andrew-healey commited on
Commit
00b98b6
·
verified ·
1 Parent(s): d42c17d

Upload folder using huggingface_hub

Browse files
lr4e-4_total_batch_size40960_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr4e-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": 0.0004, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr4e-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
lr4e-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=1.3986e-06) (hash(x)=20782690)
5
+ 1 train 11.784650 (lr=2.7972e-06) (hash(x)=23890908)
6
+ 2 train 11.771741 (lr=4.1958e-06) (hash(x)=24611628)
7
+ 3 train 11.746619 (lr=5.5944e-06) (hash(x)=25767553)
8
+ 4 train 11.734269 (lr=6.9930e-06) (hash(x)=22512166)
9
+ 5 train 11.724540 (lr=8.3916e-06) (hash(x)=24732833)
10
+ 6 train 11.717943 (lr=9.7902e-06) (hash(x)=25129410)
11
+ 7 train 11.689986 (lr=1.1189e-05) (hash(x)=26582391)
12
+ 8 train 11.662333 (lr=1.2587e-05) (hash(x)=26472336)
13
+ 9 train 11.648644 (lr=1.3986e-05) (hash(x)=27654289)
14
+ 10 train 11.616327 (lr=1.5385e-05) (hash(x)=26559876)
15
+ 11 train 11.598208 (lr=1.6783e-05) (hash(x)=21770329)
16
+ 12 train 11.559298 (lr=1.8182e-05) (hash(x)=22723459)
17
+ 13 train 11.501315 (lr=1.9580e-05) (hash(x)=30338342)
18
+ 14 train 11.474544 (lr=2.0979e-05) (hash(x)=25678059)
19
+ 15 train 11.434258 (lr=2.2378e-05) (hash(x)=26109335)
20
+ 16 train 11.388500 (lr=2.3776e-05) (hash(x)=25044885)
21
+ 17 train 11.349714 (lr=2.5175e-05) (hash(x)=28870690)
22
+ 18 train 11.323692 (lr=2.6573e-05) (hash(x)=35279941)
23
+ 19 train 11.241452 (lr=2.7972e-05) (hash(x)=25015923)
24
+ 20 train 11.172580 (lr=2.9371e-05) (hash(x)=26740855)
25
+ 21 train 11.096758 (lr=3.0769e-05) (hash(x)=22262151)
26
+ 22 train 11.024281 (lr=3.2168e-05) (hash(x)=26621834)
27
+ 23 train 10.969557 (lr=3.3566e-05) (hash(x)=23234741)
28
+ 24 train 10.965103 (lr=3.4965e-05) (hash(x)=25470563)
29
+ 25 train 10.878447 (lr=3.6364e-05) (hash(x)=21799102)
30
+ 26 train 10.807779 (lr=3.7762e-05) (hash(x)=23275479)
31
+ 27 train 10.795661 (lr=3.9161e-05) (hash(x)=22091862)
32
+ 28 train 10.675905 (lr=4.0559e-05) (hash(x)=17430373)
33
+ 29 train 10.606292 (lr=4.1958e-05) (hash(x)=22954861)
34
+ 30 train 10.644531 (lr=4.3357e-05) (hash(x)=26671799)
35
+ 31 train 10.591021 (lr=4.4755e-05) (hash(x)=26924723)
36
+ 32 train 10.579751 (lr=4.6154e-05) (hash(x)=23557495)
37
+ 33 train 10.422199 (lr=4.7552e-05) (hash(x)=27804380)
38
+ 34 train 10.544160 (lr=4.8951e-05) (hash(x)=26654908)
39
+ 35 train 10.473736 (lr=5.0350e-05) (hash(x)=23627518)
40
+ 36 train 10.448035 (lr=5.1748e-05) (hash(x)=21889990)
41
+ 37 train 10.469918 (lr=5.3147e-05) (hash(x)=22528001)
42
+ 38 train 10.469602 (lr=5.4545e-05) (hash(x)=29248942)
43
+ 39 train 10.397206 (lr=5.5944e-05) (hash(x)=25533417)
44
+ 40 train 10.347758 (lr=5.7343e-05) (hash(x)=20612533)
45
+ 41 train 10.362356 (lr=5.8741e-05) (hash(x)=24677740)
46
+ 42 train 10.375150 (lr=6.0140e-05) (hash(x)=25347203)
47
+ 43 train 10.332895 (lr=6.1538e-05) (hash(x)=28922363)
48
+ 44 train 10.326017 (lr=6.2937e-05) (hash(x)=25631397)
49
+ 45 train 10.314990 (lr=6.4336e-05) (hash(x)=26237983)
50
+ 46 train 10.327802 (lr=6.5734e-05) (hash(x)=24742645)
51
+ 47 train 10.346034 (lr=6.7133e-05) (hash(x)=26157660)
52
+ 48 train 10.235748 (lr=6.8531e-05) (hash(x)=22582169)
53
+ 49 train 10.259607 (lr=6.9930e-05) (hash(x)=26431960)
54
+ 50 val loss 10.3031
55
+ 50 val perplexity 29824.1816
56
+ 50 train 10.328391 (lr=7.1329e-05) (hash(x)=25716176)
57
+ 51 train 10.299902 (lr=7.2727e-05) (hash(x)=28019788)
58
+ 52 train 10.277354 (lr=7.4126e-05) (hash(x)=20382963)
59
+ 53 train 10.230413 (lr=7.5524e-05) (hash(x)=24489647)
60
+ 54 train 10.229125 (lr=7.6923e-05) (hash(x)=25128502)
61
+ 55 train 10.226540 (lr=7.8322e-05) (hash(x)=25231390)
62
+ 56 train 10.139033 (lr=7.9720e-05) (hash(x)=22708977)
63
+ 57 train 10.233564 (lr=8.1119e-05) (hash(x)=24364920)
64
+ 58 train 10.201653 (lr=8.2517e-05) (hash(x)=24932925)
65
+ 59 train 10.171653 (lr=8.3916e-05) (hash(x)=26913684)
66
+ 60 train 10.092138 (lr=8.5315e-05) (hash(x)=18017792)
67
+ 61 train 10.121547 (lr=8.6713e-05) (hash(x)=28613882)
68
+ 62 train 10.120727 (lr=8.8112e-05) (hash(x)=22652243)
69
+ 63 train 10.045502 (lr=8.9510e-05) (hash(x)=27194521)
70
+ 64 train 10.040846 (lr=9.0909e-05) (hash(x)=24936836)
71
+ 65 train 10.057040 (lr=9.2308e-05) (hash(x)=24765578)
72
+ 66 train 10.087700 (lr=9.3706e-05) (hash(x)=23107416)
73
+ 67 train 9.953025 (lr=9.5105e-05) (hash(x)=29314255)
74
+ 68 train 9.935004 (lr=9.6503e-05) (hash(x)=24013079)
75
+ 69 train 9.899267 (lr=9.7902e-05) (hash(x)=25322001)
76
+ 70 train 9.994378 (lr=9.9301e-05) (hash(x)=26690440)
77
+ 71 train 9.910471 (lr=1.0070e-04) (hash(x)=22598512)
78
+ 72 train 9.863027 (lr=1.0210e-04) (hash(x)=18473911)
79
+ 73 train 9.912343 (lr=1.0350e-04) (hash(x)=22322442)
80
+ 74 train 9.874385 (lr=1.0490e-04) (hash(x)=26324153)
81
+ 75 train 9.812431 (lr=1.0629e-04) (hash(x)=23446058)
82
+ 76 train 9.851996 (lr=1.0769e-04) (hash(x)=24843741)
83
+ 77 train 9.788888 (lr=1.0909e-04) (hash(x)=23197102)
84
+ 78 train 9.861073 (lr=1.1049e-04) (hash(x)=23054940)
85
+ 79 train 9.765963 (lr=1.1189e-04) (hash(x)=27140757)
86
+ 80 train 9.615582 (lr=1.1329e-04) (hash(x)=25149419)
87
+ 81 train 9.679656 (lr=1.1469e-04) (hash(x)=24727764)
88
+ 82 train 9.494093 (lr=1.1608e-04) (hash(x)=21935931)
89
+ 83 train 9.577534 (lr=1.1748e-04) (hash(x)=27606073)
90
+ 84 train 9.600612 (lr=1.1888e-04) (hash(x)=22450341)
91
+ 85 train 9.602433 (lr=1.2028e-04) (hash(x)=25945859)
92
+ 86 train 9.491938 (lr=1.2168e-04) (hash(x)=21613707)
93
+ 87 train 9.488298 (lr=1.2308e-04) (hash(x)=24396519)
94
+ 88 train 9.467055 (lr=1.2448e-04) (hash(x)=29138828)
95
+ 89 train 9.508711 (lr=1.2587e-04) (hash(x)=24685515)
96
+ 90 train 9.369195 (lr=1.2727e-04) (hash(x)=22825749)
97
+ 91 train 9.386017 (lr=1.2867e-04) (hash(x)=28767869)
98
+ 92 train 9.345555 (lr=1.3007e-04) (hash(x)=21159060)
99
+ 93 train 9.291492 (lr=1.3147e-04) (hash(x)=25045397)
100
+ 94 train 9.377343 (lr=1.3287e-04) (hash(x)=25560928)
101
+ 95 train 9.307971 (lr=1.3427e-04) (hash(x)=28959222)
102
+ 96 train 9.208468 (lr=1.3566e-04) (hash(x)=21248405)
103
+ 97 train 9.189060 (lr=1.3706e-04) (hash(x)=25830182)
104
+ 98 train 9.097737 (lr=1.3846e-04) (hash(x)=25173113)
105
+ 99 train 9.125483 (lr=1.3986e-04) (hash(x)=23486277)
106
+ 100 val loss 9.1011
107
+ 100 val perplexity 8964.7207
108
+ 100 train 9.080393 (lr=1.4126e-04) (hash(x)=28916006)
109
+ 101 train 9.155622 (lr=1.4266e-04) (hash(x)=25556929)
110
+ 102 train 9.061739 (lr=1.4406e-04) (hash(x)=29648798)
111
+ 103 train 8.991920 (lr=1.4545e-04) (hash(x)=24976217)
112
+ 104 train 9.002022 (lr=1.4685e-04) (hash(x)=31019606)
113
+ 105 train 8.986533 (lr=1.4825e-04) (hash(x)=28269421)
114
+ 106 train 8.896849 (lr=1.4965e-04) (hash(x)=24865358)
115
+ 107 train 8.798367 (lr=1.5105e-04) (hash(x)=26984272)
116
+ 108 train 8.883488 (lr=1.5245e-04) (hash(x)=24544116)
117
+ 109 train 8.836961 (lr=1.5385e-04) (hash(x)=22059850)
118
+ 110 train 8.812685 (lr=1.5524e-04) (hash(x)=23926632)
119
+ 111 train 8.739335 (lr=1.5664e-04) (hash(x)=27025333)
120
+ 112 train 8.699478 (lr=1.5804e-04) (hash(x)=29120407)
121
+ 113 train 8.754530 (lr=1.5944e-04) (hash(x)=23827393)
122
+ 114 train 8.646648 (lr=1.6084e-04) (hash(x)=27230027)
123
+ 115 train 8.630233 (lr=1.6224e-04) (hash(x)=24968260)
124
+ 116 train 8.624271 (lr=1.6364e-04) (hash(x)=27544803)
125
+ 117 train 8.596155 (lr=1.6503e-04) (hash(x)=26540663)
126
+ 118 train 8.623992 (lr=1.6643e-04) (hash(x)=25856625)
127
+ 119 train 8.535542 (lr=1.6783e-04) (hash(x)=23164356)
128
+ 120 train 8.552845 (lr=1.6923e-04) (hash(x)=25646282)
129
+ 121 train 8.540582 (lr=1.7063e-04) (hash(x)=23843233)
130
+ 122 train 8.384264 (lr=1.7203e-04) (hash(x)=24506028)
131
+ 123 train 8.429166 (lr=1.7343e-04) (hash(x)=22781277)
132
+ 124 train 8.436218 (lr=1.7483e-04) (hash(x)=24948650)
133
+ 125 train 8.383878 (lr=1.7622e-04) (hash(x)=26373991)
134
+ 126 train 8.429679 (lr=1.7762e-04) (hash(x)=23402396)
135
+ 127 train 8.373508 (lr=1.7902e-04) (hash(x)=21562133)
136
+ 128 train 8.406108 (lr=1.8042e-04) (hash(x)=25009210)
137
+ 129 train 8.281431 (lr=1.8182e-04) (hash(x)=27161811)
138
+ 130 train 8.270213 (lr=1.8322e-04) (hash(x)=29515100)
139
+ 131 train 8.234408 (lr=1.8462e-04) (hash(x)=25305929)
140
+ 132 train 8.228130 (lr=1.8601e-04) (hash(x)=26808543)
141
+ 133 train 8.147568 (lr=1.8741e-04) (hash(x)=25669509)
142
+ 134 train 8.203366 (lr=1.8881e-04) (hash(x)=28435805)
143
+ 135 train 8.177262 (lr=1.9021e-04) (hash(x)=26482758)
144
+ 136 train 8.096091 (lr=1.9161e-04) (hash(x)=24929929)
145
+ 137 train 8.094942 (lr=1.9301e-04) (hash(x)=25043037)
146
+ 138 train 8.222082 (lr=1.9441e-04) (hash(x)=29151257)
147
+ 139 train 8.040565 (lr=1.9580e-04) (hash(x)=26400365)
148
+ 140 train 8.019926 (lr=1.9720e-04) (hash(x)=24098812)
149
+ 141 train 8.013710 (lr=1.9860e-04) (hash(x)=20967666)
150
+ 142 train 8.129231 (lr=2.0000e-04) (hash(x)=28667963)
151
+ 143 train 8.055783 (lr=2.0140e-04) (hash(x)=27083417)
152
+ 144 train 7.940279 (lr=2.0280e-04) (hash(x)=23649001)
153
+ 145 train 7.957363 (lr=2.0420e-04) (hash(x)=31338300)
154
+ 146 train 7.929686 (lr=2.0559e-04) (hash(x)=30603174)
155
+ 147 train 7.939366 (lr=2.0699e-04) (hash(x)=24830965)
156
+ 148 train 8.029182 (lr=2.0839e-04) (hash(x)=27269893)
157
+ 149 train 7.869070 (lr=2.0979e-04) (hash(x)=24825041)
158
+ 150 val loss 7.8941
159
+ 150 val perplexity 2681.3186
160
+ 150 train 7.846231 (lr=2.1119e-04) (hash(x)=22724871)
161
+ 151 train 7.781201 (lr=2.1259e-04) (hash(x)=24683717)
162
+ 152 train 7.902162 (lr=2.1399e-04) (hash(x)=26190337)
163
+ 153 train 7.866138 (lr=2.1538e-04) (hash(x)=23086289)
164
+ 154 train 7.895522 (lr=2.1678e-04) (hash(x)=23593875)
165
+ 155 train 7.736624 (lr=2.1818e-04) (hash(x)=22414190)
166
+ 156 train 7.797123 (lr=2.1958e-04) (hash(x)=26758114)
167
+ 157 train 7.737235 (lr=2.2098e-04) (hash(x)=21563184)
168
+ 158 train 7.786695 (lr=2.2238e-04) (hash(x)=19218470)
169
+ 159 train 7.866757 (lr=2.2378e-04) (hash(x)=25588469)
170
+ 160 train 7.824136 (lr=2.2517e-04) (hash(x)=30433767)
171
+ 161 train 7.880255 (lr=2.2657e-04) (hash(x)=17851621)
172
+ 162 train 7.935395 (lr=2.2797e-04) (hash(x)=25183690)
173
+ 163 train 7.992916 (lr=2.2937e-04) (hash(x)=27040642)
174
+ 164 train 7.736604 (lr=2.3077e-04) (hash(x)=22183303)
175
+ 165 train 8.036337 (lr=2.3217e-04) (hash(x)=27537063)
176
+ 166 train 7.746436 (lr=2.3357e-04) (hash(x)=19492832)
177
+ 167 train 7.717609 (lr=2.3497e-04) (hash(x)=25951629)
178
+ 168 train 7.754804 (lr=2.3636e-04) (hash(x)=22408682)
179
+ 169 train 7.726074 (lr=2.3776e-04) (hash(x)=27633457)
180
+ 170 train 7.727928 (lr=2.3916e-04) (hash(x)=27635080)
181
+ 171 train 7.720882 (lr=2.4056e-04) (hash(x)=22287596)
182
+ 172 train 7.839679 (lr=2.4196e-04) (hash(x)=35895440)
183
+ 173 train 7.740313 (lr=2.4336e-04) (hash(x)=25888847)
184
+ 174 train 7.780423 (lr=2.4476e-04) (hash(x)=26518800)
185
+ 175 train 7.747272 (lr=2.4615e-04) (hash(x)=37519283)
186
+ 176 train 7.904023 (lr=2.4755e-04) (hash(x)=25141945)
187
+ 177 train 7.913914 (lr=2.4895e-04) (hash(x)=28527939)
188
+ 178 train 7.712204 (lr=2.5035e-04) (hash(x)=21942471)
189
+ 179 train 7.768369 (lr=2.5175e-04) (hash(x)=24955630)
190
+ 180 train 7.707573 (lr=2.5315e-04) (hash(x)=25541230)
191
+ 181 train 7.569569 (lr=2.5455e-04) (hash(x)=23501481)
192
+ 182 train 7.623980 (lr=2.5594e-04) (hash(x)=22754988)
193
+ 183 train 7.520475 (lr=2.5734e-04) (hash(x)=23475070)
194
+ 184 train 7.670856 (lr=2.5874e-04) (hash(x)=23609959)
195
+ 185 train 7.563438 (lr=2.6014e-04) (hash(x)=23441847)
196
+ 186 train 7.634137 (lr=2.6154e-04) (hash(x)=23096292)
197
+ 187 train 7.686985 (lr=2.6294e-04) (hash(x)=26171093)
198
+ 188 train 7.684378 (lr=2.6434e-04) (hash(x)=20060838)
199
+ 189 train 7.522353 (lr=2.6573e-04) (hash(x)=25451055)
200
+ 190 train 7.371145 (lr=2.6713e-04) (hash(x)=24934792)
201
+ 191 train 7.505332 (lr=2.6853e-04) (hash(x)=25367597)
202
+ 192 train 7.659813 (lr=2.6993e-04) (hash(x)=25022148)
203
+ 193 train 7.656339 (lr=2.7133e-04) (hash(x)=25431365)
204
+ 194 train 7.670822 (lr=2.7273e-04) (hash(x)=29307788)
205
+ 195 train 7.539510 (lr=2.7413e-04) (hash(x)=25651929)
206
+ 196 train 7.626486 (lr=2.7552e-04) (hash(x)=26312387)
207
+ 197 train 7.591259 (lr=2.7692e-04) (hash(x)=23804913)
208
+ 198 train 7.641756 (lr=2.7832e-04) (hash(x)=26115519)
209
+ 199 train 7.618354 (lr=2.7972e-04) (hash(x)=23210747)
210
+ 200 val loss 7.6250
211
+ 200 val perplexity 2048.6965
212
+ 200 train 7.578430 (lr=2.8112e-04) (hash(x)=18463712)
213
+ 201 train 7.597409 (lr=2.8252e-04) (hash(x)=23245699)
214
+ 202 train 7.535478 (lr=2.8392e-04) (hash(x)=21770300)
215
+ 203 train 7.809773 (lr=2.8531e-04) (hash(x)=25387532)
216
+ 204 train 7.618398 (lr=2.8671e-04) (hash(x)=27718703)
217
+ 205 train 7.669211 (lr=2.8811e-04) (hash(x)=27661052)
218
+ 206 train 7.538675 (lr=2.8951e-04) (hash(x)=26538928)
219
+ 207 train 7.654949 (lr=2.9091e-04) (hash(x)=21772654)
220
+ 208 train 7.747488 (lr=2.9231e-04) (hash(x)=26263254)
221
+ 209 train 7.591910 (lr=2.9371e-04) (hash(x)=23552356)
222
+ 210 train 7.795640 (lr=2.9510e-04) (hash(x)=22707141)
223
+ 211 train 7.839300 (lr=2.9650e-04) (hash(x)=34234481)
224
+ 212 train 7.661935 (lr=2.9790e-04) (hash(x)=35008670)
225
+ 213 train 7.770103 (lr=2.9930e-04) (hash(x)=25891543)
226
+ 214 train 7.540637 (lr=3.0070e-04) (hash(x)=21971224)
227
+ 215 train 7.652177 (lr=3.0210e-04) (hash(x)=27312186)
228
+ 216 train 7.749994 (lr=3.0350e-04) (hash(x)=22163749)
229
+ 217 train 7.522154 (lr=3.0490e-04) (hash(x)=25942888)
230
+ 218 train 7.622047 (lr=3.0629e-04) (hash(x)=25493916)
231
+ 219 train 7.809395 (lr=3.0769e-04) (hash(x)=24289951)
232
+ 220 train 7.612022 (lr=3.0909e-04) (hash(x)=23648671)
233
+ 221 train 7.457561 (lr=3.1049e-04) (hash(x)=24992055)
234
+ 222 train 7.732053 (lr=3.1189e-04) (hash(x)=26005549)
235
+ 223 train 7.586495 (lr=3.1329e-04) (hash(x)=21046554)
236
+ 224 train 7.586429 (lr=3.1469e-04) (hash(x)=26321539)
237
+ 225 train 7.597862 (lr=3.1608e-04) (hash(x)=27807872)
238
+ 226 train 7.597433 (lr=3.1748e-04) (hash(x)=21555364)
239
+ 227 train 7.571504 (lr=3.1888e-04) (hash(x)=26325324)
240
+ 228 train 7.609000 (lr=3.2028e-04) (hash(x)=24956294)
241
+ 229 train 7.557729 (lr=3.2168e-04) (hash(x)=24429389)
242
+ 230 train 7.652004 (lr=3.2308e-04) (hash(x)=27929846)
243
+ 231 train 7.614391 (lr=3.2448e-04) (hash(x)=25282897)
244
+ 232 train 7.462361 (lr=3.2587e-04) (hash(x)=26802508)
245
+ 233 train 7.649542 (lr=3.2727e-04) (hash(x)=28390416)
246
+ 234 train 7.777160 (lr=3.2867e-04) (hash(x)=26292451)
247
+ 235 train 7.684589 (lr=3.3007e-04) (hash(x)=22915344)
248
+ 236 train 7.601676 (lr=3.3147e-04) (hash(x)=24744383)
249
+ 237 train 7.512599 (lr=3.3287e-04) (hash(x)=26284678)
250
+ 238 train 7.483334 (lr=3.3427e-04) (hash(x)=21654269)
251
+ 239 train 7.950963 (lr=3.3566e-04) (hash(x)=29466268)
252
+ 240 train 8.215373 (lr=3.3706e-04) (hash(x)=32616762)
253
+ 241 train 8.196033 (lr=3.3846e-04) (hash(x)=32201138)
254
+ 242 train 8.250498 (lr=3.3986e-04) (hash(x)=33683468)
255
+ 243 train 7.879391 (lr=3.4126e-04) (hash(x)=27498125)
256
+ 244 train 7.704616 (lr=3.4266e-04) (hash(x)=25164318)
257
+ 245 train 7.547084 (lr=3.4406e-04) (hash(x)=25690221)
258
+ 246 train 7.570585 (lr=3.4545e-04) (hash(x)=29031550)
259
+ 247 train 7.672284 (lr=3.4685e-04) (hash(x)=24094093)
260
+ 248 train 8.330826 (lr=3.4825e-04) (hash(x)=61455960)
261
+ 249 train 11.715074 (lr=3.4965e-04) (hash(x)=61104710)
262
+ 250 val loss 7.6135
263
+ 250 val perplexity 2025.4358
264
+ 250 train 7.767569 (lr=3.5105e-04) (hash(x)=20881126)
265
+ 251 train 7.649746 (lr=3.5245e-04) (hash(x)=26920243)
266
+ 252 train 7.518490 (lr=3.5385e-04) (hash(x)=21970343)
267
+ 253 train 7.597096 (lr=3.5524e-04) (hash(x)=31676352)
268
+ 254 train 7.573557 (lr=3.5664e-04) (hash(x)=25600562)
269
+ 255 train 7.548319 (lr=3.5804e-04) (hash(x)=22022924)
270
+ 256 train 7.481236 (lr=3.5944e-04) (hash(x)=22777085)
271
+ 257 train 7.621839 (lr=3.6084e-04) (hash(x)=21727744)
272
+ 258 train 7.650042 (lr=3.6224e-04) (hash(x)=20701998)
273
+ 259 train 7.666699 (lr=3.6364e-04) (hash(x)=20863982)
274
+ 260 train 7.556705 (lr=3.6503e-04) (hash(x)=25084926)
275
+ 261 train 7.557322 (lr=3.6643e-04) (hash(x)=24412897)
276
+ 262 train 7.546864 (lr=3.6783e-04) (hash(x)=25434913)
277
+ 263 train 7.579038 (lr=3.6923e-04) (hash(x)=26302544)
278
+ 264 train 7.665436 (lr=3.7063e-04) (hash(x)=34976197)
279
+ 265 train 7.548098 (lr=3.7203e-04) (hash(x)=23870805)
280
+ 266 train 7.536252 (lr=3.7343e-04) (hash(x)=25041103)
281
+ 267 train 7.560432 (lr=3.7483e-04) (hash(x)=26041745)
282
+ 268 train 7.526058 (lr=3.7622e-04) (hash(x)=24068619)
283
+ 269 train 7.567034 (lr=3.7762e-04) (hash(x)=25441262)
284
+ 270 train 7.638949 (lr=3.7902e-04) (hash(x)=29181867)
285
+ 271 train 7.610388 (lr=3.8042e-04) (hash(x)=27093757)
286
+ 272 train 7.498973 (lr=3.8182e-04) (hash(x)=25150275)
287
+ 273 train 7.642481 (lr=3.8322e-04) (hash(x)=33457102)
288
+ 274 train 7.475198 (lr=3.8462e-04) (hash(x)=25755322)
289
+ 275 train 7.523257 (lr=3.8601e-04) (hash(x)=23630794)
290
+ 276 train 7.402236 (lr=3.8741e-04) (hash(x)=25167987)
291
+ 277 train 7.558375 (lr=3.8881e-04) (hash(x)=24043224)
292
+ 278 train 7.547144 (lr=3.9021e-04) (hash(x)=24665411)
293
+ 279 train 7.506107 (lr=3.9161e-04) (hash(x)=17321657)
294
+ 280 train 7.498788 (lr=3.9301e-04) (hash(x)=27932591)
295
+ 281 train 7.819557 (lr=3.9441e-04) (hash(x)=29447954)
296
+ 282 train 7.528435 (lr=3.9580e-04) (hash(x)=24621816)
297
+ 283 train 7.510150 (lr=3.9720e-04) (hash(x)=24682294)
298
+ 284 train 7.510372 (lr=3.9860e-04) (hash(x)=28057095)
299
+ 285 train 7.547693 (lr=4.0000e-04) (hash(x)=24547536)
300
+ 286 train 7.357195 (lr=4.0000e-04) (hash(x)=18987545)
301
+ 287 train 7.435756 (lr=4.0000e-04) (hash(x)=26745063)
302
+ 288 train 7.331068 (lr=3.9999e-04) (hash(x)=21447406)
303
+ 289 train 7.213363 (lr=3.9998e-04) (hash(x)=24650655)
304
+ 290 train 7.615647 (lr=3.9997e-04) (hash(x)=26519976)
305
+ 291 train 7.908910 (lr=3.9996e-04) (hash(x)=20106514)
306
+ 292 train 7.390336 (lr=3.9994e-04) (hash(x)=20117808)
307
+ 293 train 7.523059 (lr=3.9991e-04) (hash(x)=25510334)
308
+ 294 train 7.407591 (lr=3.9989e-04) (hash(x)=25749120)
309
+ 295 train 7.491505 (lr=3.9986e-04) (hash(x)=24357241)
310
+ 296 train 7.479247 (lr=3.9983e-04) (hash(x)=25584945)
311
+ 297 train 7.340627 (lr=3.9979e-04) (hash(x)=25198897)
312
+ 298 train 7.437020 (lr=3.9975e-04) (hash(x)=23475891)
313
+ 299 train 7.362899 (lr=3.9971e-04) (hash(x)=23644804)
314
+ 300 val loss 7.4338
315
+ 300 val perplexity 1692.3020
316
+ 300 train 7.590929 (lr=3.9966e-04) (hash(x)=25629584)
317
+ 301 train 7.707306 (lr=3.9961e-04) (hash(x)=29797280)
318
+ 302 train 7.786726 (lr=3.9955e-04) (hash(x)=30026394)
319
+ 303 train 7.787806 (lr=3.9950e-04) (hash(x)=26594177)
320
+ 304 train 7.554981 (lr=3.9944e-04) (hash(x)=23766256)
321
+ 305 train 7.461482 (lr=3.9937e-04) (hash(x)=24250150)
322
+ 306 train 7.398866 (lr=3.9930e-04) (hash(x)=26108817)
323
+ 307 train 7.327718 (lr=3.9923e-04) (hash(x)=18582688)
324
+ 308 train 7.324193 (lr=3.9916e-04) (hash(x)=23272705)
325
+ 309 train 7.331034 (lr=3.9908e-04) (hash(x)=24793412)
326
+ 310 train 7.328262 (lr=3.9900e-04) (hash(x)=19280989)
327
+ 311 train 7.336274 (lr=3.9891e-04) (hash(x)=22751150)
328
+ 312 train 7.432644 (lr=3.9882e-04) (hash(x)=23537942)
329
+ 313 train 7.288974 (lr=3.9873e-04) (hash(x)=23748375)
330
+ 314 train 7.477032 (lr=3.9864e-04) (hash(x)=27413825)
331
+ 315 train 7.556324 (lr=3.9854e-04) (hash(x)=21682445)
332
+ 316 train 7.387760 (lr=3.9843e-04) (hash(x)=22040184)
333
+ 317 train 7.347260 (lr=3.9833e-04) (hash(x)=24463229)
334
+ 318 train 7.466955 (lr=3.9822e-04) (hash(x)=24091954)
335
+ 319 train 7.315214 (lr=3.9811e-04) (hash(x)=26670219)
336
+ 320 train 7.336398 (lr=3.9799e-04) (hash(x)=25558986)
337
+ 321 train 7.336956 (lr=3.9787e-04) (hash(x)=19218943)
338
+ 322 train 7.421833 (lr=3.9775e-04) (hash(x)=24551402)
339
+ 323 train 7.308145 (lr=3.9762e-04) (hash(x)=25229281)
340
+ 324 train 7.048651 (lr=3.9749e-04) (hash(x)=21251127)
341
+ 325 train 7.372543 (lr=3.9736e-04) (hash(x)=25800857)
342
+ 326 train 7.245156 (lr=3.9722e-04) (hash(x)=25836189)
343
+ 327 train 7.282561 (lr=3.9708e-04) (hash(x)=25999255)
344
+ 328 train 7.369167 (lr=3.9694e-04) (hash(x)=20149394)
345
+ 329 train 7.270473 (lr=3.9679e-04) (hash(x)=21999338)
346
+ 330 train 7.369465 (lr=3.9664e-04) (hash(x)=27611302)
347
+ 331 train 7.326139 (lr=3.9648e-04) (hash(x)=22778356)
348
+ 332 train 7.290041 (lr=3.9633e-04) (hash(x)=24195959)
349
+ 333 train 7.581809 (lr=3.9616e-04) (hash(x)=19127465)
350
+ 334 train 7.425421 (lr=3.9600e-04) (hash(x)=22426274)
351
+ 335 train 7.235216 (lr=3.9583e-04) (hash(x)=23785205)
352
+ 336 train 7.386600 (lr=3.9566e-04) (hash(x)=28505676)
353
+ 337 train 7.376935 (lr=3.9549e-04) (hash(x)=27706477)
354
+ 338 train 7.390443 (lr=3.9531e-04) (hash(x)=28684175)
355
+ 339 train 7.359751 (lr=3.9513e-04) (hash(x)=28003600)
356
+ 340 train 7.356381 (lr=3.9494e-04) (hash(x)=27586845)
357
+ 341 train 7.276884 (lr=3.9475e-04) (hash(x)=21176405)
358
+ 342 train 7.226163 (lr=3.9456e-04) (hash(x)=25888433)
359
+ 343 train 7.397944 (lr=3.9437e-04) (hash(x)=27916982)
360
+ 344 train 7.202490 (lr=3.9417e-04) (hash(x)=24334708)
361
+ 345 train 7.295249 (lr=3.9397e-04) (hash(x)=25617709)
362
+ 346 train 7.415897 (lr=3.9376e-04) (hash(x)=25152362)
363
+ 347 train 7.294838 (lr=3.9356e-04) (hash(x)=24276512)
364
+ 348 train 7.231175 (lr=3.9334e-04) (hash(x)=26776133)
365
+ 349 train 7.116739 (lr=3.9313e-04) (hash(x)=20254159)
366
+ 350 val loss 7.3482
367
+ 350 val perplexity 1553.4376
368
+ 350 train 7.222023 (lr=3.9291e-04) (hash(x)=27873855)
369
+ 351 train 7.299384 (lr=3.9269e-04) (hash(x)=25113614)
370
+ 352 train 7.357775 (lr=3.9246e-04) (hash(x)=26192886)
371
+ 353 train 7.221242 (lr=3.9223e-04) (hash(x)=24614912)
372
+ 354 train 7.247258 (lr=3.9200e-04) (hash(x)=25397093)
373
+ 355 train 7.290273 (lr=3.9177e-04) (hash(x)=19811802)
374
+ 356 train 7.276080 (lr=3.9153e-04) (hash(x)=25579655)
375
+ 357 train 7.390073 (lr=3.9129e-04) (hash(x)=26096514)
376
+ 358 train 7.292040 (lr=3.9104e-04) (hash(x)=22589633)
377
+ 359 train 7.296351 (lr=3.9079e-04) (hash(x)=26757147)
378
+ 360 train 7.375120 (lr=3.9054e-04) (hash(x)=24963730)
379
+ 361 train 7.223266 (lr=3.9029e-04) (hash(x)=27703124)
380
+ 362 train 7.151486 (lr=3.9003e-04) (hash(x)=24781713)
381
+ 363 train 7.351826 (lr=3.8977e-04) (hash(x)=24720476)
382
+ 364 train 7.310544 (lr=3.8950e-04) (hash(x)=23481301)
383
+ 365 train 7.226436 (lr=3.8923e-04) (hash(x)=25641033)
384
+ 366 train 7.281304 (lr=3.8896e-04) (hash(x)=27369437)
385
+ 367 train 7.329542 (lr=3.8869e-04) (hash(x)=24760900)
386
+ 368 train 7.207193 (lr=3.8841e-04) (hash(x)=21692546)
387
+ 369 train 7.270567 (lr=3.8813e-04) (hash(x)=22720391)
388
+ 370 train 7.217208 (lr=3.8785e-04) (hash(x)=24376979)
389
+ 371 train 7.232458 (lr=3.8756e-04) (hash(x)=22244509)
390
+ 372 train 7.203949 (lr=3.8727e-04) (hash(x)=21831960)
391
+ 373 train 7.163178 (lr=3.8697e-04) (hash(x)=24921535)
392
+ 374 train 7.216475 (lr=3.8667e-04) (hash(x)=25381251)
393
+ 375 train 7.219894 (lr=3.8637e-04) (hash(x)=25687352)
394
+ 376 train 7.233547 (lr=3.8607e-04) (hash(x)=25221654)
395
+ 377 train 7.270916 (lr=3.8576e-04) (hash(x)=24469863)
396
+ 378 train 7.336977 (lr=3.8545e-04) (hash(x)=25291413)
397
+ 379 train 7.219201 (lr=3.8514e-04) (hash(x)=25744274)
398
+ 380 train 7.166041 (lr=3.8482e-04) (hash(x)=24246963)
399
+ 381 train 7.176661 (lr=3.8450e-04) (hash(x)=19724058)
400
+ 382 train 7.077043 (lr=3.8418e-04) (hash(x)=20705649)
401
+ 383 train 7.385333 (lr=3.8385e-04) (hash(x)=24938685)
402
+ 384 train 7.421759 (lr=3.8352e-04) (hash(x)=21707000)
403
+ 385 train 7.184870 (lr=3.8319e-04) (hash(x)=24948980)
404
+ 386 train 7.485686 (lr=3.8286e-04) (hash(x)=30773958)
405
+ 387 train 7.585213 (lr=3.8252e-04) (hash(x)=26769046)
406
+ 388 train 7.222593 (lr=3.8217e-04) (hash(x)=18849656)
407
+ 389 train 7.777078 (lr=3.8183e-04) (hash(x)=24624950)
408
+ 390 train 7.338020 (lr=3.8148e-04) (hash(x)=27299605)
409
+ 391 train 7.343889 (lr=3.8113e-04) (hash(x)=24113445)
410
+ 392 train 7.339664 (lr=3.8077e-04) (hash(x)=32644465)
411
+ 393 train 7.447757 (lr=3.8042e-04) (hash(x)=25829219)
412
+ 394 train 7.266203 (lr=3.8006e-04) (hash(x)=25194550)
413
+ 395 train 7.306701 (lr=3.7969e-04) (hash(x)=25941804)
414
+ 396 train 7.304032 (lr=3.7933e-04) (hash(x)=24075727)
415
+ 397 train 7.341809 (lr=3.7896e-04) (hash(x)=23104539)
416
+ 398 train 7.289226 (lr=3.7858e-04) (hash(x)=23018983)
417
+ 399 train 7.238637 (lr=3.7821e-04) (hash(x)=24422929)
418
+ 400 val loss 7.3237
419
+ 400 val perplexity 1515.7701
420
+ 400 train 7.286127 (lr=3.7783e-04) (hash(x)=23389696)
421
+ 401 train 7.266464 (lr=3.7744e-04) (hash(x)=29102969)
422
+ 402 train 7.402743 (lr=3.7706e-04) (hash(x)=28359859)
423
+ 403 train 7.451504 (lr=3.7667e-04) (hash(x)=24630955)
424
+ 404 train 7.457879 (lr=3.7628e-04) (hash(x)=22969113)
425
+ 405 train 7.264542 (lr=3.7588e-04) (hash(x)=27004415)
426
+ 406 train 7.185369 (lr=3.7549e-04) (hash(x)=22265063)
427
+ 407 train 7.207208 (lr=3.7509e-04) (hash(x)=26905990)
428
+ 408 train 7.276162 (lr=3.7468e-04) (hash(x)=24047679)
429
+ 409 train 7.405029 (lr=3.7428e-04) (hash(x)=26905582)
430
+ 410 train 7.150514 (lr=3.7387e-04) (hash(x)=29018970)
431
+ 411 train 7.169543 (lr=3.7345e-04) (hash(x)=24503315)
432
+ 412 train 7.152348 (lr=3.7304e-04) (hash(x)=22607537)
433
+ 413 train 8.001145 (lr=3.7262e-04) (hash(x)=26690225)
434
+ 414 train 7.271646 (lr=3.7220e-04) (hash(x)=24226255)
435
+ 415 train 7.206914 (lr=3.7177e-04) (hash(x)=23622702)
436
+ 416 train 7.251554 (lr=3.7135e-04) (hash(x)=23910425)
437
+ 417 train 7.512989 (lr=3.7092e-04) (hash(x)=27213318)
438
+ 418 train 7.275665 (lr=3.7048e-04) (hash(x)=26701577)
439
+ 419 train 7.210460 (lr=3.7005e-04) (hash(x)=26216794)
440
+ 420 train 7.441942 (lr=3.6961e-04) (hash(x)=27579623)
441
+ 421 train 7.311334 (lr=3.6917e-04) (hash(x)=26086198)
442
+ 422 train 7.256748 (lr=3.6872e-04) (hash(x)=24863441)
443
+ 423 train 7.392657 (lr=3.6828e-04) (hash(x)=25236385)
444
+ 424 train 7.476214 (lr=3.6782e-04) (hash(x)=25330803)
445
+ 425 train 7.310636 (lr=3.6737e-04) (hash(x)=22508532)
446
+ 426 train 7.273331 (lr=3.6692e-04) (hash(x)=21799261)
447
+ 427 train 7.363615 (lr=3.6646e-04) (hash(x)=26311168)
448
+ 428 train 7.163663 (lr=3.6599e-04) (hash(x)=20977430)
449
+ 429 train 7.229524 (lr=3.6553e-04) (hash(x)=26941617)
450
+ 430 train 7.511751 (lr=3.6506e-04) (hash(x)=30008957)
451
+ 431 train 7.343771 (lr=3.6459e-04) (hash(x)=27125962)
452
+ 432 train 7.183892 (lr=3.6412e-04) (hash(x)=25888358)
453
+ 433 train 7.340558 (lr=3.6364e-04) (hash(x)=21725719)
454
+ 434 train 7.207319 (lr=3.6316e-04) (hash(x)=21669704)
455
+ 435 train 7.152303 (lr=3.6268e-04) (hash(x)=23166993)
456
+ 436 train 7.238644 (lr=3.6220e-04) (hash(x)=26235974)
457
+ 437 train 7.415007 (lr=3.6171e-04) (hash(x)=25426430)
458
+ 438 train 7.142289 (lr=3.6122e-04) (hash(x)=22966314)
459
+ 439 train 7.304042 (lr=3.6073e-04) (hash(x)=23448505)
460
+ 440 train 7.173251 (lr=3.6023e-04) (hash(x)=26330693)
461
+ 441 train 7.114500 (lr=3.5974e-04) (hash(x)=29550596)
462
+ 442 train 7.176942 (lr=3.5924e-04) (hash(x)=22124892)
463
+ 443 train 7.178204 (lr=3.5873e-04) (hash(x)=25607640)
464
+ 444 train 7.059385 (lr=3.5823e-04) (hash(x)=23795894)
465
+ 445 train 6.956395 (lr=3.5772e-04) (hash(x)=24165449)
466
+ 446 train 7.210876 (lr=3.5721e-04) (hash(x)=24697756)
467
+ 447 train 7.296562 (lr=3.5669e-04) (hash(x)=27218899)
468
+ 448 train 7.174352 (lr=3.5618e-04) (hash(x)=24748522)
469
+ 449 train 7.275671 (lr=3.5566e-04) (hash(x)=24965317)
470
+ 450 val loss 7.3110
471
+ 450 val perplexity 1496.6229
472
+ 450 train 7.049213 (lr=3.5514e-04) (hash(x)=22031210)
473
+ 451 train 7.160579 (lr=3.5461e-04) (hash(x)=25877990)
474
+ 452 train 7.226542 (lr=3.5408e-04) (hash(x)=24382381)
475
+ 453 train 7.134096 (lr=3.5355e-04) (hash(x)=24551999)
476
+ 454 train 7.212431 (lr=3.5302e-04) (hash(x)=25765516)
477
+ 455 train 7.248083 (lr=3.5249e-04) (hash(x)=25204247)
478
+ 456 train 7.307763 (lr=3.5195e-04) (hash(x)=26321813)
479
+ 457 train 7.222360 (lr=3.5141e-04) (hash(x)=25528323)
480
+ 458 train 7.233037 (lr=3.5087e-04) (hash(x)=25114729)
481
+ 459 train 7.210031 (lr=3.5032e-04) (hash(x)=31341964)
482
+ 460 train 7.266203 (lr=3.4977e-04) (hash(x)=26326024)
483
+ 461 train 7.135107 (lr=3.4922e-04) (hash(x)=27273311)
484
+ 462 train 7.141449 (lr=3.4867e-04) (hash(x)=17819313)
485
+ 463 train 7.326743 (lr=3.4812e-04) (hash(x)=23769243)
486
+ 464 train 7.253543 (lr=3.4756e-04) (hash(x)=23788738)
487
+ 465 train 7.188145 (lr=3.4700e-04) (hash(x)=24818744)
488
+ 466 train 7.125351 (lr=3.4644e-04) (hash(x)=23472795)
489
+ 467 train 7.561700 (lr=3.4587e-04) (hash(x)=24991832)
490
+ 468 train 7.281077 (lr=3.4530e-04) (hash(x)=22424066)
491
+ 469 train 7.173289 (lr=3.4473e-04) (hash(x)=26474878)
492
+ 470 train 7.267947 (lr=3.4416e-04) (hash(x)=22704349)
493
+ 471 train 7.228820 (lr=3.4359e-04) (hash(x)=22293673)
494
+ 472 train 7.213423 (lr=3.4301e-04) (hash(x)=21785487)
495
+ 473 train 7.232505 (lr=3.4243e-04) (hash(x)=25251063)
496
+ 474 train 7.128202 (lr=3.4185e-04) (hash(x)=22722211)
497
+ 475 train 7.206573 (lr=3.4127e-04) (hash(x)=24522269)
498
+ 476 train 7.136857 (lr=3.4068e-04) (hash(x)=23329475)
499
+ 477 train 7.172847 (lr=3.4009e-04) (hash(x)=24735165)
500
+ 478 train 6.982765 (lr=3.3950e-04) (hash(x)=27002892)
501
+ 479 train 7.018597 (lr=3.3891e-04) (hash(x)=23499349)
502
+ 480 train 7.376822 (lr=3.3831e-04) (hash(x)=22461589)
503
+ 481 train 7.158376 (lr=3.3771e-04) (hash(x)=26415040)
504
+ 482 train 7.381897 (lr=3.3711e-04) (hash(x)=26349465)
505
+ 483 train 7.038072 (lr=3.3651e-04) (hash(x)=21596677)
506
+ 484 train 7.096442 (lr=3.3590e-04) (hash(x)=24117012)
507
+ 485 train 7.171072 (lr=3.3530e-04) (hash(x)=24105761)
508
+ 486 train 7.106800 (lr=3.3469e-04) (hash(x)=25725049)
509
+ 487 train 6.972893 (lr=3.3408e-04) (hash(x)=22892776)
510
+ 488 train 7.085896 (lr=3.3346e-04) (hash(x)=21630383)
511
+ 489 train 7.115988 (lr=3.3285e-04) (hash(x)=24727094)
512
+ 490 train 7.026695 (lr=3.3223e-04) (hash(x)=22835000)
513
+ 491 train 7.212759 (lr=3.3161e-04) (hash(x)=22839049)
514
+ 492 train 7.054127 (lr=3.3099e-04) (hash(x)=20654136)
515
+ 493 train 7.301900 (lr=3.3036e-04) (hash(x)=26776000)
516
+ 494 train 6.942503 (lr=3.2973e-04) (hash(x)=22560658)
517
+ 495 train 7.377705 (lr=3.2911e-04) (hash(x)=25995933)
518
+ 496 train 7.303614 (lr=3.2847e-04) (hash(x)=24635394)
519
+ 497 train 7.398937 (lr=3.2784e-04) (hash(x)=27537228)
520
+ 498 train 7.301142 (lr=3.2721e-04) (hash(x)=26496730)
521
+ 499 train 7.024153 (lr=3.2657e-04) (hash(x)=16533310)
522
+ 500 val loss 7.2660
523
+ 500 val perplexity 1430.7765
524
+ 500 train 7.135050 (lr=3.2593e-04) (hash(x)=24881176)
525
+ 501 train 7.347652 (lr=3.2529e-04) (hash(x)=25593613)
526
+ 502 train 7.083936 (lr=3.2464e-04) (hash(x)=22886646)
527
+ 503 train 7.339124 (lr=3.2400e-04) (hash(x)=21407676)
528
+ 504 train 7.228539 (lr=3.2335e-04) (hash(x)=21720304)
529
+ 505 train 7.379178 (lr=3.2270e-04) (hash(x)=27263416)
530
+ 506 train 7.127178 (lr=3.2205e-04) (hash(x)=25840801)
531
+ 507 train 7.280003 (lr=3.2140e-04) (hash(x)=31424568)
532
+ 508 train 7.291379 (lr=3.2074e-04) (hash(x)=23403065)
533
+ 509 train 7.113430 (lr=3.2008e-04) (hash(x)=22656033)
534
+ 510 train 7.230016 (lr=3.1943e-04) (hash(x)=26869295)
535
+ 511 train 7.502090 (lr=3.1876e-04) (hash(x)=25504151)
536
+ 512 train 7.295562 (lr=3.1810e-04) (hash(x)=25781518)
537
+ 513 train 7.233940 (lr=3.1744e-04) (hash(x)=26432850)
538
+ 514 train 7.110104 (lr=3.1677e-04) (hash(x)=20155894)
539
+ 515 train 7.008984 (lr=3.1610e-04) (hash(x)=22270169)
540
+ 516 train 7.150598 (lr=3.1543e-04) (hash(x)=22108461)
541
+ 517 train 7.225751 (lr=3.1476e-04) (hash(x)=24802848)
542
+ 518 train 7.317863 (lr=3.1408e-04) (hash(x)=25086604)
543
+ 519 train 7.319765 (lr=3.1341e-04) (hash(x)=26983186)
544
+ 520 train 7.205115 (lr=3.1273e-04) (hash(x)=27208507)
545
+ 521 train 7.384169 (lr=3.1205e-04) (hash(x)=23862845)
546
+ 522 train 7.206310 (lr=3.1137e-04) (hash(x)=22664494)
547
+ 523 train 7.178551 (lr=3.1069e-04) (hash(x)=24092927)
548
+ 524 train 7.132616 (lr=3.1000e-04) (hash(x)=27213812)
549
+ 525 train 7.108545 (lr=3.0931e-04) (hash(x)=23929955)
550
+ 526 train 7.098413 (lr=3.0862e-04) (hash(x)=23025790)
551
+ 527 train 7.045512 (lr=3.0793e-04) (hash(x)=20452272)
552
+ 528 train 7.262859 (lr=3.0724e-04) (hash(x)=26473464)
553
+ 529 train 7.231150 (lr=3.0655e-04) (hash(x)=28276603)
554
+ 530 train 7.060543 (lr=3.0585e-04) (hash(x)=21480367)
555
+ 531 train 7.246252 (lr=3.0516e-04) (hash(x)=24658642)
556
+ 532 train 7.213847 (lr=3.0446e-04) (hash(x)=26854801)
557
+ 533 train 7.047395 (lr=3.0376e-04) (hash(x)=17862226)
558
+ 534 train 7.161891 (lr=3.0306e-04) (hash(x)=28519182)
559
+ 535 train 7.143240 (lr=3.0235e-04) (hash(x)=25193694)
560
+ 536 train 7.186128 (lr=3.0165e-04) (hash(x)=25642620)
561
+ 537 train 7.223233 (lr=3.0094e-04) (hash(x)=25765909)
562
+ 538 train 7.011407 (lr=3.0023e-04) (hash(x)=20750531)
563
+ 539 train 7.135511 (lr=2.9952e-04) (hash(x)=24917098)
564
+ 540 train 7.077359 (lr=2.9881e-04) (hash(x)=24499766)
565
+ 541 train 7.117239 (lr=2.9810e-04) (hash(x)=25067358)
566
+ 542 train 7.220417 (lr=2.9738e-04) (hash(x)=27479796)
567
+ 543 train 7.104497 (lr=2.9667e-04) (hash(x)=23136080)
568
+ 544 train 7.193572 (lr=2.9595e-04) (hash(x)=29810074)
569
+ 545 train 7.082312 (lr=2.9523e-04) (hash(x)=24765539)
570
+ 546 train 7.136913 (lr=2.9451e-04) (hash(x)=22216309)
571
+ 547 train 7.072034 (lr=2.9379e-04) (hash(x)=20613861)
572
+ 548 train 7.042239 (lr=2.9307e-04) (hash(x)=23462447)
573
+ 549 train 7.115380 (lr=2.9234e-04) (hash(x)=26739455)
574
+ 550 val loss 7.2208
575
+ 550 val perplexity 1367.6086
576
+ 550 train 7.129200 (lr=2.9162e-04) (hash(x)=23222996)
577
+ 551 train 7.195045 (lr=2.9089e-04) (hash(x)=26489285)
578
+ 552 train 6.983945 (lr=2.9016e-04) (hash(x)=25185942)
579
+ 553 train 7.314059 (lr=2.8943e-04) (hash(x)=27476158)
580
+ 554 train 7.091814 (lr=2.8870e-04) (hash(x)=23653175)
581
+ 555 train 7.060676 (lr=2.8797e-04) (hash(x)=22405076)
582
+ 556 train 7.084158 (lr=2.8723e-04) (hash(x)=25673545)
583
+ 557 train 7.190689 (lr=2.8650e-04) (hash(x)=22675428)
584
+ 558 train 7.299218 (lr=2.8576e-04) (hash(x)=28168456)
585
+ 559 train 7.404202 (lr=2.8502e-04) (hash(x)=26072586)
586
+ 560 train 7.189232 (lr=2.8428e-04) (hash(x)=24974331)
587
+ 561 train 6.974414 (lr=2.8354e-04) (hash(x)=21951984)
588
+ 562 train 7.362377 (lr=2.8280e-04) (hash(x)=25403751)
589
+ 563 train 7.150120 (lr=2.8206e-04) (hash(x)=26396012)
590
+ 564 train 7.123554 (lr=2.8132e-04) (hash(x)=24827190)
591
+ 565 train 7.240861 (lr=2.8057e-04) (hash(x)=24847458)
592
+ 566 train 7.209997 (lr=2.7982e-04) (hash(x)=22159088)
593
+ 567 train 7.135837 (lr=2.7908e-04) (hash(x)=22140034)
594
+ 568 train 7.180117 (lr=2.7833e-04) (hash(x)=19044797)
595
+ 569 train 7.102286 (lr=2.7758e-04) (hash(x)=23932794)
596
+ 570 train 7.200238 (lr=2.7683e-04) (hash(x)=25256890)
597
+ 571 train 7.269563 (lr=2.7607e-04) (hash(x)=24161711)
598
+ 572 train 7.192078 (lr=2.7532e-04) (hash(x)=22036817)
599
+ 573 train 7.031141 (lr=2.7457e-04) (hash(x)=26663208)
600
+ 574 train 7.086411 (lr=2.7381e-04) (hash(x)=24072213)
601
+ 575 train 7.078128 (lr=2.7306e-04) (hash(x)=22766044)
602
+ 576 train 7.183905 (lr=2.7230e-04) (hash(x)=27871153)
603
+ 577 train 7.171895 (lr=2.7154e-04) (hash(x)=24037280)
604
+ 578 train 7.324442 (lr=2.7078e-04) (hash(x)=24121850)
605
+ 579 train 7.114185 (lr=2.7002e-04) (hash(x)=24380031)
606
+ 580 train 7.072943 (lr=2.6926e-04) (hash(x)=23055215)
607
+ 581 train 7.285850 (lr=2.6850e-04) (hash(x)=25329724)
608
+ 582 train 7.050620 (lr=2.6773e-04) (hash(x)=25662408)
609
+ 583 train 7.150427 (lr=2.6697e-04) (hash(x)=26346814)
610
+ 584 train 7.261900 (lr=2.6620e-04) (hash(x)=25283256)
611
+ 585 train 7.084950 (lr=2.6544e-04) (hash(x)=24021771)
612
+ 586 train 7.081648 (lr=2.6467e-04) (hash(x)=23576840)
613
+ 587 train 7.247495 (lr=2.6390e-04) (hash(x)=26414858)
614
+ 588 train 7.008790 (lr=2.6314e-04) (hash(x)=20099261)
615
+ 589 train 7.002177 (lr=2.6237e-04) (hash(x)=22453718)
616
+ 590 train 6.964484 (lr=2.6160e-04) (hash(x)=22529445)
617
+ 591 train 7.049136 (lr=2.6083e-04) (hash(x)=24484543)
618
+ 592 train 6.972813 (lr=2.6005e-04) (hash(x)=24043998)
619
+ 593 train 7.113982 (lr=2.5928e-04) (hash(x)=24301906)
620
+ 594 train 7.056588 (lr=2.5851e-04) (hash(x)=26328881)
621
+ 595 train 7.104700 (lr=2.5773e-04) (hash(x)=23101508)
622
+ 596 train 7.094124 (lr=2.5696e-04) (hash(x)=23527176)
623
+ 597 train 7.540867 (lr=2.5618e-04) (hash(x)=22705673)
624
+ 598 train 7.165732 (lr=2.5541e-04) (hash(x)=28664796)
625
+ 599 train 7.261077 (lr=2.5463e-04) (hash(x)=27102890)
626
+ 600 val loss 7.2058
627
+ 600 val perplexity 1347.2787
628
+ 600 train 7.388664 (lr=2.5385e-04) (hash(x)=26787259)
629
+ 601 train 7.383162 (lr=2.5307e-04) (hash(x)=23396088)
630
+ 602 train 7.123705 (lr=2.5230e-04) (hash(x)=24911831)
631
+ 603 train 7.165345 (lr=2.5152e-04) (hash(x)=23453491)
632
+ 604 train 7.211850 (lr=2.5074e-04) (hash(x)=23453396)
633
+ 605 train 7.284858 (lr=2.4996e-04) (hash(x)=23869612)
634
+ 606 train 7.373194 (lr=2.4917e-04) (hash(x)=26581590)
635
+ 607 train 7.203884 (lr=2.4839e-04) (hash(x)=28056342)
636
+ 608 train 7.244086 (lr=2.4761e-04) (hash(x)=25900866)
637
+ 609 train 7.267764 (lr=2.4683e-04) (hash(x)=24549177)
638
+ 610 train 7.199110 (lr=2.4604e-04) (hash(x)=28239583)
639
+ 611 train 7.230692 (lr=2.4526e-04) (hash(x)=23223554)
640
+ 612 train 7.276765 (lr=2.4448e-04) (hash(x)=24793480)
641
+ 613 train 7.185047 (lr=2.4369e-04) (hash(x)=25415570)
642
+ 614 train 7.256577 (lr=2.4291e-04) (hash(x)=24425760)
643
+ 615 train 7.312207 (lr=2.4212e-04) (hash(x)=23707134)
644
+ 616 train 7.256007 (lr=2.4133e-04) (hash(x)=28321697)
645
+ 617 train 7.187799 (lr=2.4055e-04) (hash(x)=23434031)
646
+ 618 train 7.161068 (lr=2.3976e-04) (hash(x)=27003481)
647
+ 619 train 7.162149 (lr=2.3897e-04) (hash(x)=22236893)
648
+ 620 train 7.184744 (lr=2.3818e-04) (hash(x)=26857458)
649
+ 621 train 7.192053 (lr=2.3740e-04) (hash(x)=26413122)
650
+ 622 train 7.502143 (lr=2.3661e-04) (hash(x)=27696537)
651
+ 623 train 7.197651 (lr=2.3582e-04) (hash(x)=21950234)
652
+ 624 train 7.221639 (lr=2.3503e-04) (hash(x)=24051952)
653
+ 625 train 7.251692 (lr=2.3424e-04) (hash(x)=24029939)
654
+ 626 train 7.322046 (lr=2.3345e-04) (hash(x)=24393346)
655
+ 627 train 7.271921 (lr=2.3266e-04) (hash(x)=26985352)
656
+ 628 train 7.043303 (lr=2.3187e-04) (hash(x)=19493265)
657
+ 629 train 7.115475 (lr=2.3108e-04) (hash(x)=23944010)
658
+ 630 train 7.320976 (lr=2.3029e-04) (hash(x)=21517435)
659
+ 631 train 7.091166 (lr=2.2950e-04) (hash(x)=25025868)
660
+ 632 train 7.093050 (lr=2.2871e-04) (hash(x)=22940997)
661
+ 633 train 7.217952 (lr=2.2792e-04) (hash(x)=25864840)
662
+ 634 train 7.160185 (lr=2.2713e-04) (hash(x)=24682151)
663
+ 635 train 7.077282 (lr=2.2633e-04) (hash(x)=27055330)
664
+ 636 train 7.151564 (lr=2.2554e-04) (hash(x)=22705710)
665
+ 637 train 7.246555 (lr=2.2475e-04) (hash(x)=26775238)
666
+ 638 train 7.375243 (lr=2.2396e-04) (hash(x)=21734331)
667
+ 639 train 7.122048 (lr=2.2317e-04) (hash(x)=25693442)
668
+ 640 train 7.208173 (lr=2.2238e-04) (hash(x)=25290454)
669
+ 641 train 7.165658 (lr=2.2158e-04) (hash(x)=25699860)
670
+ 642 train 7.351058 (lr=2.2079e-04) (hash(x)=19919115)
671
+ 643 train 7.395107 (lr=2.2000e-04) (hash(x)=25586967)
672
+ 644 train 7.145013 (lr=2.1921e-04) (hash(x)=25373506)
673
+ 645 train 7.131990 (lr=2.1842e-04) (hash(x)=26496148)
674
+ 646 train 7.225438 (lr=2.1762e-04) (hash(x)=27395083)
675
+ 647 train 7.406580 (lr=2.1683e-04) (hash(x)=27389709)
676
+ 648 train 7.276235 (lr=2.1604e-04) (hash(x)=29172968)
677
+ 649 train 7.196164 (lr=2.1525e-04) (hash(x)=25464920)
678
+ 650 val loss 7.1766
679
+ 650 val perplexity 1308.4232
680
+ 650 train 7.263709 (lr=2.1446e-04) (hash(x)=25992320)
681
+ 651 train 7.216295 (lr=2.1367e-04) (hash(x)=24949509)
682
+ 652 train 7.249476 (lr=2.1287e-04) (hash(x)=25615151)
683
+ 653 train 7.213801 (lr=2.1208e-04) (hash(x)=27923548)
684
+ 654 train 7.259383 (lr=2.1129e-04) (hash(x)=26387848)
685
+ 655 train 7.202737 (lr=2.1050e-04) (hash(x)=22863518)
686
+ 656 train 7.156873 (lr=2.0971e-04) (hash(x)=20869581)
687
+ 657 train 7.305210 (lr=2.0892e-04) (hash(x)=18133606)
688
+ 658 train 7.192937 (lr=2.0813e-04) (hash(x)=28736122)
689
+ 659 train 7.176794 (lr=2.0734e-04) (hash(x)=25528329)
690
+ 660 train 7.214803 (lr=2.0655e-04) (hash(x)=24020765)
691
+ 661 train 7.274973 (lr=2.0576e-04) (hash(x)=27792245)
692
+ 662 train 7.289534 (lr=2.0497e-04) (hash(x)=28287635)
693
+ 663 train 7.070947 (lr=2.0418e-04) (hash(x)=20951206)
694
+ 664 train 7.094005 (lr=2.0339e-04) (hash(x)=22356452)
695
+ 665 train 7.151302 (lr=2.0260e-04) (hash(x)=24385170)
696
+ 666 train 7.247064 (lr=2.0182e-04) (hash(x)=23455854)
697
+ 667 train 7.200358 (lr=2.0103e-04) (hash(x)=21589405)
698
+ 668 train 7.195724 (lr=2.0024e-04) (hash(x)=24698150)
699
+ 669 train 7.179058 (lr=1.9945e-04) (hash(x)=28010836)
700
+ 670 train 7.333903 (lr=1.9867e-04) (hash(x)=25427381)
701
+ 671 train 7.261661 (lr=1.9788e-04) (hash(x)=27478557)
702
+ 672 train 7.193837 (lr=1.9709e-04) (hash(x)=23865569)
703
+ 673 train 7.301037 (lr=1.9631e-04) (hash(x)=23407132)
704
+ 674 train 7.146557 (lr=1.9552e-04) (hash(x)=20253664)
705
+ 675 train 7.134809 (lr=1.9474e-04) (hash(x)=21087928)
706
+ 676 train 7.188758 (lr=1.9396e-04) (hash(x)=25957589)
707
+ 677 train 7.148079 (lr=1.9317e-04) (hash(x)=27008610)
708
+ 678 train 7.241390 (lr=1.9239e-04) (hash(x)=27580265)
709
+ 679 train 7.249039 (lr=1.9161e-04) (hash(x)=24244798)
710
+ 680 train 7.131508 (lr=1.9083e-04) (hash(x)=24535871)
711
+ 681 train 7.263510 (lr=1.9004e-04) (hash(x)=27558681)
712
+ 682 train 7.112218 (lr=1.8926e-04) (hash(x)=21152793)
713
+ 683 train 7.111492 (lr=1.8848e-04) (hash(x)=28147005)
714
+ 684 train 7.186238 (lr=1.8770e-04) (hash(x)=21933180)
715
+ 685 train 7.177848 (lr=1.8693e-04) (hash(x)=24735895)
716
+ 686 train 7.209908 (lr=1.8615e-04) (hash(x)=26209948)
717
+ 687 train 7.209423 (lr=1.8537e-04) (hash(x)=23358313)
718
+ 688 train 7.137181 (lr=1.8459e-04) (hash(x)=22398001)
719
+ 689 train 7.031833 (lr=1.8382e-04) (hash(x)=23724957)
720
+ 690 train 7.092579 (lr=1.8304e-04) (hash(x)=23083514)
721
+ 691 train 6.983320 (lr=1.8227e-04) (hash(x)=16016924)
722
+ 692 train 7.249783 (lr=1.8149e-04) (hash(x)=27173674)
723
+ 693 train 7.032318 (lr=1.8072e-04) (hash(x)=28056250)
724
+ 694 train 7.157750 (lr=1.7995e-04) (hash(x)=22233617)
725
+ 695 train 7.240007 (lr=1.7917e-04) (hash(x)=23777908)
726
+ 696 train 7.206275 (lr=1.7840e-04) (hash(x)=27842849)
727
+ 697 train 7.231604 (lr=1.7763e-04) (hash(x)=23711679)
728
+ 698 train 7.054009 (lr=1.7686e-04) (hash(x)=17204843)
729
+ 699 train 7.363975 (lr=1.7610e-04) (hash(x)=23909800)
730
+ 700 val loss 7.1432
731
+ 700 val perplexity 1265.4152
732
+ 700 train 7.193203 (lr=1.7533e-04) (hash(x)=25033286)
733
+ 701 train 7.501172 (lr=1.7456e-04) (hash(x)=23608371)
734
+ 702 train 7.272604 (lr=1.7380e-04) (hash(x)=23654625)
735
+ 703 train 7.113399 (lr=1.7303e-04) (hash(x)=22633313)
736
+ 704 train 7.174962 (lr=1.7227e-04) (hash(x)=24267011)
737
+ 705 train 7.126307 (lr=1.7150e-04) (hash(x)=23449416)
738
+ 706 train 7.378852 (lr=1.7074e-04) (hash(x)=28527906)
739
+ 707 train 7.054196 (lr=1.6998e-04) (hash(x)=21189075)
740
+ 708 train 7.117593 (lr=1.6922e-04) (hash(x)=25278645)
741
+ 709 train 7.556139 (lr=1.6846e-04) (hash(x)=23654290)
742
+ 710 train 7.112774 (lr=1.6770e-04) (hash(x)=24571139)
743
+ 711 train 7.189974 (lr=1.6694e-04) (hash(x)=26203172)
744
+ 712 train 7.132493 (lr=1.6619e-04) (hash(x)=19623503)
745
+ 713 train 7.097471 (lr=1.6543e-04) (hash(x)=23970545)
746
+ 714 train 7.294400 (lr=1.6468e-04) (hash(x)=25724900)
747
+ 715 train 7.215743 (lr=1.6393e-04) (hash(x)=23245421)
748
+ 716 train 7.141586 (lr=1.6317e-04) (hash(x)=24087537)
749
+ 717 train 7.075061 (lr=1.6242e-04) (hash(x)=25549646)
750
+ 718 train 7.155168 (lr=1.6167e-04) (hash(x)=26406563)
751
+ 719 train 7.199550 (lr=1.6092e-04) (hash(x)=23972429)
752
+ 720 train 7.243266 (lr=1.6018e-04) (hash(x)=26378040)
753
+ 721 train 7.116014 (lr=1.5943e-04) (hash(x)=24618531)
754
+ 722 train 7.126006 (lr=1.5868e-04) (hash(x)=20135714)
755
+ 723 train 7.260794 (lr=1.5794e-04) (hash(x)=26490981)
756
+ 724 train 7.288964 (lr=1.5720e-04) (hash(x)=23927567)
757
+ 725 train 7.125206 (lr=1.5646e-04) (hash(x)=27943609)
758
+ 726 train 7.167618 (lr=1.5572e-04) (hash(x)=22962561)
759
+ 727 train 7.216525 (lr=1.5498e-04) (hash(x)=24772257)
760
+ 728 train 7.143816 (lr=1.5424e-04) (hash(x)=25539761)
761
+ 729 train 7.281848 (lr=1.5350e-04) (hash(x)=23316262)
762
+ 730 train 7.057839 (lr=1.5277e-04) (hash(x)=23857558)
763
+ 731 train 7.240856 (lr=1.5203e-04) (hash(x)=30252690)
764
+ 732 train 7.299968 (lr=1.5130e-04) (hash(x)=30590162)
765
+ 733 train 7.197959 (lr=1.5057e-04) (hash(x)=25015721)
766
+ 734 train 7.276554 (lr=1.4984e-04) (hash(x)=28639679)
767
+ 735 train 7.163798 (lr=1.4911e-04) (hash(x)=26199058)
768
+ 736 train 7.215036 (lr=1.4838e-04) (hash(x)=28907461)
769
+ 737 train 7.198717 (lr=1.4766e-04) (hash(x)=26466915)
770
+ 738 train 7.268178 (lr=1.4693e-04) (hash(x)=20332613)
771
+ 739 train 7.153404 (lr=1.4621e-04) (hash(x)=26766470)
772
+ 740 train 7.099974 (lr=1.4549e-04) (hash(x)=26722145)
773
+ 741 train 7.153911 (lr=1.4477e-04) (hash(x)=21222489)
774
+ 742 train 7.280080 (lr=1.4405e-04) (hash(x)=24662435)
775
+ 743 train 7.104404 (lr=1.4333e-04) (hash(x)=24308927)
776
+ 744 train 7.077472 (lr=1.4262e-04) (hash(x)=28442502)
777
+ 745 train 7.111529 (lr=1.4190e-04) (hash(x)=22810435)
778
+ 746 train 7.243184 (lr=1.4119e-04) (hash(x)=30486518)
779
+ 747 train 7.232344 (lr=1.4048e-04) (hash(x)=22688712)
780
+ 748 train 7.121601 (lr=1.3977e-04) (hash(x)=20712962)
781
+ 749 train 7.014725 (lr=1.3906e-04) (hash(x)=27490086)
782
+ 750 val loss 7.1176
783
+ 750 val perplexity 1233.5182
784
+ 750 train 7.008354 (lr=1.3835e-04) (hash(x)=23973908)
785
+ 751 train 7.141920 (lr=1.3765e-04) (hash(x)=29132401)
786
+ 752 train 7.141520 (lr=1.3694e-04) (hash(x)=27386700)
787
+ 753 train 7.192436 (lr=1.3624e-04) (hash(x)=24268614)
788
+ 754 train 7.237129 (lr=1.3554e-04) (hash(x)=27378387)
789
+ 755 train 7.024290 (lr=1.3484e-04) (hash(x)=22529400)
790
+ 756 train 7.098614 (lr=1.3415e-04) (hash(x)=27586989)
791
+ 757 train 7.051932 (lr=1.3345e-04) (hash(x)=22336349)
792
+ 758 train 7.084560 (lr=1.3276e-04) (hash(x)=26147854)
793
+ 759 train 7.171667 (lr=1.3207e-04) (hash(x)=25346788)
794
+ 760 train 7.243577 (lr=1.3138e-04) (hash(x)=25956855)
795
+ 761 train 7.177638 (lr=1.3069e-04) (hash(x)=24367780)
796
+ 762 train 7.087391 (lr=1.3000e-04) (hash(x)=26084462)
797
+ 763 train 7.146277 (lr=1.2931e-04) (hash(x)=26702459)
798
+ 764 train 7.089565 (lr=1.2863e-04) (hash(x)=25717976)
799
+ 765 train 7.033863 (lr=1.2795e-04) (hash(x)=22141366)
800
+ 766 train 7.252100 (lr=1.2727e-04) (hash(x)=24652423)
801
+ 767 train 7.187285 (lr=1.2659e-04) (hash(x)=30911699)
802
+ 768 train 7.167632 (lr=1.2592e-04) (hash(x)=22076954)
803
+ 769 train 7.324314 (lr=1.2524e-04) (hash(x)=21586769)
804
+ 770 train 7.173502 (lr=1.2457e-04) (hash(x)=21994337)
805
+ 771 train 7.156310 (lr=1.2390e-04) (hash(x)=25454145)
806
+ 772 train 7.209539 (lr=1.2323e-04) (hash(x)=23239550)
807
+ 773 train 7.247126 (lr=1.2256e-04) (hash(x)=23565926)
808
+ 774 train 7.090254 (lr=1.2190e-04) (hash(x)=24705563)
809
+ 775 train 7.184941 (lr=1.2124e-04) (hash(x)=27671095)
810
+ 776 train 7.300205 (lr=1.2057e-04) (hash(x)=26254303)
811
+ 777 train 7.048709 (lr=1.1992e-04) (hash(x)=27454020)
812
+ 778 train 7.043571 (lr=1.1926e-04) (hash(x)=24394565)
813
+ 779 train 6.974378 (lr=1.1860e-04) (hash(x)=21780199)
814
+ 780 train 7.073063 (lr=1.1795e-04) (hash(x)=28296896)
815
+ 781 train 7.218161 (lr=1.1730e-04) (hash(x)=26766303)
816
+ 782 train 7.080850 (lr=1.1665e-04) (hash(x)=21599973)
817
+ 783 train 7.369593 (lr=1.1600e-04) (hash(x)=24969360)
818
+ 784 train 7.339558 (lr=1.1536e-04) (hash(x)=24382006)
819
+ 785 train 7.020406 (lr=1.1471e-04) (hash(x)=22009299)
820
+ 786 train 6.870216 (lr=1.1407e-04) (hash(x)=22376569)
821
+ 787 train 7.045311 (lr=1.1343e-04) (hash(x)=28166843)
822
+ 788 train 7.324308 (lr=1.1279e-04) (hash(x)=28230119)
823
+ 789 train 7.268019 (lr=1.1216e-04) (hash(x)=26207465)
824
+ 790 train 7.338013 (lr=1.1153e-04) (hash(x)=24748118)
825
+ 791 train 7.247057 (lr=1.1089e-04) (hash(x)=27033802)
826
+ 792 train 7.183656 (lr=1.1027e-04) (hash(x)=28788047)
827
+ 793 train 7.034350 (lr=1.0964e-04) (hash(x)=22185099)
828
+ 794 train 7.236173 (lr=1.0901e-04) (hash(x)=21585136)
829
+ 795 train 7.058391 (lr=1.0839e-04) (hash(x)=27691968)
830
+ 796 train 7.080796 (lr=1.0777e-04) (hash(x)=23706103)
831
+ 797 train 6.940464 (lr=1.0715e-04) (hash(x)=27918030)
832
+ 798 train 7.154420 (lr=1.0654e-04) (hash(x)=23127886)
833
+ 799 train 7.178724 (lr=1.0592e-04) (hash(x)=26197070)
834
+ 800 val loss 7.0980
835
+ 800 val perplexity 1209.5577
836
+ 800 train 7.222209 (lr=1.0531e-04) (hash(x)=26463005)
837
+ 801 train 7.107811 (lr=1.0470e-04) (hash(x)=19160134)
838
+ 802 train 7.082170 (lr=1.0410e-04) (hash(x)=23699136)
839
+ 803 train 7.135591 (lr=1.0349e-04) (hash(x)=25353051)
840
+ 804 train 7.077827 (lr=1.0289e-04) (hash(x)=21558484)
841
+ 805 train 7.008379 (lr=1.0229e-04) (hash(x)=22828380)
842
+ 806 train 7.110964 (lr=1.0169e-04) (hash(x)=24623811)
843
+ 807 train 7.256065 (lr=1.0109e-04) (hash(x)=25410299)
844
+ 808 train 7.151887 (lr=1.0050e-04) (hash(x)=26046091)
845
+ 809 train 7.260276 (lr=9.9910e-05) (hash(x)=29887482)
846
+ 810 train 7.002439 (lr=9.9321e-05) (hash(x)=21809706)
847
+ 811 train 7.139990 (lr=9.8735e-05) (hash(x)=23712872)
848
+ 812 train 7.033081 (lr=9.8151e-05) (hash(x)=25875103)
849
+ 813 train 6.984247 (lr=9.7569e-05) (hash(x)=23736441)
850
+ 814 train 7.078032 (lr=9.6990e-05) (hash(x)=25093883)
851
+ 815 train 7.025978 (lr=9.6413e-05) (hash(x)=23692325)
852
+ 816 train 6.908782 (lr=9.5838e-05) (hash(x)=23061082)
853
+ 817 train 7.059279 (lr=9.5266e-05) (hash(x)=26094253)
854
+ 818 train 7.187233 (lr=9.4696e-05) (hash(x)=27327886)
855
+ 819 train 7.088994 (lr=9.4129e-05) (hash(x)=26186143)
856
+ 820 train 7.065170 (lr=9.3564e-05) (hash(x)=25336043)
857
+ 821 train 7.114448 (lr=9.3001e-05) (hash(x)=22811468)
858
+ 822 train 7.104336 (lr=9.2441e-05) (hash(x)=27925831)
859
+ 823 train 7.197074 (lr=9.1884e-05) (hash(x)=35528108)
860
+ 824 train 7.178261 (lr=9.1328e-05) (hash(x)=24949421)
861
+ 825 train 6.951069 (lr=9.0776e-05) (hash(x)=22845735)
862
+ 826 train 7.063087 (lr=9.0226e-05) (hash(x)=28138281)
863
+ 827 train 7.115876 (lr=8.9678e-05) (hash(x)=29038360)
864
+ 828 train 7.086456 (lr=8.9133e-05) (hash(x)=22827903)
865
+ 829 train 7.158072 (lr=8.8591e-05) (hash(x)=24647015)
866
+ 830 train 7.047551 (lr=8.8051e-05) (hash(x)=20477804)
867
+ 831 train 7.023748 (lr=8.7513e-05) (hash(x)=26692640)
868
+ 832 train 7.100020 (lr=8.6978e-05) (hash(x)=27388794)
869
+ 833 train 7.346889 (lr=8.6446e-05) (hash(x)=25265875)
870
+ 834 train 7.018174 (lr=8.5916e-05) (hash(x)=21841888)
871
+ 835 train 7.123296 (lr=8.5389e-05) (hash(x)=19809196)
872
+ 836 train 7.204536 (lr=8.4865e-05) (hash(x)=23007945)
873
+ 837 train 7.118589 (lr=8.4343e-05) (hash(x)=24079409)
874
+ 838 train 7.065272 (lr=8.3824e-05) (hash(x)=25874533)
875
+ 839 train 6.986068 (lr=8.3307e-05) (hash(x)=23660106)
876
+ 840 train 7.457134 (lr=8.2793e-05) (hash(x)=41821329)
877
+ 841 train 7.100047 (lr=8.2282e-05) (hash(x)=25763899)
878
+ 842 train 6.934858 (lr=8.1773e-05) (hash(x)=28584111)
879
+ 843 train 7.364761 (lr=8.1267e-05) (hash(x)=25559887)
880
+ 844 train 7.174936 (lr=8.0764e-05) (hash(x)=26886377)
881
+ 845 train 7.134036 (lr=8.0263e-05) (hash(x)=26461895)
882
+ 846 train 7.175381 (lr=7.9765e-05) (hash(x)=23018827)
883
+ 847 train 6.992710 (lr=7.9270e-05) (hash(x)=24375035)
884
+ 848 train 7.138222 (lr=7.8778e-05) (hash(x)=26431670)
885
+ 849 train 7.007517 (lr=7.8288e-05) (hash(x)=25033511)
886
+ 850 val loss 7.0878
887
+ 850 val perplexity 1197.2623
888
+ 850 train 7.113752 (lr=7.7801e-05) (hash(x)=25602577)
889
+ 851 train 7.061126 (lr=7.7317e-05) (hash(x)=23316582)
890
+ 852 train 7.167081 (lr=7.6836e-05) (hash(x)=22470095)
891
+ 853 train 7.043619 (lr=7.6357e-05) (hash(x)=25089202)
892
+ 854 train 6.993402 (lr=7.5881e-05) (hash(x)=26243247)
893
+ 855 train 7.096493 (lr=7.5408e-05) (hash(x)=22231144)
894
+ 856 train 6.998126 (lr=7.4938e-05) (hash(x)=27646564)
895
+ 857 train 7.158533 (lr=7.4470e-05) (hash(x)=23889875)
896
+ 858 train 7.065603 (lr=7.4005e-05) (hash(x)=25255240)
897
+ 859 train 7.188045 (lr=7.3544e-05) (hash(x)=23883299)
898
+ 860 train 7.142145 (lr=7.3085e-05) (hash(x)=18463841)
899
+ 861 train 7.180216 (lr=7.2628e-05) (hash(x)=30657762)
900
+ 862 train 7.067801 (lr=7.2175e-05) (hash(x)=25288079)
901
+ 863 train 7.122161 (lr=7.1725e-05) (hash(x)=28979310)
902
+ 864 train 7.327227 (lr=7.1277e-05) (hash(x)=30254991)
903
+ 865 train 7.139729 (lr=7.0832e-05) (hash(x)=27274653)
904
+ 866 train 7.131728 (lr=7.0390e-05) (hash(x)=23026066)
905
+ 867 train 7.076217 (lr=6.9952e-05) (hash(x)=14001963)
906
+ 868 train 6.982872 (lr=6.9516e-05) (hash(x)=21988054)
907
+ 869 train 7.095751 (lr=6.9082e-05) (hash(x)=26241099)
908
+ 870 train 7.164630 (lr=6.8652e-05) (hash(x)=26209722)
909
+ 871 train 7.029360 (lr=6.8225e-05) (hash(x)=26928312)
910
+ 872 train 7.103787 (lr=6.7801e-05) (hash(x)=23563272)
911
+ 873 train 7.045320 (lr=6.7379e-05) (hash(x)=25704828)
912
+ 874 train 7.282993 (lr=6.6961e-05) (hash(x)=22498992)
913
+ 875 train 7.432914 (lr=6.6545e-05) (hash(x)=22895918)
914
+ 876 train 6.995058 (lr=6.6133e-05) (hash(x)=24404283)
915
+ 877 train 7.032101 (lr=6.5723e-05) (hash(x)=26663932)
916
+ 878 train 7.154165 (lr=6.5317e-05) (hash(x)=22805316)
917
+ 879 train 7.160897 (lr=6.4913e-05) (hash(x)=23344466)
918
+ 880 train 7.083907 (lr=6.4513e-05) (hash(x)=20966940)
919
+ 881 train 7.082876 (lr=6.4115e-05) (hash(x)=27422664)
920
+ 882 train 7.120934 (lr=6.3721e-05) (hash(x)=22116510)
921
+ 883 train 7.002798 (lr=6.3329e-05) (hash(x)=27269633)
922
+ 884 train 7.109317 (lr=6.2941e-05) (hash(x)=27157711)
923
+ 885 train 7.156308 (lr=6.2556e-05) (hash(x)=22774645)
924
+ 886 train 7.046669 (lr=6.2173e-05) (hash(x)=22143146)
925
+ 887 train 7.060269 (lr=6.1794e-05) (hash(x)=22356438)
926
+ 888 train 6.809245 (lr=6.1418e-05) (hash(x)=23172158)
927
+ 889 train 7.069403 (lr=6.1045e-05) (hash(x)=26971690)
928
+ 890 train 7.110359 (lr=6.0675e-05) (hash(x)=24851922)
929
+ 891 train 6.953431 (lr=6.0308e-05) (hash(x)=21017641)
930
+ 892 train 7.104090 (lr=5.9944e-05) (hash(x)=28186957)
931
+ 893 train 7.104548 (lr=5.9583e-05) (hash(x)=25661082)
932
+ 894 train 7.009439 (lr=5.9225e-05) (hash(x)=25462781)
933
+ 895 train 6.952760 (lr=5.8871e-05) (hash(x)=21774374)
934
+ 896 train 6.967298 (lr=5.8519e-05) (hash(x)=23657055)
935
+ 897 train 7.014244 (lr=5.8171e-05) (hash(x)=23604050)
936
+ 898 train 7.409962 (lr=5.7826e-05) (hash(x)=31534071)
937
+ 899 train 7.205088 (lr=5.7484e-05) (hash(x)=24183732)
938
+ 900 val loss 7.0758
939
+ 900 val perplexity 1183.0345
940
+ 900 train 7.064973 (lr=5.7145e-05) (hash(x)=23384280)
941
+ 901 train 6.933934 (lr=5.6809e-05) (hash(x)=21975928)
942
+ 902 train 6.936323 (lr=5.6476e-05) (hash(x)=25281674)
943
+ 903 train 6.932977 (lr=5.6147e-05) (hash(x)=22947712)
944
+ 904 train 7.008844 (lr=5.5821e-05) (hash(x)=24853977)
945
+ 905 train 7.061584 (lr=5.5497e-05) (hash(x)=23696470)
946
+ 906 train 7.077976 (lr=5.5178e-05) (hash(x)=27489711)
947
+ 907 train 6.995270 (lr=5.4861e-05) (hash(x)=28050163)
948
+ 908 train 6.861397 (lr=5.4547e-05) (hash(x)=26996807)
949
+ 909 train 7.260010 (lr=5.4237e-05) (hash(x)=28626495)
950
+ 910 train 6.999020 (lr=5.3930e-05) (hash(x)=24783242)
951
+ 911 train 6.999202 (lr=5.3626e-05) (hash(x)=25832479)
952
+ 912 train 6.861785 (lr=5.3325e-05) (hash(x)=23522137)
953
+ 913 train 7.183571 (lr=5.3028e-05) (hash(x)=26731973)
954
+ 914 train 7.223503 (lr=5.2734e-05) (hash(x)=26019142)
955
+ 915 train 7.026177 (lr=5.2443e-05) (hash(x)=25081877)
956
+ 916 train 7.116035 (lr=5.2155e-05) (hash(x)=23843571)
957
+ 917 train 7.144435 (lr=5.1871e-05) (hash(x)=26805739)
958
+ 918 train 7.041110 (lr=5.1589e-05) (hash(x)=27169927)
959
+ 919 train 7.094075 (lr=5.1311e-05) (hash(x)=26545491)
960
+ 920 train 7.094398 (lr=5.1037e-05) (hash(x)=25179497)
961
+ 921 train 7.036853 (lr=5.0765e-05) (hash(x)=27085003)
962
+ 922 train 7.053725 (lr=5.0497e-05) (hash(x)=24558202)
963
+ 923 train 7.179406 (lr=5.0232e-05) (hash(x)=25317684)
964
+ 924 train 7.075161 (lr=4.9971e-05) (hash(x)=27104868)
965
+ 925 train 7.118635 (lr=4.9712e-05) (hash(x)=24806274)
966
+ 926 train 7.157857 (lr=4.9457e-05) (hash(x)=24341937)
967
+ 927 train 6.974164 (lr=4.9206e-05) (hash(x)=22736558)
968
+ 928 train 6.992149 (lr=4.8957e-05) (hash(x)=23843744)
969
+ 929 train 7.097245 (lr=4.8712e-05) (hash(x)=27218468)
970
+ 930 train 7.057538 (lr=4.8470e-05) (hash(x)=26042053)
971
+ 931 train 7.081615 (lr=4.8232e-05) (hash(x)=25776082)
972
+ 932 train 6.955910 (lr=4.7997e-05) (hash(x)=21594447)
973
+ 933 train 7.005305 (lr=4.7765e-05) (hash(x)=21674469)
974
+ 934 train 6.932823 (lr=4.7537e-05) (hash(x)=26213966)
975
+ 935 train 6.935889 (lr=4.7312e-05) (hash(x)=23770702)
976
+ 936 train 7.051751 (lr=4.7090e-05) (hash(x)=27829878)
977
+ 937 train 7.144333 (lr=4.6871e-05) (hash(x)=29177278)
978
+ 938 train 7.096163 (lr=4.6656e-05) (hash(x)=26020089)
979
+ 939 train 6.986700 (lr=4.6445e-05) (hash(x)=24676475)
980
+ 940 train 7.036417 (lr=4.6236e-05) (hash(x)=24056174)
981
+ 941 train 7.072096 (lr=4.6031e-05) (hash(x)=21887514)
982
+ 942 train 7.065656 (lr=4.5830e-05) (hash(x)=24196100)
983
+ 943 train 7.050309 (lr=4.5631e-05) (hash(x)=25700093)
984
+ 944 train 7.161655 (lr=4.5437e-05) (hash(x)=24365834)
985
+ 945 train 6.981146 (lr=4.5245e-05) (hash(x)=25631786)
986
+ 946 train 7.118807 (lr=4.5057e-05) (hash(x)=24452345)
987
+ 947 train 6.825552 (lr=4.4872e-05) (hash(x)=21465330)
988
+ 948 train 6.925424 (lr=4.4691e-05) (hash(x)=20427493)
989
+ 949 train 6.988474 (lr=4.4513e-05) (hash(x)=25439720)
990
+ 950 val loss 7.0724
991
+ 950 val perplexity 1178.9203
992
+ 950 train 7.061757 (lr=4.4338e-05) (hash(x)=21652988)
993
+ 951 train 6.994700 (lr=4.4167e-05) (hash(x)=22546779)
994
+ 952 train 6.845499 (lr=4.4000e-05) (hash(x)=20855736)
995
+ 953 train 7.036118 (lr=4.3835e-05) (hash(x)=27725873)
996
+ 954 train 7.119771 (lr=4.3674e-05) (hash(x)=24582989)
997
+ 955 train 6.993190 (lr=4.3517e-05) (hash(x)=24822372)
998
+ 956 train 7.040912 (lr=4.3363e-05) (hash(x)=25624588)
999
+ 957 train 6.999645 (lr=4.3212e-05) (hash(x)=23566489)
1000
+ 958 train 7.162243 (lr=4.3065e-05) (hash(x)=28117387)
1001
+ 959 train 6.944362 (lr=4.2921e-05) (hash(x)=26874283)
1002
+ 960 train 6.958692 (lr=4.2781e-05) (hash(x)=28570123)
1003
+ 961 train 7.074913 (lr=4.2644e-05) (hash(x)=24807017)
1004
+ 962 train 6.980035 (lr=4.2510e-05) (hash(x)=25189993)
1005
+ 963 train 7.099875 (lr=4.2380e-05) (hash(x)=29202226)
1006
+ 964 train 7.281800 (lr=4.2253e-05) (hash(x)=26060713)
1007
+ 965 train 7.193967 (lr=4.2130e-05) (hash(x)=23053285)
1008
+ 966 train 7.117478 (lr=4.2010e-05) (hash(x)=21968481)
1009
+ 967 train 7.025183 (lr=4.1894e-05) (hash(x)=24492698)
1010
+ 968 train 7.011426 (lr=4.1781e-05) (hash(x)=25760806)
1011
+ 969 train 7.059545 (lr=4.1672e-05) (hash(x)=27417392)
1012
+ 970 train 7.025636 (lr=4.1566e-05) (hash(x)=26767105)
1013
+ 971 train 7.087362 (lr=4.1463e-05) (hash(x)=25117678)
1014
+ 972 train 6.983567 (lr=4.1364e-05) (hash(x)=24233758)
1015
+ 973 train 7.046432 (lr=4.1269e-05) (hash(x)=29907634)
1016
+ 974 train 6.841383 (lr=4.1177e-05) (hash(x)=23552436)
1017
+ 975 train 7.170393 (lr=4.1088e-05) (hash(x)=28610406)
1018
+ 976 train 7.020385 (lr=4.1003e-05) (hash(x)=23858212)
1019
+ 977 train 6.927962 (lr=4.0921e-05) (hash(x)=20567764)
1020
+ 978 train 6.934706 (lr=4.0843e-05) (hash(x)=29631106)
1021
+ 979 train 6.912985 (lr=4.0768e-05) (hash(x)=19843564)
1022
+ 980 train 7.043250 (lr=4.0697e-05) (hash(x)=22589891)
1023
+ 981 train 6.849034 (lr=4.0629e-05) (hash(x)=21183943)
1024
+ 982 train 6.989940 (lr=4.0564e-05) (hash(x)=26152241)
1025
+ 983 train 7.015242 (lr=4.0503e-05) (hash(x)=22700785)
1026
+ 984 train 7.207182 (lr=4.0446e-05) (hash(x)=29611280)
1027
+ 985 train 7.280855 (lr=4.0392e-05) (hash(x)=25972823)
1028
+ 986 train 7.029654 (lr=4.0341e-05) (hash(x)=23447782)
1029
+ 987 train 6.975093 (lr=4.0294e-05) (hash(x)=22578453)
1030
+ 988 train 6.988976 (lr=4.0251e-05) (hash(x)=25474622)
1031
+ 989 train 7.044487 (lr=4.0211e-05) (hash(x)=22965762)
1032
+ 990 train 7.102362 (lr=4.0174e-05) (hash(x)=26116321)
1033
+ 991 train 6.996910 (lr=4.0141e-05) (hash(x)=25934549)
1034
+ 992 train 6.923571 (lr=4.0112e-05) (hash(x)=23937495)
1035
+ 993 train 7.108010 (lr=4.0085e-05) (hash(x)=26286442)
1036
+ 994 train 7.343407 (lr=4.0063e-05) (hash(x)=32737711)
1037
+ 995 train 6.961495 (lr=4.0044e-05) (hash(x)=20563613)
1038
+ 996 train 7.215710 (lr=4.0028e-05) (hash(x)=26234770)
1039
+ 997 train 7.129551 (lr=4.0016e-05) (hash(x)=26179422)
1040
+ 998 train 7.048817 (lr=4.0007e-05) (hash(x)=23289651)
1041
+ 999 val loss 7.0692
1042
+ 999 val perplexity 1175.2170
1043
+ 999 train 7.402036 (lr=4.0002e-05) (hash(x)=36436617)
lr4e-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:d8134f4b86bb66f9752946a785bcea2bfa2b3605737a1c6cd02879d61d8b2b22
3
+ size 48156226
lr4e-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 sha256:81001e42f4b049d3d88c78e2a00d99ee53b2b3a75bdc637c471e1fb640f07a7b
3
+ size 71157574