andrew-healey commited on
Commit
7b296f2
·
verified ·
1 Parent(s): 027decb

Upload folder using huggingface_hub

Browse files
lr2e-4_total_batch_size10240_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr2e-4_total_batch_size10240_seq_len128", "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": 128, "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_len128/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_len128/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.6870
3
+ 0 val perplexity 119016.5000
4
+ 0 train 11.691129 (lr=6.9930e-07) (hash(x)=6605789)
5
+ 1 train 11.712670 (lr=1.3986e-06) (hash(x)=5113283)
6
+ 2 train 11.681951 (lr=2.0979e-06) (hash(x)=7007748)
7
+ 3 train 11.681428 (lr=2.7972e-06) (hash(x)=5534142)
8
+ 4 train 11.676117 (lr=3.4965e-06) (hash(x)=5974274)
9
+ 5 train 11.688631 (lr=4.1958e-06) (hash(x)=6763231)
10
+ 6 train 11.647173 (lr=4.8951e-06) (hash(x)=6706513)
11
+ 7 train 11.656890 (lr=5.5944e-06) (hash(x)=6328012)
12
+ 8 train 11.612038 (lr=6.2937e-06) (hash(x)=6183012)
13
+ 9 train 11.613967 (lr=6.9930e-06) (hash(x)=7052161)
14
+ 10 train 11.573435 (lr=7.6923e-06) (hash(x)=6586466)
15
+ 11 train 11.584437 (lr=8.3916e-06) (hash(x)=5556098)
16
+ 12 train 11.578300 (lr=9.0909e-06) (hash(x)=8155572)
17
+ 13 train 11.502565 (lr=9.7902e-06) (hash(x)=4696943)
18
+ 14 train 11.522008 (lr=1.0490e-05) (hash(x)=7649389)
19
+ 15 train 11.438739 (lr=1.1189e-05) (hash(x)=6219138)
20
+ 16 train 11.398885 (lr=1.1888e-05) (hash(x)=7075794)
21
+ 17 train 11.364608 (lr=1.2587e-05) (hash(x)=6049878)
22
+ 18 train 11.379427 (lr=1.3287e-05) (hash(x)=6385927)
23
+ 19 train 11.317413 (lr=1.3986e-05) (hash(x)=5491972)
24
+ 20 train 11.283849 (lr=1.4685e-05) (hash(x)=6061420)
25
+ 21 train 11.232837 (lr=1.5385e-05) (hash(x)=8075458)
26
+ 22 train 11.113583 (lr=1.6084e-05) (hash(x)=7238769)
27
+ 23 train 11.181250 (lr=1.6783e-05) (hash(x)=5917741)
28
+ 24 train 11.120145 (lr=1.7483e-05) (hash(x)=5430605)
29
+ 25 train 11.099159 (lr=1.8182e-05) (hash(x)=5750403)
30
+ 26 train 11.028784 (lr=1.8881e-05) (hash(x)=4792640)
31
+ 27 train 10.992192 (lr=1.9580e-05) (hash(x)=6112458)
32
+ 28 train 10.907507 (lr=2.0280e-05) (hash(x)=5801122)
33
+ 29 train 10.963168 (lr=2.0979e-05) (hash(x)=5806490)
34
+ 30 train 10.852732 (lr=2.1678e-05) (hash(x)=5754140)
35
+ 31 train 10.756038 (lr=2.2378e-05) (hash(x)=6568379)
36
+ 32 train 10.679668 (lr=2.3077e-05) (hash(x)=7134932)
37
+ 33 train 10.753081 (lr=2.3776e-05) (hash(x)=7212403)
38
+ 34 train 10.635824 (lr=2.4476e-05) (hash(x)=7720703)
39
+ 35 train 10.654977 (lr=2.5175e-05) (hash(x)=7386082)
40
+ 36 train 10.564970 (lr=2.5874e-05) (hash(x)=6848513)
41
+ 37 train 10.580398 (lr=2.6573e-05) (hash(x)=6359743)
42
+ 38 train 10.476182 (lr=2.7273e-05) (hash(x)=7536538)
43
+ 39 train 10.555054 (lr=2.7972e-05) (hash(x)=6151674)
44
+ 40 train 10.517968 (lr=2.8671e-05) (hash(x)=6317099)
45
+ 41 train 10.469595 (lr=2.9371e-05) (hash(x)=7168704)
46
+ 42 train 10.508398 (lr=3.0070e-05) (hash(x)=6736092)
47
+ 43 train 10.408889 (lr=3.0769e-05) (hash(x)=7859060)
48
+ 44 train 10.441653 (lr=3.1469e-05) (hash(x)=6958223)
49
+ 45 train 10.315385 (lr=3.2168e-05) (hash(x)=6976239)
50
+ 46 train 10.358286 (lr=3.2867e-05) (hash(x)=6956845)
51
+ 47 train 10.420304 (lr=3.3566e-05) (hash(x)=6391506)
52
+ 48 train 10.483868 (lr=3.4266e-05) (hash(x)=6113701)
53
+ 49 train 10.268287 (lr=3.4965e-05) (hash(x)=5245999)
54
+ 50 val loss 10.3391
55
+ 50 val perplexity 30917.8633
56
+ 50 train 10.320937 (lr=3.5664e-05) (hash(x)=5774189)
57
+ 51 train 10.347459 (lr=3.6364e-05) (hash(x)=5985675)
58
+ 52 train 10.295653 (lr=3.7063e-05) (hash(x)=6451776)
59
+ 53 train 10.193882 (lr=3.7762e-05) (hash(x)=4355733)
60
+ 54 train 10.290896 (lr=3.8462e-05) (hash(x)=6669188)
61
+ 55 train 10.562056 (lr=3.9161e-05) (hash(x)=6984772)
62
+ 56 train 10.224245 (lr=3.9860e-05) (hash(x)=7440580)
63
+ 57 train 10.288433 (lr=4.0559e-05) (hash(x)=6244096)
64
+ 58 train 10.316704 (lr=4.1259e-05) (hash(x)=8381560)
65
+ 59 train 10.238070 (lr=4.1958e-05) (hash(x)=6603717)
66
+ 60 train 10.173225 (lr=4.2657e-05) (hash(x)=6437516)
67
+ 61 train 10.179184 (lr=4.3357e-05) (hash(x)=5669522)
68
+ 62 train 10.116738 (lr=4.4056e-05) (hash(x)=5505889)
69
+ 63 train 10.178967 (lr=4.4755e-05) (hash(x)=6190579)
70
+ 64 train 10.185919 (lr=4.5455e-05) (hash(x)=5153196)
71
+ 65 train 10.215465 (lr=4.6154e-05) (hash(x)=4912983)
72
+ 66 train 10.144527 (lr=4.6853e-05) (hash(x)=5109125)
73
+ 67 train 9.981178 (lr=4.7552e-05) (hash(x)=7276303)
74
+ 68 train 10.162324 (lr=4.8252e-05) (hash(x)=6645716)
75
+ 69 train 10.074345 (lr=4.8951e-05) (hash(x)=5479947)
76
+ 70 train 10.202129 (lr=4.9650e-05) (hash(x)=7850360)
77
+ 71 train 10.225736 (lr=5.0350e-05) (hash(x)=6630378)
78
+ 72 train 10.089186 (lr=5.1049e-05) (hash(x)=6295358)
79
+ 73 train 10.166859 (lr=5.1748e-05) (hash(x)=8097010)
80
+ 74 train 10.074977 (lr=5.2448e-05) (hash(x)=3958188)
81
+ 75 train 10.281184 (lr=5.3147e-05) (hash(x)=9989744)
82
+ 76 train 10.061103 (lr=5.3846e-05) (hash(x)=6343832)
83
+ 77 train 10.015673 (lr=5.4545e-05) (hash(x)=6694932)
84
+ 78 train 9.967336 (lr=5.5245e-05) (hash(x)=6429137)
85
+ 79 train 9.927411 (lr=5.5944e-05) (hash(x)=6317894)
86
+ 80 train 9.819329 (lr=5.6643e-05) (hash(x)=5165595)
87
+ 81 train 9.864036 (lr=5.7343e-05) (hash(x)=5046327)
88
+ 82 train 9.956305 (lr=5.8042e-05) (hash(x)=7060363)
89
+ 83 train 9.838555 (lr=5.8741e-05) (hash(x)=7947703)
90
+ 84 train 9.750762 (lr=5.9441e-05) (hash(x)=5971240)
91
+ 85 train 9.727393 (lr=6.0140e-05) (hash(x)=5916068)
92
+ 86 train 9.777673 (lr=6.0839e-05) (hash(x)=7426277)
93
+ 87 train 9.863911 (lr=6.1538e-05) (hash(x)=5777498)
94
+ 88 train 9.815003 (lr=6.2238e-05) (hash(x)=6403331)
95
+ 89 train 9.649207 (lr=6.2937e-05) (hash(x)=5753092)
96
+ 90 train 9.682980 (lr=6.3636e-05) (hash(x)=5981486)
97
+ 91 train 9.694901 (lr=6.4336e-05) (hash(x)=7723226)
98
+ 92 train 9.619762 (lr=6.5035e-05) (hash(x)=6069558)
99
+ 93 train 9.674298 (lr=6.5734e-05) (hash(x)=7007016)
100
+ 94 train 9.550246 (lr=6.6434e-05) (hash(x)=5027958)
101
+ 95 train 9.556903 (lr=6.7133e-05) (hash(x)=6491229)
102
+ 96 train 9.545940 (lr=6.7832e-05) (hash(x)=7036653)
103
+ 97 train 9.601658 (lr=6.8531e-05) (hash(x)=6390752)
104
+ 98 train 9.694002 (lr=6.9231e-05) (hash(x)=7758167)
105
+ 99 train 9.637982 (lr=6.9930e-05) (hash(x)=6019389)
106
+ 100 val loss 9.5200
107
+ 100 val perplexity 13630.1113
108
+ 100 train 9.592540 (lr=7.0629e-05) (hash(x)=6503191)
109
+ 101 train 9.417792 (lr=7.1329e-05) (hash(x)=7662067)
110
+ 102 train 9.459435 (lr=7.2028e-05) (hash(x)=7901992)
111
+ 103 train 9.479414 (lr=7.2727e-05) (hash(x)=4648609)
112
+ 104 train 9.392120 (lr=7.3427e-05) (hash(x)=5897288)
113
+ 105 train 9.349004 (lr=7.4126e-05) (hash(x)=4725966)
114
+ 106 train 9.282923 (lr=7.4825e-05) (hash(x)=6243442)
115
+ 107 train 9.311500 (lr=7.5524e-05) (hash(x)=5558355)
116
+ 108 train 9.352753 (lr=7.6224e-05) (hash(x)=5400407)
117
+ 109 train 9.281687 (lr=7.6923e-05) (hash(x)=6524933)
118
+ 110 train 9.269538 (lr=7.7622e-05) (hash(x)=5889724)
119
+ 111 train 9.213450 (lr=7.8322e-05) (hash(x)=6369326)
120
+ 112 train 9.123732 (lr=7.9021e-05) (hash(x)=5911906)
121
+ 113 train 9.241519 (lr=7.9720e-05) (hash(x)=6022625)
122
+ 114 train 9.051148 (lr=8.0420e-05) (hash(x)=6584810)
123
+ 115 train 9.120297 (lr=8.1119e-05) (hash(x)=4906853)
124
+ 116 train 8.869202 (lr=8.1818e-05) (hash(x)=4564564)
125
+ 117 train 8.862937 (lr=8.2517e-05) (hash(x)=5589269)
126
+ 118 train 8.990231 (lr=8.3217e-05) (hash(x)=6969180)
127
+ 119 train 8.968222 (lr=8.3916e-05) (hash(x)=6437376)
128
+ 120 train 9.033092 (lr=8.4615e-05) (hash(x)=7116258)
129
+ 121 train 8.989401 (lr=8.5315e-05) (hash(x)=7304892)
130
+ 122 train 8.957708 (lr=8.6014e-05) (hash(x)=7366721)
131
+ 123 train 8.933330 (lr=8.6713e-05) (hash(x)=7084093)
132
+ 124 train 8.999436 (lr=8.7413e-05) (hash(x)=7197652)
133
+ 125 train 8.880999 (lr=8.8112e-05) (hash(x)=5908178)
134
+ 126 train 8.681630 (lr=8.8811e-05) (hash(x)=6349974)
135
+ 127 train 8.701337 (lr=8.9510e-05) (hash(x)=7048804)
136
+ 128 train 8.732218 (lr=9.0210e-05) (hash(x)=7909039)
137
+ 129 train 8.753483 (lr=9.0909e-05) (hash(x)=6085549)
138
+ 130 train 8.815530 (lr=9.1608e-05) (hash(x)=7463181)
139
+ 131 train 8.699938 (lr=9.2308e-05) (hash(x)=5752594)
140
+ 132 train 8.515303 (lr=9.3007e-05) (hash(x)=6114601)
141
+ 133 train 8.501603 (lr=9.3706e-05) (hash(x)=7662769)
142
+ 134 train 8.486214 (lr=9.4406e-05) (hash(x)=7485666)
143
+ 135 train 8.240755 (lr=9.5105e-05) (hash(x)=6393520)
144
+ 136 train 8.522907 (lr=9.5804e-05) (hash(x)=5586291)
145
+ 137 train 8.700139 (lr=9.6503e-05) (hash(x)=5942867)
146
+ 138 train 8.629080 (lr=9.7203e-05) (hash(x)=7574027)
147
+ 139 train 8.557504 (lr=9.7902e-05) (hash(x)=6793550)
148
+ 140 train 8.387895 (lr=9.8601e-05) (hash(x)=6264789)
149
+ 141 train 8.403167 (lr=9.9301e-05) (hash(x)=7105976)
150
+ 142 train 8.480000 (lr=1.0000e-04) (hash(x)=7244279)
151
+ 143 train 8.348940 (lr=1.0070e-04) (hash(x)=6581348)
152
+ 144 train 8.328314 (lr=1.0140e-04) (hash(x)=6543825)
153
+ 145 train 8.281722 (lr=1.0210e-04) (hash(x)=5555598)
154
+ 146 train 8.377936 (lr=1.0280e-04) (hash(x)=7254232)
155
+ 147 train 8.213572 (lr=1.0350e-04) (hash(x)=6235837)
156
+ 148 train 8.206971 (lr=1.0420e-04) (hash(x)=4738550)
157
+ 149 train 8.215639 (lr=1.0490e-04) (hash(x)=5591770)
158
+ 150 val loss 8.2410
159
+ 150 val perplexity 3793.2202
160
+ 150 train 8.738932 (lr=1.0559e-04) (hash(x)=8012491)
161
+ 151 train 8.197873 (lr=1.0629e-04) (hash(x)=7083021)
162
+ 152 train 8.306237 (lr=1.0699e-04) (hash(x)=6456022)
163
+ 153 train 8.267984 (lr=1.0769e-04) (hash(x)=7281348)
164
+ 154 train 8.360527 (lr=1.0839e-04) (hash(x)=7824794)
165
+ 155 train 8.393332 (lr=1.0909e-04) (hash(x)=6155821)
166
+ 156 train 8.239111 (lr=1.0979e-04) (hash(x)=7807564)
167
+ 157 train 8.082906 (lr=1.1049e-04) (hash(x)=5736241)
168
+ 158 train 8.173496 (lr=1.1119e-04) (hash(x)=5688415)
169
+ 159 train 7.980140 (lr=1.1189e-04) (hash(x)=5657123)
170
+ 160 train 8.035019 (lr=1.1259e-04) (hash(x)=6102985)
171
+ 161 train 8.146736 (lr=1.1329e-04) (hash(x)=6452095)
172
+ 162 train 8.075621 (lr=1.1399e-04) (hash(x)=5848921)
173
+ 163 train 7.910108 (lr=1.1469e-04) (hash(x)=5312267)
174
+ 164 train 7.817029 (lr=1.1538e-04) (hash(x)=5025297)
175
+ 165 train 7.776151 (lr=1.1608e-04) (hash(x)=5434493)
176
+ 166 train 8.171447 (lr=1.1678e-04) (hash(x)=5197474)
177
+ 167 train 8.015289 (lr=1.1748e-04) (hash(x)=6830813)
178
+ 168 train 7.905487 (lr=1.1818e-04) (hash(x)=5242548)
179
+ 169 train 7.824608 (lr=1.1888e-04) (hash(x)=4752771)
180
+ 170 train 8.041288 (lr=1.1958e-04) (hash(x)=6691249)
181
+ 171 train 8.050759 (lr=1.2028e-04) (hash(x)=6424933)
182
+ 172 train 8.001429 (lr=1.2098e-04) (hash(x)=6960120)
183
+ 173 train 7.943762 (lr=1.2168e-04) (hash(x)=5786517)
184
+ 174 train 7.981680 (lr=1.2238e-04) (hash(x)=7229689)
185
+ 175 train 8.159220 (lr=1.2308e-04) (hash(x)=6680023)
186
+ 176 train 8.025226 (lr=1.2378e-04) (hash(x)=7688393)
187
+ 177 train 7.952527 (lr=1.2448e-04) (hash(x)=6474729)
188
+ 178 train 8.032599 (lr=1.2517e-04) (hash(x)=5946650)
189
+ 179 train 7.953331 (lr=1.2587e-04) (hash(x)=7523279)
190
+ 180 train 7.839624 (lr=1.2657e-04) (hash(x)=5531605)
191
+ 181 train 7.728240 (lr=1.2727e-04) (hash(x)=5508768)
192
+ 182 train 8.068892 (lr=1.2797e-04) (hash(x)=6415982)
193
+ 183 train 7.964187 (lr=1.2867e-04) (hash(x)=6635925)
194
+ 184 train 7.874468 (lr=1.2937e-04) (hash(x)=5923706)
195
+ 185 train 7.870440 (lr=1.3007e-04) (hash(x)=6663801)
196
+ 186 train 7.795080 (lr=1.3077e-04) (hash(x)=6113540)
197
+ 187 train 7.916042 (lr=1.3147e-04) (hash(x)=5869441)
198
+ 188 train 8.001571 (lr=1.3217e-04) (hash(x)=6913454)
199
+ 189 train 7.871938 (lr=1.3287e-04) (hash(x)=6698878)
200
+ 190 train 7.906436 (lr=1.3357e-04) (hash(x)=6094899)
201
+ 191 train 7.775713 (lr=1.3427e-04) (hash(x)=6349004)
202
+ 192 train 7.816682 (lr=1.3497e-04) (hash(x)=6045933)
203
+ 193 train 7.882951 (lr=1.3566e-04) (hash(x)=4986137)
204
+ 194 train 7.592698 (lr=1.3636e-04) (hash(x)=4798229)
205
+ 195 train 7.688888 (lr=1.3706e-04) (hash(x)=6008454)
206
+ 196 train 7.751460 (lr=1.3776e-04) (hash(x)=5938362)
207
+ 197 train 7.690753 (lr=1.3846e-04) (hash(x)=6550770)
208
+ 198 train 7.798873 (lr=1.3916e-04) (hash(x)=7055344)
209
+ 199 train 7.988805 (lr=1.3986e-04) (hash(x)=7682741)
210
+ 200 val loss 7.8344
211
+ 200 val perplexity 2525.9487
212
+ 200 train 7.798297 (lr=1.4056e-04) (hash(x)=7372095)
213
+ 201 train 8.083208 (lr=1.4126e-04) (hash(x)=7892158)
214
+ 202 train 7.809347 (lr=1.4196e-04) (hash(x)=5929740)
215
+ 203 train 7.795606 (lr=1.4266e-04) (hash(x)=7794028)
216
+ 204 train 7.982548 (lr=1.4336e-04) (hash(x)=6839923)
217
+ 205 train 7.628150 (lr=1.4406e-04) (hash(x)=7093523)
218
+ 206 train 7.802186 (lr=1.4476e-04) (hash(x)=5341312)
219
+ 207 train 7.911708 (lr=1.4545e-04) (hash(x)=6166062)
220
+ 208 train 7.871403 (lr=1.4615e-04) (hash(x)=5992164)
221
+ 209 train 7.725142 (lr=1.4685e-04) (hash(x)=6645781)
222
+ 210 train 7.866336 (lr=1.4755e-04) (hash(x)=6207724)
223
+ 211 train 7.667377 (lr=1.4825e-04) (hash(x)=4811192)
224
+ 212 train 7.675628 (lr=1.4895e-04) (hash(x)=6906335)
225
+ 213 train 7.685686 (lr=1.4965e-04) (hash(x)=5970866)
226
+ 214 train 7.671600 (lr=1.5035e-04) (hash(x)=5238015)
227
+ 215 train 7.712323 (lr=1.5105e-04) (hash(x)=7822680)
228
+ 216 train 7.649664 (lr=1.5175e-04) (hash(x)=5478019)
229
+ 217 train 7.690408 (lr=1.5245e-04) (hash(x)=6317317)
230
+ 218 train 7.797443 (lr=1.5315e-04) (hash(x)=5832344)
231
+ 219 train 7.785646 (lr=1.5385e-04) (hash(x)=6431409)
232
+ 220 train 7.748334 (lr=1.5455e-04) (hash(x)=5900130)
233
+ 221 train 7.845036 (lr=1.5524e-04) (hash(x)=6653337)
234
+ 222 train 7.723774 (lr=1.5594e-04) (hash(x)=6696784)
235
+ 223 train 7.774099 (lr=1.5664e-04) (hash(x)=5323032)
236
+ 224 train 7.623947 (lr=1.5734e-04) (hash(x)=5372100)
237
+ 225 train 7.632366 (lr=1.5804e-04) (hash(x)=5125339)
238
+ 226 train 7.539998 (lr=1.5874e-04) (hash(x)=5664463)
239
+ 227 train 7.571677 (lr=1.5944e-04) (hash(x)=7363286)
240
+ 228 train 8.065365 (lr=1.6014e-04) (hash(x)=7444468)
241
+ 229 train 7.878375 (lr=1.6084e-04) (hash(x)=7670050)
242
+ 230 train 7.754068 (lr=1.6154e-04) (hash(x)=8189420)
243
+ 231 train 7.670938 (lr=1.6224e-04) (hash(x)=5616075)
244
+ 232 train 7.670694 (lr=1.6294e-04) (hash(x)=6360924)
245
+ 233 train 7.888234 (lr=1.6364e-04) (hash(x)=6593764)
246
+ 234 train 8.031884 (lr=1.6434e-04) (hash(x)=6764377)
247
+ 235 train 7.676063 (lr=1.6503e-04) (hash(x)=5672608)
248
+ 236 train 7.483418 (lr=1.6573e-04) (hash(x)=6985256)
249
+ 237 train 7.859094 (lr=1.6643e-04) (hash(x)=6828653)
250
+ 238 train 7.575392 (lr=1.6713e-04) (hash(x)=5199845)
251
+ 239 train 7.714341 (lr=1.6783e-04) (hash(x)=6839867)
252
+ 240 train 7.852492 (lr=1.6853e-04) (hash(x)=6928712)
253
+ 241 train 7.779597 (lr=1.6923e-04) (hash(x)=8426992)
254
+ 242 train 7.720500 (lr=1.6993e-04) (hash(x)=5330355)
255
+ 243 train 7.209429 (lr=1.7063e-04) (hash(x)=4916928)
256
+ 244 train 7.637279 (lr=1.7133e-04) (hash(x)=7287928)
257
+ 245 train 7.713524 (lr=1.7203e-04) (hash(x)=6484050)
258
+ 246 train 7.792243 (lr=1.7273e-04) (hash(x)=5062850)
259
+ 247 train 7.775901 (lr=1.7343e-04) (hash(x)=8353379)
260
+ 248 train 7.678964 (lr=1.7413e-04) (hash(x)=6805358)
261
+ 249 train 7.707503 (lr=1.7483e-04) (hash(x)=6552510)
262
+ 250 val loss 7.6912
263
+ 250 val perplexity 2189.0850
264
+ 250 train 7.698315 (lr=1.7552e-04) (hash(x)=5105545)
265
+ 251 train 7.661218 (lr=1.7622e-04) (hash(x)=4838871)
266
+ 252 train 7.787342 (lr=1.7692e-04) (hash(x)=5945286)
267
+ 253 train 7.056741 (lr=1.7762e-04) (hash(x)=3021697)
268
+ 254 train 7.526209 (lr=1.7832e-04) (hash(x)=6443070)
269
+ 255 train 7.640157 (lr=1.7902e-04) (hash(x)=6795665)
270
+ 256 train 7.796952 (lr=1.7972e-04) (hash(x)=7683068)
271
+ 257 train 7.392519 (lr=1.8042e-04) (hash(x)=5486935)
272
+ 258 train 7.725664 (lr=1.8112e-04) (hash(x)=6115232)
273
+ 259 train 7.837753 (lr=1.8182e-04) (hash(x)=6678038)
274
+ 260 train 7.828147 (lr=1.8252e-04) (hash(x)=5747629)
275
+ 261 train 7.811316 (lr=1.8322e-04) (hash(x)=6657714)
276
+ 262 train 7.643857 (lr=1.8392e-04) (hash(x)=6190642)
277
+ 263 train 7.751606 (lr=1.8462e-04) (hash(x)=6112215)
278
+ 264 train 7.691644 (lr=1.8531e-04) (hash(x)=7314989)
279
+ 265 train 7.808522 (lr=1.8601e-04) (hash(x)=8353143)
280
+ 266 train 7.827618 (lr=1.8671e-04) (hash(x)=5912570)
281
+ 267 train 7.773612 (lr=1.8741e-04) (hash(x)=5915361)
282
+ 268 train 7.600836 (lr=1.8811e-04) (hash(x)=6369494)
283
+ 269 train 7.801116 (lr=1.8881e-04) (hash(x)=6254885)
284
+ 270 train 7.675181 (lr=1.8951e-04) (hash(x)=5425120)
285
+ 271 train 7.709579 (lr=1.9021e-04) (hash(x)=5771588)
286
+ 272 train 7.718955 (lr=1.9091e-04) (hash(x)=6763967)
287
+ 273 train 7.625972 (lr=1.9161e-04) (hash(x)=5561507)
288
+ 274 train 7.574045 (lr=1.9231e-04) (hash(x)=7081151)
289
+ 275 train 7.640127 (lr=1.9301e-04) (hash(x)=5688829)
290
+ 276 train 7.645526 (lr=1.9371e-04) (hash(x)=7821554)
291
+ 277 train 7.620264 (lr=1.9441e-04) (hash(x)=5371951)
292
+ 278 train 7.767745 (lr=1.9510e-04) (hash(x)=8120149)
293
+ 279 train 7.849814 (lr=1.9580e-04) (hash(x)=5443305)
294
+ 280 train 7.622212 (lr=1.9650e-04) (hash(x)=7066030)
295
+ 281 train 7.928026 (lr=1.9720e-04) (hash(x)=5016757)
296
+ 282 train 7.775947 (lr=1.9790e-04) (hash(x)=6296163)
297
+ 283 train 7.797986 (lr=1.9860e-04) (hash(x)=6901933)
298
+ 284 train 7.688839 (lr=1.9930e-04) (hash(x)=8056633)
299
+ 285 train 7.815433 (lr=2.0000e-04) (hash(x)=6019085)
300
+ 286 train 7.617684 (lr=2.0000e-04) (hash(x)=7003893)
301
+ 287 train 7.551760 (lr=2.0000e-04) (hash(x)=4294425)
302
+ 288 train 7.604907 (lr=2.0000e-04) (hash(x)=6568540)
303
+ 289 train 7.786672 (lr=1.9999e-04) (hash(x)=6095229)
304
+ 290 train 7.713283 (lr=1.9999e-04) (hash(x)=5512150)
305
+ 291 train 7.300828 (lr=1.9998e-04) (hash(x)=4363941)
306
+ 292 train 7.617440 (lr=1.9997e-04) (hash(x)=6457783)
307
+ 293 train 7.613905 (lr=1.9996e-04) (hash(x)=4693798)
308
+ 294 train 8.198577 (lr=1.9994e-04) (hash(x)=8759806)
309
+ 295 train 8.210944 (lr=1.9993e-04) (hash(x)=5486335)
310
+ 296 train 7.825127 (lr=1.9991e-04) (hash(x)=6964993)
311
+ 297 train 8.123847 (lr=1.9989e-04) (hash(x)=7907450)
312
+ 298 train 7.849515 (lr=1.9987e-04) (hash(x)=5236642)
313
+ 299 train 7.743367 (lr=1.9985e-04) (hash(x)=7295165)
314
+ 300 val loss 7.6796
315
+ 300 val perplexity 2163.7166
316
+ 300 train 7.708810 (lr=1.9983e-04) (hash(x)=4913602)
317
+ 301 train 7.745741 (lr=1.9980e-04) (hash(x)=6814026)
318
+ 302 train 7.860341 (lr=1.9978e-04) (hash(x)=7042051)
319
+ 303 train 7.749226 (lr=1.9975e-04) (hash(x)=6267424)
320
+ 304 train 7.835481 (lr=1.9972e-04) (hash(x)=5614727)
321
+ 305 train 7.806735 (lr=1.9969e-04) (hash(x)=8086437)
322
+ 306 train 7.908227 (lr=1.9965e-04) (hash(x)=5900143)
323
+ 307 train 7.840741 (lr=1.9962e-04) (hash(x)=5270452)
324
+ 308 train 7.755070 (lr=1.9958e-04) (hash(x)=8617707)
325
+ 309 train 7.649565 (lr=1.9954e-04) (hash(x)=6420820)
326
+ 310 train 7.527202 (lr=1.9950e-04) (hash(x)=6215930)
327
+ 311 train 7.719905 (lr=1.9946e-04) (hash(x)=6305297)
328
+ 312 train 7.934155 (lr=1.9941e-04) (hash(x)=4026682)
329
+ 313 train 7.615021 (lr=1.9937e-04) (hash(x)=7889849)
330
+ 314 train 7.884044 (lr=1.9932e-04) (hash(x)=6141721)
331
+ 315 train 8.440273 (lr=1.9927e-04) (hash(x)=6040077)
332
+ 316 train 8.398265 (lr=1.9922e-04) (hash(x)=6155449)
333
+ 317 train 7.900534 (lr=1.9916e-04) (hash(x)=5652497)
334
+ 318 train 7.797987 (lr=1.9911e-04) (hash(x)=6961314)
335
+ 319 train 7.663014 (lr=1.9905e-04) (hash(x)=7130251)
336
+ 320 train 7.505059 (lr=1.9899e-04) (hash(x)=5927461)
337
+ 321 train 7.686394 (lr=1.9893e-04) (hash(x)=6610177)
338
+ 322 train 7.734031 (lr=1.9887e-04) (hash(x)=7277580)
339
+ 323 train 7.562392 (lr=1.9881e-04) (hash(x)=6508350)
340
+ 324 train 7.839149 (lr=1.9874e-04) (hash(x)=6536034)
341
+ 325 train 7.742727 (lr=1.9868e-04) (hash(x)=6994983)
342
+ 326 train 7.765797 (lr=1.9861e-04) (hash(x)=8700721)
343
+ 327 train 7.706434 (lr=1.9854e-04) (hash(x)=5859576)
344
+ 328 train 7.318797 (lr=1.9847e-04) (hash(x)=5264962)
345
+ 329 train 7.303150 (lr=1.9839e-04) (hash(x)=5487065)
346
+ 330 train 7.449162 (lr=1.9832e-04) (hash(x)=5731624)
347
+ 331 train 7.598080 (lr=1.9824e-04) (hash(x)=4892065)
348
+ 332 train 7.494036 (lr=1.9816e-04) (hash(x)=5489496)
349
+ 333 train 7.345603 (lr=1.9808e-04) (hash(x)=4933674)
350
+ 334 train 7.608705 (lr=1.9800e-04) (hash(x)=7021003)
351
+ 335 train 7.817288 (lr=1.9792e-04) (hash(x)=7317289)
352
+ 336 train 7.839263 (lr=1.9783e-04) (hash(x)=7869305)
353
+ 337 train 7.791542 (lr=1.9774e-04) (hash(x)=6811522)
354
+ 338 train 7.685787 (lr=1.9765e-04) (hash(x)=6601423)
355
+ 339 train 7.575094 (lr=1.9756e-04) (hash(x)=5510218)
356
+ 340 train 7.928952 (lr=1.9747e-04) (hash(x)=7180298)
357
+ 341 train 7.672410 (lr=1.9738e-04) (hash(x)=7304235)
358
+ 342 train 7.838900 (lr=1.9728e-04) (hash(x)=8111920)
359
+ 343 train 8.101269 (lr=1.9718e-04) (hash(x)=6752265)
360
+ 344 train 7.570523 (lr=1.9709e-04) (hash(x)=6788720)
361
+ 345 train 7.691226 (lr=1.9698e-04) (hash(x)=5994476)
362
+ 346 train 7.558375 (lr=1.9688e-04) (hash(x)=5862533)
363
+ 347 train 7.663682 (lr=1.9678e-04) (hash(x)=6530867)
364
+ 348 train 7.649126 (lr=1.9667e-04) (hash(x)=5914734)
365
+ 349 train 7.770808 (lr=1.9656e-04) (hash(x)=5746260)
366
+ 350 val loss 7.6691
367
+ 350 val perplexity 2141.1104
368
+ 350 train 7.666231 (lr=1.9646e-04) (hash(x)=6856097)
369
+ 351 train 7.646579 (lr=1.9634e-04) (hash(x)=5790078)
370
+ 352 train 7.798625 (lr=1.9623e-04) (hash(x)=5805002)
371
+ 353 train 7.576107 (lr=1.9612e-04) (hash(x)=6430135)
372
+ 354 train 7.579481 (lr=1.9600e-04) (hash(x)=6156715)
373
+ 355 train 7.932182 (lr=1.9588e-04) (hash(x)=7614023)
374
+ 356 train 7.829601 (lr=1.9576e-04) (hash(x)=4861744)
375
+ 357 train 7.645341 (lr=1.9564e-04) (hash(x)=5244861)
376
+ 358 train 7.718120 (lr=1.9552e-04) (hash(x)=7163697)
377
+ 359 train 7.735084 (lr=1.9540e-04) (hash(x)=5245146)
378
+ 360 train 7.586048 (lr=1.9527e-04) (hash(x)=6153866)
379
+ 361 train 7.642743 (lr=1.9514e-04) (hash(x)=6415565)
380
+ 362 train 7.942325 (lr=1.9501e-04) (hash(x)=7835853)
381
+ 363 train 7.386051 (lr=1.9488e-04) (hash(x)=5786972)
382
+ 364 train 7.765721 (lr=1.9475e-04) (hash(x)=6733560)
383
+ 365 train 7.634493 (lr=1.9462e-04) (hash(x)=6131703)
384
+ 366 train 7.495063 (lr=1.9448e-04) (hash(x)=6001331)
385
+ 367 train 8.005271 (lr=1.9434e-04) (hash(x)=6346459)
386
+ 368 train 7.800947 (lr=1.9421e-04) (hash(x)=8943770)
387
+ 369 train 7.745122 (lr=1.9406e-04) (hash(x)=6504131)
388
+ 370 train 7.673328 (lr=1.9392e-04) (hash(x)=5869339)
389
+ 371 train 7.785290 (lr=1.9378e-04) (hash(x)=5209234)
390
+ 372 train 7.634972 (lr=1.9363e-04) (hash(x)=7678937)
391
+ 373 train 7.684781 (lr=1.9349e-04) (hash(x)=6163272)
392
+ 374 train 7.688353 (lr=1.9334e-04) (hash(x)=5774172)
393
+ 375 train 7.644206 (lr=1.9319e-04) (hash(x)=5676768)
394
+ 376 train 7.298154 (lr=1.9304e-04) (hash(x)=6443895)
395
+ 377 train 8.182186 (lr=1.9288e-04) (hash(x)=6090561)
396
+ 378 train 7.777568 (lr=1.9273e-04) (hash(x)=7788628)
397
+ 379 train 7.959432 (lr=1.9257e-04) (hash(x)=5863092)
398
+ 380 train 7.692801 (lr=1.9241e-04) (hash(x)=7141354)
399
+ 381 train 7.870003 (lr=1.9225e-04) (hash(x)=7201312)
400
+ 382 train 7.836949 (lr=1.9209e-04) (hash(x)=6414278)
401
+ 383 train 7.888556 (lr=1.9193e-04) (hash(x)=7019489)
402
+ 384 train 7.628554 (lr=1.9176e-04) (hash(x)=7914626)
403
+ 385 train 7.623000 (lr=1.9160e-04) (hash(x)=6225608)
404
+ 386 train 7.712932 (lr=1.9143e-04) (hash(x)=5998078)
405
+ 387 train 7.687317 (lr=1.9126e-04) (hash(x)=5483019)
406
+ 388 train 7.511609 (lr=1.9109e-04) (hash(x)=5911642)
407
+ 389 train 7.763136 (lr=1.9091e-04) (hash(x)=5903258)
408
+ 390 train 7.770864 (lr=1.9074e-04) (hash(x)=5632006)
409
+ 391 train 7.618561 (lr=1.9056e-04) (hash(x)=5968716)
410
+ 392 train 7.711931 (lr=1.9039e-04) (hash(x)=6401968)
411
+ 393 train 7.663370 (lr=1.9021e-04) (hash(x)=5778017)
412
+ 394 train 7.189019 (lr=1.9003e-04) (hash(x)=5772567)
413
+ 395 train 7.590845 (lr=1.8985e-04) (hash(x)=7780194)
414
+ 396 train 7.693964 (lr=1.8966e-04) (hash(x)=6424771)
415
+ 397 train 7.543945 (lr=1.8948e-04) (hash(x)=6140998)
416
+ 398 train 7.555213 (lr=1.8929e-04) (hash(x)=7859566)
417
+ 399 train 7.665734 (lr=1.8910e-04) (hash(x)=3784321)
418
+ 400 val loss 7.6349
419
+ 400 val perplexity 2069.0676
420
+ 400 train 7.567648 (lr=1.8891e-04) (hash(x)=5557891)
421
+ 401 train 7.693021 (lr=1.8872e-04) (hash(x)=5860821)
422
+ 402 train 7.686289 (lr=1.8853e-04) (hash(x)=5674826)
423
+ 403 train 7.660503 (lr=1.8834e-04) (hash(x)=7774109)
424
+ 404 train 7.705628 (lr=1.8814e-04) (hash(x)=5466449)
425
+ 405 train 8.001887 (lr=1.8794e-04) (hash(x)=7997664)
426
+ 406 train 7.877735 (lr=1.8774e-04) (hash(x)=7006517)
427
+ 407 train 7.733673 (lr=1.8754e-04) (hash(x)=6576749)
428
+ 408 train 7.641888 (lr=1.8734e-04) (hash(x)=6896198)
429
+ 409 train 7.829796 (lr=1.8714e-04) (hash(x)=7266655)
430
+ 410 train 7.612694 (lr=1.8693e-04) (hash(x)=5696696)
431
+ 411 train 7.755228 (lr=1.8673e-04) (hash(x)=7072404)
432
+ 412 train 7.596433 (lr=1.8652e-04) (hash(x)=6590657)
433
+ 413 train 7.622679 (lr=1.8631e-04) (hash(x)=6626459)
434
+ 414 train 7.677356 (lr=1.8610e-04) (hash(x)=5644977)
435
+ 415 train 7.547370 (lr=1.8589e-04) (hash(x)=6556025)
436
+ 416 train 7.757295 (lr=1.8567e-04) (hash(x)=6407492)
437
+ 417 train 7.534813 (lr=1.8546e-04) (hash(x)=5942129)
438
+ 418 train 7.626450 (lr=1.8524e-04) (hash(x)=6012163)
439
+ 419 train 7.902880 (lr=1.8502e-04) (hash(x)=7325027)
440
+ 420 train 7.846186 (lr=1.8480e-04) (hash(x)=4425520)
441
+ 421 train 7.691623 (lr=1.8458e-04) (hash(x)=7322467)
442
+ 422 train 7.635442 (lr=1.8436e-04) (hash(x)=7482800)
443
+ 423 train 7.732132 (lr=1.8414e-04) (hash(x)=6373412)
444
+ 424 train 7.361012 (lr=1.8391e-04) (hash(x)=6981426)
445
+ 425 train 7.559134 (lr=1.8369e-04) (hash(x)=7332255)
446
+ 426 train 7.754871 (lr=1.8346e-04) (hash(x)=5930894)
447
+ 427 train 7.756703 (lr=1.8323e-04) (hash(x)=5262868)
448
+ 428 train 7.500230 (lr=1.8300e-04) (hash(x)=5585769)
449
+ 429 train 7.415083 (lr=1.8276e-04) (hash(x)=5688247)
450
+ 430 train 7.708945 (lr=1.8253e-04) (hash(x)=7462079)
451
+ 431 train 7.537834 (lr=1.8230e-04) (hash(x)=8055563)
452
+ 432 train 7.647192 (lr=1.8206e-04) (hash(x)=6221701)
453
+ 433 train 7.652141 (lr=1.8182e-04) (hash(x)=5352405)
454
+ 434 train 7.695446 (lr=1.8158e-04) (hash(x)=6730666)
455
+ 435 train 7.714622 (lr=1.8134e-04) (hash(x)=5932790)
456
+ 436 train 7.748716 (lr=1.8110e-04) (hash(x)=6223114)
457
+ 437 train 7.572248 (lr=1.8086e-04) (hash(x)=6249312)
458
+ 438 train 7.710807 (lr=1.8061e-04) (hash(x)=7242827)
459
+ 439 train 7.485700 (lr=1.8036e-04) (hash(x)=5680154)
460
+ 440 train 7.740619 (lr=1.8012e-04) (hash(x)=6332332)
461
+ 441 train 7.661716 (lr=1.7987e-04) (hash(x)=8156280)
462
+ 442 train 7.641844 (lr=1.7962e-04) (hash(x)=6659781)
463
+ 443 train 7.637813 (lr=1.7937e-04) (hash(x)=5914217)
464
+ 444 train 7.529335 (lr=1.7911e-04) (hash(x)=6714110)
465
+ 445 train 7.415503 (lr=1.7886e-04) (hash(x)=7120215)
466
+ 446 train 7.710623 (lr=1.7860e-04) (hash(x)=6593413)
467
+ 447 train 7.646263 (lr=1.7835e-04) (hash(x)=6424128)
468
+ 448 train 7.457904 (lr=1.7809e-04) (hash(x)=5903043)
469
+ 449 train 7.368762 (lr=1.7783e-04) (hash(x)=4390027)
470
+ 450 val loss 7.5900
471
+ 450 val perplexity 1978.2761
472
+ 450 train 7.612141 (lr=1.7757e-04) (hash(x)=6945760)
473
+ 451 train 7.661142 (lr=1.7731e-04) (hash(x)=7804089)
474
+ 452 train 7.718168 (lr=1.7704e-04) (hash(x)=6273417)
475
+ 453 train 7.758194 (lr=1.7678e-04) (hash(x)=7130267)
476
+ 454 train 7.519240 (lr=1.7651e-04) (hash(x)=6140697)
477
+ 455 train 7.484344 (lr=1.7624e-04) (hash(x)=6490149)
478
+ 456 train 7.401560 (lr=1.7597e-04) (hash(x)=4847733)
479
+ 457 train 7.778947 (lr=1.7570e-04) (hash(x)=9276378)
480
+ 458 train 7.526563 (lr=1.7543e-04) (hash(x)=7082053)
481
+ 459 train 7.697488 (lr=1.7516e-04) (hash(x)=6283540)
482
+ 460 train 7.526554 (lr=1.7489e-04) (hash(x)=5200956)
483
+ 461 train 7.657871 (lr=1.7461e-04) (hash(x)=5781680)
484
+ 462 train 7.632443 (lr=1.7434e-04) (hash(x)=8652907)
485
+ 463 train 7.399150 (lr=1.7406e-04) (hash(x)=5728337)
486
+ 464 train 7.540676 (lr=1.7378e-04) (hash(x)=5832401)
487
+ 465 train 7.494296 (lr=1.7350e-04) (hash(x)=6278202)
488
+ 466 train 7.580884 (lr=1.7322e-04) (hash(x)=7403868)
489
+ 467 train 7.605662 (lr=1.7294e-04) (hash(x)=7400272)
490
+ 468 train 7.581446 (lr=1.7265e-04) (hash(x)=5943413)
491
+ 469 train 7.587706 (lr=1.7237e-04) (hash(x)=5795128)
492
+ 470 train 7.609163 (lr=1.7208e-04) (hash(x)=5519980)
493
+ 471 train 7.473267 (lr=1.7179e-04) (hash(x)=6473432)
494
+ 472 train 7.507545 (lr=1.7151e-04) (hash(x)=5484669)
495
+ 473 train 7.605585 (lr=1.7122e-04) (hash(x)=5201502)
496
+ 474 train 7.691759 (lr=1.7092e-04) (hash(x)=7227057)
497
+ 475 train 7.884198 (lr=1.7063e-04) (hash(x)=6922646)
498
+ 476 train 7.417528 (lr=1.7034e-04) (hash(x)=6056673)
499
+ 477 train 7.566247 (lr=1.7004e-04) (hash(x)=5484641)
500
+ 478 train 7.678273 (lr=1.6975e-04) (hash(x)=7100865)
501
+ 479 train 7.432081 (lr=1.6945e-04) (hash(x)=7277141)
502
+ 480 train 7.519135 (lr=1.6915e-04) (hash(x)=6407603)
503
+ 481 train 7.604016 (lr=1.6886e-04) (hash(x)=5604614)
504
+ 482 train 7.641371 (lr=1.6856e-04) (hash(x)=6200612)
505
+ 483 train 7.619506 (lr=1.6825e-04) (hash(x)=7443124)
506
+ 484 train 7.567224 (lr=1.6795e-04) (hash(x)=6728764)
507
+ 485 train 7.694747 (lr=1.6765e-04) (hash(x)=5601968)
508
+ 486 train 7.419085 (lr=1.6734e-04) (hash(x)=6877761)
509
+ 487 train 7.630326 (lr=1.6704e-04) (hash(x)=5632735)
510
+ 488 train 7.649678 (lr=1.6673e-04) (hash(x)=5658845)
511
+ 489 train 7.398854 (lr=1.6642e-04) (hash(x)=5594810)
512
+ 490 train 7.171398 (lr=1.6611e-04) (hash(x)=4738523)
513
+ 491 train 7.556986 (lr=1.6580e-04) (hash(x)=5881015)
514
+ 492 train 7.555234 (lr=1.6549e-04) (hash(x)=5263325)
515
+ 493 train 7.501265 (lr=1.6518e-04) (hash(x)=6865361)
516
+ 494 train 7.715881 (lr=1.6487e-04) (hash(x)=5644619)
517
+ 495 train 7.527929 (lr=1.6455e-04) (hash(x)=5809411)
518
+ 496 train 7.578063 (lr=1.6424e-04) (hash(x)=7044374)
519
+ 497 train 7.627397 (lr=1.6392e-04) (hash(x)=6293025)
520
+ 498 train 7.591089 (lr=1.6360e-04) (hash(x)=5189023)
521
+ 499 train 7.501734 (lr=1.6328e-04) (hash(x)=7232964)
522
+ 500 val loss 7.5414
523
+ 500 val perplexity 1884.5446
524
+ 500 train 7.532164 (lr=1.6296e-04) (hash(x)=5806257)
525
+ 501 train 7.404531 (lr=1.6264e-04) (hash(x)=5646619)
526
+ 502 train 7.564762 (lr=1.6232e-04) (hash(x)=6966594)
527
+ 503 train 7.600958 (lr=1.6200e-04) (hash(x)=6171829)
528
+ 504 train 7.508771 (lr=1.6168e-04) (hash(x)=5945953)
529
+ 505 train 7.770303 (lr=1.6135e-04) (hash(x)=7787552)
530
+ 506 train 7.665814 (lr=1.6103e-04) (hash(x)=7184332)
531
+ 507 train 7.521362 (lr=1.6070e-04) (hash(x)=5884456)
532
+ 508 train 7.624187 (lr=1.6037e-04) (hash(x)=6443281)
533
+ 509 train 7.607260 (lr=1.6004e-04) (hash(x)=6752162)
534
+ 510 train 7.613330 (lr=1.5971e-04) (hash(x)=7140204)
535
+ 511 train 7.451076 (lr=1.5938e-04) (hash(x)=5077031)
536
+ 512 train 7.684568 (lr=1.5905e-04) (hash(x)=6490444)
537
+ 513 train 7.717364 (lr=1.5872e-04) (hash(x)=6883481)
538
+ 514 train 7.557893 (lr=1.5838e-04) (hash(x)=7381497)
539
+ 515 train 7.594733 (lr=1.5805e-04) (hash(x)=6248895)
540
+ 516 train 7.474553 (lr=1.5772e-04) (hash(x)=6655640)
541
+ 517 train 7.419809 (lr=1.5738e-04) (hash(x)=6636615)
542
+ 518 train 7.742638 (lr=1.5704e-04) (hash(x)=5452148)
543
+ 519 train 7.548080 (lr=1.5670e-04) (hash(x)=6918004)
544
+ 520 train 7.552815 (lr=1.5636e-04) (hash(x)=6547295)
545
+ 521 train 7.583791 (lr=1.5602e-04) (hash(x)=7302874)
546
+ 522 train 7.748670 (lr=1.5568e-04) (hash(x)=5637695)
547
+ 523 train 7.602750 (lr=1.5534e-04) (hash(x)=6482466)
548
+ 524 train 7.533391 (lr=1.5500e-04) (hash(x)=7612126)
549
+ 525 train 7.471854 (lr=1.5466e-04) (hash(x)=7298317)
550
+ 526 train 7.631557 (lr=1.5431e-04) (hash(x)=5635777)
551
+ 527 train 7.482061 (lr=1.5397e-04) (hash(x)=6588573)
552
+ 528 train 7.668786 (lr=1.5362e-04) (hash(x)=6634173)
553
+ 529 train 7.432822 (lr=1.5327e-04) (hash(x)=6229630)
554
+ 530 train 7.731978 (lr=1.5293e-04) (hash(x)=6867600)
555
+ 531 train 7.489731 (lr=1.5258e-04) (hash(x)=6136939)
556
+ 532 train 7.339288 (lr=1.5223e-04) (hash(x)=6014704)
557
+ 533 train 7.337971 (lr=1.5188e-04) (hash(x)=6394645)
558
+ 534 train 7.710047 (lr=1.5153e-04) (hash(x)=6548145)
559
+ 535 train 7.502839 (lr=1.5118e-04) (hash(x)=7040157)
560
+ 536 train 7.661766 (lr=1.5082e-04) (hash(x)=5642718)
561
+ 537 train 7.589428 (lr=1.5047e-04) (hash(x)=7041858)
562
+ 538 train 7.552312 (lr=1.5012e-04) (hash(x)=6392891)
563
+ 539 train 7.559540 (lr=1.4976e-04) (hash(x)=6455595)
564
+ 540 train 7.526014 (lr=1.4941e-04) (hash(x)=6245990)
565
+ 541 train 7.399398 (lr=1.4905e-04) (hash(x)=6480443)
566
+ 542 train 7.577604 (lr=1.4869e-04) (hash(x)=7680751)
567
+ 543 train 7.709227 (lr=1.4833e-04) (hash(x)=4122983)
568
+ 544 train 7.404023 (lr=1.4798e-04) (hash(x)=5996900)
569
+ 545 train 7.536460 (lr=1.4762e-04) (hash(x)=7672190)
570
+ 546 train 7.435640 (lr=1.4726e-04) (hash(x)=8149488)
571
+ 547 train 7.563015 (lr=1.4690e-04) (hash(x)=4724052)
572
+ 548 train 7.506015 (lr=1.4653e-04) (hash(x)=6812761)
573
+ 549 train 7.393397 (lr=1.4617e-04) (hash(x)=6654267)
574
+ 550 val loss 7.4660
575
+ 550 val perplexity 1747.6449
576
+ 550 train 7.447736 (lr=1.4581e-04) (hash(x)=6756431)
577
+ 551 train 7.419641 (lr=1.4545e-04) (hash(x)=6820797)
578
+ 552 train 7.526567 (lr=1.4508e-04) (hash(x)=5645711)
579
+ 553 train 7.484479 (lr=1.4472e-04) (hash(x)=6458662)
580
+ 554 train 7.512565 (lr=1.4435e-04) (hash(x)=5300771)
581
+ 555 train 8.306000 (lr=1.4398e-04) (hash(x)=7477594)
582
+ 556 train 7.597541 (lr=1.4362e-04) (hash(x)=5157748)
583
+ 557 train 7.543076 (lr=1.4325e-04) (hash(x)=5876574)
584
+ 558 train 7.309888 (lr=1.4288e-04) (hash(x)=6821258)
585
+ 559 train 7.277355 (lr=1.4251e-04) (hash(x)=6146305)
586
+ 560 train 7.566547 (lr=1.4214e-04) (hash(x)=5278517)
587
+ 561 train 7.367651 (lr=1.4177e-04) (hash(x)=6162080)
588
+ 562 train 7.345643 (lr=1.4140e-04) (hash(x)=5908198)
589
+ 563 train 7.488574 (lr=1.4103e-04) (hash(x)=6814831)
590
+ 564 train 7.561722 (lr=1.4066e-04) (hash(x)=5632357)
591
+ 565 train 7.570646 (lr=1.4029e-04) (hash(x)=7138295)
592
+ 566 train 7.177122 (lr=1.3991e-04) (hash(x)=4678562)
593
+ 567 train 7.459792 (lr=1.3954e-04) (hash(x)=5548298)
594
+ 568 train 7.317865 (lr=1.3916e-04) (hash(x)=6495991)
595
+ 569 train 7.557619 (lr=1.3879e-04) (hash(x)=6265078)
596
+ 570 train 7.739120 (lr=1.3841e-04) (hash(x)=5746797)
597
+ 571 train 7.888378 (lr=1.3804e-04) (hash(x)=7609416)
598
+ 572 train 7.620268 (lr=1.3766e-04) (hash(x)=5593869)
599
+ 573 train 7.395796 (lr=1.3728e-04) (hash(x)=6909367)
600
+ 574 train 7.440425 (lr=1.3691e-04) (hash(x)=4756034)
601
+ 575 train 7.793761 (lr=1.3653e-04) (hash(x)=7574342)
602
+ 576 train 7.209661 (lr=1.3615e-04) (hash(x)=4502769)
603
+ 577 train 7.629901 (lr=1.3577e-04) (hash(x)=8859748)
604
+ 578 train 7.388550 (lr=1.3539e-04) (hash(x)=5683686)
605
+ 579 train 7.375715 (lr=1.3501e-04) (hash(x)=5906937)
606
+ 580 train 7.494939 (lr=1.3463e-04) (hash(x)=6051136)
607
+ 581 train 7.248359 (lr=1.3425e-04) (hash(x)=7243024)
608
+ 582 train 7.309214 (lr=1.3387e-04) (hash(x)=5382761)
609
+ 583 train 7.755361 (lr=1.3348e-04) (hash(x)=8013471)
610
+ 584 train 7.367386 (lr=1.3310e-04) (hash(x)=5165270)
611
+ 585 train 7.490107 (lr=1.3272e-04) (hash(x)=5659461)
612
+ 586 train 7.552708 (lr=1.3234e-04) (hash(x)=5767800)
613
+ 587 train 7.378236 (lr=1.3195e-04) (hash(x)=6146213)
614
+ 588 train 7.737607 (lr=1.3157e-04) (hash(x)=7238407)
615
+ 589 train 7.284665 (lr=1.3118e-04) (hash(x)=6602917)
616
+ 590 train 7.479624 (lr=1.3080e-04) (hash(x)=6287715)
617
+ 591 train 7.367173 (lr=1.3041e-04) (hash(x)=6709902)
618
+ 592 train 7.484422 (lr=1.3003e-04) (hash(x)=6902255)
619
+ 593 train 7.556876 (lr=1.2964e-04) (hash(x)=8154341)
620
+ 594 train 7.789976 (lr=1.2925e-04) (hash(x)=5220688)
621
+ 595 train 7.564498 (lr=1.2887e-04) (hash(x)=6402571)
622
+ 596 train 7.369101 (lr=1.2848e-04) (hash(x)=5895167)
623
+ 597 train 7.490163 (lr=1.2809e-04) (hash(x)=5605466)
624
+ 598 train 7.430229 (lr=1.2770e-04) (hash(x)=6404902)
625
+ 599 train 7.359136 (lr=1.2732e-04) (hash(x)=7683614)
626
+ 600 val loss 7.4405
627
+ 600 val perplexity 1703.6161
628
+ 600 train 7.545094 (lr=1.2693e-04) (hash(x)=6103542)
629
+ 601 train 7.432226 (lr=1.2654e-04) (hash(x)=5421572)
630
+ 602 train 7.338719 (lr=1.2615e-04) (hash(x)=5808144)
631
+ 603 train 7.310085 (lr=1.2576e-04) (hash(x)=5453026)
632
+ 604 train 7.341038 (lr=1.2537e-04) (hash(x)=6337335)
633
+ 605 train 7.457764 (lr=1.2498e-04) (hash(x)=6437746)
634
+ 606 train 7.424649 (lr=1.2459e-04) (hash(x)=8782554)
635
+ 607 train 7.188511 (lr=1.2420e-04) (hash(x)=6260017)
636
+ 608 train 7.349472 (lr=1.2381e-04) (hash(x)=6917394)
637
+ 609 train 7.457290 (lr=1.2341e-04) (hash(x)=6119053)
638
+ 610 train 7.549181 (lr=1.2302e-04) (hash(x)=6089308)
639
+ 611 train 7.595739 (lr=1.2263e-04) (hash(x)=6591466)
640
+ 612 train 7.521837 (lr=1.2224e-04) (hash(x)=5858271)
641
+ 613 train 7.443123 (lr=1.2185e-04) (hash(x)=5652701)
642
+ 614 train 7.463419 (lr=1.2145e-04) (hash(x)=7058229)
643
+ 615 train 7.455755 (lr=1.2106e-04) (hash(x)=5501472)
644
+ 616 train 7.556617 (lr=1.2067e-04) (hash(x)=6232515)
645
+ 617 train 7.729628 (lr=1.2027e-04) (hash(x)=6160695)
646
+ 618 train 7.463953 (lr=1.1988e-04) (hash(x)=6020418)
647
+ 619 train 7.327691 (lr=1.1949e-04) (hash(x)=5871698)
648
+ 620 train 7.349918 (lr=1.1909e-04) (hash(x)=6796897)
649
+ 621 train 7.419732 (lr=1.1870e-04) (hash(x)=7301184)
650
+ 622 train 7.343307 (lr=1.1830e-04) (hash(x)=5748484)
651
+ 623 train 7.250406 (lr=1.1791e-04) (hash(x)=4282052)
652
+ 624 train 7.405371 (lr=1.1752e-04) (hash(x)=4758217)
653
+ 625 train 7.639436 (lr=1.1712e-04) (hash(x)=5679330)
654
+ 626 train 7.284977 (lr=1.1673e-04) (hash(x)=6726349)
655
+ 627 train 7.431141 (lr=1.1633e-04) (hash(x)=7824660)
656
+ 628 train 7.327853 (lr=1.1594e-04) (hash(x)=6255315)
657
+ 629 train 7.480049 (lr=1.1554e-04) (hash(x)=4857318)
658
+ 630 train 7.438340 (lr=1.1515e-04) (hash(x)=6974405)
659
+ 631 train 7.222921 (lr=1.1475e-04) (hash(x)=5954950)
660
+ 632 train 7.463291 (lr=1.1435e-04) (hash(x)=5981982)
661
+ 633 train 7.346389 (lr=1.1396e-04) (hash(x)=5758178)
662
+ 634 train 7.574354 (lr=1.1356e-04) (hash(x)=7816503)
663
+ 635 train 7.275682 (lr=1.1317e-04) (hash(x)=5484656)
664
+ 636 train 7.390398 (lr=1.1277e-04) (hash(x)=5513418)
665
+ 637 train 7.450412 (lr=1.1238e-04) (hash(x)=6818913)
666
+ 638 train 7.692029 (lr=1.1198e-04) (hash(x)=7444913)
667
+ 639 train 7.530225 (lr=1.1158e-04) (hash(x)=7031328)
668
+ 640 train 7.365683 (lr=1.1119e-04) (hash(x)=5202673)
669
+ 641 train 7.536321 (lr=1.1079e-04) (hash(x)=6837621)
670
+ 642 train 7.293514 (lr=1.1040e-04) (hash(x)=6590985)
671
+ 643 train 7.657380 (lr=1.1000e-04) (hash(x)=9157069)
672
+ 644 train 7.902295 (lr=1.0960e-04) (hash(x)=5123556)
673
+ 645 train 7.533281 (lr=1.0921e-04) (hash(x)=6509536)
674
+ 646 train 7.653763 (lr=1.0881e-04) (hash(x)=6252013)
675
+ 647 train 7.206101 (lr=1.0842e-04) (hash(x)=5996883)
676
+ 648 train 7.383912 (lr=1.0802e-04) (hash(x)=5942379)
677
+ 649 train 7.971305 (lr=1.0762e-04) (hash(x)=6850868)
678
+ 650 val loss 7.4150
679
+ 650 val perplexity 1660.7004
680
+ 650 train 7.558649 (lr=1.0723e-04) (hash(x)=8017172)
681
+ 651 train 7.673163 (lr=1.0683e-04) (hash(x)=7645447)
682
+ 652 train 7.534888 (lr=1.0644e-04) (hash(x)=8751332)
683
+ 653 train 7.678119 (lr=1.0604e-04) (hash(x)=6717515)
684
+ 654 train 7.824278 (lr=1.0565e-04) (hash(x)=6521028)
685
+ 655 train 7.652367 (lr=1.0525e-04) (hash(x)=6336164)
686
+ 656 train 7.434928 (lr=1.0485e-04) (hash(x)=6289337)
687
+ 657 train 7.467825 (lr=1.0446e-04) (hash(x)=6897445)
688
+ 658 train 7.582687 (lr=1.0406e-04) (hash(x)=7037289)
689
+ 659 train 7.138896 (lr=1.0367e-04) (hash(x)=5224240)
690
+ 660 train 7.474417 (lr=1.0327e-04) (hash(x)=6834715)
691
+ 661 train 7.537750 (lr=1.0288e-04) (hash(x)=6612945)
692
+ 662 train 8.162536 (lr=1.0248e-04) (hash(x)=8320588)
693
+ 663 train 7.794092 (lr=1.0209e-04) (hash(x)=5447801)
694
+ 664 train 7.463457 (lr=1.0170e-04) (hash(x)=6656376)
695
+ 665 train 7.532824 (lr=1.0130e-04) (hash(x)=7308579)
696
+ 666 train 7.588437 (lr=1.0091e-04) (hash(x)=5344256)
697
+ 667 train 7.183955 (lr=1.0051e-04) (hash(x)=4155598)
698
+ 668 train 7.408799 (lr=1.0012e-04) (hash(x)=5177976)
699
+ 669 train 7.376862 (lr=9.9726e-05) (hash(x)=6106773)
700
+ 670 train 7.360868 (lr=9.9333e-05) (hash(x)=8334761)
701
+ 671 train 7.513438 (lr=9.8940e-05) (hash(x)=7482744)
702
+ 672 train 7.789665 (lr=9.8547e-05) (hash(x)=6435528)
703
+ 673 train 7.197334 (lr=9.8154e-05) (hash(x)=6310687)
704
+ 674 train 7.224880 (lr=9.7762e-05) (hash(x)=6917344)
705
+ 675 train 7.527946 (lr=9.7370e-05) (hash(x)=4733414)
706
+ 676 train 7.284383 (lr=9.6978e-05) (hash(x)=5146340)
707
+ 677 train 7.713058 (lr=9.6586e-05) (hash(x)=6979734)
708
+ 678 train 7.231119 (lr=9.6195e-05) (hash(x)=5847143)
709
+ 679 train 7.321202 (lr=9.5804e-05) (hash(x)=7806284)
710
+ 680 train 7.225080 (lr=9.5413e-05) (hash(x)=6367699)
711
+ 681 train 7.385952 (lr=9.5022e-05) (hash(x)=5080536)
712
+ 682 train 7.574567 (lr=9.4632e-05) (hash(x)=7218763)
713
+ 683 train 7.460165 (lr=9.4242e-05) (hash(x)=7236478)
714
+ 684 train 7.423580 (lr=9.3852e-05) (hash(x)=5497206)
715
+ 685 train 7.496721 (lr=9.3463e-05) (hash(x)=6619318)
716
+ 686 train 7.429164 (lr=9.3073e-05) (hash(x)=5916845)
717
+ 687 train 7.275301 (lr=9.2685e-05) (hash(x)=4656255)
718
+ 688 train 7.279352 (lr=9.2296e-05) (hash(x)=5788889)
719
+ 689 train 7.313172 (lr=9.1908e-05) (hash(x)=5546888)
720
+ 690 train 7.473600 (lr=9.1520e-05) (hash(x)=8072073)
721
+ 691 train 8.035438 (lr=9.1133e-05) (hash(x)=9298325)
722
+ 692 train 7.505507 (lr=9.0746e-05) (hash(x)=5553198)
723
+ 693 train 7.372512 (lr=9.0359e-05) (hash(x)=5994232)
724
+ 694 train 7.415750 (lr=8.9973e-05) (hash(x)=7133014)
725
+ 695 train 7.459253 (lr=8.9587e-05) (hash(x)=6287524)
726
+ 696 train 7.340704 (lr=8.9202e-05) (hash(x)=4968095)
727
+ 697 train 7.751530 (lr=8.8817e-05) (hash(x)=6200049)
728
+ 698 train 7.416398 (lr=8.8432e-05) (hash(x)=6569299)
729
+ 699 train 7.483715 (lr=8.8048e-05) (hash(x)=7895970)
730
+ 700 val loss 7.3860
731
+ 700 val perplexity 1613.1682
732
+ 700 train 7.454215 (lr=8.7664e-05) (hash(x)=5286600)
733
+ 701 train 7.280346 (lr=8.7281e-05) (hash(x)=6051847)
734
+ 702 train 7.319325 (lr=8.6898e-05) (hash(x)=5156899)
735
+ 703 train 7.778841 (lr=8.6515e-05) (hash(x)=8847992)
736
+ 704 train 7.837794 (lr=8.6133e-05) (hash(x)=5343653)
737
+ 705 train 7.974396 (lr=8.5751e-05) (hash(x)=5123883)
738
+ 706 train 7.543467 (lr=8.5370e-05) (hash(x)=6462379)
739
+ 707 train 7.406644 (lr=8.4990e-05) (hash(x)=6338998)
740
+ 708 train 7.583449 (lr=8.4610e-05) (hash(x)=5803128)
741
+ 709 train 7.360124 (lr=8.4230e-05) (hash(x)=5765866)
742
+ 710 train 7.955447 (lr=8.3851e-05) (hash(x)=7297901)
743
+ 711 train 7.685194 (lr=8.3472e-05) (hash(x)=7104011)
744
+ 712 train 7.324003 (lr=8.3094e-05) (hash(x)=6511383)
745
+ 713 train 7.804550 (lr=8.2716e-05) (hash(x)=7438267)
746
+ 714 train 7.296929 (lr=8.2339e-05) (hash(x)=5600336)
747
+ 715 train 7.323614 (lr=8.1963e-05) (hash(x)=5573214)
748
+ 716 train 7.409211 (lr=8.1587e-05) (hash(x)=5751679)
749
+ 717 train 7.792267 (lr=8.1211e-05) (hash(x)=6897283)
750
+ 718 train 7.377743 (lr=8.0836e-05) (hash(x)=6011576)
751
+ 719 train 7.351501 (lr=8.0462e-05) (hash(x)=6882897)
752
+ 720 train 7.459682 (lr=8.0088e-05) (hash(x)=7223798)
753
+ 721 train 7.563742 (lr=7.9715e-05) (hash(x)=5343615)
754
+ 722 train 7.455894 (lr=7.9342e-05) (hash(x)=5500678)
755
+ 723 train 7.343488 (lr=7.8970e-05) (hash(x)=5854242)
756
+ 724 train 7.415850 (lr=7.8599e-05) (hash(x)=6087004)
757
+ 725 train 7.214909 (lr=7.8228e-05) (hash(x)=6317324)
758
+ 726 train 7.221733 (lr=7.7858e-05) (hash(x)=5502743)
759
+ 727 train 7.267812 (lr=7.7488e-05) (hash(x)=6972494)
760
+ 728 train 7.468316 (lr=7.7119e-05) (hash(x)=5138900)
761
+ 729 train 7.187045 (lr=7.6751e-05) (hash(x)=5661167)
762
+ 730 train 7.238564 (lr=7.6383e-05) (hash(x)=5204562)
763
+ 731 train 7.439147 (lr=7.6016e-05) (hash(x)=6796805)
764
+ 732 train 7.174314 (lr=7.5650e-05) (hash(x)=4915685)
765
+ 733 train 7.185508 (lr=7.5284e-05) (hash(x)=5817209)
766
+ 734 train 7.356492 (lr=7.4919e-05) (hash(x)=5270999)
767
+ 735 train 7.140251 (lr=7.4555e-05) (hash(x)=6139120)
768
+ 736 train 7.400005 (lr=7.4191e-05) (hash(x)=6348780)
769
+ 737 train 7.445010 (lr=7.3828e-05) (hash(x)=6121296)
770
+ 738 train 7.345733 (lr=7.3466e-05) (hash(x)=5527309)
771
+ 739 train 7.430100 (lr=7.3105e-05) (hash(x)=6486796)
772
+ 740 train 7.413500 (lr=7.2744e-05) (hash(x)=5859868)
773
+ 741 train 7.214375 (lr=7.2384e-05) (hash(x)=5514645)
774
+ 742 train 7.388248 (lr=7.2024e-05) (hash(x)=5740298)
775
+ 743 train 7.107208 (lr=7.1666e-05) (hash(x)=5578797)
776
+ 744 train 7.244393 (lr=7.1308e-05) (hash(x)=3634442)
777
+ 745 train 7.240322 (lr=7.0950e-05) (hash(x)=7651738)
778
+ 746 train 7.434947 (lr=7.0594e-05) (hash(x)=7118416)
779
+ 747 train 7.500448 (lr=7.0238e-05) (hash(x)=6895326)
780
+ 748 train 7.519388 (lr=6.9884e-05) (hash(x)=8607990)
781
+ 749 train 7.399383 (lr=6.9529e-05) (hash(x)=5490256)
782
+ 750 val loss 7.3695
783
+ 750 val perplexity 1586.8071
784
+ 750 train 7.350731 (lr=6.9176e-05) (hash(x)=6918464)
785
+ 751 train 7.310596 (lr=6.8824e-05) (hash(x)=7783160)
786
+ 752 train 7.313467 (lr=6.8472e-05) (hash(x)=4910007)
787
+ 753 train 7.727997 (lr=6.8121e-05) (hash(x)=5608424)
788
+ 754 train 7.517766 (lr=6.7771e-05) (hash(x)=5176743)
789
+ 755 train 7.105943 (lr=6.7422e-05) (hash(x)=4867625)
790
+ 756 train 7.233555 (lr=6.7073e-05) (hash(x)=4859474)
791
+ 757 train 7.200327 (lr=6.6725e-05) (hash(x)=4655472)
792
+ 758 train 7.231444 (lr=6.6379e-05) (hash(x)=4985126)
793
+ 759 train 7.316587 (lr=6.6033e-05) (hash(x)=6631123)
794
+ 760 train 7.122736 (lr=6.5688e-05) (hash(x)=3652957)
795
+ 761 train 6.768360 (lr=6.5343e-05) (hash(x)=5751990)
796
+ 762 train 7.144338 (lr=6.5000e-05) (hash(x)=6798916)
797
+ 763 train 7.391134 (lr=6.4657e-05) (hash(x)=6710965)
798
+ 764 train 7.293493 (lr=6.4316e-05) (hash(x)=5490166)
799
+ 765 train 7.143958 (lr=6.3975e-05) (hash(x)=5293800)
800
+ 766 train 7.154714 (lr=6.3635e-05) (hash(x)=5884835)
801
+ 767 train 7.336812 (lr=6.3296e-05) (hash(x)=6294657)
802
+ 768 train 7.522257 (lr=6.2958e-05) (hash(x)=7085154)
803
+ 769 train 7.309388 (lr=6.2621e-05) (hash(x)=6999538)
804
+ 770 train 7.473662 (lr=6.2285e-05) (hash(x)=5244581)
805
+ 771 train 7.308619 (lr=6.1950e-05) (hash(x)=5443606)
806
+ 772 train 7.423372 (lr=6.1615e-05) (hash(x)=7028408)
807
+ 773 train 7.297877 (lr=6.1282e-05) (hash(x)=5650401)
808
+ 774 train 7.382840 (lr=6.0949e-05) (hash(x)=6770125)
809
+ 775 train 7.368053 (lr=6.0618e-05) (hash(x)=7035770)
810
+ 776 train 7.235838 (lr=6.0287e-05) (hash(x)=6405982)
811
+ 777 train 7.433779 (lr=5.9958e-05) (hash(x)=6678271)
812
+ 778 train 7.341571 (lr=5.9629e-05) (hash(x)=6088760)
813
+ 779 train 7.710944 (lr=5.9301e-05) (hash(x)=4899541)
814
+ 780 train 7.227730 (lr=5.8974e-05) (hash(x)=5154342)
815
+ 781 train 7.303314 (lr=5.8649e-05) (hash(x)=5359260)
816
+ 782 train 7.120378 (lr=5.8324e-05) (hash(x)=5513509)
817
+ 783 train 7.397349 (lr=5.8000e-05) (hash(x)=6227120)
818
+ 784 train 7.289814 (lr=5.7678e-05) (hash(x)=5165184)
819
+ 785 train 7.351325 (lr=5.7356e-05) (hash(x)=6643916)
820
+ 786 train 7.387877 (lr=5.7035e-05) (hash(x)=7048323)
821
+ 787 train 7.391709 (lr=5.6716e-05) (hash(x)=5981648)
822
+ 788 train 7.256435 (lr=5.6397e-05) (hash(x)=4827872)
823
+ 789 train 7.451088 (lr=5.6079e-05) (hash(x)=6467540)
824
+ 790 train 7.321057 (lr=5.5763e-05) (hash(x)=6542722)
825
+ 791 train 7.256625 (lr=5.5447e-05) (hash(x)=5558993)
826
+ 792 train 7.475604 (lr=5.5133e-05) (hash(x)=5310679)
827
+ 793 train 7.456367 (lr=5.4820e-05) (hash(x)=6698959)
828
+ 794 train 7.151150 (lr=5.4507e-05) (hash(x)=6858783)
829
+ 795 train 7.364870 (lr=5.4196e-05) (hash(x)=7015671)
830
+ 796 train 7.295730 (lr=5.3886e-05) (hash(x)=4483663)
831
+ 797 train 7.257546 (lr=5.3577e-05) (hash(x)=6800890)
832
+ 798 train 7.268049 (lr=5.3269e-05) (hash(x)=6730827)
833
+ 799 train 7.625160 (lr=5.2962e-05) (hash(x)=5931234)
834
+ 800 val loss 7.3493
835
+ 800 val perplexity 1555.0948
836
+ 800 train 7.458741 (lr=5.2656e-05) (hash(x)=4866660)
837
+ 801 train 7.408505 (lr=5.2352e-05) (hash(x)=3995070)
838
+ 802 train 7.505974 (lr=5.2048e-05) (hash(x)=4488617)
839
+ 803 train 7.279483 (lr=5.1746e-05) (hash(x)=4582159)
840
+ 804 train 7.458854 (lr=5.1444e-05) (hash(x)=5164570)
841
+ 805 train 7.511346 (lr=5.1144e-05) (hash(x)=7081906)
842
+ 806 train 7.253983 (lr=5.0845e-05) (hash(x)=5480570)
843
+ 807 train 7.351478 (lr=5.0547e-05) (hash(x)=5029472)
844
+ 808 train 7.349893 (lr=5.0251e-05) (hash(x)=8207798)
845
+ 809 train 7.256050 (lr=4.9955e-05) (hash(x)=3946232)
846
+ 810 train 7.172788 (lr=4.9661e-05) (hash(x)=5652175)
847
+ 811 train 7.271476 (lr=4.9367e-05) (hash(x)=5434102)
848
+ 812 train 8.463399 (lr=4.9075e-05) (hash(x)=6558885)
849
+ 813 train 7.509143 (lr=4.8784e-05) (hash(x)=5340646)
850
+ 814 train 7.063549 (lr=4.8495e-05) (hash(x)=6403613)
851
+ 815 train 7.114406 (lr=4.8206e-05) (hash(x)=6810638)
852
+ 816 train 7.039555 (lr=4.7919e-05) (hash(x)=4382313)
853
+ 817 train 7.692449 (lr=4.7633e-05) (hash(x)=6768026)
854
+ 818 train 7.325496 (lr=4.7348e-05) (hash(x)=6423059)
855
+ 819 train 7.376756 (lr=4.7064e-05) (hash(x)=6521562)
856
+ 820 train 7.334204 (lr=4.6782e-05) (hash(x)=5653405)
857
+ 821 train 7.346684 (lr=4.6501e-05) (hash(x)=4429434)
858
+ 822 train 7.463165 (lr=4.6221e-05) (hash(x)=6885745)
859
+ 823 train 7.347682 (lr=4.5942e-05) (hash(x)=7226932)
860
+ 824 train 7.130199 (lr=4.5664e-05) (hash(x)=5371082)
861
+ 825 train 7.245062 (lr=4.5388e-05) (hash(x)=5065932)
862
+ 826 train 7.262284 (lr=4.5113e-05) (hash(x)=5384078)
863
+ 827 train 7.408626 (lr=4.4839e-05) (hash(x)=6957849)
864
+ 828 train 7.490227 (lr=4.4567e-05) (hash(x)=6049310)
865
+ 829 train 7.180797 (lr=4.4295e-05) (hash(x)=6431457)
866
+ 830 train 7.513694 (lr=4.4025e-05) (hash(x)=6479516)
867
+ 831 train 7.297242 (lr=4.3757e-05) (hash(x)=5334281)
868
+ 832 train 7.234174 (lr=4.3489e-05) (hash(x)=6811166)
869
+ 833 train 7.625687 (lr=4.3223e-05) (hash(x)=5471747)
870
+ 834 train 7.513314 (lr=4.2958e-05) (hash(x)=6733641)
871
+ 835 train 7.488129 (lr=4.2695e-05) (hash(x)=5887983)
872
+ 836 train 7.503080 (lr=4.2432e-05) (hash(x)=5366857)
873
+ 837 train 7.310686 (lr=4.2171e-05) (hash(x)=5075909)
874
+ 838 train 7.186604 (lr=4.1912e-05) (hash(x)=4961569)
875
+ 839 train 7.248739 (lr=4.1654e-05) (hash(x)=6393576)
876
+ 840 train 7.363343 (lr=4.1397e-05) (hash(x)=5520210)
877
+ 841 train 7.311482 (lr=4.1141e-05) (hash(x)=6608979)
878
+ 842 train 7.643979 (lr=4.0887e-05) (hash(x)=6223763)
879
+ 843 train 7.836648 (lr=4.0634e-05) (hash(x)=5057959)
880
+ 844 train 7.649400 (lr=4.0382e-05) (hash(x)=8268189)
881
+ 845 train 7.720472 (lr=4.0132e-05) (hash(x)=6414858)
882
+ 846 train 7.264541 (lr=3.9883e-05) (hash(x)=6468918)
883
+ 847 train 7.648538 (lr=3.9635e-05) (hash(x)=8595271)
884
+ 848 train 7.310149 (lr=3.9389e-05) (hash(x)=5833443)
885
+ 849 train 7.305192 (lr=3.9144e-05) (hash(x)=6500949)
886
+ 850 val loss 7.3392
887
+ 850 val perplexity 1539.4948
888
+ 850 train 7.299557 (lr=3.8901e-05) (hash(x)=5866207)
889
+ 851 train 7.662872 (lr=3.8659e-05) (hash(x)=8395167)
890
+ 852 train 7.671334 (lr=3.8418e-05) (hash(x)=8155803)
891
+ 853 train 7.554940 (lr=3.8178e-05) (hash(x)=6288420)
892
+ 854 train 7.520384 (lr=3.7941e-05) (hash(x)=6770885)
893
+ 855 train 7.317709 (lr=3.7704e-05) (hash(x)=7634417)
894
+ 856 train 7.251218 (lr=3.7469e-05) (hash(x)=6102598)
895
+ 857 train 7.333034 (lr=3.7235e-05) (hash(x)=6795561)
896
+ 858 train 7.271451 (lr=3.7003e-05) (hash(x)=6178545)
897
+ 859 train 7.141239 (lr=3.6772e-05) (hash(x)=5257771)
898
+ 860 train 7.084524 (lr=3.6542e-05) (hash(x)=5794385)
899
+ 861 train 7.620108 (lr=3.6314e-05) (hash(x)=7506860)
900
+ 862 train 7.432599 (lr=3.6088e-05) (hash(x)=4617636)
901
+ 863 train 7.405635 (lr=3.5862e-05) (hash(x)=7674238)
902
+ 864 train 7.862397 (lr=3.5639e-05) (hash(x)=7502317)
903
+ 865 train 7.228305 (lr=3.5416e-05) (hash(x)=5353794)
904
+ 866 train 7.744471 (lr=3.5195e-05) (hash(x)=8474964)
905
+ 867 train 7.170532 (lr=3.4976e-05) (hash(x)=5300555)
906
+ 868 train 7.049745 (lr=3.4758e-05) (hash(x)=6638039)
907
+ 869 train 7.231029 (lr=3.4541e-05) (hash(x)=5564419)
908
+ 870 train 7.559051 (lr=3.4326e-05) (hash(x)=6466311)
909
+ 871 train 7.285535 (lr=3.4113e-05) (hash(x)=6066139)
910
+ 872 train 7.427622 (lr=3.3900e-05) (hash(x)=6790356)
911
+ 873 train 7.328134 (lr=3.3690e-05) (hash(x)=6863709)
912
+ 874 train 7.480557 (lr=3.3480e-05) (hash(x)=6872703)
913
+ 875 train 7.264747 (lr=3.3273e-05) (hash(x)=6688212)
914
+ 876 train 7.451789 (lr=3.3066e-05) (hash(x)=4838260)
915
+ 877 train 7.605741 (lr=3.2862e-05) (hash(x)=7095090)
916
+ 878 train 7.793611 (lr=3.2658e-05) (hash(x)=6775236)
917
+ 879 train 7.435330 (lr=3.2457e-05) (hash(x)=6642404)
918
+ 880 train 7.261833 (lr=3.2256e-05) (hash(x)=6365570)
919
+ 881 train 7.394926 (lr=3.2058e-05) (hash(x)=6539227)
920
+ 882 train 7.354235 (lr=3.1860e-05) (hash(x)=5854007)
921
+ 883 train 7.299989 (lr=3.1665e-05) (hash(x)=4961570)
922
+ 884 train 7.110744 (lr=3.1471e-05) (hash(x)=3961674)
923
+ 885 train 7.415573 (lr=3.1278e-05) (hash(x)=6606391)
924
+ 886 train 6.855542 (lr=3.1087e-05) (hash(x)=6179396)
925
+ 887 train 7.271055 (lr=3.0897e-05) (hash(x)=5105190)
926
+ 888 train 7.747058 (lr=3.0709e-05) (hash(x)=5970936)
927
+ 889 train 7.538048 (lr=3.0522e-05) (hash(x)=5695997)
928
+ 890 train 7.275749 (lr=3.0337e-05) (hash(x)=6171397)
929
+ 891 train 7.429681 (lr=3.0154e-05) (hash(x)=7155991)
930
+ 892 train 7.176868 (lr=2.9972e-05) (hash(x)=6739404)
931
+ 893 train 7.485420 (lr=2.9791e-05) (hash(x)=7249727)
932
+ 894 train 7.411972 (lr=2.9613e-05) (hash(x)=6782303)
933
+ 895 train 7.204674 (lr=2.9435e-05) (hash(x)=5132972)
934
+ 896 train 7.413833 (lr=2.9260e-05) (hash(x)=5962092)
935
+ 897 train 7.197115 (lr=2.9085e-05) (hash(x)=5381802)
936
+ 898 train 7.222078 (lr=2.8913e-05) (hash(x)=6546040)
937
+ 899 train 7.466074 (lr=2.8742e-05) (hash(x)=6269946)
938
+ 900 val loss 7.3256
939
+ 900 val perplexity 1518.7249
940
+ 900 train 7.366619 (lr=2.8572e-05) (hash(x)=5427088)
941
+ 901 train 7.250206 (lr=2.8404e-05) (hash(x)=4976762)
942
+ 902 train 7.396802 (lr=2.8238e-05) (hash(x)=5758632)
943
+ 903 train 7.392920 (lr=2.8073e-05) (hash(x)=5396133)
944
+ 904 train 7.434168 (lr=2.7910e-05) (hash(x)=6805706)
945
+ 905 train 7.267515 (lr=2.7749e-05) (hash(x)=6173473)
946
+ 906 train 7.447976 (lr=2.7589e-05) (hash(x)=6565677)
947
+ 907 train 7.199090 (lr=2.7430e-05) (hash(x)=6394083)
948
+ 908 train 7.370712 (lr=2.7274e-05) (hash(x)=6445589)
949
+ 909 train 7.390798 (lr=2.7119e-05) (hash(x)=6436030)
950
+ 910 train 7.257186 (lr=2.6965e-05) (hash(x)=7220327)
951
+ 911 train 7.240999 (lr=2.6813e-05) (hash(x)=5237804)
952
+ 912 train 7.197953 (lr=2.6663e-05) (hash(x)=4906982)
953
+ 913 train 7.250177 (lr=2.6514e-05) (hash(x)=6540397)
954
+ 914 train 7.445331 (lr=2.6367e-05) (hash(x)=5799609)
955
+ 915 train 7.526650 (lr=2.6221e-05) (hash(x)=5216638)
956
+ 916 train 7.074869 (lr=2.6077e-05) (hash(x)=5646696)
957
+ 917 train 7.509258 (lr=2.5935e-05) (hash(x)=5750537)
958
+ 918 train 7.268204 (lr=2.5795e-05) (hash(x)=6625042)
959
+ 919 train 7.320524 (lr=2.5656e-05) (hash(x)=6735565)
960
+ 920 train 7.264712 (lr=2.5518e-05) (hash(x)=6600526)
961
+ 921 train 7.356437 (lr=2.5383e-05) (hash(x)=7297488)
962
+ 922 train 7.449017 (lr=2.5249e-05) (hash(x)=8039568)
963
+ 923 train 7.626235 (lr=2.5116e-05) (hash(x)=5983575)
964
+ 924 train 7.347427 (lr=2.4985e-05) (hash(x)=5322135)
965
+ 925 train 7.307502 (lr=2.4856e-05) (hash(x)=4808543)
966
+ 926 train 7.374880 (lr=2.4729e-05) (hash(x)=6387352)
967
+ 927 train 7.479345 (lr=2.4603e-05) (hash(x)=5831619)
968
+ 928 train 7.205617 (lr=2.4479e-05) (hash(x)=5418207)
969
+ 929 train 7.190817 (lr=2.4356e-05) (hash(x)=6280052)
970
+ 930 train 7.312872 (lr=2.4235e-05) (hash(x)=5668790)
971
+ 931 train 7.165340 (lr=2.4116e-05) (hash(x)=9109083)
972
+ 932 train 7.930820 (lr=2.3998e-05) (hash(x)=7957194)
973
+ 933 train 7.036334 (lr=2.3883e-05) (hash(x)=5848604)
974
+ 934 train 7.299820 (lr=2.3768e-05) (hash(x)=6466898)
975
+ 935 train 7.517517 (lr=2.3656e-05) (hash(x)=6912967)
976
+ 936 train 7.635291 (lr=2.3545e-05) (hash(x)=6802045)
977
+ 937 train 7.552060 (lr=2.3436e-05) (hash(x)=6615365)
978
+ 938 train 7.731026 (lr=2.3328e-05) (hash(x)=5956271)
979
+ 939 train 7.604625 (lr=2.3222e-05) (hash(x)=5965674)
980
+ 940 train 7.419561 (lr=2.3118e-05) (hash(x)=5925388)
981
+ 941 train 7.552092 (lr=2.3016e-05) (hash(x)=6879965)
982
+ 942 train 7.668989 (lr=2.2915e-05) (hash(x)=6850068)
983
+ 943 train 7.206676 (lr=2.2816e-05) (hash(x)=5822533)
984
+ 944 train 7.239583 (lr=2.2718e-05) (hash(x)=5175909)
985
+ 945 train 7.359796 (lr=2.2623e-05) (hash(x)=5794462)
986
+ 946 train 7.299229 (lr=2.2528e-05) (hash(x)=5451297)
987
+ 947 train 7.454557 (lr=2.2436e-05) (hash(x)=5268216)
988
+ 948 train 7.228423 (lr=2.2345e-05) (hash(x)=5828351)
989
+ 949 train 7.166238 (lr=2.2256e-05) (hash(x)=7058396)
990
+ 950 val loss 7.3174
991
+ 950 val perplexity 1506.2653
992
+ 950 train 7.175211 (lr=2.2169e-05) (hash(x)=6202824)
993
+ 951 train 7.443611 (lr=2.2084e-05) (hash(x)=6531900)
994
+ 952 train 7.320154 (lr=2.2000e-05) (hash(x)=6989972)
995
+ 953 train 7.332245 (lr=2.1918e-05) (hash(x)=5878912)
996
+ 954 train 7.128870 (lr=2.1837e-05) (hash(x)=7862929)
997
+ 955 train 7.157946 (lr=2.1758e-05) (hash(x)=5426499)
998
+ 956 train 7.127827 (lr=2.1681e-05) (hash(x)=5926459)
999
+ 957 train 7.573450 (lr=2.1606e-05) (hash(x)=8107004)
1000
+ 958 train 8.060073 (lr=2.1532e-05) (hash(x)=8057219)
1001
+ 959 train 8.048660 (lr=2.1461e-05) (hash(x)=7455372)
1002
+ 960 train 7.980464 (lr=2.1390e-05) (hash(x)=7083241)
1003
+ 961 train 8.033175 (lr=2.1322e-05) (hash(x)=7727959)
1004
+ 962 train 7.958523 (lr=2.1255e-05) (hash(x)=8069628)
1005
+ 963 train 7.980753 (lr=2.1190e-05) (hash(x)=8319119)
1006
+ 964 train 7.934962 (lr=2.1127e-05) (hash(x)=8080739)
1007
+ 965 train 7.928865 (lr=2.1065e-05) (hash(x)=7581656)
1008
+ 966 train 7.977360 (lr=2.1005e-05) (hash(x)=7644116)
1009
+ 967 train 8.031235 (lr=2.0947e-05) (hash(x)=7592128)
1010
+ 968 train 7.900897 (lr=2.0891e-05) (hash(x)=7555224)
1011
+ 969 train 8.108923 (lr=2.0836e-05) (hash(x)=10506633)
1012
+ 970 train 7.927726 (lr=2.0783e-05) (hash(x)=7171723)
1013
+ 971 train 8.066471 (lr=2.0732e-05) (hash(x)=8704170)
1014
+ 972 train 7.973075 (lr=2.0682e-05) (hash(x)=6236578)
1015
+ 973 train 7.409932 (lr=2.0634e-05) (hash(x)=7461138)
1016
+ 974 train 7.402748 (lr=2.0588e-05) (hash(x)=6802094)
1017
+ 975 train 7.519301 (lr=2.0544e-05) (hash(x)=7165037)
1018
+ 976 train 7.753916 (lr=2.0501e-05) (hash(x)=7619348)
1019
+ 977 train 7.418281 (lr=2.0460e-05) (hash(x)=4976998)
1020
+ 978 train 7.362858 (lr=2.0421e-05) (hash(x)=7270148)
1021
+ 979 train 7.199983 (lr=2.0384e-05) (hash(x)=7678000)
1022
+ 980 train 7.380810 (lr=2.0348e-05) (hash(x)=5853099)
1023
+ 981 train 7.287700 (lr=2.0314e-05) (hash(x)=6299409)
1024
+ 982 train 7.305512 (lr=2.0282e-05) (hash(x)=5422994)
1025
+ 983 train 7.211154 (lr=2.0252e-05) (hash(x)=7414788)
1026
+ 984 train 7.545966 (lr=2.0223e-05) (hash(x)=5094219)
1027
+ 985 train 7.205543 (lr=2.0196e-05) (hash(x)=5063061)
1028
+ 986 train 7.251679 (lr=2.0171e-05) (hash(x)=5964965)
1029
+ 987 train 7.354722 (lr=2.0147e-05) (hash(x)=7326216)
1030
+ 988 train 7.470905 (lr=2.0125e-05) (hash(x)=7677783)
1031
+ 989 train 7.400064 (lr=2.0105e-05) (hash(x)=7492012)
1032
+ 990 train 7.365852 (lr=2.0087e-05) (hash(x)=5898822)
1033
+ 991 train 7.444882 (lr=2.0071e-05) (hash(x)=5857162)
1034
+ 992 train 7.166565 (lr=2.0056e-05) (hash(x)=5675101)
1035
+ 993 train 7.253083 (lr=2.0043e-05) (hash(x)=6169001)
1036
+ 994 train 7.388286 (lr=2.0031e-05) (hash(x)=6691027)
1037
+ 995 train 10.628089 (lr=2.0022e-05) (hash(x)=14715669)
1038
+ 996 train 11.674992 (lr=2.0014e-05) (hash(x)=14724033)
1039
+ 997 train 11.618856 (lr=2.0008e-05) (hash(x)=18256706)
1040
+ 998 train 12.142720 (lr=2.0003e-05) (hash(x)=16949966)
1041
+ 999 val loss 7.3127
1042
+ 999 val perplexity 1499.2050
1043
+ 999 train 11.061088 (lr=2.0001e-05) (hash(x)=16950560)
lr2e-4_total_batch_size10240_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d0ffe02431d9b6f2e86c0d584e6b5ff59920f6ebf5303a019054c15d6d4850e
3
+ size 36163010
lr2e-4_total_batch_size10240_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31e9dc383ba6224a3e7012ca761db67b3bff2547edec125c5bd7aaab58cf29b0
3
+ size 70764358