andrew-healey commited on
Commit
8d9beb3
·
verified ·
1 Parent(s): c6c05b8

Upload folder using huggingface_hub

Browse files
lr8e-4_total_batch_size40960_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr8e-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_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 512, "batch_size": null, "total_batch_size": 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": 0.0008, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr8e-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
lr8e-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=2.7972e-06) (hash(x)=20782690)
5
+ 1 train 11.770680 (lr=5.5944e-06) (hash(x)=23890908)
6
+ 2 train 11.759589 (lr=8.3916e-06) (hash(x)=24611628)
7
+ 3 train 11.726483 (lr=1.1189e-05) (hash(x)=25767553)
8
+ 4 train 11.707442 (lr=1.3986e-05) (hash(x)=22512166)
9
+ 5 train 11.680038 (lr=1.6783e-05) (hash(x)=24732833)
10
+ 6 train 11.644572 (lr=1.9580e-05) (hash(x)=25129410)
11
+ 7 train 11.608064 (lr=2.2378e-05) (hash(x)=26582391)
12
+ 8 train 11.565802 (lr=2.5175e-05) (hash(x)=26472336)
13
+ 9 train 11.513274 (lr=2.7972e-05) (hash(x)=27654289)
14
+ 10 train 11.453381 (lr=3.0769e-05) (hash(x)=26559876)
15
+ 11 train 11.393353 (lr=3.3566e-05) (hash(x)=21770329)
16
+ 12 train 11.340712 (lr=3.6364e-05) (hash(x)=22723459)
17
+ 13 train 11.253342 (lr=3.9161e-05) (hash(x)=30338342)
18
+ 14 train 11.189358 (lr=4.1958e-05) (hash(x)=25678059)
19
+ 15 train 11.071432 (lr=4.4755e-05) (hash(x)=26109335)
20
+ 16 train 11.018415 (lr=4.7552e-05) (hash(x)=25044885)
21
+ 17 train 10.959982 (lr=5.0350e-05) (hash(x)=28870690)
22
+ 18 train 10.914984 (lr=5.3147e-05) (hash(x)=35279941)
23
+ 19 train 10.791675 (lr=5.5944e-05) (hash(x)=25015923)
24
+ 20 train 10.695947 (lr=5.8741e-05) (hash(x)=26740855)
25
+ 21 train 10.612270 (lr=6.1538e-05) (hash(x)=22262151)
26
+ 22 train 10.546526 (lr=6.4336e-05) (hash(x)=26621834)
27
+ 23 train 10.515039 (lr=6.7133e-05) (hash(x)=23234741)
28
+ 24 train 10.560863 (lr=6.9930e-05) (hash(x)=25470563)
29
+ 25 train 10.505759 (lr=7.2727e-05) (hash(x)=21799102)
30
+ 26 train 10.432468 (lr=7.5524e-05) (hash(x)=23275479)
31
+ 27 train 10.484335 (lr=7.8322e-05) (hash(x)=22091862)
32
+ 28 train 10.364900 (lr=8.1119e-05) (hash(x)=17430373)
33
+ 29 train 10.304455 (lr=8.3916e-05) (hash(x)=22954861)
34
+ 30 train 10.395491 (lr=8.6713e-05) (hash(x)=26671799)
35
+ 31 train 10.345235 (lr=8.9510e-05) (hash(x)=26924723)
36
+ 32 train 10.364185 (lr=9.2308e-05) (hash(x)=23557495)
37
+ 33 train 10.206500 (lr=9.5105e-05) (hash(x)=27804380)
38
+ 34 train 10.362177 (lr=9.7902e-05) (hash(x)=26654908)
39
+ 35 train 10.294909 (lr=1.0070e-04) (hash(x)=23627518)
40
+ 36 train 10.262306 (lr=1.0350e-04) (hash(x)=21889990)
41
+ 37 train 10.291734 (lr=1.0629e-04) (hash(x)=22528001)
42
+ 38 train 10.287610 (lr=1.0909e-04) (hash(x)=29248942)
43
+ 39 train 10.210621 (lr=1.1189e-04) (hash(x)=25533417)
44
+ 40 train 10.149147 (lr=1.1469e-04) (hash(x)=20612533)
45
+ 41 train 10.144782 (lr=1.1748e-04) (hash(x)=24677740)
46
+ 42 train 10.150124 (lr=1.2028e-04) (hash(x)=25347203)
47
+ 43 train 10.096593 (lr=1.2308e-04) (hash(x)=28922363)
48
+ 44 train 10.066442 (lr=1.2587e-04) (hash(x)=25631397)
49
+ 45 train 10.040096 (lr=1.2867e-04) (hash(x)=26237983)
50
+ 46 train 10.042944 (lr=1.3147e-04) (hash(x)=24742645)
51
+ 47 train 10.044178 (lr=1.3427e-04) (hash(x)=26157660)
52
+ 48 train 9.902966 (lr=1.3706e-04) (hash(x)=22582169)
53
+ 49 train 9.906336 (lr=1.3986e-04) (hash(x)=26431960)
54
+ 50 val loss 9.9372
55
+ 50 val perplexity 20686.1289
56
+ 50 train 9.957220 (lr=1.4266e-04) (hash(x)=25716176)
57
+ 51 train 9.915504 (lr=1.4545e-04) (hash(x)=28019788)
58
+ 52 train 9.870030 (lr=1.4825e-04) (hash(x)=20382963)
59
+ 53 train 9.795561 (lr=1.5105e-04) (hash(x)=24489647)
60
+ 54 train 9.769114 (lr=1.5385e-04) (hash(x)=25128502)
61
+ 55 train 9.747393 (lr=1.5664e-04) (hash(x)=25231390)
62
+ 56 train 9.631413 (lr=1.5944e-04) (hash(x)=22708977)
63
+ 57 train 9.714092 (lr=1.6224e-04) (hash(x)=24364920)
64
+ 58 train 9.661685 (lr=1.6503e-04) (hash(x)=24932925)
65
+ 59 train 9.593024 (lr=1.6783e-04) (hash(x)=26913684)
66
+ 60 train 9.486526 (lr=1.7063e-04) (hash(x)=18017792)
67
+ 61 train 9.491199 (lr=1.7343e-04) (hash(x)=28613882)
68
+ 62 train 9.462656 (lr=1.7622e-04) (hash(x)=22652243)
69
+ 63 train 9.348833 (lr=1.7902e-04) (hash(x)=27194521)
70
+ 64 train 9.317839 (lr=1.8182e-04) (hash(x)=24936836)
71
+ 65 train 9.317128 (lr=1.8462e-04) (hash(x)=24765578)
72
+ 66 train 9.328730 (lr=1.8741e-04) (hash(x)=23107416)
73
+ 67 train 9.165711 (lr=1.9021e-04) (hash(x)=29314255)
74
+ 68 train 9.114615 (lr=1.9301e-04) (hash(x)=24013079)
75
+ 69 train 9.050832 (lr=1.9580e-04) (hash(x)=25322001)
76
+ 70 train 9.133453 (lr=1.9860e-04) (hash(x)=26690440)
77
+ 71 train 9.015752 (lr=2.0140e-04) (hash(x)=22598512)
78
+ 72 train 8.929531 (lr=2.0420e-04) (hash(x)=18473911)
79
+ 73 train 8.995079 (lr=2.0699e-04) (hash(x)=22322442)
80
+ 74 train 8.952206 (lr=2.0979e-04) (hash(x)=26324153)
81
+ 75 train 8.848536 (lr=2.1259e-04) (hash(x)=23446058)
82
+ 76 train 8.890267 (lr=2.1538e-04) (hash(x)=24843741)
83
+ 77 train 8.778149 (lr=2.1818e-04) (hash(x)=23197102)
84
+ 78 train 8.898396 (lr=2.2098e-04) (hash(x)=23054940)
85
+ 79 train 8.777189 (lr=2.2378e-04) (hash(x)=27140757)
86
+ 80 train 8.589031 (lr=2.2657e-04) (hash(x)=25149419)
87
+ 81 train 8.689938 (lr=2.2937e-04) (hash(x)=24727764)
88
+ 82 train 8.369203 (lr=2.3217e-04) (hash(x)=21935931)
89
+ 83 train 8.488997 (lr=2.3497e-04) (hash(x)=27606073)
90
+ 84 train 8.551130 (lr=2.3776e-04) (hash(x)=22450341)
91
+ 85 train 8.613591 (lr=2.4056e-04) (hash(x)=25945859)
92
+ 86 train 8.418242 (lr=2.4336e-04) (hash(x)=21613707)
93
+ 87 train 8.449243 (lr=2.4615e-04) (hash(x)=24396519)
94
+ 88 train 8.447019 (lr=2.4895e-04) (hash(x)=29138828)
95
+ 89 train 8.464678 (lr=2.5175e-04) (hash(x)=24685515)
96
+ 90 train 8.336527 (lr=2.5455e-04) (hash(x)=22825749)
97
+ 91 train 8.366172 (lr=2.5734e-04) (hash(x)=28767869)
98
+ 92 train 8.366078 (lr=2.6014e-04) (hash(x)=21159060)
99
+ 93 train 8.283638 (lr=2.6294e-04) (hash(x)=25045397)
100
+ 94 train 8.412509 (lr=2.6573e-04) (hash(x)=25560928)
101
+ 95 train 8.374347 (lr=2.6853e-04) (hash(x)=28959222)
102
+ 96 train 8.235563 (lr=2.7133e-04) (hash(x)=21248405)
103
+ 97 train 8.230323 (lr=2.7413e-04) (hash(x)=25830182)
104
+ 98 train 8.148446 (lr=2.7692e-04) (hash(x)=25173113)
105
+ 99 train 8.166368 (lr=2.7972e-04) (hash(x)=23486277)
106
+ 100 val loss 8.1505
107
+ 100 val perplexity 3465.0857
108
+ 100 train 8.163050 (lr=2.8252e-04) (hash(x)=28916006)
109
+ 101 train 8.291272 (lr=2.8531e-04) (hash(x)=25556929)
110
+ 102 train 8.167394 (lr=2.8811e-04) (hash(x)=29648798)
111
+ 103 train 8.083234 (lr=2.9091e-04) (hash(x)=24976217)
112
+ 104 train 8.134238 (lr=2.9371e-04) (hash(x)=31019606)
113
+ 105 train 8.138225 (lr=2.9650e-04) (hash(x)=28269421)
114
+ 106 train 8.030826 (lr=2.9930e-04) (hash(x)=24865358)
115
+ 107 train 7.965565 (lr=3.0210e-04) (hash(x)=26984272)
116
+ 108 train 8.067903 (lr=3.0490e-04) (hash(x)=24544116)
117
+ 109 train 8.012407 (lr=3.0769e-04) (hash(x)=22059850)
118
+ 110 train 8.027818 (lr=3.1049e-04) (hash(x)=23926632)
119
+ 111 train 7.941473 (lr=3.1329e-04) (hash(x)=27025333)
120
+ 112 train 7.893347 (lr=3.1608e-04) (hash(x)=29120407)
121
+ 113 train 7.971215 (lr=3.1888e-04) (hash(x)=23827393)
122
+ 114 train 7.913682 (lr=3.2168e-04) (hash(x)=27230027)
123
+ 115 train 7.888906 (lr=3.2448e-04) (hash(x)=24968260)
124
+ 116 train 7.877968 (lr=3.2727e-04) (hash(x)=27544803)
125
+ 117 train 7.889184 (lr=3.3007e-04) (hash(x)=26540663)
126
+ 118 train 7.954833 (lr=3.3287e-04) (hash(x)=25856625)
127
+ 119 train 7.866508 (lr=3.3566e-04) (hash(x)=23164356)
128
+ 120 train 7.899280 (lr=3.3846e-04) (hash(x)=25646282)
129
+ 121 train 7.894170 (lr=3.4126e-04) (hash(x)=23843233)
130
+ 122 train 7.707137 (lr=3.4406e-04) (hash(x)=24506028)
131
+ 123 train 7.844332 (lr=3.4685e-04) (hash(x)=22781277)
132
+ 124 train 7.832813 (lr=3.4965e-04) (hash(x)=24948650)
133
+ 125 train 7.799654 (lr=3.5245e-04) (hash(x)=26373991)
134
+ 126 train 7.882482 (lr=3.5524e-04) (hash(x)=23402396)
135
+ 127 train 7.820685 (lr=3.5804e-04) (hash(x)=21562133)
136
+ 128 train 7.865142 (lr=3.6084e-04) (hash(x)=25009210)
137
+ 129 train 7.801294 (lr=3.6364e-04) (hash(x)=27161811)
138
+ 130 train 7.834145 (lr=3.6643e-04) (hash(x)=29515100)
139
+ 131 train 7.752590 (lr=3.6923e-04) (hash(x)=25305929)
140
+ 132 train 7.791063 (lr=3.7203e-04) (hash(x)=26808543)
141
+ 133 train 7.706482 (lr=3.7483e-04) (hash(x)=25669509)
142
+ 134 train 7.792989 (lr=3.7762e-04) (hash(x)=28435805)
143
+ 135 train 7.768349 (lr=3.8042e-04) (hash(x)=26482758)
144
+ 136 train 7.690908 (lr=3.8322e-04) (hash(x)=24929929)
145
+ 137 train 7.677174 (lr=3.8601e-04) (hash(x)=25043037)
146
+ 138 train 7.896683 (lr=3.8881e-04) (hash(x)=29151257)
147
+ 139 train 7.657455 (lr=3.9161e-04) (hash(x)=26400365)
148
+ 140 train 7.669600 (lr=3.9441e-04) (hash(x)=24098812)
149
+ 141 train 7.658542 (lr=3.9720e-04) (hash(x)=20967666)
150
+ 142 train 7.800637 (lr=4.0000e-04) (hash(x)=28667963)
151
+ 143 train 7.745078 (lr=4.0280e-04) (hash(x)=27083417)
152
+ 144 train 7.607168 (lr=4.0559e-04) (hash(x)=23649001)
153
+ 145 train 7.661515 (lr=4.0839e-04) (hash(x)=31338300)
154
+ 146 train 7.645726 (lr=4.1119e-04) (hash(x)=30603174)
155
+ 147 train 7.665558 (lr=4.1399e-04) (hash(x)=24830965)
156
+ 148 train 7.799366 (lr=4.1678e-04) (hash(x)=27269893)
157
+ 149 train 7.633232 (lr=4.1958e-04) (hash(x)=24825041)
158
+ 150 val loss 7.6640
159
+ 150 val perplexity 2130.2129
160
+ 150 train 7.619266 (lr=4.2238e-04) (hash(x)=22724871)
161
+ 151 train 7.567100 (lr=4.2517e-04) (hash(x)=24683717)
162
+ 152 train 7.697691 (lr=4.2797e-04) (hash(x)=26190337)
163
+ 153 train 7.661534 (lr=4.3077e-04) (hash(x)=23086289)
164
+ 154 train 7.711994 (lr=4.3357e-04) (hash(x)=23593875)
165
+ 155 train 7.551075 (lr=4.3636e-04) (hash(x)=22414190)
166
+ 156 train 7.636968 (lr=4.3916e-04) (hash(x)=26758114)
167
+ 157 train 7.581555 (lr=4.4196e-04) (hash(x)=21563184)
168
+ 158 train 7.633489 (lr=4.4476e-04) (hash(x)=19218470)
169
+ 159 train 7.733710 (lr=4.4755e-04) (hash(x)=25588469)
170
+ 160 train 7.708292 (lr=4.5035e-04) (hash(x)=30433767)
171
+ 161 train 7.777137 (lr=4.5315e-04) (hash(x)=17851621)
172
+ 162 train 7.834107 (lr=4.5594e-04) (hash(x)=25183690)
173
+ 163 train 7.888913 (lr=4.5874e-04) (hash(x)=27040642)
174
+ 164 train 7.638313 (lr=4.6154e-04) (hash(x)=22183303)
175
+ 165 train 7.949181 (lr=4.6434e-04) (hash(x)=27537063)
176
+ 166 train 7.641860 (lr=4.6713e-04) (hash(x)=19492832)
177
+ 167 train 7.643179 (lr=4.6993e-04) (hash(x)=25951629)
178
+ 168 train 7.671164 (lr=4.7273e-04) (hash(x)=22408682)
179
+ 169 train 7.657295 (lr=4.7552e-04) (hash(x)=27633457)
180
+ 170 train 7.638868 (lr=4.7832e-04) (hash(x)=27635080)
181
+ 171 train 7.637440 (lr=4.8112e-04) (hash(x)=22287596)
182
+ 172 train 7.751910 (lr=4.8392e-04) (hash(x)=35895440)
183
+ 173 train 7.660897 (lr=4.8671e-04) (hash(x)=25888847)
184
+ 174 train 7.697286 (lr=4.8951e-04) (hash(x)=26518800)
185
+ 175 train 7.672818 (lr=4.9231e-04) (hash(x)=37519283)
186
+ 176 train 7.859879 (lr=4.9510e-04) (hash(x)=25141945)
187
+ 177 train 7.867149 (lr=4.9790e-04) (hash(x)=28527939)
188
+ 178 train 7.665414 (lr=5.0070e-04) (hash(x)=21942471)
189
+ 179 train 7.694840 (lr=5.0350e-04) (hash(x)=24955630)
190
+ 180 train 7.663059 (lr=5.0629e-04) (hash(x)=25541230)
191
+ 181 train 7.497454 (lr=5.0909e-04) (hash(x)=23501481)
192
+ 182 train 7.561071 (lr=5.1189e-04) (hash(x)=22754988)
193
+ 183 train 7.468179 (lr=5.1469e-04) (hash(x)=23475070)
194
+ 184 train 7.611307 (lr=5.1748e-04) (hash(x)=23609959)
195
+ 185 train 7.498202 (lr=5.2028e-04) (hash(x)=23441847)
196
+ 186 train 7.559904 (lr=5.2308e-04) (hash(x)=23096292)
197
+ 187 train 7.599545 (lr=5.2587e-04) (hash(x)=26171093)
198
+ 188 train 7.624224 (lr=5.2867e-04) (hash(x)=20060838)
199
+ 189 train 7.477706 (lr=5.3147e-04) (hash(x)=25451055)
200
+ 190 train 7.326308 (lr=5.3427e-04) (hash(x)=24934792)
201
+ 191 train 7.445419 (lr=5.3706e-04) (hash(x)=25367597)
202
+ 192 train 7.597335 (lr=5.3986e-04) (hash(x)=25022148)
203
+ 193 train 7.582376 (lr=5.4266e-04) (hash(x)=25431365)
204
+ 194 train 7.624408 (lr=5.4545e-04) (hash(x)=29307788)
205
+ 195 train 7.507672 (lr=5.4825e-04) (hash(x)=25651929)
206
+ 196 train 7.564185 (lr=5.5105e-04) (hash(x)=26312387)
207
+ 197 train 7.523287 (lr=5.5385e-04) (hash(x)=23804913)
208
+ 198 train 7.563671 (lr=5.5664e-04) (hash(x)=26115519)
209
+ 199 train 7.542455 (lr=5.5944e-04) (hash(x)=23210747)
210
+ 200 val loss 7.5611
211
+ 200 val perplexity 1921.9099
212
+ 200 train 7.561044 (lr=5.6224e-04) (hash(x)=18463712)
213
+ 201 train 7.550570 (lr=5.6503e-04) (hash(x)=23245699)
214
+ 202 train 7.463957 (lr=5.6783e-04) (hash(x)=21770300)
215
+ 203 train 7.730957 (lr=5.7063e-04) (hash(x)=25387532)
216
+ 204 train 7.549559 (lr=5.7343e-04) (hash(x)=27718703)
217
+ 205 train 7.599236 (lr=5.7622e-04) (hash(x)=27661052)
218
+ 206 train 7.482681 (lr=5.7902e-04) (hash(x)=26538928)
219
+ 207 train 7.584099 (lr=5.8182e-04) (hash(x)=21772654)
220
+ 208 train 7.687113 (lr=5.8462e-04) (hash(x)=26263254)
221
+ 209 train 7.546593 (lr=5.8741e-04) (hash(x)=23552356)
222
+ 210 train 7.762096 (lr=5.9021e-04) (hash(x)=22707141)
223
+ 211 train 7.790479 (lr=5.9301e-04) (hash(x)=34234481)
224
+ 212 train 7.599404 (lr=5.9580e-04) (hash(x)=35008670)
225
+ 213 train 7.709475 (lr=5.9860e-04) (hash(x)=25891543)
226
+ 214 train 7.462895 (lr=6.0140e-04) (hash(x)=21971224)
227
+ 215 train 7.586832 (lr=6.0420e-04) (hash(x)=27312186)
228
+ 216 train 7.688448 (lr=6.0699e-04) (hash(x)=22163749)
229
+ 217 train 7.436416 (lr=6.0979e-04) (hash(x)=25942888)
230
+ 218 train 7.555321 (lr=6.1259e-04) (hash(x)=25493916)
231
+ 219 train 7.737813 (lr=6.1538e-04) (hash(x)=24289951)
232
+ 220 train 7.520095 (lr=6.1818e-04) (hash(x)=23648671)
233
+ 221 train 7.380739 (lr=6.2098e-04) (hash(x)=24992055)
234
+ 222 train 7.662001 (lr=6.2378e-04) (hash(x)=26005549)
235
+ 223 train 7.510076 (lr=6.2657e-04) (hash(x)=21046554)
236
+ 224 train 7.513074 (lr=6.2937e-04) (hash(x)=26321539)
237
+ 225 train 7.528290 (lr=6.3217e-04) (hash(x)=27807872)
238
+ 226 train 7.516778 (lr=6.3497e-04) (hash(x)=21555364)
239
+ 227 train 7.510825 (lr=6.3776e-04) (hash(x)=26325324)
240
+ 228 train 7.543692 (lr=6.4056e-04) (hash(x)=24956294)
241
+ 229 train 7.462821 (lr=6.4336e-04) (hash(x)=24429389)
242
+ 230 train 7.604789 (lr=6.4615e-04) (hash(x)=27929846)
243
+ 231 train 7.539295 (lr=6.4895e-04) (hash(x)=25282897)
244
+ 232 train 7.391565 (lr=6.5175e-04) (hash(x)=26802508)
245
+ 233 train 7.592510 (lr=6.5455e-04) (hash(x)=28390416)
246
+ 234 train 7.734347 (lr=6.5734e-04) (hash(x)=26292451)
247
+ 235 train 7.625930 (lr=6.6014e-04) (hash(x)=22915344)
248
+ 236 train 7.508337 (lr=6.6294e-04) (hash(x)=24744383)
249
+ 237 train 7.442818 (lr=6.6573e-04) (hash(x)=26284678)
250
+ 238 train 7.418664 (lr=6.6853e-04) (hash(x)=21654269)
251
+ 239 train 7.867980 (lr=6.7133e-04) (hash(x)=29466268)
252
+ 240 train 8.127902 (lr=6.7413e-04) (hash(x)=32616762)
253
+ 241 train 8.117806 (lr=6.7692e-04) (hash(x)=32201138)
254
+ 242 train 8.095565 (lr=6.7972e-04) (hash(x)=33683468)
255
+ 243 train 7.786491 (lr=6.8252e-04) (hash(x)=27498125)
256
+ 244 train 7.637064 (lr=6.8531e-04) (hash(x)=25164318)
257
+ 245 train 7.494011 (lr=6.8811e-04) (hash(x)=25690221)
258
+ 246 train 7.545064 (lr=6.9091e-04) (hash(x)=29031550)
259
+ 247 train 7.688444 (lr=6.9371e-04) (hash(x)=24094093)
260
+ 248 train 8.344934 (lr=6.9650e-04) (hash(x)=61455960)
261
+ 249 train 11.687546 (lr=6.9930e-04) (hash(x)=61104710)
262
+ 250 val loss 7.5852
263
+ 250 val perplexity 1968.7433
264
+ 250 train 7.728424 (lr=7.0210e-04) (hash(x)=20881126)
265
+ 251 train 7.678052 (lr=7.0490e-04) (hash(x)=26920243)
266
+ 252 train 7.563182 (lr=7.0769e-04) (hash(x)=21970343)
267
+ 253 train 7.623492 (lr=7.1049e-04) (hash(x)=31676352)
268
+ 254 train 7.616559 (lr=7.1329e-04) (hash(x)=25600562)
269
+ 255 train 7.556277 (lr=7.1608e-04) (hash(x)=22022924)
270
+ 256 train 7.460969 (lr=7.1888e-04) (hash(x)=22777085)
271
+ 257 train 7.607826 (lr=7.2168e-04) (hash(x)=21727744)
272
+ 258 train 7.613549 (lr=7.2448e-04) (hash(x)=20701998)
273
+ 259 train 7.697864 (lr=7.2727e-04) (hash(x)=20863982)
274
+ 260 train 7.569373 (lr=7.3007e-04) (hash(x)=25084926)
275
+ 261 train 7.553705 (lr=7.3287e-04) (hash(x)=24412897)
276
+ 262 train 7.516550 (lr=7.3566e-04) (hash(x)=25434913)
277
+ 263 train 7.555886 (lr=7.3846e-04) (hash(x)=26302544)
278
+ 264 train 7.623404 (lr=7.4126e-04) (hash(x)=34976197)
279
+ 265 train 7.514687 (lr=7.4406e-04) (hash(x)=23870805)
280
+ 266 train 7.528773 (lr=7.4685e-04) (hash(x)=25041103)
281
+ 267 train 7.564066 (lr=7.4965e-04) (hash(x)=26041745)
282
+ 268 train 7.542047 (lr=7.5245e-04) (hash(x)=24068619)
283
+ 269 train 7.561988 (lr=7.5524e-04) (hash(x)=25441262)
284
+ 270 train 7.622159 (lr=7.5804e-04) (hash(x)=29181867)
285
+ 271 train 7.613667 (lr=7.6084e-04) (hash(x)=27093757)
286
+ 272 train 7.511677 (lr=7.6364e-04) (hash(x)=25150275)
287
+ 273 train 7.654326 (lr=7.6643e-04) (hash(x)=33457102)
288
+ 274 train 7.462945 (lr=7.6923e-04) (hash(x)=25755322)
289
+ 275 train 7.535835 (lr=7.7203e-04) (hash(x)=23630794)
290
+ 276 train 7.453933 (lr=7.7483e-04) (hash(x)=25167987)
291
+ 277 train 7.562566 (lr=7.7762e-04) (hash(x)=24043224)
292
+ 278 train 7.528729 (lr=7.8042e-04) (hash(x)=24665411)
293
+ 279 train 7.488905 (lr=7.8322e-04) (hash(x)=17321657)
294
+ 280 train 7.491410 (lr=7.8601e-04) (hash(x)=27932591)
295
+ 281 train 7.796443 (lr=7.8881e-04) (hash(x)=29447954)
296
+ 282 train 7.515444 (lr=7.9161e-04) (hash(x)=24621816)
297
+ 283 train 7.480319 (lr=7.9441e-04) (hash(x)=24682294)
298
+ 284 train 7.504038 (lr=7.9720e-04) (hash(x)=28057095)
299
+ 285 train 7.557436 (lr=8.0000e-04) (hash(x)=24547536)
300
+ 286 train 7.407906 (lr=8.0000e-04) (hash(x)=18987545)
301
+ 287 train 7.476029 (lr=8.0000e-04) (hash(x)=26745063)
302
+ 288 train 7.340400 (lr=7.9999e-04) (hash(x)=21447406)
303
+ 289 train 7.241962 (lr=7.9997e-04) (hash(x)=24650655)
304
+ 290 train 7.609638 (lr=7.9994e-04) (hash(x)=26519976)
305
+ 291 train 7.912451 (lr=7.9991e-04) (hash(x)=20106514)
306
+ 292 train 7.362346 (lr=7.9987e-04) (hash(x)=20117808)
307
+ 293 train 7.507954 (lr=7.9983e-04) (hash(x)=25510334)
308
+ 294 train 7.412135 (lr=7.9978e-04) (hash(x)=25749120)
309
+ 295 train 7.514361 (lr=7.9972e-04) (hash(x)=24357241)
310
+ 296 train 7.496769 (lr=7.9965e-04) (hash(x)=25584945)
311
+ 297 train 7.381025 (lr=7.9958e-04) (hash(x)=25198897)
312
+ 298 train 7.479185 (lr=7.9950e-04) (hash(x)=23475891)
313
+ 299 train 7.411061 (lr=7.9941e-04) (hash(x)=23644804)
314
+ 300 val loss 7.5084
315
+ 300 val perplexity 1823.2217
316
+ 300 train 7.648129 (lr=7.9932e-04) (hash(x)=25629584)
317
+ 301 train 7.753440 (lr=7.9922e-04) (hash(x)=29797280)
318
+ 302 train 7.814864 (lr=7.9911e-04) (hash(x)=30026394)
319
+ 303 train 7.779068 (lr=7.9899e-04) (hash(x)=26594177)
320
+ 304 train 7.575085 (lr=7.9887e-04) (hash(x)=23766256)
321
+ 305 train 7.506543 (lr=7.9874e-04) (hash(x)=24250150)
322
+ 306 train 7.508531 (lr=7.9861e-04) (hash(x)=26108817)
323
+ 307 train 7.477237 (lr=7.9846e-04) (hash(x)=18582688)
324
+ 308 train 7.465552 (lr=7.9831e-04) (hash(x)=23272705)
325
+ 309 train 7.446856 (lr=7.9816e-04) (hash(x)=24793412)
326
+ 310 train 7.445893 (lr=7.9799e-04) (hash(x)=19280989)
327
+ 311 train 7.432438 (lr=7.9782e-04) (hash(x)=22751150)
328
+ 312 train 7.543500 (lr=7.9765e-04) (hash(x)=23537942)
329
+ 313 train 7.383286 (lr=7.9746e-04) (hash(x)=23748375)
330
+ 314 train 7.557084 (lr=7.9727e-04) (hash(x)=27413825)
331
+ 315 train 7.607321 (lr=7.9707e-04) (hash(x)=21682445)
332
+ 316 train 7.473935 (lr=7.9687e-04) (hash(x)=22040184)
333
+ 317 train 7.468248 (lr=7.9666e-04) (hash(x)=24463229)
334
+ 318 train 7.590176 (lr=7.9644e-04) (hash(x)=24091954)
335
+ 319 train 7.470326 (lr=7.9621e-04) (hash(x)=26670219)
336
+ 320 train 7.498962 (lr=7.9598e-04) (hash(x)=25558986)
337
+ 321 train 7.482043 (lr=7.9574e-04) (hash(x)=19218943)
338
+ 322 train 7.582918 (lr=7.9549e-04) (hash(x)=24551402)
339
+ 323 train 7.459771 (lr=7.9524e-04) (hash(x)=25229281)
340
+ 324 train 7.225049 (lr=7.9498e-04) (hash(x)=21251127)
341
+ 325 train 7.515058 (lr=7.9471e-04) (hash(x)=25800857)
342
+ 326 train 7.390404 (lr=7.9444e-04) (hash(x)=25836189)
343
+ 327 train 7.404307 (lr=7.9416e-04) (hash(x)=25999255)
344
+ 328 train 7.489329 (lr=7.9387e-04) (hash(x)=20149394)
345
+ 329 train 7.407252 (lr=7.9358e-04) (hash(x)=21999338)
346
+ 330 train 7.498377 (lr=7.9327e-04) (hash(x)=27611302)
347
+ 331 train 7.470280 (lr=7.9297e-04) (hash(x)=22778356)
348
+ 332 train 7.447369 (lr=7.9265e-04) (hash(x)=24195959)
349
+ 333 train 7.702252 (lr=7.9233e-04) (hash(x)=19127465)
350
+ 334 train 7.558182 (lr=7.9200e-04) (hash(x)=22426274)
351
+ 335 train 7.380886 (lr=7.9167e-04) (hash(x)=23785205)
352
+ 336 train 7.526174 (lr=7.9132e-04) (hash(x)=28505676)
353
+ 337 train 7.489627 (lr=7.9097e-04) (hash(x)=27706477)
354
+ 338 train 7.522386 (lr=7.9062e-04) (hash(x)=28684175)
355
+ 339 train 7.496419 (lr=7.9026e-04) (hash(x)=28003600)
356
+ 340 train 7.509902 (lr=7.8989e-04) (hash(x)=27586845)
357
+ 341 train 7.426126 (lr=7.8951e-04) (hash(x)=21176405)
358
+ 342 train 7.388934 (lr=7.8913e-04) (hash(x)=25888433)
359
+ 343 train 7.567827 (lr=7.8874e-04) (hash(x)=27916982)
360
+ 344 train 7.411689 (lr=7.8834e-04) (hash(x)=24334708)
361
+ 345 train 7.522184 (lr=7.8794e-04) (hash(x)=25617709)
362
+ 346 train 7.684250 (lr=7.8753e-04) (hash(x)=25152362)
363
+ 347 train 7.572710 (lr=7.8711e-04) (hash(x)=24276512)
364
+ 348 train 7.509953 (lr=7.8669e-04) (hash(x)=26776133)
365
+ 349 train 7.430198 (lr=7.8626e-04) (hash(x)=20254159)
366
+ 350 val loss 7.5986
367
+ 350 val perplexity 1995.3088
368
+ 350 train 7.472949 (lr=7.8582e-04) (hash(x)=27873855)
369
+ 351 train 7.529803 (lr=7.8538e-04) (hash(x)=25113614)
370
+ 352 train 7.576036 (lr=7.8493e-04) (hash(x)=26192886)
371
+ 353 train 7.458933 (lr=7.8447e-04) (hash(x)=24614912)
372
+ 354 train 7.458200 (lr=7.8401e-04) (hash(x)=25397093)
373
+ 355 train 7.493171 (lr=7.8354e-04) (hash(x)=19811802)
374
+ 356 train 7.476529 (lr=7.8306e-04) (hash(x)=25579655)
375
+ 357 train 7.606736 (lr=7.8258e-04) (hash(x)=26096514)
376
+ 358 train 7.508423 (lr=7.8209e-04) (hash(x)=22589633)
377
+ 359 train 7.529704 (lr=7.8159e-04) (hash(x)=26757147)
378
+ 360 train 7.601413 (lr=7.8109e-04) (hash(x)=24963730)
379
+ 361 train 7.469187 (lr=7.8058e-04) (hash(x)=27703124)
380
+ 362 train 7.381424 (lr=7.8006e-04) (hash(x)=24781713)
381
+ 363 train 7.564112 (lr=7.7954e-04) (hash(x)=24720476)
382
+ 364 train 7.564489 (lr=7.7901e-04) (hash(x)=23481301)
383
+ 365 train 7.456219 (lr=7.7847e-04) (hash(x)=25641033)
384
+ 366 train 7.532192 (lr=7.7793e-04) (hash(x)=27369437)
385
+ 367 train 7.622892 (lr=7.7738e-04) (hash(x)=24760900)
386
+ 368 train 7.442381 (lr=7.7682e-04) (hash(x)=21692546)
387
+ 369 train 7.520278 (lr=7.7626e-04) (hash(x)=22720391)
388
+ 370 train 7.480254 (lr=7.7569e-04) (hash(x)=24376979)
389
+ 371 train 7.491879 (lr=7.7511e-04) (hash(x)=22244509)
390
+ 372 train 7.423918 (lr=7.7453e-04) (hash(x)=21831960)
391
+ 373 train 7.408247 (lr=7.7394e-04) (hash(x)=24921535)
392
+ 374 train 7.476384 (lr=7.7335e-04) (hash(x)=25381251)
393
+ 375 train 7.482609 (lr=7.7275e-04) (hash(x)=25687352)
394
+ 376 train 7.515439 (lr=7.7214e-04) (hash(x)=25221654)
395
+ 377 train 7.546837 (lr=7.7153e-04) (hash(x)=24469863)
396
+ 378 train 7.600894 (lr=7.7091e-04) (hash(x)=25291413)
397
+ 379 train 7.474998 (lr=7.7028e-04) (hash(x)=25744274)
398
+ 380 train 7.400367 (lr=7.6964e-04) (hash(x)=24246963)
399
+ 381 train 7.417056 (lr=7.6901e-04) (hash(x)=19724058)
400
+ 382 train 7.349379 (lr=7.6836e-04) (hash(x)=20705649)
401
+ 383 train 7.623726 (lr=7.6771e-04) (hash(x)=24938685)
402
+ 384 train 7.643369 (lr=7.6705e-04) (hash(x)=21707000)
403
+ 385 train 7.440569 (lr=7.6638e-04) (hash(x)=24948980)
404
+ 386 train 7.720186 (lr=7.6571e-04) (hash(x)=30773958)
405
+ 387 train 7.814635 (lr=7.6503e-04) (hash(x)=26769046)
406
+ 388 train 7.494871 (lr=7.6435e-04) (hash(x)=18849656)
407
+ 389 train 7.994859 (lr=7.6366e-04) (hash(x)=24624950)
408
+ 390 train 7.593480 (lr=7.6296e-04) (hash(x)=27299605)
409
+ 391 train 7.602159 (lr=7.6226e-04) (hash(x)=24113445)
410
+ 392 train 7.569585 (lr=7.6155e-04) (hash(x)=32644465)
411
+ 393 train 7.672410 (lr=7.6083e-04) (hash(x)=25829219)
412
+ 394 train 7.510962 (lr=7.6011e-04) (hash(x)=25194550)
413
+ 395 train 7.553936 (lr=7.5938e-04) (hash(x)=25941804)
414
+ 396 train 7.549150 (lr=7.5865e-04) (hash(x)=24075727)
415
+ 397 train 7.606919 (lr=7.5791e-04) (hash(x)=23104539)
416
+ 398 train 7.550572 (lr=7.5716e-04) (hash(x)=23018983)
417
+ 399 train 7.510014 (lr=7.5641e-04) (hash(x)=24422929)
418
+ 400 val loss 7.5561
419
+ 400 val perplexity 1912.4291
420
+ 400 train 7.517997 (lr=7.5565e-04) (hash(x)=23389696)
421
+ 401 train 7.498629 (lr=7.5489e-04) (hash(x)=29102969)
422
+ 402 train 7.645426 (lr=7.5412e-04) (hash(x)=28359859)
423
+ 403 train 7.651928 (lr=7.5334e-04) (hash(x)=24630955)
424
+ 404 train 7.675635 (lr=7.5256e-04) (hash(x)=22969113)
425
+ 405 train 7.510665 (lr=7.5177e-04) (hash(x)=27004415)
426
+ 406 train 7.432611 (lr=7.5097e-04) (hash(x)=22265063)
427
+ 407 train 7.449023 (lr=7.5017e-04) (hash(x)=26905990)
428
+ 408 train 7.508177 (lr=7.4937e-04) (hash(x)=24047679)
429
+ 409 train 7.633502 (lr=7.4855e-04) (hash(x)=26905582)
430
+ 410 train 7.393932 (lr=7.4773e-04) (hash(x)=29018970)
431
+ 411 train 7.443156 (lr=7.4691e-04) (hash(x)=24503315)
432
+ 412 train 7.455684 (lr=7.4608e-04) (hash(x)=22607537)
433
+ 413 train 8.129822 (lr=7.4524e-04) (hash(x)=26690225)
434
+ 414 train 7.597682 (lr=7.4440e-04) (hash(x)=24226255)
435
+ 415 train 7.540190 (lr=7.4355e-04) (hash(x)=23622702)
436
+ 416 train 7.583941 (lr=7.4270e-04) (hash(x)=23910425)
437
+ 417 train 7.793859 (lr=7.4184e-04) (hash(x)=27213318)
438
+ 418 train 7.586904 (lr=7.4097e-04) (hash(x)=26701577)
439
+ 419 train 7.501926 (lr=7.4010e-04) (hash(x)=26216794)
440
+ 420 train 7.683312 (lr=7.3922e-04) (hash(x)=27579623)
441
+ 421 train 7.571878 (lr=7.3834e-04) (hash(x)=26086198)
442
+ 422 train 7.526848 (lr=7.3745e-04) (hash(x)=24863441)
443
+ 423 train 7.640356 (lr=7.3655e-04) (hash(x)=25236385)
444
+ 424 train 7.716364 (lr=7.3565e-04) (hash(x)=25330803)
445
+ 425 train 7.564563 (lr=7.3474e-04) (hash(x)=22508532)
446
+ 426 train 7.558948 (lr=7.3383e-04) (hash(x)=21799261)
447
+ 427 train 7.640644 (lr=7.3291e-04) (hash(x)=26311168)
448
+ 428 train 7.467375 (lr=7.3199e-04) (hash(x)=20977430)
449
+ 429 train 7.529807 (lr=7.3106e-04) (hash(x)=26941617)
450
+ 430 train 7.759141 (lr=7.3012e-04) (hash(x)=30008957)
451
+ 431 train 7.610386 (lr=7.2918e-04) (hash(x)=27125962)
452
+ 432 train 7.465990 (lr=7.2824e-04) (hash(x)=25888358)
453
+ 433 train 7.635141 (lr=7.2729e-04) (hash(x)=21725719)
454
+ 434 train 7.492973 (lr=7.2633e-04) (hash(x)=21669704)
455
+ 435 train 7.463477 (lr=7.2537e-04) (hash(x)=23166993)
456
+ 436 train 7.542214 (lr=7.2440e-04) (hash(x)=26235974)
457
+ 437 train 7.720412 (lr=7.2342e-04) (hash(x)=25426430)
458
+ 438 train 7.471004 (lr=7.2244e-04) (hash(x)=22966314)
459
+ 439 train 7.616211 (lr=7.2146e-04) (hash(x)=23448505)
460
+ 440 train 7.500141 (lr=7.2047e-04) (hash(x)=26330693)
461
+ 441 train 7.461365 (lr=7.1947e-04) (hash(x)=29550596)
462
+ 442 train 7.538635 (lr=7.1847e-04) (hash(x)=22124892)
463
+ 443 train 7.562909 (lr=7.1747e-04) (hash(x)=25607640)
464
+ 444 train 7.436483 (lr=7.1645e-04) (hash(x)=23795894)
465
+ 445 train 7.342511 (lr=7.1544e-04) (hash(x)=24165449)
466
+ 446 train 7.582679 (lr=7.1441e-04) (hash(x)=24697756)
467
+ 447 train 7.641857 (lr=7.1339e-04) (hash(x)=27218899)
468
+ 448 train 7.525071 (lr=7.1235e-04) (hash(x)=24748522)
469
+ 449 train 7.639192 (lr=7.1131e-04) (hash(x)=24965317)
470
+ 450 val loss 7.6418
471
+ 450 val perplexity 2083.5461
472
+ 450 train 7.424412 (lr=7.1027e-04) (hash(x)=22031210)
473
+ 451 train 7.554656 (lr=7.0922e-04) (hash(x)=25877990)
474
+ 452 train 7.596772 (lr=7.0817e-04) (hash(x)=24382381)
475
+ 453 train 7.499175 (lr=7.0711e-04) (hash(x)=24551999)
476
+ 454 train 7.595683 (lr=7.0604e-04) (hash(x)=25765516)
477
+ 455 train 7.601314 (lr=7.0497e-04) (hash(x)=25204247)
478
+ 456 train 7.656246 (lr=7.0390e-04) (hash(x)=26321813)
479
+ 457 train 7.596511 (lr=7.0282e-04) (hash(x)=25528323)
480
+ 458 train 7.598839 (lr=7.0173e-04) (hash(x)=25114729)
481
+ 459 train 7.562001 (lr=7.0064e-04) (hash(x)=31341964)
482
+ 460 train 7.627997 (lr=6.9955e-04) (hash(x)=26326024)
483
+ 461 train 7.540641 (lr=6.9845e-04) (hash(x)=27273311)
484
+ 462 train 7.534899 (lr=6.9734e-04) (hash(x)=17819313)
485
+ 463 train 7.684968 (lr=6.9623e-04) (hash(x)=23769243)
486
+ 464 train 7.652684 (lr=6.9512e-04) (hash(x)=23788738)
487
+ 465 train 7.606667 (lr=6.9400e-04) (hash(x)=24818744)
488
+ 466 train 7.523011 (lr=6.9287e-04) (hash(x)=23472795)
489
+ 467 train 7.898815 (lr=6.9174e-04) (hash(x)=24991832)
490
+ 468 train 7.667626 (lr=6.9061e-04) (hash(x)=22424066)
491
+ 469 train 7.562103 (lr=6.8947e-04) (hash(x)=26474878)
492
+ 470 train 7.647557 (lr=6.8832e-04) (hash(x)=22704349)
493
+ 471 train 7.589141 (lr=6.8717e-04) (hash(x)=22293673)
494
+ 472 train 7.620806 (lr=6.8602e-04) (hash(x)=21785487)
495
+ 473 train 7.622032 (lr=6.8486e-04) (hash(x)=25251063)
496
+ 474 train 7.532951 (lr=6.8370e-04) (hash(x)=22722211)
497
+ 475 train 7.585145 (lr=6.8253e-04) (hash(x)=24522269)
498
+ 476 train 7.552441 (lr=6.8136e-04) (hash(x)=23329475)
499
+ 477 train 7.584072 (lr=6.8018e-04) (hash(x)=24735165)
500
+ 478 train 7.413013 (lr=6.7900e-04) (hash(x)=27002892)
501
+ 479 train 7.461753 (lr=6.7781e-04) (hash(x)=23499349)
502
+ 480 train 7.767808 (lr=6.7662e-04) (hash(x)=22461589)
503
+ 481 train 7.542376 (lr=6.7542e-04) (hash(x)=26415040)
504
+ 482 train 7.760121 (lr=6.7422e-04) (hash(x)=26349465)
505
+ 483 train 7.449440 (lr=6.7302e-04) (hash(x)=21596677)
506
+ 484 train 7.490698 (lr=6.7181e-04) (hash(x)=24117012)
507
+ 485 train 7.570382 (lr=6.7059e-04) (hash(x)=24105761)
508
+ 486 train 7.536638 (lr=6.6937e-04) (hash(x)=25725049)
509
+ 487 train 7.386079 (lr=6.6815e-04) (hash(x)=22892776)
510
+ 488 train 7.485522 (lr=6.6692e-04) (hash(x)=21630383)
511
+ 489 train 7.532235 (lr=6.6569e-04) (hash(x)=24727094)
512
+ 490 train 7.464844 (lr=6.6446e-04) (hash(x)=22835000)
513
+ 491 train 7.623252 (lr=6.6322e-04) (hash(x)=22839049)
514
+ 492 train 7.476864 (lr=6.6197e-04) (hash(x)=20654136)
515
+ 493 train 7.700920 (lr=6.6072e-04) (hash(x)=26776000)
516
+ 494 train 7.338257 (lr=6.5947e-04) (hash(x)=22560658)
517
+ 495 train 7.752440 (lr=6.5821e-04) (hash(x)=25995933)
518
+ 496 train 7.658994 (lr=6.5695e-04) (hash(x)=24635394)
519
+ 497 train 7.768842 (lr=6.5568e-04) (hash(x)=27537228)
520
+ 498 train 7.710566 (lr=6.5441e-04) (hash(x)=26496730)
521
+ 499 train 7.424687 (lr=6.5314e-04) (hash(x)=16533310)
522
+ 500 val loss 7.6669
523
+ 500 val perplexity 2136.4832
524
+ 500 train 7.568194 (lr=6.5186e-04) (hash(x)=24881176)
525
+ 501 train 7.763610 (lr=6.5058e-04) (hash(x)=25593613)
526
+ 502 train 7.538302 (lr=6.4929e-04) (hash(x)=22886646)
527
+ 503 train 7.796624 (lr=6.4800e-04) (hash(x)=21407676)
528
+ 504 train 7.639430 (lr=6.4670e-04) (hash(x)=21720304)
529
+ 505 train 7.767704 (lr=6.4540e-04) (hash(x)=27263416)
530
+ 506 train 7.527770 (lr=6.4410e-04) (hash(x)=25840801)
531
+ 507 train 7.667857 (lr=6.4280e-04) (hash(x)=31424568)
532
+ 508 train 7.677245 (lr=6.4148e-04) (hash(x)=23403065)
533
+ 509 train 7.505692 (lr=6.4017e-04) (hash(x)=22656033)
534
+ 510 train 7.665759 (lr=6.3885e-04) (hash(x)=26869295)
535
+ 511 train 7.872464 (lr=6.3753e-04) (hash(x)=25504151)
536
+ 512 train 7.651931 (lr=6.3620e-04) (hash(x)=25781518)
537
+ 513 train 7.634694 (lr=6.3487e-04) (hash(x)=26432850)
538
+ 514 train 7.522344 (lr=6.3354e-04) (hash(x)=20155894)
539
+ 515 train 7.456902 (lr=6.3220e-04) (hash(x)=22270169)
540
+ 516 train 7.566882 (lr=6.3086e-04) (hash(x)=22108461)
541
+ 517 train 7.604220 (lr=6.2952e-04) (hash(x)=24802848)
542
+ 518 train 7.729218 (lr=6.2817e-04) (hash(x)=25086604)
543
+ 519 train 7.743475 (lr=6.2681e-04) (hash(x)=26983186)
544
+ 520 train 7.651695 (lr=6.2546e-04) (hash(x)=27208507)
545
+ 521 train 7.817132 (lr=6.2410e-04) (hash(x)=23862845)
546
+ 522 train 7.699908 (lr=6.2274e-04) (hash(x)=22664494)
547
+ 523 train 7.727138 (lr=6.2137e-04) (hash(x)=24092927)
548
+ 524 train 7.700470 (lr=6.2000e-04) (hash(x)=27213812)
549
+ 525 train 7.661048 (lr=6.1863e-04) (hash(x)=23929955)
550
+ 526 train 7.657619 (lr=6.1725e-04) (hash(x)=23025790)
551
+ 527 train 7.547294 (lr=6.1587e-04) (hash(x)=20452272)
552
+ 528 train 7.690085 (lr=6.1449e-04) (hash(x)=26473464)
553
+ 529 train 7.658860 (lr=6.1310e-04) (hash(x)=28276603)
554
+ 530 train 7.485226 (lr=6.1171e-04) (hash(x)=21480367)
555
+ 531 train 7.661448 (lr=6.1031e-04) (hash(x)=24658642)
556
+ 532 train 7.563164 (lr=6.0892e-04) (hash(x)=26854801)
557
+ 533 train 7.472210 (lr=6.0752e-04) (hash(x)=17862226)
558
+ 534 train 7.582506 (lr=6.0611e-04) (hash(x)=28519182)
559
+ 535 train 7.578450 (lr=6.0471e-04) (hash(x)=25193694)
560
+ 536 train 7.548943 (lr=6.0330e-04) (hash(x)=25642620)
561
+ 537 train 7.636295 (lr=6.0188e-04) (hash(x)=25765909)
562
+ 538 train 7.444387 (lr=6.0047e-04) (hash(x)=20750531)
563
+ 539 train 7.561673 (lr=5.9905e-04) (hash(x)=24917098)
564
+ 540 train 7.519141 (lr=5.9762e-04) (hash(x)=24499766)
565
+ 541 train 7.560224 (lr=5.9620e-04) (hash(x)=25067358)
566
+ 542 train 7.672600 (lr=5.9477e-04) (hash(x)=27479796)
567
+ 543 train 7.551093 (lr=5.9334e-04) (hash(x)=23136080)
568
+ 544 train 7.644979 (lr=5.9190e-04) (hash(x)=29810074)
569
+ 545 train 7.551685 (lr=5.9047e-04) (hash(x)=24765539)
570
+ 546 train 7.601232 (lr=5.8903e-04) (hash(x)=22216309)
571
+ 547 train 7.520514 (lr=5.8758e-04) (hash(x)=20613861)
572
+ 548 train 7.503346 (lr=5.8614e-04) (hash(x)=23462447)
573
+ 549 train 7.563414 (lr=5.8469e-04) (hash(x)=26739455)
574
+ 550 val loss 7.6622
575
+ 550 val perplexity 2126.4092
576
+ 550 train 7.621760 (lr=5.8323e-04) (hash(x)=23222996)
577
+ 551 train 7.621488 (lr=5.8178e-04) (hash(x)=26489285)
578
+ 552 train 7.458960 (lr=5.8032e-04) (hash(x)=25185942)
579
+ 553 train 7.745366 (lr=5.7886e-04) (hash(x)=27476158)
580
+ 554 train 7.555231 (lr=5.7740e-04) (hash(x)=23653175)
581
+ 555 train 7.496541 (lr=5.7593e-04) (hash(x)=22405076)
582
+ 556 train 7.535743 (lr=5.7447e-04) (hash(x)=25673545)
583
+ 557 train 7.593813 (lr=5.7300e-04) (hash(x)=22675428)
584
+ 558 train 7.698732 (lr=5.7152e-04) (hash(x)=28168456)
585
+ 559 train 7.824698 (lr=5.7005e-04) (hash(x)=26072586)
586
+ 560 train 7.614642 (lr=5.6857e-04) (hash(x)=24974331)
587
+ 561 train 7.390826 (lr=5.6709e-04) (hash(x)=21951984)
588
+ 562 train 7.762730 (lr=5.6560e-04) (hash(x)=25403751)
589
+ 563 train 7.583069 (lr=5.6412e-04) (hash(x)=26396012)
590
+ 564 train 7.531212 (lr=5.6263e-04) (hash(x)=24827190)
591
+ 565 train 7.636532 (lr=5.6114e-04) (hash(x)=24847458)
592
+ 566 train 7.622351 (lr=5.5965e-04) (hash(x)=22159088)
593
+ 567 train 7.580211 (lr=5.5815e-04) (hash(x)=22140034)
594
+ 568 train 7.588968 (lr=5.5666e-04) (hash(x)=19044797)
595
+ 569 train 7.519558 (lr=5.5516e-04) (hash(x)=23932794)
596
+ 570 train 7.616195 (lr=5.5365e-04) (hash(x)=25256890)
597
+ 571 train 7.673284 (lr=5.5215e-04) (hash(x)=24161711)
598
+ 572 train 7.601187 (lr=5.5064e-04) (hash(x)=22036817)
599
+ 573 train 7.472616 (lr=5.4913e-04) (hash(x)=26663208)
600
+ 574 train 7.527707 (lr=5.4762e-04) (hash(x)=24072213)
601
+ 575 train 7.522589 (lr=5.4611e-04) (hash(x)=22766044)
602
+ 576 train 7.641778 (lr=5.4460e-04) (hash(x)=27871153)
603
+ 577 train 7.584463 (lr=5.4308e-04) (hash(x)=24037280)
604
+ 578 train 7.705688 (lr=5.4156e-04) (hash(x)=24121850)
605
+ 579 train 7.561292 (lr=5.4004e-04) (hash(x)=24380031)
606
+ 580 train 7.516613 (lr=5.3852e-04) (hash(x)=23055215)
607
+ 581 train 7.731693 (lr=5.3699e-04) (hash(x)=25329724)
608
+ 582 train 7.502999 (lr=5.3547e-04) (hash(x)=25662408)
609
+ 583 train 7.604637 (lr=5.3394e-04) (hash(x)=26346814)
610
+ 584 train 7.704065 (lr=5.3241e-04) (hash(x)=25283256)
611
+ 585 train 7.522517 (lr=5.3088e-04) (hash(x)=24021771)
612
+ 586 train 7.531789 (lr=5.2934e-04) (hash(x)=23576840)
613
+ 587 train 7.664512 (lr=5.2781e-04) (hash(x)=26414858)
614
+ 588 train 7.465583 (lr=5.2627e-04) (hash(x)=20099261)
615
+ 589 train 7.462104 (lr=5.2473e-04) (hash(x)=22453718)
616
+ 590 train 7.415373 (lr=5.2319e-04) (hash(x)=22529445)
617
+ 591 train 7.500037 (lr=5.2165e-04) (hash(x)=24484543)
618
+ 592 train 7.437870 (lr=5.2011e-04) (hash(x)=24043998)
619
+ 593 train 7.574516 (lr=5.1856e-04) (hash(x)=24301906)
620
+ 594 train 7.526821 (lr=5.1702e-04) (hash(x)=26328881)
621
+ 595 train 7.558410 (lr=5.1547e-04) (hash(x)=23101508)
622
+ 596 train 7.534391 (lr=5.1392e-04) (hash(x)=23527176)
623
+ 597 train 7.974708 (lr=5.1237e-04) (hash(x)=22705673)
624
+ 598 train 7.568842 (lr=5.1082e-04) (hash(x)=28664796)
625
+ 599 train 7.670704 (lr=5.0926e-04) (hash(x)=27102890)
626
+ 600 val loss 7.6226
627
+ 600 val perplexity 2043.8217
628
+ 600 train 7.751816 (lr=5.0771e-04) (hash(x)=26787259)
629
+ 601 train 7.785450 (lr=5.0615e-04) (hash(x)=23396088)
630
+ 602 train 7.535449 (lr=5.0459e-04) (hash(x)=24911831)
631
+ 603 train 7.554314 (lr=5.0303e-04) (hash(x)=23453491)
632
+ 604 train 7.599481 (lr=5.0147e-04) (hash(x)=23453396)
633
+ 605 train 7.690997 (lr=4.9991e-04) (hash(x)=23869612)
634
+ 606 train 7.758685 (lr=4.9835e-04) (hash(x)=26581590)
635
+ 607 train 7.594559 (lr=4.9679e-04) (hash(x)=28056342)
636
+ 608 train 7.638861 (lr=4.9522e-04) (hash(x)=25900866)
637
+ 609 train 7.676452 (lr=4.9366e-04) (hash(x)=24549177)
638
+ 610 train 7.608613 (lr=4.9209e-04) (hash(x)=28239583)
639
+ 611 train 7.612027 (lr=4.9052e-04) (hash(x)=23223554)
640
+ 612 train 7.677086 (lr=4.8895e-04) (hash(x)=24793480)
641
+ 613 train 7.608883 (lr=4.8738e-04) (hash(x)=25415570)
642
+ 614 train 7.648485 (lr=4.8581e-04) (hash(x)=24425760)
643
+ 615 train 7.734621 (lr=4.8424e-04) (hash(x)=23707134)
644
+ 616 train 7.663237 (lr=4.8267e-04) (hash(x)=28321697)
645
+ 617 train 7.608747 (lr=4.8109e-04) (hash(x)=23434031)
646
+ 618 train 7.594747 (lr=4.7952e-04) (hash(x)=27003481)
647
+ 619 train 7.553249 (lr=4.7795e-04) (hash(x)=22236893)
648
+ 620 train 7.612430 (lr=4.7637e-04) (hash(x)=26857458)
649
+ 621 train 7.606282 (lr=4.7479e-04) (hash(x)=26413122)
650
+ 622 train 7.854669 (lr=4.7322e-04) (hash(x)=27696537)
651
+ 623 train 7.633525 (lr=4.7164e-04) (hash(x)=21950234)
652
+ 624 train 7.656473 (lr=4.7006e-04) (hash(x)=24051952)
653
+ 625 train 7.650985 (lr=4.6848e-04) (hash(x)=24029939)
654
+ 626 train 7.742830 (lr=4.6690e-04) (hash(x)=24393346)
655
+ 627 train 7.711442 (lr=4.6532e-04) (hash(x)=26985352)
656
+ 628 train 7.441906 (lr=4.6374e-04) (hash(x)=19493265)
657
+ 629 train 7.541669 (lr=4.6216e-04) (hash(x)=23944010)
658
+ 630 train 7.740088 (lr=4.6058e-04) (hash(x)=21517435)
659
+ 631 train 7.509793 (lr=4.5900e-04) (hash(x)=25025868)
660
+ 632 train 7.543462 (lr=4.5742e-04) (hash(x)=22940997)
661
+ 633 train 7.676443 (lr=4.5583e-04) (hash(x)=25864840)
662
+ 634 train 7.578753 (lr=4.5425e-04) (hash(x)=24682151)
663
+ 635 train 7.516493 (lr=4.5267e-04) (hash(x)=27055330)
664
+ 636 train 7.567687 (lr=4.5109e-04) (hash(x)=22705710)
665
+ 637 train 7.651609 (lr=4.4950e-04) (hash(x)=26775238)
666
+ 638 train 7.751189 (lr=4.4792e-04) (hash(x)=21734331)
667
+ 639 train 7.547499 (lr=4.4634e-04) (hash(x)=25693442)
668
+ 640 train 7.615499 (lr=4.4475e-04) (hash(x)=25290454)
669
+ 641 train 7.587174 (lr=4.4317e-04) (hash(x)=25699860)
670
+ 642 train 7.748094 (lr=4.4158e-04) (hash(x)=19919115)
671
+ 643 train 7.799127 (lr=4.4000e-04) (hash(x)=25586967)
672
+ 644 train 7.585311 (lr=4.3842e-04) (hash(x)=25373506)
673
+ 645 train 7.593597 (lr=4.3683e-04) (hash(x)=26496148)
674
+ 646 train 7.651932 (lr=4.3525e-04) (hash(x)=27395083)
675
+ 647 train 7.822460 (lr=4.3366e-04) (hash(x)=27389709)
676
+ 648 train 7.719601 (lr=4.3208e-04) (hash(x)=29172968)
677
+ 649 train 7.657162 (lr=4.3050e-04) (hash(x)=25464920)
678
+ 650 val loss 7.6180
679
+ 650 val perplexity 2034.4338
680
+ 650 train 7.686028 (lr=4.2891e-04) (hash(x)=25992320)
681
+ 651 train 7.656298 (lr=4.2733e-04) (hash(x)=24949509)
682
+ 652 train 7.682298 (lr=4.2575e-04) (hash(x)=25615151)
683
+ 653 train 7.672678 (lr=4.2417e-04) (hash(x)=27923548)
684
+ 654 train 7.699358 (lr=4.2258e-04) (hash(x)=26387848)
685
+ 655 train 7.619361 (lr=4.2100e-04) (hash(x)=22863518)
686
+ 656 train 7.575901 (lr=4.1942e-04) (hash(x)=20869581)
687
+ 657 train 7.727091 (lr=4.1784e-04) (hash(x)=18133606)
688
+ 658 train 7.624393 (lr=4.1626e-04) (hash(x)=28736122)
689
+ 659 train 7.587476 (lr=4.1468e-04) (hash(x)=25528329)
690
+ 660 train 7.652987 (lr=4.1310e-04) (hash(x)=24020765)
691
+ 661 train 7.686389 (lr=4.1152e-04) (hash(x)=27792245)
692
+ 662 train 7.706572 (lr=4.0994e-04) (hash(x)=28287635)
693
+ 663 train 7.455310 (lr=4.0836e-04) (hash(x)=20951206)
694
+ 664 train 7.526803 (lr=4.0678e-04) (hash(x)=22356452)
695
+ 665 train 7.600543 (lr=4.0521e-04) (hash(x)=24385170)
696
+ 666 train 7.673573 (lr=4.0363e-04) (hash(x)=23455854)
697
+ 667 train 7.648390 (lr=4.0205e-04) (hash(x)=21589405)
698
+ 668 train 7.632666 (lr=4.0048e-04) (hash(x)=24698150)
699
+ 669 train 7.635265 (lr=3.9891e-04) (hash(x)=28010836)
700
+ 670 train 7.763627 (lr=3.9733e-04) (hash(x)=25427381)
701
+ 671 train 7.708281 (lr=3.9576e-04) (hash(x)=27478557)
702
+ 672 train 7.654237 (lr=3.9419e-04) (hash(x)=23865569)
703
+ 673 train 7.712451 (lr=3.9262e-04) (hash(x)=23407132)
704
+ 674 train 7.596667 (lr=3.9105e-04) (hash(x)=20253664)
705
+ 675 train 7.575116 (lr=3.8948e-04) (hash(x)=21087928)
706
+ 676 train 7.674585 (lr=3.8791e-04) (hash(x)=25957589)
707
+ 677 train 7.604998 (lr=3.8634e-04) (hash(x)=27008610)
708
+ 678 train 7.682994 (lr=3.8478e-04) (hash(x)=27580265)
709
+ 679 train 7.695330 (lr=3.8321e-04) (hash(x)=24244798)
710
+ 680 train 7.593230 (lr=3.8165e-04) (hash(x)=24535871)
711
+ 681 train 7.686321 (lr=3.8009e-04) (hash(x)=27558681)
712
+ 682 train 7.572512 (lr=3.7853e-04) (hash(x)=21152793)
713
+ 683 train 7.537177 (lr=3.7697e-04) (hash(x)=28147005)
714
+ 684 train 7.637247 (lr=3.7541e-04) (hash(x)=21933180)
715
+ 685 train 7.615443 (lr=3.7385e-04) (hash(x)=24735895)
716
+ 686 train 7.647284 (lr=3.7229e-04) (hash(x)=26209948)
717
+ 687 train 7.634003 (lr=3.7074e-04) (hash(x)=23358313)
718
+ 688 train 7.584732 (lr=3.6918e-04) (hash(x)=22398001)
719
+ 689 train 7.511104 (lr=3.6763e-04) (hash(x)=23724957)
720
+ 690 train 7.571434 (lr=3.6608e-04) (hash(x)=23083514)
721
+ 691 train 7.442668 (lr=3.6453e-04) (hash(x)=16016924)
722
+ 692 train 7.698529 (lr=3.6298e-04) (hash(x)=27173674)
723
+ 693 train 7.501496 (lr=3.6144e-04) (hash(x)=28056250)
724
+ 694 train 7.620018 (lr=3.5989e-04) (hash(x)=22233617)
725
+ 695 train 7.699221 (lr=3.5835e-04) (hash(x)=23777908)
726
+ 696 train 7.669135 (lr=3.5681e-04) (hash(x)=27842849)
727
+ 697 train 7.662814 (lr=3.5527e-04) (hash(x)=23711679)
728
+ 698 train 7.508156 (lr=3.5373e-04) (hash(x)=17204843)
729
+ 699 train 7.817569 (lr=3.5219e-04) (hash(x)=23909800)
730
+ 700 val loss 7.6042
731
+ 700 val perplexity 2006.7043
732
+ 700 train 7.657484 (lr=3.5066e-04) (hash(x)=25033286)
733
+ 701 train 7.923980 (lr=3.4912e-04) (hash(x)=23608371)
734
+ 702 train 7.728893 (lr=3.4759e-04) (hash(x)=23654625)
735
+ 703 train 7.583189 (lr=3.4606e-04) (hash(x)=22633313)
736
+ 704 train 7.628647 (lr=3.4453e-04) (hash(x)=24267011)
737
+ 705 train 7.586051 (lr=3.4301e-04) (hash(x)=23449416)
738
+ 706 train 7.802535 (lr=3.4148e-04) (hash(x)=28527906)
739
+ 707 train 7.519048 (lr=3.3996e-04) (hash(x)=21189075)
740
+ 708 train 7.549572 (lr=3.3844e-04) (hash(x)=25278645)
741
+ 709 train 8.036489 (lr=3.3692e-04) (hash(x)=23654290)
742
+ 710 train 7.574094 (lr=3.3540e-04) (hash(x)=24571139)
743
+ 711 train 7.648633 (lr=3.3389e-04) (hash(x)=26203172)
744
+ 712 train 7.594803 (lr=3.3238e-04) (hash(x)=19623503)
745
+ 713 train 7.593645 (lr=3.3087e-04) (hash(x)=23970545)
746
+ 714 train 7.749001 (lr=3.2936e-04) (hash(x)=25724900)
747
+ 715 train 7.676687 (lr=3.2785e-04) (hash(x)=23245421)
748
+ 716 train 7.639895 (lr=3.2635e-04) (hash(x)=24087537)
749
+ 717 train 7.556362 (lr=3.2484e-04) (hash(x)=25549646)
750
+ 718 train 7.621048 (lr=3.2334e-04) (hash(x)=26406563)
751
+ 719 train 7.683167 (lr=3.2185e-04) (hash(x)=23972429)
752
+ 720 train 7.716583 (lr=3.2035e-04) (hash(x)=26378040)
753
+ 721 train 7.590287 (lr=3.1886e-04) (hash(x)=24618531)
754
+ 722 train 7.617709 (lr=3.1737e-04) (hash(x)=20135714)
755
+ 723 train 7.720834 (lr=3.1588e-04) (hash(x)=26490981)
756
+ 724 train 7.738691 (lr=3.1440e-04) (hash(x)=23927567)
757
+ 725 train 7.606580 (lr=3.1291e-04) (hash(x)=27943609)
758
+ 726 train 7.645923 (lr=3.1143e-04) (hash(x)=22962561)
759
+ 727 train 7.667097 (lr=3.0995e-04) (hash(x)=24772257)
760
+ 728 train 7.621945 (lr=3.0848e-04) (hash(x)=25539761)
761
+ 729 train 7.741674 (lr=3.0700e-04) (hash(x)=23316262)
762
+ 730 train 7.557631 (lr=3.0553e-04) (hash(x)=23857558)
763
+ 731 train 7.724688 (lr=3.0407e-04) (hash(x)=30252690)
764
+ 732 train 7.762346 (lr=3.0260e-04) (hash(x)=30590162)
765
+ 733 train 7.671602 (lr=3.0114e-04) (hash(x)=25015721)
766
+ 734 train 7.726585 (lr=2.9968e-04) (hash(x)=28639679)
767
+ 735 train 7.643742 (lr=2.9822e-04) (hash(x)=26199058)
768
+ 736 train 7.689322 (lr=2.9677e-04) (hash(x)=28907461)
769
+ 737 train 7.669342 (lr=2.9531e-04) (hash(x)=26466915)
770
+ 738 train 7.738862 (lr=2.9386e-04) (hash(x)=20332613)
771
+ 739 train 7.632781 (lr=2.9242e-04) (hash(x)=26766470)
772
+ 740 train 7.601734 (lr=2.9097e-04) (hash(x)=26722145)
773
+ 741 train 7.635471 (lr=2.8953e-04) (hash(x)=21222489)
774
+ 742 train 7.765924 (lr=2.8810e-04) (hash(x)=24662435)
775
+ 743 train 7.587499 (lr=2.8666e-04) (hash(x)=24308927)
776
+ 744 train 7.550857 (lr=2.8523e-04) (hash(x)=28442502)
777
+ 745 train 7.601732 (lr=2.8380e-04) (hash(x)=22810435)
778
+ 746 train 7.704366 (lr=2.8238e-04) (hash(x)=30486518)
779
+ 747 train 7.683333 (lr=2.8095e-04) (hash(x)=22688712)
780
+ 748 train 7.619670 (lr=2.7953e-04) (hash(x)=20712962)
781
+ 749 train 7.514622 (lr=2.7812e-04) (hash(x)=27490086)
782
+ 750 val loss 7.6082
783
+ 750 val perplexity 2014.6813
784
+ 750 train 7.499670 (lr=2.7670e-04) (hash(x)=23973908)
785
+ 751 train 7.669379 (lr=2.7529e-04) (hash(x)=29132401)
786
+ 752 train 7.615711 (lr=2.7389e-04) (hash(x)=27386700)
787
+ 753 train 7.671254 (lr=2.7248e-04) (hash(x)=24268614)
788
+ 754 train 7.686059 (lr=2.7108e-04) (hash(x)=27378387)
789
+ 755 train 7.520214 (lr=2.6969e-04) (hash(x)=22529400)
790
+ 756 train 7.597469 (lr=2.6829e-04) (hash(x)=27586989)
791
+ 757 train 7.539778 (lr=2.6690e-04) (hash(x)=22336349)
792
+ 758 train 7.577286 (lr=2.6551e-04) (hash(x)=26147854)
793
+ 759 train 7.659555 (lr=2.6413e-04) (hash(x)=25346788)
794
+ 760 train 7.732388 (lr=2.6275e-04) (hash(x)=25956855)
795
+ 761 train 7.676356 (lr=2.6137e-04) (hash(x)=24367780)
796
+ 762 train 7.572974 (lr=2.6000e-04) (hash(x)=26084462)
797
+ 763 train 7.639929 (lr=2.5863e-04) (hash(x)=26702459)
798
+ 764 train 7.582914 (lr=2.5726e-04) (hash(x)=25717976)
799
+ 765 train 7.550413 (lr=2.5590e-04) (hash(x)=22141366)
800
+ 766 train 7.737177 (lr=2.5454e-04) (hash(x)=24652423)
801
+ 767 train 7.657237 (lr=2.5319e-04) (hash(x)=30911699)
802
+ 768 train 7.644424 (lr=2.5183e-04) (hash(x)=22076954)
803
+ 769 train 7.816966 (lr=2.5048e-04) (hash(x)=21586769)
804
+ 770 train 7.656290 (lr=2.4914e-04) (hash(x)=21994337)
805
+ 771 train 7.610101 (lr=2.4780e-04) (hash(x)=25454145)
806
+ 772 train 7.679800 (lr=2.4646e-04) (hash(x)=23239550)
807
+ 773 train 7.707846 (lr=2.4513e-04) (hash(x)=23565926)
808
+ 774 train 7.581442 (lr=2.4380e-04) (hash(x)=24705563)
809
+ 775 train 7.667150 (lr=2.4247e-04) (hash(x)=27671095)
810
+ 776 train 7.777799 (lr=2.4115e-04) (hash(x)=26254303)
811
+ 777 train 7.559124 (lr=2.3983e-04) (hash(x)=27454020)
812
+ 778 train 7.556102 (lr=2.3852e-04) (hash(x)=24394565)
813
+ 779 train 7.498204 (lr=2.3720e-04) (hash(x)=21780199)
814
+ 780 train 7.582950 (lr=2.3590e-04) (hash(x)=28296896)
815
+ 781 train 7.690670 (lr=2.3460e-04) (hash(x)=26766303)
816
+ 782 train 7.568882 (lr=2.3330e-04) (hash(x)=21599973)
817
+ 783 train 7.844289 (lr=2.3200e-04) (hash(x)=24969360)
818
+ 784 train 7.790802 (lr=2.3071e-04) (hash(x)=24382006)
819
+ 785 train 7.513860 (lr=2.2942e-04) (hash(x)=22009299)
820
+ 786 train 7.386002 (lr=2.2814e-04) (hash(x)=22376569)
821
+ 787 train 7.543223 (lr=2.2686e-04) (hash(x)=28166843)
822
+ 788 train 7.780671 (lr=2.2559e-04) (hash(x)=28230119)
823
+ 789 train 7.754766 (lr=2.2432e-04) (hash(x)=26207465)
824
+ 790 train 7.812477 (lr=2.2305e-04) (hash(x)=24748118)
825
+ 791 train 7.753173 (lr=2.2179e-04) (hash(x)=27033802)
826
+ 792 train 7.681068 (lr=2.2053e-04) (hash(x)=28788047)
827
+ 793 train 7.567931 (lr=2.1928e-04) (hash(x)=22185099)
828
+ 794 train 7.732545 (lr=2.1803e-04) (hash(x)=21585136)
829
+ 795 train 7.566338 (lr=2.1678e-04) (hash(x)=27691968)
830
+ 796 train 7.594959 (lr=2.1554e-04) (hash(x)=23706103)
831
+ 797 train 7.457735 (lr=2.1431e-04) (hash(x)=27918030)
832
+ 798 train 7.661190 (lr=2.1308e-04) (hash(x)=23127886)
833
+ 799 train 7.679038 (lr=2.1185e-04) (hash(x)=26197070)
834
+ 800 val loss 7.5987
835
+ 800 val perplexity 1995.6077
836
+ 800 train 7.721543 (lr=2.1063e-04) (hash(x)=26463005)
837
+ 801 train 7.623165 (lr=2.0941e-04) (hash(x)=19160134)
838
+ 802 train 7.574898 (lr=2.0819e-04) (hash(x)=23699136)
839
+ 803 train 7.641213 (lr=2.0698e-04) (hash(x)=25353051)
840
+ 804 train 7.568929 (lr=2.0578e-04) (hash(x)=21558484)
841
+ 805 train 7.501099 (lr=2.0458e-04) (hash(x)=22828380)
842
+ 806 train 7.615319 (lr=2.0338e-04) (hash(x)=24623811)
843
+ 807 train 7.767680 (lr=2.0219e-04) (hash(x)=25410299)
844
+ 808 train 7.648706 (lr=2.0100e-04) (hash(x)=26046091)
845
+ 809 train 7.767862 (lr=1.9982e-04) (hash(x)=29887482)
846
+ 810 train 7.517380 (lr=1.9864e-04) (hash(x)=21809706)
847
+ 811 train 7.630578 (lr=1.9747e-04) (hash(x)=23712872)
848
+ 812 train 7.571935 (lr=1.9630e-04) (hash(x)=25875103)
849
+ 813 train 7.513557 (lr=1.9514e-04) (hash(x)=23736441)
850
+ 814 train 7.598577 (lr=1.9398e-04) (hash(x)=25093883)
851
+ 815 train 7.549140 (lr=1.9283e-04) (hash(x)=23692325)
852
+ 816 train 7.427201 (lr=1.9168e-04) (hash(x)=23061082)
853
+ 817 train 7.601085 (lr=1.9053e-04) (hash(x)=26094253)
854
+ 818 train 7.711091 (lr=1.8939e-04) (hash(x)=27327886)
855
+ 819 train 7.574717 (lr=1.8826e-04) (hash(x)=26186143)
856
+ 820 train 7.572273 (lr=1.8713e-04) (hash(x)=25336043)
857
+ 821 train 7.634582 (lr=1.8600e-04) (hash(x)=22811468)
858
+ 822 train 7.606449 (lr=1.8488e-04) (hash(x)=27925831)
859
+ 823 train 7.695093 (lr=1.8377e-04) (hash(x)=35528108)
860
+ 824 train 7.695332 (lr=1.8266e-04) (hash(x)=24949421)
861
+ 825 train 7.481486 (lr=1.8155e-04) (hash(x)=22845735)
862
+ 826 train 7.587382 (lr=1.8045e-04) (hash(x)=28138281)
863
+ 827 train 7.649101 (lr=1.7936e-04) (hash(x)=29038360)
864
+ 828 train 7.596130 (lr=1.7827e-04) (hash(x)=22827903)
865
+ 829 train 7.677898 (lr=1.7718e-04) (hash(x)=24647015)
866
+ 830 train 7.568243 (lr=1.7610e-04) (hash(x)=20477804)
867
+ 831 train 7.550918 (lr=1.7503e-04) (hash(x)=26692640)
868
+ 832 train 7.614859 (lr=1.7396e-04) (hash(x)=27388794)
869
+ 833 train 7.846033 (lr=1.7289e-04) (hash(x)=25265875)
870
+ 834 train 7.544761 (lr=1.7183e-04) (hash(x)=21841888)
871
+ 835 train 7.619591 (lr=1.7078e-04) (hash(x)=19809196)
872
+ 836 train 7.693951 (lr=1.6973e-04) (hash(x)=23007945)
873
+ 837 train 7.634729 (lr=1.6869e-04) (hash(x)=24079409)
874
+ 838 train 7.535986 (lr=1.6765e-04) (hash(x)=25874533)
875
+ 839 train 7.521673 (lr=1.6661e-04) (hash(x)=23660106)
876
+ 840 train 7.928670 (lr=1.6559e-04) (hash(x)=41821329)
877
+ 841 train 7.586507 (lr=1.6456e-04) (hash(x)=25763899)
878
+ 842 train 7.432256 (lr=1.6355e-04) (hash(x)=28584111)
879
+ 843 train 7.860564 (lr=1.6253e-04) (hash(x)=25559887)
880
+ 844 train 7.687017 (lr=1.6153e-04) (hash(x)=26886377)
881
+ 845 train 7.646035 (lr=1.6053e-04) (hash(x)=26461895)
882
+ 846 train 7.681071 (lr=1.5953e-04) (hash(x)=23018827)
883
+ 847 train 7.508821 (lr=1.5854e-04) (hash(x)=24375035)
884
+ 848 train 7.638976 (lr=1.5756e-04) (hash(x)=26431670)
885
+ 849 train 7.539228 (lr=1.5658e-04) (hash(x)=25033511)
886
+ 850 val loss 7.6019
887
+ 850 val perplexity 2002.0574
888
+ 850 train 7.649432 (lr=1.5560e-04) (hash(x)=25602577)
889
+ 851 train 7.581991 (lr=1.5463e-04) (hash(x)=23316582)
890
+ 852 train 7.692994 (lr=1.5367e-04) (hash(x)=22470095)
891
+ 853 train 7.575059 (lr=1.5271e-04) (hash(x)=25089202)
892
+ 854 train 7.546660 (lr=1.5176e-04) (hash(x)=26243247)
893
+ 855 train 7.589221 (lr=1.5082e-04) (hash(x)=22231144)
894
+ 856 train 7.518424 (lr=1.4988e-04) (hash(x)=27646564)
895
+ 857 train 7.662960 (lr=1.4894e-04) (hash(x)=23889875)
896
+ 858 train 7.591718 (lr=1.4801e-04) (hash(x)=25255240)
897
+ 859 train 7.694792 (lr=1.4709e-04) (hash(x)=23883299)
898
+ 860 train 7.675210 (lr=1.4617e-04) (hash(x)=18463841)
899
+ 861 train 7.683481 (lr=1.4526e-04) (hash(x)=30657762)
900
+ 862 train 7.580599 (lr=1.4435e-04) (hash(x)=25288079)
901
+ 863 train 7.609830 (lr=1.4345e-04) (hash(x)=28979310)
902
+ 864 train 7.787637 (lr=1.4255e-04) (hash(x)=30254991)
903
+ 865 train 7.654351 (lr=1.4166e-04) (hash(x)=27274653)
904
+ 866 train 7.639599 (lr=1.4078e-04) (hash(x)=23026066)
905
+ 867 train 7.614757 (lr=1.3990e-04) (hash(x)=14001963)
906
+ 868 train 7.527188 (lr=1.3903e-04) (hash(x)=21988054)
907
+ 869 train 7.599045 (lr=1.3816e-04) (hash(x)=26241099)
908
+ 870 train 7.669240 (lr=1.3730e-04) (hash(x)=26209722)
909
+ 871 train 7.563887 (lr=1.3645e-04) (hash(x)=26928312)
910
+ 872 train 7.618113 (lr=1.3560e-04) (hash(x)=23563272)
911
+ 873 train 7.558492 (lr=1.3476e-04) (hash(x)=25704828)
912
+ 874 train 7.787706 (lr=1.3392e-04) (hash(x)=22498992)
913
+ 875 train 7.984493 (lr=1.3309e-04) (hash(x)=22895918)
914
+ 876 train 7.545657 (lr=1.3227e-04) (hash(x)=24404283)
915
+ 877 train 7.578594 (lr=1.3145e-04) (hash(x)=26663932)
916
+ 878 train 7.689528 (lr=1.3063e-04) (hash(x)=22805316)
917
+ 879 train 7.658391 (lr=1.2983e-04) (hash(x)=23344466)
918
+ 880 train 7.611262 (lr=1.2903e-04) (hash(x)=20966940)
919
+ 881 train 7.608794 (lr=1.2823e-04) (hash(x)=27422664)
920
+ 882 train 7.658308 (lr=1.2744e-04) (hash(x)=22116510)
921
+ 883 train 7.554638 (lr=1.2666e-04) (hash(x)=27269633)
922
+ 884 train 7.632807 (lr=1.2588e-04) (hash(x)=27157711)
923
+ 885 train 7.628246 (lr=1.2511e-04) (hash(x)=22774645)
924
+ 886 train 7.562450 (lr=1.2435e-04) (hash(x)=22143146)
925
+ 887 train 7.594866 (lr=1.2359e-04) (hash(x)=22356438)
926
+ 888 train 7.364458 (lr=1.2284e-04) (hash(x)=23172158)
927
+ 889 train 7.604044 (lr=1.2209e-04) (hash(x)=26971690)
928
+ 890 train 7.640532 (lr=1.2135e-04) (hash(x)=24851922)
929
+ 891 train 7.488728 (lr=1.2062e-04) (hash(x)=21017641)
930
+ 892 train 7.600739 (lr=1.1989e-04) (hash(x)=28186957)
931
+ 893 train 7.625137 (lr=1.1917e-04) (hash(x)=25661082)
932
+ 894 train 7.531492 (lr=1.1845e-04) (hash(x)=25462781)
933
+ 895 train 7.500123 (lr=1.1774e-04) (hash(x)=21774374)
934
+ 896 train 7.506494 (lr=1.1704e-04) (hash(x)=23657055)
935
+ 897 train 7.509987 (lr=1.1634e-04) (hash(x)=23604050)
936
+ 898 train 7.931087 (lr=1.1565e-04) (hash(x)=31534071)
937
+ 899 train 7.751813 (lr=1.1497e-04) (hash(x)=24183732)
938
+ 900 val loss 7.5948
939
+ 900 val perplexity 1987.7488
940
+ 900 train 7.611561 (lr=1.1429e-04) (hash(x)=23384280)
941
+ 901 train 7.495098 (lr=1.1362e-04) (hash(x)=21975928)
942
+ 902 train 7.494906 (lr=1.1295e-04) (hash(x)=25281674)
943
+ 903 train 7.477794 (lr=1.1229e-04) (hash(x)=22947712)
944
+ 904 train 7.567415 (lr=1.1164e-04) (hash(x)=24853977)
945
+ 905 train 7.592305 (lr=1.1099e-04) (hash(x)=23696470)
946
+ 906 train 7.587409 (lr=1.1036e-04) (hash(x)=27489711)
947
+ 907 train 7.528471 (lr=1.0972e-04) (hash(x)=28050163)
948
+ 908 train 7.436489 (lr=1.0909e-04) (hash(x)=26996807)
949
+ 909 train 7.785674 (lr=1.0847e-04) (hash(x)=28626495)
950
+ 910 train 7.522654 (lr=1.0786e-04) (hash(x)=24783242)
951
+ 911 train 7.558923 (lr=1.0725e-04) (hash(x)=25832479)
952
+ 912 train 7.412366 (lr=1.0665e-04) (hash(x)=23522137)
953
+ 913 train 7.698617 (lr=1.0606e-04) (hash(x)=26731973)
954
+ 914 train 7.748580 (lr=1.0547e-04) (hash(x)=26019142)
955
+ 915 train 7.565815 (lr=1.0489e-04) (hash(x)=25081877)
956
+ 916 train 7.653770 (lr=1.0431e-04) (hash(x)=23843571)
957
+ 917 train 7.679042 (lr=1.0374e-04) (hash(x)=26805739)
958
+ 918 train 7.595867 (lr=1.0318e-04) (hash(x)=27169927)
959
+ 919 train 7.616117 (lr=1.0262e-04) (hash(x)=26545491)
960
+ 920 train 7.619985 (lr=1.0207e-04) (hash(x)=25179497)
961
+ 921 train 7.583425 (lr=1.0153e-04) (hash(x)=27085003)
962
+ 922 train 7.587474 (lr=1.0099e-04) (hash(x)=24558202)
963
+ 923 train 7.732684 (lr=1.0046e-04) (hash(x)=25317684)
964
+ 924 train 7.610097 (lr=9.9941e-05) (hash(x)=27104868)
965
+ 925 train 7.651410 (lr=9.9425e-05) (hash(x)=24806274)
966
+ 926 train 7.680593 (lr=9.8915e-05) (hash(x)=24341937)
967
+ 927 train 7.494908 (lr=9.8411e-05) (hash(x)=22736558)
968
+ 928 train 7.524287 (lr=9.7915e-05) (hash(x)=23843744)
969
+ 929 train 7.625574 (lr=9.7424e-05) (hash(x)=27218468)
970
+ 930 train 7.585509 (lr=9.6941e-05) (hash(x)=26042053)
971
+ 931 train 7.584401 (lr=9.6464e-05) (hash(x)=25776082)
972
+ 932 train 7.493101 (lr=9.5994e-05) (hash(x)=21594447)
973
+ 933 train 7.531569 (lr=9.5530e-05) (hash(x)=21674469)
974
+ 934 train 7.496267 (lr=9.5073e-05) (hash(x)=26213966)
975
+ 935 train 7.489398 (lr=9.4623e-05) (hash(x)=23770702)
976
+ 936 train 7.569094 (lr=9.4180e-05) (hash(x)=27829878)
977
+ 937 train 7.685509 (lr=9.3743e-05) (hash(x)=29177278)
978
+ 938 train 7.624414 (lr=9.3313e-05) (hash(x)=26020089)
979
+ 939 train 7.528142 (lr=9.2889e-05) (hash(x)=24676475)
980
+ 940 train 7.589476 (lr=9.2473e-05) (hash(x)=24056174)
981
+ 941 train 7.593522 (lr=9.2063e-05) (hash(x)=21887514)
982
+ 942 train 7.608597 (lr=9.1659e-05) (hash(x)=24196100)
983
+ 943 train 7.567665 (lr=9.1263e-05) (hash(x)=25700093)
984
+ 944 train 7.680118 (lr=9.0873e-05) (hash(x)=24365834)
985
+ 945 train 7.513217 (lr=9.0490e-05) (hash(x)=25631786)
986
+ 946 train 7.627189 (lr=9.0114e-05) (hash(x)=24452345)
987
+ 947 train 7.385305 (lr=8.9744e-05) (hash(x)=21465330)
988
+ 948 train 7.513659 (lr=8.9382e-05) (hash(x)=20427493)
989
+ 949 train 7.535254 (lr=8.9026e-05) (hash(x)=25439720)
990
+ 950 val loss 7.5921
991
+ 950 val perplexity 1982.4773
992
+ 950 train 7.603315 (lr=8.8677e-05) (hash(x)=21652988)
993
+ 951 train 7.534760 (lr=8.8335e-05) (hash(x)=22546779)
994
+ 952 train 7.410166 (lr=8.7999e-05) (hash(x)=20855736)
995
+ 953 train 7.590824 (lr=8.7670e-05) (hash(x)=27725873)
996
+ 954 train 7.639692 (lr=8.7349e-05) (hash(x)=24582989)
997
+ 955 train 7.512088 (lr=8.7034e-05) (hash(x)=24822372)
998
+ 956 train 7.566226 (lr=8.6725e-05) (hash(x)=25624588)
999
+ 957 train 7.538709 (lr=8.6424e-05) (hash(x)=23566489)
1000
+ 958 train 7.713848 (lr=8.6130e-05) (hash(x)=28117387)
1001
+ 959 train 7.514665 (lr=8.5842e-05) (hash(x)=26874283)
1002
+ 960 train 7.490788 (lr=8.5561e-05) (hash(x)=28570123)
1003
+ 961 train 7.614540 (lr=8.5287e-05) (hash(x)=24807017)
1004
+ 962 train 7.536523 (lr=8.5020e-05) (hash(x)=25189993)
1005
+ 963 train 7.640121 (lr=8.4760e-05) (hash(x)=29202226)
1006
+ 964 train 7.784266 (lr=8.4507e-05) (hash(x)=26060713)
1007
+ 965 train 7.699479 (lr=8.4260e-05) (hash(x)=23053285)
1008
+ 966 train 7.650479 (lr=8.4021e-05) (hash(x)=21968481)
1009
+ 967 train 7.595428 (lr=8.3788e-05) (hash(x)=24492698)
1010
+ 968 train 7.543886 (lr=8.3563e-05) (hash(x)=25760806)
1011
+ 969 train 7.578426 (lr=8.3344e-05) (hash(x)=27417392)
1012
+ 970 train 7.561257 (lr=8.3132e-05) (hash(x)=26767105)
1013
+ 971 train 7.625906 (lr=8.2927e-05) (hash(x)=25117678)
1014
+ 972 train 7.527821 (lr=8.2729e-05) (hash(x)=24233758)
1015
+ 973 train 7.596088 (lr=8.2537e-05) (hash(x)=29907634)
1016
+ 974 train 7.413289 (lr=8.2353e-05) (hash(x)=23552436)
1017
+ 975 train 7.704455 (lr=8.2176e-05) (hash(x)=28610406)
1018
+ 976 train 7.580215 (lr=8.2005e-05) (hash(x)=23858212)
1019
+ 977 train 7.461874 (lr=8.1842e-05) (hash(x)=20567764)
1020
+ 978 train 7.489018 (lr=8.1685e-05) (hash(x)=29631106)
1021
+ 979 train 7.456705 (lr=8.1536e-05) (hash(x)=19843564)
1022
+ 980 train 7.584284 (lr=8.1393e-05) (hash(x)=22589891)
1023
+ 981 train 7.365616 (lr=8.1257e-05) (hash(x)=21183943)
1024
+ 982 train 7.546909 (lr=8.1128e-05) (hash(x)=26152241)
1025
+ 983 train 7.545840 (lr=8.1007e-05) (hash(x)=22700785)
1026
+ 984 train 7.746047 (lr=8.0892e-05) (hash(x)=29611280)
1027
+ 985 train 7.853671 (lr=8.0784e-05) (hash(x)=25972823)
1028
+ 986 train 7.572216 (lr=8.0683e-05) (hash(x)=23447782)
1029
+ 987 train 7.552329 (lr=8.0589e-05) (hash(x)=22578453)
1030
+ 988 train 7.532523 (lr=8.0502e-05) (hash(x)=25474622)
1031
+ 989 train 7.601736 (lr=8.0422e-05) (hash(x)=22965762)
1032
+ 990 train 7.610272 (lr=8.0348e-05) (hash(x)=26116321)
1033
+ 991 train 7.537670 (lr=8.0282e-05) (hash(x)=25934549)
1034
+ 992 train 7.467483 (lr=8.0223e-05) (hash(x)=23937495)
1035
+ 993 train 7.606417 (lr=8.0171e-05) (hash(x)=26286442)
1036
+ 994 train 7.847108 (lr=8.0125e-05) (hash(x)=32737711)
1037
+ 995 train 7.494240 (lr=8.0087e-05) (hash(x)=20563613)
1038
+ 996 train 7.708313 (lr=8.0056e-05) (hash(x)=26234770)
1039
+ 997 train 7.643804 (lr=8.0031e-05) (hash(x)=26179422)
1040
+ 998 train 7.587658 (lr=8.0014e-05) (hash(x)=23289651)
1041
+ 999 val loss 7.5904
1042
+ 999 val perplexity 1979.1857
1043
+ 999 train 7.919212 (lr=8.0003e-05) (hash(x)=36436617)
lr8e-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:aec0b9a378f0c0c1bb7724e55be4f64c01ecc79381a0080fa1a051833a7e3a3b
3
+ size 48156226
lr8e-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 sha256:49e6383e92f6d79fd32f961893878fb9e439cbc0c3c9f6f3d89a7465bc619818
3
+ size 71157574