andrew-healey commited on
Commit
3073177
·
verified ·
1 Parent(s): 89f9b7f

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size5120_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size5120_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", "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": 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": 0.0001, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1e-4_total_batch_size5120_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr1e-4_total_batch_size5120_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 11.683437 (lr=3.4965e-07) (hash(x)=6154740)
5
+ 1 train 11.701447 (lr=6.9930e-07) (hash(x)=6605789)
6
+ 2 train 11.709952 (lr=1.0490e-06) (hash(x)=6696707)
7
+ 3 train 11.728244 (lr=1.3986e-06) (hash(x)=5113283)
8
+ 4 train 11.673847 (lr=1.7483e-06) (hash(x)=5607069)
9
+ 5 train 11.670259 (lr=2.0979e-06) (hash(x)=7007748)
10
+ 6 train 11.695046 (lr=2.4476e-06) (hash(x)=7643299)
11
+ 7 train 11.662629 (lr=2.7972e-06) (hash(x)=5534142)
12
+ 8 train 11.666996 (lr=3.1469e-06) (hash(x)=6539293)
13
+ 9 train 11.685373 (lr=3.4965e-06) (hash(x)=5974274)
14
+ 10 train 11.685617 (lr=3.8462e-06) (hash(x)=7199085)
15
+ 11 train 11.642393 (lr=4.1958e-06) (hash(x)=6763231)
16
+ 12 train 11.633266 (lr=4.5455e-06) (hash(x)=6720398)
17
+ 13 train 11.602768 (lr=4.8951e-06) (hash(x)=6706513)
18
+ 14 train 11.632250 (lr=5.2448e-06) (hash(x)=5727925)
19
+ 15 train 11.626501 (lr=5.5944e-06) (hash(x)=6328012)
20
+ 16 train 11.565728 (lr=5.9441e-06) (hash(x)=5812722)
21
+ 17 train 11.569908 (lr=6.2937e-06) (hash(x)=6183012)
22
+ 18 train 11.581360 (lr=6.6434e-06) (hash(x)=5952539)
23
+ 19 train 11.552618 (lr=6.9930e-06) (hash(x)=7052161)
24
+ 20 train 11.499270 (lr=7.3427e-06) (hash(x)=5810871)
25
+ 21 train 11.514858 (lr=7.6923e-06) (hash(x)=6586466)
26
+ 22 train 11.460441 (lr=8.0420e-06) (hash(x)=5061781)
27
+ 23 train 11.505838 (lr=8.3916e-06) (hash(x)=5556098)
28
+ 24 train 11.504732 (lr=8.7413e-06) (hash(x)=7268535)
29
+ 25 train 11.448575 (lr=9.0909e-06) (hash(x)=8155572)
30
+ 26 train 11.364543 (lr=9.4406e-06) (hash(x)=5019590)
31
+ 27 train 11.377727 (lr=9.7902e-06) (hash(x)=4696943)
32
+ 28 train 11.318115 (lr=1.0140e-05) (hash(x)=5984010)
33
+ 29 train 11.371592 (lr=1.0490e-05) (hash(x)=7649389)
34
+ 30 train 11.279446 (lr=1.0839e-05) (hash(x)=6332817)
35
+ 31 train 11.224400 (lr=1.1189e-05) (hash(x)=6219138)
36
+ 32 train 11.282112 (lr=1.1538e-05) (hash(x)=6895681)
37
+ 33 train 11.176207 (lr=1.1888e-05) (hash(x)=7075794)
38
+ 34 train 11.142393 (lr=1.2238e-05) (hash(x)=5024163)
39
+ 35 train 11.191124 (lr=1.2587e-05) (hash(x)=6049878)
40
+ 36 train 11.158174 (lr=1.2937e-05) (hash(x)=4774943)
41
+ 37 train 11.157990 (lr=1.3287e-05) (hash(x)=6385927)
42
+ 38 train 11.043331 (lr=1.3636e-05) (hash(x)=5070217)
43
+ 39 train 10.995816 (lr=1.3986e-05) (hash(x)=5491972)
44
+ 40 train 11.010164 (lr=1.4336e-05) (hash(x)=5695751)
45
+ 41 train 10.960633 (lr=1.4685e-05) (hash(x)=6061420)
46
+ 42 train 10.926920 (lr=1.5035e-05) (hash(x)=7836710)
47
+ 43 train 10.937121 (lr=1.5385e-05) (hash(x)=8075458)
48
+ 44 train 10.776705 (lr=1.5734e-05) (hash(x)=7190584)
49
+ 45 train 10.764411 (lr=1.6084e-05) (hash(x)=7238769)
50
+ 46 train 10.938816 (lr=1.6434e-05) (hash(x)=4931205)
51
+ 47 train 10.744882 (lr=1.6783e-05) (hash(x)=5917741)
52
+ 48 train 10.701169 (lr=1.7133e-05) (hash(x)=5530785)
53
+ 49 train 10.761693 (lr=1.7483e-05) (hash(x)=5430605)
54
+ 50 val loss 10.7112
55
+ 50 val perplexity 44854.2266
56
+ 50 train 10.730789 (lr=1.7832e-05) (hash(x)=6056774)
57
+ 51 train 10.655834 (lr=1.8182e-05) (hash(x)=5750403)
58
+ 52 train 10.623432 (lr=1.8531e-05) (hash(x)=6187497)
59
+ 53 train 10.626368 (lr=1.8881e-05) (hash(x)=4792640)
60
+ 54 train 10.683257 (lr=1.9231e-05) (hash(x)=6868430)
61
+ 55 train 10.562780 (lr=1.9580e-05) (hash(x)=6112458)
62
+ 56 train 10.492647 (lr=1.9930e-05) (hash(x)=5523601)
63
+ 57 train 10.492303 (lr=2.0280e-05) (hash(x)=5801122)
64
+ 58 train 10.549990 (lr=2.0629e-05) (hash(x)=6826295)
65
+ 59 train 10.634004 (lr=2.0979e-05) (hash(x)=5806490)
66
+ 60 train 10.497989 (lr=2.1329e-05) (hash(x)=6223100)
67
+ 61 train 10.451726 (lr=2.1678e-05) (hash(x)=5754140)
68
+ 62 train 10.427583 (lr=2.2028e-05) (hash(x)=6675367)
69
+ 63 train 10.337934 (lr=2.2378e-05) (hash(x)=6568379)
70
+ 64 train 10.237189 (lr=2.2727e-05) (hash(x)=5648233)
71
+ 65 train 10.362542 (lr=2.3077e-05) (hash(x)=7134932)
72
+ 66 train 10.390593 (lr=2.3427e-05) (hash(x)=5242536)
73
+ 67 train 10.390552 (lr=2.3776e-05) (hash(x)=7212403)
74
+ 68 train 10.244838 (lr=2.4126e-05) (hash(x)=5342967)
75
+ 69 train 10.368650 (lr=2.4476e-05) (hash(x)=7720703)
76
+ 70 train 10.287737 (lr=2.4825e-05) (hash(x)=7351337)
77
+ 71 train 10.354877 (lr=2.5175e-05) (hash(x)=7386082)
78
+ 72 train 10.239516 (lr=2.5524e-05) (hash(x)=6931318)
79
+ 73 train 10.259445 (lr=2.5874e-05) (hash(x)=6848513)
80
+ 74 train 10.270724 (lr=2.6224e-05) (hash(x)=7714274)
81
+ 75 train 10.323357 (lr=2.6573e-05) (hash(x)=6359743)
82
+ 76 train 10.178976 (lr=2.6923e-05) (hash(x)=6801650)
83
+ 77 train 10.155712 (lr=2.7273e-05) (hash(x)=7536538)
84
+ 78 train 10.270412 (lr=2.7622e-05) (hash(x)=6957410)
85
+ 79 train 10.270091 (lr=2.7972e-05) (hash(x)=6151674)
86
+ 80 train 10.300437 (lr=2.8322e-05) (hash(x)=6033844)
87
+ 81 train 10.200167 (lr=2.8671e-05) (hash(x)=6317099)
88
+ 82 train 10.181614 (lr=2.9021e-05) (hash(x)=5357844)
89
+ 83 train 10.235994 (lr=2.9371e-05) (hash(x)=7168704)
90
+ 84 train 10.286210 (lr=2.9720e-05) (hash(x)=7157938)
91
+ 85 train 10.205201 (lr=3.0070e-05) (hash(x)=6736092)
92
+ 86 train 10.131433 (lr=3.0420e-05) (hash(x)=5383012)
93
+ 87 train 10.169333 (lr=3.0769e-05) (hash(x)=7859060)
94
+ 88 train 10.183014 (lr=3.1119e-05) (hash(x)=5869288)
95
+ 89 train 10.185976 (lr=3.1469e-05) (hash(x)=6958223)
96
+ 90 train 10.095232 (lr=3.1818e-05) (hash(x)=6731344)
97
+ 91 train 10.001469 (lr=3.2168e-05) (hash(x)=6976239)
98
+ 92 train 10.096917 (lr=3.2517e-05) (hash(x)=7769441)
99
+ 93 train 10.089683 (lr=3.2867e-05) (hash(x)=6956845)
100
+ 94 train 10.241543 (lr=3.3217e-05) (hash(x)=6638037)
101
+ 95 train 10.054369 (lr=3.3566e-05) (hash(x)=6391506)
102
+ 96 train 10.229466 (lr=3.3916e-05) (hash(x)=7319928)
103
+ 97 train 10.216371 (lr=3.4266e-05) (hash(x)=6113701)
104
+ 98 train 9.964349 (lr=3.4615e-05) (hash(x)=6246453)
105
+ 99 train 10.011649 (lr=3.4965e-05) (hash(x)=5245999)
106
+ 100 val loss 10.0697
107
+ 100 val perplexity 23616.5293
108
+ 100 train 10.027280 (lr=3.5315e-05) (hash(x)=6227256)
109
+ 101 train 10.049640 (lr=3.5664e-05) (hash(x)=5774189)
110
+ 102 train 9.972362 (lr=3.6014e-05) (hash(x)=5588220)
111
+ 103 train 10.146553 (lr=3.6364e-05) (hash(x)=5985675)
112
+ 104 train 9.987696 (lr=3.6713e-05) (hash(x)=6259214)
113
+ 105 train 10.024349 (lr=3.7063e-05) (hash(x)=6451776)
114
+ 106 train 9.835068 (lr=3.7413e-05) (hash(x)=4940406)
115
+ 107 train 9.950304 (lr=3.7762e-05) (hash(x)=4355733)
116
+ 108 train 9.989186 (lr=3.8112e-05) (hash(x)=6186065)
117
+ 109 train 9.950101 (lr=3.8462e-05) (hash(x)=6669188)
118
+ 110 train 10.549043 (lr=3.8811e-05) (hash(x)=7118997)
119
+ 111 train 9.955672 (lr=3.9161e-05) (hash(x)=6984772)
120
+ 112 train 9.848570 (lr=3.9510e-05) (hash(x)=6472877)
121
+ 113 train 9.954233 (lr=3.9860e-05) (hash(x)=7440580)
122
+ 114 train 9.993848 (lr=4.0210e-05) (hash(x)=6378428)
123
+ 115 train 9.937982 (lr=4.0559e-05) (hash(x)=6244096)
124
+ 116 train 9.919327 (lr=4.0909e-05) (hash(x)=9100578)
125
+ 117 train 10.051030 (lr=4.1259e-05) (hash(x)=8381560)
126
+ 118 train 9.919803 (lr=4.1608e-05) (hash(x)=6066072)
127
+ 119 train 9.836422 (lr=4.1958e-05) (hash(x)=6603717)
128
+ 120 train 9.831662 (lr=4.2308e-05) (hash(x)=5138438)
129
+ 121 train 9.763198 (lr=4.2657e-05) (hash(x)=6437516)
130
+ 122 train 9.866451 (lr=4.3007e-05) (hash(x)=5743807)
131
+ 123 train 9.738591 (lr=4.3357e-05) (hash(x)=5669522)
132
+ 124 train 9.784032 (lr=4.3706e-05) (hash(x)=5963623)
133
+ 125 train 9.650594 (lr=4.4056e-05) (hash(x)=5505889)
134
+ 126 train 9.820402 (lr=4.4406e-05) (hash(x)=7607512)
135
+ 127 train 9.722525 (lr=4.4755e-05) (hash(x)=6190579)
136
+ 128 train 9.768591 (lr=4.5105e-05) (hash(x)=7444834)
137
+ 129 train 9.792058 (lr=4.5455e-05) (hash(x)=5153196)
138
+ 130 train 9.740148 (lr=4.5804e-05) (hash(x)=6495326)
139
+ 131 train 9.840230 (lr=4.6154e-05) (hash(x)=4912983)
140
+ 132 train 9.792465 (lr=4.6503e-05) (hash(x)=7094046)
141
+ 133 train 9.636225 (lr=4.6853e-05) (hash(x)=5109125)
142
+ 134 train 9.517854 (lr=4.7203e-05) (hash(x)=7119136)
143
+ 135 train 9.548765 (lr=4.7552e-05) (hash(x)=7276303)
144
+ 136 train 9.691981 (lr=4.7902e-05) (hash(x)=9147123)
145
+ 137 train 9.734625 (lr=4.8252e-05) (hash(x)=6645716)
146
+ 138 train 9.683846 (lr=4.8601e-05) (hash(x)=6347875)
147
+ 139 train 9.541714 (lr=4.8951e-05) (hash(x)=5479947)
148
+ 140 train 9.711014 (lr=4.9301e-05) (hash(x)=5832595)
149
+ 141 train 9.761479 (lr=4.9650e-05) (hash(x)=7850360)
150
+ 142 train 9.923840 (lr=5.0000e-05) (hash(x)=8499683)
151
+ 143 train 9.596619 (lr=5.0350e-05) (hash(x)=6630378)
152
+ 144 train 9.623301 (lr=5.0699e-05) (hash(x)=5340846)
153
+ 145 train 9.566620 (lr=5.1049e-05) (hash(x)=6295358)
154
+ 146 train 9.618697 (lr=5.1399e-05) (hash(x)=5974729)
155
+ 147 train 9.746862 (lr=5.1748e-05) (hash(x)=8097010)
156
+ 148 train 9.608899 (lr=5.2098e-05) (hash(x)=6681250)
157
+ 149 train 9.489750 (lr=5.2448e-05) (hash(x)=3958188)
158
+ 150 val loss 9.4858
159
+ 150 val perplexity 13171.9844
160
+ 150 train 9.503361 (lr=5.2797e-05) (hash(x)=5975539)
161
+ 151 train 10.241243 (lr=5.3147e-05) (hash(x)=9989744)
162
+ 152 train 9.823895 (lr=5.3497e-05) (hash(x)=6748487)
163
+ 153 train 9.354080 (lr=5.3846e-05) (hash(x)=6343832)
164
+ 154 train 9.317882 (lr=5.4196e-05) (hash(x)=5688949)
165
+ 155 train 9.645885 (lr=5.4545e-05) (hash(x)=6694932)
166
+ 156 train 9.402643 (lr=5.4895e-05) (hash(x)=5872835)
167
+ 157 train 9.389061 (lr=5.5245e-05) (hash(x)=6429137)
168
+ 158 train 9.419403 (lr=5.5594e-05) (hash(x)=6999390)
169
+ 159 train 9.311655 (lr=5.5944e-05) (hash(x)=6317894)
170
+ 160 train 9.195327 (lr=5.6294e-05) (hash(x)=5813657)
171
+ 161 train 9.222479 (lr=5.6643e-05) (hash(x)=5165595)
172
+ 162 train 9.235453 (lr=5.6993e-05) (hash(x)=5682633)
173
+ 163 train 9.336295 (lr=5.7343e-05) (hash(x)=5046327)
174
+ 164 train 9.483249 (lr=5.7692e-05) (hash(x)=6195808)
175
+ 165 train 9.323012 (lr=5.8042e-05) (hash(x)=7060363)
176
+ 166 train 9.237650 (lr=5.8392e-05) (hash(x)=5837126)
177
+ 167 train 9.242146 (lr=5.8741e-05) (hash(x)=7947703)
178
+ 168 train 9.013480 (lr=5.9091e-05) (hash(x)=5277865)
179
+ 169 train 9.160038 (lr=5.9441e-05) (hash(x)=5971240)
180
+ 170 train 9.020721 (lr=5.9790e-05) (hash(x)=6590408)
181
+ 171 train 9.154794 (lr=6.0140e-05) (hash(x)=5916068)
182
+ 172 train 9.153493 (lr=6.0490e-05) (hash(x)=6069727)
183
+ 173 train 9.077753 (lr=6.0839e-05) (hash(x)=7426277)
184
+ 174 train 9.255347 (lr=6.1189e-05) (hash(x)=5356513)
185
+ 175 train 9.119928 (lr=6.1538e-05) (hash(x)=5777498)
186
+ 176 train 9.203900 (lr=6.1888e-05) (hash(x)=5989756)
187
+ 177 train 9.037120 (lr=6.2238e-05) (hash(x)=6403331)
188
+ 178 train 8.950569 (lr=6.2587e-05) (hash(x)=7650667)
189
+ 179 train 8.998775 (lr=6.2937e-05) (hash(x)=5753092)
190
+ 180 train 9.054375 (lr=6.3287e-05) (hash(x)=5751457)
191
+ 181 train 8.921463 (lr=6.3636e-05) (hash(x)=5981486)
192
+ 182 train 8.962008 (lr=6.3986e-05) (hash(x)=5918229)
193
+ 183 train 9.032824 (lr=6.4336e-05) (hash(x)=7723226)
194
+ 184 train 8.977828 (lr=6.4685e-05) (hash(x)=6111721)
195
+ 185 train 8.839044 (lr=6.5035e-05) (hash(x)=6069558)
196
+ 186 train 8.906497 (lr=6.5385e-05) (hash(x)=5626336)
197
+ 187 train 9.068098 (lr=6.5734e-05) (hash(x)=7007016)
198
+ 188 train 8.816548 (lr=6.6084e-05) (hash(x)=5883013)
199
+ 189 train 8.784252 (lr=6.6434e-05) (hash(x)=5027958)
200
+ 190 train 8.892821 (lr=6.6783e-05) (hash(x)=5474948)
201
+ 191 train 8.739122 (lr=6.7133e-05) (hash(x)=6491229)
202
+ 192 train 8.831616 (lr=6.7483e-05) (hash(x)=5518341)
203
+ 193 train 8.757521 (lr=6.7832e-05) (hash(x)=7036653)
204
+ 194 train 8.837955 (lr=6.8182e-05) (hash(x)=5527616)
205
+ 195 train 8.958906 (lr=6.8531e-05) (hash(x)=6390752)
206
+ 196 train 8.809910 (lr=6.8881e-05) (hash(x)=6032557)
207
+ 197 train 9.196596 (lr=6.9231e-05) (hash(x)=7758167)
208
+ 198 train 9.086859 (lr=6.9580e-05) (hash(x)=7968070)
209
+ 199 train 8.778289 (lr=6.9930e-05) (hash(x)=6019389)
210
+ 200 val loss 8.7732
211
+ 200 val perplexity 6458.8569
212
+ 200 train 8.906675 (lr=7.0280e-05) (hash(x)=7276744)
213
+ 201 train 8.787192 (lr=7.0629e-05) (hash(x)=6503191)
214
+ 202 train 8.622018 (lr=7.0979e-05) (hash(x)=6006880)
215
+ 203 train 8.795397 (lr=7.1329e-05) (hash(x)=7662067)
216
+ 204 train 8.624496 (lr=7.1678e-05) (hash(x)=6297345)
217
+ 205 train 8.812636 (lr=7.2028e-05) (hash(x)=7901992)
218
+ 206 train 8.890238 (lr=7.2378e-05) (hash(x)=6579950)
219
+ 207 train 8.566961 (lr=7.2727e-05) (hash(x)=4648609)
220
+ 208 train 8.710080 (lr=7.3077e-05) (hash(x)=6903216)
221
+ 209 train 8.555922 (lr=7.3427e-05) (hash(x)=5897288)
222
+ 210 train 8.618887 (lr=7.3776e-05) (hash(x)=7300160)
223
+ 211 train 8.544352 (lr=7.4126e-05) (hash(x)=4725966)
224
+ 212 train 8.282263 (lr=7.4476e-05) (hash(x)=5060060)
225
+ 213 train 8.764868 (lr=7.4825e-05) (hash(x)=6243442)
226
+ 214 train 8.519241 (lr=7.5175e-05) (hash(x)=5893816)
227
+ 215 train 8.555033 (lr=7.5524e-05) (hash(x)=5558355)
228
+ 216 train 8.650086 (lr=7.5874e-05) (hash(x)=4378747)
229
+ 217 train 8.517515 (lr=7.6224e-05) (hash(x)=5400407)
230
+ 218 train 8.518856 (lr=7.6573e-05) (hash(x)=6900554)
231
+ 219 train 8.605632 (lr=7.6923e-05) (hash(x)=6524933)
232
+ 220 train 8.501142 (lr=7.7273e-05) (hash(x)=5409944)
233
+ 221 train 8.599393 (lr=7.7622e-05) (hash(x)=5889724)
234
+ 222 train 8.574244 (lr=7.7972e-05) (hash(x)=4970496)
235
+ 223 train 8.387056 (lr=7.8322e-05) (hash(x)=6369326)
236
+ 224 train 8.391987 (lr=7.8671e-05) (hash(x)=6563975)
237
+ 225 train 8.412841 (lr=7.9021e-05) (hash(x)=5911906)
238
+ 226 train 8.522840 (lr=7.9371e-05) (hash(x)=6111462)
239
+ 227 train 8.532510 (lr=7.9720e-05) (hash(x)=6022625)
240
+ 228 train 8.336929 (lr=8.0070e-05) (hash(x)=5884663)
241
+ 229 train 8.265284 (lr=8.0420e-05) (hash(x)=6584810)
242
+ 230 train 8.384203 (lr=8.0769e-05) (hash(x)=5289998)
243
+ 231 train 8.338122 (lr=8.1119e-05) (hash(x)=4906853)
244
+ 232 train 8.274113 (lr=8.1469e-05) (hash(x)=5862071)
245
+ 233 train 7.889960 (lr=8.1818e-05) (hash(x)=4564564)
246
+ 234 train 8.119931 (lr=8.2168e-05) (hash(x)=5723971)
247
+ 235 train 8.158014 (lr=8.2517e-05) (hash(x)=5589269)
248
+ 236 train 8.156556 (lr=8.2867e-05) (hash(x)=5683368)
249
+ 237 train 8.392100 (lr=8.3217e-05) (hash(x)=6969180)
250
+ 238 train 8.292452 (lr=8.3566e-05) (hash(x)=5341421)
251
+ 239 train 8.181167 (lr=8.3916e-05) (hash(x)=6437376)
252
+ 240 train 8.502455 (lr=8.4266e-05) (hash(x)=7527007)
253
+ 241 train 8.331061 (lr=8.4615e-05) (hash(x)=7116258)
254
+ 242 train 8.337444 (lr=8.4965e-05) (hash(x)=6463930)
255
+ 243 train 8.357821 (lr=8.5315e-05) (hash(x)=7304892)
256
+ 244 train 8.412012 (lr=8.5664e-05) (hash(x)=5881808)
257
+ 245 train 8.253085 (lr=8.6014e-05) (hash(x)=7366721)
258
+ 246 train 8.146955 (lr=8.6364e-05) (hash(x)=6457521)
259
+ 247 train 8.502764 (lr=8.6713e-05) (hash(x)=7084093)
260
+ 248 train 8.349351 (lr=8.7063e-05) (hash(x)=6015300)
261
+ 249 train 8.431327 (lr=8.7413e-05) (hash(x)=7197652)
262
+ 250 val loss 8.1955
263
+ 250 val perplexity 3624.7493
264
+ 250 train 8.390546 (lr=8.7762e-05) (hash(x)=4830828)
265
+ 251 train 8.215960 (lr=8.8112e-05) (hash(x)=5908178)
266
+ 252 train 8.139013 (lr=8.8462e-05) (hash(x)=6299092)
267
+ 253 train 8.064160 (lr=8.8811e-05) (hash(x)=6349974)
268
+ 254 train 8.133615 (lr=8.9161e-05) (hash(x)=5641494)
269
+ 255 train 8.122807 (lr=8.9510e-05) (hash(x)=7048804)
270
+ 256 train 8.248555 (lr=8.9860e-05) (hash(x)=7035244)
271
+ 257 train 8.048696 (lr=9.0210e-05) (hash(x)=7909039)
272
+ 258 train 8.249951 (lr=9.0559e-05) (hash(x)=5470939)
273
+ 259 train 8.119371 (lr=9.0909e-05) (hash(x)=6085549)
274
+ 260 train 8.272176 (lr=9.1259e-05) (hash(x)=5882649)
275
+ 261 train 8.297365 (lr=9.1608e-05) (hash(x)=7463181)
276
+ 262 train 8.383703 (lr=9.1958e-05) (hash(x)=5215574)
277
+ 263 train 7.996406 (lr=9.2308e-05) (hash(x)=5752594)
278
+ 264 train 8.088535 (lr=9.2657e-05) (hash(x)=6416619)
279
+ 265 train 7.998667 (lr=9.3007e-05) (hash(x)=6114601)
280
+ 266 train 7.979347 (lr=9.3357e-05) (hash(x)=5646171)
281
+ 267 train 8.006567 (lr=9.3706e-05) (hash(x)=7662769)
282
+ 268 train 8.136276 (lr=9.4056e-05) (hash(x)=6394376)
283
+ 269 train 7.873127 (lr=9.4406e-05) (hash(x)=7485666)
284
+ 270 train 7.864069 (lr=9.4755e-05) (hash(x)=6636823)
285
+ 271 train 7.797162 (lr=9.5105e-05) (hash(x)=6393520)
286
+ 272 train 8.024079 (lr=9.5455e-05) (hash(x)=5235659)
287
+ 273 train 8.085794 (lr=9.5804e-05) (hash(x)=5586291)
288
+ 274 train 8.281189 (lr=9.6154e-05) (hash(x)=7034674)
289
+ 275 train 8.332518 (lr=9.6503e-05) (hash(x)=5942867)
290
+ 276 train 8.111574 (lr=9.6853e-05) (hash(x)=7812344)
291
+ 277 train 8.299762 (lr=9.7203e-05) (hash(x)=7574027)
292
+ 278 train 8.243958 (lr=9.7552e-05) (hash(x)=5960216)
293
+ 279 train 8.079975 (lr=9.7902e-05) (hash(x)=6793550)
294
+ 280 train 8.077901 (lr=9.8252e-05) (hash(x)=6879965)
295
+ 281 train 7.822279 (lr=9.8601e-05) (hash(x)=6264789)
296
+ 282 train 7.969892 (lr=9.8951e-05) (hash(x)=5175356)
297
+ 283 train 8.098160 (lr=9.9301e-05) (hash(x)=7105976)
298
+ 284 train 8.095515 (lr=9.9650e-05) (hash(x)=7286386)
299
+ 285 train 8.242472 (lr=1.0000e-04) (hash(x)=7244279)
300
+ 286 train 7.910916 (lr=1.0000e-04) (hash(x)=4877247)
301
+ 287 train 7.853118 (lr=1.0000e-04) (hash(x)=6581348)
302
+ 288 train 7.881147 (lr=9.9998e-05) (hash(x)=6397208)
303
+ 289 train 7.917977 (lr=9.9996e-05) (hash(x)=6543825)
304
+ 290 train 7.710927 (lr=9.9993e-05) (hash(x)=6088993)
305
+ 291 train 8.016994 (lr=9.9989e-05) (hash(x)=5555598)
306
+ 292 train 8.014889 (lr=9.9984e-05) (hash(x)=6924240)
307
+ 293 train 8.148224 (lr=9.9979e-05) (hash(x)=7254232)
308
+ 294 train 7.845081 (lr=9.9972e-05) (hash(x)=7689743)
309
+ 295 train 7.818564 (lr=9.9965e-05) (hash(x)=6235837)
310
+ 296 train 7.982429 (lr=9.9956e-05) (hash(x)=5216883)
311
+ 297 train 7.751475 (lr=9.9947e-05) (hash(x)=4738550)
312
+ 298 train 7.825702 (lr=9.9937e-05) (hash(x)=5126918)
313
+ 299 train 7.968163 (lr=9.9926e-05) (hash(x)=5591770)
314
+ 300 val loss 7.9220
315
+ 300 val perplexity 2757.2471
316
+ 300 train 8.660337 (lr=9.9915e-05) (hash(x)=8001573)
317
+ 301 train 8.436928 (lr=9.9902e-05) (hash(x)=8012491)
318
+ 302 train 8.200777 (lr=9.9889e-05) (hash(x)=3997428)
319
+ 303 train 7.551160 (lr=9.9874e-05) (hash(x)=7083021)
320
+ 304 train 8.138569 (lr=9.9859e-05) (hash(x)=7177371)
321
+ 305 train 7.914383 (lr=9.9843e-05) (hash(x)=6456022)
322
+ 306 train 7.989046 (lr=9.9826e-05) (hash(x)=5238861)
323
+ 307 train 7.977997 (lr=9.9808e-05) (hash(x)=7281348)
324
+ 308 train 8.109767 (lr=9.9789e-05) (hash(x)=6848441)
325
+ 309 train 8.180537 (lr=9.9770e-05) (hash(x)=7824794)
326
+ 310 train 8.232085 (lr=9.9749e-05) (hash(x)=7440208)
327
+ 311 train 8.076976 (lr=9.9728e-05) (hash(x)=6155821)
328
+ 312 train 7.885950 (lr=9.9706e-05) (hash(x)=5350461)
329
+ 313 train 8.055326 (lr=9.9683e-05) (hash(x)=7807564)
330
+ 314 train 7.882973 (lr=9.9659e-05) (hash(x)=5992165)
331
+ 315 train 7.837400 (lr=9.9634e-05) (hash(x)=5736241)
332
+ 316 train 7.856054 (lr=9.9609e-05) (hash(x)=6536253)
333
+ 317 train 8.031982 (lr=9.9582e-05) (hash(x)=5688415)
334
+ 318 train 7.645087 (lr=9.9555e-05) (hash(x)=5839705)
335
+ 319 train 7.868800 (lr=9.9526e-05) (hash(x)=5657123)
336
+ 320 train 7.839230 (lr=9.9497e-05) (hash(x)=6354364)
337
+ 321 train 7.870350 (lr=9.9467e-05) (hash(x)=6102985)
338
+ 322 train 8.082251 (lr=9.9437e-05) (hash(x)=7161967)
339
+ 323 train 8.017052 (lr=9.9405e-05) (hash(x)=6452095)
340
+ 324 train 7.774369 (lr=9.9372e-05) (hash(x)=5547396)
341
+ 325 train 8.058315 (lr=9.9339e-05) (hash(x)=5848921)
342
+ 326 train 7.862647 (lr=9.9305e-05) (hash(x)=4890294)
343
+ 327 train 7.475873 (lr=9.9270e-05) (hash(x)=5312267)
344
+ 328 train 7.538931 (lr=9.9234e-05) (hash(x)=5348929)
345
+ 329 train 7.492502 (lr=9.9197e-05) (hash(x)=5025297)
346
+ 330 train 7.545488 (lr=9.9159e-05) (hash(x)=5670052)
347
+ 331 train 7.434057 (lr=9.9121e-05) (hash(x)=5434493)
348
+ 332 train 8.144669 (lr=9.9081e-05) (hash(x)=9122453)
349
+ 333 train 7.908087 (lr=9.9041e-05) (hash(x)=5197474)
350
+ 334 train 7.778298 (lr=9.9000e-05) (hash(x)=5550786)
351
+ 335 train 7.823593 (lr=9.8958e-05) (hash(x)=6830813)
352
+ 336 train 7.671535 (lr=9.8915e-05) (hash(x)=5258435)
353
+ 337 train 7.690409 (lr=9.8872e-05) (hash(x)=5242548)
354
+ 338 train 7.586702 (lr=9.8827e-05) (hash(x)=3754886)
355
+ 339 train 7.549487 (lr=9.8782e-05) (hash(x)=4752771)
356
+ 340 train 7.649672 (lr=9.8736e-05) (hash(x)=5926875)
357
+ 341 train 8.062690 (lr=9.8689e-05) (hash(x)=6691249)
358
+ 342 train 7.863413 (lr=9.8641e-05) (hash(x)=7514623)
359
+ 343 train 7.876762 (lr=9.8592e-05) (hash(x)=6424933)
360
+ 344 train 7.864631 (lr=9.8543e-05) (hash(x)=6347757)
361
+ 345 train 7.836298 (lr=9.8492e-05) (hash(x)=6960120)
362
+ 346 train 7.811934 (lr=9.8441e-05) (hash(x)=7597197)
363
+ 347 train 7.769913 (lr=9.8389e-05) (hash(x)=5786517)
364
+ 348 train 7.799911 (lr=9.8336e-05) (hash(x)=6424363)
365
+ 349 train 7.890680 (lr=9.8282e-05) (hash(x)=7229689)
366
+ 350 val loss 7.7940
367
+ 350 val perplexity 2426.0281
368
+ 350 train 8.014269 (lr=9.8228e-05) (hash(x)=6772284)
369
+ 351 train 8.098228 (lr=9.8172e-05) (hash(x)=6680023)
370
+ 352 train 7.980227 (lr=9.8116e-05) (hash(x)=6218520)
371
+ 353 train 8.006706 (lr=9.8059e-05) (hash(x)=7688393)
372
+ 354 train 7.874406 (lr=9.8001e-05) (hash(x)=6327977)
373
+ 355 train 7.761965 (lr=9.7942e-05) (hash(x)=6474729)
374
+ 356 train 7.889835 (lr=9.7882e-05) (hash(x)=6657266)
375
+ 357 train 7.922829 (lr=9.7822e-05) (hash(x)=5946650)
376
+ 358 train 7.836857 (lr=9.7761e-05) (hash(x)=6734878)
377
+ 359 train 7.835686 (lr=9.7699e-05) (hash(x)=7523279)
378
+ 360 train 7.716060 (lr=9.7636e-05) (hash(x)=6179478)
379
+ 361 train 7.728075 (lr=9.7572e-05) (hash(x)=5531605)
380
+ 362 train 7.739129 (lr=9.7507e-05) (hash(x)=5844980)
381
+ 363 train 7.432608 (lr=9.7442e-05) (hash(x)=5508768)
382
+ 364 train 7.318184 (lr=9.7376e-05) (hash(x)=6337190)
383
+ 365 train 8.728627 (lr=9.7309e-05) (hash(x)=6415982)
384
+ 366 train 7.883586 (lr=9.7241e-05) (hash(x)=6828959)
385
+ 367 train 7.843473 (lr=9.7172e-05) (hash(x)=6635925)
386
+ 368 train 7.591561 (lr=9.7103e-05) (hash(x)=6358540)
387
+ 369 train 7.938766 (lr=9.7032e-05) (hash(x)=5923706)
388
+ 370 train 7.607249 (lr=9.6961e-05) (hash(x)=4843320)
389
+ 371 train 7.866499 (lr=9.6889e-05) (hash(x)=6663801)
390
+ 372 train 7.690751 (lr=9.6817e-05) (hash(x)=5583287)
391
+ 373 train 7.664770 (lr=9.6743e-05) (hash(x)=6113540)
392
+ 374 train 7.785736 (lr=9.6669e-05) (hash(x)=6389936)
393
+ 375 train 7.765574 (lr=9.6593e-05) (hash(x)=5869441)
394
+ 376 train 7.811047 (lr=9.6518e-05) (hash(x)=5148617)
395
+ 377 train 8.039547 (lr=9.6441e-05) (hash(x)=6913454)
396
+ 378 train 7.801489 (lr=9.6363e-05) (hash(x)=5356233)
397
+ 379 train 7.738236 (lr=9.6285e-05) (hash(x)=6698878)
398
+ 380 train 7.711089 (lr=9.6206e-05) (hash(x)=6280665)
399
+ 381 train 7.931552 (lr=9.6126e-05) (hash(x)=6094899)
400
+ 382 train 7.493138 (lr=9.6045e-05) (hash(x)=6249746)
401
+ 383 train 7.785317 (lr=9.5963e-05) (hash(x)=6349004)
402
+ 384 train 7.719461 (lr=9.5881e-05) (hash(x)=7013468)
403
+ 385 train 7.677660 (lr=9.5798e-05) (hash(x)=6045933)
404
+ 386 train 7.979627 (lr=9.5714e-05) (hash(x)=6571735)
405
+ 387 train 7.638318 (lr=9.5629e-05) (hash(x)=4986137)
406
+ 388 train 7.278827 (lr=9.5544e-05) (hash(x)=5244912)
407
+ 389 train 7.655870 (lr=9.5457e-05) (hash(x)=4798229)
408
+ 390 train 7.673829 (lr=9.5370e-05) (hash(x)=5815783)
409
+ 391 train 7.500525 (lr=9.5282e-05) (hash(x)=6008454)
410
+ 392 train 7.723032 (lr=9.5194e-05) (hash(x)=6407333)
411
+ 393 train 7.604087 (lr=9.5104e-05) (hash(x)=5938362)
412
+ 394 train 7.460007 (lr=9.5014e-05) (hash(x)=6077124)
413
+ 395 train 7.752751 (lr=9.4923e-05) (hash(x)=6550770)
414
+ 396 train 7.590045 (lr=9.4831e-05) (hash(x)=6181528)
415
+ 397 train 7.854177 (lr=9.4739e-05) (hash(x)=7055344)
416
+ 398 train 7.952874 (lr=9.4646e-05) (hash(x)=7689348)
417
+ 399 train 7.888442 (lr=9.4551e-05) (hash(x)=7682741)
418
+ 400 val loss 7.7542
419
+ 400 val perplexity 2331.3867
420
+ 400 train 7.715816 (lr=9.4457e-05) (hash(x)=5189401)
421
+ 401 train 7.718100 (lr=9.4361e-05) (hash(x)=7372095)
422
+ 402 train 7.801532 (lr=9.4265e-05) (hash(x)=6838219)
423
+ 403 train 8.289096 (lr=9.4168e-05) (hash(x)=7892158)
424
+ 404 train 7.929001 (lr=9.4070e-05) (hash(x)=5677077)
425
+ 405 train 7.559111 (lr=9.3971e-05) (hash(x)=5929740)
426
+ 406 train 7.702172 (lr=9.3872e-05) (hash(x)=5285055)
427
+ 407 train 7.785645 (lr=9.3772e-05) (hash(x)=7794028)
428
+ 408 train 7.894628 (lr=9.3671e-05) (hash(x)=6640802)
429
+ 409 train 7.976838 (lr=9.3569e-05) (hash(x)=6839923)
430
+ 410 train 7.442285 (lr=9.3467e-05) (hash(x)=4020997)
431
+ 411 train 7.619873 (lr=9.3364e-05) (hash(x)=7093523)
432
+ 412 train 7.896111 (lr=9.3260e-05) (hash(x)=7370045)
433
+ 413 train 7.616610 (lr=9.3155e-05) (hash(x)=5341312)
434
+ 414 train 7.655666 (lr=9.3050e-05) (hash(x)=5589312)
435
+ 415 train 8.124109 (lr=9.2944e-05) (hash(x)=6166062)
436
+ 416 train 8.038710 (lr=9.2837e-05) (hash(x)=8354149)
437
+ 417 train 7.651749 (lr=9.2729e-05) (hash(x)=5992164)
438
+ 418 train 7.653467 (lr=9.2621e-05) (hash(x)=4748657)
439
+ 419 train 7.695851 (lr=9.2512e-05) (hash(x)=6645781)
440
+ 420 train 7.926841 (lr=9.2402e-05) (hash(x)=6689147)
441
+ 421 train 7.798294 (lr=9.2292e-05) (hash(x)=6207724)
442
+ 422 train 7.714084 (lr=9.2181e-05) (hash(x)=6232863)
443
+ 423 train 7.473205 (lr=9.2069e-05) (hash(x)=4811192)
444
+ 424 train 7.514046 (lr=9.1956e-05) (hash(x)=6211761)
445
+ 425 train 7.708770 (lr=9.1843e-05) (hash(x)=6906335)
446
+ 426 train 7.632403 (lr=9.1729e-05) (hash(x)=4922365)
447
+ 427 train 7.657398 (lr=9.1614e-05) (hash(x)=5970866)
448
+ 428 train 7.585715 (lr=9.1499e-05) (hash(x)=6417833)
449
+ 429 train 7.689444 (lr=9.1382e-05) (hash(x)=5238015)
450
+ 430 train 7.608083 (lr=9.1266e-05) (hash(x)=5375451)
451
+ 431 train 7.727693 (lr=9.1148e-05) (hash(x)=7822680)
452
+ 432 train 7.747825 (lr=9.1030e-05) (hash(x)=6173200)
453
+ 433 train 7.512224 (lr=9.0911e-05) (hash(x)=5478019)
454
+ 434 train 7.737761 (lr=9.0791e-05) (hash(x)=5597859)
455
+ 435 train 7.547593 (lr=9.0671e-05) (hash(x)=6317317)
456
+ 436 train 7.738361 (lr=9.0550e-05) (hash(x)=7120467)
457
+ 437 train 7.861116 (lr=9.0428e-05) (hash(x)=5832344)
458
+ 438 train 7.749628 (lr=9.0306e-05) (hash(x)=6813802)
459
+ 439 train 7.802041 (lr=9.0182e-05) (hash(x)=6431409)
460
+ 440 train 7.864882 (lr=9.0059e-05) (hash(x)=5886861)
461
+ 441 train 7.601057 (lr=8.9934e-05) (hash(x)=5900130)
462
+ 442 train 7.870883 (lr=8.9809e-05) (hash(x)=6061563)
463
+ 443 train 7.814676 (lr=8.9683e-05) (hash(x)=6653337)
464
+ 444 train 7.726514 (lr=8.9557e-05) (hash(x)=6575992)
465
+ 445 train 7.722081 (lr=8.9430e-05) (hash(x)=6696784)
466
+ 446 train 7.849238 (lr=8.9302e-05) (hash(x)=6242161)
467
+ 447 train 7.708985 (lr=8.9173e-05) (hash(x)=5323032)
468
+ 448 train 7.470171 (lr=8.9044e-05) (hash(x)=6439367)
469
+ 449 train 7.735622 (lr=8.8914e-05) (hash(x)=5372100)
470
+ 450 val loss 7.7306
471
+ 450 val perplexity 2277.0034
472
+ 450 train 7.783698 (lr=8.8784e-05) (hash(x)=6344849)
473
+ 451 train 7.410220 (lr=8.8653e-05) (hash(x)=5125339)
474
+ 452 train 7.413230 (lr=8.8521e-05) (hash(x)=4876156)
475
+ 453 train 7.568225 (lr=8.8388e-05) (hash(x)=5664463)
476
+ 454 train 7.450185 (lr=8.8255e-05) (hash(x)=6010042)
477
+ 455 train 7.626992 (lr=8.8122e-05) (hash(x)=7363286)
478
+ 456 train 8.251743 (lr=8.7987e-05) (hash(x)=4873268)
479
+ 457 train 7.974854 (lr=8.7852e-05) (hash(x)=7444468)
480
+ 458 train 7.990011 (lr=8.7717e-05) (hash(x)=6698134)
481
+ 459 train 7.812050 (lr=8.7580e-05) (hash(x)=7670050)
482
+ 460 train 7.824468 (lr=8.7444e-05) (hash(x)=6280866)
483
+ 461 train 7.680571 (lr=8.7306e-05) (hash(x)=8189420)
484
+ 462 train 7.808397 (lr=8.7168e-05) (hash(x)=6878292)
485
+ 463 train 7.528366 (lr=8.7029e-05) (hash(x)=5616075)
486
+ 464 train 7.741541 (lr=8.6890e-05) (hash(x)=5839939)
487
+ 465 train 7.601060 (lr=8.6750e-05) (hash(x)=6360924)
488
+ 466 train 8.037739 (lr=8.6609e-05) (hash(x)=6466475)
489
+ 467 train 7.803490 (lr=8.6468e-05) (hash(x)=6593764)
490
+ 468 train 8.250538 (lr=8.6326e-05) (hash(x)=7389698)
491
+ 469 train 7.926255 (lr=8.6184e-05) (hash(x)=6764377)
492
+ 470 train 7.659231 (lr=8.6041e-05) (hash(x)=7293570)
493
+ 471 train 7.697008 (lr=8.5897e-05) (hash(x)=5672608)
494
+ 472 train 7.512698 (lr=8.5753e-05) (hash(x)=5587975)
495
+ 473 train 7.418221 (lr=8.5608e-05) (hash(x)=6985256)
496
+ 474 train 8.043233 (lr=8.5462e-05) (hash(x)=6549231)
497
+ 475 train 7.734573 (lr=8.5316e-05) (hash(x)=6828653)
498
+ 476 train 7.555032 (lr=8.5170e-05) (hash(x)=5478970)
499
+ 477 train 7.601398 (lr=8.5022e-05) (hash(x)=5199845)
500
+ 478 train 7.673420 (lr=8.4875e-05) (hash(x)=6482073)
501
+ 479 train 7.771177 (lr=8.4726e-05) (hash(x)=6839867)
502
+ 480 train 7.749074 (lr=8.4577e-05) (hash(x)=7073454)
503
+ 481 train 8.012592 (lr=8.4428e-05) (hash(x)=6928712)
504
+ 482 train 7.745379 (lr=8.4278e-05) (hash(x)=5962982)
505
+ 483 train 7.839441 (lr=8.4127e-05) (hash(x)=8426992)
506
+ 484 train 7.767354 (lr=8.3976e-05) (hash(x)=6422394)
507
+ 485 train 7.709857 (lr=8.3824e-05) (hash(x)=5330355)
508
+ 486 train 7.198331 (lr=8.3672e-05) (hash(x)=4111491)
509
+ 487 train 7.199775 (lr=8.3519e-05) (hash(x)=4916928)
510
+ 488 train 7.381323 (lr=8.3366e-05) (hash(x)=6610244)
511
+ 489 train 7.878328 (lr=8.3212e-05) (hash(x)=7287928)
512
+ 490 train 7.772500 (lr=8.3057e-05) (hash(x)=6525083)
513
+ 491 train 7.713719 (lr=8.2902e-05) (hash(x)=6484050)
514
+ 492 train 7.876857 (lr=8.2746e-05) (hash(x)=6573424)
515
+ 493 train 7.725871 (lr=8.2590e-05) (hash(x)=5062850)
516
+ 494 train 7.698461 (lr=8.2434e-05) (hash(x)=6215995)
517
+ 495 train 7.842815 (lr=8.2276e-05) (hash(x)=8353379)
518
+ 496 train 7.717014 (lr=8.2119e-05) (hash(x)=6344823)
519
+ 497 train 7.685719 (lr=8.1960e-05) (hash(x)=6805358)
520
+ 498 train 7.570426 (lr=8.1801e-05) (hash(x)=6018070)
521
+ 499 train 7.869565 (lr=8.1642e-05) (hash(x)=6552510)
522
+ 500 val loss 7.7081
523
+ 500 val perplexity 2226.3865
524
+ 500 train 7.613410 (lr=8.1482e-05) (hash(x)=7218108)
525
+ 501 train 7.810271 (lr=8.1322e-05) (hash(x)=5105545)
526
+ 502 train 7.785931 (lr=8.1161e-05) (hash(x)=5997860)
527
+ 503 train 7.608711 (lr=8.1000e-05) (hash(x)=4838871)
528
+ 504 train 7.655112 (lr=8.0838e-05) (hash(x)=7105888)
529
+ 505 train 7.966247 (lr=8.0676e-05) (hash(x)=5945286)
530
+ 506 train 7.205777 (lr=8.0513e-05) (hash(x)=4381720)
531
+ 507 train 6.874617 (lr=8.0349e-05) (hash(x)=3021697)
532
+ 508 train 7.268000 (lr=8.0186e-05) (hash(x)=6826973)
533
+ 509 train 7.836849 (lr=8.0021e-05) (hash(x)=6443070)
534
+ 510 train 7.521781 (lr=7.9856e-05) (hash(x)=5520637)
535
+ 511 train 7.794570 (lr=7.9691e-05) (hash(x)=6795665)
536
+ 512 train 7.568655 (lr=7.9525e-05) (hash(x)=6063891)
537
+ 513 train 8.054665 (lr=7.9359e-05) (hash(x)=7683068)
538
+ 514 train 7.362846 (lr=7.9192e-05) (hash(x)=5526694)
539
+ 515 train 7.402991 (lr=7.9025e-05) (hash(x)=5486935)
540
+ 516 train 7.765610 (lr=7.8858e-05) (hash(x)=6002252)
541
+ 517 train 7.711968 (lr=7.8689e-05) (hash(x)=6115232)
542
+ 518 train 8.046393 (lr=7.8521e-05) (hash(x)=9039136)
543
+ 519 train 7.657831 (lr=7.8352e-05) (hash(x)=6678038)
544
+ 520 train 7.828762 (lr=7.8182e-05) (hash(x)=5383926)
545
+ 521 train 7.897987 (lr=7.8012e-05) (hash(x)=5747629)
546
+ 522 train 7.736970 (lr=7.7842e-05) (hash(x)=7090040)
547
+ 523 train 7.949093 (lr=7.7671e-05) (hash(x)=6657714)
548
+ 524 train 7.643119 (lr=7.7500e-05) (hash(x)=5414034)
549
+ 525 train 7.693629 (lr=7.7328e-05) (hash(x)=6190642)
550
+ 526 train 7.842270 (lr=7.7156e-05) (hash(x)=7431177)
551
+ 527 train 7.730779 (lr=7.6984e-05) (hash(x)=6112215)
552
+ 528 train 7.853877 (lr=7.6811e-05) (hash(x)=6785874)
553
+ 529 train 7.589388 (lr=7.6637e-05) (hash(x)=7314989)
554
+ 530 train 7.801018 (lr=7.6463e-05) (hash(x)=6222614)
555
+ 531 train 7.891786 (lr=7.6289e-05) (hash(x)=8353143)
556
+ 532 train 7.970550 (lr=7.6115e-05) (hash(x)=6752498)
557
+ 533 train 7.723977 (lr=7.5940e-05) (hash(x)=5912570)
558
+ 534 train 7.863116 (lr=7.5764e-05) (hash(x)=5621785)
559
+ 535 train 7.752716 (lr=7.5588e-05) (hash(x)=5915361)
560
+ 536 train 7.621918 (lr=7.5412e-05) (hash(x)=6853672)
561
+ 537 train 7.591694 (lr=7.5235e-05) (hash(x)=6369494)
562
+ 538 train 7.747393 (lr=7.5058e-05) (hash(x)=6039652)
563
+ 539 train 7.885455 (lr=7.4881e-05) (hash(x)=6254885)
564
+ 540 train 7.738880 (lr=7.4703e-05) (hash(x)=5829860)
565
+ 541 train 7.638913 (lr=7.4525e-05) (hash(x)=5425120)
566
+ 542 train 7.413933 (lr=7.4346e-05) (hash(x)=5209746)
567
+ 543 train 8.018750 (lr=7.4167e-05) (hash(x)=5771588)
568
+ 544 train 7.726708 (lr=7.3988e-05) (hash(x)=8298790)
569
+ 545 train 7.790674 (lr=7.3808e-05) (hash(x)=6763967)
570
+ 546 train 7.749174 (lr=7.3628e-05) (hash(x)=4882397)
571
+ 547 train 7.528704 (lr=7.3448e-05) (hash(x)=5561507)
572
+ 548 train 7.541356 (lr=7.3267e-05) (hash(x)=5762755)
573
+ 549 train 7.650998 (lr=7.3086e-05) (hash(x)=7081151)
574
+ 550 val loss 7.6997
575
+ 550 val perplexity 2207.7559
576
+ 550 train 7.585122 (lr=7.2904e-05) (hash(x)=5047520)
577
+ 551 train 7.726511 (lr=7.2723e-05) (hash(x)=5688829)
578
+ 552 train 7.608286 (lr=7.2540e-05) (hash(x)=7008925)
579
+ 553 train 7.700274 (lr=7.2358e-05) (hash(x)=7821554)
580
+ 554 train 7.742696 (lr=7.2175e-05) (hash(x)=6469035)
581
+ 555 train 7.565489 (lr=7.1992e-05) (hash(x)=5371951)
582
+ 556 train 7.837085 (lr=7.1808e-05) (hash(x)=6436050)
583
+ 557 train 7.763275 (lr=7.1624e-05) (hash(x)=8120149)
584
+ 558 train 8.010471 (lr=7.1440e-05) (hash(x)=6189935)
585
+ 559 train 7.702793 (lr=7.1256e-05) (hash(x)=5443305)
586
+ 560 train 7.808412 (lr=7.1071e-05) (hash(x)=6294312)
587
+ 561 train 7.434697 (lr=7.0886e-05) (hash(x)=7066030)
588
+ 562 train 7.862287 (lr=7.0701e-05) (hash(x)=7744989)
589
+ 563 train 8.015177 (lr=7.0515e-05) (hash(x)=5016757)
590
+ 564 train 7.812621 (lr=7.0329e-05) (hash(x)=8001284)
591
+ 565 train 7.745141 (lr=7.0143e-05) (hash(x)=6296163)
592
+ 566 train 7.839051 (lr=6.9956e-05) (hash(x)=6027688)
593
+ 567 train 7.765294 (lr=6.9769e-05) (hash(x)=6901933)
594
+ 568 train 7.622947 (lr=6.9582e-05) (hash(x)=5124305)
595
+ 569 train 7.763837 (lr=6.9394e-05) (hash(x)=8056633)
596
+ 570 train 7.810954 (lr=6.9207e-05) (hash(x)=6677566)
597
+ 571 train 7.833131 (lr=6.9019e-05) (hash(x)=6019085)
598
+ 572 train 7.577094 (lr=6.8830e-05) (hash(x)=5924495)
599
+ 573 train 7.679523 (lr=6.8642e-05) (hash(x)=7003893)
600
+ 574 train 7.642677 (lr=6.8453e-05) (hash(x)=5563075)
601
+ 575 train 7.487240 (lr=6.8264e-05) (hash(x)=4294425)
602
+ 576 train 7.681340 (lr=6.8075e-05) (hash(x)=5677870)
603
+ 577 train 7.566463 (lr=6.7885e-05) (hash(x)=6568540)
604
+ 578 train 7.815696 (lr=6.7695e-05) (hash(x)=7473394)
605
+ 579 train 7.827282 (lr=6.7505e-05) (hash(x)=6095229)
606
+ 580 train 7.813845 (lr=6.7315e-05) (hash(x)=6946168)
607
+ 581 train 7.628621 (lr=6.7124e-05) (hash(x)=5512150)
608
+ 582 train 7.266946 (lr=6.6933e-05) (hash(x)=4380268)
609
+ 583 train 7.335700 (lr=6.6742e-05) (hash(x)=4363941)
610
+ 584 train 7.482478 (lr=6.6551e-05) (hash(x)=4836009)
611
+ 585 train 7.791262 (lr=6.6360e-05) (hash(x)=6457783)
612
+ 586 train 7.655459 (lr=6.6168e-05) (hash(x)=6378784)
613
+ 587 train 7.571392 (lr=6.5976e-05) (hash(x)=4693798)
614
+ 588 train 7.895886 (lr=6.5784e-05) (hash(x)=9193310)
615
+ 589 train 8.573421 (lr=6.5592e-05) (hash(x)=8759806)
616
+ 590 train 8.921786 (lr=6.5399e-05) (hash(x)=9176140)
617
+ 591 train 7.622018 (lr=6.5206e-05) (hash(x)=5486335)
618
+ 592 train 7.903359 (lr=6.5013e-05) (hash(x)=7868436)
619
+ 593 train 7.776805 (lr=6.4820e-05) (hash(x)=6964993)
620
+ 594 train 8.002827 (lr=6.4627e-05) (hash(x)=6741270)
621
+ 595 train 8.289627 (lr=6.4433e-05) (hash(x)=7907450)
622
+ 596 train 8.112411 (lr=6.4240e-05) (hash(x)=6443452)
623
+ 597 train 7.639886 (lr=6.4046e-05) (hash(x)=5236642)
624
+ 598 train 7.793125 (lr=6.3852e-05) (hash(x)=5604795)
625
+ 599 train 7.737425 (lr=6.3658e-05) (hash(x)=7295165)
626
+ 600 val loss 7.6890
627
+ 600 val perplexity 2184.2615
628
+ 600 train 7.711761 (lr=6.3463e-05) (hash(x)=6034549)
629
+ 601 train 7.713058 (lr=6.3269e-05) (hash(x)=4913602)
630
+ 602 train 7.684858 (lr=6.3074e-05) (hash(x)=6271294)
631
+ 603 train 7.841843 (lr=6.2879e-05) (hash(x)=6814026)
632
+ 604 train 7.737566 (lr=6.2684e-05) (hash(x)=6848040)
633
+ 605 train 8.049725 (lr=6.2489e-05) (hash(x)=7042051)
634
+ 606 train 7.942205 (lr=6.2294e-05) (hash(x)=6481002)
635
+ 607 train 7.585926 (lr=6.2098e-05) (hash(x)=6267424)
636
+ 608 train 7.884603 (lr=6.1903e-05) (hash(x)=7665306)
637
+ 609 train 7.772612 (lr=6.1707e-05) (hash(x)=5614727)
638
+ 610 train 7.785593 (lr=6.1511e-05) (hash(x)=7039197)
639
+ 611 train 7.823216 (lr=6.1315e-05) (hash(x)=8086437)
640
+ 612 train 7.916790 (lr=6.1119e-05) (hash(x)=6993846)
641
+ 613 train 7.925679 (lr=6.0923e-05) (hash(x)=5900143)
642
+ 614 train 7.829940 (lr=6.0726e-05) (hash(x)=4841318)
643
+ 615 train 7.817242 (lr=6.0530e-05) (hash(x)=5270452)
644
+ 616 train 7.705128 (lr=6.0333e-05) (hash(x)=5955026)
645
+ 617 train 7.816060 (lr=6.0137e-05) (hash(x)=8617707)
646
+ 618 train 7.658469 (lr=5.9940e-05) (hash(x)=5159401)
647
+ 619 train 7.656473 (lr=5.9743e-05) (hash(x)=6420820)
648
+ 620 train 7.637937 (lr=5.9546e-05) (hash(x)=6628863)
649
+ 621 train 7.444483 (lr=5.9349e-05) (hash(x)=6215930)
650
+ 622 train 7.787798 (lr=5.9152e-05) (hash(x)=5248645)
651
+ 623 train 7.671401 (lr=5.8955e-05) (hash(x)=6305297)
652
+ 624 train 8.214602 (lr=5.8758e-05) (hash(x)=6989704)
653
+ 625 train 7.731333 (lr=5.8560e-05) (hash(x)=4026682)
654
+ 626 train 7.568751 (lr=5.8363e-05) (hash(x)=6367688)
655
+ 627 train 7.669050 (lr=5.8165e-05) (hash(x)=7889849)
656
+ 628 train 7.845494 (lr=5.7968e-05) (hash(x)=7334719)
657
+ 629 train 7.970498 (lr=5.7770e-05) (hash(x)=6141721)
658
+ 630 train 8.756197 (lr=5.7573e-05) (hash(x)=5056572)
659
+ 631 train 8.171079 (lr=5.7375e-05) (hash(x)=6040077)
660
+ 632 train 8.286074 (lr=5.7177e-05) (hash(x)=6866651)
661
+ 633 train 8.559497 (lr=5.6979e-05) (hash(x)=6155449)
662
+ 634 train 8.099715 (lr=5.6782e-05) (hash(x)=6690174)
663
+ 635 train 7.738471 (lr=5.6584e-05) (hash(x)=5652497)
664
+ 636 train 7.760779 (lr=5.6386e-05) (hash(x)=5931050)
665
+ 637 train 7.835916 (lr=5.6188e-05) (hash(x)=6961314)
666
+ 638 train 7.666307 (lr=5.5990e-05) (hash(x)=5973636)
667
+ 639 train 7.713588 (lr=5.5792e-05) (hash(x)=7130251)
668
+ 640 train 7.487885 (lr=5.5594e-05) (hash(x)=7503390)
669
+ 641 train 7.570742 (lr=5.5396e-05) (hash(x)=5927461)
670
+ 642 train 7.720950 (lr=5.5198e-05) (hash(x)=6196741)
671
+ 643 train 7.651439 (lr=5.5000e-05) (hash(x)=6610177)
672
+ 644 train 7.636801 (lr=5.4802e-05) (hash(x)=6635147)
673
+ 645 train 7.853774 (lr=5.4604e-05) (hash(x)=7277580)
674
+ 646 train 7.475645 (lr=5.4406e-05) (hash(x)=5050080)
675
+ 647 train 7.577787 (lr=5.4208e-05) (hash(x)=6508350)
676
+ 648 train 7.780419 (lr=5.4010e-05) (hash(x)=5276338)
677
+ 649 train 7.911895 (lr=5.3812e-05) (hash(x)=6536034)
678
+ 650 val loss 7.6984
679
+ 650 val perplexity 2204.8533
680
+ 650 train 7.729506 (lr=5.3614e-05) (hash(x)=6944772)
681
+ 651 train 7.770952 (lr=5.3416e-05) (hash(x)=6994983)
682
+ 652 train 7.728178 (lr=5.3218e-05) (hash(x)=7172017)
683
+ 653 train 7.808686 (lr=5.3021e-05) (hash(x)=8700721)
684
+ 654 train 7.787653 (lr=5.2823e-05) (hash(x)=6774360)
685
+ 655 train 7.640214 (lr=5.2625e-05) (hash(x)=5859576)
686
+ 656 train 7.339822 (lr=5.2427e-05) (hash(x)=5899275)
687
+ 657 train 7.283192 (lr=5.2230e-05) (hash(x)=5264962)
688
+ 658 train 7.082268 (lr=5.2032e-05) (hash(x)=5679861)
689
+ 659 train 7.507595 (lr=5.1835e-05) (hash(x)=5487065)
690
+ 660 train 7.276791 (lr=5.1637e-05) (hash(x)=4239476)
691
+ 661 train 7.593810 (lr=5.1440e-05) (hash(x)=5731624)
692
+ 662 train 7.696978 (lr=5.1242e-05) (hash(x)=5883465)
693
+ 663 train 7.491001 (lr=5.1045e-05) (hash(x)=4892065)
694
+ 664 train 7.333651 (lr=5.0848e-05) (hash(x)=5858782)
695
+ 665 train 7.625510 (lr=5.0651e-05) (hash(x)=5489496)
696
+ 666 train 7.312204 (lr=5.0454e-05) (hash(x)=4485195)
697
+ 667 train 7.345871 (lr=5.0257e-05) (hash(x)=4933674)
698
+ 668 train 7.596763 (lr=5.0060e-05) (hash(x)=5746292)
699
+ 669 train 7.627125 (lr=4.9863e-05) (hash(x)=7021003)
700
+ 670 train 7.782218 (lr=4.9667e-05) (hash(x)=5876710)
701
+ 671 train 7.816663 (lr=4.9470e-05) (hash(x)=7317289)
702
+ 672 train 7.918533 (lr=4.9274e-05) (hash(x)=5598226)
703
+ 673 train 7.732220 (lr=4.9077e-05) (hash(x)=7869305)
704
+ 674 train 7.803168 (lr=4.8881e-05) (hash(x)=6611408)
705
+ 675 train 7.777740 (lr=4.8685e-05) (hash(x)=6811522)
706
+ 676 train 7.734209 (lr=4.8489e-05) (hash(x)=6704714)
707
+ 677 train 7.608278 (lr=4.8293e-05) (hash(x)=6601423)
708
+ 678 train 7.853662 (lr=4.8097e-05) (hash(x)=6726071)
709
+ 679 train 7.316934 (lr=4.7902e-05) (hash(x)=5510218)
710
+ 680 train 8.074477 (lr=4.7706e-05) (hash(x)=7950952)
711
+ 681 train 7.799041 (lr=4.7511e-05) (hash(x)=7180298)
712
+ 682 train 7.601662 (lr=4.7316e-05) (hash(x)=6068813)
713
+ 683 train 7.742239 (lr=4.7121e-05) (hash(x)=7304235)
714
+ 684 train 7.781285 (lr=4.6926e-05) (hash(x)=7441806)
715
+ 685 train 7.889324 (lr=4.6731e-05) (hash(x)=8111920)
716
+ 686 train 8.512427 (lr=4.6537e-05) (hash(x)=6222783)
717
+ 687 train 7.731240 (lr=4.6342e-05) (hash(x)=6752265)
718
+ 688 train 7.533483 (lr=4.6148e-05) (hash(x)=6147634)
719
+ 689 train 7.610834 (lr=4.5954e-05) (hash(x)=6788720)
720
+ 690 train 7.751420 (lr=4.5760e-05) (hash(x)=6413518)
721
+ 691 train 7.633505 (lr=4.5567e-05) (hash(x)=5994476)
722
+ 692 train 7.560313 (lr=4.5373e-05) (hash(x)=5462082)
723
+ 693 train 7.555311 (lr=4.5180e-05) (hash(x)=5862533)
724
+ 694 train 7.709771 (lr=4.4987e-05) (hash(x)=7132796)
725
+ 695 train 7.610755 (lr=4.4794e-05) (hash(x)=6530867)
726
+ 696 train 7.685654 (lr=4.4601e-05) (hash(x)=7932207)
727
+ 697 train 7.625782 (lr=4.4408e-05) (hash(x)=5914734)
728
+ 698 train 7.803936 (lr=4.4216e-05) (hash(x)=6361594)
729
+ 699 train 7.764895 (lr=4.4024e-05) (hash(x)=5746260)
730
+ 700 val loss 7.6638
731
+ 700 val perplexity 2129.7861
732
+ 700 train 7.706650 (lr=4.3832e-05) (hash(x)=7065743)
733
+ 701 train 7.621795 (lr=4.3640e-05) (hash(x)=6856097)
734
+ 702 train 7.692538 (lr=4.3449e-05) (hash(x)=7588648)
735
+ 703 train 7.595893 (lr=4.3258e-05) (hash(x)=5790078)
736
+ 704 train 7.640171 (lr=4.3067e-05) (hash(x)=6069422)
737
+ 705 train 7.982376 (lr=4.2876e-05) (hash(x)=5805002)
738
+ 706 train 7.670902 (lr=4.2685e-05) (hash(x)=5344711)
739
+ 707 train 7.528575 (lr=4.2495e-05) (hash(x)=6430135)
740
+ 708 train 7.619657 (lr=4.2305e-05) (hash(x)=6317763)
741
+ 709 train 7.526407 (lr=4.2115e-05) (hash(x)=6156715)
742
+ 710 train 7.918369 (lr=4.1925e-05) (hash(x)=6321494)
743
+ 711 train 7.952781 (lr=4.1736e-05) (hash(x)=7614023)
744
+ 712 train 7.907634 (lr=4.1547e-05) (hash(x)=6740380)
745
+ 713 train 7.726121 (lr=4.1358e-05) (hash(x)=4861744)
746
+ 714 train 7.608973 (lr=4.1170e-05) (hash(x)=6542179)
747
+ 715 train 7.706766 (lr=4.0981e-05) (hash(x)=5244861)
748
+ 716 train 7.795297 (lr=4.0793e-05) (hash(x)=7306636)
749
+ 717 train 7.645782 (lr=4.0606e-05) (hash(x)=7163697)
750
+ 718 train 7.856658 (lr=4.0418e-05) (hash(x)=6421642)
751
+ 719 train 7.619659 (lr=4.0231e-05) (hash(x)=5245146)
752
+ 720 train 7.694344 (lr=4.0044e-05) (hash(x)=6046027)
753
+ 721 train 7.482431 (lr=3.9857e-05) (hash(x)=6153866)
754
+ 722 train 7.655572 (lr=3.9671e-05) (hash(x)=5827481)
755
+ 723 train 7.612278 (lr=3.9485e-05) (hash(x)=6415565)
756
+ 724 train 7.747008 (lr=3.9299e-05) (hash(x)=6409570)
757
+ 725 train 8.191133 (lr=3.9114e-05) (hash(x)=7835853)
758
+ 726 train 7.333227 (lr=3.8929e-05) (hash(x)=4827589)
759
+ 727 train 7.473877 (lr=3.8744e-05) (hash(x)=5786972)
760
+ 728 train 7.862127 (lr=3.8560e-05) (hash(x)=6736612)
761
+ 729 train 7.710213 (lr=3.8376e-05) (hash(x)=6733560)
762
+ 730 train 7.895542 (lr=3.8192e-05) (hash(x)=4877208)
763
+ 731 train 7.404007 (lr=3.8008e-05) (hash(x)=6131703)
764
+ 732 train 7.534698 (lr=3.7825e-05) (hash(x)=6533769)
765
+ 733 train 7.468703 (lr=3.7642e-05) (hash(x)=6001331)
766
+ 734 train 8.018875 (lr=3.7460e-05) (hash(x)=10602643)
767
+ 735 train 7.929778 (lr=3.7277e-05) (hash(x)=6346459)
768
+ 736 train 7.629424 (lr=3.7096e-05) (hash(x)=6728215)
769
+ 737 train 7.984586 (lr=3.6914e-05) (hash(x)=8943770)
770
+ 738 train 7.763153 (lr=3.6733e-05) (hash(x)=7141912)
771
+ 739 train 7.720656 (lr=3.6552e-05) (hash(x)=6504131)
772
+ 740 train 7.583533 (lr=3.6372e-05) (hash(x)=6461667)
773
+ 741 train 7.804521 (lr=3.6192e-05) (hash(x)=5869339)
774
+ 742 train 8.186316 (lr=3.6012e-05) (hash(x)=7948065)
775
+ 743 train 7.420292 (lr=3.5833e-05) (hash(x)=5209234)
776
+ 744 train 7.534084 (lr=3.5654e-05) (hash(x)=6372244)
777
+ 745 train 7.737566 (lr=3.5475e-05) (hash(x)=7678937)
778
+ 746 train 7.649832 (lr=3.5297e-05) (hash(x)=6519438)
779
+ 747 train 7.724603 (lr=3.5119e-05) (hash(x)=6163272)
780
+ 748 train 7.760015 (lr=3.4942e-05) (hash(x)=7025209)
781
+ 749 train 7.602235 (lr=3.4765e-05) (hash(x)=5774172)
782
+ 750 val loss 7.6545
783
+ 750 val perplexity 2110.0444
784
+ 750 train 7.760481 (lr=3.4588e-05) (hash(x)=5327301)
785
+ 751 train 7.520357 (lr=3.4412e-05) (hash(x)=5676768)
786
+ 752 train 7.231507 (lr=3.4236e-05) (hash(x)=4823304)
787
+ 753 train 7.352864 (lr=3.4060e-05) (hash(x)=6443895)
788
+ 754 train 8.174284 (lr=3.3885e-05) (hash(x)=8032890)
789
+ 755 train 8.235389 (lr=3.3711e-05) (hash(x)=6090561)
790
+ 756 train 7.558037 (lr=3.3537e-05) (hash(x)=4514262)
791
+ 757 train 8.027495 (lr=3.3363e-05) (hash(x)=7788628)
792
+ 758 train 7.949027 (lr=3.3189e-05) (hash(x)=7239456)
793
+ 759 train 7.990885 (lr=3.3016e-05) (hash(x)=5863092)
794
+ 760 train 7.653976 (lr=3.2844e-05) (hash(x)=5572628)
795
+ 761 train 7.734884 (lr=3.2672e-05) (hash(x)=7141354)
796
+ 762 train 7.957465 (lr=3.2500e-05) (hash(x)=7272638)
797
+ 763 train 7.783489 (lr=3.2329e-05) (hash(x)=7201312)
798
+ 764 train 7.811661 (lr=3.2158e-05) (hash(x)=6015388)
799
+ 765 train 7.886372 (lr=3.1988e-05) (hash(x)=6414278)
800
+ 766 train 7.633945 (lr=3.1818e-05) (hash(x)=7137466)
801
+ 767 train 8.201916 (lr=3.1648e-05) (hash(x)=7019489)
802
+ 768 train 7.759357 (lr=3.1479e-05) (hash(x)=7233453)
803
+ 769 train 7.522752 (lr=3.1311e-05) (hash(x)=7914626)
804
+ 770 train 7.609814 (lr=3.1142e-05) (hash(x)=5764080)
805
+ 771 train 7.635783 (lr=3.0975e-05) (hash(x)=6225608)
806
+ 772 train 7.734044 (lr=3.0808e-05) (hash(x)=8097255)
807
+ 773 train 7.732870 (lr=3.0641e-05) (hash(x)=5998078)
808
+ 774 train 7.850370 (lr=3.0475e-05) (hash(x)=5416254)
809
+ 775 train 7.552961 (lr=3.0309e-05) (hash(x)=5483019)
810
+ 776 train 7.372802 (lr=3.0144e-05) (hash(x)=4702208)
811
+ 777 train 7.677642 (lr=2.9979e-05) (hash(x)=5911642)
812
+ 778 train 7.721630 (lr=2.9814e-05) (hash(x)=6132487)
813
+ 779 train 7.863702 (lr=2.9651e-05) (hash(x)=5903258)
814
+ 780 train 7.791454 (lr=2.9487e-05) (hash(x)=7915382)
815
+ 781 train 7.794741 (lr=2.9324e-05) (hash(x)=5632006)
816
+ 782 train 7.588705 (lr=2.9162e-05) (hash(x)=6518211)
817
+ 783 train 7.669781 (lr=2.9000e-05) (hash(x)=5968716)
818
+ 784 train 7.795304 (lr=2.8839e-05) (hash(x)=7344525)
819
+ 785 train 7.686398 (lr=2.8678e-05) (hash(x)=6401968)
820
+ 786 train 7.613751 (lr=2.8518e-05) (hash(x)=6276127)
821
+ 787 train 7.730937 (lr=2.8358e-05) (hash(x)=5778017)
822
+ 788 train 7.252354 (lr=2.8199e-05) (hash(x)=5387306)
823
+ 789 train 7.197388 (lr=2.8040e-05) (hash(x)=5772567)
824
+ 790 train 7.442766 (lr=2.7881e-05) (hash(x)=6383748)
825
+ 791 train 7.766801 (lr=2.7724e-05) (hash(x)=7780194)
826
+ 792 train 7.635818 (lr=2.7566e-05) (hash(x)=7119030)
827
+ 793 train 7.758094 (lr=2.7410e-05) (hash(x)=6424771)
828
+ 794 train 7.568174 (lr=2.7254e-05) (hash(x)=6540151)
829
+ 795 train 7.543458 (lr=2.7098e-05) (hash(x)=6140998)
830
+ 796 train 7.627117 (lr=2.6943e-05) (hash(x)=6208271)
831
+ 797 train 7.481494 (lr=2.6788e-05) (hash(x)=7859566)
832
+ 798 train 8.071901 (lr=2.6634e-05) (hash(x)=7064477)
833
+ 799 train 7.316650 (lr=2.6481e-05) (hash(x)=3784321)
834
+ 800 val loss 7.6390
835
+ 800 val perplexity 2077.7373
836
+ 800 train 7.613373 (lr=2.6328e-05) (hash(x)=4472758)
837
+ 801 train 7.556256 (lr=2.6176e-05) (hash(x)=5557891)
838
+ 802 train 7.718231 (lr=2.6024e-05) (hash(x)=7969325)
839
+ 803 train 7.744553 (lr=2.5873e-05) (hash(x)=5860821)
840
+ 804 train 7.752196 (lr=2.5722e-05) (hash(x)=6750848)
841
+ 805 train 7.675200 (lr=2.5572e-05) (hash(x)=5674826)
842
+ 806 train 7.464351 (lr=2.5423e-05) (hash(x)=5529163)
843
+ 807 train 7.904300 (lr=2.5274e-05) (hash(x)=7774109)
844
+ 808 train 7.645289 (lr=2.5125e-05) (hash(x)=6762509)
845
+ 809 train 7.755661 (lr=2.4978e-05) (hash(x)=5466449)
846
+ 810 train 7.614519 (lr=2.4830e-05) (hash(x)=7046935)
847
+ 811 train 8.423788 (lr=2.4684e-05) (hash(x)=7997664)
848
+ 812 train 7.802542 (lr=2.4538e-05) (hash(x)=6882594)
849
+ 813 train 7.975195 (lr=2.4392e-05) (hash(x)=7006517)
850
+ 814 train 7.828537 (lr=2.4247e-05) (hash(x)=6832530)
851
+ 815 train 7.666993 (lr=2.4103e-05) (hash(x)=6576749)
852
+ 816 train 7.767770 (lr=2.3959e-05) (hash(x)=8276629)
853
+ 817 train 7.544612 (lr=2.3816e-05) (hash(x)=6896198)
854
+ 818 train 7.764786 (lr=2.3674e-05) (hash(x)=5829252)
855
+ 819 train 7.943740 (lr=2.3532e-05) (hash(x)=7266655)
856
+ 820 train 7.644639 (lr=2.3391e-05) (hash(x)=6015975)
857
+ 821 train 7.643714 (lr=2.3250e-05) (hash(x)=5696696)
858
+ 822 train 7.653847 (lr=2.3110e-05) (hash(x)=5411666)
859
+ 823 train 7.917379 (lr=2.2971e-05) (hash(x)=7072404)
860
+ 824 train 7.545630 (lr=2.2832e-05) (hash(x)=4910095)
861
+ 825 train 7.676866 (lr=2.2694e-05) (hash(x)=6590657)
862
+ 826 train 7.625092 (lr=2.2556e-05) (hash(x)=7665574)
863
+ 827 train 7.653700 (lr=2.2420e-05) (hash(x)=6626459)
864
+ 828 train 7.692810 (lr=2.2283e-05) (hash(x)=6731971)
865
+ 829 train 7.697874 (lr=2.2148e-05) (hash(x)=5644977)
866
+ 830 train 7.651792 (lr=2.2013e-05) (hash(x)=6238654)
867
+ 831 train 7.536035 (lr=2.1878e-05) (hash(x)=6556025)
868
+ 832 train 7.869701 (lr=2.1745e-05) (hash(x)=7175563)
869
+ 833 train 7.747357 (lr=2.1612e-05) (hash(x)=6407492)
870
+ 834 train 7.547194 (lr=2.1479e-05) (hash(x)=4934335)
871
+ 835 train 7.569844 (lr=2.1347e-05) (hash(x)=5942129)
872
+ 836 train 7.612617 (lr=2.1216e-05) (hash(x)=5995643)
873
+ 837 train 7.668581 (lr=2.1086e-05) (hash(x)=6012163)
874
+ 838 train 7.780588 (lr=2.0956e-05) (hash(x)=8464831)
875
+ 839 train 8.026137 (lr=2.0827e-05) (hash(x)=7325027)
876
+ 840 train 7.866978 (lr=2.0698e-05) (hash(x)=6785865)
877
+ 841 train 7.787920 (lr=2.0570e-05) (hash(x)=4425520)
878
+ 842 train 7.563994 (lr=2.0443e-05) (hash(x)=5388267)
879
+ 843 train 7.843695 (lr=2.0317e-05) (hash(x)=7322467)
880
+ 844 train 7.630819 (lr=2.0191e-05) (hash(x)=6681766)
881
+ 845 train 7.707881 (lr=2.0066e-05) (hash(x)=7482800)
882
+ 846 train 7.641542 (lr=1.9941e-05) (hash(x)=5554493)
883
+ 847 train 7.844727 (lr=1.9818e-05) (hash(x)=6373412)
884
+ 848 train 7.224220 (lr=1.9694e-05) (hash(x)=5117517)
885
+ 849 train 7.511316 (lr=1.9572e-05) (hash(x)=6981426)
886
+ 850 val loss 7.6332
887
+ 850 val perplexity 2065.5840
888
+ 850 train 7.484658 (lr=1.9450e-05) (hash(x)=6886188)
889
+ 851 train 7.703585 (lr=1.9329e-05) (hash(x)=7332255)
890
+ 852 train 7.701629 (lr=1.9209e-05) (hash(x)=6172042)
891
+ 853 train 7.831582 (lr=1.9089e-05) (hash(x)=5930894)
892
+ 854 train 7.905203 (lr=1.8970e-05) (hash(x)=7448958)
893
+ 855 train 7.657560 (lr=1.8852e-05) (hash(x)=5262868)
894
+ 856 train 7.558040 (lr=1.8734e-05) (hash(x)=5558427)
895
+ 857 train 7.450369 (lr=1.8618e-05) (hash(x)=5585769)
896
+ 858 train 7.357178 (lr=1.8501e-05) (hash(x)=5838081)
897
+ 859 train 7.494993 (lr=1.8386e-05) (hash(x)=5688247)
898
+ 860 train 7.639703 (lr=1.8271e-05) (hash(x)=5162020)
899
+ 861 train 7.808998 (lr=1.8157e-05) (hash(x)=7462079)
900
+ 862 train 7.657056 (lr=1.8044e-05) (hash(x)=6516108)
901
+ 863 train 7.453790 (lr=1.7931e-05) (hash(x)=8055563)
902
+ 864 train 7.650894 (lr=1.7819e-05) (hash(x)=6271901)
903
+ 865 train 7.700618 (lr=1.7708e-05) (hash(x)=6221701)
904
+ 866 train 7.671622 (lr=1.7598e-05) (hash(x)=5772861)
905
+ 867 train 7.663960 (lr=1.7488e-05) (hash(x)=5352405)
906
+ 868 train 7.704040 (lr=1.7379e-05) (hash(x)=6111630)
907
+ 869 train 7.744565 (lr=1.7271e-05) (hash(x)=6730666)
908
+ 870 train 7.905500 (lr=1.7163e-05) (hash(x)=8698315)
909
+ 871 train 7.582049 (lr=1.7056e-05) (hash(x)=5932790)
910
+ 872 train 7.950242 (lr=1.6950e-05) (hash(x)=8099892)
911
+ 873 train 7.608383 (lr=1.6845e-05) (hash(x)=6223114)
912
+ 874 train 7.724115 (lr=1.6740e-05) (hash(x)=5798363)
913
+ 875 train 7.578392 (lr=1.6636e-05) (hash(x)=6249312)
914
+ 876 train 7.691341 (lr=1.6533e-05) (hash(x)=6929692)
915
+ 877 train 7.779414 (lr=1.6431e-05) (hash(x)=7242827)
916
+ 878 train 7.549809 (lr=1.6329e-05) (hash(x)=6332123)
917
+ 879 train 7.446431 (lr=1.6228e-05) (hash(x)=5680154)
918
+ 880 train 7.861735 (lr=1.6128e-05) (hash(x)=6352331)
919
+ 881 train 7.642223 (lr=1.6029e-05) (hash(x)=6332332)
920
+ 882 train 7.739895 (lr=1.5930e-05) (hash(x)=6180674)
921
+ 883 train 7.662111 (lr=1.5832e-05) (hash(x)=8156280)
922
+ 884 train 7.728573 (lr=1.5735e-05) (hash(x)=5874247)
923
+ 885 train 7.646024 (lr=1.5639e-05) (hash(x)=6659781)
924
+ 886 train 7.771195 (lr=1.5543e-05) (hash(x)=5780147)
925
+ 887 train 7.569872 (lr=1.5449e-05) (hash(x)=5914217)
926
+ 888 train 7.413466 (lr=1.5354e-05) (hash(x)=5411762)
927
+ 889 train 7.716517 (lr=1.5261e-05) (hash(x)=6714110)
928
+ 890 train 7.399083 (lr=1.5169e-05) (hash(x)=5999685)
929
+ 891 train 7.613256 (lr=1.5077e-05) (hash(x)=7120215)
930
+ 892 train 7.865321 (lr=1.4986e-05) (hash(x)=6546587)
931
+ 893 train 7.609962 (lr=1.4896e-05) (hash(x)=6593413)
932
+ 894 train 7.720051 (lr=1.4806e-05) (hash(x)=6369927)
933
+ 895 train 7.619521 (lr=1.4718e-05) (hash(x)=6424128)
934
+ 896 train 7.580497 (lr=1.4630e-05) (hash(x)=5736158)
935
+ 897 train 7.386103 (lr=1.4543e-05) (hash(x)=5903043)
936
+ 898 train 7.520309 (lr=1.4456e-05) (hash(x)=4419128)
937
+ 899 train 7.322476 (lr=1.4371e-05) (hash(x)=4390027)
938
+ 900 val loss 7.6208
939
+ 900 val perplexity 2040.2112
940
+ 900 train 7.581107 (lr=1.4286e-05) (hash(x)=6728135)
941
+ 901 train 7.679765 (lr=1.4202e-05) (hash(x)=6945760)
942
+ 902 train 7.662554 (lr=1.4119e-05) (hash(x)=6081534)
943
+ 903 train 7.731725 (lr=1.4037e-05) (hash(x)=7804089)
944
+ 904 train 7.752877 (lr=1.3955e-05) (hash(x)=6225832)
945
+ 905 train 7.814371 (lr=1.3874e-05) (hash(x)=6273417)
946
+ 906 train 7.853147 (lr=1.3794e-05) (hash(x)=7775633)
947
+ 907 train 7.828918 (lr=1.3715e-05) (hash(x)=7130267)
948
+ 908 train 7.610059 (lr=1.3637e-05) (hash(x)=6554076)
949
+ 909 train 7.722689 (lr=1.3559e-05) (hash(x)=6140697)
950
+ 910 train 7.545349 (lr=1.3482e-05) (hash(x)=6128181)
951
+ 911 train 7.611325 (lr=1.3407e-05) (hash(x)=6490149)
952
+ 912 train 7.206694 (lr=1.3331e-05) (hash(x)=5422426)
953
+ 913 train 7.624938 (lr=1.3257e-05) (hash(x)=4847733)
954
+ 914 train 7.637171 (lr=1.3183e-05) (hash(x)=6349747)
955
+ 915 train 7.892220 (lr=1.3111e-05) (hash(x)=9276378)
956
+ 916 train 7.525454 (lr=1.3039e-05) (hash(x)=6869524)
957
+ 917 train 7.567428 (lr=1.2968e-05) (hash(x)=7082053)
958
+ 918 train 7.616925 (lr=1.2897e-05) (hash(x)=7799351)
959
+ 919 train 7.784974 (lr=1.2828e-05) (hash(x)=6283540)
960
+ 920 train 7.567987 (lr=1.2759e-05) (hash(x)=6660860)
961
+ 921 train 7.517439 (lr=1.2691e-05) (hash(x)=5200956)
962
+ 922 train 7.606358 (lr=1.2624e-05) (hash(x)=5653241)
963
+ 923 train 7.746596 (lr=1.2558e-05) (hash(x)=5781680)
964
+ 924 train 7.740694 (lr=1.2493e-05) (hash(x)=6146666)
965
+ 925 train 7.567626 (lr=1.2428e-05) (hash(x)=8652907)
966
+ 926 train 7.372133 (lr=1.2364e-05) (hash(x)=5311546)
967
+ 927 train 7.534654 (lr=1.2301e-05) (hash(x)=5728337)
968
+ 928 train 7.510073 (lr=1.2239e-05) (hash(x)=5413240)
969
+ 929 train 7.619895 (lr=1.2178e-05) (hash(x)=5832401)
970
+ 930 train 7.346562 (lr=1.2118e-05) (hash(x)=7420230)
971
+ 931 train 7.694197 (lr=1.2058e-05) (hash(x)=6278202)
972
+ 932 train 7.596210 (lr=1.1999e-05) (hash(x)=6189873)
973
+ 933 train 7.695422 (lr=1.1941e-05) (hash(x)=7403868)
974
+ 934 train 7.566601 (lr=1.1884e-05) (hash(x)=6867459)
975
+ 935 train 7.738658 (lr=1.1828e-05) (hash(x)=7400272)
976
+ 936 train 7.764461 (lr=1.1772e-05) (hash(x)=7936275)
977
+ 937 train 7.474885 (lr=1.1718e-05) (hash(x)=5943413)
978
+ 938 train 7.704277 (lr=1.1664e-05) (hash(x)=5442166)
979
+ 939 train 7.612360 (lr=1.1611e-05) (hash(x)=5795128)
980
+ 940 train 7.822320 (lr=1.1559e-05) (hash(x)=6831680)
981
+ 941 train 7.505625 (lr=1.1508e-05) (hash(x)=5519980)
982
+ 942 train 7.525065 (lr=1.1457e-05) (hash(x)=6765461)
983
+ 943 train 7.490722 (lr=1.1408e-05) (hash(x)=6473432)
984
+ 944 train 7.465393 (lr=1.1359e-05) (hash(x)=6490395)
985
+ 945 train 7.668106 (lr=1.1311e-05) (hash(x)=5484669)
986
+ 946 train 7.652201 (lr=1.1264e-05) (hash(x)=7278970)
987
+ 947 train 7.638943 (lr=1.1218e-05) (hash(x)=5201502)
988
+ 948 train 7.883686 (lr=1.1173e-05) (hash(x)=7277915)
989
+ 949 train 7.574278 (lr=1.1128e-05) (hash(x)=7227057)
990
+ 950 val loss 7.6126
991
+ 950 val perplexity 2023.5592
992
+ 950 train 7.816874 (lr=1.1085e-05) (hash(x)=8178996)
993
+ 951 train 7.943238 (lr=1.1042e-05) (hash(x)=6922646)
994
+ 952 train 7.490936 (lr=1.1000e-05) (hash(x)=5941772)
995
+ 953 train 7.450648 (lr=1.0959e-05) (hash(x)=6056673)
996
+ 954 train 7.846048 (lr=1.0919e-05) (hash(x)=6912178)
997
+ 955 train 7.370865 (lr=1.0879e-05) (hash(x)=5484641)
998
+ 956 train 7.709154 (lr=1.0841e-05) (hash(x)=5121117)
999
+ 957 train 7.756188 (lr=1.0803e-05) (hash(x)=7100865)
1000
+ 958 train 7.451260 (lr=1.0766e-05) (hash(x)=5334280)
1001
+ 959 train 7.519156 (lr=1.0730e-05) (hash(x)=7277141)
1002
+ 960 train 7.513563 (lr=1.0695e-05) (hash(x)=6401960)
1003
+ 961 train 7.586699 (lr=1.0661e-05) (hash(x)=6407603)
1004
+ 962 train 7.734699 (lr=1.0628e-05) (hash(x)=5786089)
1005
+ 963 train 7.592095 (lr=1.0595e-05) (hash(x)=5604614)
1006
+ 964 train 7.613576 (lr=1.0563e-05) (hash(x)=5454287)
1007
+ 965 train 7.711183 (lr=1.0533e-05) (hash(x)=6200612)
1008
+ 966 train 7.699334 (lr=1.0503e-05) (hash(x)=5691759)
1009
+ 967 train 7.649145 (lr=1.0474e-05) (hash(x)=7443124)
1010
+ 968 train 7.682613 (lr=1.0445e-05) (hash(x)=6711366)
1011
+ 969 train 7.596057 (lr=1.0418e-05) (hash(x)=6728764)
1012
+ 970 train 7.881885 (lr=1.0391e-05) (hash(x)=6870240)
1013
+ 971 train 7.569551 (lr=1.0366e-05) (hash(x)=5601968)
1014
+ 972 train 7.376374 (lr=1.0341e-05) (hash(x)=5311504)
1015
+ 973 train 7.607464 (lr=1.0317e-05) (hash(x)=6877761)
1016
+ 974 train 7.706609 (lr=1.0294e-05) (hash(x)=6457033)
1017
+ 975 train 7.652069 (lr=1.0272e-05) (hash(x)=5632735)
1018
+ 976 train 7.830911 (lr=1.0251e-05) (hash(x)=6024880)
1019
+ 977 train 7.535464 (lr=1.0230e-05) (hash(x)=5658845)
1020
+ 978 train 7.415770 (lr=1.0211e-05) (hash(x)=6089461)
1021
+ 979 train 7.500862 (lr=1.0192e-05) (hash(x)=5594810)
1022
+ 980 train 7.277288 (lr=1.0174e-05) (hash(x)=4667026)
1023
+ 981 train 7.139160 (lr=1.0157e-05) (hash(x)=4738523)
1024
+ 982 train 7.447637 (lr=1.0141e-05) (hash(x)=7451876)
1025
+ 983 train 7.724800 (lr=1.0126e-05) (hash(x)=5881015)
1026
+ 984 train 7.693962 (lr=1.0111e-05) (hash(x)=6255031)
1027
+ 985 train 7.523418 (lr=1.0098e-05) (hash(x)=5263325)
1028
+ 986 train 7.519240 (lr=1.0085e-05) (hash(x)=6654585)
1029
+ 987 train 7.613875 (lr=1.0074e-05) (hash(x)=6865361)
1030
+ 988 train 7.703878 (lr=1.0063e-05) (hash(x)=6978986)
1031
+ 989 train 7.798944 (lr=1.0053e-05) (hash(x)=5644619)
1032
+ 990 train 7.607509 (lr=1.0044e-05) (hash(x)=6551023)
1033
+ 991 train 7.585578 (lr=1.0035e-05) (hash(x)=5809411)
1034
+ 992 train 7.633374 (lr=1.0028e-05) (hash(x)=5501837)
1035
+ 993 train 7.642830 (lr=1.0021e-05) (hash(x)=7044374)
1036
+ 994 train 7.621263 (lr=1.0016e-05) (hash(x)=6526130)
1037
+ 995 train 7.740894 (lr=1.0011e-05) (hash(x)=6293025)
1038
+ 996 train 7.720303 (lr=1.0007e-05) (hash(x)=6295768)
1039
+ 997 train 7.535273 (lr=1.0004e-05) (hash(x)=5189023)
1040
+ 998 train 7.353145 (lr=1.0002e-05) (hash(x)=6255467)
1041
+ 999 val loss 7.6086
1042
+ 999 val perplexity 2015.3846
1043
+ 999 train 7.723579 (lr=1.0000e-05) (hash(x)=7232964)
lr1e-4_total_batch_size5120_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04aa835b83db0c2c1bce1164bf5426b63c597fd82a3588c41ab75791b00dc114
3
+ size 36163010
lr1e-4_total_batch_size5120_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1138eb1ef679f3179a8ea17df62284728f1a68e8be154900713a3f3bdf87273c
3
+ size 70764358