andrew-healey commited on
Commit
8a337aa
·
verified ·
1 Parent(s): 22ecfe3

Upload folder using huggingface_hub

Browse files
lr8e-4_total_batch_size5120_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr8e-4_total_batch_size5120_seq_len512", "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": 512, "batch_size": null, "total_batch_size": 5120, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0008, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr8e-4_total_batch_size5120_seq_len512/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr8e-4_total_batch_size5120_seq_len512/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7800
3
+ 0 val perplexity 130608.6406
4
+ 0 train 11.762018 (lr=2.7972e-06) (hash(x)=22886834)
5
+ 1 train 11.788344 (lr=5.5944e-06) (hash(x)=26375038)
6
+ 2 train 11.714007 (lr=8.3916e-06) (hash(x)=30598777)
7
+ 3 train 11.752528 (lr=1.1189e-05) (hash(x)=27234506)
8
+ 4 train 11.738993 (lr=1.3986e-05) (hash(x)=27767880)
9
+ 5 train 11.670406 (lr=1.6783e-05) (hash(x)=23702020)
10
+ 6 train 11.693433 (lr=1.9580e-05) (hash(x)=31986844)
11
+ 7 train 11.677820 (lr=2.2378e-05) (hash(x)=20782690)
12
+ 8 train 11.645739 (lr=2.5175e-05) (hash(x)=25201599)
13
+ 9 train 11.660674 (lr=2.7972e-05) (hash(x)=23094976)
14
+ 10 train 11.608644 (lr=3.0769e-05) (hash(x)=23841096)
15
+ 11 train 11.557910 (lr=3.3566e-05) (hash(x)=26532095)
16
+ 12 train 11.561889 (lr=3.6364e-05) (hash(x)=24432298)
17
+ 13 train 11.484296 (lr=3.9161e-05) (hash(x)=27151649)
18
+ 14 train 11.422485 (lr=4.1958e-05) (hash(x)=24596846)
19
+ 15 train 11.399806 (lr=4.4755e-05) (hash(x)=23890908)
20
+ 16 train 11.347547 (lr=4.7552e-05) (hash(x)=28913955)
21
+ 17 train 11.229208 (lr=5.0350e-05) (hash(x)=25588236)
22
+ 18 train 11.225406 (lr=5.3147e-05) (hash(x)=23770034)
23
+ 19 train 11.130153 (lr=5.5944e-05) (hash(x)=24011372)
24
+ 20 train 10.922532 (lr=5.8741e-05) (hash(x)=25441898)
25
+ 21 train 10.954045 (lr=6.1538e-05) (hash(x)=28375581)
26
+ 22 train 10.857135 (lr=6.4336e-05) (hash(x)=24046679)
27
+ 23 train 10.827314 (lr=6.7133e-05) (hash(x)=24611628)
28
+ 24 train 10.791519 (lr=6.9930e-05) (hash(x)=26169030)
29
+ 25 train 10.705917 (lr=7.2727e-05) (hash(x)=30298407)
30
+ 26 train 10.599252 (lr=7.5524e-05) (hash(x)=23711112)
31
+ 27 train 10.428332 (lr=7.8322e-05) (hash(x)=19245352)
32
+ 28 train 10.380673 (lr=8.1119e-05) (hash(x)=21529136)
33
+ 29 train 10.585643 (lr=8.3916e-05) (hash(x)=28936608)
34
+ 30 train 10.282206 (lr=8.6713e-05) (hash(x)=24339013)
35
+ 31 train 10.138211 (lr=8.9510e-05) (hash(x)=25767553)
36
+ 32 train 10.220189 (lr=9.2308e-05) (hash(x)=26439905)
37
+ 33 train 10.435468 (lr=9.5105e-05) (hash(x)=31093473)
38
+ 34 train 10.166007 (lr=9.7902e-05) (hash(x)=25450374)
39
+ 35 train 10.176581 (lr=1.0070e-04) (hash(x)=24809873)
40
+ 36 train 10.332622 (lr=1.0350e-04) (hash(x)=23253252)
41
+ 37 train 10.351113 (lr=1.0629e-04) (hash(x)=27852919)
42
+ 38 train 10.168527 (lr=1.0909e-04) (hash(x)=23327497)
43
+ 39 train 10.145299 (lr=1.1189e-04) (hash(x)=22512166)
44
+ 40 train 10.251644 (lr=1.1469e-04) (hash(x)=22859419)
45
+ 41 train 10.039390 (lr=1.1748e-04) (hash(x)=27620338)
46
+ 42 train 10.048766 (lr=1.2028e-04) (hash(x)=26397837)
47
+ 43 train 10.077471 (lr=1.2308e-04) (hash(x)=28092148)
48
+ 44 train 9.990936 (lr=1.2587e-04) (hash(x)=24662703)
49
+ 45 train 10.115264 (lr=1.2867e-04) (hash(x)=27938767)
50
+ 46 train 10.169042 (lr=1.3147e-04) (hash(x)=26037988)
51
+ 47 train 9.908460 (lr=1.3427e-04) (hash(x)=24732833)
52
+ 48 train 9.927807 (lr=1.3706e-04) (hash(x)=25259526)
53
+ 49 train 9.908470 (lr=1.3986e-04) (hash(x)=23200230)
54
+ 50 val loss 9.8647
55
+ 50 val perplexity 19238.8320
56
+ 50 train 9.863123 (lr=1.4266e-04) (hash(x)=26721357)
57
+ 51 train 9.843738 (lr=1.4545e-04) (hash(x)=22694718)
58
+ 52 train 9.792109 (lr=1.4825e-04) (hash(x)=28066766)
59
+ 53 train 9.741092 (lr=1.5105e-04) (hash(x)=23125151)
60
+ 54 train 9.798204 (lr=1.5385e-04) (hash(x)=27193725)
61
+ 55 train 9.672717 (lr=1.5664e-04) (hash(x)=25129410)
62
+ 56 train 9.713124 (lr=1.5944e-04) (hash(x)=24263988)
63
+ 57 train 9.519842 (lr=1.6224e-04) (hash(x)=23059154)
64
+ 58 train 9.670671 (lr=1.6503e-04) (hash(x)=26063864)
65
+ 59 train 9.820041 (lr=1.6783e-04) (hash(x)=27858570)
66
+ 60 train 9.561538 (lr=1.7063e-04) (hash(x)=23874620)
67
+ 61 train 9.502345 (lr=1.7343e-04) (hash(x)=22402617)
68
+ 62 train 9.514735 (lr=1.7622e-04) (hash(x)=23600822)
69
+ 63 train 9.298546 (lr=1.7902e-04) (hash(x)=26582391)
70
+ 64 train 9.290473 (lr=1.8182e-04) (hash(x)=23225283)
71
+ 65 train 9.502402 (lr=1.8462e-04) (hash(x)=26075451)
72
+ 66 train 9.376815 (lr=1.8741e-04) (hash(x)=24723419)
73
+ 67 train 9.331706 (lr=1.9021e-04) (hash(x)=27279806)
74
+ 68 train 9.249269 (lr=1.9301e-04) (hash(x)=25870391)
75
+ 69 train 9.331085 (lr=1.9580e-04) (hash(x)=26188136)
76
+ 70 train 9.235641 (lr=1.9860e-04) (hash(x)=30373443)
77
+ 71 train 9.286670 (lr=2.0140e-04) (hash(x)=26472336)
78
+ 72 train 9.186545 (lr=2.0420e-04) (hash(x)=26651572)
79
+ 73 train 9.122893 (lr=2.0699e-04) (hash(x)=26376212)
80
+ 74 train 9.101300 (lr=2.0979e-04) (hash(x)=26733350)
81
+ 75 train 9.260167 (lr=2.1259e-04) (hash(x)=28301589)
82
+ 76 train 9.061399 (lr=2.1538e-04) (hash(x)=27599559)
83
+ 77 train 9.009951 (lr=2.1818e-04) (hash(x)=28035221)
84
+ 78 train 9.045893 (lr=2.2098e-04) (hash(x)=25016783)
85
+ 79 train 9.068748 (lr=2.2378e-04) (hash(x)=27654289)
86
+ 80 train 8.989839 (lr=2.2657e-04) (hash(x)=24597558)
87
+ 81 train 8.895994 (lr=2.2937e-04) (hash(x)=21560904)
88
+ 82 train 8.676881 (lr=2.3217e-04) (hash(x)=21983837)
89
+ 83 train 8.795328 (lr=2.3497e-04) (hash(x)=24995715)
90
+ 84 train 8.956858 (lr=2.3776e-04) (hash(x)=29876413)
91
+ 85 train 8.698668 (lr=2.4056e-04) (hash(x)=23792508)
92
+ 86 train 8.586703 (lr=2.4336e-04) (hash(x)=25509120)
93
+ 87 train 8.840113 (lr=2.4615e-04) (hash(x)=26559876)
94
+ 88 train 8.750717 (lr=2.4895e-04) (hash(x)=23569647)
95
+ 89 train 8.693580 (lr=2.5175e-04) (hash(x)=25758852)
96
+ 90 train 8.542712 (lr=2.5455e-04) (hash(x)=25706298)
97
+ 91 train 8.524730 (lr=2.5734e-04) (hash(x)=28364895)
98
+ 92 train 8.428611 (lr=2.6014e-04) (hash(x)=25304663)
99
+ 93 train 8.559031 (lr=2.6294e-04) (hash(x)=25269299)
100
+ 94 train 8.637262 (lr=2.6573e-04) (hash(x)=25870566)
101
+ 95 train 8.274873 (lr=2.6853e-04) (hash(x)=21770329)
102
+ 96 train 8.633975 (lr=2.7133e-04) (hash(x)=27595900)
103
+ 97 train 8.580072 (lr=2.7413e-04) (hash(x)=24785397)
104
+ 98 train 8.255664 (lr=2.7692e-04) (hash(x)=21521480)
105
+ 99 train 8.339123 (lr=2.7972e-04) (hash(x)=24628606)
106
+ 100 val loss 8.3593
107
+ 100 val perplexity 4269.6709
108
+ 100 train 8.421602 (lr=2.8252e-04) (hash(x)=24670150)
109
+ 101 train 8.344586 (lr=2.8531e-04) (hash(x)=23181910)
110
+ 102 train 8.169065 (lr=2.8811e-04) (hash(x)=22714991)
111
+ 103 train 8.199615 (lr=2.9091e-04) (hash(x)=22723459)
112
+ 104 train 8.125424 (lr=2.9371e-04) (hash(x)=21524316)
113
+ 105 train 8.263425 (lr=2.9650e-04) (hash(x)=25506632)
114
+ 106 train 7.938436 (lr=2.9930e-04) (hash(x)=21675672)
115
+ 107 train 8.069987 (lr=3.0210e-04) (hash(x)=22897919)
116
+ 108 train 8.079854 (lr=3.0490e-04) (hash(x)=23321631)
117
+ 109 train 8.182334 (lr=3.0769e-04) (hash(x)=26546719)
118
+ 110 train 9.236870 (lr=3.1049e-04) (hash(x)=31962348)
119
+ 111 train 8.422080 (lr=3.1329e-04) (hash(x)=30338342)
120
+ 112 train 8.092641 (lr=3.1608e-04) (hash(x)=23724471)
121
+ 113 train 8.175478 (lr=3.1888e-04) (hash(x)=29175888)
122
+ 114 train 8.207407 (lr=3.2168e-04) (hash(x)=23256716)
123
+ 115 train 8.242843 (lr=3.2448e-04) (hash(x)=27063280)
124
+ 116 train 8.248553 (lr=3.2727e-04) (hash(x)=31057659)
125
+ 117 train 8.482058 (lr=3.3007e-04) (hash(x)=32915097)
126
+ 118 train 8.232648 (lr=3.3287e-04) (hash(x)=28842717)
127
+ 119 train 7.957227 (lr=3.3566e-04) (hash(x)=25678059)
128
+ 120 train 7.820350 (lr=3.3846e-04) (hash(x)=21593510)
129
+ 121 train 7.697166 (lr=3.4126e-04) (hash(x)=20083773)
130
+ 122 train 7.925789 (lr=3.4406e-04) (hash(x)=23002820)
131
+ 123 train 7.776803 (lr=3.4685e-04) (hash(x)=21853028)
132
+ 124 train 7.997765 (lr=3.4965e-04) (hash(x)=26985625)
133
+ 125 train 7.755432 (lr=3.5245e-04) (hash(x)=21808483)
134
+ 126 train 8.152184 (lr=3.5524e-04) (hash(x)=28873251)
135
+ 127 train 7.971179 (lr=3.5804e-04) (hash(x)=26109335)
136
+ 128 train 8.159985 (lr=3.6084e-04) (hash(x)=26334674)
137
+ 129 train 8.001292 (lr=3.6364e-04) (hash(x)=24916754)
138
+ 130 train 7.972189 (lr=3.6643e-04) (hash(x)=25449624)
139
+ 131 train 8.030248 (lr=3.6923e-04) (hash(x)=25334848)
140
+ 132 train 8.312885 (lr=3.7203e-04) (hash(x)=27484863)
141
+ 133 train 7.840910 (lr=3.7483e-04) (hash(x)=24917705)
142
+ 134 train 7.782190 (lr=3.7762e-04) (hash(x)=25143449)
143
+ 135 train 7.840505 (lr=3.8042e-04) (hash(x)=25044885)
144
+ 136 train 8.143399 (lr=3.8322e-04) (hash(x)=27821028)
145
+ 137 train 8.086606 (lr=3.8601e-04) (hash(x)=28747022)
146
+ 138 train 8.018038 (lr=3.8881e-04) (hash(x)=27182888)
147
+ 139 train 7.830898 (lr=3.9161e-04) (hash(x)=23678349)
148
+ 140 train 7.916873 (lr=3.9441e-04) (hash(x)=23593235)
149
+ 141 train 8.220747 (lr=3.9720e-04) (hash(x)=28529813)
150
+ 142 train 8.563222 (lr=4.0000e-04) (hash(x)=32074661)
151
+ 143 train 8.099444 (lr=4.0280e-04) (hash(x)=28870690)
152
+ 144 train 8.058742 (lr=4.0559e-04) (hash(x)=27307705)
153
+ 145 train 7.882468 (lr=4.0839e-04) (hash(x)=25044834)
154
+ 146 train 7.914749 (lr=4.1119e-04) (hash(x)=23712023)
155
+ 147 train 8.463585 (lr=4.1399e-04) (hash(x)=32982615)
156
+ 148 train 8.174138 (lr=4.1678e-04) (hash(x)=30113660)
157
+ 149 train 7.646344 (lr=4.1958e-04) (hash(x)=20970960)
158
+ 150 val loss 7.8642
159
+ 150 val perplexity 2602.3101
160
+ 150 train 7.840605 (lr=4.2238e-04) (hash(x)=23132684)
161
+ 151 train 9.691226 (lr=4.2517e-04) (hash(x)=35279941)
162
+ 152 train 8.761647 (lr=4.2797e-04) (hash(x)=31227444)
163
+ 153 train 8.022449 (lr=4.3077e-04) (hash(x)=25529472)
164
+ 154 train 8.021708 (lr=4.3357e-04) (hash(x)=24350409)
165
+ 155 train 8.292819 (lr=4.3636e-04) (hash(x)=26400041)
166
+ 156 train 7.943471 (lr=4.3916e-04) (hash(x)=25262621)
167
+ 157 train 7.942135 (lr=4.4196e-04) (hash(x)=24656138)
168
+ 158 train 8.105050 (lr=4.4476e-04) (hash(x)=26803414)
169
+ 159 train 7.961605 (lr=4.4755e-04) (hash(x)=25015923)
170
+ 160 train 7.717406 (lr=4.5035e-04) (hash(x)=23581172)
171
+ 161 train 7.821355 (lr=4.5315e-04) (hash(x)=22924885)
172
+ 162 train 7.777078 (lr=4.5594e-04) (hash(x)=23414296)
173
+ 163 train 8.034031 (lr=4.5874e-04) (hash(x)=24853586)
174
+ 164 train 8.065930 (lr=4.6154e-04) (hash(x)=25000130)
175
+ 165 train 7.932785 (lr=4.6434e-04) (hash(x)=27004780)
176
+ 166 train 7.973836 (lr=4.6713e-04) (hash(x)=26148573)
177
+ 167 train 7.882455 (lr=4.6993e-04) (hash(x)=26740855)
178
+ 168 train 7.515782 (lr=4.7273e-04) (hash(x)=20965419)
179
+ 169 train 7.777332 (lr=4.7552e-04) (hash(x)=23950114)
180
+ 170 train 7.730685 (lr=4.7832e-04) (hash(x)=24951982)
181
+ 171 train 7.864974 (lr=4.8112e-04) (hash(x)=24584116)
182
+ 172 train 7.851701 (lr=4.8392e-04) (hash(x)=24378759)
183
+ 173 train 7.874352 (lr=4.8671e-04) (hash(x)=25718516)
184
+ 174 train 8.000737 (lr=4.8951e-04) (hash(x)=28424396)
185
+ 175 train 7.832283 (lr=4.9231e-04) (hash(x)=22262151)
186
+ 176 train 8.062624 (lr=4.9510e-04) (hash(x)=26438412)
187
+ 177 train 7.793323 (lr=4.9790e-04) (hash(x)=23025303)
188
+ 178 train 7.833280 (lr=5.0070e-04) (hash(x)=24190770)
189
+ 179 train 7.981001 (lr=5.0350e-04) (hash(x)=26627860)
190
+ 180 train 7.756132 (lr=5.0629e-04) (hash(x)=23663439)
191
+ 181 train 7.655196 (lr=5.0909e-04) (hash(x)=23700532)
192
+ 182 train 7.668510 (lr=5.1189e-04) (hash(x)=23075676)
193
+ 183 train 7.936311 (lr=5.1469e-04) (hash(x)=26621834)
194
+ 184 train 8.020643 (lr=5.1748e-04) (hash(x)=29426269)
195
+ 185 train 7.688787 (lr=5.2028e-04) (hash(x)=21821465)
196
+ 186 train 7.958414 (lr=5.2308e-04) (hash(x)=26506130)
197
+ 187 train 8.231274 (lr=5.2587e-04) (hash(x)=26930630)
198
+ 188 train 7.776676 (lr=5.2867e-04) (hash(x)=22993793)
199
+ 189 train 7.663370 (lr=5.3147e-04) (hash(x)=19557946)
200
+ 190 train 7.900905 (lr=5.3427e-04) (hash(x)=23572891)
201
+ 191 train 7.703701 (lr=5.3706e-04) (hash(x)=23234741)
202
+ 192 train 7.799783 (lr=5.3986e-04) (hash(x)=25547951)
203
+ 193 train 7.871212 (lr=5.4266e-04) (hash(x)=26713563)
204
+ 194 train 7.909348 (lr=5.4545e-04) (hash(x)=25913622)
205
+ 195 train 8.213284 (lr=5.4825e-04) (hash(x)=28102443)
206
+ 196 train 7.799465 (lr=5.5105e-04) (hash(x)=23093351)
207
+ 197 train 8.612424 (lr=5.5385e-04) (hash(x)=31689122)
208
+ 198 train 8.472438 (lr=5.5664e-04) (hash(x)=32252517)
209
+ 199 train 8.018157 (lr=5.5944e-04) (hash(x)=25470563)
210
+ 200 val loss 7.9574
211
+ 200 val perplexity 2856.7104
212
+ 200 train 8.017978 (lr=5.6224e-04) (hash(x)=25597614)
213
+ 201 train 8.042575 (lr=5.6503e-04) (hash(x)=23757479)
214
+ 202 train 7.804260 (lr=5.6783e-04) (hash(x)=25140048)
215
+ 203 train 8.048356 (lr=5.7063e-04) (hash(x)=28282861)
216
+ 204 train 7.767929 (lr=5.7343e-04) (hash(x)=24754885)
217
+ 205 train 8.035678 (lr=5.7622e-04) (hash(x)=26731964)
218
+ 206 train 8.160533 (lr=5.7902e-04) (hash(x)=26660561)
219
+ 207 train 7.663044 (lr=5.8182e-04) (hash(x)=21799102)
220
+ 208 train 7.902650 (lr=5.8462e-04) (hash(x)=27025986)
221
+ 209 train 7.768052 (lr=5.8741e-04) (hash(x)=24672077)
222
+ 210 train 7.841220 (lr=5.9021e-04) (hash(x)=25322984)
223
+ 211 train 7.690701 (lr=5.9301e-04) (hash(x)=23471769)
224
+ 212 train 7.491029 (lr=5.9580e-04) (hash(x)=20766491)
225
+ 213 train 8.005263 (lr=5.9860e-04) (hash(x)=24058931)
226
+ 214 train 7.691373 (lr=6.0140e-04) (hash(x)=23872843)
227
+ 215 train 7.738661 (lr=6.0420e-04) (hash(x)=23275479)
228
+ 216 train 7.865913 (lr=6.0699e-04) (hash(x)=24914695)
229
+ 217 train 7.710634 (lr=6.0979e-04) (hash(x)=24364396)
230
+ 218 train 7.976550 (lr=6.1259e-04) (hash(x)=27986474)
231
+ 219 train 7.879230 (lr=6.1538e-04) (hash(x)=24473581)
232
+ 220 train 7.795339 (lr=6.1818e-04) (hash(x)=22974689)
233
+ 221 train 7.920116 (lr=6.2098e-04) (hash(x)=23774644)
234
+ 222 train 7.840755 (lr=6.2378e-04) (hash(x)=23245327)
235
+ 223 train 7.598310 (lr=6.2657e-04) (hash(x)=22091862)
236
+ 224 train 7.712978 (lr=6.2937e-04) (hash(x)=24362839)
237
+ 225 train 7.873438 (lr=6.3217e-04) (hash(x)=25482303)
238
+ 226 train 7.963933 (lr=6.3497e-04) (hash(x)=24911853)
239
+ 227 train 7.930900 (lr=6.3776e-04) (hash(x)=26018202)
240
+ 228 train 7.794469 (lr=6.4056e-04) (hash(x)=26124495)
241
+ 229 train 7.918449 (lr=6.4336e-04) (hash(x)=24560096)
242
+ 230 train 7.904563 (lr=6.4615e-04) (hash(x)=24695331)
243
+ 231 train 7.603589 (lr=6.4895e-04) (hash(x)=17430373)
244
+ 232 train 7.622127 (lr=6.5175e-04) (hash(x)=21813345)
245
+ 233 train 7.243373 (lr=6.5455e-04) (hash(x)=20098681)
246
+ 234 train 7.622121 (lr=6.5734e-04) (hash(x)=25095928)
247
+ 235 train 7.740546 (lr=6.6014e-04) (hash(x)=24078083)
248
+ 236 train 7.750760 (lr=6.6294e-04) (hash(x)=22901505)
249
+ 237 train 7.964634 (lr=6.6573e-04) (hash(x)=26595592)
250
+ 238 train 7.868276 (lr=6.6853e-04) (hash(x)=27663196)
251
+ 239 train 7.657093 (lr=6.7133e-04) (hash(x)=22954861)
252
+ 240 train 8.081882 (lr=6.7413e-04) (hash(x)=30159234)
253
+ 241 train 7.970140 (lr=6.7692e-04) (hash(x)=29294271)
254
+ 242 train 7.916879 (lr=6.7972e-04) (hash(x)=23728322)
255
+ 243 train 7.974487 (lr=6.8252e-04) (hash(x)=28695016)
256
+ 244 train 8.135062 (lr=6.8531e-04) (hash(x)=26702728)
257
+ 245 train 7.957509 (lr=6.8811e-04) (hash(x)=27100115)
258
+ 246 train 7.872787 (lr=6.9091e-04) (hash(x)=25082752)
259
+ 247 train 8.197650 (lr=6.9371e-04) (hash(x)=26671799)
260
+ 248 train 7.872856 (lr=6.9650e-04) (hash(x)=23718946)
261
+ 249 train 8.097557 (lr=6.9930e-04) (hash(x)=28137394)
262
+ 250 val loss 7.8564
263
+ 250 val perplexity 2582.1597
264
+ 250 train 8.141696 (lr=7.0210e-04) (hash(x)=23893495)
265
+ 251 train 7.863445 (lr=7.0490e-04) (hash(x)=23166092)
266
+ 252 train 7.883599 (lr=7.0769e-04) (hash(x)=25907665)
267
+ 253 train 7.864685 (lr=7.1049e-04) (hash(x)=25885986)
268
+ 254 train 7.886959 (lr=7.1329e-04) (hash(x)=25262712)
269
+ 255 train 7.852839 (lr=7.1608e-04) (hash(x)=26924723)
270
+ 256 train 8.098969 (lr=7.1888e-04) (hash(x)=28744736)
271
+ 257 train 7.767800 (lr=7.2168e-04) (hash(x)=26140590)
272
+ 258 train 8.043932 (lr=7.2448e-04) (hash(x)=25780449)
273
+ 259 train 7.867133 (lr=7.2727e-04) (hash(x)=25138659)
274
+ 260 train 8.155873 (lr=7.3007e-04) (hash(x)=27244046)
275
+ 261 train 7.906598 (lr=7.3287e-04) (hash(x)=27224685)
276
+ 262 train 8.237020 (lr=7.3566e-04) (hash(x)=28274477)
277
+ 263 train 7.712274 (lr=7.3846e-04) (hash(x)=23557495)
278
+ 264 train 7.851491 (lr=7.4126e-04) (hash(x)=24680596)
279
+ 265 train 7.840305 (lr=7.4406e-04) (hash(x)=23928957)
280
+ 266 train 7.796228 (lr=7.4685e-04) (hash(x)=23761390)
281
+ 267 train 7.716026 (lr=7.4965e-04) (hash(x)=25288123)
282
+ 268 train 7.963743 (lr=7.5245e-04) (hash(x)=28705502)
283
+ 269 train 7.612598 (lr=7.5524e-04) (hash(x)=23246294)
284
+ 270 train 7.710668 (lr=7.5804e-04) (hash(x)=28639079)
285
+ 271 train 7.600383 (lr=7.6084e-04) (hash(x)=27804380)
286
+ 272 train 7.785028 (lr=7.6364e-04) (hash(x)=24172235)
287
+ 273 train 7.842323 (lr=7.6643e-04) (hash(x)=23089140)
288
+ 274 train 8.025490 (lr=7.6923e-04) (hash(x)=27163701)
289
+ 275 train 8.181913 (lr=7.7203e-04) (hash(x)=26993263)
290
+ 276 train 7.915608 (lr=7.7483e-04) (hash(x)=28224233)
291
+ 277 train 8.167254 (lr=7.7762e-04) (hash(x)=27397203)
292
+ 278 train 8.101922 (lr=7.8042e-04) (hash(x)=27923882)
293
+ 279 train 7.857222 (lr=7.8322e-04) (hash(x)=26654908)
294
+ 280 train 7.838658 (lr=7.8601e-04) (hash(x)=24213147)
295
+ 281 train 7.567994 (lr=7.8881e-04) (hash(x)=21965022)
296
+ 282 train 7.690363 (lr=7.9161e-04) (hash(x)=25465685)
297
+ 283 train 7.954621 (lr=7.9441e-04) (hash(x)=27347722)
298
+ 284 train 7.950071 (lr=7.9720e-04) (hash(x)=26732050)
299
+ 285 train 8.091448 (lr=8.0000e-04) (hash(x)=28314127)
300
+ 286 train 7.708301 (lr=8.0000e-04) (hash(x)=21471186)
301
+ 287 train 7.632243 (lr=8.0000e-04) (hash(x)=23627518)
302
+ 288 train 7.610782 (lr=7.9999e-04) (hash(x)=20870353)
303
+ 289 train 7.682548 (lr=7.9997e-04) (hash(x)=25024764)
304
+ 290 train 7.472862 (lr=7.9994e-04) (hash(x)=20683822)
305
+ 291 train 7.792346 (lr=7.9991e-04) (hash(x)=21768671)
306
+ 292 train 7.867378 (lr=7.9987e-04) (hash(x)=25557309)
307
+ 293 train 8.001612 (lr=7.9983e-04) (hash(x)=25076667)
308
+ 294 train 7.677911 (lr=7.9978e-04) (hash(x)=23765822)
309
+ 295 train 7.590454 (lr=7.9972e-04) (hash(x)=21889990)
310
+ 296 train 7.817246 (lr=7.9965e-04) (hash(x)=26339893)
311
+ 297 train 7.491063 (lr=7.9958e-04) (hash(x)=20932794)
312
+ 298 train 7.655568 (lr=7.9950e-04) (hash(x)=21750070)
313
+ 299 train 7.754472 (lr=7.9941e-04) (hash(x)=23665838)
314
+ 300 val loss 7.7775
315
+ 300 val perplexity 2386.2725
316
+ 300 train 8.576934 (lr=7.9932e-04) (hash(x)=32888061)
317
+ 301 train 8.351179 (lr=7.9922e-04) (hash(x)=30223582)
318
+ 302 train 8.128295 (lr=7.9911e-04) (hash(x)=26908418)
319
+ 303 train 7.469078 (lr=7.9899e-04) (hash(x)=22528001)
320
+ 304 train 8.016008 (lr=7.9887e-04) (hash(x)=27452187)
321
+ 305 train 7.787511 (lr=7.9874e-04) (hash(x)=25181641)
322
+ 306 train 7.892481 (lr=7.9861e-04) (hash(x)=25546593)
323
+ 307 train 7.827933 (lr=7.9846e-04) (hash(x)=22487328)
324
+ 308 train 7.977264 (lr=7.9831e-04) (hash(x)=27804274)
325
+ 309 train 8.037114 (lr=7.9816e-04) (hash(x)=26544630)
326
+ 310 train 8.195479 (lr=7.9799e-04) (hash(x)=27738934)
327
+ 311 train 8.097330 (lr=7.9782e-04) (hash(x)=29248942)
328
+ 312 train 7.952436 (lr=7.9765e-04) (hash(x)=25103452)
329
+ 313 train 8.146689 (lr=7.9746e-04) (hash(x)=25052066)
330
+ 314 train 8.098853 (lr=7.9727e-04) (hash(x)=24481302)
331
+ 315 train 7.924493 (lr=7.9707e-04) (hash(x)=23543273)
332
+ 316 train 7.791909 (lr=7.9687e-04) (hash(x)=25608244)
333
+ 317 train 7.964328 (lr=7.9666e-04) (hash(x)=27451288)
334
+ 318 train 7.577507 (lr=7.9644e-04) (hash(x)=22806491)
335
+ 319 train 7.755146 (lr=7.9621e-04) (hash(x)=25533417)
336
+ 320 train 7.757766 (lr=7.9598e-04) (hash(x)=24557997)
337
+ 321 train 7.761224 (lr=7.9574e-04) (hash(x)=24432899)
338
+ 322 train 8.000078 (lr=7.9549e-04) (hash(x)=27583287)
339
+ 323 train 7.924997 (lr=7.9524e-04) (hash(x)=25552036)
340
+ 324 train 7.632097 (lr=7.9498e-04) (hash(x)=24201868)
341
+ 325 train 7.977160 (lr=7.9471e-04) (hash(x)=28149782)
342
+ 326 train 7.799114 (lr=7.9444e-04) (hash(x)=25529698)
343
+ 327 train 7.569270 (lr=7.9416e-04) (hash(x)=20612533)
344
+ 328 train 7.602509 (lr=7.9387e-04) (hash(x)=20699000)
345
+ 329 train 7.541306 (lr=7.9358e-04) (hash(x)=19774173)
346
+ 330 train 7.556673 (lr=7.9327e-04) (hash(x)=21681646)
347
+ 331 train 7.359681 (lr=7.9297e-04) (hash(x)=20216795)
348
+ 332 train 8.133517 (lr=7.9265e-04) (hash(x)=27697998)
349
+ 333 train 7.824211 (lr=7.9233e-04) (hash(x)=25896435)
350
+ 334 train 7.674326 (lr=7.9200e-04) (hash(x)=21585310)
351
+ 335 train 7.757842 (lr=7.9167e-04) (hash(x)=24677740)
352
+ 336 train 7.585959 (lr=7.9132e-04) (hash(x)=22027900)
353
+ 337 train 7.578889 (lr=7.9097e-04) (hash(x)=21835643)
354
+ 338 train 7.477405 (lr=7.9062e-04) (hash(x)=20558462)
355
+ 339 train 7.395654 (lr=7.9026e-04) (hash(x)=16722715)
356
+ 340 train 7.595539 (lr=7.8989e-04) (hash(x)=20972655)
357
+ 341 train 8.050769 (lr=7.8951e-04) (hash(x)=26303975)
358
+ 342 train 7.730185 (lr=7.8913e-04) (hash(x)=22938170)
359
+ 343 train 7.866648 (lr=7.8874e-04) (hash(x)=25347203)
360
+ 344 train 7.798642 (lr=7.8834e-04) (hash(x)=27398686)
361
+ 345 train 7.755324 (lr=7.8794e-04) (hash(x)=25973417)
362
+ 346 train 7.738847 (lr=7.8753e-04) (hash(x)=26918389)
363
+ 347 train 7.734116 (lr=7.8711e-04) (hash(x)=24272489)
364
+ 348 train 7.731317 (lr=7.8669e-04) (hash(x)=25593714)
365
+ 349 train 7.790157 (lr=7.8626e-04) (hash(x)=29260846)
366
+ 350 val loss 7.7331
367
+ 350 val perplexity 2282.7175
368
+ 350 train 7.896770 (lr=7.8582e-04) (hash(x)=27951602)
369
+ 351 train 7.981031 (lr=7.8538e-04) (hash(x)=28922363)
370
+ 352 train 7.865451 (lr=7.8493e-04) (hash(x)=27210734)
371
+ 353 train 7.877768 (lr=7.8447e-04) (hash(x)=26322572)
372
+ 354 train 7.782646 (lr=7.8401e-04) (hash(x)=27084665)
373
+ 355 train 7.684190 (lr=7.8354e-04) (hash(x)=25510798)
374
+ 356 train 7.740617 (lr=7.8306e-04) (hash(x)=24970921)
375
+ 357 train 7.819368 (lr=7.8258e-04) (hash(x)=24138948)
376
+ 358 train 7.708192 (lr=7.8209e-04) (hash(x)=24790211)
377
+ 359 train 7.858117 (lr=7.8159e-04) (hash(x)=25631397)
378
+ 360 train 7.628566 (lr=7.8109e-04) (hash(x)=23226625)
379
+ 361 train 7.741333 (lr=7.8058e-04) (hash(x)=24001903)
380
+ 362 train 7.735635 (lr=7.8006e-04) (hash(x)=24587948)
381
+ 363 train 7.414525 (lr=7.7954e-04) (hash(x)=21333676)
382
+ 364 train 7.218723 (lr=7.7901e-04) (hash(x)=23673779)
383
+ 365 train 8.807135 (lr=7.7847e-04) (hash(x)=30770484)
384
+ 366 train 7.834210 (lr=7.7793e-04) (hash(x)=26564899)
385
+ 367 train 7.846120 (lr=7.7738e-04) (hash(x)=26237983)
386
+ 368 train 7.599641 (lr=7.7682e-04) (hash(x)=23764356)
387
+ 369 train 7.929819 (lr=7.7626e-04) (hash(x)=26205744)
388
+ 370 train 7.642391 (lr=7.7569e-04) (hash(x)=19208770)
389
+ 371 train 7.803739 (lr=7.7511e-04) (hash(x)=25976502)
390
+ 372 train 7.669033 (lr=7.7453e-04) (hash(x)=23983933)
391
+ 373 train 7.680843 (lr=7.7394e-04) (hash(x)=24080636)
392
+ 374 train 7.774732 (lr=7.7335e-04) (hash(x)=24404047)
393
+ 375 train 7.720113 (lr=7.7275e-04) (hash(x)=24742645)
394
+ 376 train 7.779361 (lr=7.7214e-04) (hash(x)=24159600)
395
+ 377 train 8.038014 (lr=7.7153e-04) (hash(x)=28677257)
396
+ 378 train 7.751854 (lr=7.7091e-04) (hash(x)=25604111)
397
+ 379 train 7.740615 (lr=7.7028e-04) (hash(x)=27086333)
398
+ 380 train 7.690839 (lr=7.6964e-04) (hash(x)=25188207)
399
+ 381 train 7.911790 (lr=7.6901e-04) (hash(x)=27855233)
400
+ 382 train 7.474239 (lr=7.6836e-04) (hash(x)=19470039)
401
+ 383 train 7.812398 (lr=7.6771e-04) (hash(x)=26157660)
402
+ 384 train 7.697671 (lr=7.6705e-04) (hash(x)=25291570)
403
+ 385 train 7.693808 (lr=7.6638e-04) (hash(x)=25046062)
404
+ 386 train 7.934536 (lr=7.6571e-04) (hash(x)=27020337)
405
+ 387 train 7.626514 (lr=7.6503e-04) (hash(x)=23616370)
406
+ 388 train 7.335904 (lr=7.6435e-04) (hash(x)=19113218)
407
+ 389 train 7.630908 (lr=7.6366e-04) (hash(x)=24302232)
408
+ 390 train 7.652987 (lr=7.6296e-04) (hash(x)=22188949)
409
+ 391 train 7.446413 (lr=7.6226e-04) (hash(x)=22582169)
410
+ 392 train 7.631700 (lr=7.6155e-04) (hash(x)=24700570)
411
+ 393 train 7.545644 (lr=7.6083e-04) (hash(x)=22773833)
412
+ 394 train 7.402556 (lr=7.6011e-04) (hash(x)=21875928)
413
+ 395 train 7.702667 (lr=7.5938e-04) (hash(x)=26233189)
414
+ 396 train 7.545315 (lr=7.5865e-04) (hash(x)=24321467)
415
+ 397 train 7.830301 (lr=7.5791e-04) (hash(x)=26431507)
416
+ 398 train 7.942723 (lr=7.5716e-04) (hash(x)=28690877)
417
+ 399 train 7.834900 (lr=7.5641e-04) (hash(x)=26431960)
418
+ 400 val loss 7.7169
419
+ 400 val perplexity 2246.0032
420
+ 400 train 7.672882 (lr=7.5565e-04) (hash(x)=24580300)
421
+ 401 train 7.663965 (lr=7.5489e-04) (hash(x)=25112360)
422
+ 402 train 7.759009 (lr=7.5412e-04) (hash(x)=27597243)
423
+ 403 train 8.182242 (lr=7.5334e-04) (hash(x)=30707498)
424
+ 404 train 7.837644 (lr=7.5256e-04) (hash(x)=28485465)
425
+ 405 train 7.520735 (lr=7.5177e-04) (hash(x)=22586447)
426
+ 406 train 7.675550 (lr=7.5097e-04) (hash(x)=23175270)
427
+ 407 train 7.775589 (lr=7.5017e-04) (hash(x)=25716176)
428
+ 408 train 7.845477 (lr=7.4937e-04) (hash(x)=26861373)
429
+ 409 train 7.893079 (lr=7.4855e-04) (hash(x)=25118971)
430
+ 410 train 7.418460 (lr=7.4773e-04) (hash(x)=19829066)
431
+ 411 train 7.561564 (lr=7.4691e-04) (hash(x)=26256420)
432
+ 412 train 7.828972 (lr=7.4608e-04) (hash(x)=27796153)
433
+ 413 train 7.615864 (lr=7.4524e-04) (hash(x)=22633318)
434
+ 414 train 7.651105 (lr=7.4440e-04) (hash(x)=22589383)
435
+ 415 train 8.081419 (lr=7.4355e-04) (hash(x)=28019788)
436
+ 416 train 8.007912 (lr=7.4270e-04) (hash(x)=28970440)
437
+ 417 train 7.662087 (lr=7.4184e-04) (hash(x)=27396089)
438
+ 418 train 7.699080 (lr=7.4097e-04) (hash(x)=21183513)
439
+ 419 train 7.668869 (lr=7.4010e-04) (hash(x)=23510110)
440
+ 420 train 7.900928 (lr=7.3922e-04) (hash(x)=28833467)
441
+ 421 train 7.734310 (lr=7.3834e-04) (hash(x)=23646926)
442
+ 422 train 7.706089 (lr=7.3745e-04) (hash(x)=24697272)
443
+ 423 train 7.490810 (lr=7.3655e-04) (hash(x)=20382963)
444
+ 424 train 7.469598 (lr=7.3565e-04) (hash(x)=23467595)
445
+ 425 train 7.671999 (lr=7.3474e-04) (hash(x)=24304768)
446
+ 426 train 7.598923 (lr=7.3383e-04) (hash(x)=21392328)
447
+ 427 train 7.653069 (lr=7.3291e-04) (hash(x)=25339466)
448
+ 428 train 7.552856 (lr=7.3199e-04) (hash(x)=22092542)
449
+ 429 train 7.640275 (lr=7.3106e-04) (hash(x)=22088696)
450
+ 430 train 7.567098 (lr=7.3012e-04) (hash(x)=22184471)
451
+ 431 train 7.672014 (lr=7.2918e-04) (hash(x)=24489647)
452
+ 432 train 7.689391 (lr=7.2824e-04) (hash(x)=26794132)
453
+ 433 train 7.483659 (lr=7.2729e-04) (hash(x)=22940357)
454
+ 434 train 7.658186 (lr=7.2633e-04) (hash(x)=23719522)
455
+ 435 train 7.491032 (lr=7.2537e-04) (hash(x)=22927699)
456
+ 436 train 7.703458 (lr=7.2440e-04) (hash(x)=26068576)
457
+ 437 train 7.840613 (lr=7.2342e-04) (hash(x)=27631132)
458
+ 438 train 7.708693 (lr=7.2244e-04) (hash(x)=26739991)
459
+ 439 train 7.818534 (lr=7.2146e-04) (hash(x)=25128502)
460
+ 440 train 7.850517 (lr=7.2047e-04) (hash(x)=25657260)
461
+ 441 train 7.544271 (lr=7.1947e-04) (hash(x)=23576982)
462
+ 442 train 7.781532 (lr=7.1847e-04) (hash(x)=27117886)
463
+ 443 train 7.750961 (lr=7.1747e-04) (hash(x)=25808969)
464
+ 444 train 7.644546 (lr=7.1645e-04) (hash(x)=24738238)
465
+ 445 train 7.647129 (lr=7.1544e-04) (hash(x)=23429962)
466
+ 446 train 7.828409 (lr=7.1441e-04) (hash(x)=25075165)
467
+ 447 train 7.631224 (lr=7.1339e-04) (hash(x)=25231390)
468
+ 448 train 7.469218 (lr=7.1235e-04) (hash(x)=22055054)
469
+ 449 train 7.683580 (lr=7.1131e-04) (hash(x)=25395441)
470
+ 450 val loss 7.7056
471
+ 450 val perplexity 2220.8008
472
+ 450 train 7.761360 (lr=7.1027e-04) (hash(x)=25863209)
473
+ 451 train 7.404840 (lr=7.0922e-04) (hash(x)=21154388)
474
+ 452 train 7.371719 (lr=7.0817e-04) (hash(x)=21600876)
475
+ 453 train 7.476564 (lr=7.0711e-04) (hash(x)=24278611)
476
+ 454 train 7.374765 (lr=7.0604e-04) (hash(x)=23221720)
477
+ 455 train 7.667328 (lr=7.0497e-04) (hash(x)=22708977)
478
+ 456 train 8.317062 (lr=7.0390e-04) (hash(x)=23637758)
479
+ 457 train 8.027709 (lr=7.0282e-04) (hash(x)=28228490)
480
+ 458 train 8.011183 (lr=7.0173e-04) (hash(x)=28638071)
481
+ 459 train 7.840356 (lr=7.0064e-04) (hash(x)=27258353)
482
+ 460 train 7.821271 (lr=6.9955e-04) (hash(x)=26604728)
483
+ 461 train 7.775870 (lr=6.9845e-04) (hash(x)=23252199)
484
+ 462 train 7.869496 (lr=6.9734e-04) (hash(x)=26441427)
485
+ 463 train 7.607005 (lr=6.9623e-04) (hash(x)=24364920)
486
+ 464 train 7.706796 (lr=6.9512e-04) (hash(x)=25623792)
487
+ 465 train 7.611350 (lr=6.9400e-04) (hash(x)=23283905)
488
+ 466 train 8.018950 (lr=6.9287e-04) (hash(x)=26025267)
489
+ 467 train 7.771983 (lr=6.9174e-04) (hash(x)=27243972)
490
+ 468 train 8.112730 (lr=6.9061e-04) (hash(x)=30449945)
491
+ 469 train 7.792244 (lr=6.8947e-04) (hash(x)=28113043)
492
+ 470 train 7.615073 (lr=6.8832e-04) (hash(x)=25182521)
493
+ 471 train 7.662032 (lr=6.8717e-04) (hash(x)=24932925)
494
+ 472 train 7.467474 (lr=6.8602e-04) (hash(x)=20353098)
495
+ 473 train 7.394199 (lr=6.8486e-04) (hash(x)=19001259)
496
+ 474 train 7.963696 (lr=6.8370e-04) (hash(x)=27585685)
497
+ 475 train 7.644670 (lr=6.8253e-04) (hash(x)=26371091)
498
+ 476 train 7.510098 (lr=6.8136e-04) (hash(x)=24891798)
499
+ 477 train 7.538738 (lr=6.8018e-04) (hash(x)=24258817)
500
+ 478 train 7.617583 (lr=6.7900e-04) (hash(x)=24330263)
501
+ 479 train 7.747804 (lr=6.7781e-04) (hash(x)=26913684)
502
+ 480 train 7.699680 (lr=6.7662e-04) (hash(x)=26338455)
503
+ 481 train 7.971358 (lr=6.7542e-04) (hash(x)=27753043)
504
+ 482 train 7.729463 (lr=6.7422e-04) (hash(x)=26123289)
505
+ 483 train 7.877473 (lr=6.7302e-04) (hash(x)=29239611)
506
+ 484 train 7.806308 (lr=6.7181e-04) (hash(x)=26553003)
507
+ 485 train 7.719978 (lr=6.7059e-04) (hash(x)=22984557)
508
+ 486 train 7.319560 (lr=6.6937e-04) (hash(x)=16947491)
509
+ 487 train 7.298081 (lr=6.6815e-04) (hash(x)=18017792)
510
+ 488 train 7.440715 (lr=6.6692e-04) (hash(x)=19918608)
511
+ 489 train 7.904072 (lr=6.6569e-04) (hash(x)=23374526)
512
+ 490 train 7.857007 (lr=6.6446e-04) (hash(x)=25009505)
513
+ 491 train 7.823639 (lr=6.6322e-04) (hash(x)=27574089)
514
+ 492 train 7.967720 (lr=6.6197e-04) (hash(x)=24122664)
515
+ 493 train 7.822350 (lr=6.6072e-04) (hash(x)=26154906)
516
+ 494 train 7.773486 (lr=6.5947e-04) (hash(x)=25192767)
517
+ 495 train 7.977979 (lr=6.5821e-04) (hash(x)=28613882)
518
+ 496 train 7.902569 (lr=6.5695e-04) (hash(x)=23547219)
519
+ 497 train 7.840771 (lr=6.5568e-04) (hash(x)=25272182)
520
+ 498 train 7.752533 (lr=6.5441e-04) (hash(x)=24992761)
521
+ 499 train 8.020627 (lr=6.5314e-04) (hash(x)=26981914)
522
+ 500 val loss 7.8523
523
+ 500 val perplexity 2571.7212
524
+ 500 train 7.730431 (lr=6.5186e-04) (hash(x)=22051933)
525
+ 501 train 7.944839 (lr=6.5058e-04) (hash(x)=24232348)
526
+ 502 train 7.903593 (lr=6.4929e-04) (hash(x)=23158331)
527
+ 503 train 7.726024 (lr=6.4800e-04) (hash(x)=22652243)
528
+ 504 train 7.765051 (lr=6.4670e-04) (hash(x)=23805602)
529
+ 505 train 8.041995 (lr=6.4540e-04) (hash(x)=25411991)
530
+ 506 train 7.388538 (lr=6.4410e-04) (hash(x)=18827215)
531
+ 507 train 7.103041 (lr=6.4280e-04) (hash(x)=15446025)
532
+ 508 train 7.428703 (lr=6.4148e-04) (hash(x)=20516263)
533
+ 509 train 7.901276 (lr=6.4017e-04) (hash(x)=27846176)
534
+ 510 train 7.566938 (lr=6.3885e-04) (hash(x)=23342449)
535
+ 511 train 7.844957 (lr=6.3753e-04) (hash(x)=27194521)
536
+ 512 train 7.646736 (lr=6.3620e-04) (hash(x)=23008284)
537
+ 513 train 8.093014 (lr=6.3487e-04) (hash(x)=29430001)
538
+ 514 train 7.458516 (lr=6.3354e-04) (hash(x)=22579319)
539
+ 515 train 7.525160 (lr=6.3220e-04) (hash(x)=25264518)
540
+ 516 train 7.859096 (lr=6.3086e-04) (hash(x)=25359075)
541
+ 517 train 7.793195 (lr=6.2952e-04) (hash(x)=25568956)
542
+ 518 train 8.109798 (lr=6.2817e-04) (hash(x)=32004108)
543
+ 519 train 7.749269 (lr=6.2681e-04) (hash(x)=24936836)
544
+ 520 train 7.862515 (lr=6.2546e-04) (hash(x)=27263338)
545
+ 521 train 7.870036 (lr=6.2410e-04) (hash(x)=27452099)
546
+ 522 train 7.772457 (lr=6.2274e-04) (hash(x)=25965406)
547
+ 523 train 7.969261 (lr=6.2137e-04) (hash(x)=28197282)
548
+ 524 train 7.640119 (lr=6.2000e-04) (hash(x)=22466209)
549
+ 525 train 7.741972 (lr=6.1863e-04) (hash(x)=22931889)
550
+ 526 train 7.870201 (lr=6.1725e-04) (hash(x)=26903920)
551
+ 527 train 7.716482 (lr=6.1587e-04) (hash(x)=24765578)
552
+ 528 train 7.862207 (lr=6.1449e-04) (hash(x)=27811359)
553
+ 529 train 7.639336 (lr=6.1310e-04) (hash(x)=25078649)
554
+ 530 train 7.829195 (lr=6.1171e-04) (hash(x)=25572416)
555
+ 531 train 7.934850 (lr=6.1031e-04) (hash(x)=27448185)
556
+ 532 train 7.976680 (lr=6.0892e-04) (hash(x)=25923719)
557
+ 533 train 7.786470 (lr=6.0752e-04) (hash(x)=24804856)
558
+ 534 train 7.935441 (lr=6.0611e-04) (hash(x)=23207829)
559
+ 535 train 7.832587 (lr=6.0471e-04) (hash(x)=23107416)
560
+ 536 train 7.667039 (lr=6.0330e-04) (hash(x)=26739531)
561
+ 537 train 7.667488 (lr=6.0188e-04) (hash(x)=24960796)
562
+ 538 train 7.796093 (lr=6.0047e-04) (hash(x)=24667802)
563
+ 539 train 7.952707 (lr=5.9905e-04) (hash(x)=26755138)
564
+ 540 train 7.771085 (lr=5.9762e-04) (hash(x)=25537132)
565
+ 541 train 7.721744 (lr=5.9620e-04) (hash(x)=24542526)
566
+ 542 train 7.539455 (lr=5.9477e-04) (hash(x)=21296355)
567
+ 543 train 7.982269 (lr=5.9334e-04) (hash(x)=29314255)
568
+ 544 train 7.701555 (lr=5.9190e-04) (hash(x)=26001799)
569
+ 545 train 7.811852 (lr=5.9047e-04) (hash(x)=27347755)
570
+ 546 train 7.779440 (lr=5.8903e-04) (hash(x)=25107798)
571
+ 547 train 7.579586 (lr=5.8758e-04) (hash(x)=22112669)
572
+ 548 train 7.544932 (lr=5.8614e-04) (hash(x)=21897967)
573
+ 549 train 7.656406 (lr=5.8469e-04) (hash(x)=25161929)
574
+ 550 val loss 7.6863
575
+ 550 val perplexity 2178.2954
576
+ 550 train 7.543018 (lr=5.8323e-04) (hash(x)=27465106)
577
+ 551 train 7.688510 (lr=5.8178e-04) (hash(x)=24013079)
578
+ 552 train 7.597805 (lr=5.8032e-04) (hash(x)=23142015)
579
+ 553 train 7.673297 (lr=5.7886e-04) (hash(x)=26768629)
580
+ 554 train 7.702567 (lr=5.7740e-04) (hash(x)=26393383)
581
+ 555 train 7.492744 (lr=5.7593e-04) (hash(x)=22537194)
582
+ 556 train 7.759174 (lr=5.7447e-04) (hash(x)=24046036)
583
+ 557 train 7.700624 (lr=5.7300e-04) (hash(x)=24974360)
584
+ 558 train 7.932451 (lr=5.7152e-04) (hash(x)=28379928)
585
+ 559 train 7.701679 (lr=5.7005e-04) (hash(x)=25322001)
586
+ 560 train 7.770366 (lr=5.6857e-04) (hash(x)=26622031)
587
+ 561 train 7.476768 (lr=5.6709e-04) (hash(x)=20562247)
588
+ 562 train 7.827218 (lr=5.6560e-04) (hash(x)=27381885)
589
+ 563 train 7.943812 (lr=5.6412e-04) (hash(x)=27028126)
590
+ 564 train 7.768546 (lr=5.6263e-04) (hash(x)=28882928)
591
+ 565 train 7.701894 (lr=5.6114e-04) (hash(x)=25666355)
592
+ 566 train 7.785081 (lr=5.5965e-04) (hash(x)=24330810)
593
+ 567 train 7.726898 (lr=5.5815e-04) (hash(x)=26690440)
594
+ 568 train 7.570568 (lr=5.5666e-04) (hash(x)=22923592)
595
+ 569 train 7.689852 (lr=5.5516e-04) (hash(x)=27348418)
596
+ 570 train 7.714531 (lr=5.5365e-04) (hash(x)=28849848)
597
+ 571 train 7.760826 (lr=5.5215e-04) (hash(x)=26967331)
598
+ 572 train 7.514410 (lr=5.5064e-04) (hash(x)=22831467)
599
+ 573 train 7.635793 (lr=5.4913e-04) (hash(x)=24765121)
600
+ 574 train 7.649658 (lr=5.4762e-04) (hash(x)=24331857)
601
+ 575 train 7.484460 (lr=5.4611e-04) (hash(x)=22598512)
602
+ 576 train 7.684183 (lr=5.4460e-04) (hash(x)=25149353)
603
+ 577 train 7.572140 (lr=5.4308e-04) (hash(x)=23725598)
604
+ 578 train 7.759624 (lr=5.4156e-04) (hash(x)=26449557)
605
+ 579 train 7.764091 (lr=5.4004e-04) (hash(x)=24697985)
606
+ 580 train 7.781351 (lr=5.3852e-04) (hash(x)=26923059)
607
+ 581 train 7.551102 (lr=5.3699e-04) (hash(x)=25201962)
608
+ 582 train 7.202026 (lr=5.3547e-04) (hash(x)=20931520)
609
+ 583 train 7.335045 (lr=5.3394e-04) (hash(x)=18473911)
610
+ 584 train 7.517091 (lr=5.3241e-04) (hash(x)=21306267)
611
+ 585 train 7.727769 (lr=5.3088e-04) (hash(x)=25982840)
612
+ 586 train 7.599980 (lr=5.2934e-04) (hash(x)=25364874)
613
+ 587 train 7.460021 (lr=5.2781e-04) (hash(x)=23172124)
614
+ 588 train 7.795931 (lr=5.2627e-04) (hash(x)=27876897)
615
+ 589 train 8.409831 (lr=5.2473e-04) (hash(x)=34646114)
616
+ 590 train 8.655646 (lr=5.2319e-04) (hash(x)=35153576)
617
+ 591 train 7.629370 (lr=5.2165e-04) (hash(x)=22322442)
618
+ 592 train 7.828239 (lr=5.2011e-04) (hash(x)=27907331)
619
+ 593 train 7.707060 (lr=5.1856e-04) (hash(x)=26211794)
620
+ 594 train 7.895902 (lr=5.1702e-04) (hash(x)=29291512)
621
+ 595 train 8.081199 (lr=5.1547e-04) (hash(x)=29659121)
622
+ 596 train 7.957455 (lr=5.1392e-04) (hash(x)=29674399)
623
+ 597 train 7.623165 (lr=5.1237e-04) (hash(x)=23538306)
624
+ 598 train 7.710436 (lr=5.1082e-04) (hash(x)=21991524)
625
+ 599 train 7.619271 (lr=5.0926e-04) (hash(x)=26324153)
626
+ 600 val loss 7.6176
627
+ 600 val perplexity 2033.7753
628
+ 600 train 7.631342 (lr=5.0771e-04) (hash(x)=23712082)
629
+ 601 train 7.688214 (lr=5.0615e-04) (hash(x)=24910403)
630
+ 602 train 7.636192 (lr=5.0459e-04) (hash(x)=26737205)
631
+ 603 train 7.753280 (lr=5.0303e-04) (hash(x)=26939970)
632
+ 604 train 7.627880 (lr=5.0147e-04) (hash(x)=27651943)
633
+ 605 train 7.905376 (lr=4.9991e-04) (hash(x)=27515446)
634
+ 606 train 7.826280 (lr=4.9835e-04) (hash(x)=26753129)
635
+ 607 train 7.560742 (lr=4.9679e-04) (hash(x)=23446058)
636
+ 608 train 7.816426 (lr=4.9522e-04) (hash(x)=27587849)
637
+ 609 train 7.724239 (lr=4.9366e-04) (hash(x)=25308253)
638
+ 610 train 7.718523 (lr=4.9209e-04) (hash(x)=26615098)
639
+ 611 train 7.758392 (lr=4.9052e-04) (hash(x)=29981801)
640
+ 612 train 7.840088 (lr=4.8895e-04) (hash(x)=29592345)
641
+ 613 train 7.908283 (lr=4.8738e-04) (hash(x)=23470413)
642
+ 614 train 7.766068 (lr=4.8581e-04) (hash(x)=24742370)
643
+ 615 train 7.740089 (lr=4.8424e-04) (hash(x)=24843741)
644
+ 616 train 7.652130 (lr=4.8267e-04) (hash(x)=25192548)
645
+ 617 train 7.767859 (lr=4.8109e-04) (hash(x)=27176996)
646
+ 618 train 7.610803 (lr=4.7952e-04) (hash(x)=23964552)
647
+ 619 train 7.581560 (lr=4.7795e-04) (hash(x)=22855363)
648
+ 620 train 7.559329 (lr=4.7637e-04) (hash(x)=26332996)
649
+ 621 train 7.396117 (lr=4.7479e-04) (hash(x)=22960957)
650
+ 622 train 7.741891 (lr=4.7322e-04) (hash(x)=22752597)
651
+ 623 train 7.632192 (lr=4.7164e-04) (hash(x)=23197102)
652
+ 624 train 8.067077 (lr=4.7006e-04) (hash(x)=27383319)
653
+ 625 train 7.673409 (lr=4.6848e-04) (hash(x)=26132276)
654
+ 626 train 7.610291 (lr=4.6690e-04) (hash(x)=19810497)
655
+ 627 train 7.629117 (lr=4.6532e-04) (hash(x)=25704919)
656
+ 628 train 7.830702 (lr=4.6374e-04) (hash(x)=27174264)
657
+ 629 train 7.980893 (lr=4.6216e-04) (hash(x)=22280814)
658
+ 630 train 8.653076 (lr=4.6058e-04) (hash(x)=24279448)
659
+ 631 train 8.154045 (lr=4.5900e-04) (hash(x)=23054940)
660
+ 632 train 8.220322 (lr=4.5742e-04) (hash(x)=23942400)
661
+ 633 train 8.435189 (lr=4.5583e-04) (hash(x)=24712416)
662
+ 634 train 8.068440 (lr=4.5425e-04) (hash(x)=24564658)
663
+ 635 train 7.803092 (lr=4.5267e-04) (hash(x)=24909904)
664
+ 636 train 7.847820 (lr=4.5109e-04) (hash(x)=23049534)
665
+ 637 train 7.888532 (lr=4.4950e-04) (hash(x)=24321591)
666
+ 638 train 7.675288 (lr=4.4792e-04) (hash(x)=26153298)
667
+ 639 train 7.712720 (lr=4.4634e-04) (hash(x)=27140757)
668
+ 640 train 7.474725 (lr=4.4475e-04) (hash(x)=25115907)
669
+ 641 train 7.449550 (lr=4.4317e-04) (hash(x)=26563770)
670
+ 642 train 7.659669 (lr=4.4158e-04) (hash(x)=28089252)
671
+ 643 train 7.596198 (lr=4.4000e-04) (hash(x)=25993110)
672
+ 644 train 7.583232 (lr=4.3842e-04) (hash(x)=25150008)
673
+ 645 train 7.828732 (lr=4.3683e-04) (hash(x)=28520222)
674
+ 646 train 7.465116 (lr=4.3525e-04) (hash(x)=21349943)
675
+ 647 train 7.508747 (lr=4.3366e-04) (hash(x)=25149419)
676
+ 648 train 7.753227 (lr=4.3208e-04) (hash(x)=25730641)
677
+ 649 train 7.856267 (lr=4.3050e-04) (hash(x)=26112813)
678
+ 650 val loss 7.6305
679
+ 650 val perplexity 2060.1296
680
+ 650 train 7.676654 (lr=4.2891e-04) (hash(x)=25907805)
681
+ 651 train 7.702614 (lr=4.2733e-04) (hash(x)=27623643)
682
+ 652 train 7.642177 (lr=4.2575e-04) (hash(x)=26484959)
683
+ 653 train 7.732927 (lr=4.2417e-04) (hash(x)=29199854)
684
+ 654 train 7.721803 (lr=4.2258e-04) (hash(x)=28369628)
685
+ 655 train 7.532369 (lr=4.2100e-04) (hash(x)=24727764)
686
+ 656 train 7.292552 (lr=4.1942e-04) (hash(x)=22610673)
687
+ 657 train 7.214331 (lr=4.1784e-04) (hash(x)=22667179)
688
+ 658 train 7.035042 (lr=4.1626e-04) (hash(x)=18477300)
689
+ 659 train 7.461417 (lr=4.1468e-04) (hash(x)=23155773)
690
+ 660 train 7.243999 (lr=4.1310e-04) (hash(x)=19461032)
691
+ 661 train 7.514318 (lr=4.1152e-04) (hash(x)=23453788)
692
+ 662 train 7.610628 (lr=4.0994e-04) (hash(x)=24543466)
693
+ 663 train 7.413083 (lr=4.0836e-04) (hash(x)=21935931)
694
+ 664 train 7.258331 (lr=4.0678e-04) (hash(x)=19910292)
695
+ 665 train 7.535789 (lr=4.0521e-04) (hash(x)=24481079)
696
+ 666 train 7.240422 (lr=4.0363e-04) (hash(x)=18922411)
697
+ 667 train 7.284522 (lr=4.0205e-04) (hash(x)=20054917)
698
+ 668 train 7.520946 (lr=4.0048e-04) (hash(x)=24850470)
699
+ 669 train 7.528824 (lr=3.9891e-04) (hash(x)=25907741)
700
+ 670 train 7.683520 (lr=3.9733e-04) (hash(x)=26873522)
701
+ 671 train 7.734401 (lr=3.9576e-04) (hash(x)=27606073)
702
+ 672 train 7.851099 (lr=3.9419e-04) (hash(x)=26209645)
703
+ 673 train 7.634325 (lr=3.9262e-04) (hash(x)=25202001)
704
+ 674 train 7.724946 (lr=3.9105e-04) (hash(x)=25569462)
705
+ 675 train 7.681394 (lr=3.8948e-04) (hash(x)=26534487)
706
+ 676 train 7.664291 (lr=3.8791e-04) (hash(x)=26455057)
707
+ 677 train 7.543324 (lr=3.8634e-04) (hash(x)=24095850)
708
+ 678 train 7.786176 (lr=3.8478e-04) (hash(x)=25287752)
709
+ 679 train 7.278767 (lr=3.8321e-04) (hash(x)=22450341)
710
+ 680 train 7.995637 (lr=3.8165e-04) (hash(x)=29004853)
711
+ 681 train 7.693821 (lr=3.8009e-04) (hash(x)=27993763)
712
+ 682 train 7.520468 (lr=3.7853e-04) (hash(x)=26382658)
713
+ 683 train 7.663476 (lr=3.7697e-04) (hash(x)=25013073)
714
+ 684 train 7.712439 (lr=3.7541e-04) (hash(x)=30595809)
715
+ 685 train 7.773958 (lr=3.7385e-04) (hash(x)=30934371)
716
+ 686 train 8.373900 (lr=3.7229e-04) (hash(x)=33060834)
717
+ 687 train 7.618799 (lr=3.7074e-04) (hash(x)=25945859)
718
+ 688 train 7.495952 (lr=3.6918e-04) (hash(x)=23375678)
719
+ 689 train 7.511031 (lr=3.6763e-04) (hash(x)=25218689)
720
+ 690 train 7.655021 (lr=3.6608e-04) (hash(x)=28127397)
721
+ 691 train 7.524625 (lr=3.6453e-04) (hash(x)=24418091)
722
+ 692 train 7.451604 (lr=3.6298e-04) (hash(x)=22761099)
723
+ 693 train 7.497445 (lr=3.6144e-04) (hash(x)=24615466)
724
+ 694 train 7.654570 (lr=3.5989e-04) (hash(x)=24699240)
725
+ 695 train 7.518694 (lr=3.5835e-04) (hash(x)=21613707)
726
+ 696 train 7.588291 (lr=3.5681e-04) (hash(x)=24977554)
727
+ 697 train 7.527724 (lr=3.5527e-04) (hash(x)=24348175)
728
+ 698 train 7.753681 (lr=3.5373e-04) (hash(x)=25102767)
729
+ 699 train 7.697758 (lr=3.5219e-04) (hash(x)=26386157)
730
+ 700 val loss 7.5832
731
+ 700 val perplexity 1964.9421
732
+ 700 train 7.588155 (lr=3.5066e-04) (hash(x)=26423460)
733
+ 701 train 7.520961 (lr=3.4912e-04) (hash(x)=25771047)
734
+ 702 train 7.631700 (lr=3.4759e-04) (hash(x)=26999875)
735
+ 703 train 7.545857 (lr=3.4606e-04) (hash(x)=24396519)
736
+ 704 train 7.551044 (lr=3.4453e-04) (hash(x)=22588122)
737
+ 705 train 7.859184 (lr=3.4301e-04) (hash(x)=25142399)
738
+ 706 train 7.604138 (lr=3.4148e-04) (hash(x)=20440214)
739
+ 707 train 7.437064 (lr=3.3996e-04) (hash(x)=23265507)
740
+ 708 train 7.556749 (lr=3.3844e-04) (hash(x)=24563470)
741
+ 709 train 7.479530 (lr=3.3692e-04) (hash(x)=22514858)
742
+ 710 train 7.815952 (lr=3.3540e-04) (hash(x)=26691212)
743
+ 711 train 7.836912 (lr=3.3389e-04) (hash(x)=29138828)
744
+ 712 train 7.809328 (lr=3.3238e-04) (hash(x)=28028528)
745
+ 713 train 7.688264 (lr=3.3087e-04) (hash(x)=20531210)
746
+ 714 train 7.558179 (lr=3.2936e-04) (hash(x)=25075352)
747
+ 715 train 7.589893 (lr=3.2785e-04) (hash(x)=24265353)
748
+ 716 train 7.705373 (lr=3.2635e-04) (hash(x)=24635726)
749
+ 717 train 7.545704 (lr=3.2484e-04) (hash(x)=24999726)
750
+ 718 train 7.738439 (lr=3.2334e-04) (hash(x)=27412910)
751
+ 719 train 7.521271 (lr=3.2185e-04) (hash(x)=24685515)
752
+ 720 train 7.613613 (lr=3.2035e-04) (hash(x)=23780329)
753
+ 721 train 7.413539 (lr=3.1886e-04) (hash(x)=25071701)
754
+ 722 train 7.577625 (lr=3.1737e-04) (hash(x)=23767130)
755
+ 723 train 7.536301 (lr=3.1588e-04) (hash(x)=24876269)
756
+ 724 train 7.647758 (lr=3.1440e-04) (hash(x)=26405773)
757
+ 725 train 8.056955 (lr=3.1291e-04) (hash(x)=31733180)
758
+ 726 train 7.257443 (lr=3.1143e-04) (hash(x)=21337509)
759
+ 727 train 7.359027 (lr=3.0995e-04) (hash(x)=22825749)
760
+ 728 train 7.765396 (lr=3.0848e-04) (hash(x)=28638695)
761
+ 729 train 7.631400 (lr=3.0700e-04) (hash(x)=26393943)
762
+ 730 train 7.791526 (lr=3.0553e-04) (hash(x)=27563583)
763
+ 731 train 7.371138 (lr=3.0407e-04) (hash(x)=21239652)
764
+ 732 train 7.454707 (lr=3.0260e-04) (hash(x)=23986428)
765
+ 733 train 7.356722 (lr=3.0114e-04) (hash(x)=24943881)
766
+ 734 train 7.903755 (lr=2.9968e-04) (hash(x)=29691448)
767
+ 735 train 7.845333 (lr=2.9822e-04) (hash(x)=28767869)
768
+ 736 train 7.532405 (lr=2.9677e-04) (hash(x)=23628188)
769
+ 737 train 7.869692 (lr=2.9531e-04) (hash(x)=29341482)
770
+ 738 train 7.657808 (lr=2.9386e-04) (hash(x)=30336570)
771
+ 739 train 7.608553 (lr=2.9242e-04) (hash(x)=25614301)
772
+ 740 train 7.488235 (lr=2.9097e-04) (hash(x)=24160500)
773
+ 741 train 7.674870 (lr=2.8953e-04) (hash(x)=26030058)
774
+ 742 train 7.990932 (lr=2.8810e-04) (hash(x)=29243936)
775
+ 743 train 7.393469 (lr=2.8666e-04) (hash(x)=21159060)
776
+ 744 train 7.478491 (lr=2.8523e-04) (hash(x)=23701853)
777
+ 745 train 7.634654 (lr=2.8380e-04) (hash(x)=24629937)
778
+ 746 train 7.551548 (lr=2.8238e-04) (hash(x)=25110108)
779
+ 747 train 7.613504 (lr=2.8095e-04) (hash(x)=26751788)
780
+ 748 train 7.676363 (lr=2.7953e-04) (hash(x)=26430427)
781
+ 749 train 7.550113 (lr=2.7812e-04) (hash(x)=26012353)
782
+ 750 val loss 7.5679
783
+ 750 val perplexity 1934.9924
784
+ 750 train 7.643803 (lr=2.7670e-04) (hash(x)=22735910)
785
+ 751 train 7.461554 (lr=2.7529e-04) (hash(x)=25045397)
786
+ 752 train 7.180844 (lr=2.7389e-04) (hash(x)=21554427)
787
+ 753 train 7.289801 (lr=2.7248e-04) (hash(x)=23751143)
788
+ 754 train 8.062931 (lr=2.7108e-04) (hash(x)=28602273)
789
+ 755 train 8.096751 (lr=2.6969e-04) (hash(x)=29989709)
790
+ 756 train 7.494996 (lr=2.6829e-04) (hash(x)=22331648)
791
+ 757 train 7.899054 (lr=2.6690e-04) (hash(x)=31017246)
792
+ 758 train 7.836470 (lr=2.6551e-04) (hash(x)=29709045)
793
+ 759 train 7.863954 (lr=2.6413e-04) (hash(x)=25560928)
794
+ 760 train 7.568710 (lr=2.6275e-04) (hash(x)=25075464)
795
+ 761 train 7.656059 (lr=2.6137e-04) (hash(x)=27352253)
796
+ 762 train 7.839081 (lr=2.6000e-04) (hash(x)=28187891)
797
+ 763 train 7.709825 (lr=2.5863e-04) (hash(x)=26062687)
798
+ 764 train 7.690404 (lr=2.5726e-04) (hash(x)=27427811)
799
+ 765 train 7.760766 (lr=2.5590e-04) (hash(x)=27614522)
800
+ 766 train 7.558891 (lr=2.5454e-04) (hash(x)=26129544)
801
+ 767 train 8.104906 (lr=2.5319e-04) (hash(x)=28959222)
802
+ 768 train 7.677136 (lr=2.5183e-04) (hash(x)=26860067)
803
+ 769 train 7.485062 (lr=2.5048e-04) (hash(x)=25122598)
804
+ 770 train 7.561226 (lr=2.4914e-04) (hash(x)=25245030)
805
+ 771 train 7.561200 (lr=2.4780e-04) (hash(x)=25434884)
806
+ 772 train 7.641778 (lr=2.4646e-04) (hash(x)=27732790)
807
+ 773 train 7.650968 (lr=2.4513e-04) (hash(x)=27824438)
808
+ 774 train 7.777805 (lr=2.4380e-04) (hash(x)=27201953)
809
+ 775 train 7.484239 (lr=2.4247e-04) (hash(x)=21248405)
810
+ 776 train 7.331871 (lr=2.4115e-04) (hash(x)=22805934)
811
+ 777 train 7.625768 (lr=2.3983e-04) (hash(x)=26482588)
812
+ 778 train 7.605461 (lr=2.3852e-04) (hash(x)=24153691)
813
+ 779 train 7.737486 (lr=2.3720e-04) (hash(x)=25044192)
814
+ 780 train 7.677638 (lr=2.3590e-04) (hash(x)=25910078)
815
+ 781 train 7.695659 (lr=2.3460e-04) (hash(x)=28645524)
816
+ 782 train 7.517026 (lr=2.3330e-04) (hash(x)=24368498)
817
+ 783 train 7.571799 (lr=2.3200e-04) (hash(x)=25830182)
818
+ 784 train 7.686057 (lr=2.3071e-04) (hash(x)=29181807)
819
+ 785 train 7.613077 (lr=2.2942e-04) (hash(x)=25585137)
820
+ 786 train 7.519020 (lr=2.2814e-04) (hash(x)=24798246)
821
+ 787 train 7.660114 (lr=2.2686e-04) (hash(x)=26621419)
822
+ 788 train 7.202465 (lr=2.2559e-04) (hash(x)=21446891)
823
+ 789 train 7.181139 (lr=2.2432e-04) (hash(x)=22165286)
824
+ 790 train 7.378574 (lr=2.2305e-04) (hash(x)=23477219)
825
+ 791 train 7.668839 (lr=2.2179e-04) (hash(x)=25173113)
826
+ 792 train 7.542339 (lr=2.2053e-04) (hash(x)=25853788)
827
+ 793 train 7.650847 (lr=2.1928e-04) (hash(x)=27267091)
828
+ 794 train 7.490808 (lr=2.1803e-04) (hash(x)=23743694)
829
+ 795 train 7.473039 (lr=2.1678e-04) (hash(x)=24400133)
830
+ 796 train 7.543204 (lr=2.1554e-04) (hash(x)=23663639)
831
+ 797 train 7.418899 (lr=2.1431e-04) (hash(x)=23103223)
832
+ 798 train 7.896310 (lr=2.1308e-04) (hash(x)=28748411)
833
+ 799 train 7.256450 (lr=2.1185e-04) (hash(x)=23486277)
834
+ 800 val loss 7.5516
835
+ 800 val perplexity 1903.8044
836
+ 800 train 7.507089 (lr=2.1063e-04) (hash(x)=25678518)
837
+ 801 train 7.476920 (lr=2.0941e-04) (hash(x)=23421286)
838
+ 802 train 7.616274 (lr=2.0819e-04) (hash(x)=26054104)
839
+ 803 train 7.621419 (lr=2.0698e-04) (hash(x)=25978130)
840
+ 804 train 7.642684 (lr=2.0578e-04) (hash(x)=26006525)
841
+ 805 train 7.586324 (lr=2.0458e-04) (hash(x)=25769432)
842
+ 806 train 7.399048 (lr=2.0338e-04) (hash(x)=22430795)
843
+ 807 train 7.765576 (lr=2.0219e-04) (hash(x)=28916006)
844
+ 808 train 7.569942 (lr=2.0100e-04) (hash(x)=25166800)
845
+ 809 train 7.668936 (lr=1.9982e-04) (hash(x)=24226056)
846
+ 810 train 7.522010 (lr=1.9864e-04) (hash(x)=25057934)
847
+ 811 train 8.287386 (lr=1.9747e-04) (hash(x)=33195100)
848
+ 812 train 7.717071 (lr=1.9630e-04) (hash(x)=26312888)
849
+ 813 train 7.867256 (lr=1.9514e-04) (hash(x)=27730410)
850
+ 814 train 7.742851 (lr=1.9398e-04) (hash(x)=27372474)
851
+ 815 train 7.580007 (lr=1.9283e-04) (hash(x)=25556929)
852
+ 816 train 7.656108 (lr=1.9168e-04) (hash(x)=26909985)
853
+ 817 train 7.494391 (lr=1.9053e-04) (hash(x)=25991247)
854
+ 818 train 7.692978 (lr=1.8939e-04) (hash(x)=27438141)
855
+ 819 train 7.825996 (lr=1.8826e-04) (hash(x)=29536986)
856
+ 820 train 7.598423 (lr=1.8713e-04) (hash(x)=24478391)
857
+ 821 train 7.570776 (lr=1.8600e-04) (hash(x)=26125216)
858
+ 822 train 7.581394 (lr=1.8488e-04) (hash(x)=26422130)
859
+ 823 train 7.787670 (lr=1.8377e-04) (hash(x)=29648798)
860
+ 824 train 7.493511 (lr=1.8266e-04) (hash(x)=21247770)
861
+ 825 train 7.601787 (lr=1.8155e-04) (hash(x)=23195388)
862
+ 826 train 7.538207 (lr=1.8045e-04) (hash(x)=25796725)
863
+ 827 train 7.563684 (lr=1.7936e-04) (hash(x)=23124767)
864
+ 828 train 7.603641 (lr=1.7827e-04) (hash(x)=25233464)
865
+ 829 train 7.582232 (lr=1.7718e-04) (hash(x)=25713275)
866
+ 830 train 7.523696 (lr=1.7610e-04) (hash(x)=25550167)
867
+ 831 train 7.440241 (lr=1.7503e-04) (hash(x)=24976217)
868
+ 832 train 7.725430 (lr=1.7396e-04) (hash(x)=28536827)
869
+ 833 train 7.645870 (lr=1.7289e-04) (hash(x)=27500801)
870
+ 834 train 7.464616 (lr=1.7183e-04) (hash(x)=25545765)
871
+ 835 train 7.502928 (lr=1.7078e-04) (hash(x)=23632825)
872
+ 836 train 7.534663 (lr=1.6973e-04) (hash(x)=25708009)
873
+ 837 train 7.577198 (lr=1.6869e-04) (hash(x)=24456276)
874
+ 838 train 7.662659 (lr=1.6765e-04) (hash(x)=29189855)
875
+ 839 train 7.896054 (lr=1.6661e-04) (hash(x)=31019606)
876
+ 840 train 7.765890 (lr=1.6559e-04) (hash(x)=26328013)
877
+ 841 train 7.682037 (lr=1.6456e-04) (hash(x)=25027904)
878
+ 842 train 7.475315 (lr=1.6355e-04) (hash(x)=23734189)
879
+ 843 train 7.731221 (lr=1.6253e-04) (hash(x)=28236580)
880
+ 844 train 7.520881 (lr=1.6153e-04) (hash(x)=26509780)
881
+ 845 train 7.566965 (lr=1.6053e-04) (hash(x)=25386473)
882
+ 846 train 7.527994 (lr=1.5953e-04) (hash(x)=24052671)
883
+ 847 train 7.706694 (lr=1.5854e-04) (hash(x)=28269421)
884
+ 848 train 7.165405 (lr=1.5756e-04) (hash(x)=22251724)
885
+ 849 train 7.424680 (lr=1.5658e-04) (hash(x)=24308447)
886
+ 850 val loss 7.5386
887
+ 850 val perplexity 1879.1919
888
+ 850 train 7.410348 (lr=1.5560e-04) (hash(x)=24242830)
889
+ 851 train 7.589027 (lr=1.5463e-04) (hash(x)=25563279)
890
+ 852 train 7.581572 (lr=1.5367e-04) (hash(x)=26354481)
891
+ 853 train 7.723355 (lr=1.5271e-04) (hash(x)=26152637)
892
+ 854 train 7.748899 (lr=1.5176e-04) (hash(x)=28051025)
893
+ 855 train 7.566444 (lr=1.5082e-04) (hash(x)=24865358)
894
+ 856 train 7.460689 (lr=1.4988e-04) (hash(x)=24288911)
895
+ 857 train 7.358409 (lr=1.4894e-04) (hash(x)=22230964)
896
+ 858 train 7.262041 (lr=1.4801e-04) (hash(x)=21303832)
897
+ 859 train 7.397797 (lr=1.4709e-04) (hash(x)=22155546)
898
+ 860 train 7.528834 (lr=1.4617e-04) (hash(x)=25296428)
899
+ 861 train 7.700087 (lr=1.4526e-04) (hash(x)=29142319)
900
+ 862 train 7.501176 (lr=1.4435e-04) (hash(x)=25545430)
901
+ 863 train 7.315697 (lr=1.4345e-04) (hash(x)=26984272)
902
+ 864 train 7.512764 (lr=1.4255e-04) (hash(x)=25429005)
903
+ 865 train 7.563079 (lr=1.4166e-04) (hash(x)=27077032)
904
+ 866 train 7.575536 (lr=1.4078e-04) (hash(x)=26494424)
905
+ 867 train 7.564958 (lr=1.3990e-04) (hash(x)=23193673)
906
+ 868 train 7.598187 (lr=1.3903e-04) (hash(x)=25075134)
907
+ 869 train 7.610200 (lr=1.3816e-04) (hash(x)=27112558)
908
+ 870 train 7.753952 (lr=1.3730e-04) (hash(x)=27436608)
909
+ 871 train 7.479683 (lr=1.3645e-04) (hash(x)=24544116)
910
+ 872 train 7.852112 (lr=1.3560e-04) (hash(x)=31632686)
911
+ 873 train 7.498715 (lr=1.3476e-04) (hash(x)=25890184)
912
+ 874 train 7.638007 (lr=1.3392e-04) (hash(x)=22887555)
913
+ 875 train 7.483808 (lr=1.3309e-04) (hash(x)=24547533)
914
+ 876 train 7.612001 (lr=1.3227e-04) (hash(x)=26553496)
915
+ 877 train 7.718784 (lr=1.3145e-04) (hash(x)=27467688)
916
+ 878 train 7.466613 (lr=1.3063e-04) (hash(x)=24766934)
917
+ 879 train 7.371673 (lr=1.2983e-04) (hash(x)=22059850)
918
+ 880 train 7.719420 (lr=1.2903e-04) (hash(x)=22871702)
919
+ 881 train 7.573467 (lr=1.2823e-04) (hash(x)=23893130)
920
+ 882 train 7.616336 (lr=1.2744e-04) (hash(x)=25125691)
921
+ 883 train 7.550086 (lr=1.2666e-04) (hash(x)=25994573)
922
+ 884 train 7.601233 (lr=1.2588e-04) (hash(x)=26076345)
923
+ 885 train 7.535115 (lr=1.2511e-04) (hash(x)=26577783)
924
+ 886 train 7.655323 (lr=1.2435e-04) (hash(x)=27395225)
925
+ 887 train 7.478864 (lr=1.2359e-04) (hash(x)=23926632)
926
+ 888 train 7.337152 (lr=1.2284e-04) (hash(x)=21737239)
927
+ 889 train 7.599586 (lr=1.2209e-04) (hash(x)=23574207)
928
+ 890 train 7.301548 (lr=1.2135e-04) (hash(x)=24365231)
929
+ 891 train 7.478979 (lr=1.2062e-04) (hash(x)=27111369)
930
+ 892 train 7.726985 (lr=1.1989e-04) (hash(x)=27290015)
931
+ 893 train 7.525814 (lr=1.1917e-04) (hash(x)=23979820)
932
+ 894 train 7.623068 (lr=1.1845e-04) (hash(x)=26450121)
933
+ 895 train 7.531061 (lr=1.1774e-04) (hash(x)=27025333)
934
+ 896 train 7.507591 (lr=1.1704e-04) (hash(x)=23624605)
935
+ 897 train 7.313787 (lr=1.1634e-04) (hash(x)=22846386)
936
+ 898 train 7.398183 (lr=1.1565e-04) (hash(x)=22970561)
937
+ 899 train 7.256033 (lr=1.1497e-04) (hash(x)=16908068)
938
+ 900 val loss 7.5211
939
+ 900 val perplexity 1846.5104
940
+ 900 train 7.473626 (lr=1.1429e-04) (hash(x)=24661446)
941
+ 901 train 7.582211 (lr=1.1362e-04) (hash(x)=25664727)
942
+ 902 train 7.559648 (lr=1.1295e-04) (hash(x)=25667011)
943
+ 903 train 7.602252 (lr=1.1229e-04) (hash(x)=29120407)
944
+ 904 train 7.661061 (lr=1.1164e-04) (hash(x)=23385735)
945
+ 905 train 7.673601 (lr=1.1099e-04) (hash(x)=25564213)
946
+ 906 train 7.710779 (lr=1.1036e-04) (hash(x)=25413898)
947
+ 907 train 7.687609 (lr=1.0972e-04) (hash(x)=27092710)
948
+ 908 train 7.464441 (lr=1.0909e-04) (hash(x)=25789923)
949
+ 909 train 7.567976 (lr=1.0847e-04) (hash(x)=28533197)
950
+ 910 train 7.427011 (lr=1.0786e-04) (hash(x)=22982996)
951
+ 911 train 7.524367 (lr=1.0725e-04) (hash(x)=23827393)
952
+ 912 train 7.165169 (lr=1.0665e-04) (hash(x)=21242640)
953
+ 913 train 7.550535 (lr=1.0606e-04) (hash(x)=24154233)
954
+ 914 train 7.526055 (lr=1.0547e-04) (hash(x)=24331967)
955
+ 915 train 7.777141 (lr=1.0489e-04) (hash(x)=32812727)
956
+ 916 train 7.432430 (lr=1.0431e-04) (hash(x)=23572994)
957
+ 917 train 7.477477 (lr=1.0374e-04) (hash(x)=26305435)
958
+ 918 train 7.498130 (lr=1.0318e-04) (hash(x)=26268355)
959
+ 919 train 7.676999 (lr=1.0262e-04) (hash(x)=27230027)
960
+ 920 train 7.478352 (lr=1.0207e-04) (hash(x)=23885377)
961
+ 921 train 7.428955 (lr=1.0153e-04) (hash(x)=23532437)
962
+ 922 train 7.514291 (lr=1.0099e-04) (hash(x)=25577034)
963
+ 923 train 7.659935 (lr=1.0046e-04) (hash(x)=25703381)
964
+ 924 train 7.636822 (lr=9.9941e-05) (hash(x)=27113866)
965
+ 925 train 7.464108 (lr=9.9425e-05) (hash(x)=26961429)
966
+ 926 train 7.287851 (lr=9.8915e-05) (hash(x)=21355372)
967
+ 927 train 7.450452 (lr=9.8411e-05) (hash(x)=24968260)
968
+ 928 train 7.436380 (lr=9.7915e-05) (hash(x)=25357517)
969
+ 929 train 7.510117 (lr=9.7424e-05) (hash(x)=24854265)
970
+ 930 train 7.266597 (lr=9.6941e-05) (hash(x)=21102770)
971
+ 931 train 7.600950 (lr=9.6464e-05) (hash(x)=25676468)
972
+ 932 train 7.481898 (lr=9.5994e-05) (hash(x)=22809869)
973
+ 933 train 7.591363 (lr=9.5530e-05) (hash(x)=25503865)
974
+ 934 train 7.452654 (lr=9.5073e-05) (hash(x)=24853995)
975
+ 935 train 7.622903 (lr=9.4623e-05) (hash(x)=27544803)
976
+ 936 train 7.653581 (lr=9.4180e-05) (hash(x)=25981933)
977
+ 937 train 7.401806 (lr=9.3743e-05) (hash(x)=24658683)
978
+ 938 train 7.579965 (lr=9.3313e-05) (hash(x)=23855201)
979
+ 939 train 7.513447 (lr=9.2889e-05) (hash(x)=24331407)
980
+ 940 train 7.707307 (lr=9.2473e-05) (hash(x)=29265551)
981
+ 941 train 7.424454 (lr=9.2063e-05) (hash(x)=21892556)
982
+ 942 train 7.468751 (lr=9.1659e-05) (hash(x)=27183405)
983
+ 943 train 7.423522 (lr=9.1263e-05) (hash(x)=26540663)
984
+ 944 train 7.404305 (lr=9.0873e-05) (hash(x)=25718393)
985
+ 945 train 7.546010 (lr=9.0490e-05) (hash(x)=26819462)
986
+ 946 train 7.583229 (lr=9.0114e-05) (hash(x)=27427540)
987
+ 947 train 7.540302 (lr=8.9744e-05) (hash(x)=25532657)
988
+ 948 train 7.737018 (lr=8.9382e-05) (hash(x)=27641372)
989
+ 949 train 7.491671 (lr=8.9026e-05) (hash(x)=26515570)
990
+ 950 val loss 7.5113
991
+ 950 val perplexity 1828.6364
992
+ 950 train 7.683830 (lr=8.8677e-05) (hash(x)=26911957)
993
+ 951 train 7.804724 (lr=8.8335e-05) (hash(x)=25856625)
994
+ 952 train 7.377818 (lr=8.7999e-05) (hash(x)=25219129)
995
+ 953 train 7.351259 (lr=8.7670e-05) (hash(x)=25260471)
996
+ 954 train 7.702522 (lr=8.7349e-05) (hash(x)=29373370)
997
+ 955 train 7.283093 (lr=8.7034e-05) (hash(x)=23437426)
998
+ 956 train 7.621381 (lr=8.6725e-05) (hash(x)=23769521)
999
+ 957 train 7.641384 (lr=8.6424e-05) (hash(x)=25961833)
1000
+ 958 train 7.352495 (lr=8.6130e-05) (hash(x)=23582666)
1001
+ 959 train 7.452064 (lr=8.5842e-05) (hash(x)=23164356)
1002
+ 960 train 7.419547 (lr=8.5561e-05) (hash(x)=24443114)
1003
+ 961 train 7.476512 (lr=8.5287e-05) (hash(x)=25052665)
1004
+ 962 train 7.599847 (lr=8.5020e-05) (hash(x)=27802272)
1005
+ 963 train 7.471274 (lr=8.4760e-05) (hash(x)=25957896)
1006
+ 964 train 7.498709 (lr=8.4507e-05) (hash(x)=26737251)
1007
+ 965 train 7.629146 (lr=8.4260e-05) (hash(x)=24723263)
1008
+ 966 train 7.574878 (lr=8.4021e-05) (hash(x)=24707011)
1009
+ 967 train 7.548302 (lr=8.3788e-05) (hash(x)=25646282)
1010
+ 968 train 7.564161 (lr=8.3563e-05) (hash(x)=27544665)
1011
+ 969 train 7.478656 (lr=8.3344e-05) (hash(x)=25851993)
1012
+ 970 train 7.770664 (lr=8.3132e-05) (hash(x)=29059700)
1013
+ 971 train 7.495431 (lr=8.2927e-05) (hash(x)=21513584)
1014
+ 972 train 7.268135 (lr=8.2729e-05) (hash(x)=23151267)
1015
+ 973 train 7.523670 (lr=8.2537e-05) (hash(x)=26017176)
1016
+ 974 train 7.591654 (lr=8.2353e-05) (hash(x)=26979518)
1017
+ 975 train 7.560774 (lr=8.2176e-05) (hash(x)=23843233)
1018
+ 976 train 7.718054 (lr=8.2005e-05) (hash(x)=24193010)
1019
+ 977 train 7.472030 (lr=8.1842e-05) (hash(x)=21476847)
1020
+ 978 train 7.370245 (lr=8.1685e-05) (hash(x)=21366504)
1021
+ 979 train 7.414275 (lr=8.1536e-05) (hash(x)=23226697)
1022
+ 980 train 7.215604 (lr=8.1393e-05) (hash(x)=19961773)
1023
+ 981 train 7.114618 (lr=8.1257e-05) (hash(x)=19772969)
1024
+ 982 train 7.378058 (lr=8.1128e-05) (hash(x)=23110142)
1025
+ 983 train 7.614204 (lr=8.1007e-05) (hash(x)=24506028)
1026
+ 984 train 7.593784 (lr=8.0892e-05) (hash(x)=25480731)
1027
+ 985 train 7.438121 (lr=8.0784e-05) (hash(x)=21077417)
1028
+ 986 train 7.418439 (lr=8.0683e-05) (hash(x)=23686713)
1029
+ 987 train 7.512817 (lr=8.0589e-05) (hash(x)=26024321)
1030
+ 988 train 7.565305 (lr=8.0502e-05) (hash(x)=27424109)
1031
+ 989 train 7.659481 (lr=8.0422e-05) (hash(x)=27786174)
1032
+ 990 train 7.483014 (lr=8.0348e-05) (hash(x)=25232502)
1033
+ 991 train 7.474936 (lr=8.0282e-05) (hash(x)=22781277)
1034
+ 992 train 7.551094 (lr=8.0223e-05) (hash(x)=26184527)
1035
+ 993 train 7.544542 (lr=8.0171e-05) (hash(x)=24459895)
1036
+ 994 train 7.525168 (lr=8.0125e-05) (hash(x)=25244624)
1037
+ 995 train 7.617669 (lr=8.0087e-05) (hash(x)=24451843)
1038
+ 996 train 7.609920 (lr=8.0056e-05) (hash(x)=22129897)
1039
+ 997 train 7.447691 (lr=8.0031e-05) (hash(x)=21116390)
1040
+ 998 train 7.253922 (lr=8.0014e-05) (hash(x)=20650070)
1041
+ 999 val loss 7.5030
1042
+ 999 val perplexity 1813.4725
1043
+ 999 train 7.592977 (lr=8.0003e-05) (hash(x)=24948650)
lr8e-4_total_batch_size5120_seq_len512/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61c714b236ad209c486905a583d06d2b364cf21e22a930ea39c8f661b89c947b
3
+ size 48156226
lr8e-4_total_batch_size5120_seq_len512/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:986c6277022e16ddb16342b11cf0e2d3dec51dfbb4fdde4ecb58584ddd8d9b99
3
+ size 71157574