andrew-healey commited on
Commit
e209ba9
·
verified ·
1 Parent(s): 7b296f2

Upload folder using huggingface_hub

Browse files
lr2e-4_total_batch_size10240_seq_len256/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr2e-4_total_batch_size10240_seq_len256", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 256, "batch_size": null, "total_batch_size": 10240, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0002, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr2e-4_total_batch_size10240_seq_len256/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b9fd0bfa27ab267155e5ae2bb7e7ccf3640df9cfec6890472ce93f6d886a52e
3
+ size 964
lr2e-4_total_batch_size10240_seq_len256/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7488
3
+ 0 val perplexity 126607.6172
4
+ 0 train 11.749471 (lr=6.9930e-07) (hash(x)=12683395)
5
+ 1 train 11.743154 (lr=1.3986e-06) (hash(x)=12136211)
6
+ 2 train 11.733978 (lr=2.0979e-06) (hash(x)=13068552)
7
+ 3 train 11.750169 (lr=2.7972e-06) (hash(x)=9915468)
8
+ 4 train 11.752558 (lr=3.4965e-06) (hash(x)=12164183)
9
+ 5 train 11.750552 (lr=4.1958e-06) (hash(x)=13629277)
10
+ 6 train 11.726274 (lr=4.8951e-06) (hash(x)=13413021)
11
+ 7 train 11.725346 (lr=5.5944e-06) (hash(x)=12684830)
12
+ 8 train 11.688498 (lr=6.2937e-06) (hash(x)=13262550)
13
+ 9 train 11.694875 (lr=6.9930e-06) (hash(x)=12003126)
14
+ 10 train 11.672234 (lr=7.6923e-06) (hash(x)=14144015)
15
+ 11 train 11.664067 (lr=8.3916e-06) (hash(x)=12014715)
16
+ 12 train 11.613773 (lr=9.0909e-06) (hash(x)=14912984)
17
+ 13 train 11.589005 (lr=9.7902e-06) (hash(x)=9492167)
18
+ 14 train 11.592712 (lr=1.0490e-05) (hash(x)=15241928)
19
+ 15 train 11.516011 (lr=1.1189e-05) (hash(x)=12595649)
20
+ 16 train 11.510752 (lr=1.1888e-05) (hash(x)=15162520)
21
+ 17 train 11.505452 (lr=1.2587e-05) (hash(x)=11980631)
22
+ 18 train 11.502942 (lr=1.3287e-05) (hash(x)=13172222)
23
+ 19 train 11.474338 (lr=1.3986e-05) (hash(x)=11101031)
24
+ 20 train 11.431174 (lr=1.4685e-05) (hash(x)=13275746)
25
+ 21 train 11.373371 (lr=1.5385e-05) (hash(x)=14316349)
26
+ 22 train 11.361119 (lr=1.6084e-05) (hash(x)=14143687)
27
+ 23 train 11.350359 (lr=1.6783e-05) (hash(x)=12036401)
28
+ 24 train 11.302754 (lr=1.7483e-05) (hash(x)=11143776)
29
+ 25 train 11.254465 (lr=1.8182e-05) (hash(x)=12337125)
30
+ 26 train 11.242458 (lr=1.8881e-05) (hash(x)=10700109)
31
+ 27 train 11.173590 (lr=1.9580e-05) (hash(x)=11823810)
32
+ 28 train 11.164010 (lr=2.0280e-05) (hash(x)=12070591)
33
+ 29 train 11.088997 (lr=2.0979e-05) (hash(x)=13033959)
34
+ 30 train 11.063822 (lr=2.1678e-05) (hash(x)=10786770)
35
+ 31 train 10.988603 (lr=2.2378e-05) (hash(x)=13549925)
36
+ 32 train 10.833302 (lr=2.3077e-05) (hash(x)=14101838)
37
+ 33 train 10.938073 (lr=2.3776e-05) (hash(x)=14220701)
38
+ 34 train 10.810691 (lr=2.4476e-05) (hash(x)=14571180)
39
+ 35 train 10.786020 (lr=2.5175e-05) (hash(x)=13252262)
40
+ 36 train 10.680337 (lr=2.5874e-05) (hash(x)=13676730)
41
+ 37 train 10.682432 (lr=2.6573e-05) (hash(x)=14638145)
42
+ 38 train 10.593128 (lr=2.7273e-05) (hash(x)=14385711)
43
+ 39 train 10.657967 (lr=2.7972e-05) (hash(x)=13749054)
44
+ 40 train 10.635027 (lr=2.8671e-05) (hash(x)=12469703)
45
+ 41 train 10.565075 (lr=2.9371e-05) (hash(x)=12681760)
46
+ 42 train 10.555275 (lr=3.0070e-05) (hash(x)=12423697)
47
+ 43 train 10.442893 (lr=3.0769e-05) (hash(x)=14461706)
48
+ 44 train 10.473587 (lr=3.1469e-05) (hash(x)=13727016)
49
+ 45 train 10.349246 (lr=3.2168e-05) (hash(x)=14235218)
50
+ 46 train 10.411022 (lr=3.2867e-05) (hash(x)=12501568)
51
+ 47 train 10.461914 (lr=3.3566e-05) (hash(x)=11735995)
52
+ 48 train 10.496268 (lr=3.4266e-05) (hash(x)=11569945)
53
+ 49 train 10.336614 (lr=3.4965e-05) (hash(x)=11641112)
54
+ 50 val loss 10.3398
55
+ 50 val perplexity 30940.9590
56
+ 50 train 10.346021 (lr=3.5664e-05) (hash(x)=11536784)
57
+ 51 train 10.397942 (lr=3.6364e-05) (hash(x)=12231701)
58
+ 52 train 10.321541 (lr=3.7063e-05) (hash(x)=13521301)
59
+ 53 train 10.209512 (lr=3.7762e-05) (hash(x)=10175470)
60
+ 54 train 10.311617 (lr=3.8462e-05) (hash(x)=13903066)
61
+ 55 train 10.509489 (lr=3.9161e-05) (hash(x)=14117945)
62
+ 56 train 10.251172 (lr=3.9860e-05) (hash(x)=14851356)
63
+ 57 train 10.279968 (lr=4.0559e-05) (hash(x)=13116544)
64
+ 58 train 10.238316 (lr=4.1259e-05) (hash(x)=16366718)
65
+ 59 train 10.225586 (lr=4.1958e-05) (hash(x)=12678303)
66
+ 60 train 10.124987 (lr=4.2657e-05) (hash(x)=11144768)
67
+ 61 train 10.228955 (lr=4.3357e-05) (hash(x)=11927462)
68
+ 62 train 10.141272 (lr=4.4056e-05) (hash(x)=10418366)
69
+ 63 train 10.209117 (lr=4.4755e-05) (hash(x)=12758461)
70
+ 64 train 10.231982 (lr=4.5455e-05) (hash(x)=11680802)
71
+ 65 train 10.238916 (lr=4.6154e-05) (hash(x)=11946230)
72
+ 66 train 10.147930 (lr=4.6853e-05) (hash(x)=11704519)
73
+ 67 train 10.051651 (lr=4.7552e-05) (hash(x)=13701311)
74
+ 68 train 10.110293 (lr=4.8252e-05) (hash(x)=12751615)
75
+ 69 train 10.110296 (lr=4.8951e-05) (hash(x)=11198382)
76
+ 70 train 10.166554 (lr=4.9650e-05) (hash(x)=16075171)
77
+ 71 train 10.147088 (lr=5.0350e-05) (hash(x)=14346433)
78
+ 72 train 10.122236 (lr=5.1049e-05) (hash(x)=12888806)
79
+ 73 train 10.209002 (lr=5.1748e-05) (hash(x)=17855984)
80
+ 74 train 10.154438 (lr=5.2448e-05) (hash(x)=9616381)
81
+ 75 train 10.320233 (lr=5.3147e-05) (hash(x)=18910052)
82
+ 76 train 10.131425 (lr=5.3846e-05) (hash(x)=12769596)
83
+ 77 train 10.090733 (lr=5.4545e-05) (hash(x)=11885403)
84
+ 78 train 10.086129 (lr=5.5245e-05) (hash(x)=12476514)
85
+ 79 train 10.065707 (lr=5.5944e-05) (hash(x)=12508350)
86
+ 80 train 9.959964 (lr=5.6643e-05) (hash(x)=10745637)
87
+ 81 train 10.015217 (lr=5.7343e-05) (hash(x)=10862804)
88
+ 82 train 10.030217 (lr=5.8042e-05) (hash(x)=15075084)
89
+ 83 train 9.944746 (lr=5.8741e-05) (hash(x)=14043988)
90
+ 84 train 9.816319 (lr=5.9441e-05) (hash(x)=12667924)
91
+ 85 train 9.823545 (lr=6.0140e-05) (hash(x)=11916406)
92
+ 86 train 9.854385 (lr=6.0839e-05) (hash(x)=13406195)
93
+ 87 train 9.873839 (lr=6.1538e-05) (hash(x)=11968227)
94
+ 88 train 9.926235 (lr=6.2238e-05) (hash(x)=11927490)
95
+ 89 train 9.846236 (lr=6.2937e-05) (hash(x)=13290191)
96
+ 90 train 9.819974 (lr=6.3636e-05) (hash(x)=11771253)
97
+ 91 train 9.853929 (lr=6.4336e-05) (hash(x)=15670776)
98
+ 92 train 9.750486 (lr=6.5035e-05) (hash(x)=11306133)
99
+ 93 train 9.847816 (lr=6.5734e-05) (hash(x)=14358384)
100
+ 94 train 9.733951 (lr=6.6434e-05) (hash(x)=10267428)
101
+ 95 train 9.803350 (lr=6.7133e-05) (hash(x)=11289341)
102
+ 96 train 9.725645 (lr=6.7832e-05) (hash(x)=14062293)
103
+ 97 train 9.810717 (lr=6.8531e-05) (hash(x)=14736057)
104
+ 98 train 9.894498 (lr=6.9231e-05) (hash(x)=16036112)
105
+ 99 train 9.869472 (lr=6.9930e-05) (hash(x)=11482477)
106
+ 100 val loss 9.7314
107
+ 100 val perplexity 16837.6836
108
+ 100 train 9.785767 (lr=7.0629e-05) (hash(x)=12742648)
109
+ 101 train 9.693295 (lr=7.1329e-05) (hash(x)=15429973)
110
+ 102 train 9.772370 (lr=7.2028e-05) (hash(x)=13972190)
111
+ 103 train 9.730985 (lr=7.2727e-05) (hash(x)=10345057)
112
+ 104 train 9.695408 (lr=7.3427e-05) (hash(x)=11609727)
113
+ 105 train 9.654381 (lr=7.4126e-05) (hash(x)=10816510)
114
+ 106 train 9.581069 (lr=7.4825e-05) (hash(x)=11801604)
115
+ 107 train 9.672201 (lr=7.5524e-05) (hash(x)=11157124)
116
+ 108 train 9.708123 (lr=7.6224e-05) (hash(x)=11666699)
117
+ 109 train 9.572017 (lr=7.6923e-05) (hash(x)=12349897)
118
+ 110 train 9.613996 (lr=7.7622e-05) (hash(x)=11648826)
119
+ 111 train 9.601323 (lr=7.8322e-05) (hash(x)=10750452)
120
+ 112 train 9.526505 (lr=7.9021e-05) (hash(x)=12114386)
121
+ 113 train 9.608311 (lr=7.9720e-05) (hash(x)=11859950)
122
+ 114 train 9.456744 (lr=8.0420e-05) (hash(x)=12837162)
123
+ 115 train 9.595715 (lr=8.1119e-05) (hash(x)=8921601)
124
+ 116 train 9.352226 (lr=8.1818e-05) (hash(x)=9125058)
125
+ 117 train 9.387005 (lr=8.2517e-05) (hash(x)=11333049)
126
+ 118 train 9.447655 (lr=8.3217e-05) (hash(x)=13807539)
127
+ 119 train 9.432055 (lr=8.3916e-05) (hash(x)=12165500)
128
+ 120 train 9.509651 (lr=8.4615e-05) (hash(x)=14343620)
129
+ 121 train 9.457852 (lr=8.5315e-05) (hash(x)=13153948)
130
+ 122 train 9.449279 (lr=8.6014e-05) (hash(x)=14497853)
131
+ 123 train 9.428129 (lr=8.6713e-05) (hash(x)=13569316)
132
+ 124 train 9.460892 (lr=8.7413e-05) (hash(x)=14212041)
133
+ 125 train 9.453527 (lr=8.8112e-05) (hash(x)=11456278)
134
+ 126 train 9.269216 (lr=8.8811e-05) (hash(x)=13563812)
135
+ 127 train 9.331639 (lr=8.9510e-05) (hash(x)=14043713)
136
+ 128 train 9.361615 (lr=9.0210e-05) (hash(x)=13968786)
137
+ 129 train 9.378619 (lr=9.0909e-05) (hash(x)=11896948)
138
+ 130 train 9.349016 (lr=9.1608e-05) (hash(x)=13298324)
139
+ 131 train 9.330015 (lr=9.2308e-05) (hash(x)=11969650)
140
+ 132 train 9.206083 (lr=9.3007e-05) (hash(x)=12021926)
141
+ 133 train 9.170892 (lr=9.3706e-05) (hash(x)=14387351)
142
+ 134 train 9.185492 (lr=9.4406e-05) (hash(x)=13939547)
143
+ 135 train 8.938717 (lr=9.5105e-05) (hash(x)=13680255)
144
+ 136 train 9.213625 (lr=9.5804e-05) (hash(x)=10982564)
145
+ 137 train 9.315022 (lr=9.6503e-05) (hash(x)=11875534)
146
+ 138 train 9.291365 (lr=9.7203e-05) (hash(x)=13390260)
147
+ 139 train 9.238653 (lr=9.7902e-05) (hash(x)=13871675)
148
+ 140 train 9.082945 (lr=9.8601e-05) (hash(x)=10728444)
149
+ 141 train 9.092201 (lr=9.9301e-05) (hash(x)=13389922)
150
+ 142 train 9.149725 (lr=1.0000e-04) (hash(x)=13458742)
151
+ 143 train 9.079385 (lr=1.0070e-04) (hash(x)=12263431)
152
+ 144 train 9.129459 (lr=1.0140e-04) (hash(x)=12377887)
153
+ 145 train 9.077869 (lr=1.0210e-04) (hash(x)=10141365)
154
+ 146 train 9.085869 (lr=1.0280e-04) (hash(x)=13965689)
155
+ 147 train 9.015737 (lr=1.0350e-04) (hash(x)=12414778)
156
+ 148 train 8.965915 (lr=1.0420e-04) (hash(x)=9727144)
157
+ 149 train 9.016699 (lr=1.0490e-04) (hash(x)=11738134)
158
+ 150 val loss 8.9729
159
+ 150 val perplexity 7886.8335
160
+ 150 train 9.310808 (lr=1.0559e-04) (hash(x)=14924380)
161
+ 151 train 8.940922 (lr=1.0629e-04) (hash(x)=14443665)
162
+ 152 train 9.032997 (lr=1.0699e-04) (hash(x)=12873827)
163
+ 153 train 8.956676 (lr=1.0769e-04) (hash(x)=13435160)
164
+ 154 train 9.031758 (lr=1.0839e-04) (hash(x)=13024409)
165
+ 155 train 9.087739 (lr=1.0909e-04) (hash(x)=13100449)
166
+ 156 train 8.967399 (lr=1.0979e-04) (hash(x)=13304906)
167
+ 157 train 8.855659 (lr=1.1049e-04) (hash(x)=11690244)
168
+ 158 train 8.931254 (lr=1.1119e-04) (hash(x)=12027838)
169
+ 159 train 8.825787 (lr=1.1189e-04) (hash(x)=13953073)
170
+ 160 train 8.777500 (lr=1.1259e-04) (hash(x)=12066345)
171
+ 161 train 8.795609 (lr=1.1329e-04) (hash(x)=12657159)
172
+ 162 train 8.796979 (lr=1.1399e-04) (hash(x)=11335971)
173
+ 163 train 8.819024 (lr=1.1469e-04) (hash(x)=10956041)
174
+ 164 train 8.755186 (lr=1.1538e-04) (hash(x)=9766608)
175
+ 165 train 8.705883 (lr=1.1608e-04) (hash(x)=11060230)
176
+ 166 train 8.891259 (lr=1.1678e-04) (hash(x)=11396268)
177
+ 167 train 8.777832 (lr=1.1748e-04) (hash(x)=12840879)
178
+ 168 train 8.742261 (lr=1.1818e-04) (hash(x)=10987186)
179
+ 169 train 8.717211 (lr=1.1888e-04) (hash(x)=8916599)
180
+ 170 train 8.781445 (lr=1.1958e-04) (hash(x)=13235433)
181
+ 171 train 8.677981 (lr=1.2028e-04) (hash(x)=14428050)
182
+ 172 train 8.653564 (lr=1.2098e-04) (hash(x)=12631329)
183
+ 173 train 8.668625 (lr=1.2168e-04) (hash(x)=12787108)
184
+ 174 train 8.656042 (lr=1.2238e-04) (hash(x)=14499940)
185
+ 175 train 8.745576 (lr=1.2308e-04) (hash(x)=14104576)
186
+ 176 train 8.566192 (lr=1.2378e-04) (hash(x)=13943168)
187
+ 177 train 8.582632 (lr=1.2448e-04) (hash(x)=12623743)
188
+ 178 train 8.758264 (lr=1.2517e-04) (hash(x)=12689741)
189
+ 179 train 8.628689 (lr=1.2587e-04) (hash(x)=14453314)
190
+ 180 train 8.486784 (lr=1.2657e-04) (hash(x)=10824391)
191
+ 181 train 8.394562 (lr=1.2727e-04) (hash(x)=10470424)
192
+ 182 train 8.641977 (lr=1.2797e-04) (hash(x)=15186277)
193
+ 183 train 8.516485 (lr=1.2867e-04) (hash(x)=12918172)
194
+ 184 train 8.478453 (lr=1.2937e-04) (hash(x)=11936107)
195
+ 185 train 8.509807 (lr=1.3007e-04) (hash(x)=13259459)
196
+ 186 train 8.398291 (lr=1.3077e-04) (hash(x)=12918385)
197
+ 187 train 8.579128 (lr=1.3147e-04) (hash(x)=12368572)
198
+ 188 train 8.539043 (lr=1.3217e-04) (hash(x)=13455578)
199
+ 189 train 8.419795 (lr=1.3287e-04) (hash(x)=13505144)
200
+ 190 train 8.475689 (lr=1.3357e-04) (hash(x)=13370946)
201
+ 191 train 8.463377 (lr=1.3427e-04) (hash(x)=13340443)
202
+ 192 train 8.470586 (lr=1.3497e-04) (hash(x)=12806141)
203
+ 193 train 8.438359 (lr=1.3566e-04) (hash(x)=11469900)
204
+ 194 train 8.243909 (lr=1.3636e-04) (hash(x)=11591533)
205
+ 195 train 8.299441 (lr=1.3706e-04) (hash(x)=10742101)
206
+ 196 train 8.323741 (lr=1.3776e-04) (hash(x)=11781196)
207
+ 197 train 8.278149 (lr=1.3846e-04) (hash(x)=13261517)
208
+ 198 train 8.353140 (lr=1.3916e-04) (hash(x)=13592858)
209
+ 199 train 8.450720 (lr=1.3986e-04) (hash(x)=13025742)
210
+ 200 val loss 8.3073
211
+ 200 val perplexity 4053.1631
212
+ 200 train 8.298951 (lr=1.4056e-04) (hash(x)=12152419)
213
+ 201 train 8.552971 (lr=1.4126e-04) (hash(x)=15100390)
214
+ 202 train 8.291928 (lr=1.4196e-04) (hash(x)=10708197)
215
+ 203 train 8.271540 (lr=1.4266e-04) (hash(x)=14351068)
216
+ 204 train 8.405098 (lr=1.4336e-04) (hash(x)=12619936)
217
+ 205 train 8.149250 (lr=1.4406e-04) (hash(x)=13445545)
218
+ 206 train 8.243004 (lr=1.4476e-04) (hash(x)=11375930)
219
+ 207 train 8.327666 (lr=1.4545e-04) (hash(x)=13155663)
220
+ 208 train 8.249639 (lr=1.4615e-04) (hash(x)=13404244)
221
+ 209 train 8.260911 (lr=1.4685e-04) (hash(x)=13309898)
222
+ 210 train 8.274372 (lr=1.4755e-04) (hash(x)=12356526)
223
+ 211 train 8.113143 (lr=1.4825e-04) (hash(x)=9157946)
224
+ 212 train 8.151432 (lr=1.4895e-04) (hash(x)=11837703)
225
+ 213 train 8.072153 (lr=1.4965e-04) (hash(x)=12222186)
226
+ 214 train 8.111135 (lr=1.5035e-04) (hash(x)=12588150)
227
+ 215 train 8.143168 (lr=1.5105e-04) (hash(x)=13872294)
228
+ 216 train 8.026382 (lr=1.5175e-04) (hash(x)=10621792)
229
+ 217 train 8.099649 (lr=1.5245e-04) (hash(x)=11824400)
230
+ 218 train 8.126266 (lr=1.5315e-04) (hash(x)=12495032)
231
+ 219 train 8.119047 (lr=1.5385e-04) (hash(x)=12773612)
232
+ 220 train 8.134309 (lr=1.5455e-04) (hash(x)=10808441)
233
+ 221 train 8.140932 (lr=1.5524e-04) (hash(x)=12794078)
234
+ 222 train 8.026840 (lr=1.5594e-04) (hash(x)=11328897)
235
+ 223 train 8.109459 (lr=1.5664e-04) (hash(x)=12235713)
236
+ 224 train 8.017649 (lr=1.5734e-04) (hash(x)=11512364)
237
+ 225 train 7.994483 (lr=1.5804e-04) (hash(x)=10270693)
238
+ 226 train 7.862049 (lr=1.5874e-04) (hash(x)=12380104)
239
+ 227 train 7.923952 (lr=1.5944e-04) (hash(x)=13686441)
240
+ 228 train 8.323906 (lr=1.6014e-04) (hash(x)=14151557)
241
+ 229 train 8.162612 (lr=1.6084e-04) (hash(x)=13567687)
242
+ 230 train 8.024878 (lr=1.6154e-04) (hash(x)=12718282)
243
+ 231 train 7.979778 (lr=1.6224e-04) (hash(x)=13214962)
244
+ 232 train 7.972181 (lr=1.6294e-04) (hash(x)=10855282)
245
+ 233 train 8.099891 (lr=1.6364e-04) (hash(x)=12404167)
246
+ 234 train 8.173978 (lr=1.6434e-04) (hash(x)=13017256)
247
+ 235 train 7.932282 (lr=1.6503e-04) (hash(x)=12602681)
248
+ 236 train 7.829520 (lr=1.6573e-04) (hash(x)=11916856)
249
+ 237 train 8.061350 (lr=1.6643e-04) (hash(x)=13850551)
250
+ 238 train 7.832509 (lr=1.6713e-04) (hash(x)=11382479)
251
+ 239 train 7.933521 (lr=1.6783e-04) (hash(x)=13829219)
252
+ 240 train 8.083911 (lr=1.6853e-04) (hash(x)=13540459)
253
+ 241 train 7.955428 (lr=1.6923e-04) (hash(x)=15485762)
254
+ 242 train 7.938529 (lr=1.6993e-04) (hash(x)=12268519)
255
+ 243 train 7.577065 (lr=1.7063e-04) (hash(x)=9335220)
256
+ 244 train 7.879964 (lr=1.7133e-04) (hash(x)=13359270)
257
+ 245 train 7.887694 (lr=1.7203e-04) (hash(x)=13162637)
258
+ 246 train 7.972208 (lr=1.7273e-04) (hash(x)=13305408)
259
+ 247 train 7.912307 (lr=1.7343e-04) (hash(x)=15713183)
260
+ 248 train 7.856433 (lr=1.7413e-04) (hash(x)=13536419)
261
+ 249 train 7.872691 (lr=1.7483e-04) (hash(x)=13326615)
262
+ 250 val loss 7.8331
263
+ 250 val perplexity 2522.7397
264
+ 250 train 7.878263 (lr=1.7552e-04) (hash(x)=12112694)
265
+ 251 train 7.890191 (lr=1.7622e-04) (hash(x)=9917297)
266
+ 252 train 7.961455 (lr=1.7692e-04) (hash(x)=12656639)
267
+ 253 train 7.323850 (lr=1.7762e-04) (hash(x)=6829764)
268
+ 254 train 7.700480 (lr=1.7832e-04) (hash(x)=12964320)
269
+ 255 train 7.777606 (lr=1.7902e-04) (hash(x)=12688412)
270
+ 256 train 7.900908 (lr=1.7972e-04) (hash(x)=14906439)
271
+ 257 train 7.523143 (lr=1.8042e-04) (hash(x)=13039934)
272
+ 258 train 7.849216 (lr=1.8112e-04) (hash(x)=12020357)
273
+ 259 train 7.936213 (lr=1.8182e-04) (hash(x)=12108991)
274
+ 260 train 7.919735 (lr=1.8252e-04) (hash(x)=12328329)
275
+ 261 train 7.921086 (lr=1.8322e-04) (hash(x)=14740147)
276
+ 262 train 7.777666 (lr=1.8392e-04) (hash(x)=11559413)
277
+ 263 train 7.848758 (lr=1.8462e-04) (hash(x)=12152216)
278
+ 264 train 7.775743 (lr=1.8531e-04) (hash(x)=12708922)
279
+ 265 train 7.897610 (lr=1.8601e-04) (hash(x)=14469749)
280
+ 266 train 7.897973 (lr=1.8671e-04) (hash(x)=12459368)
281
+ 267 train 7.871258 (lr=1.8741e-04) (hash(x)=11421483)
282
+ 268 train 7.691977 (lr=1.8811e-04) (hash(x)=12939881)
283
+ 269 train 7.851340 (lr=1.8881e-04) (hash(x)=12362234)
284
+ 270 train 7.740550 (lr=1.8951e-04) (hash(x)=12322333)
285
+ 271 train 7.745141 (lr=1.9021e-04) (hash(x)=13448155)
286
+ 272 train 7.779603 (lr=1.9091e-04) (hash(x)=12928941)
287
+ 273 train 7.705258 (lr=1.9161e-04) (hash(x)=10971430)
288
+ 274 train 7.670495 (lr=1.9231e-04) (hash(x)=13627629)
289
+ 275 train 7.703697 (lr=1.9301e-04) (hash(x)=11905081)
290
+ 276 train 7.718877 (lr=1.9371e-04) (hash(x)=14339184)
291
+ 277 train 7.677378 (lr=1.9441e-04) (hash(x)=10917438)
292
+ 278 train 7.818774 (lr=1.9510e-04) (hash(x)=13804603)
293
+ 279 train 7.898653 (lr=1.9580e-04) (hash(x)=12290857)
294
+ 280 train 7.681757 (lr=1.9650e-04) (hash(x)=10211118)
295
+ 281 train 7.984233 (lr=1.9720e-04) (hash(x)=13407076)
296
+ 282 train 7.834126 (lr=1.9790e-04) (hash(x)=12249945)
297
+ 283 train 7.830030 (lr=1.9860e-04) (hash(x)=13234933)
298
+ 284 train 7.730503 (lr=1.9930e-04) (hash(x)=15194855)
299
+ 285 train 7.864285 (lr=2.0000e-04) (hash(x)=12892366)
300
+ 286 train 7.660939 (lr=2.0000e-04) (hash(x)=12609795)
301
+ 287 train 7.609954 (lr=2.0000e-04) (hash(x)=9768085)
302
+ 288 train 7.620918 (lr=2.0000e-04) (hash(x)=12252104)
303
+ 289 train 7.819005 (lr=1.9999e-04) (hash(x)=12692791)
304
+ 290 train 7.760107 (lr=1.9999e-04) (hash(x)=11957245)
305
+ 291 train 7.348352 (lr=1.9998e-04) (hash(x)=9528974)
306
+ 292 train 7.674488 (lr=1.9997e-04) (hash(x)=12231837)
307
+ 293 train 7.631418 (lr=1.9996e-04) (hash(x)=10032842)
308
+ 294 train 8.221061 (lr=1.9994e-04) (hash(x)=17178097)
309
+ 295 train 8.253121 (lr=1.9993e-04) (hash(x)=10720403)
310
+ 296 train 7.862677 (lr=1.9991e-04) (hash(x)=13541352)
311
+ 297 train 8.162098 (lr=1.9989e-04) (hash(x)=15187742)
312
+ 298 train 7.876316 (lr=1.9987e-04) (hash(x)=11971935)
313
+ 299 train 7.784756 (lr=1.9985e-04) (hash(x)=12874592)
314
+ 300 val loss 7.6752
315
+ 300 val perplexity 2154.1558
316
+ 300 train 7.754529 (lr=1.9983e-04) (hash(x)=12462802)
317
+ 301 train 7.805879 (lr=1.9980e-04) (hash(x)=13622545)
318
+ 302 train 7.900079 (lr=1.9978e-04) (hash(x)=14255974)
319
+ 303 train 7.776807 (lr=1.9975e-04) (hash(x)=11025404)
320
+ 304 train 7.872910 (lr=1.9972e-04) (hash(x)=11239636)
321
+ 305 train 7.809615 (lr=1.9969e-04) (hash(x)=15190123)
322
+ 306 train 7.924077 (lr=1.9965e-04) (hash(x)=11464046)
323
+ 307 train 7.851423 (lr=1.9962e-04) (hash(x)=12567779)
324
+ 308 train 7.787735 (lr=1.9958e-04) (hash(x)=16046446)
325
+ 309 train 7.657711 (lr=1.9954e-04) (hash(x)=12439519)
326
+ 310 train 7.541646 (lr=1.9950e-04) (hash(x)=11917658)
327
+ 311 train 7.727854 (lr=1.9946e-04) (hash(x)=11826091)
328
+ 312 train 7.978692 (lr=1.9941e-04) (hash(x)=10807703)
329
+ 313 train 7.658695 (lr=1.9937e-04) (hash(x)=14723866)
330
+ 314 train 7.919335 (lr=1.9932e-04) (hash(x)=11315230)
331
+ 315 train 8.446399 (lr=1.9927e-04) (hash(x)=11259652)
332
+ 316 train 8.421213 (lr=1.9922e-04) (hash(x)=11702582)
333
+ 317 train 7.942480 (lr=1.9916e-04) (hash(x)=11135951)
334
+ 318 train 7.832261 (lr=1.9911e-04) (hash(x)=11570885)
335
+ 319 train 7.695963 (lr=1.9905e-04) (hash(x)=14642293)
336
+ 320 train 7.539567 (lr=1.9899e-04) (hash(x)=13037070)
337
+ 321 train 7.714262 (lr=1.9893e-04) (hash(x)=12485426)
338
+ 322 train 7.774962 (lr=1.9887e-04) (hash(x)=15142358)
339
+ 323 train 7.572247 (lr=1.9881e-04) (hash(x)=12923132)
340
+ 324 train 7.878205 (lr=1.9874e-04) (hash(x)=13055942)
341
+ 325 train 7.769583 (lr=1.9868e-04) (hash(x)=13559280)
342
+ 326 train 7.782434 (lr=1.9861e-04) (hash(x)=15313622)
343
+ 327 train 7.749083 (lr=1.9854e-04) (hash(x)=12169964)
344
+ 328 train 7.348492 (lr=1.9847e-04) (hash(x)=10902629)
345
+ 329 train 7.344389 (lr=1.9839e-04) (hash(x)=12000182)
346
+ 330 train 7.485283 (lr=1.9832e-04) (hash(x)=11345877)
347
+ 331 train 7.628186 (lr=1.9824e-04) (hash(x)=10948640)
348
+ 332 train 7.508961 (lr=1.9816e-04) (hash(x)=13037309)
349
+ 333 train 7.363987 (lr=1.9808e-04) (hash(x)=9666433)
350
+ 334 train 7.635446 (lr=1.9800e-04) (hash(x)=14049895)
351
+ 335 train 7.824716 (lr=1.9792e-04) (hash(x)=14103013)
352
+ 336 train 7.855577 (lr=1.9783e-04) (hash(x)=14058870)
353
+ 337 train 7.836854 (lr=1.9774e-04) (hash(x)=13553506)
354
+ 338 train 7.716927 (lr=1.9765e-04) (hash(x)=12902075)
355
+ 339 train 7.619915 (lr=1.9756e-04) (hash(x)=11135042)
356
+ 340 train 7.965980 (lr=1.9747e-04) (hash(x)=13650023)
357
+ 341 train 7.717850 (lr=1.9738e-04) (hash(x)=14930403)
358
+ 342 train 7.881103 (lr=1.9728e-04) (hash(x)=15485935)
359
+ 343 train 8.141835 (lr=1.9718e-04) (hash(x)=12404200)
360
+ 344 train 7.619013 (lr=1.9709e-04) (hash(x)=12392953)
361
+ 345 train 7.721322 (lr=1.9698e-04) (hash(x)=12280314)
362
+ 346 train 7.593549 (lr=1.9688e-04) (hash(x)=10900559)
363
+ 347 train 7.695595 (lr=1.9678e-04) (hash(x)=11534055)
364
+ 348 train 7.694337 (lr=1.9667e-04) (hash(x)=11997160)
365
+ 349 train 7.812058 (lr=1.9656e-04) (hash(x)=11308531)
366
+ 350 val loss 7.6552
367
+ 350 val perplexity 2111.6155
368
+ 350 train 7.717064 (lr=1.9646e-04) (hash(x)=12759751)
369
+ 351 train 7.693947 (lr=1.9634e-04) (hash(x)=10914085)
370
+ 352 train 7.832214 (lr=1.9623e-04) (hash(x)=12776027)
371
+ 353 train 7.648399 (lr=1.9612e-04) (hash(x)=12357751)
372
+ 354 train 7.628036 (lr=1.9600e-04) (hash(x)=11040798)
373
+ 355 train 7.967459 (lr=1.9588e-04) (hash(x)=14570713)
374
+ 356 train 7.851944 (lr=1.9576e-04) (hash(x)=11492564)
375
+ 357 train 7.687325 (lr=1.9564e-04) (hash(x)=12212797)
376
+ 358 train 7.760355 (lr=1.9552e-04) (hash(x)=13496879)
377
+ 359 train 7.765784 (lr=1.9540e-04) (hash(x)=11680329)
378
+ 360 train 7.634298 (lr=1.9527e-04) (hash(x)=11948575)
379
+ 361 train 7.682464 (lr=1.9514e-04) (hash(x)=13116772)
380
+ 362 train 7.973674 (lr=1.9501e-04) (hash(x)=15527273)
381
+ 363 train 7.423680 (lr=1.9488e-04) (hash(x)=12437096)
382
+ 364 train 7.818903 (lr=1.9475e-04) (hash(x)=12518287)
383
+ 365 train 7.684490 (lr=1.9462e-04) (hash(x)=11220338)
384
+ 366 train 7.535876 (lr=1.9448e-04) (hash(x)=12034192)
385
+ 367 train 7.996837 (lr=1.9434e-04) (hash(x)=12654843)
386
+ 368 train 7.843126 (lr=1.9421e-04) (hash(x)=16116174)
387
+ 369 train 7.777497 (lr=1.9406e-04) (hash(x)=13505076)
388
+ 370 train 7.725749 (lr=1.9392e-04) (hash(x)=11820607)
389
+ 371 train 7.839240 (lr=1.9378e-04) (hash(x)=10606850)
390
+ 372 train 7.657317 (lr=1.9363e-04) (hash(x)=13759635)
391
+ 373 train 7.719068 (lr=1.9349e-04) (hash(x)=15357227)
392
+ 374 train 7.727069 (lr=1.9334e-04) (hash(x)=11597062)
393
+ 375 train 7.668932 (lr=1.9319e-04) (hash(x)=11281537)
394
+ 376 train 7.337694 (lr=1.9304e-04) (hash(x)=13294828)
395
+ 377 train 8.220462 (lr=1.9288e-04) (hash(x)=13253249)
396
+ 378 train 7.828575 (lr=1.9273e-04) (hash(x)=16101250)
397
+ 379 train 7.983093 (lr=1.9257e-04) (hash(x)=12363856)
398
+ 380 train 7.731238 (lr=1.9241e-04) (hash(x)=12206540)
399
+ 381 train 7.916253 (lr=1.9225e-04) (hash(x)=13580623)
400
+ 382 train 7.874535 (lr=1.9209e-04) (hash(x)=12509602)
401
+ 383 train 7.951620 (lr=1.9193e-04) (hash(x)=14001265)
402
+ 384 train 7.673168 (lr=1.9176e-04) (hash(x)=14290431)
403
+ 385 train 7.668699 (lr=1.9160e-04) (hash(x)=13045278)
404
+ 386 train 7.777698 (lr=1.9143e-04) (hash(x)=13371383)
405
+ 387 train 7.732079 (lr=1.9126e-04) (hash(x)=10444420)
406
+ 388 train 7.562488 (lr=1.9109e-04) (hash(x)=12732892)
407
+ 389 train 7.822778 (lr=1.9091e-04) (hash(x)=11832964)
408
+ 390 train 7.807354 (lr=1.9074e-04) (hash(x)=11477958)
409
+ 391 train 7.662290 (lr=1.9056e-04) (hash(x)=12249732)
410
+ 392 train 7.744616 (lr=1.9039e-04) (hash(x)=12132933)
411
+ 393 train 7.696017 (lr=1.9021e-04) (hash(x)=12307821)
412
+ 394 train 7.255061 (lr=1.9003e-04) (hash(x)=11266264)
413
+ 395 train 7.629808 (lr=1.8985e-04) (hash(x)=13803785)
414
+ 396 train 7.710684 (lr=1.8966e-04) (hash(x)=13256773)
415
+ 397 train 7.586926 (lr=1.8948e-04) (hash(x)=12493316)
416
+ 398 train 7.592146 (lr=1.8929e-04) (hash(x)=12991346)
417
+ 399 train 7.718314 (lr=1.8910e-04) (hash(x)=11689978)
418
+ 400 val loss 7.6230
419
+ 400 val perplexity 2044.6552
420
+ 400 train 7.601723 (lr=1.8891e-04) (hash(x)=12056273)
421
+ 401 train 7.734839 (lr=1.8872e-04) (hash(x)=11906763)
422
+ 402 train 7.737982 (lr=1.8853e-04) (hash(x)=12621559)
423
+ 403 train 7.694338 (lr=1.8834e-04) (hash(x)=13865869)
424
+ 404 train 7.733510 (lr=1.8814e-04) (hash(x)=11425874)
425
+ 405 train 8.046168 (lr=1.8794e-04) (hash(x)=16513375)
426
+ 406 train 7.911267 (lr=1.8774e-04) (hash(x)=14492920)
427
+ 407 train 7.771743 (lr=1.8754e-04) (hash(x)=13030618)
428
+ 408 train 7.654598 (lr=1.8734e-04) (hash(x)=12816692)
429
+ 409 train 7.871097 (lr=1.8714e-04) (hash(x)=13701949)
430
+ 410 train 7.654761 (lr=1.8693e-04) (hash(x)=12904618)
431
+ 411 train 7.794955 (lr=1.8673e-04) (hash(x)=14299833)
432
+ 412 train 7.630018 (lr=1.8652e-04) (hash(x)=12586511)
433
+ 413 train 7.681787 (lr=1.8631e-04) (hash(x)=11818854)
434
+ 414 train 7.727222 (lr=1.8610e-04) (hash(x)=13063115)
435
+ 415 train 7.576703 (lr=1.8589e-04) (hash(x)=12908875)
436
+ 416 train 7.811481 (lr=1.8567e-04) (hash(x)=12627550)
437
+ 417 train 7.580979 (lr=1.8546e-04) (hash(x)=12634656)
438
+ 418 train 7.676210 (lr=1.8524e-04) (hash(x)=13466941)
439
+ 419 train 7.913406 (lr=1.8502e-04) (hash(x)=13859304)
440
+ 420 train 7.825897 (lr=1.8480e-04) (hash(x)=9637450)
441
+ 421 train 7.719577 (lr=1.8458e-04) (hash(x)=14940204)
442
+ 422 train 7.671168 (lr=1.8436e-04) (hash(x)=13686717)
443
+ 423 train 7.757048 (lr=1.8414e-04) (hash(x)=14396964)
444
+ 424 train 7.401555 (lr=1.8391e-04) (hash(x)=12949082)
445
+ 425 train 7.609304 (lr=1.8369e-04) (hash(x)=13577749)
446
+ 426 train 7.790442 (lr=1.8346e-04) (hash(x)=13031728)
447
+ 427 train 7.792082 (lr=1.8323e-04) (hash(x)=11269801)
448
+ 428 train 7.522133 (lr=1.8300e-04) (hash(x)=11106556)
449
+ 429 train 7.440247 (lr=1.8276e-04) (hash(x)=11222906)
450
+ 430 train 7.745721 (lr=1.8253e-04) (hash(x)=14052269)
451
+ 431 train 7.567030 (lr=1.8230e-04) (hash(x)=14486071)
452
+ 432 train 7.677009 (lr=1.8206e-04) (hash(x)=13026662)
453
+ 433 train 7.681513 (lr=1.8182e-04) (hash(x)=11065744)
454
+ 434 train 7.729194 (lr=1.8158e-04) (hash(x)=13318947)
455
+ 435 train 7.747689 (lr=1.8134e-04) (hash(x)=11997361)
456
+ 436 train 7.789652 (lr=1.8110e-04) (hash(x)=12437178)
457
+ 437 train 7.607648 (lr=1.8086e-04) (hash(x)=12705537)
458
+ 438 train 7.761497 (lr=1.8061e-04) (hash(x)=15790088)
459
+ 439 train 7.531714 (lr=1.8036e-04) (hash(x)=11010336)
460
+ 440 train 7.776014 (lr=1.8012e-04) (hash(x)=11401216)
461
+ 441 train 7.691857 (lr=1.7987e-04) (hash(x)=15106564)
462
+ 442 train 7.675643 (lr=1.7962e-04) (hash(x)=13019227)
463
+ 443 train 7.672286 (lr=1.7937e-04) (hash(x)=10864248)
464
+ 444 train 7.583384 (lr=1.7911e-04) (hash(x)=12107995)
465
+ 445 train 7.473984 (lr=1.7886e-04) (hash(x)=13284954)
466
+ 446 train 7.731310 (lr=1.7860e-04) (hash(x)=12430135)
467
+ 447 train 7.663692 (lr=1.7835e-04) (hash(x)=11906151)
468
+ 448 train 7.507871 (lr=1.7809e-04) (hash(x)=11697970)
469
+ 449 train 7.425570 (lr=1.7783e-04) (hash(x)=8629295)
470
+ 450 val loss 7.5794
471
+ 450 val perplexity 1957.3666
472
+ 450 train 7.641700 (lr=1.7757e-04) (hash(x)=13977261)
473
+ 451 train 7.680037 (lr=1.7731e-04) (hash(x)=15166897)
474
+ 452 train 7.755672 (lr=1.7704e-04) (hash(x)=13444474)
475
+ 453 train 7.802022 (lr=1.7678e-04) (hash(x)=13108020)
476
+ 454 train 7.577708 (lr=1.7651e-04) (hash(x)=12559086)
477
+ 455 train 7.525719 (lr=1.7624e-04) (hash(x)=12338963)
478
+ 456 train 7.423385 (lr=1.7597e-04) (hash(x)=10411127)
479
+ 457 train 7.785961 (lr=1.7570e-04) (hash(x)=17176736)
480
+ 458 train 7.546532 (lr=1.7543e-04) (hash(x)=13586859)
481
+ 459 train 7.706274 (lr=1.7516e-04) (hash(x)=13235312)
482
+ 460 train 7.539670 (lr=1.7489e-04) (hash(x)=10862031)
483
+ 461 train 7.679858 (lr=1.7461e-04) (hash(x)=12498316)
484
+ 462 train 7.669574 (lr=1.7434e-04) (hash(x)=15109534)
485
+ 463 train 7.454619 (lr=1.7406e-04) (hash(x)=11031343)
486
+ 464 train 7.567317 (lr=1.7378e-04) (hash(x)=13030101)
487
+ 465 train 7.521578 (lr=1.7350e-04) (hash(x)=12787203)
488
+ 466 train 7.663948 (lr=1.7322e-04) (hash(x)=13192715)
489
+ 467 train 7.653156 (lr=1.7294e-04) (hash(x)=14597757)
490
+ 468 train 7.609468 (lr=1.7265e-04) (hash(x)=11548718)
491
+ 469 train 7.610665 (lr=1.7237e-04) (hash(x)=12017158)
492
+ 470 train 7.655280 (lr=1.7208e-04) (hash(x)=10164479)
493
+ 471 train 7.550844 (lr=1.7179e-04) (hash(x)=13571036)
494
+ 472 train 7.540527 (lr=1.7151e-04) (hash(x)=11609529)
495
+ 473 train 7.649351 (lr=1.7122e-04) (hash(x)=12289875)
496
+ 474 train 7.739728 (lr=1.7092e-04) (hash(x)=13086437)
497
+ 475 train 7.893467 (lr=1.7063e-04) (hash(x)=13225100)
498
+ 476 train 7.461298 (lr=1.7034e-04) (hash(x)=12863127)
499
+ 477 train 7.614952 (lr=1.7004e-04) (hash(x)=11162887)
500
+ 478 train 7.721237 (lr=1.6975e-04) (hash(x)=12364743)
501
+ 479 train 7.490405 (lr=1.6945e-04) (hash(x)=12281916)
502
+ 480 train 7.557831 (lr=1.6915e-04) (hash(x)=12249214)
503
+ 481 train 7.658423 (lr=1.6886e-04) (hash(x)=12348477)
504
+ 482 train 7.681965 (lr=1.6856e-04) (hash(x)=12468719)
505
+ 483 train 7.662239 (lr=1.6825e-04) (hash(x)=13634084)
506
+ 484 train 7.618810 (lr=1.6795e-04) (hash(x)=11740722)
507
+ 485 train 7.747998 (lr=1.6765e-04) (hash(x)=11378506)
508
+ 486 train 7.489720 (lr=1.6734e-04) (hash(x)=13417634)
509
+ 487 train 7.698050 (lr=1.6704e-04) (hash(x)=11801649)
510
+ 488 train 7.705902 (lr=1.6673e-04) (hash(x)=10917325)
511
+ 489 train 7.476423 (lr=1.6642e-04) (hash(x)=10552788)
512
+ 490 train 7.248501 (lr=1.6611e-04) (hash(x)=9421716)
513
+ 491 train 7.611014 (lr=1.6580e-04) (hash(x)=11746289)
514
+ 492 train 7.598160 (lr=1.6549e-04) (hash(x)=9589864)
515
+ 493 train 7.538913 (lr=1.6518e-04) (hash(x)=13826293)
516
+ 494 train 7.737317 (lr=1.6487e-04) (hash(x)=12720603)
517
+ 495 train 7.600025 (lr=1.6455e-04) (hash(x)=10783313)
518
+ 496 train 7.647229 (lr=1.6424e-04) (hash(x)=13077539)
519
+ 497 train 7.681565 (lr=1.6392e-04) (hash(x)=12913270)
520
+ 498 train 7.622911 (lr=1.6360e-04) (hash(x)=11228160)
521
+ 499 train 7.527336 (lr=1.6328e-04) (hash(x)=14212467)
522
+ 500 val loss 7.5543
523
+ 500 val perplexity 1908.9769
524
+ 500 train 7.570513 (lr=1.6296e-04) (hash(x)=12714536)
525
+ 501 train 7.470831 (lr=1.6264e-04) (hash(x)=11495162)
526
+ 502 train 7.630595 (lr=1.6232e-04) (hash(x)=13571819)
527
+ 503 train 7.649988 (lr=1.6200e-04) (hash(x)=12994231)
528
+ 504 train 7.573901 (lr=1.6168e-04) (hash(x)=12786058)
529
+ 505 train 7.823289 (lr=1.6135e-04) (hash(x)=15319532)
530
+ 506 train 7.710766 (lr=1.6103e-04) (hash(x)=12432415)
531
+ 507 train 7.569649 (lr=1.6070e-04) (hash(x)=11485137)
532
+ 508 train 7.689421 (lr=1.6037e-04) (hash(x)=13106040)
533
+ 509 train 7.669937 (lr=1.6004e-04) (hash(x)=14010329)
534
+ 510 train 7.666778 (lr=1.5971e-04) (hash(x)=14627424)
535
+ 511 train 7.506310 (lr=1.5938e-04) (hash(x)=10470487)
536
+ 512 train 7.738064 (lr=1.5905e-04) (hash(x)=12735976)
537
+ 513 train 7.789033 (lr=1.5872e-04) (hash(x)=14077502)
538
+ 514 train 7.621404 (lr=1.5838e-04) (hash(x)=13758986)
539
+ 515 train 7.654015 (lr=1.5805e-04) (hash(x)=12205695)
540
+ 516 train 7.548759 (lr=1.5772e-04) (hash(x)=12101077)
541
+ 517 train 7.454276 (lr=1.5738e-04) (hash(x)=12096952)
542
+ 518 train 7.803391 (lr=1.5704e-04) (hash(x)=10823414)
543
+ 519 train 7.603261 (lr=1.5670e-04) (hash(x)=13985657)
544
+ 520 train 7.593972 (lr=1.5636e-04) (hash(x)=13657835)
545
+ 521 train 7.641784 (lr=1.5602e-04) (hash(x)=14868589)
546
+ 522 train 7.808705 (lr=1.5568e-04) (hash(x)=12440659)
547
+ 523 train 7.641877 (lr=1.5534e-04) (hash(x)=12944090)
548
+ 524 train 7.585098 (lr=1.5500e-04) (hash(x)=14152385)
549
+ 525 train 7.535565 (lr=1.5466e-04) (hash(x)=14652164)
550
+ 526 train 7.673131 (lr=1.5431e-04) (hash(x)=11671931)
551
+ 527 train 7.530089 (lr=1.5397e-04) (hash(x)=12831661)
552
+ 528 train 7.734603 (lr=1.5362e-04) (hash(x)=13612890)
553
+ 529 train 7.485812 (lr=1.5327e-04) (hash(x)=13472998)
554
+ 530 train 7.770911 (lr=1.5293e-04) (hash(x)=12912339)
555
+ 531 train 7.555176 (lr=1.5258e-04) (hash(x)=12806095)
556
+ 532 train 7.402361 (lr=1.5223e-04) (hash(x)=10738555)
557
+ 533 train 7.414106 (lr=1.5188e-04) (hash(x)=13410818)
558
+ 534 train 7.756829 (lr=1.5153e-04) (hash(x)=13997116)
559
+ 535 train 7.544933 (lr=1.5118e-04) (hash(x)=14617414)
560
+ 536 train 7.716651 (lr=1.5082e-04) (hash(x)=12788088)
561
+ 537 train 7.648048 (lr=1.5047e-04) (hash(x)=15066658)
562
+ 538 train 7.600406 (lr=1.5012e-04) (hash(x)=13331118)
563
+ 539 train 7.616180 (lr=1.4976e-04) (hash(x)=13221393)
564
+ 540 train 7.602322 (lr=1.4941e-04) (hash(x)=13355413)
565
+ 541 train 7.478397 (lr=1.4905e-04) (hash(x)=11528563)
566
+ 542 train 7.650199 (lr=1.4869e-04) (hash(x)=15327421)
567
+ 543 train 7.756787 (lr=1.4833e-04) (hash(x)=11041023)
568
+ 544 train 7.464303 (lr=1.4798e-04) (hash(x)=13600125)
569
+ 545 train 7.615798 (lr=1.4762e-04) (hash(x)=14885399)
570
+ 546 train 7.506554 (lr=1.4726e-04) (hash(x)=14786577)
571
+ 547 train 7.634468 (lr=1.4690e-04) (hash(x)=11362709)
572
+ 548 train 7.541002 (lr=1.4653e-04) (hash(x)=13426228)
573
+ 549 train 7.477359 (lr=1.4617e-04) (hash(x)=13083412)
574
+ 550 val loss 7.4923
575
+ 550 val perplexity 1794.0919
576
+ 550 train 7.516134 (lr=1.4581e-04) (hash(x)=13873840)
577
+ 551 train 7.484634 (lr=1.4545e-04) (hash(x)=13186073)
578
+ 552 train 7.605412 (lr=1.4508e-04) (hash(x)=13488437)
579
+ 553 train 7.538831 (lr=1.4472e-04) (hash(x)=13988224)
580
+ 554 train 7.577554 (lr=1.4435e-04) (hash(x)=11786791)
581
+ 555 train 8.368612 (lr=1.4398e-04) (hash(x)=13274177)
582
+ 556 train 7.658959 (lr=1.4362e-04) (hash(x)=10204377)
583
+ 557 train 7.594363 (lr=1.4325e-04) (hash(x)=12363265)
584
+ 558 train 7.387480 (lr=1.4288e-04) (hash(x)=12918001)
585
+ 559 train 7.346763 (lr=1.4251e-04) (hash(x)=13507868)
586
+ 560 train 7.622850 (lr=1.4214e-04) (hash(x)=10580307)
587
+ 561 train 7.439390 (lr=1.4177e-04) (hash(x)=12531978)
588
+ 562 train 7.410974 (lr=1.4140e-04) (hash(x)=11570432)
589
+ 563 train 7.540801 (lr=1.4103e-04) (hash(x)=12512404)
590
+ 564 train 7.618252 (lr=1.4066e-04) (hash(x)=12011112)
591
+ 565 train 7.626017 (lr=1.4029e-04) (hash(x)=13439040)
592
+ 566 train 7.248649 (lr=1.3991e-04) (hash(x)=9321383)
593
+ 567 train 7.522358 (lr=1.3954e-04) (hash(x)=11221721)
594
+ 568 train 7.361130 (lr=1.3916e-04) (hash(x)=12048885)
595
+ 569 train 7.616123 (lr=1.3879e-04) (hash(x)=13854047)
596
+ 570 train 7.794003 (lr=1.3841e-04) (hash(x)=13041387)
597
+ 571 train 7.997252 (lr=1.3804e-04) (hash(x)=15372354)
598
+ 572 train 7.666510 (lr=1.3766e-04) (hash(x)=11895595)
599
+ 573 train 7.448882 (lr=1.3728e-04) (hash(x)=13070460)
600
+ 574 train 7.505824 (lr=1.3691e-04) (hash(x)=11507165)
601
+ 575 train 7.828852 (lr=1.3653e-04) (hash(x)=14689778)
602
+ 576 train 7.261027 (lr=1.3615e-04) (hash(x)=10149008)
603
+ 577 train 7.677993 (lr=1.3577e-04) (hash(x)=15364407)
604
+ 578 train 7.442108 (lr=1.3539e-04) (hash(x)=11676391)
605
+ 579 train 7.433780 (lr=1.3501e-04) (hash(x)=13156509)
606
+ 580 train 7.558657 (lr=1.3463e-04) (hash(x)=12911319)
607
+ 581 train 7.323328 (lr=1.3425e-04) (hash(x)=13358322)
608
+ 582 train 7.361831 (lr=1.3387e-04) (hash(x)=11462782)
609
+ 583 train 7.804766 (lr=1.3348e-04) (hash(x)=15734207)
610
+ 584 train 7.416653 (lr=1.3310e-04) (hash(x)=9117677)
611
+ 585 train 7.547590 (lr=1.3272e-04) (hash(x)=12334129)
612
+ 586 train 7.617244 (lr=1.3234e-04) (hash(x)=14097844)
613
+ 587 train 7.466152 (lr=1.3195e-04) (hash(x)=13758140)
614
+ 588 train 7.829444 (lr=1.3157e-04) (hash(x)=15284545)
615
+ 589 train 7.340835 (lr=1.3118e-04) (hash(x)=11809021)
616
+ 590 train 7.529563 (lr=1.3080e-04) (hash(x)=12752849)
617
+ 591 train 7.441621 (lr=1.3041e-04) (hash(x)=12925955)
618
+ 592 train 7.549499 (lr=1.3003e-04) (hash(x)=12026092)
619
+ 593 train 7.636655 (lr=1.2964e-04) (hash(x)=15708918)
620
+ 594 train 7.888101 (lr=1.2925e-04) (hash(x)=13954437)
621
+ 595 train 7.624602 (lr=1.2887e-04) (hash(x)=12750849)
622
+ 596 train 7.422124 (lr=1.2848e-04) (hash(x)=12076716)
623
+ 597 train 7.560092 (lr=1.2809e-04) (hash(x)=12269439)
624
+ 598 train 7.492734 (lr=1.2770e-04) (hash(x)=9706332)
625
+ 599 train 7.425945 (lr=1.2732e-04) (hash(x)=14327624)
626
+ 600 val loss 7.4601
627
+ 600 val perplexity 1737.4041
628
+ 600 train 7.599716 (lr=1.2693e-04) (hash(x)=12001611)
629
+ 601 train 7.489417 (lr=1.2654e-04) (hash(x)=12664301)
630
+ 602 train 7.398048 (lr=1.2615e-04) (hash(x)=10457844)
631
+ 603 train 7.372234 (lr=1.2576e-04) (hash(x)=11819313)
632
+ 604 train 7.397732 (lr=1.2537e-04) (hash(x)=12304915)
633
+ 605 train 7.517466 (lr=1.2498e-04) (hash(x)=13250863)
634
+ 606 train 7.493772 (lr=1.2459e-04) (hash(x)=15885206)
635
+ 607 train 7.246219 (lr=1.2420e-04) (hash(x)=12559422)
636
+ 608 train 7.422775 (lr=1.2381e-04) (hash(x)=12136733)
637
+ 609 train 7.524293 (lr=1.2341e-04) (hash(x)=12183244)
638
+ 610 train 7.620301 (lr=1.2302e-04) (hash(x)=12682752)
639
+ 611 train 7.653262 (lr=1.2263e-04) (hash(x)=12816359)
640
+ 612 train 7.585143 (lr=1.2224e-04) (hash(x)=11628021)
641
+ 613 train 7.494077 (lr=1.2185e-04) (hash(x)=11408330)
642
+ 614 train 7.528630 (lr=1.2145e-04) (hash(x)=12901064)
643
+ 615 train 7.494481 (lr=1.2106e-04) (hash(x)=10162896)
644
+ 616 train 7.606795 (lr=1.2067e-04) (hash(x)=12416355)
645
+ 617 train 7.828153 (lr=1.2027e-04) (hash(x)=12904775)
646
+ 618 train 7.518283 (lr=1.1988e-04) (hash(x)=12697259)
647
+ 619 train 7.363958 (lr=1.1949e-04) (hash(x)=12041552)
648
+ 620 train 7.413340 (lr=1.1909e-04) (hash(x)=13429059)
649
+ 621 train 7.460439 (lr=1.1870e-04) (hash(x)=14176315)
650
+ 622 train 7.409813 (lr=1.1830e-04) (hash(x)=11386804)
651
+ 623 train 7.301888 (lr=1.1791e-04) (hash(x)=10256290)
652
+ 624 train 7.442668 (lr=1.1752e-04) (hash(x)=11669288)
653
+ 625 train 7.697031 (lr=1.1712e-04) (hash(x)=12287950)
654
+ 626 train 7.335518 (lr=1.1673e-04) (hash(x)=12384271)
655
+ 627 train 7.476335 (lr=1.1633e-04) (hash(x)=14312559)
656
+ 628 train 7.380740 (lr=1.1594e-04) (hash(x)=13864960)
657
+ 629 train 7.550436 (lr=1.1554e-04) (hash(x)=10078167)
658
+ 630 train 7.492968 (lr=1.1515e-04) (hash(x)=14121709)
659
+ 631 train 7.278607 (lr=1.1475e-04) (hash(x)=11443472)
660
+ 632 train 7.521988 (lr=1.1435e-04) (hash(x)=11816027)
661
+ 633 train 7.399258 (lr=1.1396e-04) (hash(x)=11630498)
662
+ 634 train 7.611786 (lr=1.1356e-04) (hash(x)=12453204)
663
+ 635 train 7.342435 (lr=1.1317e-04) (hash(x)=10480150)
664
+ 636 train 7.466508 (lr=1.1277e-04) (hash(x)=10854685)
665
+ 637 train 7.517278 (lr=1.1238e-04) (hash(x)=13365961)
666
+ 638 train 7.782605 (lr=1.1198e-04) (hash(x)=13227220)
667
+ 639 train 7.598751 (lr=1.1158e-04) (hash(x)=13927238)
668
+ 640 train 7.426774 (lr=1.1119e-04) (hash(x)=11183563)
669
+ 641 train 7.626401 (lr=1.1079e-04) (hash(x)=12661142)
670
+ 642 train 7.347363 (lr=1.1040e-04) (hash(x)=12116909)
671
+ 643 train 7.748536 (lr=1.1000e-04) (hash(x)=16448258)
672
+ 644 train 7.973157 (lr=1.0960e-04) (hash(x)=11230472)
673
+ 645 train 7.579286 (lr=1.0921e-04) (hash(x)=12689745)
674
+ 646 train 7.744392 (lr=1.0881e-04) (hash(x)=12535866)
675
+ 647 train 7.273182 (lr=1.0842e-04) (hash(x)=10701487)
676
+ 648 train 7.436188 (lr=1.0802e-04) (hash(x)=12572372)
677
+ 649 train 8.038168 (lr=1.0762e-04) (hash(x)=13938367)
678
+ 650 val loss 7.4305
679
+ 650 val perplexity 1686.6033
680
+ 650 train 7.608813 (lr=1.0723e-04) (hash(x)=15954548)
681
+ 651 train 7.711209 (lr=1.0683e-04) (hash(x)=13034568)
682
+ 652 train 7.613459 (lr=1.0644e-04) (hash(x)=17184553)
683
+ 653 train 7.756634 (lr=1.0604e-04) (hash(x)=13398831)
684
+ 654 train 7.880007 (lr=1.0565e-04) (hash(x)=15245462)
685
+ 655 train 7.703962 (lr=1.0525e-04) (hash(x)=14575913)
686
+ 656 train 7.482484 (lr=1.0485e-04) (hash(x)=11389089)
687
+ 657 train 7.525179 (lr=1.0446e-04) (hash(x)=12121790)
688
+ 658 train 7.650610 (lr=1.0406e-04) (hash(x)=11991841)
689
+ 659 train 7.206368 (lr=1.0367e-04) (hash(x)=10875163)
690
+ 660 train 7.519822 (lr=1.0327e-04) (hash(x)=13641028)
691
+ 661 train 7.614529 (lr=1.0288e-04) (hash(x)=13782256)
692
+ 662 train 8.227384 (lr=1.0248e-04) (hash(x)=16715370)
693
+ 663 train 7.855869 (lr=1.0209e-04) (hash(x)=13744329)
694
+ 664 train 7.518046 (lr=1.0170e-04) (hash(x)=12865885)
695
+ 665 train 7.596542 (lr=1.0130e-04) (hash(x)=14443216)
696
+ 666 train 7.639358 (lr=1.0091e-04) (hash(x)=10361766)
697
+ 667 train 7.238644 (lr=1.0051e-04) (hash(x)=8860388)
698
+ 668 train 7.471066 (lr=1.0012e-04) (hash(x)=12188781)
699
+ 669 train 7.421829 (lr=9.9726e-05) (hash(x)=13546558)
700
+ 670 train 7.391385 (lr=9.9333e-05) (hash(x)=14731723)
701
+ 671 train 7.573414 (lr=9.8940e-05) (hash(x)=14097210)
702
+ 672 train 7.870723 (lr=9.8547e-05) (hash(x)=13277664)
703
+ 673 train 7.264497 (lr=9.8154e-05) (hash(x)=12154683)
704
+ 674 train 7.289332 (lr=9.7762e-05) (hash(x)=14244571)
705
+ 675 train 7.573891 (lr=9.7370e-05) (hash(x)=11088127)
706
+ 676 train 7.335888 (lr=9.6978e-05) (hash(x)=10259005)
707
+ 677 train 7.780912 (lr=9.6586e-05) (hash(x)=15853795)
708
+ 678 train 7.296978 (lr=9.6195e-05) (hash(x)=12406215)
709
+ 679 train 7.385936 (lr=9.5804e-05) (hash(x)=14363085)
710
+ 680 train 7.274235 (lr=9.5413e-05) (hash(x)=12303156)
711
+ 681 train 7.441944 (lr=9.5022e-05) (hash(x)=12144036)
712
+ 682 train 7.636589 (lr=9.4632e-05) (hash(x)=14153850)
713
+ 683 train 7.522573 (lr=9.4242e-05) (hash(x)=14133983)
714
+ 684 train 7.485146 (lr=9.3852e-05) (hash(x)=13407720)
715
+ 685 train 7.557484 (lr=9.3463e-05) (hash(x)=13632329)
716
+ 686 train 7.453802 (lr=9.3073e-05) (hash(x)=11013484)
717
+ 687 train 7.312720 (lr=9.2685e-05) (hash(x)=9539498)
718
+ 688 train 7.326020 (lr=9.2296e-05) (hash(x)=11548592)
719
+ 689 train 7.376711 (lr=9.1908e-05) (hash(x)=11883263)
720
+ 690 train 7.525008 (lr=9.1520e-05) (hash(x)=17040815)
721
+ 691 train 8.096370 (lr=9.1133e-05) (hash(x)=17793549)
722
+ 692 train 7.550653 (lr=9.0746e-05) (hash(x)=13566067)
723
+ 693 train 7.425389 (lr=9.0359e-05) (hash(x)=11871877)
724
+ 694 train 7.461133 (lr=8.9973e-05) (hash(x)=13936711)
725
+ 695 train 7.517392 (lr=8.9587e-05) (hash(x)=14163458)
726
+ 696 train 7.400500 (lr=8.9202e-05) (hash(x)=9756608)
727
+ 697 train 7.811378 (lr=8.8817e-05) (hash(x)=15049180)
728
+ 698 train 7.468044 (lr=8.8432e-05) (hash(x)=12741546)
729
+ 699 train 7.545247 (lr=8.8048e-05) (hash(x)=14609724)
730
+ 700 val loss 7.3995
731
+ 700 val perplexity 1635.0869
732
+ 700 train 7.520119 (lr=8.7664e-05) (hash(x)=13165136)
733
+ 701 train 7.323648 (lr=8.7281e-05) (hash(x)=11157433)
734
+ 702 train 7.382147 (lr=8.6898e-05) (hash(x)=10769955)
735
+ 703 train 7.803646 (lr=8.6515e-05) (hash(x)=18154448)
736
+ 704 train 7.864119 (lr=8.6133e-05) (hash(x)=11566844)
737
+ 705 train 8.030075 (lr=8.5751e-05) (hash(x)=10541575)
738
+ 706 train 7.605852 (lr=8.5370e-05) (hash(x)=13400633)
739
+ 707 train 7.451710 (lr=8.4990e-05) (hash(x)=11995482)
740
+ 708 train 7.629969 (lr=8.4610e-05) (hash(x)=12559360)
741
+ 709 train 7.430182 (lr=8.4230e-05) (hash(x)=11153006)
742
+ 710 train 8.020091 (lr=8.3851e-05) (hash(x)=14803712)
743
+ 711 train 7.735259 (lr=8.3472e-05) (hash(x)=13713689)
744
+ 712 train 7.376389 (lr=8.3094e-05) (hash(x)=12883531)
745
+ 713 train 7.888739 (lr=8.2716e-05) (hash(x)=15113554)
746
+ 714 train 7.339868 (lr=8.2339e-05) (hash(x)=11466714)
747
+ 715 train 7.357756 (lr=8.1963e-05) (hash(x)=11555030)
748
+ 716 train 7.468003 (lr=8.1587e-05) (hash(x)=11661705)
749
+ 717 train 7.808087 (lr=8.1211e-05) (hash(x)=13374040)
750
+ 718 train 7.434581 (lr=8.0836e-05) (hash(x)=11884576)
751
+ 719 train 7.400000 (lr=8.0462e-05) (hash(x)=14115772)
752
+ 720 train 7.509164 (lr=8.0088e-05) (hash(x)=14042374)
753
+ 721 train 7.609578 (lr=7.9715e-05) (hash(x)=9720467)
754
+ 722 train 7.501239 (lr=7.9342e-05) (hash(x)=13709799)
755
+ 723 train 7.393279 (lr=7.8970e-05) (hash(x)=12279208)
756
+ 724 train 7.455206 (lr=7.8599e-05) (hash(x)=12168977)
757
+ 725 train 7.259744 (lr=7.8228e-05) (hash(x)=11796701)
758
+ 726 train 7.273384 (lr=7.7858e-05) (hash(x)=12054640)
759
+ 727 train 7.317337 (lr=7.7488e-05) (hash(x)=12021395)
760
+ 728 train 7.516764 (lr=7.7119e-05) (hash(x)=11650114)
761
+ 729 train 7.254666 (lr=7.6751e-05) (hash(x)=10845152)
762
+ 730 train 7.299281 (lr=7.6383e-05) (hash(x)=10962390)
763
+ 731 train 7.495566 (lr=7.6016e-05) (hash(x)=12617648)
764
+ 732 train 7.215428 (lr=7.5650e-05) (hash(x)=9908464)
765
+ 733 train 7.227355 (lr=7.5284e-05) (hash(x)=11035304)
766
+ 734 train 7.406920 (lr=7.4919e-05) (hash(x)=11048281)
767
+ 735 train 7.201939 (lr=7.4555e-05) (hash(x)=12798942)
768
+ 736 train 7.456601 (lr=7.4191e-05) (hash(x)=11858915)
769
+ 737 train 7.492944 (lr=7.3828e-05) (hash(x)=11344910)
770
+ 738 train 7.382166 (lr=7.3466e-05) (hash(x)=11509808)
771
+ 739 train 7.487852 (lr=7.3105e-05) (hash(x)=12366544)
772
+ 740 train 7.460274 (lr=7.2744e-05) (hash(x)=11976631)
773
+ 741 train 7.268960 (lr=7.2384e-05) (hash(x)=11321027)
774
+ 742 train 7.426065 (lr=7.2024e-05) (hash(x)=11893948)
775
+ 743 train 7.120966 (lr=7.1666e-05) (hash(x)=10886073)
776
+ 744 train 7.291236 (lr=7.1308e-05) (hash(x)=9205900)
777
+ 745 train 7.275642 (lr=7.0950e-05) (hash(x)=12334480)
778
+ 746 train 7.485856 (lr=7.0594e-05) (hash(x)=13094093)
779
+ 747 train 7.561396 (lr=7.0238e-05) (hash(x)=12555076)
780
+ 748 train 7.579319 (lr=6.9884e-05) (hash(x)=14919237)
781
+ 749 train 7.451693 (lr=6.9529e-05) (hash(x)=13525646)
782
+ 750 val loss 7.3783
783
+ 750 val perplexity 1600.8113
784
+ 750 train 7.411161 (lr=6.9176e-05) (hash(x)=12350457)
785
+ 751 train 7.368320 (lr=6.8824e-05) (hash(x)=14777879)
786
+ 752 train 7.358210 (lr=6.8472e-05) (hash(x)=9165650)
787
+ 753 train 7.818675 (lr=6.8121e-05) (hash(x)=12138318)
788
+ 754 train 7.585203 (lr=6.7771e-05) (hash(x)=12066198)
789
+ 755 train 7.135248 (lr=6.7422e-05) (hash(x)=9927010)
790
+ 756 train 7.266870 (lr=6.7073e-05) (hash(x)=10963177)
791
+ 757 train 7.253885 (lr=6.6725e-05) (hash(x)=10265465)
792
+ 758 train 7.277007 (lr=6.6379e-05) (hash(x)=10012601)
793
+ 759 train 7.379737 (lr=6.6033e-05) (hash(x)=13073054)
794
+ 760 train 7.186854 (lr=6.5688e-05) (hash(x)=8375643)
795
+ 761 train 6.860710 (lr=6.5343e-05) (hash(x)=10517302)
796
+ 762 train 7.185311 (lr=6.5000e-05) (hash(x)=11617446)
797
+ 763 train 7.440364 (lr=6.4657e-05) (hash(x)=14967464)
798
+ 764 train 7.336786 (lr=6.4316e-05) (hash(x)=11293886)
799
+ 765 train 7.199153 (lr=6.3975e-05) (hash(x)=12190853)
800
+ 766 train 7.219638 (lr=6.3635e-05) (hash(x)=10723695)
801
+ 767 train 7.390001 (lr=6.3296e-05) (hash(x)=13463198)
802
+ 768 train 7.564269 (lr=6.2958e-05) (hash(x)=13293049)
803
+ 769 train 7.347086 (lr=6.2621e-05) (hash(x)=13635928)
804
+ 770 train 7.537520 (lr=6.2285e-05) (hash(x)=11492108)
805
+ 771 train 7.352237 (lr=6.1950e-05) (hash(x)=12209984)
806
+ 772 train 7.469516 (lr=6.1615e-05) (hash(x)=13817436)
807
+ 773 train 7.348662 (lr=6.1282e-05) (hash(x)=11946913)
808
+ 774 train 7.438359 (lr=6.0949e-05) (hash(x)=13760019)
809
+ 775 train 7.396285 (lr=6.0618e-05) (hash(x)=11983929)
810
+ 776 train 7.265125 (lr=6.0287e-05) (hash(x)=12740415)
811
+ 777 train 7.470258 (lr=5.9958e-05) (hash(x)=13044287)
812
+ 778 train 7.372878 (lr=5.9629e-05) (hash(x)=12306640)
813
+ 779 train 7.773452 (lr=5.9301e-05) (hash(x)=13502211)
814
+ 780 train 7.266268 (lr=5.8974e-05) (hash(x)=11651077)
815
+ 781 train 7.347181 (lr=5.8649e-05) (hash(x)=11732764)
816
+ 782 train 7.169496 (lr=5.8324e-05) (hash(x)=12023079)
817
+ 783 train 7.459864 (lr=5.8000e-05) (hash(x)=12563030)
818
+ 784 train 7.358162 (lr=5.7678e-05) (hash(x)=11093401)
819
+ 785 train 7.399489 (lr=5.7356e-05) (hash(x)=13370486)
820
+ 786 train 7.462607 (lr=5.7035e-05) (hash(x)=13064477)
821
+ 787 train 7.465228 (lr=5.6716e-05) (hash(x)=12774663)
822
+ 788 train 7.293483 (lr=5.6397e-05) (hash(x)=10977883)
823
+ 789 train 7.486957 (lr=5.6079e-05) (hash(x)=12581170)
824
+ 790 train 7.365511 (lr=5.5763e-05) (hash(x)=11710735)
825
+ 791 train 7.307434 (lr=5.5447e-05) (hash(x)=12170905)
826
+ 792 train 7.533897 (lr=5.5133e-05) (hash(x)=11984555)
827
+ 793 train 7.483794 (lr=5.4820e-05) (hash(x)=13035845)
828
+ 794 train 7.201138 (lr=5.4507e-05) (hash(x)=13998448)
829
+ 795 train 7.437726 (lr=5.4196e-05) (hash(x)=13348784)
830
+ 796 train 7.338566 (lr=5.3886e-05) (hash(x)=11535637)
831
+ 797 train 7.291408 (lr=5.3577e-05) (hash(x)=14533849)
832
+ 798 train 7.304417 (lr=5.3269e-05) (hash(x)=12593307)
833
+ 799 train 7.674043 (lr=5.2962e-05) (hash(x)=11525433)
834
+ 800 val loss 7.3533
835
+ 800 val perplexity 1561.3518
836
+ 800 train 7.480927 (lr=5.2656e-05) (hash(x)=9198860)
837
+ 801 train 7.429761 (lr=5.2352e-05) (hash(x)=7779807)
838
+ 802 train 7.492553 (lr=5.2048e-05) (hash(x)=8327978)
839
+ 803 train 7.289938 (lr=5.1746e-05) (hash(x)=9800296)
840
+ 804 train 7.448369 (lr=5.1444e-05) (hash(x)=10408800)
841
+ 805 train 7.502490 (lr=5.1144e-05) (hash(x)=11761671)
842
+ 806 train 7.314196 (lr=5.0845e-05) (hash(x)=10137737)
843
+ 807 train 7.400724 (lr=5.0547e-05) (hash(x)=10168268)
844
+ 808 train 7.424397 (lr=5.0251e-05) (hash(x)=15372601)
845
+ 809 train 7.290130 (lr=4.9955e-05) (hash(x)=10272263)
846
+ 810 train 7.237645 (lr=4.9661e-05) (hash(x)=10560999)
847
+ 811 train 7.314849 (lr=4.9367e-05) (hash(x)=10276843)
848
+ 812 train 8.512241 (lr=4.9075e-05) (hash(x)=12452448)
849
+ 813 train 7.550393 (lr=4.8784e-05) (hash(x)=11364751)
850
+ 814 train 7.097547 (lr=4.8495e-05) (hash(x)=11485502)
851
+ 815 train 7.150331 (lr=4.8206e-05) (hash(x)=12497404)
852
+ 816 train 7.088206 (lr=4.7919e-05) (hash(x)=10013569)
853
+ 817 train 7.711329 (lr=4.7633e-05) (hash(x)=14966009)
854
+ 818 train 7.375296 (lr=4.7348e-05) (hash(x)=11562145)
855
+ 819 train 7.433075 (lr=4.7064e-05) (hash(x)=12932339)
856
+ 820 train 7.365604 (lr=4.6782e-05) (hash(x)=11563553)
857
+ 821 train 7.390022 (lr=4.6501e-05) (hash(x)=8563011)
858
+ 822 train 7.509526 (lr=4.6221e-05) (hash(x)=13409258)
859
+ 823 train 7.410320 (lr=4.5942e-05) (hash(x)=14203584)
860
+ 824 train 7.177043 (lr=4.5664e-05) (hash(x)=10360049)
861
+ 825 train 7.281394 (lr=4.5388e-05) (hash(x)=10373543)
862
+ 826 train 7.323854 (lr=4.5113e-05) (hash(x)=12645249)
863
+ 827 train 7.444703 (lr=4.4839e-05) (hash(x)=13921321)
864
+ 828 train 7.530418 (lr=4.4567e-05) (hash(x)=10868669)
865
+ 829 train 7.218770 (lr=4.4295e-05) (hash(x)=12997326)
866
+ 830 train 7.556257 (lr=4.4025e-05) (hash(x)=13553993)
867
+ 831 train 7.349843 (lr=4.3757e-05) (hash(x)=10946961)
868
+ 832 train 7.284000 (lr=4.3489e-05) (hash(x)=13714094)
869
+ 833 train 7.685631 (lr=4.3223e-05) (hash(x)=11317673)
870
+ 834 train 7.548755 (lr=4.2958e-05) (hash(x)=11555387)
871
+ 835 train 7.525081 (lr=4.2695e-05) (hash(x)=12168723)
872
+ 836 train 7.562858 (lr=4.2432e-05) (hash(x)=11310063)
873
+ 837 train 7.339433 (lr=4.2171e-05) (hash(x)=11165585)
874
+ 838 train 7.219427 (lr=4.1912e-05) (hash(x)=10049422)
875
+ 839 train 7.264568 (lr=4.1654e-05) (hash(x)=10942254)
876
+ 840 train 7.399611 (lr=4.1397e-05) (hash(x)=12383062)
877
+ 841 train 7.343269 (lr=4.1141e-05) (hash(x)=12927776)
878
+ 842 train 7.700893 (lr=4.0887e-05) (hash(x)=12835199)
879
+ 843 train 7.903824 (lr=4.0634e-05) (hash(x)=10371632)
880
+ 844 train 7.698252 (lr=4.0382e-05) (hash(x)=15118958)
881
+ 845 train 7.754153 (lr=4.0132e-05) (hash(x)=12057844)
882
+ 846 train 7.304929 (lr=3.9883e-05) (hash(x)=12276438)
883
+ 847 train 7.685699 (lr=3.9635e-05) (hash(x)=16703203)
884
+ 848 train 7.361772 (lr=3.9389e-05) (hash(x)=12033983)
885
+ 849 train 7.339394 (lr=3.9144e-05) (hash(x)=12424204)
886
+ 850 val loss 7.3460
887
+ 850 val perplexity 1549.9402
888
+ 850 train 7.334066 (lr=3.8901e-05) (hash(x)=11830025)
889
+ 851 train 7.695447 (lr=3.8659e-05) (hash(x)=17013185)
890
+ 852 train 7.716277 (lr=3.8418e-05) (hash(x)=16760833)
891
+ 853 train 7.571669 (lr=3.8178e-05) (hash(x)=11881739)
892
+ 854 train 7.542478 (lr=3.7941e-05) (hash(x)=13649855)
893
+ 855 train 7.355089 (lr=3.7704e-05) (hash(x)=15000294)
894
+ 856 train 7.301039 (lr=3.7469e-05) (hash(x)=11412424)
895
+ 857 train 7.359219 (lr=3.7235e-05) (hash(x)=13008387)
896
+ 858 train 7.317033 (lr=3.7003e-05) (hash(x)=10899640)
897
+ 859 train 7.171286 (lr=3.6772e-05) (hash(x)=10834161)
898
+ 860 train 7.124207 (lr=3.6542e-05) (hash(x)=10472237)
899
+ 861 train 7.621085 (lr=3.6314e-05) (hash(x)=15743796)
900
+ 862 train 7.447645 (lr=3.6088e-05) (hash(x)=10842000)
901
+ 863 train 7.452590 (lr=3.5862e-05) (hash(x)=14862449)
902
+ 864 train 7.916103 (lr=3.5639e-05) (hash(x)=15511051)
903
+ 865 train 7.265149 (lr=3.5416e-05) (hash(x)=11257626)
904
+ 866 train 7.786394 (lr=3.5195e-05) (hash(x)=17810791)
905
+ 867 train 7.196777 (lr=3.4976e-05) (hash(x)=10386532)
906
+ 868 train 7.076453 (lr=3.4758e-05) (hash(x)=12466709)
907
+ 869 train 7.274563 (lr=3.4541e-05) (hash(x)=11765736)
908
+ 870 train 7.596333 (lr=3.4326e-05) (hash(x)=12448217)
909
+ 871 train 7.316454 (lr=3.4113e-05) (hash(x)=13806874)
910
+ 872 train 7.474932 (lr=3.3900e-05) (hash(x)=12238204)
911
+ 873 train 7.367246 (lr=3.3690e-05) (hash(x)=12702206)
912
+ 874 train 7.529489 (lr=3.3480e-05) (hash(x)=13884223)
913
+ 875 train 7.297134 (lr=3.3273e-05) (hash(x)=12733894)
914
+ 876 train 7.515083 (lr=3.3066e-05) (hash(x)=7848745)
915
+ 877 train 7.645507 (lr=3.2862e-05) (hash(x)=13860937)
916
+ 878 train 7.831485 (lr=3.2658e-05) (hash(x)=14870569)
917
+ 879 train 7.486482 (lr=3.2457e-05) (hash(x)=12259111)
918
+ 880 train 7.315836 (lr=3.2256e-05) (hash(x)=12747643)
919
+ 881 train 7.429464 (lr=3.2058e-05) (hash(x)=14710730)
920
+ 882 train 7.379239 (lr=3.1860e-05) (hash(x)=12185557)
921
+ 883 train 7.339972 (lr=3.1665e-05) (hash(x)=10003332)
922
+ 884 train 7.142712 (lr=3.1471e-05) (hash(x)=8591840)
923
+ 885 train 7.457528 (lr=3.1278e-05) (hash(x)=13675590)
924
+ 886 train 6.886081 (lr=3.1087e-05) (hash(x)=11265029)
925
+ 887 train 7.294426 (lr=3.0897e-05) (hash(x)=11837867)
926
+ 888 train 7.812868 (lr=3.0709e-05) (hash(x)=11888580)
927
+ 889 train 7.609653 (lr=3.0522e-05) (hash(x)=11296484)
928
+ 890 train 7.313854 (lr=3.0337e-05) (hash(x)=12991870)
929
+ 891 train 7.457551 (lr=3.0154e-05) (hash(x)=13840537)
930
+ 892 train 7.202431 (lr=2.9972e-05) (hash(x)=12800791)
931
+ 893 train 7.545238 (lr=2.9791e-05) (hash(x)=14920737)
932
+ 894 train 7.456173 (lr=2.9613e-05) (hash(x)=13950257)
933
+ 895 train 7.246991 (lr=2.9435e-05) (hash(x)=12159808)
934
+ 896 train 7.445202 (lr=2.9260e-05) (hash(x)=12734109)
935
+ 897 train 7.236135 (lr=2.9085e-05) (hash(x)=9163004)
936
+ 898 train 7.272314 (lr=2.8913e-05) (hash(x)=11771048)
937
+ 899 train 7.508772 (lr=2.8742e-05) (hash(x)=13797226)
938
+ 900 val loss 7.3290
939
+ 900 val perplexity 1523.8252
940
+ 900 train 7.405778 (lr=2.8572e-05) (hash(x)=12813851)
941
+ 901 train 7.294621 (lr=2.8404e-05) (hash(x)=10204611)
942
+ 902 train 7.431230 (lr=2.8238e-05) (hash(x)=11849134)
943
+ 903 train 7.439422 (lr=2.8073e-05) (hash(x)=12372219)
944
+ 904 train 7.466256 (lr=2.7910e-05) (hash(x)=14633473)
945
+ 905 train 7.304804 (lr=2.7749e-05) (hash(x)=12434908)
946
+ 906 train 7.503487 (lr=2.7589e-05) (hash(x)=11743620)
947
+ 907 train 7.228284 (lr=2.7430e-05) (hash(x)=11371435)
948
+ 908 train 7.407928 (lr=2.7274e-05) (hash(x)=12483864)
949
+ 909 train 7.426461 (lr=2.7119e-05) (hash(x)=13523826)
950
+ 910 train 7.297338 (lr=2.6965e-05) (hash(x)=11575838)
951
+ 911 train 7.297334 (lr=2.6813e-05) (hash(x)=10910611)
952
+ 912 train 7.257138 (lr=2.6663e-05) (hash(x)=10322819)
953
+ 913 train 7.282458 (lr=2.6514e-05) (hash(x)=12905087)
954
+ 914 train 7.508273 (lr=2.6367e-05) (hash(x)=11755845)
955
+ 915 train 7.585574 (lr=2.6221e-05) (hash(x)=10985712)
956
+ 916 train 7.114016 (lr=2.6077e-05) (hash(x)=10264591)
957
+ 917 train 7.572395 (lr=2.5935e-05) (hash(x)=13245256)
958
+ 918 train 7.296788 (lr=2.5795e-05) (hash(x)=13080408)
959
+ 919 train 7.365932 (lr=2.5656e-05) (hash(x)=12509276)
960
+ 920 train 7.305144 (lr=2.5518e-05) (hash(x)=12427196)
961
+ 921 train 7.393879 (lr=2.5383e-05) (hash(x)=13538047)
962
+ 922 train 7.478158 (lr=2.5249e-05) (hash(x)=16129945)
963
+ 923 train 7.667432 (lr=2.5116e-05) (hash(x)=12953787)
964
+ 924 train 7.382128 (lr=2.4985e-05) (hash(x)=9473654)
965
+ 925 train 7.342725 (lr=2.4856e-05) (hash(x)=9468319)
966
+ 926 train 7.406217 (lr=2.4729e-05) (hash(x)=13193021)
967
+ 927 train 7.526223 (lr=2.4603e-05) (hash(x)=12720900)
968
+ 928 train 7.260283 (lr=2.4479e-05) (hash(x)=10634942)
969
+ 929 train 7.226396 (lr=2.4356e-05) (hash(x)=12030112)
970
+ 930 train 7.358836 (lr=2.4235e-05) (hash(x)=12212081)
971
+ 931 train 7.196365 (lr=2.4116e-05) (hash(x)=16075757)
972
+ 932 train 7.981443 (lr=2.3998e-05) (hash(x)=16126508)
973
+ 933 train 7.060184 (lr=2.3883e-05) (hash(x)=11540359)
974
+ 934 train 7.327433 (lr=2.3768e-05) (hash(x)=13428693)
975
+ 935 train 7.543718 (lr=2.3656e-05) (hash(x)=14048369)
976
+ 936 train 7.670028 (lr=2.3545e-05) (hash(x)=12554502)
977
+ 937 train 7.589133 (lr=2.3436e-05) (hash(x)=12961504)
978
+ 938 train 7.784477 (lr=2.3328e-05) (hash(x)=12343533)
979
+ 939 train 7.640431 (lr=2.3222e-05) (hash(x)=12602604)
980
+ 940 train 7.452373 (lr=2.3118e-05) (hash(x)=11612970)
981
+ 941 train 7.642474 (lr=2.3016e-05) (hash(x)=14173792)
982
+ 942 train 7.743091 (lr=2.2915e-05) (hash(x)=15456213)
983
+ 943 train 7.243293 (lr=2.2816e-05) (hash(x)=11552959)
984
+ 944 train 7.278809 (lr=2.2718e-05) (hash(x)=9453679)
985
+ 945 train 7.415085 (lr=2.2623e-05) (hash(x)=11986216)
986
+ 946 train 7.328720 (lr=2.2528e-05) (hash(x)=11356627)
987
+ 947 train 7.506425 (lr=2.2436e-05) (hash(x)=9890481)
988
+ 948 train 7.267688 (lr=2.2345e-05) (hash(x)=11347234)
989
+ 949 train 7.207678 (lr=2.2256e-05) (hash(x)=13052540)
990
+ 950 val loss 7.3231
991
+ 950 val perplexity 1514.8345
992
+ 950 train 7.199020 (lr=2.2169e-05) (hash(x)=13065639)
993
+ 951 train 7.486315 (lr=2.2084e-05) (hash(x)=13423367)
994
+ 952 train 7.360673 (lr=2.2000e-05) (hash(x)=13054517)
995
+ 953 train 7.377185 (lr=2.1918e-05) (hash(x)=12494434)
996
+ 954 train 7.175545 (lr=2.1837e-05) (hash(x)=12496694)
997
+ 955 train 7.179671 (lr=2.1758e-05) (hash(x)=11555634)
998
+ 956 train 7.172164 (lr=2.1681e-05) (hash(x)=10983636)
999
+ 957 train 7.622031 (lr=2.1606e-05) (hash(x)=17108448)
1000
+ 958 train 8.127798 (lr=2.1532e-05) (hash(x)=15985570)
1001
+ 959 train 8.135619 (lr=2.1461e-05) (hash(x)=14914789)
1002
+ 960 train 8.048985 (lr=2.1390e-05) (hash(x)=13097897)
1003
+ 961 train 8.072018 (lr=2.1322e-05) (hash(x)=16237319)
1004
+ 962 train 8.021012 (lr=2.1255e-05) (hash(x)=16105534)
1005
+ 963 train 8.051115 (lr=2.1190e-05) (hash(x)=16763637)
1006
+ 964 train 7.994876 (lr=2.1127e-05) (hash(x)=16598501)
1007
+ 965 train 8.005061 (lr=2.1065e-05) (hash(x)=15222896)
1008
+ 966 train 8.045486 (lr=2.1005e-05) (hash(x)=14919621)
1009
+ 967 train 8.099416 (lr=2.0947e-05) (hash(x)=15783017)
1010
+ 968 train 7.970376 (lr=2.0891e-05) (hash(x)=15370617)
1011
+ 969 train 8.179298 (lr=2.0836e-05) (hash(x)=17102921)
1012
+ 970 train 7.981600 (lr=2.0783e-05) (hash(x)=15358790)
1013
+ 971 train 8.135964 (lr=2.0732e-05) (hash(x)=16354494)
1014
+ 972 train 8.044673 (lr=2.0682e-05) (hash(x)=13580270)
1015
+ 973 train 7.450547 (lr=2.0634e-05) (hash(x)=14615340)
1016
+ 974 train 7.453959 (lr=2.0588e-05) (hash(x)=12022804)
1017
+ 975 train 7.568706 (lr=2.0544e-05) (hash(x)=13512761)
1018
+ 976 train 7.840421 (lr=2.0501e-05) (hash(x)=16681653)
1019
+ 977 train 7.491522 (lr=2.0460e-05) (hash(x)=10750782)
1020
+ 978 train 7.417721 (lr=2.0421e-05) (hash(x)=12643825)
1021
+ 979 train 7.254063 (lr=2.0384e-05) (hash(x)=13751537)
1022
+ 980 train 7.433456 (lr=2.0348e-05) (hash(x)=12575386)
1023
+ 981 train 7.356341 (lr=2.0314e-05) (hash(x)=12343566)
1024
+ 982 train 7.370157 (lr=2.0282e-05) (hash(x)=11685246)
1025
+ 983 train 7.270071 (lr=2.0252e-05) (hash(x)=13830852)
1026
+ 984 train 7.601840 (lr=2.0223e-05) (hash(x)=11998502)
1027
+ 985 train 7.271432 (lr=2.0196e-05) (hash(x)=11170832)
1028
+ 986 train 7.265491 (lr=2.0171e-05) (hash(x)=11482404)
1029
+ 987 train 7.421940 (lr=2.0147e-05) (hash(x)=14841133)
1030
+ 988 train 7.535845 (lr=2.0125e-05) (hash(x)=14960072)
1031
+ 989 train 7.444242 (lr=2.0105e-05) (hash(x)=13683107)
1032
+ 990 train 7.408623 (lr=2.0087e-05) (hash(x)=11884964)
1033
+ 991 train 7.484094 (lr=2.0071e-05) (hash(x)=11832803)
1034
+ 992 train 7.187645 (lr=2.0056e-05) (hash(x)=11624359)
1035
+ 993 train 7.281477 (lr=2.0043e-05) (hash(x)=11362245)
1036
+ 994 train 7.420945 (lr=2.0031e-05) (hash(x)=13335953)
1037
+ 995 train 10.674305 (lr=2.0022e-05) (hash(x)=27572323)
1038
+ 996 train 11.717047 (lr=2.0014e-05) (hash(x)=30981287)
1039
+ 997 train 11.691801 (lr=2.0008e-05) (hash(x)=36479466)
1040
+ 998 train 12.197124 (lr=2.0003e-05) (hash(x)=34231765)
1041
+ 999 val loss 7.3216
1042
+ 999 val perplexity 1512.6554
1043
+ 999 train 11.088883 (lr=2.0001e-05) (hash(x)=31819669)
lr2e-4_total_batch_size10240_seq_len256/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75e3ac46053da95522b01ecf3ae3631ad35817fa65c44577e8261cdffd288a84
3
+ size 38587970
lr2e-4_total_batch_size10240_seq_len256/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89cf77c8b8e796feca265433c9f68c2e6d841963b3cb749e704b0eeabaf1b126
3
+ size 70895430