andrew-healey commited on
Commit
d42c17d
·
verified ·
1 Parent(s): e35a9db

Upload folder using huggingface_hub

Browse files
lr4e-4_total_batch_size40960_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr4e-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": 0.0004, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr4e-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
lr4e-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=1.3986e-06) (hash(x)=5534142)
5
+ 1 train 11.703205 (lr=2.7972e-06) (hash(x)=6328012)
6
+ 2 train 11.693671 (lr=4.1958e-06) (hash(x)=5556098)
7
+ 3 train 11.674109 (lr=5.5944e-06) (hash(x)=6219138)
8
+ 4 train 11.644265 (lr=6.9930e-06) (hash(x)=5491972)
9
+ 5 train 11.621668 (lr=8.3916e-06) (hash(x)=5917741)
10
+ 6 train 11.590139 (lr=9.7902e-06) (hash(x)=6112458)
11
+ 7 train 11.548161 (lr=1.1189e-05) (hash(x)=6568379)
12
+ 8 train 11.491055 (lr=1.2587e-05) (hash(x)=7386082)
13
+ 9 train 11.451992 (lr=1.3986e-05) (hash(x)=6151674)
14
+ 10 train 11.421305 (lr=1.5385e-05) (hash(x)=7859060)
15
+ 11 train 11.352206 (lr=1.6783e-05) (hash(x)=6391506)
16
+ 12 train 11.296203 (lr=1.8182e-05) (hash(x)=5985675)
17
+ 13 train 11.237391 (lr=1.9580e-05) (hash(x)=6984772)
18
+ 14 train 11.164810 (lr=2.0979e-05) (hash(x)=6603717)
19
+ 15 train 11.073522 (lr=2.2378e-05) (hash(x)=6190579)
20
+ 16 train 11.004282 (lr=2.3776e-05) (hash(x)=7276303)
21
+ 17 train 10.991491 (lr=2.5175e-05) (hash(x)=6630378)
22
+ 18 train 10.943866 (lr=2.6573e-05) (hash(x)=9989744)
23
+ 19 train 10.810376 (lr=2.7972e-05) (hash(x)=6317894)
24
+ 20 train 10.716109 (lr=2.9371e-05) (hash(x)=7947703)
25
+ 21 train 10.652974 (lr=3.0769e-05) (hash(x)=5777498)
26
+ 22 train 10.623652 (lr=3.2168e-05) (hash(x)=7723226)
27
+ 23 train 10.563340 (lr=3.3566e-05) (hash(x)=6491229)
28
+ 24 train 10.606959 (lr=3.4965e-05) (hash(x)=6019389)
29
+ 25 train 10.544305 (lr=3.6364e-05) (hash(x)=4648609)
30
+ 26 train 10.480566 (lr=3.7762e-05) (hash(x)=5558355)
31
+ 27 train 10.496490 (lr=3.9161e-05) (hash(x)=6369326)
32
+ 28 train 10.436907 (lr=4.0559e-05) (hash(x)=4906853)
33
+ 29 train 10.337166 (lr=4.1958e-05) (hash(x)=6437376)
34
+ 30 train 10.424633 (lr=4.3357e-05) (hash(x)=7084093)
35
+ 31 train 10.366438 (lr=4.4755e-05) (hash(x)=7048804)
36
+ 32 train 10.380322 (lr=4.6154e-05) (hash(x)=5752594)
37
+ 33 train 10.201049 (lr=4.7552e-05) (hash(x)=6393520)
38
+ 34 train 10.399938 (lr=4.8951e-05) (hash(x)=6793550)
39
+ 35 train 10.337663 (lr=5.0350e-05) (hash(x)=6581348)
40
+ 36 train 10.318378 (lr=5.1748e-05) (hash(x)=6235837)
41
+ 37 train 10.360291 (lr=5.3147e-05) (hash(x)=7083021)
42
+ 38 train 10.342774 (lr=5.4545e-05) (hash(x)=6155821)
43
+ 39 train 10.252976 (lr=5.5944e-05) (hash(x)=5657123)
44
+ 40 train 10.163769 (lr=5.7343e-05) (hash(x)=5312267)
45
+ 41 train 10.268682 (lr=5.8741e-05) (hash(x)=6830813)
46
+ 42 train 10.233603 (lr=6.0140e-05) (hash(x)=6424933)
47
+ 43 train 10.192308 (lr=6.1538e-05) (hash(x)=6680023)
48
+ 44 train 10.154663 (lr=6.2937e-05) (hash(x)=7523279)
49
+ 45 train 10.094012 (lr=6.4336e-05) (hash(x)=6635925)
50
+ 46 train 10.130739 (lr=6.5734e-05) (hash(x)=5869441)
51
+ 47 train 10.111992 (lr=6.7133e-05) (hash(x)=6349004)
52
+ 48 train 10.017846 (lr=6.8531e-05) (hash(x)=6008454)
53
+ 49 train 10.018545 (lr=6.9930e-05) (hash(x)=7682741)
54
+ 50 val loss 10.0566
55
+ 50 val perplexity 23308.0664
56
+ 50 train 10.064154 (lr=7.1329e-05) (hash(x)=7794028)
57
+ 51 train 10.022840 (lr=7.2727e-05) (hash(x)=6166062)
58
+ 52 train 9.962205 (lr=7.4126e-05) (hash(x)=4811192)
59
+ 53 train 9.918077 (lr=7.5524e-05) (hash(x)=7822680)
60
+ 54 train 9.858716 (lr=7.6923e-05) (hash(x)=6431409)
61
+ 55 train 9.861258 (lr=7.8322e-05) (hash(x)=5323032)
62
+ 56 train 9.806050 (lr=7.9720e-05) (hash(x)=7363286)
63
+ 57 train 9.802444 (lr=8.1119e-05) (hash(x)=5616075)
64
+ 58 train 9.799147 (lr=8.2517e-05) (hash(x)=5672608)
65
+ 59 train 9.725146 (lr=8.3916e-05) (hash(x)=6839867)
66
+ 60 train 9.653217 (lr=8.5315e-05) (hash(x)=4916928)
67
+ 61 train 9.664149 (lr=8.6713e-05) (hash(x)=8353379)
68
+ 62 train 9.590437 (lr=8.8112e-05) (hash(x)=4838871)
69
+ 63 train 9.468619 (lr=8.9510e-05) (hash(x)=6795665)
70
+ 64 train 9.465662 (lr=9.0909e-05) (hash(x)=6678038)
71
+ 65 train 9.494607 (lr=9.2308e-05) (hash(x)=6112215)
72
+ 66 train 9.427754 (lr=9.3706e-05) (hash(x)=5915361)
73
+ 67 train 9.312510 (lr=9.5105e-05) (hash(x)=5771588)
74
+ 68 train 9.273679 (lr=9.6503e-05) (hash(x)=5688829)
75
+ 69 train 9.181032 (lr=9.7902e-05) (hash(x)=5443305)
76
+ 70 train 9.284546 (lr=9.9301e-05) (hash(x)=6901933)
77
+ 71 train 9.169958 (lr=1.0070e-04) (hash(x)=4294425)
78
+ 72 train 9.075689 (lr=1.0210e-04) (hash(x)=4363941)
79
+ 73 train 9.149575 (lr=1.0350e-04) (hash(x)=5486335)
80
+ 74 train 9.056443 (lr=1.0490e-04) (hash(x)=7295165)
81
+ 75 train 9.018016 (lr=1.0629e-04) (hash(x)=6267424)
82
+ 76 train 8.985387 (lr=1.0769e-04) (hash(x)=5270452)
83
+ 77 train 8.889153 (lr=1.0909e-04) (hash(x)=6305297)
84
+ 78 train 8.974252 (lr=1.1049e-04) (hash(x)=6040077)
85
+ 79 train 8.864988 (lr=1.1189e-04) (hash(x)=7130251)
86
+ 80 train 8.644584 (lr=1.1329e-04) (hash(x)=6508350)
87
+ 81 train 8.762431 (lr=1.1469e-04) (hash(x)=5859576)
88
+ 82 train 8.482652 (lr=1.1608e-04) (hash(x)=4892065)
89
+ 83 train 8.580753 (lr=1.1748e-04) (hash(x)=7317289)
90
+ 84 train 8.608212 (lr=1.1888e-04) (hash(x)=5510218)
91
+ 85 train 8.649257 (lr=1.2028e-04) (hash(x)=6752265)
92
+ 86 train 8.474513 (lr=1.2168e-04) (hash(x)=6530867)
93
+ 87 train 8.467771 (lr=1.2308e-04) (hash(x)=5790078)
94
+ 88 train 8.448227 (lr=1.2448e-04) (hash(x)=7614023)
95
+ 89 train 8.440626 (lr=1.2587e-04) (hash(x)=5245146)
96
+ 90 train 8.340357 (lr=1.2727e-04) (hash(x)=5786972)
97
+ 91 train 8.333920 (lr=1.2867e-04) (hash(x)=6346459)
98
+ 92 train 8.337217 (lr=1.3007e-04) (hash(x)=5209234)
99
+ 93 train 8.223918 (lr=1.3147e-04) (hash(x)=5676768)
100
+ 94 train 8.350408 (lr=1.3287e-04) (hash(x)=5863092)
101
+ 95 train 8.311888 (lr=1.3427e-04) (hash(x)=7019489)
102
+ 96 train 8.160555 (lr=1.3566e-04) (hash(x)=5483019)
103
+ 97 train 8.171141 (lr=1.3706e-04) (hash(x)=5968716)
104
+ 98 train 8.058758 (lr=1.3846e-04) (hash(x)=7780194)
105
+ 99 train 8.073570 (lr=1.3986e-04) (hash(x)=3784321)
106
+ 100 val loss 8.0696
107
+ 100 val perplexity 3195.9485
108
+ 100 train 8.064280 (lr=1.4126e-04) (hash(x)=7774109)
109
+ 101 train 8.180045 (lr=1.4266e-04) (hash(x)=6576749)
110
+ 102 train 8.073578 (lr=1.4406e-04) (hash(x)=7072404)
111
+ 103 train 7.956420 (lr=1.4545e-04) (hash(x)=6556025)
112
+ 104 train 8.027597 (lr=1.4685e-04) (hash(x)=7325027)
113
+ 105 train 8.020245 (lr=1.4825e-04) (hash(x)=6373412)
114
+ 106 train 7.903085 (lr=1.4965e-04) (hash(x)=5262868)
115
+ 107 train 7.826821 (lr=1.5105e-04) (hash(x)=8055563)
116
+ 108 train 7.935591 (lr=1.5245e-04) (hash(x)=5932790)
117
+ 109 train 7.868052 (lr=1.5385e-04) (hash(x)=5680154)
118
+ 110 train 7.917224 (lr=1.5524e-04) (hash(x)=5914217)
119
+ 111 train 7.827121 (lr=1.5664e-04) (hash(x)=6424128)
120
+ 112 train 7.772991 (lr=1.5804e-04) (hash(x)=7804089)
121
+ 113 train 7.871150 (lr=1.5944e-04) (hash(x)=6490149)
122
+ 114 train 7.796384 (lr=1.6084e-04) (hash(x)=6283540)
123
+ 115 train 7.758559 (lr=1.6224e-04) (hash(x)=5728337)
124
+ 116 train 7.756753 (lr=1.6364e-04) (hash(x)=7400272)
125
+ 117 train 7.764024 (lr=1.6503e-04) (hash(x)=6473432)
126
+ 118 train 7.830143 (lr=1.6643e-04) (hash(x)=6922646)
127
+ 119 train 7.720221 (lr=1.6783e-04) (hash(x)=7277141)
128
+ 120 train 7.771255 (lr=1.6923e-04) (hash(x)=7443124)
129
+ 121 train 7.761242 (lr=1.7063e-04) (hash(x)=5632735)
130
+ 122 train 7.569963 (lr=1.7203e-04) (hash(x)=5881015)
131
+ 123 train 7.719284 (lr=1.7343e-04) (hash(x)=5809411)
132
+ 124 train 7.717339 (lr=1.7483e-04) (hash(x)=7232964)
133
+ 125 train 7.694315 (lr=1.7622e-04) (hash(x)=6171829)
134
+ 126 train 7.770548 (lr=1.7762e-04) (hash(x)=5884456)
135
+ 127 train 7.719826 (lr=1.7902e-04) (hash(x)=5077031)
136
+ 128 train 7.773602 (lr=1.8042e-04) (hash(x)=6248895)
137
+ 129 train 7.691053 (lr=1.8182e-04) (hash(x)=6918004)
138
+ 130 train 7.744273 (lr=1.8322e-04) (hash(x)=6482466)
139
+ 131 train 7.666913 (lr=1.8462e-04) (hash(x)=6588573)
140
+ 132 train 7.723249 (lr=1.8601e-04) (hash(x)=6136939)
141
+ 133 train 7.631812 (lr=1.8741e-04) (hash(x)=7040157)
142
+ 134 train 7.725544 (lr=1.8881e-04) (hash(x)=6455595)
143
+ 135 train 7.708362 (lr=1.9021e-04) (hash(x)=4122983)
144
+ 136 train 7.641993 (lr=1.9161e-04) (hash(x)=4724052)
145
+ 137 train 7.603775 (lr=1.9301e-04) (hash(x)=6820797)
146
+ 138 train 7.836803 (lr=1.9441e-04) (hash(x)=7477594)
147
+ 139 train 7.581267 (lr=1.9580e-04) (hash(x)=6146305)
148
+ 140 train 7.606592 (lr=1.9720e-04) (hash(x)=6814831)
149
+ 141 train 7.608084 (lr=1.9860e-04) (hash(x)=5548298)
150
+ 142 train 7.768223 (lr=2.0000e-04) (hash(x)=7609416)
151
+ 143 train 7.703319 (lr=2.0140e-04) (hash(x)=7574342)
152
+ 144 train 7.566972 (lr=2.0280e-04) (hash(x)=5906937)
153
+ 145 train 7.609381 (lr=2.0420e-04) (hash(x)=8013471)
154
+ 146 train 7.596162 (lr=2.0559e-04) (hash(x)=6146213)
155
+ 147 train 7.629440 (lr=2.0699e-04) (hash(x)=6709902)
156
+ 148 train 7.770957 (lr=2.0839e-04) (hash(x)=6402571)
157
+ 149 train 7.589540 (lr=2.0979e-04) (hash(x)=7683614)
158
+ 150 val loss 7.6168
159
+ 150 val perplexity 2032.0188
160
+ 150 train 7.576081 (lr=2.1119e-04) (hash(x)=5453026)
161
+ 151 train 7.539551 (lr=2.1259e-04) (hash(x)=6260017)
162
+ 152 train 7.664032 (lr=2.1399e-04) (hash(x)=6591466)
163
+ 153 train 7.636136 (lr=2.1538e-04) (hash(x)=5501472)
164
+ 154 train 7.684352 (lr=2.1678e-04) (hash(x)=5871698)
165
+ 155 train 7.519178 (lr=2.1818e-04) (hash(x)=4282052)
166
+ 156 train 7.606510 (lr=2.1958e-04) (hash(x)=7824660)
167
+ 157 train 7.538606 (lr=2.2098e-04) (hash(x)=5954950)
168
+ 158 train 7.604222 (lr=2.2238e-04) (hash(x)=5484656)
169
+ 159 train 7.701617 (lr=2.2378e-04) (hash(x)=7031328)
170
+ 160 train 7.665685 (lr=2.2517e-04) (hash(x)=9157069)
171
+ 161 train 7.744207 (lr=2.2657e-04) (hash(x)=5996883)
172
+ 162 train 7.806075 (lr=2.2797e-04) (hash(x)=7645447)
173
+ 163 train 7.853721 (lr=2.2937e-04) (hash(x)=6336164)
174
+ 164 train 7.596636 (lr=2.3077e-04) (hash(x)=5224240)
175
+ 165 train 7.922683 (lr=2.3217e-04) (hash(x)=5447801)
176
+ 166 train 7.628397 (lr=2.3357e-04) (hash(x)=4155598)
177
+ 167 train 7.612489 (lr=2.3497e-04) (hash(x)=7482744)
178
+ 168 train 7.633993 (lr=2.3636e-04) (hash(x)=4733414)
179
+ 169 train 7.608324 (lr=2.3776e-04) (hash(x)=7806284)
180
+ 170 train 7.613298 (lr=2.3916e-04) (hash(x)=7236478)
181
+ 171 train 7.626079 (lr=2.4056e-04) (hash(x)=4656255)
182
+ 172 train 7.732102 (lr=2.4196e-04) (hash(x)=9298325)
183
+ 173 train 7.650301 (lr=2.4336e-04) (hash(x)=6287524)
184
+ 174 train 7.692192 (lr=2.4476e-04) (hash(x)=7895970)
185
+ 175 train 7.658209 (lr=2.4615e-04) (hash(x)=8847992)
186
+ 176 train 7.848678 (lr=2.4755e-04) (hash(x)=6338998)
187
+ 177 train 7.826061 (lr=2.4895e-04) (hash(x)=7104011)
188
+ 178 train 7.644137 (lr=2.5035e-04) (hash(x)=5573214)
189
+ 179 train 7.687749 (lr=2.5175e-04) (hash(x)=6882897)
190
+ 180 train 7.641420 (lr=2.5315e-04) (hash(x)=5854242)
191
+ 181 train 7.500236 (lr=2.5455e-04) (hash(x)=6972494)
192
+ 182 train 7.541924 (lr=2.5594e-04) (hash(x)=6796805)
193
+ 183 train 7.443000 (lr=2.5734e-04) (hash(x)=6139120)
194
+ 184 train 7.616415 (lr=2.5874e-04) (hash(x)=6486796)
195
+ 185 train 7.493693 (lr=2.6014e-04) (hash(x)=5578797)
196
+ 186 train 7.566456 (lr=2.6154e-04) (hash(x)=6895326)
197
+ 187 train 7.604846 (lr=2.6294e-04) (hash(x)=7783160)
198
+ 188 train 7.632280 (lr=2.6434e-04) (hash(x)=4867625)
199
+ 189 train 7.454949 (lr=2.6573e-04) (hash(x)=6631123)
200
+ 190 train 7.306797 (lr=2.6713e-04) (hash(x)=6710965)
201
+ 191 train 7.441695 (lr=2.6853e-04) (hash(x)=6294657)
202
+ 192 train 7.603365 (lr=2.6993e-04) (hash(x)=5443606)
203
+ 193 train 7.581280 (lr=2.7133e-04) (hash(x)=7035770)
204
+ 194 train 7.624101 (lr=2.7273e-04) (hash(x)=4899541)
205
+ 195 train 7.476818 (lr=2.7413e-04) (hash(x)=6227120)
206
+ 196 train 7.565859 (lr=2.7552e-04) (hash(x)=5981648)
207
+ 197 train 7.530153 (lr=2.7692e-04) (hash(x)=5558993)
208
+ 198 train 7.570823 (lr=2.7832e-04) (hash(x)=7015671)
209
+ 199 train 7.558886 (lr=2.7972e-04) (hash(x)=5931234)
210
+ 200 val loss 7.5594
211
+ 200 val perplexity 1918.6475
212
+ 200 train 7.528236 (lr=2.8112e-04) (hash(x)=4582159)
213
+ 201 train 7.548489 (lr=2.8252e-04) (hash(x)=5029472)
214
+ 202 train 7.468040 (lr=2.8392e-04) (hash(x)=5434102)
215
+ 203 train 7.742717 (lr=2.8531e-04) (hash(x)=6810638)
216
+ 204 train 7.556846 (lr=2.8671e-04) (hash(x)=6521562)
217
+ 205 train 7.584042 (lr=2.8811e-04) (hash(x)=7226932)
218
+ 206 train 7.480263 (lr=2.8951e-04) (hash(x)=6957849)
219
+ 207 train 7.576184 (lr=2.9091e-04) (hash(x)=5334281)
220
+ 208 train 7.674344 (lr=2.9231e-04) (hash(x)=5887983)
221
+ 209 train 7.513726 (lr=2.9371e-04) (hash(x)=6393576)
222
+ 210 train 7.725246 (lr=2.9510e-04) (hash(x)=5057959)
223
+ 211 train 7.780063 (lr=2.9650e-04) (hash(x)=8595271)
224
+ 212 train 7.599269 (lr=2.9790e-04) (hash(x)=8395167)
225
+ 213 train 7.706912 (lr=2.9930e-04) (hash(x)=7634417)
226
+ 214 train 7.450010 (lr=3.0070e-04) (hash(x)=5257771)
227
+ 215 train 7.587952 (lr=3.0210e-04) (hash(x)=7674238)
228
+ 216 train 7.684290 (lr=3.0350e-04) (hash(x)=5300555)
229
+ 217 train 7.459308 (lr=3.0490e-04) (hash(x)=6066139)
230
+ 218 train 7.557741 (lr=3.0629e-04) (hash(x)=6688212)
231
+ 219 train 7.748426 (lr=3.0769e-04) (hash(x)=6642404)
232
+ 220 train 7.517848 (lr=3.0909e-04) (hash(x)=4961570)
233
+ 221 train 7.369778 (lr=3.1049e-04) (hash(x)=5105190)
234
+ 222 train 7.681576 (lr=3.1189e-04) (hash(x)=7155991)
235
+ 223 train 7.517878 (lr=3.1329e-04) (hash(x)=5132972)
236
+ 224 train 7.510911 (lr=3.1469e-04) (hash(x)=6269946)
237
+ 225 train 7.531682 (lr=3.1608e-04) (hash(x)=5396133)
238
+ 226 train 7.528765 (lr=3.1748e-04) (hash(x)=6394083)
239
+ 227 train 7.505261 (lr=3.1888e-04) (hash(x)=5237804)
240
+ 228 train 7.536111 (lr=3.2028e-04) (hash(x)=5216638)
241
+ 229 train 7.477200 (lr=3.2168e-04) (hash(x)=6735565)
242
+ 230 train 7.583837 (lr=3.2308e-04) (hash(x)=5983575)
243
+ 231 train 7.551377 (lr=3.2448e-04) (hash(x)=5831619)
244
+ 232 train 7.388682 (lr=3.2587e-04) (hash(x)=9109083)
245
+ 233 train 7.598585 (lr=3.2727e-04) (hash(x)=6912967)
246
+ 234 train 7.749548 (lr=3.2867e-04) (hash(x)=5965674)
247
+ 235 train 7.633712 (lr=3.3007e-04) (hash(x)=5822533)
248
+ 236 train 7.532730 (lr=3.3147e-04) (hash(x)=5268216)
249
+ 237 train 7.440441 (lr=3.3287e-04) (hash(x)=6531900)
250
+ 238 train 7.416707 (lr=3.3427e-04) (hash(x)=5426499)
251
+ 239 train 7.909187 (lr=3.3566e-04) (hash(x)=7455372)
252
+ 240 train 8.181813 (lr=3.3706e-04) (hash(x)=8319119)
253
+ 241 train 8.144635 (lr=3.3846e-04) (hash(x)=7592128)
254
+ 242 train 8.154997 (lr=3.3986e-04) (hash(x)=8704170)
255
+ 243 train 7.760542 (lr=3.4126e-04) (hash(x)=7165037)
256
+ 244 train 7.640127 (lr=3.4266e-04) (hash(x)=7678000)
257
+ 245 train 7.490226 (lr=3.4406e-04) (hash(x)=7414788)
258
+ 246 train 7.517199 (lr=3.4545e-04) (hash(x)=7326216)
259
+ 247 train 7.613448 (lr=3.4685e-04) (hash(x)=5857162)
260
+ 248 train 8.262084 (lr=3.4825e-04) (hash(x)=14715669)
261
+ 249 train 11.765888 (lr=3.4965e-04) (hash(x)=16950560)
262
+ 250 val loss 7.5279
263
+ 250 val perplexity 1859.2861
264
+ 250 train 7.679810 (lr=3.5105e-04) (hash(x)=6159125)
265
+ 251 train 7.557268 (lr=3.5245e-04) (hash(x)=5908442)
266
+ 252 train 7.407663 (lr=3.5385e-04) (hash(x)=6668887)
267
+ 253 train 7.512524 (lr=3.5524e-04) (hash(x)=7662975)
268
+ 254 train 7.481692 (lr=3.5664e-04) (hash(x)=6669539)
269
+ 255 train 7.464712 (lr=3.5804e-04) (hash(x)=5015998)
270
+ 256 train 7.391434 (lr=3.5944e-04) (hash(x)=5458386)
271
+ 257 train 7.545062 (lr=3.6084e-04) (hash(x)=4849847)
272
+ 258 train 7.556354 (lr=3.6224e-04) (hash(x)=5125562)
273
+ 259 train 7.593246 (lr=3.6364e-04) (hash(x)=6012047)
274
+ 260 train 7.486858 (lr=3.6503e-04) (hash(x)=6079364)
275
+ 261 train 7.482993 (lr=3.6643e-04) (hash(x)=6030395)
276
+ 262 train 7.479014 (lr=3.6783e-04) (hash(x)=6781762)
277
+ 263 train 7.515913 (lr=3.6923e-04) (hash(x)=5238325)
278
+ 264 train 7.566898 (lr=3.7063e-04) (hash(x)=8611144)
279
+ 265 train 7.427498 (lr=3.7203e-04) (hash(x)=6901605)
280
+ 266 train 7.451635 (lr=3.7343e-04) (hash(x)=5499337)
281
+ 267 train 7.489699 (lr=3.7483e-04) (hash(x)=7134874)
282
+ 268 train 7.476024 (lr=3.7622e-04) (hash(x)=6149038)
283
+ 269 train 7.514464 (lr=3.7762e-04) (hash(x)=6609760)
284
+ 270 train 7.591351 (lr=3.7902e-04) (hash(x)=6652627)
285
+ 271 train 7.540944 (lr=3.8042e-04) (hash(x)=8618579)
286
+ 272 train 7.434485 (lr=3.8182e-04) (hash(x)=5896892)
287
+ 273 train 7.596657 (lr=3.8322e-04) (hash(x)=8945854)
288
+ 274 train 7.370127 (lr=3.8462e-04) (hash(x)=6775103)
289
+ 275 train 7.448914 (lr=3.8601e-04) (hash(x)=6200024)
290
+ 276 train 7.341398 (lr=3.8741e-04) (hash(x)=5238139)
291
+ 277 train 7.483980 (lr=3.8881e-04) (hash(x)=5780948)
292
+ 278 train 7.465149 (lr=3.9021e-04) (hash(x)=6168407)
293
+ 279 train 7.395815 (lr=3.9161e-04) (hash(x)=3775881)
294
+ 280 train 7.407447 (lr=3.9301e-04) (hash(x)=6848709)
295
+ 281 train 7.764241 (lr=3.9441e-04) (hash(x)=6478113)
296
+ 282 train 7.447802 (lr=3.9580e-04) (hash(x)=6483914)
297
+ 283 train 7.401541 (lr=3.9720e-04) (hash(x)=5927795)
298
+ 284 train 7.426169 (lr=3.9860e-04) (hash(x)=8438033)
299
+ 285 train 7.477301 (lr=4.0000e-04) (hash(x)=5684788)
300
+ 286 train 7.272407 (lr=4.0000e-04) (hash(x)=5710182)
301
+ 287 train 7.354019 (lr=4.0000e-04) (hash(x)=6160063)
302
+ 288 train 7.254322 (lr=3.9999e-04) (hash(x)=5955100)
303
+ 289 train 7.148359 (lr=3.9998e-04) (hash(x)=6672309)
304
+ 290 train 7.549190 (lr=3.9997e-04) (hash(x)=9001564)
305
+ 291 train 7.898192 (lr=3.9996e-04) (hash(x)=5454654)
306
+ 292 train 7.271064 (lr=3.9994e-04) (hash(x)=5174062)
307
+ 293 train 7.413528 (lr=3.9991e-04) (hash(x)=6199203)
308
+ 294 train 7.298938 (lr=3.9989e-04) (hash(x)=8352747)
309
+ 295 train 7.418542 (lr=3.9986e-04) (hash(x)=4547106)
310
+ 296 train 7.401461 (lr=3.9983e-04) (hash(x)=6433099)
311
+ 297 train 7.243129 (lr=3.9979e-04) (hash(x)=6424428)
312
+ 298 train 7.337133 (lr=3.9975e-04) (hash(x)=5975712)
313
+ 299 train 7.278755 (lr=3.9971e-04) (hash(x)=5458532)
314
+ 300 val loss 7.3569
315
+ 300 val perplexity 1567.0084
316
+ 300 train 7.493862 (lr=3.9966e-04) (hash(x)=7002774)
317
+ 301 train 7.624249 (lr=3.9961e-04) (hash(x)=8206173)
318
+ 302 train 7.725688 (lr=3.9955e-04) (hash(x)=7784971)
319
+ 303 train 7.652866 (lr=3.9950e-04) (hash(x)=6001843)
320
+ 304 train 7.434690 (lr=3.9944e-04) (hash(x)=5698642)
321
+ 305 train 7.349646 (lr=3.9937e-04) (hash(x)=6600751)
322
+ 306 train 7.292259 (lr=3.9930e-04) (hash(x)=5804610)
323
+ 307 train 7.236773 (lr=3.9923e-04) (hash(x)=5162652)
324
+ 308 train 7.253363 (lr=3.9916e-04) (hash(x)=5297006)
325
+ 309 train 7.243694 (lr=3.9908e-04) (hash(x)=6728323)
326
+ 310 train 7.253349 (lr=3.9900e-04) (hash(x)=3964882)
327
+ 311 train 7.246111 (lr=3.9891e-04) (hash(x)=6735993)
328
+ 312 train 7.375684 (lr=3.9882e-04) (hash(x)=5150316)
329
+ 313 train 7.188715 (lr=3.9873e-04) (hash(x)=5220448)
330
+ 314 train 7.395230 (lr=3.9864e-04) (hash(x)=6971289)
331
+ 315 train 7.474053 (lr=3.9854e-04) (hash(x)=5012069)
332
+ 316 train 7.268793 (lr=3.9843e-04) (hash(x)=6576136)
333
+ 317 train 7.226099 (lr=3.9833e-04) (hash(x)=6077647)
334
+ 318 train 7.365600 (lr=3.9822e-04) (hash(x)=6032381)
335
+ 319 train 7.218400 (lr=3.9811e-04) (hash(x)=8410716)
336
+ 320 train 7.278893 (lr=3.9799e-04) (hash(x)=6212516)
337
+ 321 train 7.239206 (lr=3.9787e-04) (hash(x)=5298545)
338
+ 322 train 7.318979 (lr=3.9775e-04) (hash(x)=5779073)
339
+ 323 train 7.214545 (lr=3.9762e-04) (hash(x)=6340947)
340
+ 324 train 6.945568 (lr=3.9749e-04) (hash(x)=6530441)
341
+ 325 train 7.303853 (lr=3.9736e-04) (hash(x)=7818007)
342
+ 326 train 7.155062 (lr=3.9722e-04) (hash(x)=6277435)
343
+ 327 train 7.185274 (lr=3.9708e-04) (hash(x)=6226320)
344
+ 328 train 7.251521 (lr=3.9694e-04) (hash(x)=5463751)
345
+ 329 train 7.146029 (lr=3.9679e-04) (hash(x)=6579329)
346
+ 330 train 7.277887 (lr=3.9664e-04) (hash(x)=6660540)
347
+ 331 train 7.224781 (lr=3.9648e-04) (hash(x)=6491272)
348
+ 332 train 7.208120 (lr=3.9633e-04) (hash(x)=6766947)
349
+ 333 train 7.501063 (lr=3.9616e-04) (hash(x)=3721218)
350
+ 334 train 7.331292 (lr=3.9600e-04) (hash(x)=5229898)
351
+ 335 train 7.137834 (lr=3.9583e-04) (hash(x)=6054428)
352
+ 336 train 7.272422 (lr=3.9566e-04) (hash(x)=6290464)
353
+ 337 train 7.290438 (lr=3.9549e-04) (hash(x)=6904498)
354
+ 338 train 7.312539 (lr=3.9531e-04) (hash(x)=7765805)
355
+ 339 train 7.254771 (lr=3.9513e-04) (hash(x)=6543764)
356
+ 340 train 7.228377 (lr=3.9494e-04) (hash(x)=7204955)
357
+ 341 train 7.158746 (lr=3.9475e-04) (hash(x)=4423711)
358
+ 342 train 7.131050 (lr=3.9456e-04) (hash(x)=5343477)
359
+ 343 train 7.296641 (lr=3.9437e-04) (hash(x)=6595856)
360
+ 344 train 7.109536 (lr=3.9417e-04) (hash(x)=7229015)
361
+ 345 train 7.186171 (lr=3.9397e-04) (hash(x)=5916034)
362
+ 346 train 7.324819 (lr=3.9376e-04) (hash(x)=6939521)
363
+ 347 train 7.207018 (lr=3.9356e-04) (hash(x)=5492185)
364
+ 348 train 7.143770 (lr=3.9334e-04) (hash(x)=6898006)
365
+ 349 train 7.019977 (lr=3.9313e-04) (hash(x)=5462008)
366
+ 350 val loss 7.2457
367
+ 350 val perplexity 1402.0508
368
+ 350 train 7.133975 (lr=3.9291e-04) (hash(x)=7821902)
369
+ 351 train 7.207252 (lr=3.9269e-04) (hash(x)=6608108)
370
+ 352 train 7.255003 (lr=3.9246e-04) (hash(x)=6295394)
371
+ 353 train 7.127368 (lr=3.9223e-04) (hash(x)=6803449)
372
+ 354 train 7.120997 (lr=3.9200e-04) (hash(x)=7444134)
373
+ 355 train 7.162943 (lr=3.9177e-04) (hash(x)=5361923)
374
+ 356 train 7.171433 (lr=3.9153e-04) (hash(x)=7803039)
375
+ 357 train 7.294749 (lr=3.9129e-04) (hash(x)=7042578)
376
+ 358 train 7.188426 (lr=3.9104e-04) (hash(x)=6647007)
377
+ 359 train 7.204747 (lr=3.9079e-04) (hash(x)=6303658)
378
+ 360 train 7.295964 (lr=3.9054e-04) (hash(x)=6158927)
379
+ 361 train 7.111096 (lr=3.9029e-04) (hash(x)=6419656)
380
+ 362 train 7.039339 (lr=3.9003e-04) (hash(x)=7891271)
381
+ 363 train 7.265607 (lr=3.8977e-04) (hash(x)=7099888)
382
+ 364 train 7.201857 (lr=3.8950e-04) (hash(x)=6352424)
383
+ 365 train 7.133401 (lr=3.8923e-04) (hash(x)=5863145)
384
+ 366 train 7.155655 (lr=3.8896e-04) (hash(x)=6844597)
385
+ 367 train 7.214378 (lr=3.8869e-04) (hash(x)=6009861)
386
+ 368 train 7.114035 (lr=3.8841e-04) (hash(x)=5374957)
387
+ 369 train 7.155708 (lr=3.8813e-04) (hash(x)=3923225)
388
+ 370 train 7.096405 (lr=3.8785e-04) (hash(x)=5848862)
389
+ 371 train 7.135873 (lr=3.8756e-04) (hash(x)=4440634)
390
+ 372 train 7.093006 (lr=3.8727e-04) (hash(x)=4788998)
391
+ 373 train 7.063994 (lr=3.8697e-04) (hash(x)=7378165)
392
+ 374 train 7.100214 (lr=3.8667e-04) (hash(x)=4771243)
393
+ 375 train 7.102592 (lr=3.8637e-04) (hash(x)=5716089)
394
+ 376 train 7.127706 (lr=3.8607e-04) (hash(x)=7991009)
395
+ 377 train 7.155115 (lr=3.8576e-04) (hash(x)=5348607)
396
+ 378 train 7.235158 (lr=3.8545e-04) (hash(x)=6036146)
397
+ 379 train 7.106081 (lr=3.8514e-04) (hash(x)=6658718)
398
+ 380 train 7.057232 (lr=3.8482e-04) (hash(x)=6176253)
399
+ 381 train 7.057700 (lr=3.8450e-04) (hash(x)=4542895)
400
+ 382 train 6.940403 (lr=3.8418e-04) (hash(x)=5549073)
401
+ 383 train 7.301069 (lr=3.8385e-04) (hash(x)=5342783)
402
+ 384 train 7.291691 (lr=3.8352e-04) (hash(x)=6549546)
403
+ 385 train 7.043332 (lr=3.8319e-04) (hash(x)=6170599)
404
+ 386 train 7.359943 (lr=3.8286e-04) (hash(x)=8483134)
405
+ 387 train 7.464865 (lr=3.8252e-04) (hash(x)=4865036)
406
+ 388 train 7.113779 (lr=3.8217e-04) (hash(x)=4433890)
407
+ 389 train 7.622947 (lr=3.8183e-04) (hash(x)=6738597)
408
+ 390 train 7.223063 (lr=3.8148e-04) (hash(x)=5459620)
409
+ 391 train 7.225796 (lr=3.8113e-04) (hash(x)=4604074)
410
+ 392 train 7.227925 (lr=3.8077e-04) (hash(x)=8045501)
411
+ 393 train 7.316499 (lr=3.8042e-04) (hash(x)=6176464)
412
+ 394 train 7.133785 (lr=3.8006e-04) (hash(x)=6127317)
413
+ 395 train 7.167578 (lr=3.7969e-04) (hash(x)=5571462)
414
+ 396 train 7.177154 (lr=3.7933e-04) (hash(x)=6640755)
415
+ 397 train 7.221729 (lr=3.7896e-04) (hash(x)=6147295)
416
+ 398 train 7.160261 (lr=3.7858e-04) (hash(x)=6120461)
417
+ 399 train 7.130022 (lr=3.7821e-04) (hash(x)=5123928)
418
+ 400 val loss 7.1993
419
+ 400 val perplexity 1338.5188
420
+ 400 train 7.182039 (lr=3.7783e-04) (hash(x)=5428813)
421
+ 401 train 7.150963 (lr=3.7744e-04) (hash(x)=6448586)
422
+ 402 train 7.304321 (lr=3.7706e-04) (hash(x)=8116987)
423
+ 403 train 7.330850 (lr=3.7667e-04) (hash(x)=5974342)
424
+ 404 train 7.315406 (lr=3.7628e-04) (hash(x)=5261336)
425
+ 405 train 7.127333 (lr=3.7588e-04) (hash(x)=5307056)
426
+ 406 train 7.081825 (lr=3.7549e-04) (hash(x)=5865119)
427
+ 407 train 7.127733 (lr=3.7509e-04) (hash(x)=6672380)
428
+ 408 train 7.168630 (lr=3.7468e-04) (hash(x)=6886810)
429
+ 409 train 7.306437 (lr=3.7428e-04) (hash(x)=6564071)
430
+ 410 train 7.038849 (lr=3.7387e-04) (hash(x)=7705089)
431
+ 411 train 7.058094 (lr=3.7345e-04) (hash(x)=5773073)
432
+ 412 train 7.057440 (lr=3.7304e-04) (hash(x)=5905976)
433
+ 413 train 7.915966 (lr=3.7262e-04) (hash(x)=6936653)
434
+ 414 train 7.147428 (lr=3.7220e-04) (hash(x)=6592773)
435
+ 415 train 7.059624 (lr=3.7177e-04) (hash(x)=6796551)
436
+ 416 train 7.117065 (lr=3.7135e-04) (hash(x)=6151589)
437
+ 417 train 7.398699 (lr=3.7092e-04) (hash(x)=7045164)
438
+ 418 train 7.179357 (lr=3.7048e-04) (hash(x)=6650506)
439
+ 419 train 7.088448 (lr=3.7005e-04) (hash(x)=7839748)
440
+ 420 train 7.312460 (lr=3.6961e-04) (hash(x)=6028671)
441
+ 421 train 7.189707 (lr=3.6917e-04) (hash(x)=6118408)
442
+ 422 train 7.124938 (lr=3.6872e-04) (hash(x)=5757526)
443
+ 423 train 7.261390 (lr=3.6828e-04) (hash(x)=6030172)
444
+ 424 train 7.327654 (lr=3.6782e-04) (hash(x)=6719882)
445
+ 425 train 7.146479 (lr=3.6737e-04) (hash(x)=5028050)
446
+ 426 train 7.145702 (lr=3.6692e-04) (hash(x)=5682910)
447
+ 427 train 7.232688 (lr=3.6646e-04) (hash(x)=5747891)
448
+ 428 train 7.047057 (lr=3.6599e-04) (hash(x)=5734208)
449
+ 429 train 7.127618 (lr=3.6553e-04) (hash(x)=7024381)
450
+ 430 train 7.425903 (lr=3.6506e-04) (hash(x)=7814361)
451
+ 431 train 7.237172 (lr=3.6459e-04) (hash(x)=5765893)
452
+ 432 train 7.071737 (lr=3.6412e-04) (hash(x)=5572048)
453
+ 433 train 7.215243 (lr=3.6364e-04) (hash(x)=5610058)
454
+ 434 train 7.083005 (lr=3.6316e-04) (hash(x)=4459483)
455
+ 435 train 7.007694 (lr=3.6268e-04) (hash(x)=6758411)
456
+ 436 train 7.123289 (lr=3.6220e-04) (hash(x)=6701381)
457
+ 437 train 7.271715 (lr=3.6171e-04) (hash(x)=5416400)
458
+ 438 train 7.014993 (lr=3.6122e-04) (hash(x)=5919569)
459
+ 439 train 7.179642 (lr=3.6073e-04) (hash(x)=5282870)
460
+ 440 train 7.011823 (lr=3.6023e-04) (hash(x)=8072181)
461
+ 441 train 6.959876 (lr=3.5974e-04) (hash(x)=7996770)
462
+ 442 train 7.038750 (lr=3.5924e-04) (hash(x)=5093519)
463
+ 443 train 7.077492 (lr=3.5873e-04) (hash(x)=6270917)
464
+ 444 train 6.980232 (lr=3.5823e-04) (hash(x)=5852780)
465
+ 445 train 6.867648 (lr=3.5772e-04) (hash(x)=5327728)
466
+ 446 train 7.087059 (lr=3.5721e-04) (hash(x)=6260765)
467
+ 447 train 7.175209 (lr=3.5669e-04) (hash(x)=6212894)
468
+ 448 train 7.056237 (lr=3.5618e-04) (hash(x)=6724010)
469
+ 449 train 7.162891 (lr=3.5566e-04) (hash(x)=6059934)
470
+ 450 val loss 7.1652
471
+ 450 val perplexity 1293.6422
472
+ 450 train 6.898933 (lr=3.5514e-04) (hash(x)=5587382)
473
+ 451 train 7.030536 (lr=3.5461e-04) (hash(x)=7274108)
474
+ 452 train 7.122969 (lr=3.5408e-04) (hash(x)=6085396)
475
+ 453 train 7.020986 (lr=3.5355e-04) (hash(x)=4784666)
476
+ 454 train 7.098415 (lr=3.5302e-04) (hash(x)=6288638)
477
+ 455 train 7.135777 (lr=3.5249e-04) (hash(x)=7714288)
478
+ 456 train 7.192381 (lr=3.5195e-04) (hash(x)=5534748)
479
+ 457 train 7.092381 (lr=3.5141e-04) (hash(x)=6361270)
480
+ 458 train 7.085051 (lr=3.5087e-04) (hash(x)=4682956)
481
+ 459 train 7.051907 (lr=3.5032e-04) (hash(x)=7568935)
482
+ 460 train 7.155283 (lr=3.4977e-04) (hash(x)=7973311)
483
+ 461 train 6.965080 (lr=3.4922e-04) (hash(x)=6987352)
484
+ 462 train 7.021971 (lr=3.4867e-04) (hash(x)=5413904)
485
+ 463 train 7.241307 (lr=3.4812e-04) (hash(x)=6732924)
486
+ 464 train 7.136016 (lr=3.4756e-04) (hash(x)=6219754)
487
+ 465 train 7.050240 (lr=3.4700e-04) (hash(x)=6839070)
488
+ 466 train 6.989139 (lr=3.4644e-04) (hash(x)=6090120)
489
+ 467 train 7.429257 (lr=3.4587e-04) (hash(x)=7115176)
490
+ 468 train 7.170084 (lr=3.4530e-04) (hash(x)=4989299)
491
+ 469 train 7.083098 (lr=3.4473e-04) (hash(x)=7497590)
492
+ 470 train 7.180089 (lr=3.4416e-04) (hash(x)=3977557)
493
+ 471 train 7.160344 (lr=3.4359e-04) (hash(x)=5524165)
494
+ 472 train 7.108642 (lr=3.4301e-04) (hash(x)=5124936)
495
+ 473 train 7.131470 (lr=3.4243e-04) (hash(x)=5522119)
496
+ 474 train 7.025152 (lr=3.4185e-04) (hash(x)=5030865)
497
+ 475 train 7.090784 (lr=3.4127e-04) (hash(x)=6411783)
498
+ 476 train 7.016685 (lr=3.4068e-04) (hash(x)=6581937)
499
+ 477 train 7.078342 (lr=3.4009e-04) (hash(x)=5116910)
500
+ 478 train 6.874388 (lr=3.3950e-04) (hash(x)=7353459)
501
+ 479 train 6.914215 (lr=3.3891e-04) (hash(x)=6344535)
502
+ 480 train 7.283129 (lr=3.3831e-04) (hash(x)=5840169)
503
+ 481 train 7.052958 (lr=3.3771e-04) (hash(x)=6754290)
504
+ 482 train 7.283436 (lr=3.3711e-04) (hash(x)=6484573)
505
+ 483 train 6.921170 (lr=3.3651e-04) (hash(x)=5260125)
506
+ 484 train 6.967510 (lr=3.3590e-04) (hash(x)=4881179)
507
+ 485 train 7.062225 (lr=3.3530e-04) (hash(x)=5001634)
508
+ 486 train 7.006595 (lr=3.3469e-04) (hash(x)=7424810)
509
+ 487 train 6.872552 (lr=3.3408e-04) (hash(x)=5176431)
510
+ 488 train 6.963641 (lr=3.3346e-04) (hash(x)=5421624)
511
+ 489 train 7.010760 (lr=3.3285e-04) (hash(x)=5674439)
512
+ 490 train 6.934192 (lr=3.3223e-04) (hash(x)=5819157)
513
+ 491 train 7.120770 (lr=3.3161e-04) (hash(x)=6506856)
514
+ 492 train 6.936605 (lr=3.3099e-04) (hash(x)=6119124)
515
+ 493 train 7.201096 (lr=3.3036e-04) (hash(x)=6241669)
516
+ 494 train 6.775741 (lr=3.2973e-04) (hash(x)=4992756)
517
+ 495 train 7.260022 (lr=3.2911e-04) (hash(x)=7277927)
518
+ 496 train 7.193001 (lr=3.2847e-04) (hash(x)=6622097)
519
+ 497 train 7.308081 (lr=3.2784e-04) (hash(x)=7026582)
520
+ 498 train 7.191180 (lr=3.2721e-04) (hash(x)=5840987)
521
+ 499 train 6.920600 (lr=3.2657e-04) (hash(x)=4351074)
522
+ 500 val loss 7.1533
523
+ 500 val perplexity 1278.2556
524
+ 500 train 7.021180 (lr=3.2593e-04) (hash(x)=7301477)
525
+ 501 train 7.248353 (lr=3.2529e-04) (hash(x)=6264167)
526
+ 502 train 6.979429 (lr=3.2464e-04) (hash(x)=4673363)
527
+ 503 train 7.241158 (lr=3.2400e-04) (hash(x)=6332240)
528
+ 504 train 7.121806 (lr=3.2335e-04) (hash(x)=5093790)
529
+ 505 train 7.258310 (lr=3.2270e-04) (hash(x)=6890713)
530
+ 506 train 7.001636 (lr=3.2205e-04) (hash(x)=6042691)
531
+ 507 train 7.166331 (lr=3.2140e-04) (hash(x)=7080746)
532
+ 508 train 7.163791 (lr=3.2074e-04) (hash(x)=7609207)
533
+ 509 train 6.976152 (lr=3.2008e-04) (hash(x)=4591273)
534
+ 510 train 7.130836 (lr=3.1943e-04) (hash(x)=8183111)
535
+ 511 train 7.391247 (lr=3.1876e-04) (hash(x)=5893095)
536
+ 512 train 7.159004 (lr=3.1810e-04) (hash(x)=5709077)
537
+ 513 train 7.121282 (lr=3.1744e-04) (hash(x)=6218715)
538
+ 514 train 7.011164 (lr=3.1677e-04) (hash(x)=4722188)
539
+ 515 train 6.906708 (lr=3.1610e-04) (hash(x)=6313422)
540
+ 516 train 7.037715 (lr=3.1543e-04) (hash(x)=6350660)
541
+ 517 train 7.120565 (lr=3.1476e-04) (hash(x)=5371913)
542
+ 518 train 7.235427 (lr=3.1408e-04) (hash(x)=6283595)
543
+ 519 train 7.234426 (lr=3.1341e-04) (hash(x)=6179167)
544
+ 520 train 7.101098 (lr=3.1273e-04) (hash(x)=7865194)
545
+ 521 train 7.269358 (lr=3.1205e-04) (hash(x)=4883760)
546
+ 522 train 7.089323 (lr=3.1137e-04) (hash(x)=5769286)
547
+ 523 train 7.073445 (lr=3.1069e-04) (hash(x)=6309912)
548
+ 524 train 7.031682 (lr=3.1000e-04) (hash(x)=7490849)
549
+ 525 train 7.019407 (lr=3.0931e-04) (hash(x)=5631134)
550
+ 526 train 7.027297 (lr=3.0862e-04) (hash(x)=4754560)
551
+ 527 train 6.953103 (lr=3.0793e-04) (hash(x)=5448702)
552
+ 528 train 7.179194 (lr=3.0724e-04) (hash(x)=7168842)
553
+ 529 train 7.165020 (lr=3.0655e-04) (hash(x)=7469195)
554
+ 530 train 6.981960 (lr=3.0585e-04) (hash(x)=5479500)
555
+ 531 train 7.156197 (lr=3.0516e-04) (hash(x)=5634177)
556
+ 532 train 7.094009 (lr=3.0446e-04) (hash(x)=9779982)
557
+ 533 train 6.951110 (lr=3.0376e-04) (hash(x)=3641890)
558
+ 534 train 7.069530 (lr=3.0306e-04) (hash(x)=7994103)
559
+ 535 train 7.064605 (lr=3.0235e-04) (hash(x)=5358547)
560
+ 536 train 7.104650 (lr=3.0165e-04) (hash(x)=6797424)
561
+ 537 train 7.132009 (lr=3.0094e-04) (hash(x)=4876580)
562
+ 538 train 6.919934 (lr=3.0023e-04) (hash(x)=4867629)
563
+ 539 train 7.045597 (lr=2.9952e-04) (hash(x)=6601967)
564
+ 540 train 6.992087 (lr=2.9881e-04) (hash(x)=5480657)
565
+ 541 train 7.028528 (lr=2.9810e-04) (hash(x)=7127346)
566
+ 542 train 7.149080 (lr=2.9738e-04) (hash(x)=7224794)
567
+ 543 train 7.036208 (lr=2.9667e-04) (hash(x)=5967252)
568
+ 544 train 7.130063 (lr=2.9595e-04) (hash(x)=7415262)
569
+ 545 train 6.998951 (lr=2.9523e-04) (hash(x)=6522544)
570
+ 546 train 7.044478 (lr=2.9451e-04) (hash(x)=5181640)
571
+ 547 train 6.993721 (lr=2.9379e-04) (hash(x)=5916880)
572
+ 548 train 6.963616 (lr=2.9307e-04) (hash(x)=5620397)
573
+ 549 train 7.052005 (lr=2.9234e-04) (hash(x)=5994970)
574
+ 550 val loss 7.1261
575
+ 550 val perplexity 1243.9720
576
+ 550 train 7.061192 (lr=2.9162e-04) (hash(x)=6532717)
577
+ 551 train 7.120857 (lr=2.9089e-04) (hash(x)=6262762)
578
+ 552 train 6.903340 (lr=2.9016e-04) (hash(x)=5495326)
579
+ 553 train 7.248234 (lr=2.8943e-04) (hash(x)=5927868)
580
+ 554 train 7.026209 (lr=2.8870e-04) (hash(x)=5485433)
581
+ 555 train 6.976710 (lr=2.8797e-04) (hash(x)=6462425)
582
+ 556 train 7.021902 (lr=2.8723e-04) (hash(x)=6299518)
583
+ 557 train 7.114375 (lr=2.8650e-04) (hash(x)=5950376)
584
+ 558 train 7.238941 (lr=2.8576e-04) (hash(x)=6296440)
585
+ 559 train 7.324310 (lr=2.8502e-04) (hash(x)=6974102)
586
+ 560 train 7.124984 (lr=2.8428e-04) (hash(x)=6228679)
587
+ 561 train 7.021175 (lr=2.8354e-04) (hash(x)=6120319)
588
+ 562 train 7.291116 (lr=2.8280e-04) (hash(x)=6289494)
589
+ 563 train 7.085501 (lr=2.8206e-04) (hash(x)=6877818)
590
+ 564 train 7.065352 (lr=2.8132e-04) (hash(x)=6490976)
591
+ 565 train 7.207963 (lr=2.8057e-04) (hash(x)=6419450)
592
+ 566 train 7.147706 (lr=2.7982e-04) (hash(x)=6840857)
593
+ 567 train 7.083073 (lr=2.7908e-04) (hash(x)=4837686)
594
+ 568 train 7.150756 (lr=2.7833e-04) (hash(x)=5561493)
595
+ 569 train 7.050315 (lr=2.7758e-04) (hash(x)=5831242)
596
+ 570 train 7.155830 (lr=2.7683e-04) (hash(x)=6851511)
597
+ 571 train 7.221403 (lr=2.7607e-04) (hash(x)=6169616)
598
+ 572 train 7.135947 (lr=2.7532e-04) (hash(x)=5357151)
599
+ 573 train 6.977868 (lr=2.7457e-04) (hash(x)=5471028)
600
+ 574 train 7.060381 (lr=2.7381e-04) (hash(x)=5051894)
601
+ 575 train 7.067574 (lr=2.7306e-04) (hash(x)=4930450)
602
+ 576 train 7.170663 (lr=2.7230e-04) (hash(x)=6537527)
603
+ 577 train 7.140301 (lr=2.7154e-04) (hash(x)=6330572)
604
+ 578 train 7.277368 (lr=2.7078e-04) (hash(x)=5315155)
605
+ 579 train 7.070692 (lr=2.7002e-04) (hash(x)=5416079)
606
+ 580 train 7.035961 (lr=2.6926e-04) (hash(x)=6255865)
607
+ 581 train 7.237593 (lr=2.6850e-04) (hash(x)=6723036)
608
+ 582 train 7.015143 (lr=2.6773e-04) (hash(x)=7107190)
609
+ 583 train 7.130934 (lr=2.6697e-04) (hash(x)=5868958)
610
+ 584 train 7.233924 (lr=2.6620e-04) (hash(x)=7385923)
611
+ 585 train 7.054811 (lr=2.6544e-04) (hash(x)=5373053)
612
+ 586 train 7.058751 (lr=2.6467e-04) (hash(x)=5371748)
613
+ 587 train 7.229062 (lr=2.6390e-04) (hash(x)=5783717)
614
+ 588 train 6.978134 (lr=2.6314e-04) (hash(x)=6388848)
615
+ 589 train 6.981009 (lr=2.6237e-04) (hash(x)=4865068)
616
+ 590 train 6.941400 (lr=2.6160e-04) (hash(x)=5844961)
617
+ 591 train 7.034984 (lr=2.6083e-04) (hash(x)=5989273)
618
+ 592 train 6.953630 (lr=2.6005e-04) (hash(x)=5136202)
619
+ 593 train 7.100945 (lr=2.5928e-04) (hash(x)=5811889)
620
+ 594 train 7.055227 (lr=2.5851e-04) (hash(x)=6554415)
621
+ 595 train 7.092797 (lr=2.5773e-04) (hash(x)=5969272)
622
+ 596 train 7.083670 (lr=2.5696e-04) (hash(x)=7121518)
623
+ 597 train 7.542042 (lr=2.5618e-04) (hash(x)=6363768)
624
+ 598 train 7.139645 (lr=2.5541e-04) (hash(x)=7030284)
625
+ 599 train 7.239227 (lr=2.5463e-04) (hash(x)=7243339)
626
+ 600 val loss 7.1678
627
+ 600 val perplexity 1296.9423
628
+ 600 train 7.357470 (lr=2.5385e-04) (hash(x)=7199493)
629
+ 601 train 7.357932 (lr=2.5307e-04) (hash(x)=5819281)
630
+ 602 train 7.098391 (lr=2.5230e-04) (hash(x)=7285909)
631
+ 603 train 7.131826 (lr=2.5152e-04) (hash(x)=6730878)
632
+ 604 train 7.160683 (lr=2.5074e-04) (hash(x)=5200141)
633
+ 605 train 7.244627 (lr=2.4996e-04) (hash(x)=5208573)
634
+ 606 train 7.346661 (lr=2.4917e-04) (hash(x)=5976779)
635
+ 607 train 7.172334 (lr=2.4839e-04) (hash(x)=5921767)
636
+ 608 train 7.212012 (lr=2.4761e-04) (hash(x)=6596032)
637
+ 609 train 7.230341 (lr=2.4683e-04) (hash(x)=5428058)
638
+ 610 train 7.170779 (lr=2.4604e-04) (hash(x)=5748050)
639
+ 611 train 7.177820 (lr=2.4526e-04) (hash(x)=6397421)
640
+ 612 train 7.251243 (lr=2.4448e-04) (hash(x)=5638644)
641
+ 613 train 7.140042 (lr=2.4369e-04) (hash(x)=7317623)
642
+ 614 train 7.213242 (lr=2.4291e-04) (hash(x)=5254932)
643
+ 615 train 7.264142 (lr=2.4212e-04) (hash(x)=4945141)
644
+ 616 train 7.229517 (lr=2.4133e-04) (hash(x)=7093994)
645
+ 617 train 7.148717 (lr=2.4055e-04) (hash(x)=7511894)
646
+ 618 train 7.130069 (lr=2.3976e-04) (hash(x)=6706832)
647
+ 619 train 7.115844 (lr=2.3897e-04) (hash(x)=6175371)
648
+ 620 train 7.135775 (lr=2.3818e-04) (hash(x)=7089897)
649
+ 621 train 7.162364 (lr=2.3740e-04) (hash(x)=6828202)
650
+ 622 train 7.500745 (lr=2.3661e-04) (hash(x)=7095301)
651
+ 623 train 7.157488 (lr=2.3582e-04) (hash(x)=5652191)
652
+ 624 train 7.197343 (lr=2.3503e-04) (hash(x)=7336431)
653
+ 625 train 7.212706 (lr=2.3424e-04) (hash(x)=6426339)
654
+ 626 train 7.302027 (lr=2.3345e-04) (hash(x)=5874982)
655
+ 627 train 7.220245 (lr=2.3266e-04) (hash(x)=6978359)
656
+ 628 train 7.015278 (lr=2.3187e-04) (hash(x)=4438173)
657
+ 629 train 7.075735 (lr=2.3108e-04) (hash(x)=6475356)
658
+ 630 train 7.292076 (lr=2.3029e-04) (hash(x)=5030069)
659
+ 631 train 7.047969 (lr=2.2950e-04) (hash(x)=6507939)
660
+ 632 train 7.064350 (lr=2.2871e-04) (hash(x)=6373406)
661
+ 633 train 7.197098 (lr=2.2792e-04) (hash(x)=6623130)
662
+ 634 train 7.096190 (lr=2.2713e-04) (hash(x)=6276741)
663
+ 635 train 7.023569 (lr=2.2633e-04) (hash(x)=7169874)
664
+ 636 train 7.101986 (lr=2.2554e-04) (hash(x)=4947348)
665
+ 637 train 7.199184 (lr=2.2475e-04) (hash(x)=7407715)
666
+ 638 train 7.328821 (lr=2.2396e-04) (hash(x)=5453341)
667
+ 639 train 7.076392 (lr=2.2317e-04) (hash(x)=6520161)
668
+ 640 train 7.155873 (lr=2.2238e-04) (hash(x)=6011882)
669
+ 641 train 7.107605 (lr=2.2158e-04) (hash(x)=5573372)
670
+ 642 train 7.302817 (lr=2.2079e-04) (hash(x)=5344827)
671
+ 643 train 7.341261 (lr=2.2000e-04) (hash(x)=6127887)
672
+ 644 train 7.081565 (lr=2.1921e-04) (hash(x)=6052304)
673
+ 645 train 7.067441 (lr=2.1842e-04) (hash(x)=7039733)
674
+ 646 train 7.180966 (lr=2.1762e-04) (hash(x)=7746451)
675
+ 647 train 7.381979 (lr=2.1683e-04) (hash(x)=7694373)
676
+ 648 train 7.220730 (lr=2.1604e-04) (hash(x)=6911004)
677
+ 649 train 7.145518 (lr=2.1525e-04) (hash(x)=8055224)
678
+ 650 val loss 7.1229
679
+ 650 val perplexity 1239.9946
680
+ 650 train 7.200710 (lr=2.1446e-04) (hash(x)=7060476)
681
+ 651 train 7.179423 (lr=2.1367e-04) (hash(x)=6002317)
682
+ 652 train 7.201079 (lr=2.1287e-04) (hash(x)=6524390)
683
+ 653 train 7.191113 (lr=2.1208e-04) (hash(x)=8331083)
684
+ 654 train 7.215149 (lr=2.1129e-04) (hash(x)=7084852)
685
+ 655 train 7.147175 (lr=2.1050e-04) (hash(x)=6190174)
686
+ 656 train 7.110544 (lr=2.0971e-04) (hash(x)=5601780)
687
+ 657 train 7.263338 (lr=2.0892e-04) (hash(x)=4764034)
688
+ 658 train 7.139706 (lr=2.0813e-04) (hash(x)=6361606)
689
+ 659 train 7.123537 (lr=2.0734e-04) (hash(x)=5697877)
690
+ 660 train 7.170410 (lr=2.0655e-04) (hash(x)=6249793)
691
+ 661 train 7.238626 (lr=2.0576e-04) (hash(x)=7424709)
692
+ 662 train 7.244258 (lr=2.0497e-04) (hash(x)=8203076)
693
+ 663 train 7.011907 (lr=2.0418e-04) (hash(x)=5290019)
694
+ 664 train 7.049381 (lr=2.0339e-04) (hash(x)=6278399)
695
+ 665 train 7.107878 (lr=2.0260e-04) (hash(x)=5709192)
696
+ 666 train 7.214690 (lr=2.0182e-04) (hash(x)=5660292)
697
+ 667 train 7.169106 (lr=2.0103e-04) (hash(x)=5563697)
698
+ 668 train 7.167667 (lr=2.0024e-04) (hash(x)=6325245)
699
+ 669 train 7.148139 (lr=1.9945e-04) (hash(x)=6259741)
700
+ 670 train 7.296114 (lr=1.9867e-04) (hash(x)=6238355)
701
+ 671 train 7.224691 (lr=1.9788e-04) (hash(x)=7778802)
702
+ 672 train 7.144229 (lr=1.9709e-04) (hash(x)=6283496)
703
+ 673 train 7.269869 (lr=1.9631e-04) (hash(x)=5325614)
704
+ 674 train 7.097204 (lr=1.9552e-04) (hash(x)=5376057)
705
+ 675 train 7.097230 (lr=1.9474e-04) (hash(x)=5203215)
706
+ 676 train 7.154555 (lr=1.9396e-04) (hash(x)=4893030)
707
+ 677 train 7.118423 (lr=1.9317e-04) (hash(x)=7595573)
708
+ 678 train 7.199433 (lr=1.9239e-04) (hash(x)=7188575)
709
+ 679 train 7.226526 (lr=1.9161e-04) (hash(x)=6858739)
710
+ 680 train 7.091007 (lr=1.9083e-04) (hash(x)=7157032)
711
+ 681 train 7.208856 (lr=1.9004e-04) (hash(x)=5871492)
712
+ 682 train 7.066514 (lr=1.8926e-04) (hash(x)=5636927)
713
+ 683 train 7.072114 (lr=1.8848e-04) (hash(x)=6135391)
714
+ 684 train 7.145304 (lr=1.8770e-04) (hash(x)=3351432)
715
+ 685 train 7.137728 (lr=1.8693e-04) (hash(x)=6781160)
716
+ 686 train 7.154644 (lr=1.8615e-04) (hash(x)=5824767)
717
+ 687 train 7.187216 (lr=1.8537e-04) (hash(x)=4512723)
718
+ 688 train 7.098910 (lr=1.8459e-04) (hash(x)=4600884)
719
+ 689 train 6.983595 (lr=1.8382e-04) (hash(x)=5079210)
720
+ 690 train 7.065967 (lr=1.8304e-04) (hash(x)=4671963)
721
+ 691 train 6.946387 (lr=1.8227e-04) (hash(x)=4169508)
722
+ 692 train 7.207753 (lr=1.8149e-04) (hash(x)=5341676)
723
+ 693 train 7.001956 (lr=1.8072e-04) (hash(x)=7593391)
724
+ 694 train 7.128521 (lr=1.7995e-04) (hash(x)=6723106)
725
+ 695 train 7.222433 (lr=1.7917e-04) (hash(x)=5136994)
726
+ 696 train 7.152708 (lr=1.7840e-04) (hash(x)=6571636)
727
+ 697 train 7.191177 (lr=1.7763e-04) (hash(x)=6975234)
728
+ 698 train 7.005063 (lr=1.7686e-04) (hash(x)=4868479)
729
+ 699 train 7.300931 (lr=1.7610e-04) (hash(x)=7086593)
730
+ 700 val loss 7.1059
731
+ 700 val perplexity 1219.0916
732
+ 700 train 7.153590 (lr=1.7533e-04) (hash(x)=6870808)
733
+ 701 train 7.454185 (lr=1.7456e-04) (hash(x)=7186398)
734
+ 702 train 7.242810 (lr=1.7380e-04) (hash(x)=4907570)
735
+ 703 train 7.075156 (lr=1.7303e-04) (hash(x)=5812649)
736
+ 704 train 7.140736 (lr=1.7227e-04) (hash(x)=6577447)
737
+ 705 train 7.081462 (lr=1.7150e-04) (hash(x)=5923453)
738
+ 706 train 7.354680 (lr=1.7074e-04) (hash(x)=7282962)
739
+ 707 train 7.029342 (lr=1.6998e-04) (hash(x)=4767725)
740
+ 708 train 7.066556 (lr=1.6922e-04) (hash(x)=6894202)
741
+ 709 train 7.511934 (lr=1.6846e-04) (hash(x)=4904377)
742
+ 710 train 7.085299 (lr=1.6770e-04) (hash(x)=5810443)
743
+ 711 train 7.175058 (lr=1.6694e-04) (hash(x)=5193944)
744
+ 712 train 7.106949 (lr=1.6619e-04) (hash(x)=5188803)
745
+ 713 train 7.075499 (lr=1.6543e-04) (hash(x)=7235549)
746
+ 714 train 7.286214 (lr=1.6468e-04) (hash(x)=5692576)
747
+ 715 train 7.195729 (lr=1.6393e-04) (hash(x)=5103770)
748
+ 716 train 7.086327 (lr=1.6317e-04) (hash(x)=6084786)
749
+ 717 train 7.049202 (lr=1.6242e-04) (hash(x)=5949126)
750
+ 718 train 7.138165 (lr=1.6167e-04) (hash(x)=6321938)
751
+ 719 train 7.161206 (lr=1.6092e-04) (hash(x)=6227413)
752
+ 720 train 7.217298 (lr=1.6018e-04) (hash(x)=6795372)
753
+ 721 train 7.096369 (lr=1.5943e-04) (hash(x)=6506400)
754
+ 722 train 7.093799 (lr=1.5868e-04) (hash(x)=4595742)
755
+ 723 train 7.229487 (lr=1.5794e-04) (hash(x)=7342185)
756
+ 724 train 7.261048 (lr=1.5720e-04) (hash(x)=5762071)
757
+ 725 train 7.105037 (lr=1.5646e-04) (hash(x)=6337670)
758
+ 726 train 7.123835 (lr=1.5572e-04) (hash(x)=6831170)
759
+ 727 train 7.187840 (lr=1.5498e-04) (hash(x)=7080870)
760
+ 728 train 7.085952 (lr=1.5424e-04) (hash(x)=6449970)
761
+ 729 train 7.263817 (lr=1.5350e-04) (hash(x)=5588514)
762
+ 730 train 7.033251 (lr=1.5277e-04) (hash(x)=5379746)
763
+ 731 train 7.218739 (lr=1.5203e-04) (hash(x)=8049901)
764
+ 732 train 7.281722 (lr=1.5130e-04) (hash(x)=8398209)
765
+ 733 train 7.181120 (lr=1.5057e-04) (hash(x)=6152707)
766
+ 734 train 7.244746 (lr=1.4984e-04) (hash(x)=7682617)
767
+ 735 train 7.120516 (lr=1.4911e-04) (hash(x)=7053352)
768
+ 736 train 7.178997 (lr=1.4838e-04) (hash(x)=8491505)
769
+ 737 train 7.158307 (lr=1.4766e-04) (hash(x)=6102853)
770
+ 738 train 7.230662 (lr=1.4693e-04) (hash(x)=4166591)
771
+ 739 train 7.125430 (lr=1.4621e-04) (hash(x)=7501152)
772
+ 740 train 7.065097 (lr=1.4549e-04) (hash(x)=8231949)
773
+ 741 train 7.128492 (lr=1.4477e-04) (hash(x)=5741690)
774
+ 742 train 7.256294 (lr=1.4405e-04) (hash(x)=5420426)
775
+ 743 train 7.065107 (lr=1.4333e-04) (hash(x)=5808766)
776
+ 744 train 7.047638 (lr=1.4262e-04) (hash(x)=7533322)
777
+ 745 train 7.076741 (lr=1.4190e-04) (hash(x)=6350197)
778
+ 746 train 7.222216 (lr=1.4119e-04) (hash(x)=6207170)
779
+ 747 train 7.218302 (lr=1.4048e-04) (hash(x)=5811552)
780
+ 748 train 7.095105 (lr=1.3977e-04) (hash(x)=4540607)
781
+ 749 train 6.961958 (lr=1.3906e-04) (hash(x)=5599396)
782
+ 750 val loss 7.0833
783
+ 750 val perplexity 1191.8630
784
+ 750 train 6.971573 (lr=1.3835e-04) (hash(x)=6180984)
785
+ 751 train 7.105026 (lr=1.3765e-04) (hash(x)=6519152)
786
+ 752 train 7.119277 (lr=1.3694e-04) (hash(x)=9018585)
787
+ 753 train 7.158452 (lr=1.3624e-04) (hash(x)=5988202)
788
+ 754 train 7.206363 (lr=1.3554e-04) (hash(x)=7243449)
789
+ 755 train 6.986727 (lr=1.3484e-04) (hash(x)=4377360)
790
+ 756 train 7.058391 (lr=1.3415e-04) (hash(x)=5928295)
791
+ 757 train 7.019558 (lr=1.3345e-04) (hash(x)=4996189)
792
+ 758 train 7.044543 (lr=1.3276e-04) (hash(x)=6829435)
793
+ 759 train 7.137823 (lr=1.3207e-04) (hash(x)=6929283)
794
+ 760 train 7.233899 (lr=1.3138e-04) (hash(x)=7733662)
795
+ 761 train 7.151871 (lr=1.3069e-04) (hash(x)=5975918)
796
+ 762 train 7.048453 (lr=1.3000e-04) (hash(x)=5566111)
797
+ 763 train 7.131865 (lr=1.2931e-04) (hash(x)=7407511)
798
+ 764 train 7.058040 (lr=1.2863e-04) (hash(x)=5896879)
799
+ 765 train 7.001778 (lr=1.2795e-04) (hash(x)=5569196)
800
+ 766 train 7.205760 (lr=1.2727e-04) (hash(x)=6235218)
801
+ 767 train 7.144052 (lr=1.2659e-04) (hash(x)=7467630)
802
+ 768 train 7.130623 (lr=1.2592e-04) (hash(x)=5268563)
803
+ 769 train 7.311975 (lr=1.2524e-04) (hash(x)=5915215)
804
+ 770 train 7.134995 (lr=1.2457e-04) (hash(x)=5439067)
805
+ 771 train 7.103903 (lr=1.2390e-04) (hash(x)=6167203)
806
+ 772 train 7.141500 (lr=1.2323e-04) (hash(x)=4940526)
807
+ 773 train 7.199246 (lr=1.2256e-04) (hash(x)=5897107)
808
+ 774 train 7.038918 (lr=1.2190e-04) (hash(x)=6780557)
809
+ 775 train 7.132605 (lr=1.2124e-04) (hash(x)=7566733)
810
+ 776 train 7.256866 (lr=1.2057e-04) (hash(x)=7939982)
811
+ 777 train 7.027801 (lr=1.1992e-04) (hash(x)=6447254)
812
+ 778 train 7.030881 (lr=1.1926e-04) (hash(x)=5525831)
813
+ 779 train 6.948378 (lr=1.1860e-04) (hash(x)=5059170)
814
+ 780 train 7.042084 (lr=1.1795e-04) (hash(x)=7485229)
815
+ 781 train 7.203197 (lr=1.1730e-04) (hash(x)=5955219)
816
+ 782 train 7.039525 (lr=1.1665e-04) (hash(x)=6625876)
817
+ 783 train 7.351377 (lr=1.1600e-04) (hash(x)=7141085)
818
+ 784 train 7.309487 (lr=1.1536e-04) (hash(x)=5777157)
819
+ 785 train 6.994884 (lr=1.1471e-04) (hash(x)=6236468)
820
+ 786 train 6.828361 (lr=1.1407e-04) (hash(x)=5381032)
821
+ 787 train 7.003125 (lr=1.1343e-04) (hash(x)=6733716)
822
+ 788 train 7.290294 (lr=1.1279e-04) (hash(x)=7073122)
823
+ 789 train 7.223703 (lr=1.1216e-04) (hash(x)=7622163)
824
+ 790 train 7.306184 (lr=1.1153e-04) (hash(x)=6866834)
825
+ 791 train 7.222014 (lr=1.1089e-04) (hash(x)=7008739)
826
+ 792 train 7.145790 (lr=1.1027e-04) (hash(x)=7839835)
827
+ 793 train 6.988474 (lr=1.0964e-04) (hash(x)=5313637)
828
+ 794 train 7.211545 (lr=1.0901e-04) (hash(x)=6410359)
829
+ 795 train 7.031748 (lr=1.0839e-04) (hash(x)=5721008)
830
+ 796 train 7.057252 (lr=1.0777e-04) (hash(x)=6137936)
831
+ 797 train 6.904137 (lr=1.0715e-04) (hash(x)=6881406)
832
+ 798 train 7.134869 (lr=1.0654e-04) (hash(x)=6088931)
833
+ 799 train 7.156132 (lr=1.0592e-04) (hash(x)=5689728)
834
+ 800 val loss 7.0664
835
+ 800 val perplexity 1171.9719
836
+ 800 train 7.192979 (lr=1.0531e-04) (hash(x)=5800778)
837
+ 801 train 7.088962 (lr=1.0470e-04) (hash(x)=4419021)
838
+ 802 train 7.048554 (lr=1.0410e-04) (hash(x)=5544294)
839
+ 803 train 7.116417 (lr=1.0349e-04) (hash(x)=5663946)
840
+ 804 train 7.051162 (lr=1.0289e-04) (hash(x)=4573911)
841
+ 805 train 6.970887 (lr=1.0229e-04) (hash(x)=5879465)
842
+ 806 train 7.072859 (lr=1.0169e-04) (hash(x)=5641517)
843
+ 807 train 7.238340 (lr=1.0109e-04) (hash(x)=6390336)
844
+ 808 train 7.119595 (lr=1.0050e-04) (hash(x)=8412819)
845
+ 809 train 7.238630 (lr=9.9910e-05) (hash(x)=7005536)
846
+ 810 train 6.965857 (lr=9.9321e-05) (hash(x)=6630048)
847
+ 811 train 7.109157 (lr=9.8735e-05) (hash(x)=7267424)
848
+ 812 train 6.991896 (lr=9.8151e-05) (hash(x)=6497645)
849
+ 813 train 6.953327 (lr=9.7569e-05) (hash(x)=5498697)
850
+ 814 train 7.063187 (lr=9.6990e-05) (hash(x)=6323524)
851
+ 815 train 7.003180 (lr=9.6413e-05) (hash(x)=6529460)
852
+ 816 train 6.893846 (lr=9.5838e-05) (hash(x)=5498587)
853
+ 817 train 7.034052 (lr=9.5266e-05) (hash(x)=6828893)
854
+ 818 train 7.171636 (lr=9.4696e-05) (hash(x)=6591055)
855
+ 819 train 7.046059 (lr=9.4129e-05) (hash(x)=6739572)
856
+ 820 train 7.051702 (lr=9.3564e-05) (hash(x)=5614231)
857
+ 821 train 7.094314 (lr=9.3001e-05) (hash(x)=5462474)
858
+ 822 train 7.084296 (lr=9.2441e-05) (hash(x)=5906742)
859
+ 823 train 7.159026 (lr=9.1884e-05) (hash(x)=8249268)
860
+ 824 train 7.155792 (lr=9.1328e-05) (hash(x)=4886590)
861
+ 825 train 6.909027 (lr=9.0776e-05) (hash(x)=6736150)
862
+ 826 train 7.030104 (lr=9.0226e-05) (hash(x)=6601014)
863
+ 827 train 7.085587 (lr=8.9678e-05) (hash(x)=8283020)
864
+ 828 train 7.062344 (lr=8.9133e-05) (hash(x)=6526568)
865
+ 829 train 7.129486 (lr=8.8591e-05) (hash(x)=6609259)
866
+ 830 train 7.021660 (lr=8.8051e-05) (hash(x)=6079648)
867
+ 831 train 6.987033 (lr=8.7513e-05) (hash(x)=6607178)
868
+ 832 train 7.077545 (lr=8.6978e-05) (hash(x)=6815798)
869
+ 833 train 7.328344 (lr=8.6446e-05) (hash(x)=7277017)
870
+ 834 train 6.986910 (lr=8.5916e-05) (hash(x)=5326415)
871
+ 835 train 7.089821 (lr=8.5389e-05) (hash(x)=4215944)
872
+ 836 train 7.166488 (lr=8.4865e-05) (hash(x)=4689010)
873
+ 837 train 7.092046 (lr=8.4343e-05) (hash(x)=5690469)
874
+ 838 train 7.019005 (lr=8.3824e-05) (hash(x)=6699212)
875
+ 839 train 6.950070 (lr=8.3307e-05) (hash(x)=5494876)
876
+ 840 train 7.433526 (lr=8.2793e-05) (hash(x)=10756054)
877
+ 841 train 7.055246 (lr=8.2282e-05) (hash(x)=6553219)
878
+ 842 train 6.885718 (lr=8.1773e-05) (hash(x)=7323893)
879
+ 843 train 7.331628 (lr=8.1267e-05) (hash(x)=5481970)
880
+ 844 train 7.136324 (lr=8.0764e-05) (hash(x)=6102855)
881
+ 845 train 7.104608 (lr=8.0263e-05) (hash(x)=6889406)
882
+ 846 train 7.149639 (lr=7.9765e-05) (hash(x)=4468013)
883
+ 847 train 6.968787 (lr=7.9270e-05) (hash(x)=6621321)
884
+ 848 train 7.111864 (lr=7.8778e-05) (hash(x)=6185717)
885
+ 849 train 6.978690 (lr=7.8288e-05) (hash(x)=5594828)
886
+ 850 val loss 7.0597
887
+ 850 val perplexity 1164.0487
888
+ 850 train 7.094741 (lr=7.7801e-05) (hash(x)=6530853)
889
+ 851 train 7.032275 (lr=7.7317e-05) (hash(x)=5688551)
890
+ 852 train 7.158369 (lr=7.6836e-05) (hash(x)=6979027)
891
+ 853 train 7.014186 (lr=7.6357e-05) (hash(x)=6065040)
892
+ 854 train 6.967831 (lr=7.5881e-05) (hash(x)=6836310)
893
+ 855 train 7.072920 (lr=7.5408e-05) (hash(x)=4871742)
894
+ 856 train 6.979170 (lr=7.4938e-05) (hash(x)=7461300)
895
+ 857 train 7.130583 (lr=7.4470e-05) (hash(x)=5073610)
896
+ 858 train 7.040239 (lr=7.4005e-05) (hash(x)=6725326)
897
+ 859 train 7.166819 (lr=7.3544e-05) (hash(x)=6114900)
898
+ 860 train 7.107960 (lr=7.3085e-05) (hash(x)=5249155)
899
+ 861 train 7.157339 (lr=7.2628e-05) (hash(x)=8179785)
900
+ 862 train 7.051825 (lr=7.2175e-05) (hash(x)=5588627)
901
+ 863 train 7.106995 (lr=7.1725e-05) (hash(x)=7084451)
902
+ 864 train 7.306528 (lr=7.1277e-05) (hash(x)=7478960)
903
+ 865 train 7.114061 (lr=7.0832e-05) (hash(x)=5609947)
904
+ 866 train 7.106631 (lr=7.0390e-05) (hash(x)=5741827)
905
+ 867 train 7.041296 (lr=6.9952e-05) (hash(x)=3359718)
906
+ 868 train 6.925842 (lr=6.9516e-05) (hash(x)=5480813)
907
+ 869 train 7.057669 (lr=6.9082e-05) (hash(x)=4774534)
908
+ 870 train 7.147145 (lr=6.8652e-05) (hash(x)=5895917)
909
+ 871 train 7.003381 (lr=6.8225e-05) (hash(x)=4902539)
910
+ 872 train 7.072736 (lr=6.7801e-05) (hash(x)=5734618)
911
+ 873 train 7.019657 (lr=6.7379e-05) (hash(x)=7086920)
912
+ 874 train 7.258543 (lr=6.6961e-05) (hash(x)=5338668)
913
+ 875 train 7.403380 (lr=6.6545e-05) (hash(x)=5238889)
914
+ 876 train 6.959765 (lr=6.6133e-05) (hash(x)=6244468)
915
+ 877 train 7.013886 (lr=6.5723e-05) (hash(x)=6715860)
916
+ 878 train 7.130364 (lr=6.5317e-05) (hash(x)=6216047)
917
+ 879 train 7.138363 (lr=6.4913e-05) (hash(x)=6565285)
918
+ 880 train 7.057242 (lr=6.4513e-05) (hash(x)=5257335)
919
+ 881 train 7.073652 (lr=6.4115e-05) (hash(x)=6474071)
920
+ 882 train 7.097051 (lr=6.3721e-05) (hash(x)=5705620)
921
+ 883 train 6.981376 (lr=6.3329e-05) (hash(x)=7193697)
922
+ 884 train 7.075822 (lr=6.2941e-05) (hash(x)=6264039)
923
+ 885 train 7.135762 (lr=6.2556e-05) (hash(x)=6303643)
924
+ 886 train 7.032229 (lr=6.2173e-05) (hash(x)=6120402)
925
+ 887 train 7.025526 (lr=6.1794e-05) (hash(x)=4158764)
926
+ 888 train 6.775064 (lr=6.1418e-05) (hash(x)=6173916)
927
+ 889 train 7.063085 (lr=6.1045e-05) (hash(x)=6211683)
928
+ 890 train 7.103407 (lr=6.0675e-05) (hash(x)=4950548)
929
+ 891 train 6.936401 (lr=6.0308e-05) (hash(x)=4885446)
930
+ 892 train 7.074884 (lr=5.9944e-05) (hash(x)=8117411)
931
+ 893 train 7.089654 (lr=5.9583e-05) (hash(x)=7446434)
932
+ 894 train 6.986457 (lr=5.9225e-05) (hash(x)=5371870)
933
+ 895 train 6.922860 (lr=5.8871e-05) (hash(x)=5022800)
934
+ 896 train 6.958308 (lr=5.8519e-05) (hash(x)=5666758)
935
+ 897 train 7.004141 (lr=5.8171e-05) (hash(x)=5872751)
936
+ 898 train 7.366642 (lr=5.7826e-05) (hash(x)=8428026)
937
+ 899 train 7.158260 (lr=5.7484e-05) (hash(x)=5222699)
938
+ 900 val loss 7.0570
939
+ 900 val perplexity 1160.9916
940
+ 900 train 7.052257 (lr=5.7145e-05) (hash(x)=6048200)
941
+ 901 train 6.920994 (lr=5.6809e-05) (hash(x)=5125556)
942
+ 902 train 6.940650 (lr=5.6476e-05) (hash(x)=6198583)
943
+ 903 train 6.924703 (lr=5.6147e-05) (hash(x)=5112805)
944
+ 904 train 7.010784 (lr=5.5821e-05) (hash(x)=6212803)
945
+ 905 train 7.053824 (lr=5.5497e-05) (hash(x)=5642978)
946
+ 906 train 7.064363 (lr=5.5178e-05) (hash(x)=7307168)
947
+ 907 train 6.977453 (lr=5.4861e-05) (hash(x)=6115590)
948
+ 908 train 6.842629 (lr=5.4547e-05) (hash(x)=7014937)
949
+ 909 train 7.248070 (lr=5.4237e-05) (hash(x)=6374852)
950
+ 910 train 6.976215 (lr=5.3930e-05) (hash(x)=5780051)
951
+ 911 train 6.973834 (lr=5.3626e-05) (hash(x)=4762818)
952
+ 912 train 6.836053 (lr=5.3325e-05) (hash(x)=6585420)
953
+ 913 train 7.181557 (lr=5.3028e-05) (hash(x)=6291590)
954
+ 914 train 7.218442 (lr=5.2734e-05) (hash(x)=5941561)
955
+ 915 train 7.018901 (lr=5.2443e-05) (hash(x)=6824617)
956
+ 916 train 7.095047 (lr=5.2155e-05) (hash(x)=5763341)
957
+ 917 train 7.132275 (lr=5.1871e-05) (hash(x)=6030150)
958
+ 918 train 7.019478 (lr=5.1589e-05) (hash(x)=5325490)
959
+ 919 train 7.081575 (lr=5.1311e-05) (hash(x)=6373236)
960
+ 920 train 7.082984 (lr=5.1037e-05) (hash(x)=6189151)
961
+ 921 train 7.025155 (lr=5.0765e-05) (hash(x)=6645157)
962
+ 922 train 7.027046 (lr=5.0497e-05) (hash(x)=6478394)
963
+ 923 train 7.155767 (lr=5.0232e-05) (hash(x)=5615918)
964
+ 924 train 7.064300 (lr=4.9971e-05) (hash(x)=6376988)
965
+ 925 train 7.099637 (lr=4.9712e-05) (hash(x)=5456350)
966
+ 926 train 7.140680 (lr=4.9457e-05) (hash(x)=5491217)
967
+ 927 train 6.950314 (lr=4.9206e-05) (hash(x)=6871163)
968
+ 928 train 6.969989 (lr=4.8957e-05) (hash(x)=7003708)
969
+ 929 train 7.072631 (lr=4.8712e-05) (hash(x)=7271154)
970
+ 930 train 7.047529 (lr=4.8470e-05) (hash(x)=5378966)
971
+ 931 train 7.055283 (lr=4.8232e-05) (hash(x)=5912082)
972
+ 932 train 6.938816 (lr=4.7997e-05) (hash(x)=5199606)
973
+ 933 train 6.984380 (lr=4.7765e-05) (hash(x)=5195464)
974
+ 934 train 6.914891 (lr=4.7537e-05) (hash(x)=7534052)
975
+ 935 train 6.925033 (lr=4.7312e-05) (hash(x)=6594135)
976
+ 936 train 7.035066 (lr=4.7090e-05) (hash(x)=6782051)
977
+ 937 train 7.118762 (lr=4.6871e-05) (hash(x)=6485577)
978
+ 938 train 7.076303 (lr=4.6656e-05) (hash(x)=6356813)
979
+ 939 train 6.960426 (lr=4.6445e-05) (hash(x)=6903068)
980
+ 940 train 7.010263 (lr=4.6236e-05) (hash(x)=6200862)
981
+ 941 train 7.061135 (lr=4.6031e-05) (hash(x)=5873929)
982
+ 942 train 7.032727 (lr=4.5830e-05) (hash(x)=5624440)
983
+ 943 train 7.029069 (lr=4.5631e-05) (hash(x)=7166972)
984
+ 944 train 7.147011 (lr=4.5437e-05) (hash(x)=6973234)
985
+ 945 train 6.968304 (lr=4.5245e-05) (hash(x)=6626956)
986
+ 946 train 7.092038 (lr=4.5057e-05) (hash(x)=5265592)
987
+ 947 train 6.795167 (lr=4.4872e-05) (hash(x)=5569439)
988
+ 948 train 6.909481 (lr=4.4691e-05) (hash(x)=4981059)
989
+ 949 train 6.965042 (lr=4.4513e-05) (hash(x)=7488974)
990
+ 950 val loss 7.0493
991
+ 950 val perplexity 1152.0797
992
+ 950 train 7.034908 (lr=4.4338e-05) (hash(x)=5160526)
993
+ 951 train 6.979475 (lr=4.4167e-05) (hash(x)=5340877)
994
+ 952 train 6.828856 (lr=4.4000e-05) (hash(x)=6333269)
995
+ 953 train 7.009393 (lr=4.3835e-05) (hash(x)=7205151)
996
+ 954 train 7.105854 (lr=4.3674e-05) (hash(x)=7256545)
997
+ 955 train 6.978057 (lr=4.3517e-05) (hash(x)=6374904)
998
+ 956 train 7.020600 (lr=4.3363e-05) (hash(x)=7124927)
999
+ 957 train 6.970525 (lr=4.3212e-05) (hash(x)=5263132)
1000
+ 958 train 7.141869 (lr=4.3065e-05) (hash(x)=8427208)
1001
+ 959 train 6.928646 (lr=4.2921e-05) (hash(x)=6884232)
1002
+ 960 train 6.946011 (lr=4.2781e-05) (hash(x)=7416977)
1003
+ 961 train 7.035114 (lr=4.2644e-05) (hash(x)=6733050)
1004
+ 962 train 6.952401 (lr=4.2510e-05) (hash(x)=5251404)
1005
+ 963 train 7.070653 (lr=4.2380e-05) (hash(x)=8162997)
1006
+ 964 train 7.249129 (lr=4.2253e-05) (hash(x)=7234382)
1007
+ 965 train 7.159603 (lr=4.2130e-05) (hash(x)=6255275)
1008
+ 966 train 7.088919 (lr=4.2010e-05) (hash(x)=4874742)
1009
+ 967 train 6.992590 (lr=4.1894e-05) (hash(x)=6303037)
1010
+ 968 train 6.994813 (lr=4.1781e-05) (hash(x)=5002696)
1011
+ 969 train 7.032016 (lr=4.1672e-05) (hash(x)=7636268)
1012
+ 970 train 6.992894 (lr=4.1566e-05) (hash(x)=5843590)
1013
+ 971 train 7.062142 (lr=4.1463e-05) (hash(x)=5122289)
1014
+ 972 train 6.963322 (lr=4.1364e-05) (hash(x)=5785819)
1015
+ 973 train 7.021972 (lr=4.1269e-05) (hash(x)=7350343)
1016
+ 974 train 6.822265 (lr=4.1177e-05) (hash(x)=7612943)
1017
+ 975 train 7.141160 (lr=4.1088e-05) (hash(x)=7955067)
1018
+ 976 train 7.003903 (lr=4.1003e-05) (hash(x)=5520886)
1019
+ 977 train 6.900310 (lr=4.0921e-05) (hash(x)=5299200)
1020
+ 978 train 6.901451 (lr=4.0843e-05) (hash(x)=7820550)
1021
+ 979 train 6.886482 (lr=4.0768e-05) (hash(x)=3895312)
1022
+ 980 train 7.033036 (lr=4.0697e-05) (hash(x)=4726896)
1023
+ 981 train 6.827602 (lr=4.0629e-05) (hash(x)=5061167)
1024
+ 982 train 6.974776 (lr=4.0564e-05) (hash(x)=7006804)
1025
+ 983 train 7.002985 (lr=4.0503e-05) (hash(x)=7392600)
1026
+ 984 train 7.197157 (lr=4.0446e-05) (hash(x)=7717172)
1027
+ 985 train 7.270103 (lr=4.0392e-05) (hash(x)=5800587)
1028
+ 986 train 7.006440 (lr=4.0341e-05) (hash(x)=5813784)
1029
+ 987 train 6.964899 (lr=4.0294e-05) (hash(x)=5513267)
1030
+ 988 train 6.979331 (lr=4.0251e-05) (hash(x)=7679326)
1031
+ 989 train 7.019550 (lr=4.0211e-05) (hash(x)=4820278)
1032
+ 990 train 7.070889 (lr=4.0174e-05) (hash(x)=5950845)
1033
+ 991 train 6.980536 (lr=4.0141e-05) (hash(x)=6988067)
1034
+ 992 train 6.900168 (lr=4.0112e-05) (hash(x)=4536923)
1035
+ 993 train 7.067915 (lr=4.0085e-05) (hash(x)=6874134)
1036
+ 994 train 7.305395 (lr=4.0063e-05) (hash(x)=7252013)
1037
+ 995 train 6.948840 (lr=4.0044e-05) (hash(x)=6180446)
1038
+ 996 train 7.202205 (lr=4.0028e-05) (hash(x)=5936769)
1039
+ 997 train 7.114357 (lr=4.0016e-05) (hash(x)=5627313)
1040
+ 998 train 7.031322 (lr=4.0007e-05) (hash(x)=7541743)
1041
+ 999 val loss 7.0475
1042
+ 999 val perplexity 1149.9557
1043
+ 999 train 7.377769 (lr=4.0002e-05) (hash(x)=9812956)
lr4e-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:d3caa22494a4ccf6ea91bd26fafc239278952455570bf5220e732a7032787349
3
+ size 36163010
lr4e-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:b056312fe68beac595e57810759ca5aae75c83aecb407c8c52d7a28db40a2ea5
3
+ size 70764358