andrew-healey commited on
Commit
027decb
·
verified ·
1 Parent(s): d196106

Upload folder using huggingface_hub

Browse files
lr2e-4_total_batch_size5120_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr2e-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", "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": 0.0002, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr2e-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
lr2e-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=6.9930e-07) (hash(x)=22886834)
5
+ 1 train 11.786522 (lr=1.3986e-06) (hash(x)=26375038)
6
+ 2 train 11.726857 (lr=2.0979e-06) (hash(x)=30598777)
7
+ 3 train 11.773165 (lr=2.7972e-06) (hash(x)=27234506)
8
+ 4 train 11.729205 (lr=3.4965e-06) (hash(x)=27767880)
9
+ 5 train 11.717231 (lr=4.1958e-06) (hash(x)=23702020)
10
+ 6 train 11.762368 (lr=4.8951e-06) (hash(x)=31986844)
11
+ 7 train 11.759068 (lr=5.5944e-06) (hash(x)=20782690)
12
+ 8 train 11.753537 (lr=6.2937e-06) (hash(x)=25201599)
13
+ 9 train 11.778136 (lr=6.9930e-06) (hash(x)=23094976)
14
+ 10 train 11.727038 (lr=7.6923e-06) (hash(x)=23841096)
15
+ 11 train 11.745360 (lr=8.3916e-06) (hash(x)=26532095)
16
+ 12 train 11.718360 (lr=9.0909e-06) (hash(x)=24432298)
17
+ 13 train 11.712705 (lr=9.7902e-06) (hash(x)=27151649)
18
+ 14 train 11.696010 (lr=1.0490e-05) (hash(x)=24596846)
19
+ 15 train 11.650504 (lr=1.1189e-05) (hash(x)=23890908)
20
+ 16 train 11.646593 (lr=1.1888e-05) (hash(x)=28913955)
21
+ 17 train 11.633705 (lr=1.2587e-05) (hash(x)=25588236)
22
+ 18 train 11.640189 (lr=1.3287e-05) (hash(x)=23770034)
23
+ 19 train 11.629020 (lr=1.3986e-05) (hash(x)=24011372)
24
+ 20 train 11.570956 (lr=1.4685e-05) (hash(x)=25441898)
25
+ 21 train 11.558985 (lr=1.5385e-05) (hash(x)=28375581)
26
+ 22 train 11.558283 (lr=1.6084e-05) (hash(x)=24046679)
27
+ 23 train 11.581992 (lr=1.6783e-05) (hash(x)=24611628)
28
+ 24 train 11.559752 (lr=1.7483e-05) (hash(x)=26169030)
29
+ 25 train 11.453639 (lr=1.8182e-05) (hash(x)=30298407)
30
+ 26 train 11.481302 (lr=1.8881e-05) (hash(x)=23711112)
31
+ 27 train 11.402436 (lr=1.9580e-05) (hash(x)=19245352)
32
+ 28 train 11.387091 (lr=2.0280e-05) (hash(x)=21529136)
33
+ 29 train 11.403493 (lr=2.0979e-05) (hash(x)=28936608)
34
+ 30 train 11.357623 (lr=2.1678e-05) (hash(x)=24339013)
35
+ 31 train 11.231040 (lr=2.2378e-05) (hash(x)=25767553)
36
+ 32 train 11.247643 (lr=2.3077e-05) (hash(x)=26439905)
37
+ 33 train 11.319613 (lr=2.3776e-05) (hash(x)=31093473)
38
+ 34 train 11.150300 (lr=2.4476e-05) (hash(x)=25450374)
39
+ 35 train 11.210859 (lr=2.5175e-05) (hash(x)=24809873)
40
+ 36 train 11.183257 (lr=2.5874e-05) (hash(x)=23253252)
41
+ 37 train 11.179575 (lr=2.6573e-05) (hash(x)=27852919)
42
+ 38 train 11.071074 (lr=2.7273e-05) (hash(x)=23327497)
43
+ 39 train 11.023358 (lr=2.7972e-05) (hash(x)=22512166)
44
+ 40 train 11.080914 (lr=2.8671e-05) (hash(x)=22859419)
45
+ 41 train 10.925023 (lr=2.9371e-05) (hash(x)=27620338)
46
+ 42 train 10.890419 (lr=3.0070e-05) (hash(x)=26397837)
47
+ 43 train 10.913859 (lr=3.0769e-05) (hash(x)=28092148)
48
+ 44 train 10.837189 (lr=3.1469e-05) (hash(x)=24662703)
49
+ 45 train 10.844550 (lr=3.2168e-05) (hash(x)=27938767)
50
+ 46 train 10.882991 (lr=3.2867e-05) (hash(x)=26037988)
51
+ 47 train 10.705787 (lr=3.3566e-05) (hash(x)=24732833)
52
+ 48 train 10.698764 (lr=3.4266e-05) (hash(x)=25259526)
53
+ 49 train 10.708899 (lr=3.4965e-05) (hash(x)=23200230)
54
+ 50 val loss 10.6598
55
+ 50 val perplexity 42609.6016
56
+ 50 train 10.672421 (lr=3.5664e-05) (hash(x)=26721357)
57
+ 51 train 10.677696 (lr=3.6364e-05) (hash(x)=22694718)
58
+ 52 train 10.581690 (lr=3.7063e-05) (hash(x)=28066766)
59
+ 53 train 10.562861 (lr=3.7762e-05) (hash(x)=23125151)
60
+ 54 train 10.591826 (lr=3.8462e-05) (hash(x)=27193725)
61
+ 55 train 10.483047 (lr=3.9161e-05) (hash(x)=25129410)
62
+ 56 train 10.513149 (lr=3.9860e-05) (hash(x)=24263988)
63
+ 57 train 10.379344 (lr=4.0559e-05) (hash(x)=23059154)
64
+ 58 train 10.476598 (lr=4.1259e-05) (hash(x)=26063864)
65
+ 59 train 10.615381 (lr=4.1958e-05) (hash(x)=27858570)
66
+ 60 train 10.403590 (lr=4.2657e-05) (hash(x)=23874620)
67
+ 61 train 10.412221 (lr=4.3357e-05) (hash(x)=22402617)
68
+ 62 train 10.397899 (lr=4.4056e-05) (hash(x)=23600822)
69
+ 63 train 10.213144 (lr=4.4755e-05) (hash(x)=26582391)
70
+ 64 train 10.241872 (lr=4.5455e-05) (hash(x)=23225283)
71
+ 65 train 10.402961 (lr=4.6154e-05) (hash(x)=26075451)
72
+ 66 train 10.354083 (lr=4.6853e-05) (hash(x)=24723419)
73
+ 67 train 10.275083 (lr=4.7552e-05) (hash(x)=27279806)
74
+ 68 train 10.215707 (lr=4.8252e-05) (hash(x)=25870391)
75
+ 69 train 10.296074 (lr=4.8951e-05) (hash(x)=26188136)
76
+ 70 train 10.219914 (lr=4.9650e-05) (hash(x)=30373443)
77
+ 71 train 10.317218 (lr=5.0350e-05) (hash(x)=26472336)
78
+ 72 train 10.218242 (lr=5.1049e-05) (hash(x)=26651572)
79
+ 73 train 10.174848 (lr=5.1748e-05) (hash(x)=26376212)
80
+ 74 train 10.173363 (lr=5.2448e-05) (hash(x)=26733350)
81
+ 75 train 10.300672 (lr=5.3147e-05) (hash(x)=28301589)
82
+ 76 train 10.172812 (lr=5.3846e-05) (hash(x)=27599559)
83
+ 77 train 10.094896 (lr=5.4545e-05) (hash(x)=28035221)
84
+ 78 train 10.174782 (lr=5.5245e-05) (hash(x)=25016783)
85
+ 79 train 10.197110 (lr=5.5944e-05) (hash(x)=27654289)
86
+ 80 train 10.226367 (lr=5.6643e-05) (hash(x)=24597558)
87
+ 81 train 10.171132 (lr=5.7343e-05) (hash(x)=21560904)
88
+ 82 train 9.989909 (lr=5.8042e-05) (hash(x)=21983837)
89
+ 83 train 10.051674 (lr=5.8741e-05) (hash(x)=24995715)
90
+ 84 train 10.109978 (lr=5.9441e-05) (hash(x)=29876413)
91
+ 85 train 10.031592 (lr=6.0140e-05) (hash(x)=23792508)
92
+ 86 train 9.953173 (lr=6.0839e-05) (hash(x)=25509120)
93
+ 87 train 10.126330 (lr=6.1538e-05) (hash(x)=26559876)
94
+ 88 train 10.043777 (lr=6.2238e-05) (hash(x)=23569647)
95
+ 89 train 10.058722 (lr=6.2937e-05) (hash(x)=25758852)
96
+ 90 train 9.943398 (lr=6.3636e-05) (hash(x)=25706298)
97
+ 91 train 9.849136 (lr=6.4336e-05) (hash(x)=28364895)
98
+ 92 train 9.893332 (lr=6.5035e-05) (hash(x)=25304663)
99
+ 93 train 9.974350 (lr=6.5734e-05) (hash(x)=25269299)
100
+ 94 train 10.112905 (lr=6.6434e-05) (hash(x)=25870566)
101
+ 95 train 9.896851 (lr=6.7133e-05) (hash(x)=21770329)
102
+ 96 train 10.107920 (lr=6.7832e-05) (hash(x)=27595900)
103
+ 97 train 10.109680 (lr=6.8531e-05) (hash(x)=24785397)
104
+ 98 train 9.869967 (lr=6.9231e-05) (hash(x)=21521480)
105
+ 99 train 9.913299 (lr=6.9930e-05) (hash(x)=24628606)
106
+ 100 val loss 9.9170
107
+ 100 val perplexity 20271.3438
108
+ 100 train 9.947840 (lr=7.0629e-05) (hash(x)=24670150)
109
+ 101 train 9.946340 (lr=7.1329e-05) (hash(x)=23181910)
110
+ 102 train 9.804832 (lr=7.2028e-05) (hash(x)=22714991)
111
+ 103 train 9.908710 (lr=7.2727e-05) (hash(x)=22723459)
112
+ 104 train 9.838172 (lr=7.3427e-05) (hash(x)=21524316)
113
+ 105 train 9.824501 (lr=7.4126e-05) (hash(x)=25506632)
114
+ 106 train 9.649211 (lr=7.4825e-05) (hash(x)=21675672)
115
+ 107 train 9.767374 (lr=7.5524e-05) (hash(x)=22897919)
116
+ 108 train 9.780364 (lr=7.6224e-05) (hash(x)=23321631)
117
+ 109 train 9.803567 (lr=7.6923e-05) (hash(x)=26546719)
118
+ 110 train 10.338003 (lr=7.7622e-05) (hash(x)=31962348)
119
+ 111 train 9.792398 (lr=7.8322e-05) (hash(x)=30338342)
120
+ 112 train 9.764856 (lr=7.9021e-05) (hash(x)=23724471)
121
+ 113 train 9.752960 (lr=7.9720e-05) (hash(x)=29175888)
122
+ 114 train 9.868297 (lr=8.0420e-05) (hash(x)=23256716)
123
+ 115 train 9.808032 (lr=8.1119e-05) (hash(x)=27063280)
124
+ 116 train 9.750420 (lr=8.1818e-05) (hash(x)=31057659)
125
+ 117 train 9.815223 (lr=8.2517e-05) (hash(x)=32915097)
126
+ 118 train 9.724854 (lr=8.3217e-05) (hash(x)=28842717)
127
+ 119 train 9.690181 (lr=8.3916e-05) (hash(x)=25678059)
128
+ 120 train 9.686526 (lr=8.4615e-05) (hash(x)=21593510)
129
+ 121 train 9.556270 (lr=8.5315e-05) (hash(x)=20083773)
130
+ 122 train 9.606078 (lr=8.6014e-05) (hash(x)=23002820)
131
+ 123 train 9.568427 (lr=8.6713e-05) (hash(x)=21853028)
132
+ 124 train 9.604660 (lr=8.7413e-05) (hash(x)=26985625)
133
+ 125 train 9.520707 (lr=8.8112e-05) (hash(x)=21808483)
134
+ 126 train 9.696350 (lr=8.8811e-05) (hash(x)=28873251)
135
+ 127 train 9.588122 (lr=8.9510e-05) (hash(x)=26109335)
136
+ 128 train 9.610017 (lr=9.0210e-05) (hash(x)=26334674)
137
+ 129 train 9.575045 (lr=9.0909e-05) (hash(x)=24916754)
138
+ 130 train 9.599522 (lr=9.1608e-05) (hash(x)=25449624)
139
+ 131 train 9.667902 (lr=9.2308e-05) (hash(x)=25334848)
140
+ 132 train 9.730344 (lr=9.3007e-05) (hash(x)=27484863)
141
+ 133 train 9.521794 (lr=9.3706e-05) (hash(x)=24917705)
142
+ 134 train 9.384773 (lr=9.4406e-05) (hash(x)=25143449)
143
+ 135 train 9.435587 (lr=9.5105e-05) (hash(x)=25044885)
144
+ 136 train 9.643442 (lr=9.5804e-05) (hash(x)=27821028)
145
+ 137 train 9.568672 (lr=9.6503e-05) (hash(x)=28747022)
146
+ 138 train 9.512908 (lr=9.7203e-05) (hash(x)=27182888)
147
+ 139 train 9.379439 (lr=9.7902e-05) (hash(x)=23678349)
148
+ 140 train 9.497525 (lr=9.8601e-05) (hash(x)=23593235)
149
+ 141 train 9.620494 (lr=9.9301e-05) (hash(x)=28529813)
150
+ 142 train 9.707245 (lr=1.0000e-04) (hash(x)=32074661)
151
+ 143 train 9.387294 (lr=1.0070e-04) (hash(x)=28870690)
152
+ 144 train 9.445533 (lr=1.0140e-04) (hash(x)=27307705)
153
+ 145 train 9.392129 (lr=1.0210e-04) (hash(x)=25044834)
154
+ 146 train 9.441619 (lr=1.0280e-04) (hash(x)=23712023)
155
+ 147 train 9.634463 (lr=1.0350e-04) (hash(x)=32982615)
156
+ 148 train 9.459651 (lr=1.0420e-04) (hash(x)=30113660)
157
+ 149 train 9.218554 (lr=1.0490e-04) (hash(x)=20970960)
158
+ 150 val loss 9.2831
159
+ 150 val perplexity 10755.0908
160
+ 150 train 9.249382 (lr=1.0559e-04) (hash(x)=23132684)
161
+ 151 train 10.230716 (lr=1.0629e-04) (hash(x)=35279941)
162
+ 152 train 9.746226 (lr=1.0699e-04) (hash(x)=31227444)
163
+ 153 train 9.202677 (lr=1.0769e-04) (hash(x)=25529472)
164
+ 154 train 9.201354 (lr=1.0839e-04) (hash(x)=24350409)
165
+ 155 train 9.459618 (lr=1.0909e-04) (hash(x)=26400041)
166
+ 156 train 9.159389 (lr=1.0979e-04) (hash(x)=25262621)
167
+ 157 train 9.154494 (lr=1.1049e-04) (hash(x)=24656138)
168
+ 158 train 9.207989 (lr=1.1119e-04) (hash(x)=26803414)
169
+ 159 train 9.144874 (lr=1.1189e-04) (hash(x)=25015923)
170
+ 160 train 8.999480 (lr=1.1259e-04) (hash(x)=23581172)
171
+ 161 train 9.057517 (lr=1.1329e-04) (hash(x)=22924885)
172
+ 162 train 9.004022 (lr=1.1399e-04) (hash(x)=23414296)
173
+ 163 train 9.282326 (lr=1.1469e-04) (hash(x)=24853586)
174
+ 164 train 9.267085 (lr=1.1538e-04) (hash(x)=25000130)
175
+ 165 train 9.086567 (lr=1.1608e-04) (hash(x)=27004780)
176
+ 166 train 9.049629 (lr=1.1678e-04) (hash(x)=26148573)
177
+ 167 train 8.988647 (lr=1.1748e-04) (hash(x)=26740855)
178
+ 168 train 8.706361 (lr=1.1818e-04) (hash(x)=20965419)
179
+ 169 train 8.852076 (lr=1.1888e-04) (hash(x)=23950114)
180
+ 170 train 8.825558 (lr=1.1958e-04) (hash(x)=24951982)
181
+ 171 train 8.930049 (lr=1.2028e-04) (hash(x)=24584116)
182
+ 172 train 8.940637 (lr=1.2098e-04) (hash(x)=24378759)
183
+ 173 train 8.826471 (lr=1.2168e-04) (hash(x)=25718516)
184
+ 174 train 8.893836 (lr=1.2238e-04) (hash(x)=28424396)
185
+ 175 train 8.759524 (lr=1.2308e-04) (hash(x)=22262151)
186
+ 176 train 8.896358 (lr=1.2378e-04) (hash(x)=26438412)
187
+ 177 train 8.743075 (lr=1.2448e-04) (hash(x)=23025303)
188
+ 178 train 8.666364 (lr=1.2517e-04) (hash(x)=24190770)
189
+ 179 train 8.775293 (lr=1.2587e-04) (hash(x)=26627860)
190
+ 180 train 8.733484 (lr=1.2657e-04) (hash(x)=23663439)
191
+ 181 train 8.531133 (lr=1.2727e-04) (hash(x)=23700532)
192
+ 182 train 8.602875 (lr=1.2797e-04) (hash(x)=23075676)
193
+ 183 train 8.675462 (lr=1.2867e-04) (hash(x)=26621834)
194
+ 184 train 8.671355 (lr=1.2937e-04) (hash(x)=29426269)
195
+ 185 train 8.554217 (lr=1.3007e-04) (hash(x)=21821465)
196
+ 186 train 8.611747 (lr=1.3077e-04) (hash(x)=26506130)
197
+ 187 train 8.801375 (lr=1.3147e-04) (hash(x)=26930630)
198
+ 188 train 8.557786 (lr=1.3217e-04) (hash(x)=22993793)
199
+ 189 train 8.517845 (lr=1.3287e-04) (hash(x)=19557946)
200
+ 190 train 8.674314 (lr=1.3357e-04) (hash(x)=23572891)
201
+ 191 train 8.456822 (lr=1.3427e-04) (hash(x)=23234741)
202
+ 192 train 8.532045 (lr=1.3497e-04) (hash(x)=25547951)
203
+ 193 train 8.523528 (lr=1.3566e-04) (hash(x)=26713563)
204
+ 194 train 8.547825 (lr=1.3636e-04) (hash(x)=25913622)
205
+ 195 train 8.718466 (lr=1.3706e-04) (hash(x)=28102443)
206
+ 196 train 8.544539 (lr=1.3776e-04) (hash(x)=23093351)
207
+ 197 train 8.994954 (lr=1.3846e-04) (hash(x)=31689122)
208
+ 198 train 8.894501 (lr=1.3916e-04) (hash(x)=32252517)
209
+ 199 train 8.532555 (lr=1.3986e-04) (hash(x)=25470563)
210
+ 200 val loss 8.4982
211
+ 200 val perplexity 4905.9277
212
+ 200 train 8.594085 (lr=1.4056e-04) (hash(x)=25597614)
213
+ 201 train 8.577597 (lr=1.4126e-04) (hash(x)=23757479)
214
+ 202 train 8.417030 (lr=1.4196e-04) (hash(x)=25140048)
215
+ 203 train 8.572240 (lr=1.4266e-04) (hash(x)=28282861)
216
+ 204 train 8.393707 (lr=1.4336e-04) (hash(x)=24754885)
217
+ 205 train 8.576873 (lr=1.4406e-04) (hash(x)=26731964)
218
+ 206 train 8.632682 (lr=1.4476e-04) (hash(x)=26660561)
219
+ 207 train 8.349551 (lr=1.4545e-04) (hash(x)=21799102)
220
+ 208 train 8.440106 (lr=1.4615e-04) (hash(x)=27025986)
221
+ 209 train 8.296119 (lr=1.4685e-04) (hash(x)=24672077)
222
+ 210 train 8.384494 (lr=1.4755e-04) (hash(x)=25322984)
223
+ 211 train 8.283589 (lr=1.4825e-04) (hash(x)=23471769)
224
+ 212 train 8.067083 (lr=1.4895e-04) (hash(x)=20766491)
225
+ 213 train 8.592277 (lr=1.4965e-04) (hash(x)=24058931)
226
+ 214 train 8.250019 (lr=1.5035e-04) (hash(x)=23872843)
227
+ 215 train 8.301787 (lr=1.5105e-04) (hash(x)=23275479)
228
+ 216 train 8.403808 (lr=1.5175e-04) (hash(x)=24914695)
229
+ 217 train 8.284886 (lr=1.5245e-04) (hash(x)=24364396)
230
+ 218 train 8.372843 (lr=1.5315e-04) (hash(x)=27986474)
231
+ 219 train 8.379944 (lr=1.5385e-04) (hash(x)=24473581)
232
+ 220 train 8.317241 (lr=1.5455e-04) (hash(x)=22974689)
233
+ 221 train 8.359950 (lr=1.5524e-04) (hash(x)=23774644)
234
+ 222 train 8.364591 (lr=1.5594e-04) (hash(x)=23245327)
235
+ 223 train 8.176744 (lr=1.5664e-04) (hash(x)=22091862)
236
+ 224 train 8.160067 (lr=1.5734e-04) (hash(x)=24362839)
237
+ 225 train 8.184381 (lr=1.5804e-04) (hash(x)=25482303)
238
+ 226 train 8.345274 (lr=1.5874e-04) (hash(x)=24911853)
239
+ 227 train 8.325546 (lr=1.5944e-04) (hash(x)=26018202)
240
+ 228 train 8.174924 (lr=1.6014e-04) (hash(x)=26124495)
241
+ 229 train 8.189546 (lr=1.6084e-04) (hash(x)=24560096)
242
+ 230 train 8.209442 (lr=1.6154e-04) (hash(x)=24695331)
243
+ 231 train 8.084067 (lr=1.6224e-04) (hash(x)=17430373)
244
+ 232 train 8.051170 (lr=1.6294e-04) (hash(x)=21813345)
245
+ 233 train 7.691124 (lr=1.6364e-04) (hash(x)=20098681)
246
+ 234 train 7.938356 (lr=1.6434e-04) (hash(x)=25095928)
247
+ 235 train 8.023400 (lr=1.6503e-04) (hash(x)=24078083)
248
+ 236 train 8.066783 (lr=1.6573e-04) (hash(x)=22901505)
249
+ 237 train 8.134725 (lr=1.6643e-04) (hash(x)=26595592)
250
+ 238 train 8.109261 (lr=1.6713e-04) (hash(x)=27663196)
251
+ 239 train 7.915150 (lr=1.6783e-04) (hash(x)=22954861)
252
+ 240 train 8.242521 (lr=1.6853e-04) (hash(x)=30159234)
253
+ 241 train 8.166611 (lr=1.6923e-04) (hash(x)=29294271)
254
+ 242 train 8.153853 (lr=1.6993e-04) (hash(x)=23728322)
255
+ 243 train 8.178614 (lr=1.7063e-04) (hash(x)=28695016)
256
+ 244 train 8.282378 (lr=1.7133e-04) (hash(x)=26702728)
257
+ 245 train 8.075462 (lr=1.7203e-04) (hash(x)=27100115)
258
+ 246 train 8.030471 (lr=1.7273e-04) (hash(x)=25082752)
259
+ 247 train 8.320981 (lr=1.7343e-04) (hash(x)=26671799)
260
+ 248 train 8.036476 (lr=1.7413e-04) (hash(x)=23718946)
261
+ 249 train 8.222040 (lr=1.7483e-04) (hash(x)=28137394)
262
+ 250 val loss 7.9936
263
+ 250 val perplexity 2961.9292
264
+ 250 train 8.233218 (lr=1.7552e-04) (hash(x)=23893495)
265
+ 251 train 8.047023 (lr=1.7622e-04) (hash(x)=23166092)
266
+ 252 train 7.957166 (lr=1.7692e-04) (hash(x)=25907665)
267
+ 253 train 7.925347 (lr=1.7762e-04) (hash(x)=25885986)
268
+ 254 train 7.991336 (lr=1.7832e-04) (hash(x)=25262712)
269
+ 255 train 7.930650 (lr=1.7902e-04) (hash(x)=26924723)
270
+ 256 train 8.111571 (lr=1.7972e-04) (hash(x)=28744736)
271
+ 257 train 7.894122 (lr=1.8042e-04) (hash(x)=26140590)
272
+ 258 train 8.074984 (lr=1.8112e-04) (hash(x)=25780449)
273
+ 259 train 7.919110 (lr=1.8182e-04) (hash(x)=25138659)
274
+ 260 train 8.184468 (lr=1.8252e-04) (hash(x)=27244046)
275
+ 261 train 8.055023 (lr=1.8322e-04) (hash(x)=27224685)
276
+ 262 train 8.264785 (lr=1.8392e-04) (hash(x)=28274477)
277
+ 263 train 7.820229 (lr=1.8462e-04) (hash(x)=23557495)
278
+ 264 train 7.952621 (lr=1.8531e-04) (hash(x)=24680596)
279
+ 265 train 7.948886 (lr=1.8601e-04) (hash(x)=23928957)
280
+ 266 train 7.861025 (lr=1.8671e-04) (hash(x)=23761390)
281
+ 267 train 7.810233 (lr=1.8741e-04) (hash(x)=25288123)
282
+ 268 train 8.035193 (lr=1.8811e-04) (hash(x)=28705502)
283
+ 269 train 7.663564 (lr=1.8881e-04) (hash(x)=23246294)
284
+ 270 train 7.761146 (lr=1.8951e-04) (hash(x)=28639079)
285
+ 271 train 7.669071 (lr=1.9021e-04) (hash(x)=27804380)
286
+ 272 train 7.872932 (lr=1.9091e-04) (hash(x)=24172235)
287
+ 273 train 7.944713 (lr=1.9161e-04) (hash(x)=23089140)
288
+ 274 train 8.133986 (lr=1.9231e-04) (hash(x)=27163701)
289
+ 275 train 8.229663 (lr=1.9301e-04) (hash(x)=26993263)
290
+ 276 train 7.914978 (lr=1.9371e-04) (hash(x)=28224233)
291
+ 277 train 8.244328 (lr=1.9441e-04) (hash(x)=27397203)
292
+ 278 train 8.158372 (lr=1.9510e-04) (hash(x)=27923882)
293
+ 279 train 7.958852 (lr=1.9580e-04) (hash(x)=26654908)
294
+ 280 train 7.944772 (lr=1.9650e-04) (hash(x)=24213147)
295
+ 281 train 7.671964 (lr=1.9720e-04) (hash(x)=21965022)
296
+ 282 train 7.833171 (lr=1.9790e-04) (hash(x)=25465685)
297
+ 283 train 8.019527 (lr=1.9860e-04) (hash(x)=27347722)
298
+ 284 train 8.030728 (lr=1.9930e-04) (hash(x)=26732050)
299
+ 285 train 8.184006 (lr=2.0000e-04) (hash(x)=28314127)
300
+ 286 train 7.727821 (lr=2.0000e-04) (hash(x)=21471186)
301
+ 287 train 7.696778 (lr=2.0000e-04) (hash(x)=23627518)
302
+ 288 train 7.679504 (lr=2.0000e-04) (hash(x)=20870353)
303
+ 289 train 7.751715 (lr=1.9999e-04) (hash(x)=25024764)
304
+ 290 train 7.557401 (lr=1.9999e-04) (hash(x)=20683822)
305
+ 291 train 7.894186 (lr=1.9998e-04) (hash(x)=21768671)
306
+ 292 train 7.888232 (lr=1.9997e-04) (hash(x)=25557309)
307
+ 293 train 8.037185 (lr=1.9996e-04) (hash(x)=25076667)
308
+ 294 train 7.683108 (lr=1.9994e-04) (hash(x)=23765822)
309
+ 295 train 7.669858 (lr=1.9993e-04) (hash(x)=21889990)
310
+ 296 train 7.860289 (lr=1.9991e-04) (hash(x)=26339893)
311
+ 297 train 7.556648 (lr=1.9989e-04) (hash(x)=20932794)
312
+ 298 train 7.691652 (lr=1.9987e-04) (hash(x)=21750070)
313
+ 299 train 7.852615 (lr=1.9985e-04) (hash(x)=23665838)
314
+ 300 val loss 7.8118
315
+ 300 val perplexity 2469.5681
316
+ 300 train 8.620625 (lr=1.9983e-04) (hash(x)=32888061)
317
+ 301 train 8.423040 (lr=1.9980e-04) (hash(x)=30223582)
318
+ 302 train 8.100951 (lr=1.9978e-04) (hash(x)=26908418)
319
+ 303 train 7.449349 (lr=1.9975e-04) (hash(x)=22528001)
320
+ 304 train 8.016787 (lr=1.9972e-04) (hash(x)=27452187)
321
+ 305 train 7.833827 (lr=1.9969e-04) (hash(x)=25181641)
322
+ 306 train 7.922005 (lr=1.9965e-04) (hash(x)=25546593)
323
+ 307 train 7.881425 (lr=1.9962e-04) (hash(x)=22487328)
324
+ 308 train 8.024761 (lr=1.9958e-04) (hash(x)=27804274)
325
+ 309 train 8.094653 (lr=1.9954e-04) (hash(x)=26544630)
326
+ 310 train 8.191751 (lr=1.9950e-04) (hash(x)=27738934)
327
+ 311 train 7.999748 (lr=1.9946e-04) (hash(x)=29248942)
328
+ 312 train 7.764293 (lr=1.9941e-04) (hash(x)=25103452)
329
+ 313 train 7.983594 (lr=1.9937e-04) (hash(x)=25052066)
330
+ 314 train 7.826106 (lr=1.9932e-04) (hash(x)=24481302)
331
+ 315 train 7.768442 (lr=1.9927e-04) (hash(x)=23543273)
332
+ 316 train 7.775046 (lr=1.9922e-04) (hash(x)=25608244)
333
+ 317 train 7.955246 (lr=1.9916e-04) (hash(x)=27451288)
334
+ 318 train 7.612047 (lr=1.9911e-04) (hash(x)=22806491)
335
+ 319 train 7.791568 (lr=1.9905e-04) (hash(x)=25533417)
336
+ 320 train 7.821728 (lr=1.9899e-04) (hash(x)=24557997)
337
+ 321 train 7.874645 (lr=1.9893e-04) (hash(x)=24432899)
338
+ 322 train 8.093554 (lr=1.9887e-04) (hash(x)=27583287)
339
+ 323 train 7.999551 (lr=1.9881e-04) (hash(x)=25552036)
340
+ 324 train 7.720852 (lr=1.9874e-04) (hash(x)=24201868)
341
+ 325 train 8.055707 (lr=1.9868e-04) (hash(x)=28149782)
342
+ 326 train 7.802312 (lr=1.9861e-04) (hash(x)=25529698)
343
+ 327 train 7.456099 (lr=1.9854e-04) (hash(x)=20612533)
344
+ 328 train 7.512290 (lr=1.9847e-04) (hash(x)=20699000)
345
+ 329 train 7.458238 (lr=1.9839e-04) (hash(x)=19774173)
346
+ 330 train 7.510390 (lr=1.9832e-04) (hash(x)=21681646)
347
+ 331 train 7.373506 (lr=1.9824e-04) (hash(x)=20216795)
348
+ 332 train 8.105064 (lr=1.9816e-04) (hash(x)=27697998)
349
+ 333 train 7.838728 (lr=1.9808e-04) (hash(x)=25896435)
350
+ 334 train 7.706117 (lr=1.9800e-04) (hash(x)=21585310)
351
+ 335 train 7.775377 (lr=1.9792e-04) (hash(x)=24677740)
352
+ 336 train 7.615940 (lr=1.9783e-04) (hash(x)=22027900)
353
+ 337 train 7.630406 (lr=1.9774e-04) (hash(x)=21835643)
354
+ 338 train 7.534153 (lr=1.9765e-04) (hash(x)=20558462)
355
+ 339 train 7.465154 (lr=1.9756e-04) (hash(x)=16722715)
356
+ 340 train 7.607401 (lr=1.9747e-04) (hash(x)=20972655)
357
+ 341 train 8.039042 (lr=1.9738e-04) (hash(x)=26303975)
358
+ 342 train 7.769916 (lr=1.9728e-04) (hash(x)=22938170)
359
+ 343 train 7.819746 (lr=1.9718e-04) (hash(x)=25347203)
360
+ 344 train 7.846312 (lr=1.9709e-04) (hash(x)=27398686)
361
+ 345 train 7.794079 (lr=1.9698e-04) (hash(x)=25973417)
362
+ 346 train 7.823333 (lr=1.9688e-04) (hash(x)=26918389)
363
+ 347 train 7.784008 (lr=1.9678e-04) (hash(x)=24272489)
364
+ 348 train 7.789529 (lr=1.9667e-04) (hash(x)=25593714)
365
+ 349 train 7.847080 (lr=1.9656e-04) (hash(x)=29260846)
366
+ 350 val loss 7.7775
367
+ 350 val perplexity 2386.3508
368
+ 350 train 7.946877 (lr=1.9646e-04) (hash(x)=27951602)
369
+ 351 train 8.021439 (lr=1.9634e-04) (hash(x)=28922363)
370
+ 352 train 7.917610 (lr=1.9623e-04) (hash(x)=27210734)
371
+ 353 train 7.971622 (lr=1.9612e-04) (hash(x)=26322572)
372
+ 354 train 7.850690 (lr=1.9600e-04) (hash(x)=27084665)
373
+ 355 train 7.765667 (lr=1.9588e-04) (hash(x)=25510798)
374
+ 356 train 7.863712 (lr=1.9576e-04) (hash(x)=24970921)
375
+ 357 train 7.872037 (lr=1.9564e-04) (hash(x)=24138948)
376
+ 358 train 7.772050 (lr=1.9552e-04) (hash(x)=24790211)
377
+ 359 train 7.843518 (lr=1.9540e-04) (hash(x)=25631397)
378
+ 360 train 7.681755 (lr=1.9527e-04) (hash(x)=23226625)
379
+ 361 train 7.778989 (lr=1.9514e-04) (hash(x)=24001903)
380
+ 362 train 7.773816 (lr=1.9501e-04) (hash(x)=24587948)
381
+ 363 train 7.478784 (lr=1.9488e-04) (hash(x)=21333676)
382
+ 364 train 7.293770 (lr=1.9475e-04) (hash(x)=23673779)
383
+ 365 train 8.708640 (lr=1.9462e-04) (hash(x)=30770484)
384
+ 366 train 7.900467 (lr=1.9448e-04) (hash(x)=26564899)
385
+ 367 train 7.907673 (lr=1.9434e-04) (hash(x)=26237983)
386
+ 368 train 7.626750 (lr=1.9421e-04) (hash(x)=23764356)
387
+ 369 train 7.950007 (lr=1.9406e-04) (hash(x)=26205744)
388
+ 370 train 7.651555 (lr=1.9392e-04) (hash(x)=19208770)
389
+ 371 train 7.884854 (lr=1.9378e-04) (hash(x)=25976502)
390
+ 372 train 7.704685 (lr=1.9363e-04) (hash(x)=23983933)
391
+ 373 train 7.695064 (lr=1.9349e-04) (hash(x)=24080636)
392
+ 374 train 7.799112 (lr=1.9334e-04) (hash(x)=24404047)
393
+ 375 train 7.761975 (lr=1.9319e-04) (hash(x)=24742645)
394
+ 376 train 7.825440 (lr=1.9304e-04) (hash(x)=24159600)
395
+ 377 train 8.057252 (lr=1.9288e-04) (hash(x)=28677257)
396
+ 378 train 7.841246 (lr=1.9273e-04) (hash(x)=25604111)
397
+ 379 train 7.746893 (lr=1.9257e-04) (hash(x)=27086333)
398
+ 380 train 7.690507 (lr=1.9241e-04) (hash(x)=25188207)
399
+ 381 train 7.916008 (lr=1.9225e-04) (hash(x)=27855233)
400
+ 382 train 7.496825 (lr=1.9209e-04) (hash(x)=19470039)
401
+ 383 train 7.809551 (lr=1.9193e-04) (hash(x)=26157660)
402
+ 384 train 7.729379 (lr=1.9176e-04) (hash(x)=25291570)
403
+ 385 train 7.735660 (lr=1.9160e-04) (hash(x)=25046062)
404
+ 386 train 7.995230 (lr=1.9143e-04) (hash(x)=27020337)
405
+ 387 train 7.599226 (lr=1.9126e-04) (hash(x)=23616370)
406
+ 388 train 7.274593 (lr=1.9109e-04) (hash(x)=19113218)
407
+ 389 train 7.639833 (lr=1.9091e-04) (hash(x)=24302232)
408
+ 390 train 7.661734 (lr=1.9074e-04) (hash(x)=22188949)
409
+ 391 train 7.479010 (lr=1.9056e-04) (hash(x)=22582169)
410
+ 392 train 7.662862 (lr=1.9039e-04) (hash(x)=24700570)
411
+ 393 train 7.595780 (lr=1.9021e-04) (hash(x)=22773833)
412
+ 394 train 7.481508 (lr=1.9003e-04) (hash(x)=21875928)
413
+ 395 train 7.779692 (lr=1.8985e-04) (hash(x)=26233189)
414
+ 396 train 7.636544 (lr=1.8966e-04) (hash(x)=24321467)
415
+ 397 train 7.885419 (lr=1.8948e-04) (hash(x)=26431507)
416
+ 398 train 7.946703 (lr=1.8929e-04) (hash(x)=28690877)
417
+ 399 train 7.923569 (lr=1.8910e-04) (hash(x)=26431960)
418
+ 400 val loss 7.7738
419
+ 400 val perplexity 2377.5042
420
+ 400 train 7.755708 (lr=1.8891e-04) (hash(x)=24580300)
421
+ 401 train 7.761201 (lr=1.8872e-04) (hash(x)=25112360)
422
+ 402 train 7.828431 (lr=1.8853e-04) (hash(x)=27597243)
423
+ 403 train 8.252640 (lr=1.8834e-04) (hash(x)=30707498)
424
+ 404 train 7.951135 (lr=1.8814e-04) (hash(x)=28485465)
425
+ 405 train 7.593062 (lr=1.8794e-04) (hash(x)=22586447)
426
+ 406 train 7.754116 (lr=1.8774e-04) (hash(x)=23175270)
427
+ 407 train 7.818353 (lr=1.8754e-04) (hash(x)=25716176)
428
+ 408 train 7.893514 (lr=1.8734e-04) (hash(x)=26861373)
429
+ 409 train 7.976079 (lr=1.8714e-04) (hash(x)=25118971)
430
+ 410 train 7.455047 (lr=1.8693e-04) (hash(x)=19829066)
431
+ 411 train 7.637711 (lr=1.8673e-04) (hash(x)=26256420)
432
+ 412 train 7.924240 (lr=1.8652e-04) (hash(x)=27796153)
433
+ 413 train 7.625217 (lr=1.8631e-04) (hash(x)=22633318)
434
+ 414 train 7.661022 (lr=1.8610e-04) (hash(x)=22589383)
435
+ 415 train 8.128585 (lr=1.8589e-04) (hash(x)=28019788)
436
+ 416 train 8.032027 (lr=1.8567e-04) (hash(x)=28970440)
437
+ 417 train 7.696302 (lr=1.8546e-04) (hash(x)=27396089)
438
+ 418 train 7.687305 (lr=1.8524e-04) (hash(x)=21183513)
439
+ 419 train 7.731539 (lr=1.8502e-04) (hash(x)=23510110)
440
+ 420 train 7.913129 (lr=1.8480e-04) (hash(x)=28833467)
441
+ 421 train 7.798448 (lr=1.8458e-04) (hash(x)=23646926)
442
+ 422 train 7.753499 (lr=1.8436e-04) (hash(x)=24697272)
443
+ 423 train 7.531423 (lr=1.8414e-04) (hash(x)=20382963)
444
+ 424 train 7.539647 (lr=1.8391e-04) (hash(x)=23467595)
445
+ 425 train 7.686554 (lr=1.8369e-04) (hash(x)=24304768)
446
+ 426 train 7.632451 (lr=1.8346e-04) (hash(x)=21392328)
447
+ 427 train 7.644069 (lr=1.8323e-04) (hash(x)=25339466)
448
+ 428 train 7.580129 (lr=1.8300e-04) (hash(x)=22092542)
449
+ 429 train 7.686469 (lr=1.8276e-04) (hash(x)=22088696)
450
+ 430 train 7.602908 (lr=1.8253e-04) (hash(x)=22184471)
451
+ 431 train 7.709032 (lr=1.8230e-04) (hash(x)=24489647)
452
+ 432 train 7.749334 (lr=1.8206e-04) (hash(x)=26794132)
453
+ 433 train 7.506369 (lr=1.8182e-04) (hash(x)=22940357)
454
+ 434 train 7.722810 (lr=1.8158e-04) (hash(x)=23719522)
455
+ 435 train 7.533243 (lr=1.8134e-04) (hash(x)=22927699)
456
+ 436 train 7.746373 (lr=1.8110e-04) (hash(x)=26068576)
457
+ 437 train 7.900070 (lr=1.8086e-04) (hash(x)=27631132)
458
+ 438 train 7.762776 (lr=1.8061e-04) (hash(x)=26739991)
459
+ 439 train 7.824598 (lr=1.8036e-04) (hash(x)=25128502)
460
+ 440 train 7.854796 (lr=1.8012e-04) (hash(x)=25657260)
461
+ 441 train 7.586406 (lr=1.7987e-04) (hash(x)=23576982)
462
+ 442 train 7.868599 (lr=1.7962e-04) (hash(x)=27117886)
463
+ 443 train 7.771770 (lr=1.7937e-04) (hash(x)=25808969)
464
+ 444 train 7.690547 (lr=1.7911e-04) (hash(x)=24738238)
465
+ 445 train 7.728097 (lr=1.7886e-04) (hash(x)=23429962)
466
+ 446 train 7.869180 (lr=1.7860e-04) (hash(x)=25075165)
467
+ 447 train 7.714848 (lr=1.7835e-04) (hash(x)=25231390)
468
+ 448 train 7.497079 (lr=1.7809e-04) (hash(x)=22055054)
469
+ 449 train 7.723441 (lr=1.7783e-04) (hash(x)=25395441)
470
+ 450 val loss 7.7149
471
+ 450 val perplexity 2241.5073
472
+ 450 train 7.764291 (lr=1.7757e-04) (hash(x)=25863209)
473
+ 451 train 7.420974 (lr=1.7731e-04) (hash(x)=21154388)
474
+ 452 train 7.407745 (lr=1.7704e-04) (hash(x)=21600876)
475
+ 453 train 7.569364 (lr=1.7678e-04) (hash(x)=24278611)
476
+ 454 train 7.437864 (lr=1.7651e-04) (hash(x)=23221720)
477
+ 455 train 7.637383 (lr=1.7624e-04) (hash(x)=22708977)
478
+ 456 train 8.251624 (lr=1.7597e-04) (hash(x)=23637758)
479
+ 457 train 7.977963 (lr=1.7570e-04) (hash(x)=28228490)
480
+ 458 train 8.015456 (lr=1.7543e-04) (hash(x)=28638071)
481
+ 459 train 7.863803 (lr=1.7516e-04) (hash(x)=27258353)
482
+ 460 train 7.859031 (lr=1.7489e-04) (hash(x)=26604728)
483
+ 461 train 7.735706 (lr=1.7461e-04) (hash(x)=23252199)
484
+ 462 train 7.817516 (lr=1.7434e-04) (hash(x)=26441427)
485
+ 463 train 7.534299 (lr=1.7406e-04) (hash(x)=24364920)
486
+ 464 train 7.670477 (lr=1.7378e-04) (hash(x)=25623792)
487
+ 465 train 7.602322 (lr=1.7350e-04) (hash(x)=23283905)
488
+ 466 train 8.034961 (lr=1.7322e-04) (hash(x)=26025267)
489
+ 467 train 7.797604 (lr=1.7294e-04) (hash(x)=27243972)
490
+ 468 train 8.175448 (lr=1.7265e-04) (hash(x)=30449945)
491
+ 469 train 7.849927 (lr=1.7237e-04) (hash(x)=28113043)
492
+ 470 train 7.676528 (lr=1.7208e-04) (hash(x)=25182521)
493
+ 471 train 7.718434 (lr=1.7179e-04) (hash(x)=24932925)
494
+ 472 train 7.528409 (lr=1.7151e-04) (hash(x)=20353098)
495
+ 473 train 7.471998 (lr=1.7122e-04) (hash(x)=19001259)
496
+ 474 train 8.024014 (lr=1.7092e-04) (hash(x)=27585685)
497
+ 475 train 7.718150 (lr=1.7063e-04) (hash(x)=26371091)
498
+ 476 train 7.545109 (lr=1.7034e-04) (hash(x)=24891798)
499
+ 477 train 7.559484 (lr=1.7004e-04) (hash(x)=24258817)
500
+ 478 train 7.662754 (lr=1.6975e-04) (hash(x)=24330263)
501
+ 479 train 7.749597 (lr=1.6945e-04) (hash(x)=26913684)
502
+ 480 train 7.733013 (lr=1.6915e-04) (hash(x)=26338455)
503
+ 481 train 7.995474 (lr=1.6886e-04) (hash(x)=27753043)
504
+ 482 train 7.717316 (lr=1.6856e-04) (hash(x)=26123289)
505
+ 483 train 7.810239 (lr=1.6825e-04) (hash(x)=29239611)
506
+ 484 train 7.727054 (lr=1.6795e-04) (hash(x)=26553003)
507
+ 485 train 7.675922 (lr=1.6765e-04) (hash(x)=22984557)
508
+ 486 train 7.222454 (lr=1.6734e-04) (hash(x)=16947491)
509
+ 487 train 7.214111 (lr=1.6704e-04) (hash(x)=18017792)
510
+ 488 train 7.369260 (lr=1.6673e-04) (hash(x)=19918608)
511
+ 489 train 7.870216 (lr=1.6642e-04) (hash(x)=23374526)
512
+ 490 train 7.748593 (lr=1.6611e-04) (hash(x)=25009505)
513
+ 491 train 7.685672 (lr=1.6580e-04) (hash(x)=27574089)
514
+ 492 train 7.869650 (lr=1.6549e-04) (hash(x)=24122664)
515
+ 493 train 7.725525 (lr=1.6518e-04) (hash(x)=26154906)
516
+ 494 train 7.663374 (lr=1.6487e-04) (hash(x)=25192767)
517
+ 495 train 7.781842 (lr=1.6455e-04) (hash(x)=28613882)
518
+ 496 train 7.728054 (lr=1.6424e-04) (hash(x)=23547219)
519
+ 497 train 7.709431 (lr=1.6392e-04) (hash(x)=25272182)
520
+ 498 train 7.565484 (lr=1.6360e-04) (hash(x)=24992761)
521
+ 499 train 7.842152 (lr=1.6328e-04) (hash(x)=26981914)
522
+ 500 val loss 7.6998
523
+ 500 val perplexity 2207.9097
524
+ 500 train 7.603104 (lr=1.6296e-04) (hash(x)=22051933)
525
+ 501 train 7.800693 (lr=1.6264e-04) (hash(x)=24232348)
526
+ 502 train 7.760748 (lr=1.6232e-04) (hash(x)=23158331)
527
+ 503 train 7.590585 (lr=1.6200e-04) (hash(x)=22652243)
528
+ 504 train 7.629569 (lr=1.6168e-04) (hash(x)=23805602)
529
+ 505 train 7.946434 (lr=1.6135e-04) (hash(x)=25411991)
530
+ 506 train 7.248652 (lr=1.6103e-04) (hash(x)=18827215)
531
+ 507 train 6.933525 (lr=1.6070e-04) (hash(x)=15446025)
532
+ 508 train 7.260715 (lr=1.6037e-04) (hash(x)=20516263)
533
+ 509 train 7.770967 (lr=1.6004e-04) (hash(x)=27846176)
534
+ 510 train 7.473714 (lr=1.5971e-04) (hash(x)=23342449)
535
+ 511 train 7.755173 (lr=1.5938e-04) (hash(x)=27194521)
536
+ 512 train 7.563705 (lr=1.5905e-04) (hash(x)=23008284)
537
+ 513 train 8.053714 (lr=1.5872e-04) (hash(x)=29430001)
538
+ 514 train 7.312722 (lr=1.5838e-04) (hash(x)=22579319)
539
+ 515 train 7.366152 (lr=1.5805e-04) (hash(x)=25264518)
540
+ 516 train 7.724942 (lr=1.5772e-04) (hash(x)=25359075)
541
+ 517 train 7.722156 (lr=1.5738e-04) (hash(x)=25568956)
542
+ 518 train 8.014643 (lr=1.5704e-04) (hash(x)=32004108)
543
+ 519 train 7.655769 (lr=1.5670e-04) (hash(x)=24936836)
544
+ 520 train 7.793204 (lr=1.5636e-04) (hash(x)=27263338)
545
+ 521 train 7.862413 (lr=1.5602e-04) (hash(x)=27452099)
546
+ 522 train 7.699762 (lr=1.5568e-04) (hash(x)=25965406)
547
+ 523 train 7.932436 (lr=1.5534e-04) (hash(x)=28197282)
548
+ 524 train 7.603921 (lr=1.5500e-04) (hash(x)=22466209)
549
+ 525 train 7.653257 (lr=1.5466e-04) (hash(x)=22931889)
550
+ 526 train 7.807931 (lr=1.5431e-04) (hash(x)=26903920)
551
+ 527 train 7.693447 (lr=1.5397e-04) (hash(x)=24765578)
552
+ 528 train 7.781635 (lr=1.5362e-04) (hash(x)=27811359)
553
+ 529 train 7.587069 (lr=1.5327e-04) (hash(x)=25078649)
554
+ 530 train 7.768027 (lr=1.5293e-04) (hash(x)=25572416)
555
+ 531 train 7.855004 (lr=1.5258e-04) (hash(x)=27448185)
556
+ 532 train 7.919146 (lr=1.5223e-04) (hash(x)=25923719)
557
+ 533 train 7.678726 (lr=1.5188e-04) (hash(x)=24804856)
558
+ 534 train 7.839749 (lr=1.5153e-04) (hash(x)=23207829)
559
+ 535 train 7.718102 (lr=1.5118e-04) (hash(x)=23107416)
560
+ 536 train 7.571668 (lr=1.5082e-04) (hash(x)=26739531)
561
+ 537 train 7.568475 (lr=1.5047e-04) (hash(x)=24960796)
562
+ 538 train 7.715461 (lr=1.5012e-04) (hash(x)=24667802)
563
+ 539 train 7.815919 (lr=1.4976e-04) (hash(x)=26755138)
564
+ 540 train 7.699287 (lr=1.4941e-04) (hash(x)=25537132)
565
+ 541 train 7.608414 (lr=1.4905e-04) (hash(x)=24542526)
566
+ 542 train 7.406418 (lr=1.4869e-04) (hash(x)=21296355)
567
+ 543 train 7.958098 (lr=1.4833e-04) (hash(x)=29314255)
568
+ 544 train 7.670983 (lr=1.4798e-04) (hash(x)=26001799)
569
+ 545 train 7.722990 (lr=1.4762e-04) (hash(x)=27347755)
570
+ 546 train 7.729439 (lr=1.4726e-04) (hash(x)=25107798)
571
+ 547 train 7.505542 (lr=1.4690e-04) (hash(x)=22112669)
572
+ 548 train 7.488782 (lr=1.4653e-04) (hash(x)=21897967)
573
+ 549 train 7.637074 (lr=1.4617e-04) (hash(x)=25161929)
574
+ 550 val loss 7.6645
575
+ 550 val perplexity 2131.3425
576
+ 550 train 7.534184 (lr=1.4581e-04) (hash(x)=27465106)
577
+ 551 train 7.654341 (lr=1.4545e-04) (hash(x)=24013079)
578
+ 552 train 7.591847 (lr=1.4508e-04) (hash(x)=23142015)
579
+ 553 train 7.669997 (lr=1.4472e-04) (hash(x)=26768629)
580
+ 554 train 7.691318 (lr=1.4435e-04) (hash(x)=26393383)
581
+ 555 train 7.553133 (lr=1.4398e-04) (hash(x)=22537194)
582
+ 556 train 7.810132 (lr=1.4362e-04) (hash(x)=24046036)
583
+ 557 train 7.734828 (lr=1.4325e-04) (hash(x)=24974360)
584
+ 558 train 7.948419 (lr=1.4288e-04) (hash(x)=28379928)
585
+ 559 train 7.690988 (lr=1.4251e-04) (hash(x)=25322001)
586
+ 560 train 7.782243 (lr=1.4214e-04) (hash(x)=26622031)
587
+ 561 train 7.432893 (lr=1.4177e-04) (hash(x)=20562247)
588
+ 562 train 7.800185 (lr=1.4140e-04) (hash(x)=27381885)
589
+ 563 train 7.957479 (lr=1.4103e-04) (hash(x)=27028126)
590
+ 564 train 7.764669 (lr=1.4066e-04) (hash(x)=28882928)
591
+ 565 train 7.673637 (lr=1.4029e-04) (hash(x)=25666355)
592
+ 566 train 7.820342 (lr=1.3991e-04) (hash(x)=24330810)
593
+ 567 train 7.717567 (lr=1.3954e-04) (hash(x)=26690440)
594
+ 568 train 7.563193 (lr=1.3916e-04) (hash(x)=22923592)
595
+ 569 train 7.704013 (lr=1.3879e-04) (hash(x)=27348418)
596
+ 570 train 7.774705 (lr=1.3841e-04) (hash(x)=28849848)
597
+ 571 train 7.816841 (lr=1.3804e-04) (hash(x)=26967331)
598
+ 572 train 7.519025 (lr=1.3766e-04) (hash(x)=22831467)
599
+ 573 train 7.638724 (lr=1.3728e-04) (hash(x)=24765121)
600
+ 574 train 7.595777 (lr=1.3691e-04) (hash(x)=24331857)
601
+ 575 train 7.420772 (lr=1.3653e-04) (hash(x)=22598512)
602
+ 576 train 7.650543 (lr=1.3615e-04) (hash(x)=25149353)
603
+ 577 train 7.456262 (lr=1.3577e-04) (hash(x)=23725598)
604
+ 578 train 7.741199 (lr=1.3539e-04) (hash(x)=26449557)
605
+ 579 train 7.755365 (lr=1.3501e-04) (hash(x)=24697985)
606
+ 580 train 7.727361 (lr=1.3463e-04) (hash(x)=26923059)
607
+ 581 train 7.562045 (lr=1.3425e-04) (hash(x)=25201962)
608
+ 582 train 7.226964 (lr=1.3387e-04) (hash(x)=20931520)
609
+ 583 train 7.349971 (lr=1.3348e-04) (hash(x)=18473911)
610
+ 584 train 7.481683 (lr=1.3310e-04) (hash(x)=21306267)
611
+ 585 train 7.712385 (lr=1.3272e-04) (hash(x)=25982840)
612
+ 586 train 7.614620 (lr=1.3234e-04) (hash(x)=25364874)
613
+ 587 train 7.458683 (lr=1.3195e-04) (hash(x)=23172124)
614
+ 588 train 7.795200 (lr=1.3157e-04) (hash(x)=27876897)
615
+ 589 train 8.450606 (lr=1.3118e-04) (hash(x)=34646114)
616
+ 590 train 8.848363 (lr=1.3080e-04) (hash(x)=35153576)
617
+ 591 train 7.561924 (lr=1.3041e-04) (hash(x)=22322442)
618
+ 592 train 7.796544 (lr=1.3003e-04) (hash(x)=27907331)
619
+ 593 train 7.689842 (lr=1.2964e-04) (hash(x)=26211794)
620
+ 594 train 7.907566 (lr=1.2925e-04) (hash(x)=29291512)
621
+ 595 train 8.201463 (lr=1.2887e-04) (hash(x)=29659121)
622
+ 596 train 8.018188 (lr=1.2848e-04) (hash(x)=29674399)
623
+ 597 train 7.690909 (lr=1.2809e-04) (hash(x)=23538306)
624
+ 598 train 7.810375 (lr=1.2770e-04) (hash(x)=21991524)
625
+ 599 train 7.703667 (lr=1.2732e-04) (hash(x)=26324153)
626
+ 600 val loss 7.6676
627
+ 600 val perplexity 2137.8884
628
+ 600 train 7.627975 (lr=1.2693e-04) (hash(x)=23712082)
629
+ 601 train 7.707057 (lr=1.2654e-04) (hash(x)=24910403)
630
+ 602 train 7.617029 (lr=1.2615e-04) (hash(x)=26737205)
631
+ 603 train 7.780415 (lr=1.2576e-04) (hash(x)=26939970)
632
+ 604 train 7.642729 (lr=1.2537e-04) (hash(x)=27651943)
633
+ 605 train 7.987737 (lr=1.2498e-04) (hash(x)=27515446)
634
+ 606 train 7.823711 (lr=1.2459e-04) (hash(x)=26753129)
635
+ 607 train 7.503444 (lr=1.2420e-04) (hash(x)=23446058)
636
+ 608 train 7.855058 (lr=1.2381e-04) (hash(x)=27587849)
637
+ 609 train 7.740616 (lr=1.2341e-04) (hash(x)=25308253)
638
+ 610 train 7.729346 (lr=1.2302e-04) (hash(x)=26615098)
639
+ 611 train 7.758736 (lr=1.2263e-04) (hash(x)=29981801)
640
+ 612 train 7.842911 (lr=1.2224e-04) (hash(x)=29592345)
641
+ 613 train 7.857415 (lr=1.2185e-04) (hash(x)=23470413)
642
+ 614 train 7.782956 (lr=1.2145e-04) (hash(x)=24742370)
643
+ 615 train 7.774812 (lr=1.2106e-04) (hash(x)=24843741)
644
+ 616 train 7.642323 (lr=1.2067e-04) (hash(x)=25192548)
645
+ 617 train 7.747602 (lr=1.2027e-04) (hash(x)=27176996)
646
+ 618 train 7.600376 (lr=1.1988e-04) (hash(x)=23964552)
647
+ 619 train 7.589353 (lr=1.1949e-04) (hash(x)=22855363)
648
+ 620 train 7.566194 (lr=1.1909e-04) (hash(x)=26332996)
649
+ 621 train 7.380164 (lr=1.1870e-04) (hash(x)=22960957)
650
+ 622 train 7.706482 (lr=1.1830e-04) (hash(x)=22752597)
651
+ 623 train 7.596294 (lr=1.1791e-04) (hash(x)=23197102)
652
+ 624 train 8.053697 (lr=1.1752e-04) (hash(x)=27383319)
653
+ 625 train 7.634480 (lr=1.1712e-04) (hash(x)=26132276)
654
+ 626 train 7.521698 (lr=1.1673e-04) (hash(x)=19810497)
655
+ 627 train 7.591949 (lr=1.1633e-04) (hash(x)=25704919)
656
+ 628 train 7.773521 (lr=1.1594e-04) (hash(x)=27174264)
657
+ 629 train 7.916601 (lr=1.1554e-04) (hash(x)=22280814)
658
+ 630 train 8.690178 (lr=1.1515e-04) (hash(x)=24279448)
659
+ 631 train 8.101439 (lr=1.1475e-04) (hash(x)=23054940)
660
+ 632 train 8.237493 (lr=1.1435e-04) (hash(x)=23942400)
661
+ 633 train 8.484448 (lr=1.1396e-04) (hash(x)=24712416)
662
+ 634 train 8.019852 (lr=1.1356e-04) (hash(x)=24564658)
663
+ 635 train 7.741986 (lr=1.1317e-04) (hash(x)=24909904)
664
+ 636 train 7.757636 (lr=1.1277e-04) (hash(x)=23049534)
665
+ 637 train 7.830408 (lr=1.1238e-04) (hash(x)=24321591)
666
+ 638 train 7.587223 (lr=1.1198e-04) (hash(x)=26153298)
667
+ 639 train 7.616875 (lr=1.1158e-04) (hash(x)=27140757)
668
+ 640 train 7.367975 (lr=1.1119e-04) (hash(x)=25115907)
669
+ 641 train 7.412740 (lr=1.1079e-04) (hash(x)=26563770)
670
+ 642 train 7.637044 (lr=1.1040e-04) (hash(x)=28089252)
671
+ 643 train 7.539066 (lr=1.1000e-04) (hash(x)=25993110)
672
+ 644 train 7.511646 (lr=1.0960e-04) (hash(x)=25150008)
673
+ 645 train 7.775749 (lr=1.0921e-04) (hash(x)=28520222)
674
+ 646 train 7.439560 (lr=1.0881e-04) (hash(x)=21349943)
675
+ 647 train 7.511539 (lr=1.0842e-04) (hash(x)=25149419)
676
+ 648 train 7.702456 (lr=1.0802e-04) (hash(x)=25730641)
677
+ 649 train 7.842803 (lr=1.0762e-04) (hash(x)=26112813)
678
+ 650 val loss 7.6128
679
+ 650 val perplexity 2023.8439
680
+ 650 train 7.652194 (lr=1.0723e-04) (hash(x)=25907805)
681
+ 651 train 7.652926 (lr=1.0683e-04) (hash(x)=27623643)
682
+ 652 train 7.628788 (lr=1.0644e-04) (hash(x)=26484959)
683
+ 653 train 7.693550 (lr=1.0604e-04) (hash(x)=29199854)
684
+ 654 train 7.670650 (lr=1.0565e-04) (hash(x)=28369628)
685
+ 655 train 7.550072 (lr=1.0525e-04) (hash(x)=24727764)
686
+ 656 train 7.273953 (lr=1.0485e-04) (hash(x)=22610673)
687
+ 657 train 7.217771 (lr=1.0446e-04) (hash(x)=22667179)
688
+ 658 train 7.008431 (lr=1.0406e-04) (hash(x)=18477300)
689
+ 659 train 7.422233 (lr=1.0367e-04) (hash(x)=23155773)
690
+ 660 train 7.202079 (lr=1.0327e-04) (hash(x)=19461032)
691
+ 661 train 7.495474 (lr=1.0288e-04) (hash(x)=23453788)
692
+ 662 train 7.628123 (lr=1.0248e-04) (hash(x)=24543466)
693
+ 663 train 7.415193 (lr=1.0209e-04) (hash(x)=21935931)
694
+ 664 train 7.236494 (lr=1.0170e-04) (hash(x)=19910292)
695
+ 665 train 7.537254 (lr=1.0130e-04) (hash(x)=24481079)
696
+ 666 train 7.204022 (lr=1.0091e-04) (hash(x)=18922411)
697
+ 667 train 7.254138 (lr=1.0051e-04) (hash(x)=20054917)
698
+ 668 train 7.488539 (lr=1.0012e-04) (hash(x)=24850470)
699
+ 669 train 7.497656 (lr=9.9726e-05) (hash(x)=25907741)
700
+ 670 train 7.697685 (lr=9.9333e-05) (hash(x)=26873522)
701
+ 671 train 7.721660 (lr=9.8940e-05) (hash(x)=27606073)
702
+ 672 train 7.817198 (lr=9.8547e-05) (hash(x)=26209645)
703
+ 673 train 7.663980 (lr=9.8154e-05) (hash(x)=25202001)
704
+ 674 train 7.729453 (lr=9.7762e-05) (hash(x)=25569462)
705
+ 675 train 7.691226 (lr=9.7370e-05) (hash(x)=26534487)
706
+ 676 train 7.671887 (lr=9.6978e-05) (hash(x)=26455057)
707
+ 677 train 7.534261 (lr=9.6586e-05) (hash(x)=24095850)
708
+ 678 train 7.748569 (lr=9.6195e-05) (hash(x)=25287752)
709
+ 679 train 7.233978 (lr=9.5804e-05) (hash(x)=22450341)
710
+ 680 train 7.966220 (lr=9.5413e-05) (hash(x)=29004853)
711
+ 681 train 7.635642 (lr=9.5022e-05) (hash(x)=27993763)
712
+ 682 train 7.486473 (lr=9.4632e-05) (hash(x)=26382658)
713
+ 683 train 7.664583 (lr=9.4242e-05) (hash(x)=25013073)
714
+ 684 train 7.692557 (lr=9.3852e-05) (hash(x)=30595809)
715
+ 685 train 7.776471 (lr=9.3463e-05) (hash(x)=30934371)
716
+ 686 train 8.409088 (lr=9.3073e-05) (hash(x)=33060834)
717
+ 687 train 7.632722 (lr=9.2685e-05) (hash(x)=25945859)
718
+ 688 train 7.482469 (lr=9.2296e-05) (hash(x)=23375678)
719
+ 689 train 7.517372 (lr=9.1908e-05) (hash(x)=25218689)
720
+ 690 train 7.629486 (lr=9.1520e-05) (hash(x)=28127397)
721
+ 691 train 7.552429 (lr=9.1133e-05) (hash(x)=24418091)
722
+ 692 train 7.457856 (lr=9.0746e-05) (hash(x)=22761099)
723
+ 693 train 7.438896 (lr=9.0359e-05) (hash(x)=24615466)
724
+ 694 train 7.600534 (lr=8.9973e-05) (hash(x)=24699240)
725
+ 695 train 7.546469 (lr=8.9587e-05) (hash(x)=21613707)
726
+ 696 train 7.605920 (lr=8.9202e-05) (hash(x)=24977554)
727
+ 697 train 7.530506 (lr=8.8817e-05) (hash(x)=24348175)
728
+ 698 train 7.679914 (lr=8.8432e-05) (hash(x)=25102767)
729
+ 699 train 7.651213 (lr=8.8048e-05) (hash(x)=26386157)
730
+ 700 val loss 7.5696
731
+ 700 val perplexity 1938.4214
732
+ 700 train 7.610486 (lr=8.7664e-05) (hash(x)=26423460)
733
+ 701 train 7.517082 (lr=8.7281e-05) (hash(x)=25771047)
734
+ 702 train 7.574867 (lr=8.6898e-05) (hash(x)=26999875)
735
+ 703 train 7.533044 (lr=8.6515e-05) (hash(x)=24396519)
736
+ 704 train 7.533234 (lr=8.6133e-05) (hash(x)=22588122)
737
+ 705 train 7.840021 (lr=8.5751e-05) (hash(x)=25142399)
738
+ 706 train 7.569984 (lr=8.5370e-05) (hash(x)=20440214)
739
+ 707 train 7.421633 (lr=8.4990e-05) (hash(x)=23265507)
740
+ 708 train 7.549704 (lr=8.4610e-05) (hash(x)=24563470)
741
+ 709 train 7.472009 (lr=8.4230e-05) (hash(x)=22514858)
742
+ 710 train 7.803635 (lr=8.3851e-05) (hash(x)=26691212)
743
+ 711 train 7.842292 (lr=8.3472e-05) (hash(x)=29138828)
744
+ 712 train 7.814781 (lr=8.3094e-05) (hash(x)=28028528)
745
+ 713 train 7.681085 (lr=8.2716e-05) (hash(x)=20531210)
746
+ 714 train 7.528381 (lr=8.2339e-05) (hash(x)=25075352)
747
+ 715 train 7.586564 (lr=8.1963e-05) (hash(x)=24265353)
748
+ 716 train 7.698923 (lr=8.1587e-05) (hash(x)=24635726)
749
+ 717 train 7.521175 (lr=8.1211e-05) (hash(x)=24999726)
750
+ 718 train 7.727352 (lr=8.0836e-05) (hash(x)=27412910)
751
+ 719 train 7.521915 (lr=8.0462e-05) (hash(x)=24685515)
752
+ 720 train 7.588367 (lr=8.0088e-05) (hash(x)=23780329)
753
+ 721 train 7.373122 (lr=7.9715e-05) (hash(x)=25071701)
754
+ 722 train 7.549147 (lr=7.9342e-05) (hash(x)=23767130)
755
+ 723 train 7.501265 (lr=7.8970e-05) (hash(x)=24876269)
756
+ 724 train 7.627699 (lr=7.8599e-05) (hash(x)=26405773)
757
+ 725 train 8.042769 (lr=7.8228e-05) (hash(x)=31733180)
758
+ 726 train 7.224096 (lr=7.7858e-05) (hash(x)=21337509)
759
+ 727 train 7.347805 (lr=7.7488e-05) (hash(x)=22825749)
760
+ 728 train 7.724216 (lr=7.7119e-05) (hash(x)=28638695)
761
+ 729 train 7.614185 (lr=7.6751e-05) (hash(x)=26393943)
762
+ 730 train 7.756969 (lr=7.6383e-05) (hash(x)=27563583)
763
+ 731 train 7.262392 (lr=7.6016e-05) (hash(x)=21239652)
764
+ 732 train 7.424767 (lr=7.5650e-05) (hash(x)=23986428)
765
+ 733 train 7.345778 (lr=7.5284e-05) (hash(x)=24943881)
766
+ 734 train 7.873682 (lr=7.4919e-05) (hash(x)=29691448)
767
+ 735 train 7.809682 (lr=7.4555e-05) (hash(x)=28767869)
768
+ 736 train 7.544102 (lr=7.4191e-05) (hash(x)=23628188)
769
+ 737 train 7.839100 (lr=7.3828e-05) (hash(x)=29341482)
770
+ 738 train 7.663057 (lr=7.3466e-05) (hash(x)=30336570)
771
+ 739 train 7.591226 (lr=7.3105e-05) (hash(x)=25614301)
772
+ 740 train 7.460135 (lr=7.2744e-05) (hash(x)=24160500)
773
+ 741 train 7.652695 (lr=7.2384e-05) (hash(x)=26030058)
774
+ 742 train 8.018718 (lr=7.2024e-05) (hash(x)=29243936)
775
+ 743 train 7.349188 (lr=7.1666e-05) (hash(x)=21159060)
776
+ 744 train 7.456842 (lr=7.1308e-05) (hash(x)=23701853)
777
+ 745 train 7.622776 (lr=7.0950e-05) (hash(x)=24629937)
778
+ 746 train 7.538063 (lr=7.0594e-05) (hash(x)=25110108)
779
+ 747 train 7.606596 (lr=7.0238e-05) (hash(x)=26751788)
780
+ 748 train 7.646847 (lr=6.9884e-05) (hash(x)=26430427)
781
+ 749 train 7.491092 (lr=6.9529e-05) (hash(x)=26012353)
782
+ 750 val loss 7.5419
783
+ 750 val perplexity 1885.3894
784
+ 750 train 7.619861 (lr=6.9176e-05) (hash(x)=22735910)
785
+ 751 train 7.415953 (lr=6.8824e-05) (hash(x)=25045397)
786
+ 752 train 7.135645 (lr=6.8472e-05) (hash(x)=21554427)
787
+ 753 train 7.254901 (lr=6.8121e-05) (hash(x)=23751143)
788
+ 754 train 7.976012 (lr=6.7771e-05) (hash(x)=28602273)
789
+ 755 train 8.037753 (lr=6.7422e-05) (hash(x)=29989709)
790
+ 756 train 7.468126 (lr=6.7073e-05) (hash(x)=22331648)
791
+ 757 train 7.841812 (lr=6.6725e-05) (hash(x)=31017246)
792
+ 758 train 7.809131 (lr=6.6379e-05) (hash(x)=29709045)
793
+ 759 train 7.883296 (lr=6.6033e-05) (hash(x)=25560928)
794
+ 760 train 7.586539 (lr=6.5688e-05) (hash(x)=25075464)
795
+ 761 train 7.643425 (lr=6.5343e-05) (hash(x)=27352253)
796
+ 762 train 7.839050 (lr=6.5000e-05) (hash(x)=28187891)
797
+ 763 train 7.696804 (lr=6.4657e-05) (hash(x)=26062687)
798
+ 764 train 7.673834 (lr=6.4316e-05) (hash(x)=27427811)
799
+ 765 train 7.778177 (lr=6.3975e-05) (hash(x)=27614522)
800
+ 766 train 7.525851 (lr=6.3635e-05) (hash(x)=26129544)
801
+ 767 train 8.084749 (lr=6.3296e-05) (hash(x)=28959222)
802
+ 768 train 7.677368 (lr=6.2958e-05) (hash(x)=26860067)
803
+ 769 train 7.391256 (lr=6.2621e-05) (hash(x)=25122598)
804
+ 770 train 7.487373 (lr=6.2285e-05) (hash(x)=25245030)
805
+ 771 train 7.496639 (lr=6.1950e-05) (hash(x)=25434884)
806
+ 772 train 7.593259 (lr=6.1615e-05) (hash(x)=27732790)
807
+ 773 train 7.630791 (lr=6.1282e-05) (hash(x)=27824438)
808
+ 774 train 7.746143 (lr=6.0949e-05) (hash(x)=27201953)
809
+ 775 train 7.446448 (lr=6.0618e-05) (hash(x)=21248405)
810
+ 776 train 7.260738 (lr=6.0287e-05) (hash(x)=22805934)
811
+ 777 train 7.550602 (lr=5.9958e-05) (hash(x)=26482588)
812
+ 778 train 7.611481 (lr=5.9629e-05) (hash(x)=24153691)
813
+ 779 train 7.687893 (lr=5.9301e-05) (hash(x)=25044192)
814
+ 780 train 7.638855 (lr=5.8974e-05) (hash(x)=25910078)
815
+ 781 train 7.632890 (lr=5.8649e-05) (hash(x)=28645524)
816
+ 782 train 7.485985 (lr=5.8324e-05) (hash(x)=24368498)
817
+ 783 train 7.552732 (lr=5.8000e-05) (hash(x)=25830182)
818
+ 784 train 7.639267 (lr=5.7678e-05) (hash(x)=29181807)
819
+ 785 train 7.575421 (lr=5.7356e-05) (hash(x)=25585137)
820
+ 786 train 7.496849 (lr=5.7035e-05) (hash(x)=24798246)
821
+ 787 train 7.598012 (lr=5.6716e-05) (hash(x)=26621419)
822
+ 788 train 7.163819 (lr=5.6397e-05) (hash(x)=21446891)
823
+ 789 train 7.111635 (lr=5.6079e-05) (hash(x)=22165286)
824
+ 790 train 7.345735 (lr=5.5763e-05) (hash(x)=23477219)
825
+ 791 train 7.613524 (lr=5.5447e-05) (hash(x)=25173113)
826
+ 792 train 7.515689 (lr=5.5133e-05) (hash(x)=25853788)
827
+ 793 train 7.605453 (lr=5.4820e-05) (hash(x)=27267091)
828
+ 794 train 7.442778 (lr=5.4507e-05) (hash(x)=23743694)
829
+ 795 train 7.433052 (lr=5.4196e-05) (hash(x)=24400133)
830
+ 796 train 7.503316 (lr=5.3886e-05) (hash(x)=23663639)
831
+ 797 train 7.404125 (lr=5.3577e-05) (hash(x)=23103223)
832
+ 798 train 7.915274 (lr=5.3269e-05) (hash(x)=28748411)
833
+ 799 train 7.196221 (lr=5.2962e-05) (hash(x)=23486277)
834
+ 800 val loss 7.5060
835
+ 800 val perplexity 1818.8557
836
+ 800 train 7.496999 (lr=5.2656e-05) (hash(x)=25678518)
837
+ 801 train 7.435771 (lr=5.2352e-05) (hash(x)=23421286)
838
+ 802 train 7.579250 (lr=5.2048e-05) (hash(x)=26054104)
839
+ 803 train 7.573879 (lr=5.1746e-05) (hash(x)=25978130)
840
+ 804 train 7.597505 (lr=5.1444e-05) (hash(x)=26006525)
841
+ 805 train 7.546604 (lr=5.1144e-05) (hash(x)=25769432)
842
+ 806 train 7.375392 (lr=5.0845e-05) (hash(x)=22430795)
843
+ 807 train 7.750615 (lr=5.0547e-05) (hash(x)=28916006)
844
+ 808 train 7.539983 (lr=5.0251e-05) (hash(x)=25166800)
845
+ 809 train 7.648313 (lr=4.9955e-05) (hash(x)=24226056)
846
+ 810 train 7.506187 (lr=4.9661e-05) (hash(x)=25057934)
847
+ 811 train 8.336482 (lr=4.9367e-05) (hash(x)=33195100)
848
+ 812 train 7.678180 (lr=4.9075e-05) (hash(x)=26312888)
849
+ 813 train 7.832169 (lr=4.8784e-05) (hash(x)=27730410)
850
+ 814 train 7.680675 (lr=4.8495e-05) (hash(x)=27372474)
851
+ 815 train 7.549817 (lr=4.8206e-05) (hash(x)=25556929)
852
+ 816 train 7.652252 (lr=4.7919e-05) (hash(x)=26909985)
853
+ 817 train 7.475259 (lr=4.7633e-05) (hash(x)=25991247)
854
+ 818 train 7.651593 (lr=4.7348e-05) (hash(x)=27438141)
855
+ 819 train 7.795113 (lr=4.7064e-05) (hash(x)=29536986)
856
+ 820 train 7.539062 (lr=4.6782e-05) (hash(x)=24478391)
857
+ 821 train 7.497285 (lr=4.6501e-05) (hash(x)=26125216)
858
+ 822 train 7.510713 (lr=4.6221e-05) (hash(x)=26422130)
859
+ 823 train 7.744484 (lr=4.5942e-05) (hash(x)=29648798)
860
+ 824 train 7.453009 (lr=4.5664e-05) (hash(x)=21247770)
861
+ 825 train 7.543437 (lr=4.5388e-05) (hash(x)=23195388)
862
+ 826 train 7.463849 (lr=4.5113e-05) (hash(x)=25796725)
863
+ 827 train 7.557454 (lr=4.4839e-05) (hash(x)=23124767)
864
+ 828 train 7.555113 (lr=4.4567e-05) (hash(x)=25233464)
865
+ 829 train 7.558540 (lr=4.4295e-05) (hash(x)=25713275)
866
+ 830 train 7.511130 (lr=4.4025e-05) (hash(x)=25550167)
867
+ 831 train 7.389587 (lr=4.3757e-05) (hash(x)=24976217)
868
+ 832 train 7.710184 (lr=4.3489e-05) (hash(x)=28536827)
869
+ 833 train 7.605587 (lr=4.3223e-05) (hash(x)=27500801)
870
+ 834 train 7.416769 (lr=4.2958e-05) (hash(x)=25545765)
871
+ 835 train 7.440898 (lr=4.2695e-05) (hash(x)=23632825)
872
+ 836 train 7.470075 (lr=4.2432e-05) (hash(x)=25708009)
873
+ 837 train 7.534590 (lr=4.2171e-05) (hash(x)=24456276)
874
+ 838 train 7.631275 (lr=4.1912e-05) (hash(x)=29189855)
875
+ 839 train 7.867588 (lr=4.1654e-05) (hash(x)=31019606)
876
+ 840 train 7.730634 (lr=4.1397e-05) (hash(x)=26328013)
877
+ 841 train 7.674356 (lr=4.1141e-05) (hash(x)=25027904)
878
+ 842 train 7.447919 (lr=4.0887e-05) (hash(x)=23734189)
879
+ 843 train 7.688663 (lr=4.0634e-05) (hash(x)=28236580)
880
+ 844 train 7.499403 (lr=4.0382e-05) (hash(x)=26509780)
881
+ 845 train 7.572858 (lr=4.0132e-05) (hash(x)=25386473)
882
+ 846 train 7.515182 (lr=3.9883e-05) (hash(x)=24052671)
883
+ 847 train 7.689115 (lr=3.9635e-05) (hash(x)=28269421)
884
+ 848 train 7.123300 (lr=3.9389e-05) (hash(x)=22251724)
885
+ 849 train 7.390975 (lr=3.9144e-05) (hash(x)=24308447)
886
+ 850 val loss 7.5023
887
+ 850 val perplexity 1812.2269
888
+ 850 train 7.366690 (lr=3.8901e-05) (hash(x)=24242830)
889
+ 851 train 7.569607 (lr=3.8659e-05) (hash(x)=25563279)
890
+ 852 train 7.578651 (lr=3.8418e-05) (hash(x)=26354481)
891
+ 853 train 7.694679 (lr=3.8178e-05) (hash(x)=26152637)
892
+ 854 train 7.755402 (lr=3.7941e-05) (hash(x)=28051025)
893
+ 855 train 7.545935 (lr=3.7704e-05) (hash(x)=24865358)
894
+ 856 train 7.436066 (lr=3.7469e-05) (hash(x)=24288911)
895
+ 857 train 7.309545 (lr=3.7235e-05) (hash(x)=22230964)
896
+ 858 train 7.239000 (lr=3.7003e-05) (hash(x)=21303832)
897
+ 859 train 7.373697 (lr=3.6772e-05) (hash(x)=22155546)
898
+ 860 train 7.517040 (lr=3.6542e-05) (hash(x)=25296428)
899
+ 861 train 7.643286 (lr=3.6314e-05) (hash(x)=29142319)
900
+ 862 train 7.464158 (lr=3.6088e-05) (hash(x)=25545430)
901
+ 863 train 7.303130 (lr=3.5862e-05) (hash(x)=26984272)
902
+ 864 train 7.510620 (lr=3.5639e-05) (hash(x)=25429005)
903
+ 865 train 7.553981 (lr=3.5416e-05) (hash(x)=27077032)
904
+ 866 train 7.531604 (lr=3.5195e-05) (hash(x)=26494424)
905
+ 867 train 7.543144 (lr=3.4976e-05) (hash(x)=23193673)
906
+ 868 train 7.551274 (lr=3.4758e-05) (hash(x)=25075134)
907
+ 869 train 7.580610 (lr=3.4541e-05) (hash(x)=27112558)
908
+ 870 train 7.775641 (lr=3.4326e-05) (hash(x)=27436608)
909
+ 871 train 7.466033 (lr=3.4113e-05) (hash(x)=24544116)
910
+ 872 train 7.809839 (lr=3.3900e-05) (hash(x)=31632686)
911
+ 873 train 7.486576 (lr=3.3690e-05) (hash(x)=25890184)
912
+ 874 train 7.612395 (lr=3.3480e-05) (hash(x)=22887555)
913
+ 875 train 7.452681 (lr=3.3273e-05) (hash(x)=24547533)
914
+ 876 train 7.578310 (lr=3.3066e-05) (hash(x)=26553496)
915
+ 877 train 7.675249 (lr=3.2862e-05) (hash(x)=27467688)
916
+ 878 train 7.453346 (lr=3.2658e-05) (hash(x)=24766934)
917
+ 879 train 7.332334 (lr=3.2457e-05) (hash(x)=22059850)
918
+ 880 train 7.707195 (lr=3.2256e-05) (hash(x)=22871702)
919
+ 881 train 7.556789 (lr=3.2058e-05) (hash(x)=23893130)
920
+ 882 train 7.564200 (lr=3.1860e-05) (hash(x)=25125691)
921
+ 883 train 7.519304 (lr=3.1665e-05) (hash(x)=25994573)
922
+ 884 train 7.544690 (lr=3.1471e-05) (hash(x)=26076345)
923
+ 885 train 7.508221 (lr=3.1278e-05) (hash(x)=26577783)
924
+ 886 train 7.606555 (lr=3.1087e-05) (hash(x)=27395225)
925
+ 887 train 7.459943 (lr=3.0897e-05) (hash(x)=23926632)
926
+ 888 train 7.297412 (lr=3.0709e-05) (hash(x)=21737239)
927
+ 889 train 7.591444 (lr=3.0522e-05) (hash(x)=23574207)
928
+ 890 train 7.280792 (lr=3.0337e-05) (hash(x)=24365231)
929
+ 891 train 7.448013 (lr=3.0154e-05) (hash(x)=27111369)
930
+ 892 train 7.706406 (lr=2.9972e-05) (hash(x)=27290015)
931
+ 893 train 7.495033 (lr=2.9791e-05) (hash(x)=23979820)
932
+ 894 train 7.591215 (lr=2.9613e-05) (hash(x)=26450121)
933
+ 895 train 7.488577 (lr=2.9435e-05) (hash(x)=27025333)
934
+ 896 train 7.473794 (lr=2.9260e-05) (hash(x)=23624605)
935
+ 897 train 7.326399 (lr=2.9085e-05) (hash(x)=22846386)
936
+ 898 train 7.378272 (lr=2.8913e-05) (hash(x)=22970561)
937
+ 899 train 7.246465 (lr=2.8742e-05) (hash(x)=16908068)
938
+ 900 val loss 7.4872
939
+ 900 val perplexity 1785.1053
940
+ 900 train 7.436440 (lr=2.8572e-05) (hash(x)=24661446)
941
+ 901 train 7.550224 (lr=2.8404e-05) (hash(x)=25664727)
942
+ 902 train 7.540842 (lr=2.8238e-05) (hash(x)=25667011)
943
+ 903 train 7.579391 (lr=2.8073e-05) (hash(x)=29120407)
944
+ 904 train 7.650917 (lr=2.7910e-05) (hash(x)=23385735)
945
+ 905 train 7.682660 (lr=2.7749e-05) (hash(x)=25564213)
946
+ 906 train 7.721912 (lr=2.7589e-05) (hash(x)=25413898)
947
+ 907 train 7.699838 (lr=2.7430e-05) (hash(x)=27092710)
948
+ 908 train 7.475684 (lr=2.7274e-05) (hash(x)=25789923)
949
+ 909 train 7.558824 (lr=2.7119e-05) (hash(x)=28533197)
950
+ 910 train 7.404026 (lr=2.6965e-05) (hash(x)=22982996)
951
+ 911 train 7.534404 (lr=2.6813e-05) (hash(x)=23827393)
952
+ 912 train 7.076814 (lr=2.6663e-05) (hash(x)=21242640)
953
+ 913 train 7.479911 (lr=2.6514e-05) (hash(x)=24154233)
954
+ 914 train 7.510229 (lr=2.6367e-05) (hash(x)=24331967)
955
+ 915 train 7.764019 (lr=2.6221e-05) (hash(x)=32812727)
956
+ 916 train 7.405843 (lr=2.6077e-05) (hash(x)=23572994)
957
+ 917 train 7.422185 (lr=2.5935e-05) (hash(x)=26305435)
958
+ 918 train 7.452371 (lr=2.5795e-05) (hash(x)=26268355)
959
+ 919 train 7.655228 (lr=2.5656e-05) (hash(x)=27230027)
960
+ 920 train 7.452167 (lr=2.5518e-05) (hash(x)=23885377)
961
+ 921 train 7.371185 (lr=2.5383e-05) (hash(x)=23532437)
962
+ 922 train 7.469969 (lr=2.5249e-05) (hash(x)=25577034)
963
+ 923 train 7.640682 (lr=2.5116e-05) (hash(x)=25703381)
964
+ 924 train 7.582704 (lr=2.4985e-05) (hash(x)=27113866)
965
+ 925 train 7.432436 (lr=2.4856e-05) (hash(x)=26961429)
966
+ 926 train 7.264868 (lr=2.4729e-05) (hash(x)=21355372)
967
+ 927 train 7.414123 (lr=2.4603e-05) (hash(x)=24968260)
968
+ 928 train 7.362894 (lr=2.4479e-05) (hash(x)=25357517)
969
+ 929 train 7.494921 (lr=2.4356e-05) (hash(x)=24854265)
970
+ 930 train 7.274008 (lr=2.4235e-05) (hash(x)=21102770)
971
+ 931 train 7.573416 (lr=2.4116e-05) (hash(x)=25676468)
972
+ 932 train 7.448094 (lr=2.3998e-05) (hash(x)=22809869)
973
+ 933 train 7.527488 (lr=2.3883e-05) (hash(x)=25503865)
974
+ 934 train 7.428623 (lr=2.3768e-05) (hash(x)=24853995)
975
+ 935 train 7.609602 (lr=2.3656e-05) (hash(x)=27544803)
976
+ 936 train 7.654929 (lr=2.3545e-05) (hash(x)=25981933)
977
+ 937 train 7.361592 (lr=2.3436e-05) (hash(x)=24658683)
978
+ 938 train 7.578220 (lr=2.3328e-05) (hash(x)=23855201)
979
+ 939 train 7.498975 (lr=2.3222e-05) (hash(x)=24331407)
980
+ 940 train 7.652050 (lr=2.3118e-05) (hash(x)=29265551)
981
+ 941 train 7.365368 (lr=2.3016e-05) (hash(x)=21892556)
982
+ 942 train 7.383709 (lr=2.2915e-05) (hash(x)=27183405)
983
+ 943 train 7.329541 (lr=2.2816e-05) (hash(x)=26540663)
984
+ 944 train 7.327266 (lr=2.2718e-05) (hash(x)=25718393)
985
+ 945 train 7.527846 (lr=2.2623e-05) (hash(x)=26819462)
986
+ 946 train 7.506734 (lr=2.2528e-05) (hash(x)=27427540)
987
+ 947 train 7.505540 (lr=2.2436e-05) (hash(x)=25532657)
988
+ 948 train 7.730583 (lr=2.2345e-05) (hash(x)=27641372)
989
+ 949 train 7.443826 (lr=2.2256e-05) (hash(x)=26515570)
990
+ 950 val loss 7.4778
991
+ 950 val perplexity 1768.3257
992
+ 950 train 7.711879 (lr=2.2169e-05) (hash(x)=26911957)
993
+ 951 train 7.774530 (lr=2.2084e-05) (hash(x)=25856625)
994
+ 952 train 7.349820 (lr=2.2000e-05) (hash(x)=25219129)
995
+ 953 train 7.307069 (lr=2.1918e-05) (hash(x)=25260471)
996
+ 954 train 7.680506 (lr=2.1837e-05) (hash(x)=29373370)
997
+ 955 train 7.287915 (lr=2.1758e-05) (hash(x)=23437426)
998
+ 956 train 7.587149 (lr=2.1681e-05) (hash(x)=23769521)
999
+ 957 train 7.587186 (lr=2.1606e-05) (hash(x)=25961833)
1000
+ 958 train 7.322324 (lr=2.1532e-05) (hash(x)=23582666)
1001
+ 959 train 7.409260 (lr=2.1461e-05) (hash(x)=23164356)
1002
+ 960 train 7.413455 (lr=2.1390e-05) (hash(x)=24443114)
1003
+ 961 train 7.451641 (lr=2.1322e-05) (hash(x)=25052665)
1004
+ 962 train 7.560568 (lr=2.1255e-05) (hash(x)=27802272)
1005
+ 963 train 7.439767 (lr=2.1190e-05) (hash(x)=25957896)
1006
+ 964 train 7.461141 (lr=2.1127e-05) (hash(x)=26737251)
1007
+ 965 train 7.583496 (lr=2.1065e-05) (hash(x)=24723263)
1008
+ 966 train 7.537825 (lr=2.1005e-05) (hash(x)=24707011)
1009
+ 967 train 7.490358 (lr=2.0947e-05) (hash(x)=25646282)
1010
+ 968 train 7.529397 (lr=2.0891e-05) (hash(x)=27544665)
1011
+ 969 train 7.456842 (lr=2.0836e-05) (hash(x)=25851993)
1012
+ 970 train 7.733618 (lr=2.0783e-05) (hash(x)=29059700)
1013
+ 971 train 7.443387 (lr=2.0732e-05) (hash(x)=21513584)
1014
+ 972 train 7.247725 (lr=2.0682e-05) (hash(x)=23151267)
1015
+ 973 train 7.479581 (lr=2.0634e-05) (hash(x)=26017176)
1016
+ 974 train 7.565686 (lr=2.0588e-05) (hash(x)=26979518)
1017
+ 975 train 7.531202 (lr=2.0544e-05) (hash(x)=23843233)
1018
+ 976 train 7.692906 (lr=2.0501e-05) (hash(x)=24193010)
1019
+ 977 train 7.418847 (lr=2.0460e-05) (hash(x)=21476847)
1020
+ 978 train 7.319603 (lr=2.0421e-05) (hash(x)=21366504)
1021
+ 979 train 7.357791 (lr=2.0384e-05) (hash(x)=23226697)
1022
+ 980 train 7.198592 (lr=2.0348e-05) (hash(x)=19961773)
1023
+ 981 train 7.061893 (lr=2.0314e-05) (hash(x)=19772969)
1024
+ 982 train 7.367008 (lr=2.0282e-05) (hash(x)=23110142)
1025
+ 983 train 7.603434 (lr=2.0252e-05) (hash(x)=24506028)
1026
+ 984 train 7.602077 (lr=2.0223e-05) (hash(x)=25480731)
1027
+ 985 train 7.448381 (lr=2.0196e-05) (hash(x)=21077417)
1028
+ 986 train 7.411280 (lr=2.0171e-05) (hash(x)=23686713)
1029
+ 987 train 7.465852 (lr=2.0147e-05) (hash(x)=26024321)
1030
+ 988 train 7.555223 (lr=2.0125e-05) (hash(x)=27424109)
1031
+ 989 train 7.652968 (lr=2.0105e-05) (hash(x)=27786174)
1032
+ 990 train 7.466365 (lr=2.0087e-05) (hash(x)=25232502)
1033
+ 991 train 7.440222 (lr=2.0071e-05) (hash(x)=22781277)
1034
+ 992 train 7.501549 (lr=2.0056e-05) (hash(x)=26184527)
1035
+ 993 train 7.514047 (lr=2.0043e-05) (hash(x)=24459895)
1036
+ 994 train 7.486927 (lr=2.0031e-05) (hash(x)=25244624)
1037
+ 995 train 7.596182 (lr=2.0022e-05) (hash(x)=24451843)
1038
+ 996 train 7.589316 (lr=2.0014e-05) (hash(x)=22129897)
1039
+ 997 train 7.404402 (lr=2.0008e-05) (hash(x)=21116390)
1040
+ 998 train 7.269816 (lr=2.0003e-05) (hash(x)=20650070)
1041
+ 999 val loss 7.4762
1042
+ 999 val perplexity 1765.5487
1043
+ 999 train 7.608737 (lr=2.0001e-05) (hash(x)=24948650)
lr2e-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 sha256:6860362c34b6ad48881acb9faa23676c2745962f854e423f1294feadac85ea10
3
+ size 48156226
lr2e-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 sha256:726f514b93717f5077765a6c0d0e23065306646bf752ab4ca3340f8e72dcd53f
3
+ size 71157574