e-04) (hash(x)=5684788)
598
+ 572 train 7.192412 (lr=2.7532e-04) (hash(x)=6272055)
599
+ 573 train 7.161539 (lr=2.7457e-04) (hash(x)=5710182)
600
+ 574 train 7.272127 (lr=2.7381e-04) (hash(x)=6219616)
601
+ 575 train 7.246496 (lr=2.7306e-04) (hash(x)=6160063)
602
+ 576 train 7.265047 (lr=2.7230e-04) (hash(x)=5710809)
603
+ 577 train 7.024844 (lr=2.7154e-04) (hash(x)=5955100)
604
+ 578 train 7.040871 (lr=2.7078e-04) (hash(x)=5356088)
605
+ 579 train 7.083298 (lr=2.7002e-04) (hash(x)=6672309)
606
+ 580 train 7.283853 (lr=2.6926e-04) (hash(x)=5841684)
607
+ 581 train 7.632938 (lr=2.6850e-04) (hash(x)=9001564)
608
+ 582 train 8.183352 (lr=2.6773e-04) (hash(x)=6897178)
609
+ 583 train 7.391829 (lr=2.6697e-04) (hash(x)=5454654)
610
+ 584 train 7.146838 (lr=2.6620e-04) (hash(x)=6391454)
611
+ 585 train 7.241993 (lr=2.6544e-04) (hash(x)=5174062)
612
+ 586 train 7.355693 (lr=2.6467e-04) (hash(x)=5197924)
613
+ 587 train 7.313063 (lr=2.6390e-04) (hash(x)=6199203)
614
+ 588 train 7.137313 (lr=2.6314e-04) (hash(x)=5786498)
615
+ 589 train 7.287375 (lr=2.6237e-04) (hash(x)=8352747)
616
+ 590 train 7.572512 (lr=2.6160e-04) (hash(x)=4829700)
617
+ 591 train 7.071661 (lr=2.6083e-04) (hash(x)=4547106)
618
+ 592 train 7.374060 (lr=2.6005e-04) (hash(x)=7686831)
619
+ 593 train 7.254583 (lr=2.5928e-04) (hash(x)=6433099)
620
+ 594 train 7.109669 (lr=2.5851e-04) (hash(x)=4826460)
621
+ 595 train 7.211390 (lr=2.5773e-04) (hash(x)=6424428)
622
+ 596 train 7.095765 (lr=2.5696e-04) (hash(x)=6073605)
623
+ 597 train 7.441885 (lr=2.5618e-04) (hash(x)=5975712)
624
+ 598 train 7.042920 (lr=2.5541e-04) (hash(x)=7030029)
625
+ 599 train 7.353360 (lr=2.5463e-04) (hash(x)=5458532)
626
+ 600 val loss 7.2687
627
+ 600 val perplexity 1434.7274
628
+ 600 train 7.190872 (lr=2.5385e-04) (hash(x)=7823919)
629
+ 601 train 7.637245 (lr=2.5307e-04) (hash(x)=7002774)
630
+ 602 train 7.428175 (lr=2.5230e-04) (hash(x)=7056399)
631
+ 603 train 7.637279 (lr=2.5152e-04) (hash(x)=8206173)
632
+ 604 train 7.469888 (lr=2.5074e-04) (hash(x)=8034846)
633
+ 605 train 7.767740 (lr=2.4996e-04) (hash(x)=7784971)
634
+ 606 train 7.912747 (lr=2.4917e-04) (hash(x)=7118076)
635
+ 607 train 7.200405 (lr=2.4839e-04) (hash(x)=6001843)
636
+ 608 train 7.297425 (lr=2.4761e-04) (hash(x)=5582437)
637
+ 609 train 7.408721 (lr=2.4683e-04) (hash(x)=5698642)
638
+ 610 train 7.256955 (lr=2.4604e-04) (hash(x)=6615301)
639
+ 611 train 7.261686 (lr=2.4526e-04) (hash(x)=6600751)
640
+ 612 train 7.176589 (lr=2.4448e-04) (hash(x)=6880955)
641
+ 613 train 7.233938 (lr=2.4369e-04) (hash(x)=5804610)
642
+ 614 train 7.166404 (lr=2.4291e-04) (hash(x)=6263347)
643
+ 615 train 7.125833 (lr=2.4212e-04) (hash(x)=5162652)
644
+ 616 train 7.184370 (lr=2.4133e-04) (hash(x)=6055710)
645
+ 617 train 7.140441 (lr=2.4055e-04) (hash(x)=5297006)
646
+ 618 train 7.168477 (lr=2.3976e-04) (hash(x)=5923648)
647
+ 619 train 7.176470 (lr=2.3897e-04) (hash(x)=6728323)
648
+ 620 train 7.185227 (lr=2.3818e-04) (hash(x)=6660239)
649
+ 621 train 7.154287 (lr=2.3740e-04) (hash(x)=3964882)
650
+ 622 train 7.203990 (lr=2.3661e-04) (hash(x)=5868942)
651
+ 623 train 7.118853 (lr=2.3582e-04) (hash(x)=6735993)
652
+ 624 train 7.468281 (lr=2.3503e-04) (hash(x)=5926190)
653
+ 625 train 7.150704 (lr=2.3424e-04) (hash(x)=5150316)
654
+ 626 train 7.166037 (lr=2.3345e-04) (hash(x)=7571467)
655
+ 627 train 7.072007 (lr=2.3266e-04) (hash(x)=5220448)
656
+ 628 train 7.086149 (lr=2.3187e-04) (hash(x)=6790397)
657
+ 629 train 7.560840 (lr=2.3108e-04) (hash(x)=6971289)
658
+ 630 train 7.514282 (lr=2.3029e-04) (hash(x)=7123359)
659
+ 631 train 7.295150 (lr=2.2950e-04) (hash(x)=5012069)
660
+ 632 train 7.258813 (lr=2.2871e-04) (hash(x)=6183574)
661
+ 633 train 7.152743 (lr=2.2792e-04) (hash(x)=6576136)
662
+ 634 train 7.178514 (lr=2.2713e-04) (hash(x)=6157346)
663
+ 635 train 7.171775 (lr=2.2633e-04) (hash(x)=6077647)
664
+ 636 train 7.413737 (lr=2.2554e-04) (hash(x)=8369488)
665
+ 637 train 7.205590 (lr=2.2475e-04) (hash(x)=6032381)
666
+ 638 train 7.130117 (lr=2.2396e-04) (hash(x)=5924756)
667
+ 639 train 7.171339 (lr=2.2317e-04) (hash(x)=8410716)
668
+ 640 train 7.155844 (lr=2.2238e-04) (hash(x)=5803440)
669
+ 641 train 7.248858 (lr=2.2158e-04) (hash(x)=6212516)
670
+ 642 train 7.325352 (lr=2.2079e-04) (hash(x)=7165945)
671
+ 643 train 7.037984 (lr=2.2000e-04) (hash(x)=5298545)
672
+ 644 train 7.292557 (lr=2.1921e-04) (hash(x)=6281739)
673
+ 645 train 7.257489 (lr=2.1842e-04) (hash(x)=5779073)
674
+ 646 train 7.117522 (lr=2.1762e-04) (hash(x)=5090107)
675
+ 647 train 7.184642 (lr=2.1683e-04) (hash(x)=6340947)
676
+ 648 train 6.941740 (lr=2.1604e-04) (hash(x)=4741418)
677
+ 649 train 6.861466 (lr=2.1525e-04) (hash(x)=6530441)
678
+ 650 val loss 7.2342
679
+ 650 val perplexity 1386.0918
680
+ 650 train 7.399755 (lr=2.1446e-04) (hash(x)=6396041)
681
+ 651 train 7.084840 (lr=2.1367e-04) (hash(x)=7818007)
682
+ 652 train 7.090919 (lr=2.1287e-04) (hash(x)=6478457)
683
+ 653 train 7.159522 (lr=2.1208e-04) (hash(x)=6277435)
684
+ 654 train 7.123230 (lr=2.1129e-04) (hash(x)=6042369)
685
+ 655 train 7.139806 (lr=2.1050e-04) (hash(x)=6226320)
686
+ 656 train 7.160164 (lr=2.0971e-04) (hash(x)=5826883)
687
+ 657 train 7.271748 (lr=2.0892e-04) (hash(x)=5463751)
688
+ 658 train 7.191085 (lr=2.0813e-04) (hash(x)=6433080)
689
+ 659 train 7.041486 (lr=2.0734e-04) (hash(x)=6579329)
690
+ 660 train 7.062831 (lr=2.0655e-04) (hash(x)=4724703)
691
+ 661 train 7.415412 (lr=2.0576e-04) (hash(x)=6660540)
692
+ 662 train 7.218508 (lr=2.0497e-04) (hash(x)=5918756)
693
+ 663 train 7.151925 (lr=2.0418e-04) (hash(x)=6491272)
694
+ 664 train 7.080535 (lr=2.0339e-04) (hash(x)=5452863)
695
+ 665 train 7.272725 (lr=2.0260e-04) (hash(x)=6766947)
696
+ 666 train 7.634573 (lr=2.0182e-04) (hash(x)=7459504)
697
+ 667 train 7.294215 (lr=2.0103e-04) (hash(x)=3721218)
698
+ 668 train 7.465971 (lr=2.0024e-04) (hash(x)=7929571)
699
+ 669 train 7.136056 (lr=1.9945e-04) (hash(x)=5229898)
700
+ 670 train 7.182530 (lr=1.9867e-04) (hash(x)=6083090)
701
+ 671 train 7.041215 (lr=1.9788e-04) (hash(x)=6054428)
702
+ 672 train 7.194268 (lr=1.9709e-04) (hash(x)=6072563)
703
+ 673 train 7.283079 (lr=1.9631e-04) (hash(x)=6290464)
704
+ 674 train 7.396589 (lr=1.9552e-04) (hash(x)=6314742)
705
+ 675 train 7.069119 (lr=1.9474e-04) (hash(x)=6904498)
706
+ 676 train 7.215933 (lr=1.9396e-04) (hash(x)=5378107)
707
+ 677 train 7.333801 (lr=1.9317e-04) (hash(x)=7765805)
708
+ 678 train 7.147527 (lr=1.9239e-04) (hash(x)=5871280)
709
+ 679 train 7.303738 (lr=1.9161e-04) (hash(x)=6543764)
710
+ 680 train 7.144301 (lr=1.9083e-04) (hash(x)=7369665)
711
+ 681 train 7.253389 (lr=1.9004e-04) (hash(x)=7204955)
712
+ 682 train 7.201522 (lr=1.8926e-04) (hash(x)=5856543)
713
+ 683 train 7.077721 (lr=1.8848e-04) (hash(x)=4423711)
714
+ 684 train 7.056482 (lr=1.8770e-04) (hash(x)=5818153)
715
+ 685 train 7.152046 (lr=1.8693e-04) (hash(x)=5343477)
716
+ 686 train 7.346313 (lr=1.8615e-04) (hash(x)=6581702)
717
+ 687 train 7.176725 (lr=1.8537e-04) (hash(x)=6595856)
718
+ 688 train 7.104743 (lr=1.8459e-04) (hash(x)=5325378)
719
+ 689 train 7.037066 (lr=1.8382e-04) (hash(x)=7229015)
720
+ 690 train 7.196552 (lr=1.8304e-04) (hash(x)=5918186)
721
+ 691 train 7.068548 (lr=1.8227e-04) (hash(x)=5916034)
722
+ 692 train 7.093777 (lr=1.8149e-04) (hash(x)=5280722)
723
+ 693 train 7.472866 (lr=1.8072e-04) (hash(x)=6939521)
724
+ 694 train 7.223158 (lr=1.7995e-04) (hash(x)=5172566)
725
+ 695 train 7.084385 (lr=1.7917e-04) (hash(x)=5492185)
726
+ 696 train 6.955538 (lr=1.7840e-04) (hash(x)=5779011)
727
+ 697 train 7.224815 (lr=1.7763e-04) (hash(x)=6898006)
728
+ 698 train 7.055333 (lr=1.7686e-04) (hash(x)=5779217)
729
+ 699 train 6.919622 (lr=1.7610e-04) (hash(x)=5462008)
730
+ 700 val loss 7.2164
731
+ 700 val perplexity 1361.5625
732
+ 700 train 7.072332 (lr=1.7533e-04) (hash(x)=6243843)
733
+ 701 train 7.131474 (lr=1.7456e-04) (hash(x)=7821902)
734
+ 702 train 7.196996 (lr=1.7380e-04) (hash(x)=5222501)
735
+ 703 train 7.155895 (lr=1.7303e-04) (hash(x)=6608108)
736
+ 704 train 7.264574 (lr=1.7227e-04) (hash(x)=7602879)
737
+ 705 train 7.193008 (lr=1.7150e-04) (hash(x)=6295394)
738
+ 706 train 7.152050 (lr=1.7074e-04) (hash(x)=6579907)
739
+ 707 train 7.041983 (lr=1.6998e-04) (hash(x)=6803449)
740
+ 708 train 7.111906 (lr=1.6922e-04) (hash(x)=5413733)
741
+ 709 train 7.073820 (lr=1.6846e-04) (hash(x)=7444134)
742
+ 710 train 7.233346 (lr=1.6770e-04) (hash(x)=7028652)
743
+ 711 train 7.037975 (lr=1.6694e-04) (hash(x)=5361923)
744
+ 712 train 7.141428 (lr=1.6619e-04) (hash(x)=5478293)
745
+ 713 train 7.123958 (lr=1.6543e-04) (hash(x)=7803039)
746
+ 714 train 7.308537 (lr=1.6468e-04) (hash(x)=6469017)
747
+ 715 train 7.208521 (lr=1.6393e-04) (hash(x)=7042578)
748
+ 716 train 7.267323 (lr=1.6317e-04) (hash(x)=6927957)
749
+ 717 train 7.069807 (lr=1.6242e-04) (hash(x)=6647007)
750
+ 718 train 7.038350 (lr=1.6167e-04) (hash(x)=6230888)
751
+ 719 train 7.319347 (lr=1.6092e-04) (hash(x)=6303658)
752
+ 720 train 7.342918 (lr=1.6018e-04) (hash(x)=6769880)
753
+ 721 train 7.181073 (lr=1.5943e-04) (hash(x)=6158927)
754
+ 722 train 7.158894 (lr=1.5868e-04) (hash(x)=5991631)
755
+ 723 train 7.033717 (lr=1.5794e-04) (hash(x)=6419656)
756
+ 724 train 7.028175 (lr=1.5720e-04) (hash(x)=7739209)
757
+ 725 train 6.981717 (lr=1.5646e-04) (hash(x)=7891271)
758
+ 726 train 7.347407 (lr=1.5572e-04) (hash(x)=7171433)
759
+ 727 train 7.125878 (lr=1.5498e-04) (hash(x)=7099888)
760
+ 728 train 7.128310 (lr=1.5424e-04) (hash(x)=6103730)
761
+ 729 train 7.235734 (lr=1.5350e-04) (hash(x)=6352424)
762
+ 730 train 7.133034 (lr=1.5277e-04) (hash(x)=4893671)
763
+ 731 train 7.091638 (lr=1.5203e-04) (hash(x)=5863145)
764
+ 732 train 7.193548 (lr=1.5130e-04) (hash(x)=5502233)
765
+ 733 train 7.107647 (lr=1.5057e-04) (hash(x)=6844597)
766
+ 734 train 7.123130 (lr=1.4984e-04) (hash(x)=6744526)
767
+ 735 train 7.267018 (lr=1.4911e-04) (hash(x)=6009861)
768
+ 736 train 7.126598 (lr=1.4838e-04) (hash(x)=6152358)
769
+ 737 train 7.054047 (lr=1.4766e-04) (hash(x)=5374957)
770
+ 738 train 7.081131 (lr=1.4693e-04) (hash(x)=6495309)
771
+ 739 train 7.229886 (lr=1.4621e-04) (hash(x)=3923225)
772
+ 740 train 7.007717 (lr=1.4549e-04) (hash(x)=6255243)
773
+ 741 train 7.162672 (lr=1.4477e-04) (hash(x)=5848862)
774
+ 742 train 7.171687 (lr=1.4405e-04) (hash(x)=4535000)
775
+ 743 train 7.044148 (lr=1.4333e-04) (hash(x)=4440634)
776
+ 744 train 7.043993 (lr=1.4262e-04) (hash(x)=5222077)
777
+ 745 train 7.089469 (lr=1.4190e-04) (hash(x)=4788998)
778
+ 746 train 7.086133 (lr=1.4119e-04) (hash(x)=6108350)
779
+ 747 train 7.002780 (lr=1.4048e-04) (hash(x)=7378165)
780
+ 748 train 7.083867 (lr=1.3977e-04) (hash(x)=5475646)
781
+ 749 train 7.081274 (lr=1.3906e-04) (hash(x)=4771243)
782
+ 750 val loss 7.1958
783
+ 750 val perplexity 1333.7659
784
+ 750 train 6.975585 (lr=1.3835e-04) (hash(x)=6668088)
785
+ 751 train 7.192849 (lr=1.3765e-04) (hash(x)=5716089)
786
+ 752 train 7.151796 (lr=1.3694e-04) (hash(x)=6991946)
787
+ 753 train 7.062603 (lr=1.3624e-04) (hash(x)=7991009)
788
+ 754 train 7.149424 (lr=1.3554e-04) (hash(x)=6037018)
789
+ 755 train 7.115553 (lr=1.3484e-04) (hash(x)=5348607)
790
+ 756 train 7.266709 (lr=1.3415e-04) (hash(x)=5190526)
791
+ 757 train 7.138197 (lr=1.3345e-04) (hash(x)=6036146)
792
+ 758 train 7.033880 (lr=1.3276e-04) (hash(x)=5329783)
793
+ 759 train 7.137833 (lr=1.3207e-04) (hash(x)=6658718)
794
+ 760 train 7.036718 (lr=1.3138e-04) (hash(x)=4820939)
795
+ 761 train 7.045386 (lr=1.3069e-04) (hash(x)=6176253)
796
+ 762 train 7.003334 (lr=1.3000e-04) (hash(x)=5316750)
797
+ 763 train 7.104775 (lr=1.2931e-04) (hash(x)=4542895)
798
+ 764 train 6.868597 (lr=1.2863e-04) (hash(x)=5395673)
799
+ 765 train 6.986865 (lr=1.2795e-04) (hash(x)=5549073)
800
+ 766 train 7.255738 (lr=1.2727e-04) (hash(x)=5233723)
801
+ 767 train 7.305648 (lr=1.2659e-04) (hash(x)=5342783)
802
+ 768 train 7.237090 (lr=1.2592e-04) (hash(x)=5552448)
803
+ 769 train 7.376866 (lr=1.2524e-04) (hash(x)=6549546)
804
+ 770 train 7.020038 (lr=1.2457e-04) (hash(x)=3816493)
805
+ 771 train 7.054595 (lr=1.2390e-04) (hash(x)=6170599)
806
+ 772 train 7.553206 (lr=1.2323e-04) (hash(x)=6708759)
807
+ 773 train 7.161798 (lr=1.2256e-04) (hash(x)=8483134)
808
+ 774 train 7.740232 (lr=1.2190e-04) (hash(x)=5583732)
809
+ 775 train 7.171287 (lr=1.2124e-04) (hash(x)=4865036)
810
+ 776 train 7.080164 (lr=1.2057e-04) (hash(x)=7234336)
811
+ 777 train 7.157468 (lr=1.1992e-04) (hash(x)=4433890)
812
+ 778 train 7.468447 (lr=1.1926e-04) (hash(x)=7484636)
813
+ 779 train 7.882348 (lr=1.1860e-04) (hash(x)=6738597)
814
+ 780 train 7.116121 (lr=1.1795e-04) (hash(x)=6752421)
815
+ 781 train 7.308029 (lr=1.1730e-04) (hash(x)=5459620)
816
+ 782 train 7.220198 (lr=1.1665e-04) (hash(x)=6718131)
817
+ 783 train 7.192760 (lr=1.1600e-04) (hash(x)=4604074)
818
+ 784 train 7.066220 (lr=1.1536e-04) (hash(x)=5946124)
819
+ 785 train 7.350662 (lr=1.1471e-04) (hash(x)=8045501)
820
+ 786 train 7.442341 (lr=1.1407e-04) (hash(x)=5712725)
821
+ 787 train 7.169938 (lr=1.1343e-04) (hash(x)=6176464)
822
+ 788 train 7.217295 (lr=1.1279e-04) (hash(x)=6276702)
823
+ 789 train 7.029304 (lr=1.1216e-04) (hash(x)=6127317)
824
+ 790 train 7.185762 (lr=1.1153e-04) (hash(x)=7221374)
825
+ 791 train 7.154645 (lr=1.1089e-04) (hash(x)=5571462)
826
+ 792 train 7.176119 (lr=1.1027e-04) (hash(x)=7518021)
827
+ 793 train 7.164327 (lr=1.0964e-04) (hash(x)=6640755)
828
+ 794 train 7.236927 (lr=1.0901e-04) (hash(x)=6870405)
829
+ 795 train 7.213480 (lr=1.0839e-04) (hash(x)=6147295)
830
+ 796 train 7.096205 (lr=1.0777e-04) (hash(x)=6782583)
831
+ 797 train 7.221269 (lr=1.0715e-04) (hash(x)=6120461)
832
+ 798 train 7.153096 (lr=1.0654e-04) (hash(x)=5893870)
833
+ 799 train 7.097794 (lr=1.0592e-04) (hash(x)=5123928)
834
+ 800 val loss 7.1858
835
+ 800 val perplexity 1320.5674
836
+ 800 train 7.167883 (lr=1.0531e-04) (hash(x)=7037251)
837
+ 801 train 7.153453 (lr=1.0470e-04) (hash(x)=5428813)
838
+ 802 train 7.104988 (lr=1.0410e-04) (hash(x)=5550387)
839
+ 803 train 7.164821 (lr=1.0349e-04) (hash(x)=6448586)
840
+ 804 train 7.310443 (lr=1.0289e-04) (hash(x)=5916934)
841
+ 805 train 7.270067 (lr=1.0229e-04) (hash(x)=8116987)
842
+ 806 train 7.249178 (lr=1.0169e-04) (hash(x)=6210899)
843
+ 807 train 7.352675 (lr=1.0109e-04) (hash(x)=5974342)
844
+ 808 train 7.341385 (lr=1.0050e-04) (hash(x)=7236199)
845
+ 809 train 7.272756 (lr=9.9910e-05) (hash(x)=5261336)
846
+ 810 train 7.091850 (lr=9.9321e-05) (hash(x)=6310720)
847
+ 811 train 7.140950 (lr=9.8735e-05) (hash(x)=5307056)
848
+ 812 train 6.960600 (lr=9.8151e-05) (hash(x)=7356633)
849
+ 813 train 7.095349 (lr=9.7569e-05) (hash(x)=5865119)
850
+ 814 train 6.991228 (lr=9.6990e-05) (hash(x)=5979205)
851
+ 815 train 7.159368 (lr=9.6413e-05) (hash(x)=6672380)
852
+ 816 train 7.143548 (lr=9.5838e-05) (hash(x)=5565030)
853
+ 817 train 7.114964 (lr=9.5266e-05) (hash(x)=6886810)
854
+ 818 train 7.167445 (lr=9.4696e-05) (hash(x)=5926478)
855
+ 819 train 7.377586 (lr=9.4129e-05) (hash(x)=6564071)
856
+ 820 train 6.997161 (lr=9.3564e-05) (hash(x)=5024280)
857
+ 821 train 6.998382 (lr=9.3001e-05) (hash(x)=7705089)
858
+ 822 train 6.974398 (lr=9.2441e-05) (hash(x)=4037371)
859
+ 823 train 7.087169 (lr=9.1884e-05) (hash(x)=5773073)
860
+ 824 train 7.102084 (lr=9.1328e-05) (hash(x)=7158139)
861
+ 825 train 6.941420 (lr=9.0776e-05) (hash(x)=5905976)
862
+ 826 train 7.371319 (lr=9.0226e-05) (hash(x)=13897369)
863
+ 827 train 8.487906 (lr=8.9678e-05) (hash(x)=6936653)
864
+ 828 train 7.124721 (lr=8.9133e-05) (hash(x)=6592246)
865
+ 829 train 7.107628 (lr=8.8591e-05) (hash(x)=6592773)
866
+ 830 train 7.165405 (lr=8.8051e-05) (hash(x)=5461178)
867
+ 831 train 6.931377 (lr=8.7513e-05) (hash(x)=6796551)
868
+ 832 train 7.151716 (lr=8.6978e-05) (hash(x)=4721691)
869
+ 833 train 7.071537 (lr=8.6446e-05) (hash(x)=6151589)
870
+ 834 train 7.216248 (lr=8.5916e-05) (hash(x)=7008995)
871
+ 835 train 7.537697 (lr=8.5389e-05) (hash(x)=7045164)
872
+ 836 train 7.138468 (lr=8.4865e-05) (hash(x)=6614221)
873
+ 837 train 7.166931 (lr=8.4343e-05) (hash(x)=6650506)
874
+ 838 train 6.991625 (lr=8.3824e-05) (hash(x)=6363762)
875
+ 839 train 7.156168 (lr=8.3307e-05) (hash(x)=7839748)
876
+ 840 train 7.092438 (lr=8.2793e-05) (hash(x)=5124077)
877
+ 841 train 7.521439 (lr=8.2282e-05) (hash(x)=6028671)
878
+ 842 train 7.002462 (lr=8.1773e-05) (hash(x)=5427175)
879
+ 843 train 7.344855 (lr=8.1267e-05) (hash(x)=6118408)
880
+ 844 train 7.103992 (lr=8.0764e-05) (hash(x)=6022552)
881
+ 845 train 7.126197 (lr=8.0263e-05) (hash(x)=5757526)
882
+ 846 train 7.593597 (lr=7.9765e-05) (hash(x)=6476549)
883
+ 847 train 6.912841 (lr=7.9270e-05) (hash(x)=6030172)
884
+ 848 train 7.301657 (lr=7.8778e-05) (hash(x)=5917229)
885
+ 849 train 7.360444 (lr=7.8288e-05) (hash(x)=6719882)
886
+ 850 val loss 7.1740
887
+ 850 val perplexity 1305.0623
888
+ 850 train 7.220369 (lr=7.7801e-05) (hash(x)=5441498)
889
+ 851 train 7.072789 (lr=7.7317e-05) (hash(x)=5028050)
890
+ 852 train 7.136975 (lr=7.6836e-05) (hash(x)=8487491)
891
+ 853 train 7.131804 (lr=7.6357e-05) (hash(x)=5682910)
892
+ 854 train 7.330167 (lr=7.5881e-05) (hash(x)=6781362)
893
+ 855 train 7.146786 (lr=7.5408e-05) (hash(x)=5747891)
894
+ 856 train 7.178394 (lr=7.4938e-05) (hash(x)=7627661)
895
+ 857 train 6.840408 (lr=7.4470e-05) (hash(x)=5734208)
896
+ 858 train 7.005817 (lr=7.4005e-05) (hash(x)=6162015)
897
+ 859 train 7.214223 (lr=7.3544e-05) (hash(x)=7024381)
898
+ 860 train 7.427608 (lr=7.3085e-05) (hash(x)=5198498)
899
+ 861 train 7.415044 (lr=7.2628e-05) (hash(x)=7814361)
900
+ 862 train 7.261909 (lr=7.2175e-05) (hash(x)=7394995)
901
+ 863 train 7.165792 (lr=7.1725e-05) (hash(x)=5765893)
902
+ 864 train 6.968561 (lr=7.1277e-05) (hash(x)=4990435)
903
+ 865 train 7.116785 (lr=7.0832e-05) (hash(x)=5572048)
904
+ 866 train 7.261439 (lr=7.0390e-05) (hash(x)=7093293)
905
+ 867 train 7.137643 (lr=6.9952e-05) (hash(x)=5610058)
906
+ 868 train 7.045588 (lr=6.9516e-05) (hash(x)=6771468)
907
+ 869 train 7.076818 (lr=6.9082e-05) (hash(x)=4459483)
908
+ 870 train 6.974629 (lr=6.8652e-05) (hash(x)=4926546)
909
+ 871 train 7.073829 (lr=6.8225e-05) (hash(x)=6758411)
910
+ 872 train 7.050490 (lr=6.7801e-05) (hash(x)=7394277)
911
+ 873 train 7.176453 (lr=6.7379e-05) (hash(x)=6701381)
912
+ 874 train 7.527780 (lr=6.6961e-05) (hash(x)=8120966)
913
+ 875 train 7.029181 (lr=6.6545e-05) (hash(x)=5416400)
914
+ 876 train 7.045394 (lr=6.6133e-05) (hash(x)=6292778)
915
+ 877 train 6.982659 (lr=6.5723e-05) (hash(x)=5919569)
916
+ 878 train 7.292915 (lr=6.5317e-05) (hash(x)=7561740)
917
+ 879 train 7.048343 (lr=6.4913e-05) (hash(x)=5282870)
918
+ 880 train 7.077844 (lr=6.4513e-05) (hash(x)=5473238)
919
+ 881 train 6.951062 (lr=6.4115e-05) (hash(x)=8072181)
920
+ 882 train 6.977979 (lr=6.3721e-05) (hash(x)=6757322)
921
+ 883 train 6.942024 (lr=6.3329e-05) (hash(x)=7996770)
922
+ 884 train 7.021532 (lr=6.2941e-05) (hash(x)=3181021)
923
+ 885 train 7.057776 (lr=6.2556e-05) (hash(x)=5093519)
924
+ 886 train 7.053750 (lr=6.2173e-05) (hash(x)=5391352)
925
+ 887 train 7.087307 (lr=6.1794e-05) (hash(x)=6270917)
926
+ 888 train 7.082528 (lr=6.1418e-05) (hash(x)=5818195)
927
+ 889 train 6.850028 (lr=6.1045e-05) (hash(x)=5852780)
928
+ 890 train 6.771029 (lr=6.0675e-05) (hash(x)=5777928)
929
+ 891 train 6.941120 (lr=6.0308e-05) (hash(x)=5327728)
930
+ 892 train 6.976896 (lr=5.9944e-05) (hash(x)=7651281)
931
+ 893 train 7.194061 (lr=5.9583e-05) (hash(x)=6260765)
932
+ 894 train 7.064901 (lr=5.9225e-05) (hash(x)=6265102)
933
+ 895 train 7.251947 (lr=5.8871e-05) (hash(x)=6212894)
934
+ 896 train 7.007589 (lr=5.8519e-05) (hash(x)=6901967)
935
+ 897 train 7.082289 (lr=5.8171e-05) (hash(x)=6724010)
936
+ 898 train 7.069149 (lr=5.7826e-05) (hash(x)=6299033)
937
+ 899 train 7.237507 (lr=5.7484e-05) (hash(x)=6059934)
938
+ 900 val loss 7.1657
939
+ 900 val perplexity 1294.3068
940
+ 900 train 6.943421 (lr=5.7145e-05) (hash(x)=6029696)
941
+ 901 train 6.854955 (lr=5.6809e-05) (hash(x)=5587382)
942
+ 902 train 7.071787 (lr=5.6476e-05) (hash(x)=5890076)
943
+ 903 train 6.993479 (lr=5.6147e-05) (hash(x)=7274108)
944
+ 904 train 7.086677 (lr=5.5821e-05) (hash(x)=7975528)
945
+ 905 train 7.135279 (lr=5.5497e-05) (hash(x)=6085396)
946
+ 906 train 7.020583 (lr=5.5178e-05) (hash(x)=6724496)
947
+ 907 train 7.017730 (lr=5.4861e-05) (hash(x)=4784666)
948
+ 908 train 7.009595 (lr=5.4547e-05) (hash(x)=7193845)
949
+ 909 train 7.177256 (lr=5.4237e-05) (hash(x)=6288638)
950
+ 910 train 7.141451 (lr=5.3930e-05) (hash(x)=5685285)
951
+ 911 train 7.123322 (lr=5.3626e-05) (hash(x)=7714288)
952
+ 912 train 7.217126 (lr=5.3325e-05) (hash(x)=5983679)
953
+ 913 train 7.159458 (lr=5.3028e-05) (hash(x)=5534748)
954
+ 914 train 7.093553 (lr=5.2734e-05) (hash(x)=6545471)
955
+ 915 train 7.090096 (lr=5.2443e-05) (hash(x)=6361270)
956
+ 916 train 7.092950 (lr=5.2155e-05) (hash(x)=6257744)
957
+ 917 train 7.096164 (lr=5.1871e-05) (hash(x)=4682956)
958
+ 918 train 6.988361 (lr=5.1589e-05) (hash(x)=4521756)
959
+ 919 train 7.126476 (lr=5.1311e-05) (hash(x)=7568935)
960
+ 920 train 7.045220 (lr=5.1037e-05) (hash(x)=7969353)
961
+ 921 train 7.253804 (lr=5.0765e-05) (hash(x)=7973311)
962
+ 922 train 6.916569 (lr=5.0497e-05) (hash(x)=4843937)
963
+ 923 train 7.070973 (lr=5.0232e-05) (hash(x)=6987352)
964
+ 924 train 7.215007 (lr=4.9971e-05) (hash(x)=7179418)
965
+ 925 train 6.840687 (lr=4.9712e-05) (hash(x)=5413904)
966
+ 926 train 7.272241 (lr=4.9457e-05) (hash(x)=6028105)
967
+ 927 train 7.176475 (lr=4.9206e-05) (hash(x)=6732924)
968
+ 928 train 7.066581 (lr=4.8957e-05) (hash(x)=4547651)
969
+ 929 train 7.206191 (lr=4.8712e-05) (hash(x)=6219754)
970
+ 930 train 6.977387 (lr=4.8470e-05) (hash(x)=7465186)
971
+ 931 train 7.135737 (lr=4.8232e-05) (hash(x)=6839070)
972
+ 932 train 7.015966 (lr=4.7997e-05) (hash(x)=5390324)
973
+ 933 train 6.991179 (lr=4.7765e-05) (hash(x)=6090120)
974
+ 934 train 7.602218 (lr=4.7537e-05) (hash(x)=7123181)
975
+ 935 train 7.291836 (lr=4.7312e-05) (hash(x)=7115176)
976
+ 936 train 7.265821 (lr=4.7090e-05) (hash(x)=7016990)
977
+ 937 train 7.070276 (lr=4.6871e-05) (hash(x)=4989299)
978
+ 938 train 7.080413 (lr=4.6656e-05) (hash(x)=6783311)
979
+ 939 train 7.054582 (lr=4.6445e-05) (hash(x)=7497590)
980
+ 940 train 7.281096 (lr=4.6236e-05) (hash(x)=6767514)
981
+ 941 train 7.032323 (lr=4.6031e-05) (hash(x)=3977557)
982
+ 942 train 7.171589 (lr=4.5830e-05) (hash(x)=4806492)
983
+ 943 train 7.065566 (lr=4.5631e-05) (hash(x)=5524165)
984
+ 944 train 7.119200 (lr=4.5437e-05) (hash(x)=4930752)
985
+ 945 train 7.072712 (lr=4.5245e-05) (hash(x)=5124936)
986
+ 946 train 6.970873 (lr=4.5057e-05) (hash(x)=6877351)
987
+ 947 train 7.258476 (lr=4.4872e-05) (hash(x)=5522119)
988
+ 948 train 7.072148 (lr=4.4691e-05) (hash(x)=6183512)
989
+ 949 train 6.988660 (lr=4.4513e-05) (hash(x)=5030865)
990
+ 950 val loss 7.1565
991
+ 950 val perplexity 1282.4493
992
+ 950 train 7.080802 (lr=4.4338e-05) (hash(x)=6619812)
993
+ 951 train 7.092521 (lr=4.4167e-05) (hash(x)=6411783)
994
+ 952 train 7.075318 (lr=4.4000e-05) (hash(x)=6748982)
995
+ 953 train 6.979934 (lr=4.3835e-05) (hash(x)=6581937)
996
+ 954 train 7.094067 (lr=4.3674e-05) (hash(x)=7881022)
997
+ 955 train 7.049211 (lr=4.3517e-05) (hash(x)=5116910)
998
+ 956 train 6.840390 (lr=4.3363e-05) (hash(x)=4880833)
999
+ 957 train 6.897750 (lr=4.3212e-05) (hash(x)=7353459)
1000
+ 958 train 6.958652 (lr=4.3065e-05) (hash(x)=4951302)
1001
+ 959 train 6.859964 (lr=4.2921e-05) (hash(x)=6344535)
1002
+ 960 train 7.482074 (lr=4.2781e-05) (hash(x)=7411763)
1003
+ 961 train 7.056205 (lr=4.2644e-05) (hash(x)=5840169)
1004
+ 962 train 7.154108 (lr=4.2510e-05) (hash(x)=5784788)
1005
+ 963 train 6.964046 (lr=4.2380e-05) (hash(x)=6754290)
1006
+ 964 train 7.243752 (lr=4.2253e-05) (hash(x)=6369963)
1007
+ 965 train 7.322361 (lr=4.2130e-05) (hash(x)=6484573)
1008
+ 966 train 6.894042 (lr=4.2010e-05) (hash(x)=6516338)
1009
+ 967 train 6.932183 (lr=4.1894e-05) (hash(x)=5260125)
1010
+ 968 train 7.070291 (lr=4.1781e-05) (hash(x)=5781624)
1011
+ 969 train 6.851396 (lr=4.1672e-05) (hash(x)=4881179)
1012
+ 970 train 6.993733 (lr=4.1566e-05) (hash(x)=6707821)
1013
+ 971 train 7.134998 (lr=4.1463e-05) (hash(x)=5001634)
1014
+ 972 train 6.999440 (lr=4.1364e-05) (hash(x)=4402949)
1015
+ 973 train 7.020264 (lr=4.1269e-05) (hash(x)=7424810)
1016
+ 974 train 6.895786 (lr=4.1177e-05) (hash(x)=4814803)
1017
+ 975 train 6.816129 (lr=4.1088e-05) (hash(x)=5176431)
1018
+ 976 train 7.014992 (lr=4.1003e-05) (hash(x)=5769393)
1019
+ 977 train 6.951116 (lr=4.0921e-05) (hash(x)=5421624)
1020
+ 978 train 7.168338 (lr=4.0843e-05) (hash(x)=5999773)
1021
+ 979 train 6.877117 (lr=4.0768e-05) (hash(x)=5674439)
1022
+ 980 train 6.972448 (lr=4.0697e-05) (hash(x)=6345351)
1023
+ 981 train 6.912955 (lr=4.0629e-05) (hash(x)=5819157)
1024
+ 982 train 7.121727 (lr=4.0564e-05) (hash(x)=11159355)
1025
+ 983 train 7.088211 (lr=4.0503e-05) (hash(x)=6506856)
1026
+ 984 train 7.000676 (lr=4.0446e-05) (hash(x)=6331023)
1027
+ 985 train 6.903489 (lr=4.0392e-05) (hash(x)=6119124)
1028
+ 986 train 7.111833 (lr=4.0341e-05) (hash(x)=8142528)
1029
+ 987 train 7.293222 (lr=4.0294e-05) (hash(x)=6241669)
1030
+ 988 train 6.915487 (lr=4.0251e-05) (hash(x)=3852991)
1031
+ 989 train 6.638735 (lr=4.0211e-05) (hash(x)=4992756)
1032
+ 990 train 7.366664 (lr=4.0174e-05) (hash(x)=7097145)
1033
+ 991 train 7.150247 (lr=4.0141e-05) (hash(x)=7277927)
1034
+ 992 train 7.241073 (lr=4.0112e-05) (hash(x)=5661400)
1035
+ 993 train 7.138251 (lr=4.0085e-05) (hash(x)=6622097)
1036
+ 994 train 7.261961 (lr=4.0063e-05) (hash(x)=5016804)
1037
+ 995 train 7.346981 (lr=4.0044e-05) (hash(x)=7026582)
1038
+ 996 train 7.134982 (lr=4.0028e-05) (hash(x)=5639136)
1039
+ 997 train 7.260903 (lr=4.0016e-05) (hash(x)=5840987)
1040
+ 998 train 6.904773 (lr=4.0007e-05) (hash(x)=5265410)
1041
+ 999 val loss 7.1602
1042
+ 999 val perplexity 1287.1317
1043
+ 999 train 6.898254 (lr=4.0002e-05) (hash(x)=4351074)
lr4e-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 sha256:27127df628df234d0adba59fb5c29db60cb074058c14aac4b6466ae15776604d
3
+ size 36163010
lr4e-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 sha256:07a0644e1f55aac7f84e481d119354796b46015a1fe461a993a7eb674b661f02
3
+ size 70764358