andrew-healey commited on
Commit
34c4f15
·
verified ·
1 Parent(s): 7b33959

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size40960_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size40960_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": 40960, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0001, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1e-4_total_batch_size40960_seq_len512/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cec289292cf51a5462e2631af59109e64df3eca2d9e62e1fbd1c53a776a0f874
3
+ size 964
lr1e-4_total_batch_size40960_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 11.768598 (lr=3.4965e-07) (hash(x)=20782690)
5
+ 1 train 11.786053 (lr=6.9930e-07) (hash(x)=23890908)
6
+ 2 train 11.768688 (lr=1.0490e-06) (hash(x)=24611628)
7
+ 3 train 11.759061 (lr=1.3986e-06) (hash(x)=25767553)
8
+ 4 train 11.747805 (lr=1.7483e-06) (hash(x)=22512166)
9
+ 5 train 11.747869 (lr=2.0979e-06) (hash(x)=24732833)
10
+ 6 train 11.754723 (lr=2.4476e-06) (hash(x)=25129410)
11
+ 7 train 11.757739 (lr=2.7972e-06) (hash(x)=26582391)
12
+ 8 train 11.737188 (lr=3.1469e-06) (hash(x)=26472336)
13
+ 9 train 11.743741 (lr=3.4965e-06) (hash(x)=27654289)
14
+ 10 train 11.730678 (lr=3.8462e-06) (hash(x)=26559876)
15
+ 11 train 11.731321 (lr=4.1958e-06) (hash(x)=21770329)
16
+ 12 train 11.707128 (lr=4.5455e-06) (hash(x)=22723459)
17
+ 13 train 11.695724 (lr=4.8951e-06) (hash(x)=30338342)
18
+ 14 train 11.703027 (lr=5.2448e-06) (hash(x)=25678059)
19
+ 15 train 11.684772 (lr=5.5944e-06) (hash(x)=26109335)
20
+ 16 train 11.678885 (lr=5.9441e-06) (hash(x)=25044885)
21
+ 17 train 11.671003 (lr=6.2937e-06) (hash(x)=28870690)
22
+ 18 train 11.645406 (lr=6.6434e-06) (hash(x)=35279941)
23
+ 19 train 11.638879 (lr=6.9930e-06) (hash(x)=25015923)
24
+ 20 train 11.619357 (lr=7.3427e-06) (hash(x)=26740855)
25
+ 21 train 11.615834 (lr=7.6923e-06) (hash(x)=22262151)
26
+ 22 train 11.581081 (lr=8.0420e-06) (hash(x)=26621834)
27
+ 23 train 11.559290 (lr=8.3916e-06) (hash(x)=23234741)
28
+ 24 train 11.568782 (lr=8.7413e-06) (hash(x)=25470563)
29
+ 25 train 11.542329 (lr=9.0909e-06) (hash(x)=21799102)
30
+ 26 train 11.524164 (lr=9.4406e-06) (hash(x)=23275479)
31
+ 27 train 11.509420 (lr=9.7902e-06) (hash(x)=22091862)
32
+ 28 train 11.466881 (lr=1.0140e-05) (hash(x)=17430373)
33
+ 29 train 11.452047 (lr=1.0490e-05) (hash(x)=22954861)
34
+ 30 train 11.440687 (lr=1.0839e-05) (hash(x)=26671799)
35
+ 31 train 11.428970 (lr=1.1189e-05) (hash(x)=26924723)
36
+ 32 train 11.409254 (lr=1.1538e-05) (hash(x)=23557495)
37
+ 33 train 11.348234 (lr=1.1888e-05) (hash(x)=27804380)
38
+ 34 train 11.359153 (lr=1.2238e-05) (hash(x)=26654908)
39
+ 35 train 11.311438 (lr=1.2587e-05) (hash(x)=23627518)
40
+ 36 train 11.292107 (lr=1.2937e-05) (hash(x)=21889990)
41
+ 37 train 11.270789 (lr=1.3287e-05) (hash(x)=22528001)
42
+ 38 train 11.264645 (lr=1.3636e-05) (hash(x)=29248942)
43
+ 39 train 11.186676 (lr=1.3986e-05) (hash(x)=25533417)
44
+ 40 train 11.140759 (lr=1.4336e-05) (hash(x)=20612533)
45
+ 41 train 11.126341 (lr=1.4685e-05) (hash(x)=24677740)
46
+ 42 train 11.111518 (lr=1.5035e-05) (hash(x)=25347203)
47
+ 43 train 11.053201 (lr=1.5385e-05) (hash(x)=28922363)
48
+ 44 train 11.030757 (lr=1.5734e-05) (hash(x)=25631397)
49
+ 45 train 10.987176 (lr=1.6084e-05) (hash(x)=26237983)
50
+ 46 train 10.952265 (lr=1.6434e-05) (hash(x)=24742645)
51
+ 47 train 10.945024 (lr=1.6783e-05) (hash(x)=26157660)
52
+ 48 train 10.868473 (lr=1.7133e-05) (hash(x)=22582169)
53
+ 49 train 10.837150 (lr=1.7483e-05) (hash(x)=26431960)
54
+ 50 val loss 10.8427
55
+ 50 val perplexity 51159.7617
56
+ 50 train 10.849695 (lr=1.7832e-05) (hash(x)=25716176)
57
+ 51 train 10.799539 (lr=1.8182e-05) (hash(x)=28019788)
58
+ 52 train 10.774885 (lr=1.8531e-05) (hash(x)=20382963)
59
+ 53 train 10.734505 (lr=1.8881e-05) (hash(x)=24489647)
60
+ 54 train 10.719847 (lr=1.9231e-05) (hash(x)=25128502)
61
+ 55 train 10.699546 (lr=1.9580e-05) (hash(x)=25231390)
62
+ 56 train 10.619297 (lr=1.9930e-05) (hash(x)=22708977)
63
+ 57 train 10.682159 (lr=2.0280e-05) (hash(x)=24364920)
64
+ 58 train 10.638140 (lr=2.0629e-05) (hash(x)=24932925)
65
+ 59 train 10.613176 (lr=2.0979e-05) (hash(x)=26913684)
66
+ 60 train 10.540243 (lr=2.1329e-05) (hash(x)=18017792)
67
+ 61 train 10.566302 (lr=2.1678e-05) (hash(x)=28613882)
68
+ 62 train 10.566837 (lr=2.2028e-05) (hash(x)=22652243)
69
+ 63 train 10.505860 (lr=2.2378e-05) (hash(x)=27194521)
70
+ 64 train 10.498785 (lr=2.2727e-05) (hash(x)=24936836)
71
+ 65 train 10.518402 (lr=2.3077e-05) (hash(x)=24765578)
72
+ 66 train 10.546158 (lr=2.3427e-05) (hash(x)=23107416)
73
+ 67 train 10.437639 (lr=2.3776e-05) (hash(x)=29314255)
74
+ 68 train 10.427565 (lr=2.4126e-05) (hash(x)=24013079)
75
+ 69 train 10.405937 (lr=2.4476e-05) (hash(x)=25322001)
76
+ 70 train 10.500807 (lr=2.4825e-05) (hash(x)=26690440)
77
+ 71 train 10.439799 (lr=2.5175e-05) (hash(x)=22598512)
78
+ 72 train 10.416533 (lr=2.5524e-05) (hash(x)=18473911)
79
+ 73 train 10.466362 (lr=2.5874e-05) (hash(x)=22322442)
80
+ 74 train 10.445258 (lr=2.6224e-05) (hash(x)=26324153)
81
+ 75 train 10.417565 (lr=2.6573e-05) (hash(x)=23446058)
82
+ 76 train 10.463790 (lr=2.6923e-05) (hash(x)=24843741)
83
+ 77 train 10.429681 (lr=2.7273e-05) (hash(x)=23197102)
84
+ 78 train 10.503236 (lr=2.7622e-05) (hash(x)=23054940)
85
+ 79 train 10.440751 (lr=2.7972e-05) (hash(x)=27140757)
86
+ 80 train 10.326969 (lr=2.8322e-05) (hash(x)=25149419)
87
+ 81 train 10.402832 (lr=2.8671e-05) (hash(x)=24727764)
88
+ 82 train 10.274497 (lr=2.9021e-05) (hash(x)=21935931)
89
+ 83 train 10.368818 (lr=2.9371e-05) (hash(x)=27606073)
90
+ 84 train 10.395518 (lr=2.9720e-05) (hash(x)=22450341)
91
+ 85 train 10.399356 (lr=3.0070e-05) (hash(x)=25945859)
92
+ 86 train 10.353920 (lr=3.0420e-05) (hash(x)=21613707)
93
+ 87 train 10.366814 (lr=3.0769e-05) (hash(x)=24396519)
94
+ 88 train 10.361895 (lr=3.1119e-05) (hash(x)=29138828)
95
+ 89 train 10.434487 (lr=3.1469e-05) (hash(x)=24685515)
96
+ 90 train 10.327500 (lr=3.1818e-05) (hash(x)=22825749)
97
+ 91 train 10.354466 (lr=3.2168e-05) (hash(x)=28767869)
98
+ 92 train 10.339799 (lr=3.2517e-05) (hash(x)=21159060)
99
+ 93 train 10.321786 (lr=3.2867e-05) (hash(x)=25045397)
100
+ 94 train 10.399552 (lr=3.3217e-05) (hash(x)=25560928)
101
+ 95 train 10.358747 (lr=3.3566e-05) (hash(x)=28959222)
102
+ 96 train 10.308291 (lr=3.3916e-05) (hash(x)=21248405)
103
+ 97 train 10.315656 (lr=3.4266e-05) (hash(x)=25830182)
104
+ 98 train 10.263159 (lr=3.4615e-05) (hash(x)=25173113)
105
+ 99 train 10.316659 (lr=3.4965e-05) (hash(x)=23486277)
106
+ 100 val loss 10.3136
107
+ 100 val perplexity 30138.7266
108
+ 100 train 10.278738 (lr=3.5315e-05) (hash(x)=28916006)
109
+ 101 train 10.347554 (lr=3.5664e-05) (hash(x)=25556929)
110
+ 102 train 10.305278 (lr=3.6014e-05) (hash(x)=29648798)
111
+ 103 train 10.257299 (lr=3.6364e-05) (hash(x)=24976217)
112
+ 104 train 10.276778 (lr=3.6713e-05) (hash(x)=31019606)
113
+ 105 train 10.283932 (lr=3.7063e-05) (hash(x)=28269421)
114
+ 106 train 10.253519 (lr=3.7413e-05) (hash(x)=24865358)
115
+ 107 train 10.158926 (lr=3.7762e-05) (hash(x)=26984272)
116
+ 108 train 10.257380 (lr=3.8112e-05) (hash(x)=24544116)
117
+ 109 train 10.240767 (lr=3.8462e-05) (hash(x)=22059850)
118
+ 110 train 10.227790 (lr=3.8811e-05) (hash(x)=23926632)
119
+ 111 train 10.198865 (lr=3.9161e-05) (hash(x)=27025333)
120
+ 112 train 10.186539 (lr=3.9510e-05) (hash(x)=29120407)
121
+ 113 train 10.234774 (lr=3.9860e-05) (hash(x)=23827393)
122
+ 114 train 10.161518 (lr=4.0210e-05) (hash(x)=27230027)
123
+ 115 train 10.172895 (lr=4.0559e-05) (hash(x)=24968260)
124
+ 116 train 10.209167 (lr=4.0909e-05) (hash(x)=27544803)
125
+ 117 train 10.185083 (lr=4.1259e-05) (hash(x)=26540663)
126
+ 118 train 10.209141 (lr=4.1608e-05) (hash(x)=25856625)
127
+ 119 train 10.159639 (lr=4.1958e-05) (hash(x)=23164356)
128
+ 120 train 10.183915 (lr=4.2308e-05) (hash(x)=25646282)
129
+ 121 train 10.201326 (lr=4.2657e-05) (hash(x)=23843233)
130
+ 122 train 10.101642 (lr=4.3007e-05) (hash(x)=24506028)
131
+ 123 train 10.141010 (lr=4.3357e-05) (hash(x)=22781277)
132
+ 124 train 10.162203 (lr=4.3706e-05) (hash(x)=24948650)
133
+ 125 train 10.117309 (lr=4.4056e-05) (hash(x)=26373991)
134
+ 126 train 10.160105 (lr=4.4406e-05) (hash(x)=23402396)
135
+ 127 train 10.119009 (lr=4.4755e-05) (hash(x)=21562133)
136
+ 128 train 10.162827 (lr=4.5105e-05) (hash(x)=25009210)
137
+ 129 train 10.060822 (lr=4.5455e-05) (hash(x)=27161811)
138
+ 130 train 10.063899 (lr=4.5804e-05) (hash(x)=29515100)
139
+ 131 train 10.056215 (lr=4.6154e-05) (hash(x)=25305929)
140
+ 132 train 10.066180 (lr=4.6503e-05) (hash(x)=26808543)
141
+ 133 train 10.030074 (lr=4.6853e-05) (hash(x)=25669509)
142
+ 134 train 10.047958 (lr=4.7203e-05) (hash(x)=28435805)
143
+ 135 train 10.047024 (lr=4.7552e-05) (hash(x)=26482758)
144
+ 136 train 9.992270 (lr=4.7902e-05) (hash(x)=24929929)
145
+ 137 train 10.049812 (lr=4.8252e-05) (hash(x)=25043037)
146
+ 138 train 10.060753 (lr=4.8601e-05) (hash(x)=29151257)
147
+ 139 train 10.021026 (lr=4.8951e-05) (hash(x)=26400365)
148
+ 140 train 9.950485 (lr=4.9301e-05) (hash(x)=24098812)
149
+ 141 train 10.005425 (lr=4.9650e-05) (hash(x)=20967666)
150
+ 142 train 10.025003 (lr=5.0000e-05) (hash(x)=28667963)
151
+ 143 train 10.000402 (lr=5.0350e-05) (hash(x)=27083417)
152
+ 144 train 9.966627 (lr=5.0699e-05) (hash(x)=23649001)
153
+ 145 train 9.948689 (lr=5.1049e-05) (hash(x)=31338300)
154
+ 146 train 9.906151 (lr=5.1399e-05) (hash(x)=30603174)
155
+ 147 train 9.904183 (lr=5.1748e-05) (hash(x)=24830965)
156
+ 148 train 9.939874 (lr=5.2098e-05) (hash(x)=27269893)
157
+ 149 train 9.906730 (lr=5.2448e-05) (hash(x)=24825041)
158
+ 150 val loss 9.8926
159
+ 150 val perplexity 19783.4316
160
+ 150 train 9.884244 (lr=5.2797e-05) (hash(x)=22724871)
161
+ 151 train 9.822634 (lr=5.3147e-05) (hash(x)=24683717)
162
+ 152 train 9.867231 (lr=5.3497e-05) (hash(x)=26190337)
163
+ 153 train 9.851982 (lr=5.3846e-05) (hash(x)=23086289)
164
+ 154 train 9.881013 (lr=5.4196e-05) (hash(x)=23593875)
165
+ 155 train 9.842882 (lr=5.4545e-05) (hash(x)=22414190)
166
+ 156 train 9.833754 (lr=5.4895e-05) (hash(x)=26758114)
167
+ 157 train 9.797348 (lr=5.5245e-05) (hash(x)=21563184)
168
+ 158 train 9.810123 (lr=5.5594e-05) (hash(x)=19218470)
169
+ 159 train 9.827530 (lr=5.5944e-05) (hash(x)=25588469)
170
+ 160 train 9.773418 (lr=5.6294e-05) (hash(x)=30433767)
171
+ 161 train 9.758258 (lr=5.6643e-05) (hash(x)=17851621)
172
+ 162 train 9.797021 (lr=5.6993e-05) (hash(x)=25183690)
173
+ 163 train 9.825255 (lr=5.7343e-05) (hash(x)=27040642)
174
+ 164 train 9.700119 (lr=5.7692e-05) (hash(x)=22183303)
175
+ 165 train 9.833939 (lr=5.8042e-05) (hash(x)=27537063)
176
+ 166 train 9.701316 (lr=5.8392e-05) (hash(x)=19492832)
177
+ 167 train 9.662508 (lr=5.8741e-05) (hash(x)=25951629)
178
+ 168 train 9.630916 (lr=5.9091e-05) (hash(x)=22408682)
179
+ 169 train 9.688861 (lr=5.9441e-05) (hash(x)=27633457)
180
+ 170 train 9.684103 (lr=5.9790e-05) (hash(x)=27635080)
181
+ 171 train 9.680268 (lr=6.0140e-05) (hash(x)=22287596)
182
+ 172 train 9.690231 (lr=6.0490e-05) (hash(x)=35895440)
183
+ 173 train 9.669487 (lr=6.0839e-05) (hash(x)=25888847)
184
+ 174 train 9.591892 (lr=6.1189e-05) (hash(x)=26518800)
185
+ 175 train 9.635827 (lr=6.1538e-05) (hash(x)=37519283)
186
+ 176 train 9.680520 (lr=6.1888e-05) (hash(x)=25141945)
187
+ 177 train 9.705601 (lr=6.2238e-05) (hash(x)=28527939)
188
+ 178 train 9.587241 (lr=6.2587e-05) (hash(x)=21942471)
189
+ 179 train 9.603369 (lr=6.2937e-05) (hash(x)=24955630)
190
+ 180 train 9.544624 (lr=6.3287e-05) (hash(x)=25541230)
191
+ 181 train 9.475161 (lr=6.3636e-05) (hash(x)=23501481)
192
+ 182 train 9.496090 (lr=6.3986e-05) (hash(x)=22754988)
193
+ 183 train 9.434362 (lr=6.4336e-05) (hash(x)=23475070)
194
+ 184 train 9.505939 (lr=6.4685e-05) (hash(x)=23609959)
195
+ 185 train 9.474903 (lr=6.5035e-05) (hash(x)=23441847)
196
+ 186 train 9.463425 (lr=6.5385e-05) (hash(x)=23096292)
197
+ 187 train 9.471959 (lr=6.5734e-05) (hash(x)=26171093)
198
+ 188 train 9.478407 (lr=6.6084e-05) (hash(x)=20060838)
199
+ 189 train 9.401480 (lr=6.6434e-05) (hash(x)=25451055)
200
+ 190 train 9.306700 (lr=6.6783e-05) (hash(x)=24934792)
201
+ 191 train 9.332983 (lr=6.7133e-05) (hash(x)=25367597)
202
+ 192 train 9.398968 (lr=6.7483e-05) (hash(x)=25022148)
203
+ 193 train 9.393510 (lr=6.7832e-05) (hash(x)=25431365)
204
+ 194 train 9.383740 (lr=6.8182e-05) (hash(x)=29307788)
205
+ 195 train 9.303251 (lr=6.8531e-05) (hash(x)=25651929)
206
+ 196 train 9.343544 (lr=6.8881e-05) (hash(x)=26312387)
207
+ 197 train 9.326360 (lr=6.9231e-05) (hash(x)=23804913)
208
+ 198 train 9.381145 (lr=6.9580e-05) (hash(x)=26115519)
209
+ 199 train 9.279168 (lr=6.9930e-05) (hash(x)=23210747)
210
+ 200 val loss 9.2906
211
+ 200 val perplexity 10835.8887
212
+ 200 train 9.307692 (lr=7.0280e-05) (hash(x)=18463712)
213
+ 201 train 9.248871 (lr=7.0629e-05) (hash(x)=23245699)
214
+ 202 train 9.257250 (lr=7.0979e-05) (hash(x)=21770300)
215
+ 203 train 9.407807 (lr=7.1329e-05) (hash(x)=25387532)
216
+ 204 train 9.254056 (lr=7.1678e-05) (hash(x)=27718703)
217
+ 205 train 9.333721 (lr=7.2028e-05) (hash(x)=27661052)
218
+ 206 train 9.207791 (lr=7.2378e-05) (hash(x)=26538928)
219
+ 207 train 9.226565 (lr=7.2727e-05) (hash(x)=21772654)
220
+ 208 train 9.227448 (lr=7.3077e-05) (hash(x)=26263254)
221
+ 209 train 9.214375 (lr=7.3427e-05) (hash(x)=23552356)
222
+ 210 train 9.296247 (lr=7.3776e-05) (hash(x)=22707141)
223
+ 211 train 9.288834 (lr=7.4126e-05) (hash(x)=34234481)
224
+ 212 train 9.183087 (lr=7.4476e-05) (hash(x)=35008670)
225
+ 213 train 9.190234 (lr=7.4825e-05) (hash(x)=25891543)
226
+ 214 train 9.096066 (lr=7.5175e-05) (hash(x)=21971224)
227
+ 215 train 9.123388 (lr=7.5524e-05) (hash(x)=27312186)
228
+ 216 train 9.164174 (lr=7.5874e-05) (hash(x)=22163749)
229
+ 217 train 9.005236 (lr=7.6224e-05) (hash(x)=25942888)
230
+ 218 train 9.101052 (lr=7.6573e-05) (hash(x)=25493916)
231
+ 219 train 9.198045 (lr=7.6923e-05) (hash(x)=24289951)
232
+ 220 train 9.056392 (lr=7.7273e-05) (hash(x)=23648671)
233
+ 221 train 9.034428 (lr=7.7622e-05) (hash(x)=24992055)
234
+ 222 train 9.113957 (lr=7.7972e-05) (hash(x)=26005549)
235
+ 223 train 9.034801 (lr=7.8322e-05) (hash(x)=21046554)
236
+ 224 train 9.012487 (lr=7.8671e-05) (hash(x)=26321539)
237
+ 225 train 9.029699 (lr=7.9021e-05) (hash(x)=27807872)
238
+ 226 train 9.051228 (lr=7.9371e-05) (hash(x)=21555364)
239
+ 227 train 9.027201 (lr=7.9720e-05) (hash(x)=26325324)
240
+ 228 train 8.983374 (lr=8.0070e-05) (hash(x)=24956294)
241
+ 229 train 8.982791 (lr=8.0420e-05) (hash(x)=24429389)
242
+ 230 train 9.016384 (lr=8.0769e-05) (hash(x)=27929846)
243
+ 231 train 8.987251 (lr=8.1119e-05) (hash(x)=25282897)
244
+ 232 train 8.908170 (lr=8.1469e-05) (hash(x)=26802508)
245
+ 233 train 8.965171 (lr=8.1818e-05) (hash(x)=28390416)
246
+ 234 train 9.025198 (lr=8.2168e-05) (hash(x)=26292451)
247
+ 235 train 8.994064 (lr=8.2517e-05) (hash(x)=22915344)
248
+ 236 train 8.934509 (lr=8.2867e-05) (hash(x)=24744383)
249
+ 237 train 8.837766 (lr=8.3217e-05) (hash(x)=26284678)
250
+ 238 train 8.864168 (lr=8.3566e-05) (hash(x)=21654269)
251
+ 239 train 9.119543 (lr=8.3916e-05) (hash(x)=29466268)
252
+ 240 train 9.278403 (lr=8.4266e-05) (hash(x)=32616762)
253
+ 241 train 9.244632 (lr=8.4615e-05) (hash(x)=32201138)
254
+ 242 train 9.227572 (lr=8.4965e-05) (hash(x)=33683468)
255
+ 243 train 8.983342 (lr=8.5315e-05) (hash(x)=27498125)
256
+ 244 train 8.884466 (lr=8.5664e-05) (hash(x)=25164318)
257
+ 245 train 8.854906 (lr=8.6014e-05) (hash(x)=25690221)
258
+ 246 train 8.865967 (lr=8.6364e-05) (hash(x)=29031550)
259
+ 247 train 8.893048 (lr=8.6713e-05) (hash(x)=24094093)
260
+ 248 train 9.176089 (lr=8.7063e-05) (hash(x)=61455960)
261
+ 249 train 10.917917 (lr=8.7413e-05) (hash(x)=61104710)
262
+ 250 val loss 8.8090
263
+ 250 val perplexity 6694.3369
264
+ 250 train 8.819347 (lr=8.7762e-05) (hash(x)=20881126)
265
+ 251 train 8.852946 (lr=8.8112e-05) (hash(x)=26920243)
266
+ 252 train 8.754860 (lr=8.8462e-05) (hash(x)=21970343)
267
+ 253 train 8.772523 (lr=8.8811e-05) (hash(x)=31676352)
268
+ 254 train 8.780556 (lr=8.9161e-05) (hash(x)=25600562)
269
+ 255 train 8.780517 (lr=8.9510e-05) (hash(x)=22022924)
270
+ 256 train 8.734633 (lr=8.9860e-05) (hash(x)=22777085)
271
+ 257 train 8.809948 (lr=9.0210e-05) (hash(x)=21727744)
272
+ 258 train 8.832301 (lr=9.0559e-05) (hash(x)=20701998)
273
+ 259 train 8.801223 (lr=9.0909e-05) (hash(x)=20863982)
274
+ 260 train 8.732726 (lr=9.1259e-05) (hash(x)=25084926)
275
+ 261 train 8.731787 (lr=9.1608e-05) (hash(x)=24412897)
276
+ 262 train 8.662621 (lr=9.1958e-05) (hash(x)=25434913)
277
+ 263 train 8.694664 (lr=9.2308e-05) (hash(x)=26302544)
278
+ 264 train 8.731335 (lr=9.2657e-05) (hash(x)=34976197)
279
+ 265 train 8.662732 (lr=9.3007e-05) (hash(x)=23870805)
280
+ 266 train 8.664909 (lr=9.3357e-05) (hash(x)=25041103)
281
+ 267 train 8.684845 (lr=9.3706e-05) (hash(x)=26041745)
282
+ 268 train 8.644638 (lr=9.4056e-05) (hash(x)=24068619)
283
+ 269 train 8.633161 (lr=9.4406e-05) (hash(x)=25441262)
284
+ 270 train 8.645530 (lr=9.4755e-05) (hash(x)=29181867)
285
+ 271 train 8.656697 (lr=9.5105e-05) (hash(x)=27093757)
286
+ 272 train 8.553125 (lr=9.5455e-05) (hash(x)=25150275)
287
+ 273 train 8.626158 (lr=9.5804e-05) (hash(x)=33457102)
288
+ 274 train 8.531854 (lr=9.6154e-05) (hash(x)=25755322)
289
+ 275 train 8.554464 (lr=9.6503e-05) (hash(x)=23630794)
290
+ 276 train 8.472621 (lr=9.6853e-05) (hash(x)=25167987)
291
+ 277 train 8.582054 (lr=9.7203e-05) (hash(x)=24043224)
292
+ 278 train 8.552578 (lr=9.7552e-05) (hash(x)=24665411)
293
+ 279 train 8.544823 (lr=9.7902e-05) (hash(x)=17321657)
294
+ 280 train 8.504908 (lr=9.8252e-05) (hash(x)=27932591)
295
+ 281 train 8.608794 (lr=9.8601e-05) (hash(x)=29447954)
296
+ 282 train 8.534290 (lr=9.8951e-05) (hash(x)=24621816)
297
+ 283 train 8.502698 (lr=9.9301e-05) (hash(x)=24682294)
298
+ 284 train 8.497841 (lr=9.9650e-05) (hash(x)=28057095)
299
+ 285 train 8.522227 (lr=1.0000e-04) (hash(x)=24547536)
300
+ 286 train 8.350670 (lr=1.0000e-04) (hash(x)=18987545)
301
+ 287 train 8.436469 (lr=1.0000e-04) (hash(x)=26745063)
302
+ 288 train 8.365735 (lr=9.9998e-05) (hash(x)=21447406)
303
+ 289 train 8.309195 (lr=9.9996e-05) (hash(x)=24650655)
304
+ 290 train 8.513535 (lr=9.9993e-05) (hash(x)=26519976)
305
+ 291 train 8.778757 (lr=9.9989e-05) (hash(x)=20106514)
306
+ 292 train 8.315075 (lr=9.9984e-05) (hash(x)=20117808)
307
+ 293 train 8.413671 (lr=9.9979e-05) (hash(x)=25510334)
308
+ 294 train 8.340737 (lr=9.9972e-05) (hash(x)=25749120)
309
+ 295 train 8.399283 (lr=9.9965e-05) (hash(x)=24357241)
310
+ 296 train 8.379560 (lr=9.9956e-05) (hash(x)=25584945)
311
+ 297 train 8.299529 (lr=9.9947e-05) (hash(x)=25198897)
312
+ 298 train 8.308249 (lr=9.9937e-05) (hash(x)=23475891)
313
+ 299 train 8.287750 (lr=9.9926e-05) (hash(x)=23644804)
314
+ 300 val loss 8.3086
315
+ 300 val perplexity 4058.6750
316
+ 300 train 8.435748 (lr=9.9915e-05) (hash(x)=25629584)
317
+ 301 train 8.480757 (lr=9.9902e-05) (hash(x)=29797280)
318
+ 302 train 8.497992 (lr=9.9889e-05) (hash(x)=30026394)
319
+ 303 train 8.419273 (lr=9.9874e-05) (hash(x)=26594177)
320
+ 304 train 8.305301 (lr=9.9859e-05) (hash(x)=23766256)
321
+ 305 train 8.259278 (lr=9.9843e-05) (hash(x)=24250150)
322
+ 306 train 8.253291 (lr=9.9826e-05) (hash(x)=26108817)
323
+ 307 train 8.200242 (lr=9.9808e-05) (hash(x)=18582688)
324
+ 308 train 8.176954 (lr=9.9789e-05) (hash(x)=23272705)
325
+ 309 train 8.152193 (lr=9.9770e-05) (hash(x)=24793412)
326
+ 310 train 8.166468 (lr=9.9749e-05) (hash(x)=19280989)
327
+ 311 train 8.157917 (lr=9.9728e-05) (hash(x)=22751150)
328
+ 312 train 8.245569 (lr=9.9706e-05) (hash(x)=23537942)
329
+ 313 train 8.102750 (lr=9.9683e-05) (hash(x)=23748375)
330
+ 314 train 8.187785 (lr=9.9659e-05) (hash(x)=27413825)
331
+ 315 train 8.191016 (lr=9.9634e-05) (hash(x)=21682445)
332
+ 316 train 8.125028 (lr=9.9609e-05) (hash(x)=22040184)
333
+ 317 train 8.069901 (lr=9.9582e-05) (hash(x)=24463229)
334
+ 318 train 8.183447 (lr=9.9555e-05) (hash(x)=24091954)
335
+ 319 train 8.102878 (lr=9.9526e-05) (hash(x)=26670219)
336
+ 320 train 8.085333 (lr=9.9497e-05) (hash(x)=25558986)
337
+ 321 train 8.049417 (lr=9.9467e-05) (hash(x)=19218943)
338
+ 322 train 8.143736 (lr=9.9437e-05) (hash(x)=24551402)
339
+ 323 train 8.042617 (lr=9.9405e-05) (hash(x)=25229281)
340
+ 324 train 7.878533 (lr=9.9372e-05) (hash(x)=21251127)
341
+ 325 train 8.064113 (lr=9.9339e-05) (hash(x)=25800857)
342
+ 326 train 7.988725 (lr=9.9305e-05) (hash(x)=25836189)
343
+ 327 train 8.010557 (lr=9.9270e-05) (hash(x)=25999255)
344
+ 328 train 8.048934 (lr=9.9234e-05) (hash(x)=20149394)
345
+ 329 train 7.997222 (lr=9.9197e-05) (hash(x)=21999338)
346
+ 330 train 8.070961 (lr=9.9159e-05) (hash(x)=27611302)
347
+ 331 train 8.037538 (lr=9.9121e-05) (hash(x)=22778356)
348
+ 332 train 7.991700 (lr=9.9081e-05) (hash(x)=24195959)
349
+ 333 train 8.208155 (lr=9.9041e-05) (hash(x)=19127465)
350
+ 334 train 8.072520 (lr=9.9000e-05) (hash(x)=22426274)
351
+ 335 train 7.929836 (lr=9.8958e-05) (hash(x)=23785205)
352
+ 336 train 8.038789 (lr=9.8915e-05) (hash(x)=28505676)
353
+ 337 train 7.982605 (lr=9.8872e-05) (hash(x)=27706477)
354
+ 338 train 8.036742 (lr=9.8827e-05) (hash(x)=28684175)
355
+ 339 train 8.014040 (lr=9.8782e-05) (hash(x)=28003600)
356
+ 340 train 8.014636 (lr=9.8736e-05) (hash(x)=27586845)
357
+ 341 train 7.955000 (lr=9.8689e-05) (hash(x)=21176405)
358
+ 342 train 7.877286 (lr=9.8641e-05) (hash(x)=25888433)
359
+ 343 train 8.016254 (lr=9.8592e-05) (hash(x)=27916982)
360
+ 344 train 7.874392 (lr=9.8543e-05) (hash(x)=24334708)
361
+ 345 train 7.941211 (lr=9.8492e-05) (hash(x)=25617709)
362
+ 346 train 8.036167 (lr=9.8441e-05) (hash(x)=25152362)
363
+ 347 train 7.900301 (lr=9.8389e-05) (hash(x)=24276512)
364
+ 348 train 7.813753 (lr=9.8336e-05) (hash(x)=26776133)
365
+ 349 train 7.775318 (lr=9.8282e-05) (hash(x)=20254159)
366
+ 350 val loss 7.9409
367
+ 350 val perplexity 2809.8557
368
+ 350 train 7.814607 (lr=9.8228e-05) (hash(x)=27873855)
369
+ 351 train 7.906783 (lr=9.8172e-05) (hash(x)=25113614)
370
+ 352 train 7.965298 (lr=9.8116e-05) (hash(x)=26192886)
371
+ 353 train 7.856303 (lr=9.8059e-05) (hash(x)=24614912)
372
+ 354 train 7.846335 (lr=9.8001e-05) (hash(x)=25397093)
373
+ 355 train 7.875048 (lr=9.7942e-05) (hash(x)=19811802)
374
+ 356 train 7.864684 (lr=9.7882e-05) (hash(x)=25579655)
375
+ 357 train 7.962258 (lr=9.7822e-05) (hash(x)=26096514)
376
+ 358 train 7.858101 (lr=9.7761e-05) (hash(x)=22589633)
377
+ 359 train 7.864480 (lr=9.7699e-05) (hash(x)=26757147)
378
+ 360 train 7.956186 (lr=9.7636e-05) (hash(x)=24963730)
379
+ 361 train 7.854289 (lr=9.7572e-05) (hash(x)=27703124)
380
+ 362 train 7.743626 (lr=9.7507e-05) (hash(x)=24781713)
381
+ 363 train 7.891638 (lr=9.7442e-05) (hash(x)=24720476)
382
+ 364 train 7.875727 (lr=9.7376e-05) (hash(x)=23481301)
383
+ 365 train 7.793710 (lr=9.7309e-05) (hash(x)=25641033)
384
+ 366 train 7.846750 (lr=9.7241e-05) (hash(x)=27369437)
385
+ 367 train 7.884250 (lr=9.7172e-05) (hash(x)=24760900)
386
+ 368 train 7.763644 (lr=9.7103e-05) (hash(x)=21692546)
387
+ 369 train 7.829751 (lr=9.7032e-05) (hash(x)=22720391)
388
+ 370 train 7.782937 (lr=9.6961e-05) (hash(x)=24376979)
389
+ 371 train 7.795093 (lr=9.6889e-05) (hash(x)=22244509)
390
+ 372 train 7.745054 (lr=9.6817e-05) (hash(x)=21831960)
391
+ 373 train 7.725542 (lr=9.6743e-05) (hash(x)=24921535)
392
+ 374 train 7.737894 (lr=9.6669e-05) (hash(x)=25381251)
393
+ 375 train 7.736143 (lr=9.6593e-05) (hash(x)=25687352)
394
+ 376 train 7.763109 (lr=9.6518e-05) (hash(x)=25221654)
395
+ 377 train 7.808217 (lr=9.6441e-05) (hash(x)=24469863)
396
+ 378 train 7.841477 (lr=9.6363e-05) (hash(x)=25291413)
397
+ 379 train 7.748312 (lr=9.6285e-05) (hash(x)=25744274)
398
+ 380 train 7.685605 (lr=9.6206e-05) (hash(x)=24246963)
399
+ 381 train 7.710941 (lr=9.6126e-05) (hash(x)=19724058)
400
+ 382 train 7.667651 (lr=9.6045e-05) (hash(x)=20705649)
401
+ 383 train 7.874396 (lr=9.5963e-05) (hash(x)=24938685)
402
+ 384 train 7.897980 (lr=9.5881e-05) (hash(x)=21707000)
403
+ 385 train 7.711274 (lr=9.5798e-05) (hash(x)=24948980)
404
+ 386 train 7.939587 (lr=9.5714e-05) (hash(x)=30773958)
405
+ 387 train 8.002645 (lr=9.5629e-05) (hash(x)=26769046)
406
+ 388 train 7.721035 (lr=9.5544e-05) (hash(x)=18849656)
407
+ 389 train 8.200633 (lr=9.5457e-05) (hash(x)=24624950)
408
+ 390 train 7.816331 (lr=9.5370e-05) (hash(x)=27299605)
409
+ 391 train 7.807438 (lr=9.5282e-05) (hash(x)=24113445)
410
+ 392 train 7.825818 (lr=9.5194e-05) (hash(x)=32644465)
411
+ 393 train 7.902504 (lr=9.5104e-05) (hash(x)=25829219)
412
+ 394 train 7.739758 (lr=9.5014e-05) (hash(x)=25194550)
413
+ 395 train 7.775651 (lr=9.4923e-05) (hash(x)=25941804)
414
+ 396 train 7.771498 (lr=9.4831e-05) (hash(x)=24075727)
415
+ 397 train 7.813376 (lr=9.4739e-05) (hash(x)=23104539)
416
+ 398 train 7.750863 (lr=9.4646e-05) (hash(x)=23018983)
417
+ 399 train 7.701500 (lr=9.4551e-05) (hash(x)=24422929)
418
+ 400 val loss 7.7568
419
+ 400 val perplexity 2337.3364
420
+ 400 train 7.736512 (lr=9.4457e-05) (hash(x)=23389696)
421
+ 401 train 7.697756 (lr=9.4361e-05) (hash(x)=29102969)
422
+ 402 train 7.838785 (lr=9.4265e-05) (hash(x)=28359859)
423
+ 403 train 7.818171 (lr=9.4168e-05) (hash(x)=24630955)
424
+ 404 train 7.868329 (lr=9.4070e-05) (hash(x)=22969113)
425
+ 405 train 7.691509 (lr=9.3971e-05) (hash(x)=27004415)
426
+ 406 train 7.626100 (lr=9.3872e-05) (hash(x)=22265063)
427
+ 407 train 7.663659 (lr=9.3772e-05) (hash(x)=26905990)
428
+ 408 train 7.703139 (lr=9.3671e-05) (hash(x)=24047679)
429
+ 409 train 7.832115 (lr=9.3569e-05) (hash(x)=26905582)
430
+ 410 train 7.611836 (lr=9.3467e-05) (hash(x)=29018970)
431
+ 411 train 7.613941 (lr=9.3364e-05) (hash(x)=24503315)
432
+ 412 train 7.623048 (lr=9.3260e-05) (hash(x)=22607537)
433
+ 413 train 8.289001 (lr=9.3155e-05) (hash(x)=26690225)
434
+ 414 train 7.682450 (lr=9.3050e-05) (hash(x)=24226255)
435
+ 415 train 7.630458 (lr=9.2944e-05) (hash(x)=23622702)
436
+ 416 train 7.686643 (lr=9.2837e-05) (hash(x)=23910425)
437
+ 417 train 7.893850 (lr=9.2729e-05) (hash(x)=27213318)
438
+ 418 train 7.703747 (lr=9.2621e-05) (hash(x)=26701577)
439
+ 419 train 7.645214 (lr=9.2512e-05) (hash(x)=26216794)
440
+ 420 train 7.847490 (lr=9.2402e-05) (hash(x)=27579623)
441
+ 421 train 7.690247 (lr=9.2292e-05) (hash(x)=26086198)
442
+ 422 train 7.662221 (lr=9.2181e-05) (hash(x)=24863441)
443
+ 423 train 7.781695 (lr=9.2069e-05) (hash(x)=25236385)
444
+ 424 train 7.803724 (lr=9.1956e-05) (hash(x)=25330803)
445
+ 425 train 7.693798 (lr=9.1843e-05) (hash(x)=22508532)
446
+ 426 train 7.672997 (lr=9.1729e-05) (hash(x)=21799261)
447
+ 427 train 7.764224 (lr=9.1614e-05) (hash(x)=26311168)
448
+ 428 train 7.560589 (lr=9.1499e-05) (hash(x)=20977430)
449
+ 429 train 7.622502 (lr=9.1382e-05) (hash(x)=26941617)
450
+ 430 train 7.873599 (lr=9.1266e-05) (hash(x)=30008957)
451
+ 431 train 7.739056 (lr=9.1148e-05) (hash(x)=27125962)
452
+ 432 train 7.598950 (lr=9.1030e-05) (hash(x)=25888358)
453
+ 433 train 7.733124 (lr=9.0911e-05) (hash(x)=21725719)
454
+ 434 train 7.584195 (lr=9.0791e-05) (hash(x)=21669704)
455
+ 435 train 7.572590 (lr=9.0671e-05) (hash(x)=23166993)
456
+ 436 train 7.638998 (lr=9.0550e-05) (hash(x)=26235974)
457
+ 437 train 7.788588 (lr=9.0428e-05) (hash(x)=25426430)
458
+ 438 train 7.534440 (lr=9.0306e-05) (hash(x)=22966314)
459
+ 439 train 7.683634 (lr=9.0182e-05) (hash(x)=23448505)
460
+ 440 train 7.550514 (lr=9.0059e-05) (hash(x)=26330693)
461
+ 441 train 7.503012 (lr=8.9934e-05) (hash(x)=29550596)
462
+ 442 train 7.569022 (lr=8.9809e-05) (hash(x)=22124892)
463
+ 443 train 7.593273 (lr=8.9683e-05) (hash(x)=25607640)
464
+ 444 train 7.440102 (lr=8.9557e-05) (hash(x)=23795894)
465
+ 445 train 7.334437 (lr=8.9430e-05) (hash(x)=24165449)
466
+ 446 train 7.597899 (lr=8.9302e-05) (hash(x)=24697756)
467
+ 447 train 7.681522 (lr=8.9173e-05) (hash(x)=27218899)
468
+ 448 train 7.567974 (lr=8.9044e-05) (hash(x)=24748522)
469
+ 449 train 7.648011 (lr=8.8914e-05) (hash(x)=24965317)
470
+ 450 val loss 7.6558
471
+ 450 val perplexity 2112.8896
472
+ 450 train 7.441152 (lr=8.8784e-05) (hash(x)=22031210)
473
+ 451 train 7.567151 (lr=8.8653e-05) (hash(x)=25877990)
474
+ 452 train 7.608446 (lr=8.8521e-05) (hash(x)=24382381)
475
+ 453 train 7.512006 (lr=8.8388e-05) (hash(x)=24551999)
476
+ 454 train 7.613674 (lr=8.8255e-05) (hash(x)=25765516)
477
+ 455 train 7.634459 (lr=8.8122e-05) (hash(x)=25204247)
478
+ 456 train 7.681192 (lr=8.7987e-05) (hash(x)=26321813)
479
+ 457 train 7.611825 (lr=8.7852e-05) (hash(x)=25528323)
480
+ 458 train 7.602286 (lr=8.7717e-05) (hash(x)=25114729)
481
+ 459 train 7.576212 (lr=8.7580e-05) (hash(x)=31341964)
482
+ 460 train 7.621799 (lr=8.7444e-05) (hash(x)=26326024)
483
+ 461 train 7.538342 (lr=8.7306e-05) (hash(x)=27273311)
484
+ 462 train 7.527748 (lr=8.7168e-05) (hash(x)=17819313)
485
+ 463 train 7.676411 (lr=8.7029e-05) (hash(x)=23769243)
486
+ 464 train 7.633408 (lr=8.6890e-05) (hash(x)=23788738)
487
+ 465 train 7.584522 (lr=8.6750e-05) (hash(x)=24818744)
488
+ 466 train 7.514266 (lr=8.6609e-05) (hash(x)=23472795)
489
+ 467 train 7.910563 (lr=8.6468e-05) (hash(x)=24991832)
490
+ 468 train 7.652891 (lr=8.6326e-05) (hash(x)=22424066)
491
+ 469 train 7.544734 (lr=8.6184e-05) (hash(x)=26474878)
492
+ 470 train 7.638586 (lr=8.6041e-05) (hash(x)=22704349)
493
+ 471 train 7.575641 (lr=8.5897e-05) (hash(x)=22293673)
494
+ 472 train 7.600972 (lr=8.5753e-05) (hash(x)=21785487)
495
+ 473 train 7.602975 (lr=8.5608e-05) (hash(x)=25251063)
496
+ 474 train 7.506220 (lr=8.5462e-05) (hash(x)=22722211)
497
+ 475 train 7.570879 (lr=8.5316e-05) (hash(x)=24522269)
498
+ 476 train 7.521543 (lr=8.5170e-05) (hash(x)=23329475)
499
+ 477 train 7.545805 (lr=8.5022e-05) (hash(x)=24735165)
500
+ 478 train 7.364789 (lr=8.4875e-05) (hash(x)=27002892)
501
+ 479 train 7.408906 (lr=8.4726e-05) (hash(x)=23499349)
502
+ 480 train 7.739799 (lr=8.4577e-05) (hash(x)=22461589)
503
+ 481 train 7.506948 (lr=8.4428e-05) (hash(x)=26415040)
504
+ 482 train 7.734367 (lr=8.4278e-05) (hash(x)=26349465)
505
+ 483 train 7.407123 (lr=8.4127e-05) (hash(x)=21596677)
506
+ 484 train 7.458468 (lr=8.3976e-05) (hash(x)=24117012)
507
+ 485 train 7.531465 (lr=8.3824e-05) (hash(x)=24105761)
508
+ 486 train 7.468420 (lr=8.3672e-05) (hash(x)=25725049)
509
+ 487 train 7.322726 (lr=8.3519e-05) (hash(x)=22892776)
510
+ 488 train 7.429528 (lr=8.3366e-05) (hash(x)=21630383)
511
+ 489 train 7.487536 (lr=8.3212e-05) (hash(x)=24727094)
512
+ 490 train 7.420637 (lr=8.3057e-05) (hash(x)=22835000)
513
+ 491 train 7.591608 (lr=8.2902e-05) (hash(x)=22839049)
514
+ 492 train 7.435911 (lr=8.2746e-05) (hash(x)=20654136)
515
+ 493 train 7.659186 (lr=8.2590e-05) (hash(x)=26776000)
516
+ 494 train 7.273840 (lr=8.2434e-05) (hash(x)=22560658)
517
+ 495 train 7.715094 (lr=8.2276e-05) (hash(x)=25995933)
518
+ 496 train 7.653127 (lr=8.2119e-05) (hash(x)=24635394)
519
+ 497 train 7.734672 (lr=8.1960e-05) (hash(x)=27537228)
520
+ 498 train 7.632674 (lr=8.1801e-05) (hash(x)=26496730)
521
+ 499 train 7.370257 (lr=8.1642e-05) (hash(x)=16533310)
522
+ 500 val loss 7.6008
523
+ 500 val perplexity 1999.7646
524
+ 500 train 7.491931 (lr=8.1482e-05) (hash(x)=24881176)
525
+ 501 train 7.699295 (lr=8.1322e-05) (hash(x)=25593613)
526
+ 502 train 7.456593 (lr=8.1161e-05) (hash(x)=22886646)
527
+ 503 train 7.718415 (lr=8.1000e-05) (hash(x)=21407676)
528
+ 504 train 7.588889 (lr=8.0838e-05) (hash(x)=21720304)
529
+ 505 train 7.715669 (lr=8.0676e-05) (hash(x)=27263416)
530
+ 506 train 7.482646 (lr=8.0513e-05) (hash(x)=25840801)
531
+ 507 train 7.624564 (lr=8.0349e-05) (hash(x)=31424568)
532
+ 508 train 7.625317 (lr=8.0186e-05) (hash(x)=23403065)
533
+ 509 train 7.449756 (lr=8.0021e-05) (hash(x)=22656033)
534
+ 510 train 7.584896 (lr=7.9856e-05) (hash(x)=26869295)
535
+ 511 train 7.835832 (lr=7.9691e-05) (hash(x)=25504151)
536
+ 512 train 7.613889 (lr=7.9525e-05) (hash(x)=25781518)
537
+ 513 train 7.579289 (lr=7.9359e-05) (hash(x)=26432850)
538
+ 514 train 7.452091 (lr=7.9192e-05) (hash(x)=20155894)
539
+ 515 train 7.367572 (lr=7.9025e-05) (hash(x)=22270169)
540
+ 516 train 7.516262 (lr=7.8858e-05) (hash(x)=22108461)
541
+ 517 train 7.582701 (lr=7.8689e-05) (hash(x)=24802848)
542
+ 518 train 7.664511 (lr=7.8521e-05) (hash(x)=25086604)
543
+ 519 train 7.665092 (lr=7.8352e-05) (hash(x)=26983186)
544
+ 520 train 7.527274 (lr=7.8182e-05) (hash(x)=27208507)
545
+ 521 train 7.712989 (lr=7.8012e-05) (hash(x)=23862845)
546
+ 522 train 7.528232 (lr=7.7842e-05) (hash(x)=22664494)
547
+ 523 train 7.505093 (lr=7.7671e-05) (hash(x)=24092927)
548
+ 524 train 7.492690 (lr=7.7500e-05) (hash(x)=27213812)
549
+ 525 train 7.467358 (lr=7.7328e-05) (hash(x)=23929955)
550
+ 526 train 7.472635 (lr=7.7156e-05) (hash(x)=23025790)
551
+ 527 train 7.418207 (lr=7.6984e-05) (hash(x)=20452272)
552
+ 528 train 7.614613 (lr=7.6811e-05) (hash(x)=26473464)
553
+ 529 train 7.583344 (lr=7.6637e-05) (hash(x)=28276603)
554
+ 530 train 7.414617 (lr=7.6463e-05) (hash(x)=21480367)
555
+ 531 train 7.611022 (lr=7.6289e-05) (hash(x)=24658642)
556
+ 532 train 7.503337 (lr=7.6115e-05) (hash(x)=26854801)
557
+ 533 train 7.397410 (lr=7.5940e-05) (hash(x)=17862226)
558
+ 534 train 7.505068 (lr=7.5764e-05) (hash(x)=28519182)
559
+ 535 train 7.485750 (lr=7.5588e-05) (hash(x)=25193694)
560
+ 536 train 7.473051 (lr=7.5412e-05) (hash(x)=25642620)
561
+ 537 train 7.547242 (lr=7.5235e-05) (hash(x)=25765909)
562
+ 538 train 7.373703 (lr=7.5058e-05) (hash(x)=20750531)
563
+ 539 train 7.479123 (lr=7.4881e-05) (hash(x)=24917098)
564
+ 540 train 7.434493 (lr=7.4703e-05) (hash(x)=24499766)
565
+ 541 train 7.482337 (lr=7.4525e-05) (hash(x)=25067358)
566
+ 542 train 7.601202 (lr=7.4346e-05) (hash(x)=27479796)
567
+ 543 train 7.478824 (lr=7.4167e-05) (hash(x)=23136080)
568
+ 544 train 7.554965 (lr=7.3988e-05) (hash(x)=29810074)
569
+ 545 train 7.448411 (lr=7.3808e-05) (hash(x)=24765539)
570
+ 546 train 7.487386 (lr=7.3628e-05) (hash(x)=22216309)
571
+ 547 train 7.414191 (lr=7.3448e-05) (hash(x)=20613861)
572
+ 548 train 7.385593 (lr=7.3267e-05) (hash(x)=23462447)
573
+ 549 train 7.452051 (lr=7.3086e-05) (hash(x)=26739455)
574
+ 550 val loss 7.5604
575
+ 550 val perplexity 1920.5990
576
+ 550 train 7.514684 (lr=7.2904e-05) (hash(x)=23222996)
577
+ 551 train 7.536718 (lr=7.2723e-05) (hash(x)=26489285)
578
+ 552 train 7.348818 (lr=7.2540e-05) (hash(x)=25185942)
579
+ 553 train 7.680166 (lr=7.2358e-05) (hash(x)=27476158)
580
+ 554 train 7.436522 (lr=7.2175e-05) (hash(x)=23653175)
581
+ 555 train 7.416105 (lr=7.1992e-05) (hash(x)=22405076)
582
+ 556 train 7.441228 (lr=7.1808e-05) (hash(x)=25673545)
583
+ 557 train 7.533946 (lr=7.1624e-05) (hash(x)=22675428)
584
+ 558 train 7.641112 (lr=7.1440e-05) (hash(x)=28168456)
585
+ 559 train 7.756699 (lr=7.1256e-05) (hash(x)=26072586)
586
+ 560 train 7.549207 (lr=7.1071e-05) (hash(x)=24974331)
587
+ 561 train 7.352679 (lr=7.0886e-05) (hash(x)=21951984)
588
+ 562 train 7.707144 (lr=7.0701e-05) (hash(x)=25403751)
589
+ 563 train 7.493234 (lr=7.0515e-05) (hash(x)=26396012)
590
+ 564 train 7.467531 (lr=7.0329e-05) (hash(x)=24827190)
591
+ 565 train 7.585389 (lr=7.0143e-05) (hash(x)=24847458)
592
+ 566 train 7.553688 (lr=6.9956e-05) (hash(x)=22159088)
593
+ 567 train 7.492432 (lr=6.9769e-05) (hash(x)=22140034)
594
+ 568 train 7.537660 (lr=6.9582e-05) (hash(x)=19044797)
595
+ 569 train 7.436035 (lr=6.9394e-05) (hash(x)=23932794)
596
+ 570 train 7.545462 (lr=6.9207e-05) (hash(x)=25256890)
597
+ 571 train 7.619053 (lr=6.9019e-05) (hash(x)=24161711)
598
+ 572 train 7.531468 (lr=6.8830e-05) (hash(x)=22036817)
599
+ 573 train 7.371822 (lr=6.8642e-05) (hash(x)=26663208)
600
+ 574 train 7.441002 (lr=6.8453e-05) (hash(x)=24072213)
601
+ 575 train 7.432795 (lr=6.8264e-05) (hash(x)=22766044)
602
+ 576 train 7.551762 (lr=6.8075e-05) (hash(x)=27871153)
603
+ 577 train 7.525133 (lr=6.7885e-05) (hash(x)=24037280)
604
+ 578 train 7.630750 (lr=6.7695e-05) (hash(x)=24121850)
605
+ 579 train 7.472374 (lr=6.7505e-05) (hash(x)=24380031)
606
+ 580 train 7.429338 (lr=6.7315e-05) (hash(x)=23055215)
607
+ 581 train 7.702837 (lr=6.7124e-05) (hash(x)=25329724)
608
+ 582 train 7.414389 (lr=6.6933e-05) (hash(x)=25662408)
609
+ 583 train 7.511388 (lr=6.6742e-05) (hash(x)=26346814)
610
+ 584 train 7.619717 (lr=6.6551e-05) (hash(x)=25283256)
611
+ 585 train 7.421248 (lr=6.6360e-05) (hash(x)=24021771)
612
+ 586 train 7.440869 (lr=6.6168e-05) (hash(x)=23576840)
613
+ 587 train 7.605403 (lr=6.5976e-05) (hash(x)=26414858)
614
+ 588 train 7.356205 (lr=6.5784e-05) (hash(x)=20099261)
615
+ 589 train 7.365086 (lr=6.5592e-05) (hash(x)=22453718)
616
+ 590 train 7.329267 (lr=6.5399e-05) (hash(x)=22529445)
617
+ 591 train 7.415786 (lr=6.5206e-05) (hash(x)=24484543)
618
+ 592 train 7.343019 (lr=6.5013e-05) (hash(x)=24043998)
619
+ 593 train 7.474518 (lr=6.4820e-05) (hash(x)=24301906)
620
+ 594 train 7.408827 (lr=6.4627e-05) (hash(x)=26328881)
621
+ 595 train 7.460378 (lr=6.4433e-05) (hash(x)=23101508)
622
+ 596 train 7.438038 (lr=6.4240e-05) (hash(x)=23527176)
623
+ 597 train 7.950079 (lr=6.4046e-05) (hash(x)=22705673)
624
+ 598 train 7.496606 (lr=6.3852e-05) (hash(x)=28664796)
625
+ 599 train 7.621476 (lr=6.3658e-05) (hash(x)=27102890)
626
+ 600 val loss 7.5413
627
+ 600 val perplexity 1884.3082
628
+ 600 train 7.697159 (lr=6.3463e-05) (hash(x)=26787259)
629
+ 601 train 7.742362 (lr=6.3269e-05) (hash(x)=23396088)
630
+ 602 train 7.458788 (lr=6.3074e-05) (hash(x)=24911831)
631
+ 603 train 7.484278 (lr=6.2879e-05) (hash(x)=23453491)
632
+ 604 train 7.530580 (lr=6.2684e-05) (hash(x)=23453396)
633
+ 605 train 7.621758 (lr=6.2489e-05) (hash(x)=23869612)
634
+ 606 train 7.710625 (lr=6.2294e-05) (hash(x)=26581590)
635
+ 607 train 7.530530 (lr=6.2098e-05) (hash(x)=28056342)
636
+ 608 train 7.592166 (lr=6.1903e-05) (hash(x)=25900866)
637
+ 609 train 7.621538 (lr=6.1707e-05) (hash(x)=24549177)
638
+ 610 train 7.532023 (lr=6.1511e-05) (hash(x)=28239583)
639
+ 611 train 7.558032 (lr=6.1315e-05) (hash(x)=23223554)
640
+ 612 train 7.627146 (lr=6.1119e-05) (hash(x)=24793480)
641
+ 613 train 7.514314 (lr=6.0923e-05) (hash(x)=25415570)
642
+ 614 train 7.585463 (lr=6.0726e-05) (hash(x)=24425760)
643
+ 615 train 7.648799 (lr=6.0530e-05) (hash(x)=23707134)
644
+ 616 train 7.598625 (lr=6.0333e-05) (hash(x)=28321697)
645
+ 617 train 7.526546 (lr=6.0137e-05) (hash(x)=23434031)
646
+ 618 train 7.507131 (lr=5.9940e-05) (hash(x)=27003481)
647
+ 619 train 7.496655 (lr=5.9743e-05) (hash(x)=22236893)
648
+ 620 train 7.521164 (lr=5.9546e-05) (hash(x)=26857458)
649
+ 621 train 7.518344 (lr=5.9349e-05) (hash(x)=26413122)
650
+ 622 train 7.823038 (lr=5.9152e-05) (hash(x)=27696537)
651
+ 623 train 7.550300 (lr=5.8955e-05) (hash(x)=21950234)
652
+ 624 train 7.573440 (lr=5.8758e-05) (hash(x)=24051952)
653
+ 625 train 7.582304 (lr=5.8560e-05) (hash(x)=24029939)
654
+ 626 train 7.670248 (lr=5.8363e-05) (hash(x)=24393346)
655
+ 627 train 7.632165 (lr=5.8165e-05) (hash(x)=26985352)
656
+ 628 train 7.384880 (lr=5.7968e-05) (hash(x)=19493265)
657
+ 629 train 7.458791 (lr=5.7770e-05) (hash(x)=23944010)
658
+ 630 train 7.673358 (lr=5.7573e-05) (hash(x)=21517435)
659
+ 631 train 7.438535 (lr=5.7375e-05) (hash(x)=25025868)
660
+ 632 train 7.457140 (lr=5.7177e-05) (hash(x)=22940997)
661
+ 633 train 7.595205 (lr=5.6979e-05) (hash(x)=25864840)
662
+ 634 train 7.494565 (lr=5.6782e-05) (hash(x)=24682151)
663
+ 635 train 7.417654 (lr=5.6584e-05) (hash(x)=27055330)
664
+ 636 train 7.468966 (lr=5.6386e-05) (hash(x)=22705710)
665
+ 637 train 7.574765 (lr=5.6188e-05) (hash(x)=26775238)
666
+ 638 train 7.712748 (lr=5.5990e-05) (hash(x)=21734331)
667
+ 639 train 7.437728 (lr=5.5792e-05) (hash(x)=25693442)
668
+ 640 train 7.551642 (lr=5.5594e-05) (hash(x)=25290454)
669
+ 641 train 7.511855 (lr=5.5396e-05) (hash(x)=25699860)
670
+ 642 train 7.678161 (lr=5.5198e-05) (hash(x)=19919115)
671
+ 643 train 7.760191 (lr=5.5000e-05) (hash(x)=25586967)
672
+ 644 train 7.481709 (lr=5.4802e-05) (hash(x)=25373506)
673
+ 645 train 7.493378 (lr=5.4604e-05) (hash(x)=26496148)
674
+ 646 train 7.566517 (lr=5.4406e-05) (hash(x)=27395083)
675
+ 647 train 7.766306 (lr=5.4208e-05) (hash(x)=27389709)
676
+ 648 train 7.630821 (lr=5.4010e-05) (hash(x)=29172968)
677
+ 649 train 7.560219 (lr=5.3812e-05) (hash(x)=25464920)
678
+ 650 val loss 7.5098
679
+ 650 val perplexity 1825.8177
680
+ 650 train 7.589197 (lr=5.3614e-05) (hash(x)=25992320)
681
+ 651 train 7.574807 (lr=5.3416e-05) (hash(x)=24949509)
682
+ 652 train 7.610744 (lr=5.3218e-05) (hash(x)=25615151)
683
+ 653 train 7.565460 (lr=5.3021e-05) (hash(x)=27923548)
684
+ 654 train 7.601680 (lr=5.2823e-05) (hash(x)=26387848)
685
+ 655 train 7.543574 (lr=5.2625e-05) (hash(x)=22863518)
686
+ 656 train 7.487079 (lr=5.2427e-05) (hash(x)=20869581)
687
+ 657 train 7.649981 (lr=5.2230e-05) (hash(x)=18133606)
688
+ 658 train 7.532894 (lr=5.2032e-05) (hash(x)=28736122)
689
+ 659 train 7.510650 (lr=5.1835e-05) (hash(x)=25528329)
690
+ 660 train 7.575254 (lr=5.1637e-05) (hash(x)=24020765)
691
+ 661 train 7.608615 (lr=5.1440e-05) (hash(x)=27792245)
692
+ 662 train 7.624794 (lr=5.1242e-05) (hash(x)=28287635)
693
+ 663 train 7.358210 (lr=5.1045e-05) (hash(x)=20951206)
694
+ 664 train 7.445130 (lr=5.0848e-05) (hash(x)=22356452)
695
+ 665 train 7.516540 (lr=5.0651e-05) (hash(x)=24385170)
696
+ 666 train 7.592459 (lr=5.0454e-05) (hash(x)=23455854)
697
+ 667 train 7.571509 (lr=5.0257e-05) (hash(x)=21589405)
698
+ 668 train 7.553639 (lr=5.0060e-05) (hash(x)=24698150)
699
+ 669 train 7.531401 (lr=4.9863e-05) (hash(x)=28010836)
700
+ 670 train 7.657823 (lr=4.9667e-05) (hash(x)=25427381)
701
+ 671 train 7.617446 (lr=4.9470e-05) (hash(x)=27478557)
702
+ 672 train 7.556267 (lr=4.9274e-05) (hash(x)=23865569)
703
+ 673 train 7.630557 (lr=4.9077e-05) (hash(x)=23407132)
704
+ 674 train 7.486390 (lr=4.8881e-05) (hash(x)=20253664)
705
+ 675 train 7.456762 (lr=4.8685e-05) (hash(x)=21087928)
706
+ 676 train 7.571109 (lr=4.8489e-05) (hash(x)=25957589)
707
+ 677 train 7.497773 (lr=4.8293e-05) (hash(x)=27008610)
708
+ 678 train 7.566462 (lr=4.8097e-05) (hash(x)=27580265)
709
+ 679 train 7.608348 (lr=4.7902e-05) (hash(x)=24244798)
710
+ 680 train 7.478534 (lr=4.7706e-05) (hash(x)=24535871)
711
+ 681 train 7.584311 (lr=4.7511e-05) (hash(x)=27558681)
712
+ 682 train 7.467221 (lr=4.7316e-05) (hash(x)=21152793)
713
+ 683 train 7.441523 (lr=4.7121e-05) (hash(x)=28147005)
714
+ 684 train 7.536576 (lr=4.6926e-05) (hash(x)=21933180)
715
+ 685 train 7.504664 (lr=4.6731e-05) (hash(x)=24735895)
716
+ 686 train 7.555192 (lr=4.6537e-05) (hash(x)=26209948)
717
+ 687 train 7.538921 (lr=4.6342e-05) (hash(x)=23358313)
718
+ 688 train 7.475364 (lr=4.6148e-05) (hash(x)=22398001)
719
+ 689 train 7.383994 (lr=4.5954e-05) (hash(x)=23724957)
720
+ 690 train 7.434212 (lr=4.5760e-05) (hash(x)=23083514)
721
+ 691 train 7.312852 (lr=4.5567e-05) (hash(x)=16016924)
722
+ 692 train 7.591164 (lr=4.5373e-05) (hash(x)=27173674)
723
+ 693 train 7.392171 (lr=4.5180e-05) (hash(x)=28056250)
724
+ 694 train 7.498176 (lr=4.4987e-05) (hash(x)=22233617)
725
+ 695 train 7.591100 (lr=4.4794e-05) (hash(x)=23777908)
726
+ 696 train 7.566017 (lr=4.4601e-05) (hash(x)=27842849)
727
+ 697 train 7.549255 (lr=4.4408e-05) (hash(x)=23711679)
728
+ 698 train 7.398949 (lr=4.4216e-05) (hash(x)=17204843)
729
+ 699 train 7.702005 (lr=4.4024e-05) (hash(x)=23909800)
730
+ 700 val loss 7.4818
731
+ 700 val perplexity 1775.3721
732
+ 700 train 7.556087 (lr=4.3832e-05) (hash(x)=25033286)
733
+ 701 train 7.843888 (lr=4.3640e-05) (hash(x)=23608371)
734
+ 702 train 7.637945 (lr=4.3449e-05) (hash(x)=23654625)
735
+ 703 train 7.452461 (lr=4.3258e-05) (hash(x)=22633313)
736
+ 704 train 7.515621 (lr=4.3067e-05) (hash(x)=24267011)
737
+ 705 train 7.466049 (lr=4.2876e-05) (hash(x)=23449416)
738
+ 706 train 7.712853 (lr=4.2685e-05) (hash(x)=28527906)
739
+ 707 train 7.405141 (lr=4.2495e-05) (hash(x)=21189075)
740
+ 708 train 7.472712 (lr=4.2305e-05) (hash(x)=25278645)
741
+ 709 train 7.998730 (lr=4.2115e-05) (hash(x)=23654290)
742
+ 710 train 7.454517 (lr=4.1925e-05) (hash(x)=24571139)
743
+ 711 train 7.553723 (lr=4.1736e-05) (hash(x)=26203172)
744
+ 712 train 7.482317 (lr=4.1547e-05) (hash(x)=19623503)
745
+ 713 train 7.469617 (lr=4.1358e-05) (hash(x)=23970545)
746
+ 714 train 7.649938 (lr=4.1170e-05) (hash(x)=25724900)
747
+ 715 train 7.563676 (lr=4.0981e-05) (hash(x)=23245421)
748
+ 716 train 7.509876 (lr=4.0793e-05) (hash(x)=24087537)
749
+ 717 train 7.436679 (lr=4.0606e-05) (hash(x)=25549646)
750
+ 718 train 7.509693 (lr=4.0418e-05) (hash(x)=26406563)
751
+ 719 train 7.550564 (lr=4.0231e-05) (hash(x)=23972429)
752
+ 720 train 7.605820 (lr=4.0044e-05) (hash(x)=26378040)
753
+ 721 train 7.458829 (lr=3.9857e-05) (hash(x)=24618531)
754
+ 722 train 7.473231 (lr=3.9671e-05) (hash(x)=20135714)
755
+ 723 train 7.614961 (lr=3.9485e-05) (hash(x)=26490981)
756
+ 724 train 7.627093 (lr=3.9299e-05) (hash(x)=23927567)
757
+ 725 train 7.468428 (lr=3.9114e-05) (hash(x)=27943609)
758
+ 726 train 7.522170 (lr=3.8929e-05) (hash(x)=22962561)
759
+ 727 train 7.559854 (lr=3.8744e-05) (hash(x)=24772257)
760
+ 728 train 7.483731 (lr=3.8560e-05) (hash(x)=25539761)
761
+ 729 train 7.647997 (lr=3.8376e-05) (hash(x)=23316262)
762
+ 730 train 7.415804 (lr=3.8192e-05) (hash(x)=23857558)
763
+ 731 train 7.597335 (lr=3.8008e-05) (hash(x)=30252690)
764
+ 732 train 7.665937 (lr=3.7825e-05) (hash(x)=30590162)
765
+ 733 train 7.552160 (lr=3.7642e-05) (hash(x)=25015721)
766
+ 734 train 7.633540 (lr=3.7460e-05) (hash(x)=28639679)
767
+ 735 train 7.505615 (lr=3.7277e-05) (hash(x)=26199058)
768
+ 736 train 7.564039 (lr=3.7096e-05) (hash(x)=28907461)
769
+ 737 train 7.546357 (lr=3.6914e-05) (hash(x)=26466915)
770
+ 738 train 7.621041 (lr=3.6733e-05) (hash(x)=20332613)
771
+ 739 train 7.506954 (lr=3.6552e-05) (hash(x)=26766470)
772
+ 740 train 7.450099 (lr=3.6372e-05) (hash(x)=26722145)
773
+ 741 train 7.500107 (lr=3.6192e-05) (hash(x)=21222489)
774
+ 742 train 7.634332 (lr=3.6012e-05) (hash(x)=24662435)
775
+ 743 train 7.442200 (lr=3.5833e-05) (hash(x)=24308927)
776
+ 744 train 7.402232 (lr=3.5654e-05) (hash(x)=28442502)
777
+ 745 train 7.444362 (lr=3.5475e-05) (hash(x)=22810435)
778
+ 746 train 7.578400 (lr=3.5297e-05) (hash(x)=30486518)
779
+ 747 train 7.559855 (lr=3.5119e-05) (hash(x)=22688712)
780
+ 748 train 7.473432 (lr=3.4942e-05) (hash(x)=20712962)
781
+ 749 train 7.337163 (lr=3.4765e-05) (hash(x)=27490086)
782
+ 750 val loss 7.4596
783
+ 750 val perplexity 1736.4541
784
+ 750 train 7.334404 (lr=3.4588e-05) (hash(x)=23973908)
785
+ 751 train 7.488537 (lr=3.4412e-05) (hash(x)=29132401)
786
+ 752 train 7.490381 (lr=3.4236e-05) (hash(x)=27386700)
787
+ 753 train 7.518881 (lr=3.4060e-05) (hash(x)=24268614)
788
+ 754 train 7.570349 (lr=3.3885e-05) (hash(x)=27378387)
789
+ 755 train 7.373429 (lr=3.3711e-05) (hash(x)=22529400)
790
+ 756 train 7.452330 (lr=3.3537e-05) (hash(x)=27586989)
791
+ 757 train 7.381913 (lr=3.3363e-05) (hash(x)=22336349)
792
+ 758 train 7.420479 (lr=3.3189e-05) (hash(x)=26147854)
793
+ 759 train 7.513012 (lr=3.3016e-05) (hash(x)=25346788)
794
+ 760 train 7.601246 (lr=3.2844e-05) (hash(x)=25956855)
795
+ 761 train 7.531219 (lr=3.2672e-05) (hash(x)=24367780)
796
+ 762 train 7.424407 (lr=3.2500e-05) (hash(x)=26084462)
797
+ 763 train 7.501590 (lr=3.2329e-05) (hash(x)=26702459)
798
+ 764 train 7.437404 (lr=3.2158e-05) (hash(x)=25717976)
799
+ 765 train 7.375256 (lr=3.1988e-05) (hash(x)=22141366)
800
+ 766 train 7.605688 (lr=3.1818e-05) (hash(x)=24652423)
801
+ 767 train 7.527318 (lr=3.1648e-05) (hash(x)=30911699)
802
+ 768 train 7.515924 (lr=3.1479e-05) (hash(x)=22076954)
803
+ 769 train 7.697556 (lr=3.1311e-05) (hash(x)=21586769)
804
+ 770 train 7.514791 (lr=3.1142e-05) (hash(x)=21994337)
805
+ 771 train 7.473822 (lr=3.0975e-05) (hash(x)=25454145)
806
+ 772 train 7.521131 (lr=3.0808e-05) (hash(x)=23239550)
807
+ 773 train 7.570200 (lr=3.0641e-05) (hash(x)=23565926)
808
+ 774 train 7.426509 (lr=3.0475e-05) (hash(x)=24705563)
809
+ 775 train 7.541328 (lr=3.0309e-05) (hash(x)=27671095)
810
+ 776 train 7.654326 (lr=3.0144e-05) (hash(x)=26254303)
811
+ 777 train 7.393283 (lr=2.9979e-05) (hash(x)=27454020)
812
+ 778 train 7.391985 (lr=2.9814e-05) (hash(x)=24394565)
813
+ 779 train 7.349934 (lr=2.9651e-05) (hash(x)=21780199)
814
+ 780 train 7.433042 (lr=2.9487e-05) (hash(x)=28296896)
815
+ 781 train 7.569510 (lr=2.9324e-05) (hash(x)=26766303)
816
+ 782 train 7.406379 (lr=2.9162e-05) (hash(x)=21599973)
817
+ 783 train 7.731707 (lr=2.9000e-05) (hash(x)=24969360)
818
+ 784 train 7.696099 (lr=2.8839e-05) (hash(x)=24382006)
819
+ 785 train 7.370741 (lr=2.8678e-05) (hash(x)=22009299)
820
+ 786 train 7.222163 (lr=2.8518e-05) (hash(x)=22376569)
821
+ 787 train 7.393046 (lr=2.8358e-05) (hash(x)=28166843)
822
+ 788 train 7.669737 (lr=2.8199e-05) (hash(x)=28230119)
823
+ 789 train 7.611578 (lr=2.8040e-05) (hash(x)=26207465)
824
+ 790 train 7.688897 (lr=2.7881e-05) (hash(x)=24748118)
825
+ 791 train 7.610233 (lr=2.7724e-05) (hash(x)=27033802)
826
+ 792 train 7.528696 (lr=2.7566e-05) (hash(x)=28788047)
827
+ 793 train 7.397250 (lr=2.7410e-05) (hash(x)=22185099)
828
+ 794 train 7.595845 (lr=2.7254e-05) (hash(x)=21585136)
829
+ 795 train 7.406788 (lr=2.7098e-05) (hash(x)=27691968)
830
+ 796 train 7.453660 (lr=2.6943e-05) (hash(x)=23706103)
831
+ 797 train 7.289941 (lr=2.6788e-05) (hash(x)=27918030)
832
+ 798 train 7.509144 (lr=2.6634e-05) (hash(x)=23127886)
833
+ 799 train 7.527495 (lr=2.6481e-05) (hash(x)=26197070)
834
+ 800 val loss 7.4417
835
+ 800 val perplexity 1705.7286
836
+ 800 train 7.574115 (lr=2.6328e-05) (hash(x)=26463005)
837
+ 801 train 7.469114 (lr=2.6176e-05) (hash(x)=19160134)
838
+ 802 train 7.417971 (lr=2.6024e-05) (hash(x)=23699136)
839
+ 803 train 7.488408 (lr=2.5873e-05) (hash(x)=25353051)
840
+ 804 train 7.415915 (lr=2.5722e-05) (hash(x)=21558484)
841
+ 805 train 7.333173 (lr=2.5572e-05) (hash(x)=22828380)
842
+ 806 train 7.464387 (lr=2.5423e-05) (hash(x)=24623811)
843
+ 807 train 7.636195 (lr=2.5274e-05) (hash(x)=25410299)
844
+ 808 train 7.492069 (lr=2.5125e-05) (hash(x)=26046091)
845
+ 809 train 7.611971 (lr=2.4978e-05) (hash(x)=29887482)
846
+ 810 train 7.349257 (lr=2.4830e-05) (hash(x)=21809706)
847
+ 811 train 7.472101 (lr=2.4684e-05) (hash(x)=23712872)
848
+ 812 train 7.361542 (lr=2.4538e-05) (hash(x)=25875103)
849
+ 813 train 7.333164 (lr=2.4392e-05) (hash(x)=23736441)
850
+ 814 train 7.424818 (lr=2.4247e-05) (hash(x)=25093883)
851
+ 815 train 7.363343 (lr=2.4103e-05) (hash(x)=23692325)
852
+ 816 train 7.262480 (lr=2.3959e-05) (hash(x)=23061082)
853
+ 817 train 7.393563 (lr=2.3816e-05) (hash(x)=26094253)
854
+ 818 train 7.550554 (lr=2.3674e-05) (hash(x)=27327886)
855
+ 819 train 7.422655 (lr=2.3532e-05) (hash(x)=26186143)
856
+ 820 train 7.419140 (lr=2.3391e-05) (hash(x)=25336043)
857
+ 821 train 7.460618 (lr=2.3250e-05) (hash(x)=22811468)
858
+ 822 train 7.441939 (lr=2.3110e-05) (hash(x)=27925831)
859
+ 823 train 7.546475 (lr=2.2971e-05) (hash(x)=35528108)
860
+ 824 train 7.537710 (lr=2.2832e-05) (hash(x)=24949421)
861
+ 825 train 7.297977 (lr=2.2694e-05) (hash(x)=22845735)
862
+ 826 train 7.411054 (lr=2.2556e-05) (hash(x)=28138281)
863
+ 827 train 7.482001 (lr=2.2420e-05) (hash(x)=29038360)
864
+ 828 train 7.425698 (lr=2.2283e-05) (hash(x)=22827903)
865
+ 829 train 7.514131 (lr=2.2148e-05) (hash(x)=24647015)
866
+ 830 train 7.399959 (lr=2.2013e-05) (hash(x)=20477804)
867
+ 831 train 7.367899 (lr=2.1878e-05) (hash(x)=26692640)
868
+ 832 train 7.446416 (lr=2.1745e-05) (hash(x)=27388794)
869
+ 833 train 7.734144 (lr=2.1612e-05) (hash(x)=25265875)
870
+ 834 train 7.370404 (lr=2.1479e-05) (hash(x)=21841888)
871
+ 835 train 7.475332 (lr=2.1347e-05) (hash(x)=19809196)
872
+ 836 train 7.534215 (lr=2.1216e-05) (hash(x)=23007945)
873
+ 837 train 7.462128 (lr=2.1086e-05) (hash(x)=24079409)
874
+ 838 train 7.383613 (lr=2.0956e-05) (hash(x)=25874533)
875
+ 839 train 7.305530 (lr=2.0827e-05) (hash(x)=23660106)
876
+ 840 train 7.794976 (lr=2.0698e-05) (hash(x)=41821329)
877
+ 841 train 7.407612 (lr=2.0570e-05) (hash(x)=25763899)
878
+ 842 train 7.245990 (lr=2.0443e-05) (hash(x)=28584111)
879
+ 843 train 7.734368 (lr=2.0317e-05) (hash(x)=25559887)
880
+ 844 train 7.494929 (lr=2.0191e-05) (hash(x)=26886377)
881
+ 845 train 7.471811 (lr=2.0066e-05) (hash(x)=26461895)
882
+ 846 train 7.517651 (lr=1.9941e-05) (hash(x)=23018827)
883
+ 847 train 7.321105 (lr=1.9818e-05) (hash(x)=24375035)
884
+ 848 train 7.491956 (lr=1.9694e-05) (hash(x)=26431670)
885
+ 849 train 7.375950 (lr=1.9572e-05) (hash(x)=25033511)
886
+ 850 val loss 7.4396
887
+ 850 val perplexity 1702.0132
888
+ 850 train 7.488484 (lr=1.9450e-05) (hash(x)=25602577)
889
+ 851 train 7.400078 (lr=1.9329e-05) (hash(x)=23316582)
890
+ 852 train 7.535356 (lr=1.9209e-05) (hash(x)=22470095)
891
+ 853 train 7.406911 (lr=1.9089e-05) (hash(x)=25089202)
892
+ 854 train 7.337296 (lr=1.8970e-05) (hash(x)=26243247)
893
+ 855 train 7.435785 (lr=1.8852e-05) (hash(x)=22231144)
894
+ 856 train 7.346595 (lr=1.8734e-05) (hash(x)=27646564)
895
+ 857 train 7.491600 (lr=1.8618e-05) (hash(x)=23889875)
896
+ 858 train 7.445905 (lr=1.8501e-05) (hash(x)=25255240)
897
+ 859 train 7.527608 (lr=1.8386e-05) (hash(x)=23883299)
898
+ 860 train 7.501969 (lr=1.8271e-05) (hash(x)=18463841)
899
+ 861 train 7.533608 (lr=1.8157e-05) (hash(x)=30657762)
900
+ 862 train 7.416885 (lr=1.8044e-05) (hash(x)=25288079)
901
+ 863 train 7.461645 (lr=1.7931e-05) (hash(x)=28979310)
902
+ 864 train 7.647059 (lr=1.7819e-05) (hash(x)=30254991)
903
+ 865 train 7.492545 (lr=1.7708e-05) (hash(x)=27274653)
904
+ 866 train 7.472941 (lr=1.7598e-05) (hash(x)=23026066)
905
+ 867 train 7.428191 (lr=1.7488e-05) (hash(x)=14001963)
906
+ 868 train 7.326510 (lr=1.7379e-05) (hash(x)=21988054)
907
+ 869 train 7.420466 (lr=1.7271e-05) (hash(x)=26241099)
908
+ 870 train 7.513637 (lr=1.7163e-05) (hash(x)=26209722)
909
+ 871 train 7.371310 (lr=1.7056e-05) (hash(x)=26928312)
910
+ 872 train 7.457225 (lr=1.6950e-05) (hash(x)=23563272)
911
+ 873 train 7.384061 (lr=1.6845e-05) (hash(x)=25704828)
912
+ 874 train 7.650157 (lr=1.6740e-05) (hash(x)=22498992)
913
+ 875 train 7.784369 (lr=1.6636e-05) (hash(x)=22895918)
914
+ 876 train 7.341893 (lr=1.6533e-05) (hash(x)=24404283)
915
+ 877 train 7.376393 (lr=1.6431e-05) (hash(x)=26663932)
916
+ 878 train 7.514564 (lr=1.6329e-05) (hash(x)=22805316)
917
+ 879 train 7.509393 (lr=1.6228e-05) (hash(x)=23344466)
918
+ 880 train 7.453814 (lr=1.6128e-05) (hash(x)=20966940)
919
+ 881 train 7.451263 (lr=1.6029e-05) (hash(x)=27422664)
920
+ 882 train 7.477421 (lr=1.5930e-05) (hash(x)=22116510)
921
+ 883 train 7.340514 (lr=1.5832e-05) (hash(x)=27269633)
922
+ 884 train 7.450755 (lr=1.5735e-05) (hash(x)=27157711)
923
+ 885 train 7.491467 (lr=1.5639e-05) (hash(x)=22774645)
924
+ 886 train 7.409890 (lr=1.5543e-05) (hash(x)=22143146)
925
+ 887 train 7.402446 (lr=1.5449e-05) (hash(x)=22356438)
926
+ 888 train 7.170747 (lr=1.5354e-05) (hash(x)=23172158)
927
+ 889 train 7.429227 (lr=1.5261e-05) (hash(x)=26971690)
928
+ 890 train 7.466310 (lr=1.5169e-05) (hash(x)=24851922)
929
+ 891 train 7.281016 (lr=1.5077e-05) (hash(x)=21017641)
930
+ 892 train 7.436307 (lr=1.4986e-05) (hash(x)=28186957)
931
+ 893 train 7.455210 (lr=1.4896e-05) (hash(x)=25661082)
932
+ 894 train 7.351072 (lr=1.4806e-05) (hash(x)=25462781)
933
+ 895 train 7.305271 (lr=1.4718e-05) (hash(x)=21774374)
934
+ 896 train 7.305682 (lr=1.4630e-05) (hash(x)=23657055)
935
+ 897 train 7.333868 (lr=1.4543e-05) (hash(x)=23604050)
936
+ 898 train 7.771733 (lr=1.4456e-05) (hash(x)=31534071)
937
+ 899 train 7.578480 (lr=1.4371e-05) (hash(x)=24183732)
938
+ 900 val loss 7.4296
939
+ 900 val perplexity 1685.2150
940
+ 900 train 7.407505 (lr=1.4286e-05) (hash(x)=23384280)
941
+ 901 train 7.293602 (lr=1.4202e-05) (hash(x)=21975928)
942
+ 902 train 7.294423 (lr=1.4119e-05) (hash(x)=25281674)
943
+ 903 train 7.279311 (lr=1.4037e-05) (hash(x)=22947712)
944
+ 904 train 7.354813 (lr=1.3955e-05) (hash(x)=24853977)
945
+ 905 train 7.403245 (lr=1.3874e-05) (hash(x)=23696470)
946
+ 906 train 7.413362 (lr=1.3794e-05) (hash(x)=27489711)
947
+ 907 train 7.342287 (lr=1.3715e-05) (hash(x)=28050163)
948
+ 908 train 7.214733 (lr=1.3637e-05) (hash(x)=26996807)
949
+ 909 train 7.621967 (lr=1.3559e-05) (hash(x)=28626495)
950
+ 910 train 7.336179 (lr=1.3482e-05) (hash(x)=24783242)
951
+ 911 train 7.337069 (lr=1.3407e-05) (hash(x)=25832479)
952
+ 912 train 7.224257 (lr=1.3331e-05) (hash(x)=23522137)
953
+ 913 train 7.514649 (lr=1.3257e-05) (hash(x)=26731973)
954
+ 914 train 7.595162 (lr=1.3183e-05) (hash(x)=26019142)
955
+ 915 train 7.370749 (lr=1.3111e-05) (hash(x)=25081877)
956
+ 916 train 7.487756 (lr=1.3039e-05) (hash(x)=23843571)
957
+ 917 train 7.493926 (lr=1.2968e-05) (hash(x)=26805739)
958
+ 918 train 7.396144 (lr=1.2897e-05) (hash(x)=27169927)
959
+ 919 train 7.439053 (lr=1.2828e-05) (hash(x)=26545491)
960
+ 920 train 7.450454 (lr=1.2759e-05) (hash(x)=25179497)
961
+ 921 train 7.389654 (lr=1.2691e-05) (hash(x)=27085003)
962
+ 922 train 7.410111 (lr=1.2624e-05) (hash(x)=24558202)
963
+ 923 train 7.564204 (lr=1.2558e-05) (hash(x)=25317684)
964
+ 924 train 7.422526 (lr=1.2493e-05) (hash(x)=27104868)
965
+ 925 train 7.461666 (lr=1.2428e-05) (hash(x)=24806274)
966
+ 926 train 7.518903 (lr=1.2364e-05) (hash(x)=24341937)
967
+ 927 train 7.316999 (lr=1.2301e-05) (hash(x)=22736558)
968
+ 928 train 7.333353 (lr=1.2239e-05) (hash(x)=23843744)
969
+ 929 train 7.460651 (lr=1.2178e-05) (hash(x)=27218468)
970
+ 930 train 7.399772 (lr=1.2118e-05) (hash(x)=26042053)
971
+ 931 train 7.419421 (lr=1.2058e-05) (hash(x)=25776082)
972
+ 932 train 7.299749 (lr=1.1999e-05) (hash(x)=21594447)
973
+ 933 train 7.338398 (lr=1.1941e-05) (hash(x)=21674469)
974
+ 934 train 7.281021 (lr=1.1884e-05) (hash(x)=26213966)
975
+ 935 train 7.290705 (lr=1.1828e-05) (hash(x)=23770702)
976
+ 936 train 7.407078 (lr=1.1772e-05) (hash(x)=27829878)
977
+ 937 train 7.501351 (lr=1.1718e-05) (hash(x)=29177278)
978
+ 938 train 7.426019 (lr=1.1664e-05) (hash(x)=26020089)
979
+ 939 train 7.322884 (lr=1.1611e-05) (hash(x)=24676475)
980
+ 940 train 7.393018 (lr=1.1559e-05) (hash(x)=24056174)
981
+ 941 train 7.414608 (lr=1.1508e-05) (hash(x)=21887514)
982
+ 942 train 7.415479 (lr=1.1457e-05) (hash(x)=24196100)
983
+ 943 train 7.372238 (lr=1.1408e-05) (hash(x)=25700093)
984
+ 944 train 7.545765 (lr=1.1359e-05) (hash(x)=24365834)
985
+ 945 train 7.332805 (lr=1.1311e-05) (hash(x)=25631786)
986
+ 946 train 7.464033 (lr=1.1264e-05) (hash(x)=24452345)
987
+ 947 train 7.171823 (lr=1.1218e-05) (hash(x)=21465330)
988
+ 948 train 7.298168 (lr=1.1173e-05) (hash(x)=20427493)
989
+ 949 train 7.335891 (lr=1.1128e-05) (hash(x)=25439720)
990
+ 950 val loss 7.4199
991
+ 950 val perplexity 1668.8562
992
+ 950 train 7.398656 (lr=1.1085e-05) (hash(x)=21652988)
993
+ 951 train 7.339907 (lr=1.1042e-05) (hash(x)=22546779)
994
+ 952 train 7.209517 (lr=1.1000e-05) (hash(x)=20855736)
995
+ 953 train 7.397276 (lr=1.0959e-05) (hash(x)=27725873)
996
+ 954 train 7.464233 (lr=1.0919e-05) (hash(x)=24582989)
997
+ 955 train 7.338234 (lr=1.0879e-05) (hash(x)=24822372)
998
+ 956 train 7.390271 (lr=1.0841e-05) (hash(x)=25624588)
999
+ 957 train 7.347393 (lr=1.0803e-05) (hash(x)=23566489)
1000
+ 958 train 7.524933 (lr=1.0766e-05) (hash(x)=28117387)
1001
+ 959 train 7.290871 (lr=1.0730e-05) (hash(x)=26874283)
1002
+ 960 train 7.297306 (lr=1.0695e-05) (hash(x)=28570123)
1003
+ 961 train 7.451760 (lr=1.0661e-05) (hash(x)=24807017)
1004
+ 962 train 7.323246 (lr=1.0628e-05) (hash(x)=25189993)
1005
+ 963 train 7.455842 (lr=1.0595e-05) (hash(x)=29202226)
1006
+ 964 train 7.645636 (lr=1.0563e-05) (hash(x)=26060713)
1007
+ 965 train 7.524036 (lr=1.0533e-05) (hash(x)=23053285)
1008
+ 966 train 7.500039 (lr=1.0503e-05) (hash(x)=21968481)
1009
+ 967 train 7.385129 (lr=1.0474e-05) (hash(x)=24492698)
1010
+ 968 train 7.363263 (lr=1.0445e-05) (hash(x)=25760806)
1011
+ 969 train 7.384084 (lr=1.0418e-05) (hash(x)=27417392)
1012
+ 970 train 7.385349 (lr=1.0391e-05) (hash(x)=26767105)
1013
+ 971 train 7.450591 (lr=1.0366e-05) (hash(x)=25117678)
1014
+ 972 train 7.330565 (lr=1.0341e-05) (hash(x)=24233758)
1015
+ 973 train 7.392282 (lr=1.0317e-05) (hash(x)=29907634)
1016
+ 974 train 7.174641 (lr=1.0294e-05) (hash(x)=23552436)
1017
+ 975 train 7.544744 (lr=1.0272e-05) (hash(x)=28610406)
1018
+ 976 train 7.364544 (lr=1.0251e-05) (hash(x)=23858212)
1019
+ 977 train 7.265794 (lr=1.0230e-05) (hash(x)=20567764)
1020
+ 978 train 7.275741 (lr=1.0211e-05) (hash(x)=29631106)
1021
+ 979 train 7.253947 (lr=1.0192e-05) (hash(x)=19843564)
1022
+ 980 train 7.392068 (lr=1.0174e-05) (hash(x)=22589891)
1023
+ 981 train 7.177075 (lr=1.0157e-05) (hash(x)=21183943)
1024
+ 982 train 7.337965 (lr=1.0141e-05) (hash(x)=26152241)
1025
+ 983 train 7.350290 (lr=1.0126e-05) (hash(x)=22700785)
1026
+ 984 train 7.585400 (lr=1.0111e-05) (hash(x)=29611280)
1027
+ 985 train 7.703375 (lr=1.0098e-05) (hash(x)=25972823)
1028
+ 986 train 7.394578 (lr=1.0085e-05) (hash(x)=23447782)
1029
+ 987 train 7.338687 (lr=1.0074e-05) (hash(x)=22578453)
1030
+ 988 train 7.334603 (lr=1.0063e-05) (hash(x)=25474622)
1031
+ 989 train 7.404408 (lr=1.0053e-05) (hash(x)=22965762)
1032
+ 990 train 7.445079 (lr=1.0044e-05) (hash(x)=26116321)
1033
+ 991 train 7.362096 (lr=1.0035e-05) (hash(x)=25934549)
1034
+ 992 train 7.276827 (lr=1.0028e-05) (hash(x)=23937495)
1035
+ 993 train 7.459218 (lr=1.0021e-05) (hash(x)=26286442)
1036
+ 994 train 7.714988 (lr=1.0016e-05) (hash(x)=32737711)
1037
+ 995 train 7.286125 (lr=1.0011e-05) (hash(x)=20563613)
1038
+ 996 train 7.547499 (lr=1.0007e-05) (hash(x)=26234770)
1039
+ 997 train 7.480315 (lr=1.0004e-05) (hash(x)=26179422)
1040
+ 998 train 7.395674 (lr=1.0002e-05) (hash(x)=23289651)
1041
+ 999 val loss 7.4181
1042
+ 999 val perplexity 1665.8024
1043
+ 999 train 7.777097 (lr=1.0000e-05) (hash(x)=36436617)
lr1e-4_total_batch_size40960_seq_len512/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f21197b92b6b0edde3f983ad03ec8d5251e84fa0f8338300e8154e9e656d20e6
3
+ size 48156226
lr1e-4_total_batch_size40960_seq_len512/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:605e9a1094a164709d40da707bd52d0f0b74e3af4eae96ec46388dfd6a69a34b
3
+ size 71157574