andrew-healey commited on
Commit
6701ffd
·
verified ·
1 Parent(s): 0693dc4

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size10240_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-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": 0.0001, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1e-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
lr1e-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=3.4965e-07) (hash(x)=6605789)
5
+ 1 train 11.715714 (lr=6.9930e-07) (hash(x)=5113283)
6
+ 2 train 11.677877 (lr=1.0490e-06) (hash(x)=7007748)
7
+ 3 train 11.696216 (lr=1.3986e-06) (hash(x)=5534142)
8
+ 4 train 11.698164 (lr=1.7483e-06) (hash(x)=5974274)
9
+ 5 train 11.693561 (lr=2.0979e-06) (hash(x)=6763231)
10
+ 6 train 11.663198 (lr=2.4476e-06) (hash(x)=6706513)
11
+ 7 train 11.692604 (lr=2.7972e-06) (hash(x)=6328012)
12
+ 8 train 11.669147 (lr=3.1469e-06) (hash(x)=6183012)
13
+ 9 train 11.655550 (lr=3.4965e-06) (hash(x)=7052161)
14
+ 10 train 11.631498 (lr=3.8462e-06) (hash(x)=6586466)
15
+ 11 train 11.661675 (lr=4.1958e-06) (hash(x)=5556098)
16
+ 12 train 11.645960 (lr=4.5455e-06) (hash(x)=8155572)
17
+ 13 train 11.602326 (lr=4.8951e-06) (hash(x)=4696943)
18
+ 14 train 11.583433 (lr=5.2448e-06) (hash(x)=7649389)
19
+ 15 train 11.568135 (lr=5.5944e-06) (hash(x)=6219138)
20
+ 16 train 11.537117 (lr=5.9441e-06) (hash(x)=7075794)
21
+ 17 train 11.528027 (lr=6.2937e-06) (hash(x)=6049878)
22
+ 18 train 11.538563 (lr=6.6434e-06) (hash(x)=6385927)
23
+ 19 train 11.504963 (lr=6.9930e-06) (hash(x)=5491972)
24
+ 20 train 11.490391 (lr=7.3427e-06) (hash(x)=6061420)
25
+ 21 train 11.492226 (lr=7.6923e-06) (hash(x)=8075458)
26
+ 22 train 11.383057 (lr=8.0420e-06) (hash(x)=7238769)
27
+ 23 train 11.447841 (lr=8.3916e-06) (hash(x)=5917741)
28
+ 24 train 11.400802 (lr=8.7413e-06) (hash(x)=5430605)
29
+ 25 train 11.395272 (lr=9.0909e-06) (hash(x)=5750403)
30
+ 26 train 11.354577 (lr=9.4406e-06) (hash(x)=4792640)
31
+ 27 train 11.341745 (lr=9.7902e-06) (hash(x)=6112458)
32
+ 28 train 11.292657 (lr=1.0140e-05) (hash(x)=5801122)
33
+ 29 train 11.317315 (lr=1.0490e-05) (hash(x)=5806490)
34
+ 30 train 11.257146 (lr=1.0839e-05) (hash(x)=5754140)
35
+ 31 train 11.199593 (lr=1.1189e-05) (hash(x)=6568379)
36
+ 32 train 11.133384 (lr=1.1538e-05) (hash(x)=7134932)
37
+ 33 train 11.164309 (lr=1.1888e-05) (hash(x)=7212403)
38
+ 34 train 11.064849 (lr=1.2238e-05) (hash(x)=7720703)
39
+ 35 train 11.096609 (lr=1.2587e-05) (hash(x)=7386082)
40
+ 36 train 11.024403 (lr=1.2937e-05) (hash(x)=6848513)
41
+ 37 train 11.035874 (lr=1.3287e-05) (hash(x)=6359743)
42
+ 38 train 10.946769 (lr=1.3636e-05) (hash(x)=7536538)
43
+ 39 train 10.977413 (lr=1.3986e-05) (hash(x)=6151674)
44
+ 40 train 10.942835 (lr=1.4336e-05) (hash(x)=6317099)
45
+ 41 train 10.886848 (lr=1.4685e-05) (hash(x)=7168704)
46
+ 42 train 10.885317 (lr=1.5035e-05) (hash(x)=6736092)
47
+ 43 train 10.812740 (lr=1.5385e-05) (hash(x)=7859060)
48
+ 44 train 10.821023 (lr=1.5734e-05) (hash(x)=6958223)
49
+ 45 train 10.686498 (lr=1.6084e-05) (hash(x)=6976239)
50
+ 46 train 10.730842 (lr=1.6434e-05) (hash(x)=6956845)
51
+ 47 train 10.738949 (lr=1.6783e-05) (hash(x)=6391506)
52
+ 48 train 10.788889 (lr=1.7133e-05) (hash(x)=6113701)
53
+ 49 train 10.591890 (lr=1.7483e-05) (hash(x)=5245999)
54
+ 50 val loss 10.6346
55
+ 50 val perplexity 41549.5352
56
+ 50 train 10.623287 (lr=1.7832e-05) (hash(x)=5774189)
57
+ 51 train 10.636392 (lr=1.8182e-05) (hash(x)=5985675)
58
+ 52 train 10.594037 (lr=1.8531e-05) (hash(x)=6451776)
59
+ 53 train 10.461292 (lr=1.8881e-05) (hash(x)=4355733)
60
+ 54 train 10.536674 (lr=1.9231e-05) (hash(x)=6669188)
61
+ 55 train 10.767937 (lr=1.9580e-05) (hash(x)=6984772)
62
+ 56 train 10.482365 (lr=1.9930e-05) (hash(x)=7440580)
63
+ 57 train 10.540380 (lr=2.0280e-05) (hash(x)=6244096)
64
+ 58 train 10.531529 (lr=2.0629e-05) (hash(x)=8381560)
65
+ 59 train 10.482718 (lr=2.0979e-05) (hash(x)=6603717)
66
+ 60 train 10.406415 (lr=2.1329e-05) (hash(x)=6437516)
67
+ 61 train 10.411435 (lr=2.1678e-05) (hash(x)=5669522)
68
+ 62 train 10.355264 (lr=2.2028e-05) (hash(x)=5505889)
69
+ 63 train 10.414718 (lr=2.2378e-05) (hash(x)=6190579)
70
+ 64 train 10.417758 (lr=2.2727e-05) (hash(x)=5153196)
71
+ 65 train 10.457509 (lr=2.3077e-05) (hash(x)=4912983)
72
+ 66 train 10.406000 (lr=2.3427e-05) (hash(x)=5109125)
73
+ 67 train 10.235314 (lr=2.3776e-05) (hash(x)=7276303)
74
+ 68 train 10.400863 (lr=2.4126e-05) (hash(x)=6645716)
75
+ 69 train 10.332122 (lr=2.4476e-05) (hash(x)=5479947)
76
+ 70 train 10.451086 (lr=2.4825e-05) (hash(x)=7850360)
77
+ 71 train 10.455929 (lr=2.5175e-05) (hash(x)=6630378)
78
+ 72 train 10.349250 (lr=2.5524e-05) (hash(x)=6295358)
79
+ 73 train 10.428650 (lr=2.5874e-05) (hash(x)=8097010)
80
+ 74 train 10.360823 (lr=2.6224e-05) (hash(x)=3958188)
81
+ 75 train 10.521447 (lr=2.6573e-05) (hash(x)=9989744)
82
+ 76 train 10.334774 (lr=2.6923e-05) (hash(x)=6343832)
83
+ 77 train 10.311204 (lr=2.7273e-05) (hash(x)=6694932)
84
+ 78 train 10.278445 (lr=2.7622e-05) (hash(x)=6429137)
85
+ 79 train 10.220717 (lr=2.7972e-05) (hash(x)=6317894)
86
+ 80 train 10.144623 (lr=2.8322e-05) (hash(x)=5165595)
87
+ 81 train 10.203988 (lr=2.8671e-05) (hash(x)=5046327)
88
+ 82 train 10.299314 (lr=2.9021e-05) (hash(x)=7060363)
89
+ 83 train 10.193643 (lr=2.9371e-05) (hash(x)=7947703)
90
+ 84 train 10.129883 (lr=2.9720e-05) (hash(x)=5971240)
91
+ 85 train 10.101086 (lr=3.0070e-05) (hash(x)=5916068)
92
+ 86 train 10.165997 (lr=3.0420e-05) (hash(x)=7426277)
93
+ 87 train 10.254584 (lr=3.0769e-05) (hash(x)=5777498)
94
+ 88 train 10.221705 (lr=3.1119e-05) (hash(x)=6403331)
95
+ 89 train 10.059228 (lr=3.1469e-05) (hash(x)=5753092)
96
+ 90 train 10.120261 (lr=3.1818e-05) (hash(x)=5981486)
97
+ 91 train 10.138000 (lr=3.2168e-05) (hash(x)=7723226)
98
+ 92 train 10.086612 (lr=3.2517e-05) (hash(x)=6069558)
99
+ 93 train 10.141511 (lr=3.2867e-05) (hash(x)=7007016)
100
+ 94 train 10.032905 (lr=3.3217e-05) (hash(x)=5027958)
101
+ 95 train 10.056123 (lr=3.3566e-05) (hash(x)=6491229)
102
+ 96 train 10.060586 (lr=3.3916e-05) (hash(x)=7036653)
103
+ 97 train 10.112430 (lr=3.4266e-05) (hash(x)=6390752)
104
+ 98 train 10.217685 (lr=3.4615e-05) (hash(x)=7758167)
105
+ 99 train 10.176164 (lr=3.4965e-05) (hash(x)=6019389)
106
+ 100 val loss 10.0864
107
+ 100 val perplexity 24014.5977
108
+ 100 train 10.158809 (lr=3.5315e-05) (hash(x)=6503191)
109
+ 101 train 9.981990 (lr=3.5664e-05) (hash(x)=7662067)
110
+ 102 train 10.049532 (lr=3.6014e-05) (hash(x)=7901992)
111
+ 103 train 10.092311 (lr=3.6364e-05) (hash(x)=4648609)
112
+ 104 train 10.026969 (lr=3.6713e-05) (hash(x)=5897288)
113
+ 105 train 9.998008 (lr=3.7063e-05) (hash(x)=4725966)
114
+ 106 train 9.951184 (lr=3.7413e-05) (hash(x)=6243442)
115
+ 107 train 10.000803 (lr=3.7762e-05) (hash(x)=5558355)
116
+ 108 train 10.049534 (lr=3.8112e-05) (hash(x)=5400407)
117
+ 109 train 9.979938 (lr=3.8462e-05) (hash(x)=6524933)
118
+ 110 train 9.982258 (lr=3.8811e-05) (hash(x)=5889724)
119
+ 111 train 9.947940 (lr=3.9161e-05) (hash(x)=6369326)
120
+ 112 train 9.881700 (lr=3.9510e-05) (hash(x)=5911906)
121
+ 113 train 9.994719 (lr=3.9860e-05) (hash(x)=6022625)
122
+ 114 train 9.846291 (lr=4.0210e-05) (hash(x)=6584810)
123
+ 115 train 9.940397 (lr=4.0559e-05) (hash(x)=4906853)
124
+ 116 train 9.738394 (lr=4.0909e-05) (hash(x)=4564564)
125
+ 117 train 9.710005 (lr=4.1259e-05) (hash(x)=5589269)
126
+ 118 train 9.845630 (lr=4.1608e-05) (hash(x)=6969180)
127
+ 119 train 9.838554 (lr=4.1958e-05) (hash(x)=6437376)
128
+ 120 train 9.881484 (lr=4.2308e-05) (hash(x)=7116258)
129
+ 121 train 9.859009 (lr=4.2657e-05) (hash(x)=7304892)
130
+ 122 train 9.838655 (lr=4.3007e-05) (hash(x)=7366721)
131
+ 123 train 9.825720 (lr=4.3357e-05) (hash(x)=7084093)
132
+ 124 train 9.906609 (lr=4.3706e-05) (hash(x)=7197652)
133
+ 125 train 9.792642 (lr=4.4056e-05) (hash(x)=5908178)
134
+ 126 train 9.634016 (lr=4.4406e-05) (hash(x)=6349974)
135
+ 127 train 9.669449 (lr=4.4755e-05) (hash(x)=7048804)
136
+ 128 train 9.718451 (lr=4.5105e-05) (hash(x)=7909039)
137
+ 129 train 9.742558 (lr=4.5455e-05) (hash(x)=6085549)
138
+ 130 train 9.775120 (lr=4.5804e-05) (hash(x)=7463181)
139
+ 131 train 9.685642 (lr=4.6154e-05) (hash(x)=5752594)
140
+ 132 train 9.536407 (lr=4.6503e-05) (hash(x)=6114601)
141
+ 133 train 9.566474 (lr=4.6853e-05) (hash(x)=7662769)
142
+ 134 train 9.543540 (lr=4.7203e-05) (hash(x)=7485666)
143
+ 135 train 9.284313 (lr=4.7552e-05) (hash(x)=6393520)
144
+ 136 train 9.599241 (lr=4.7902e-05) (hash(x)=5586291)
145
+ 137 train 9.706452 (lr=4.8252e-05) (hash(x)=5942867)
146
+ 138 train 9.675098 (lr=4.8601e-05) (hash(x)=7574027)
147
+ 139 train 9.608974 (lr=4.8951e-05) (hash(x)=6793550)
148
+ 140 train 9.516984 (lr=4.9301e-05) (hash(x)=6264789)
149
+ 141 train 9.483897 (lr=4.9650e-05) (hash(x)=7105976)
150
+ 142 train 9.519382 (lr=5.0000e-05) (hash(x)=7244279)
151
+ 143 train 9.530899 (lr=5.0350e-05) (hash(x)=6581348)
152
+ 144 train 9.495308 (lr=5.0699e-05) (hash(x)=6543825)
153
+ 145 train 9.433113 (lr=5.1049e-05) (hash(x)=5555598)
154
+ 146 train 9.428387 (lr=5.1399e-05) (hash(x)=7254232)
155
+ 147 train 9.366930 (lr=5.1748e-05) (hash(x)=6235837)
156
+ 148 train 9.337235 (lr=5.2098e-05) (hash(x)=4738550)
157
+ 149 train 9.329519 (lr=5.2448e-05) (hash(x)=5591770)
158
+ 150 val loss 9.3583
159
+ 150 val perplexity 11594.8623
160
+ 150 train 9.720510 (lr=5.2797e-05) (hash(x)=8012491)
161
+ 151 train 9.331465 (lr=5.3147e-05) (hash(x)=7083021)
162
+ 152 train 9.370584 (lr=5.3497e-05) (hash(x)=6456022)
163
+ 153 train 9.336122 (lr=5.3846e-05) (hash(x)=7281348)
164
+ 154 train 9.382425 (lr=5.4196e-05) (hash(x)=7824794)
165
+ 155 train 9.377480 (lr=5.4545e-05) (hash(x)=6155821)
166
+ 156 train 9.306238 (lr=5.4895e-05) (hash(x)=7807564)
167
+ 157 train 9.142947 (lr=5.5245e-05) (hash(x)=5736241)
168
+ 158 train 9.229507 (lr=5.5594e-05) (hash(x)=5688415)
169
+ 159 train 9.050955 (lr=5.5944e-05) (hash(x)=5657123)
170
+ 160 train 9.043358 (lr=5.6294e-05) (hash(x)=6102985)
171
+ 161 train 9.007455 (lr=5.6643e-05) (hash(x)=6452095)
172
+ 162 train 9.028298 (lr=5.6993e-05) (hash(x)=5848921)
173
+ 163 train 9.049395 (lr=5.7343e-05) (hash(x)=5312267)
174
+ 164 train 9.039939 (lr=5.7692e-05) (hash(x)=5025297)
175
+ 165 train 8.954381 (lr=5.8042e-05) (hash(x)=5434493)
176
+ 166 train 9.139805 (lr=5.8392e-05) (hash(x)=5197474)
177
+ 167 train 9.037271 (lr=5.8741e-05) (hash(x)=6830813)
178
+ 168 train 9.008745 (lr=5.9091e-05) (hash(x)=5242548)
179
+ 169 train 8.934137 (lr=5.9441e-05) (hash(x)=4752771)
180
+ 170 train 8.994386 (lr=5.9790e-05) (hash(x)=6691249)
181
+ 171 train 8.956799 (lr=6.0140e-05) (hash(x)=6424933)
182
+ 172 train 8.924302 (lr=6.0490e-05) (hash(x)=6960120)
183
+ 173 train 8.824750 (lr=6.0839e-05) (hash(x)=5786517)
184
+ 174 train 8.833849 (lr=6.1189e-05) (hash(x)=7229689)
185
+ 175 train 8.956733 (lr=6.1538e-05) (hash(x)=6680023)
186
+ 176 train 8.773468 (lr=6.1888e-05) (hash(x)=7688393)
187
+ 177 train 8.780998 (lr=6.2238e-05) (hash(x)=6474729)
188
+ 178 train 8.859513 (lr=6.2587e-05) (hash(x)=5946650)
189
+ 179 train 8.776073 (lr=6.2937e-05) (hash(x)=7523279)
190
+ 180 train 8.668302 (lr=6.3287e-05) (hash(x)=5531605)
191
+ 181 train 8.576374 (lr=6.3636e-05) (hash(x)=5508768)
192
+ 182 train 8.736878 (lr=6.3986e-05) (hash(x)=6415982)
193
+ 183 train 8.701837 (lr=6.4336e-05) (hash(x)=6635925)
194
+ 184 train 8.620385 (lr=6.4685e-05) (hash(x)=5923706)
195
+ 185 train 8.682731 (lr=6.5035e-05) (hash(x)=6663801)
196
+ 186 train 8.523819 (lr=6.5385e-05) (hash(x)=6113540)
197
+ 187 train 8.680603 (lr=6.5734e-05) (hash(x)=5869441)
198
+ 188 train 8.679300 (lr=6.6084e-05) (hash(x)=6913454)
199
+ 189 train 8.524346 (lr=6.6434e-05) (hash(x)=6698878)
200
+ 190 train 8.540435 (lr=6.6783e-05) (hash(x)=6094899)
201
+ 191 train 8.540534 (lr=6.7133e-05) (hash(x)=6349004)
202
+ 192 train 8.539870 (lr=6.7483e-05) (hash(x)=6045933)
203
+ 193 train 8.500272 (lr=6.7832e-05) (hash(x)=4986137)
204
+ 194 train 8.281454 (lr=6.8182e-05) (hash(x)=4798229)
205
+ 195 train 8.319677 (lr=6.8531e-05) (hash(x)=6008454)
206
+ 196 train 8.370437 (lr=6.8881e-05) (hash(x)=5938362)
207
+ 197 train 8.305453 (lr=6.9231e-05) (hash(x)=6550770)
208
+ 198 train 8.384578 (lr=6.9580e-05) (hash(x)=7055344)
209
+ 199 train 8.513483 (lr=6.9930e-05) (hash(x)=7682741)
210
+ 200 val loss 8.4100
211
+ 200 val perplexity 4491.9609
212
+ 200 train 8.341786 (lr=7.0280e-05) (hash(x)=7372095)
213
+ 201 train 8.598861 (lr=7.0629e-05) (hash(x)=7892158)
214
+ 202 train 8.341990 (lr=7.0979e-05) (hash(x)=5929740)
215
+ 203 train 8.357166 (lr=7.1329e-05) (hash(x)=7794028)
216
+ 204 train 8.464108 (lr=7.1678e-05) (hash(x)=6839923)
217
+ 205 train 8.203156 (lr=7.2028e-05) (hash(x)=7093523)
218
+ 206 train 8.295508 (lr=7.2378e-05) (hash(x)=5341312)
219
+ 207 train 8.347928 (lr=7.2727e-05) (hash(x)=6166062)
220
+ 208 train 8.284388 (lr=7.3077e-05) (hash(x)=5992164)
221
+ 209 train 8.226460 (lr=7.3427e-05) (hash(x)=6645781)
222
+ 210 train 8.294956 (lr=7.3776e-05) (hash(x)=6207724)
223
+ 211 train 8.174348 (lr=7.4126e-05) (hash(x)=4811192)
224
+ 212 train 8.161320 (lr=7.4476e-05) (hash(x)=6906335)
225
+ 213 train 8.119017 (lr=7.4825e-05) (hash(x)=5970866)
226
+ 214 train 8.136651 (lr=7.5175e-05) (hash(x)=5238015)
227
+ 215 train 8.193901 (lr=7.5524e-05) (hash(x)=7822680)
228
+ 216 train 8.030585 (lr=7.5874e-05) (hash(x)=5478019)
229
+ 217 train 8.130445 (lr=7.6224e-05) (hash(x)=6317317)
230
+ 218 train 8.152289 (lr=7.6573e-05) (hash(x)=5832344)
231
+ 219 train 8.153970 (lr=7.6923e-05) (hash(x)=6431409)
232
+ 220 train 8.155324 (lr=7.7273e-05) (hash(x)=5900130)
233
+ 221 train 8.180564 (lr=7.7622e-05) (hash(x)=6653337)
234
+ 222 train 8.067355 (lr=7.7972e-05) (hash(x)=6696784)
235
+ 223 train 8.116126 (lr=7.8322e-05) (hash(x)=5323032)
236
+ 224 train 7.992850 (lr=7.8671e-05) (hash(x)=5372100)
237
+ 225 train 8.009519 (lr=7.9021e-05) (hash(x)=5125339)
238
+ 226 train 7.958435 (lr=7.9371e-05) (hash(x)=5664463)
239
+ 227 train 7.962347 (lr=7.9720e-05) (hash(x)=7363286)
240
+ 228 train 8.321622 (lr=8.0070e-05) (hash(x)=7444468)
241
+ 229 train 8.161204 (lr=8.0420e-05) (hash(x)=7670050)
242
+ 230 train 8.074509 (lr=8.0769e-05) (hash(x)=8189420)
243
+ 231 train 7.987268 (lr=8.1119e-05) (hash(x)=5616075)
244
+ 232 train 8.002240 (lr=8.1469e-05) (hash(x)=6360924)
245
+ 233 train 8.146235 (lr=8.1818e-05) (hash(x)=6593764)
246
+ 234 train 8.280091 (lr=8.2168e-05) (hash(x)=6764377)
247
+ 235 train 7.952452 (lr=8.2517e-05) (hash(x)=5672608)
248
+ 236 train 7.871394 (lr=8.2867e-05) (hash(x)=6985256)
249
+ 237 train 8.116099 (lr=8.3217e-05) (hash(x)=6828653)
250
+ 238 train 7.915079 (lr=8.3566e-05) (hash(x)=5199845)
251
+ 239 train 8.002543 (lr=8.3916e-05) (hash(x)=6839867)
252
+ 240 train 8.115566 (lr=8.4266e-05) (hash(x)=6928712)
253
+ 241 train 7.993395 (lr=8.4615e-05) (hash(x)=8426992)
254
+ 242 train 7.981725 (lr=8.4965e-05) (hash(x)=5330355)
255
+ 243 train 7.560278 (lr=8.5315e-05) (hash(x)=4916928)
256
+ 244 train 7.947277 (lr=8.5664e-05) (hash(x)=7287928)
257
+ 245 train 7.976170 (lr=8.6014e-05) (hash(x)=6484050)
258
+ 246 train 8.020281 (lr=8.6364e-05) (hash(x)=5062850)
259
+ 247 train 8.033268 (lr=8.6713e-05) (hash(x)=8353379)
260
+ 248 train 7.921080 (lr=8.7063e-05) (hash(x)=6805358)
261
+ 249 train 7.926918 (lr=8.7413e-05) (hash(x)=6552510)
262
+ 250 val loss 7.9394
263
+ 250 val perplexity 2805.7080
264
+ 250 train 7.947418 (lr=8.7762e-05) (hash(x)=5105545)
265
+ 251 train 7.907321 (lr=8.8112e-05) (hash(x)=4838871)
266
+ 252 train 8.009884 (lr=8.8462e-05) (hash(x)=5945286)
267
+ 253 train 7.340507 (lr=8.8811e-05) (hash(x)=3021697)
268
+ 254 train 7.757444 (lr=8.9161e-05) (hash(x)=6443070)
269
+ 255 train 7.844752 (lr=8.9510e-05) (hash(x)=6795665)
270
+ 256 train 7.989934 (lr=8.9860e-05) (hash(x)=7683068)
271
+ 257 train 7.620282 (lr=9.0210e-05) (hash(x)=5486935)
272
+ 258 train 7.923187 (lr=9.0559e-05) (hash(x)=6115232)
273
+ 259 train 8.029565 (lr=9.0909e-05) (hash(x)=6678038)
274
+ 260 train 7.989561 (lr=9.1259e-05) (hash(x)=5747629)
275
+ 261 train 7.993327 (lr=9.1608e-05) (hash(x)=6657714)
276
+ 262 train 7.854912 (lr=9.1958e-05) (hash(x)=6190642)
277
+ 263 train 7.960157 (lr=9.2308e-05) (hash(x)=6112215)
278
+ 264 train 7.854783 (lr=9.2657e-05) (hash(x)=7314989)
279
+ 265 train 7.983354 (lr=9.3007e-05) (hash(x)=8353143)
280
+ 266 train 7.972321 (lr=9.3357e-05) (hash(x)=5912570)
281
+ 267 train 7.889248 (lr=9.3706e-05) (hash(x)=5915361)
282
+ 268 train 7.711596 (lr=9.4056e-05) (hash(x)=6369494)
283
+ 269 train 7.941582 (lr=9.4406e-05) (hash(x)=6254885)
284
+ 270 train 7.836363 (lr=9.4755e-05) (hash(x)=5425120)
285
+ 271 train 7.861965 (lr=9.5105e-05) (hash(x)=5771588)
286
+ 272 train 7.842351 (lr=9.5455e-05) (hash(x)=6763967)
287
+ 273 train 7.791145 (lr=9.5804e-05) (hash(x)=5561507)
288
+ 274 train 7.744189 (lr=9.6154e-05) (hash(x)=7081151)
289
+ 275 train 7.823861 (lr=9.6503e-05) (hash(x)=5688829)
290
+ 276 train 7.779854 (lr=9.6853e-05) (hash(x)=7821554)
291
+ 277 train 7.712911 (lr=9.7203e-05) (hash(x)=5371951)
292
+ 278 train 7.826582 (lr=9.7552e-05) (hash(x)=8120149)
293
+ 279 train 7.947105 (lr=9.7902e-05) (hash(x)=5443305)
294
+ 280 train 7.755939 (lr=9.8252e-05) (hash(x)=7066030)
295
+ 281 train 8.032635 (lr=9.8601e-05) (hash(x)=5016757)
296
+ 282 train 7.887743 (lr=9.8951e-05) (hash(x)=6296163)
297
+ 283 train 7.889939 (lr=9.9301e-05) (hash(x)=6901933)
298
+ 284 train 7.784420 (lr=9.9650e-05) (hash(x)=8056633)
299
+ 285 train 7.900005 (lr=1.0000e-04) (hash(x)=6019085)
300
+ 286 train 7.718867 (lr=1.0000e-04) (hash(x)=7003893)
301
+ 287 train 7.687271 (lr=1.0000e-04) (hash(x)=4294425)
302
+ 288 train 7.707568 (lr=9.9998e-05) (hash(x)=6568540)
303
+ 289 train 7.890723 (lr=9.9996e-05) (hash(x)=6095229)
304
+ 290 train 7.826421 (lr=9.9993e-05) (hash(x)=5512150)
305
+ 291 train 7.442386 (lr=9.9989e-05) (hash(x)=4363941)
306
+ 292 train 7.733924 (lr=9.9984e-05) (hash(x)=6457783)
307
+ 293 train 7.695461 (lr=9.9979e-05) (hash(x)=4693798)
308
+ 294 train 8.200798 (lr=9.9972e-05) (hash(x)=8759806)
309
+ 295 train 8.214126 (lr=9.9965e-05) (hash(x)=5486335)
310
+ 296 train 7.879014 (lr=9.9956e-05) (hash(x)=6964993)
311
+ 297 train 8.163002 (lr=9.9947e-05) (hash(x)=7907450)
312
+ 298 train 7.910578 (lr=9.9937e-05) (hash(x)=5236642)
313
+ 299 train 7.843867 (lr=9.9926e-05) (hash(x)=7295165)
314
+ 300 val loss 7.7771
315
+ 300 val perplexity 2385.2905
316
+ 300 train 7.780232 (lr=9.9915e-05) (hash(x)=4913602)
317
+ 301 train 7.839532 (lr=9.9902e-05) (hash(x)=6814026)
318
+ 302 train 7.930420 (lr=9.9889e-05) (hash(x)=7042051)
319
+ 303 train 7.842582 (lr=9.9874e-05) (hash(x)=6267424)
320
+ 304 train 7.889661 (lr=9.9859e-05) (hash(x)=5614727)
321
+ 305 train 7.856449 (lr=9.9843e-05) (hash(x)=8086437)
322
+ 306 train 7.962467 (lr=9.9826e-05) (hash(x)=5900143)
323
+ 307 train 7.881742 (lr=9.9808e-05) (hash(x)=5270452)
324
+ 308 train 7.833651 (lr=9.9789e-05) (hash(x)=8617707)
325
+ 309 train 7.694805 (lr=9.9770e-05) (hash(x)=6420820)
326
+ 310 train 7.576705 (lr=9.9749e-05) (hash(x)=6215930)
327
+ 311 train 7.764153 (lr=9.9728e-05) (hash(x)=6305297)
328
+ 312 train 7.990970 (lr=9.9706e-05) (hash(x)=4026682)
329
+ 313 train 7.688000 (lr=9.9683e-05) (hash(x)=7889849)
330
+ 314 train 7.930256 (lr=9.9659e-05) (hash(x)=6141721)
331
+ 315 train 8.413348 (lr=9.9634e-05) (hash(x)=6040077)
332
+ 316 train 8.392202 (lr=9.9609e-05) (hash(x)=6155449)
333
+ 317 train 7.945762 (lr=9.9582e-05) (hash(x)=5652497)
334
+ 318 train 7.841300 (lr=9.9555e-05) (hash(x)=6961314)
335
+ 319 train 7.693284 (lr=9.9526e-05) (hash(x)=7130251)
336
+ 320 train 7.557432 (lr=9.9497e-05) (hash(x)=5927461)
337
+ 321 train 7.720663 (lr=9.9467e-05) (hash(x)=6610177)
338
+ 322 train 7.783356 (lr=9.9437e-05) (hash(x)=7277580)
339
+ 323 train 7.602861 (lr=9.9405e-05) (hash(x)=6508350)
340
+ 324 train 7.889494 (lr=9.9372e-05) (hash(x)=6536034)
341
+ 325 train 7.795445 (lr=9.9339e-05) (hash(x)=6994983)
342
+ 326 train 7.807154 (lr=9.9305e-05) (hash(x)=8700721)
343
+ 327 train 7.756165 (lr=9.9270e-05) (hash(x)=5859576)
344
+ 328 train 7.351646 (lr=9.9234e-05) (hash(x)=5264962)
345
+ 329 train 7.362898 (lr=9.9197e-05) (hash(x)=5487065)
346
+ 330 train 7.498593 (lr=9.9159e-05) (hash(x)=5731624)
347
+ 331 train 7.631650 (lr=9.9121e-05) (hash(x)=4892065)
348
+ 332 train 7.528642 (lr=9.9081e-05) (hash(x)=5489496)
349
+ 333 train 7.415709 (lr=9.9041e-05) (hash(x)=4933674)
350
+ 334 train 7.677978 (lr=9.9000e-05) (hash(x)=7021003)
351
+ 335 train 7.832438 (lr=9.8958e-05) (hash(x)=7317289)
352
+ 336 train 7.859572 (lr=9.8915e-05) (hash(x)=7869305)
353
+ 337 train 7.820935 (lr=9.8872e-05) (hash(x)=6811522)
354
+ 338 train 7.731375 (lr=9.8827e-05) (hash(x)=6601423)
355
+ 339 train 7.629858 (lr=9.8782e-05) (hash(x)=5510218)
356
+ 340 train 7.958024 (lr=9.8736e-05) (hash(x)=7180298)
357
+ 341 train 7.718833 (lr=9.8689e-05) (hash(x)=7304235)
358
+ 342 train 7.887617 (lr=9.8641e-05) (hash(x)=8111920)
359
+ 343 train 8.112843 (lr=9.8592e-05) (hash(x)=6752265)
360
+ 344 train 7.617948 (lr=9.8543e-05) (hash(x)=6788720)
361
+ 345 train 7.728952 (lr=9.8492e-05) (hash(x)=5994476)
362
+ 346 train 7.594692 (lr=9.8441e-05) (hash(x)=5862533)
363
+ 347 train 7.700786 (lr=9.8389e-05) (hash(x)=6530867)
364
+ 348 train 7.701354 (lr=9.8336e-05) (hash(x)=5914734)
365
+ 349 train 7.800994 (lr=9.8282e-05) (hash(x)=5746260)
366
+ 350 val loss 7.7044
367
+ 350 val perplexity 2217.9963
368
+ 350 train 7.708432 (lr=9.8228e-05) (hash(x)=6856097)
369
+ 351 train 7.698062 (lr=9.8172e-05) (hash(x)=5790078)
370
+ 352 train 7.829632 (lr=9.8116e-05) (hash(x)=5805002)
371
+ 353 train 7.626110 (lr=9.8059e-05) (hash(x)=6430135)
372
+ 354 train 7.624830 (lr=9.8001e-05) (hash(x)=6156715)
373
+ 355 train 7.945642 (lr=9.7942e-05) (hash(x)=7614023)
374
+ 356 train 7.858819 (lr=9.7882e-05) (hash(x)=4861744)
375
+ 357 train 7.688467 (lr=9.7822e-05) (hash(x)=5244861)
376
+ 358 train 7.739106 (lr=9.7761e-05) (hash(x)=7163697)
377
+ 359 train 7.775827 (lr=9.7699e-05) (hash(x)=5245146)
378
+ 360 train 7.632038 (lr=9.7636e-05) (hash(x)=6153866)
379
+ 361 train 7.681840 (lr=9.7572e-05) (hash(x)=6415565)
380
+ 362 train 7.951396 (lr=9.7507e-05) (hash(x)=7835853)
381
+ 363 train 7.465362 (lr=9.7442e-05) (hash(x)=5786972)
382
+ 364 train 7.806993 (lr=9.7376e-05) (hash(x)=6733560)
383
+ 365 train 7.683011 (lr=9.7309e-05) (hash(x)=6131703)
384
+ 366 train 7.541628 (lr=9.7241e-05) (hash(x)=6001331)
385
+ 367 train 7.989435 (lr=9.7172e-05) (hash(x)=6346459)
386
+ 368 train 7.824390 (lr=9.7103e-05) (hash(x)=8943770)
387
+ 369 train 7.774294 (lr=9.7032e-05) (hash(x)=6504131)
388
+ 370 train 7.724070 (lr=9.6961e-05) (hash(x)=5869339)
389
+ 371 train 7.809279 (lr=9.6889e-05) (hash(x)=5209234)
390
+ 372 train 7.660731 (lr=9.6817e-05) (hash(x)=7678937)
391
+ 373 train 7.710608 (lr=9.6743e-05) (hash(x)=6163272)
392
+ 374 train 7.712218 (lr=9.6669e-05) (hash(x)=5774172)
393
+ 375 train 7.669598 (lr=9.6593e-05) (hash(x)=5676768)
394
+ 376 train 7.349015 (lr=9.6518e-05) (hash(x)=6443895)
395
+ 377 train 8.201034 (lr=9.6441e-05) (hash(x)=6090561)
396
+ 378 train 7.812503 (lr=9.6363e-05) (hash(x)=7788628)
397
+ 379 train 7.969537 (lr=9.6285e-05) (hash(x)=5863092)
398
+ 380 train 7.729538 (lr=9.6206e-05) (hash(x)=7141354)
399
+ 381 train 7.901978 (lr=9.6126e-05) (hash(x)=7201312)
400
+ 382 train 7.867665 (lr=9.6045e-05) (hash(x)=6414278)
401
+ 383 train 7.927705 (lr=9.5963e-05) (hash(x)=7019489)
402
+ 384 train 7.662621 (lr=9.5881e-05) (hash(x)=7914626)
403
+ 385 train 7.661434 (lr=9.5798e-05) (hash(x)=6225608)
404
+ 386 train 7.760068 (lr=9.5714e-05) (hash(x)=5998078)
405
+ 387 train 7.716781 (lr=9.5629e-05) (hash(x)=5483019)
406
+ 388 train 7.565473 (lr=9.5544e-05) (hash(x)=5911642)
407
+ 389 train 7.813452 (lr=9.5457e-05) (hash(x)=5903258)
408
+ 390 train 7.816289 (lr=9.5370e-05) (hash(x)=5632006)
409
+ 391 train 7.662690 (lr=9.5282e-05) (hash(x)=5968716)
410
+ 392 train 7.743196 (lr=9.5194e-05) (hash(x)=6401968)
411
+ 393 train 7.696720 (lr=9.5104e-05) (hash(x)=5778017)
412
+ 394 train 7.284863 (lr=9.5014e-05) (hash(x)=5772567)
413
+ 395 train 7.642454 (lr=9.4923e-05) (hash(x)=7780194)
414
+ 396 train 7.714479 (lr=9.4831e-05) (hash(x)=6424771)
415
+ 397 train 7.598411 (lr=9.4739e-05) (hash(x)=6140998)
416
+ 398 train 7.581070 (lr=9.4646e-05) (hash(x)=7859566)
417
+ 399 train 7.713297 (lr=9.4551e-05) (hash(x)=3784321)
418
+ 400 val loss 7.6699
419
+ 400 val perplexity 2142.7986
420
+ 400 train 7.599418 (lr=9.4457e-05) (hash(x)=5557891)
421
+ 401 train 7.730908 (lr=9.4361e-05) (hash(x)=5860821)
422
+ 402 train 7.726912 (lr=9.4265e-05) (hash(x)=5674826)
423
+ 403 train 7.704773 (lr=9.4168e-05) (hash(x)=7774109)
424
+ 404 train 7.735427 (lr=9.4070e-05) (hash(x)=5466449)
425
+ 405 train 8.021085 (lr=9.3971e-05) (hash(x)=7997664)
426
+ 406 train 7.918145 (lr=9.3872e-05) (hash(x)=7006517)
427
+ 407 train 7.796578 (lr=9.3772e-05) (hash(x)=6576749)
428
+ 408 train 7.672090 (lr=9.3671e-05) (hash(x)=6896198)
429
+ 409 train 7.881176 (lr=9.3569e-05) (hash(x)=7266655)
430
+ 410 train 7.674552 (lr=9.3467e-05) (hash(x)=5696696)
431
+ 411 train 7.788144 (lr=9.3364e-05) (hash(x)=7072404)
432
+ 412 train 7.625633 (lr=9.3260e-05) (hash(x)=6590657)
433
+ 413 train 7.678952 (lr=9.3155e-05) (hash(x)=6626459)
434
+ 414 train 7.733379 (lr=9.3050e-05) (hash(x)=5644977)
435
+ 415 train 7.630622 (lr=9.2944e-05) (hash(x)=6556025)
436
+ 416 train 7.834709 (lr=9.2837e-05) (hash(x)=6407492)
437
+ 417 train 7.597487 (lr=9.2729e-05) (hash(x)=5942129)
438
+ 418 train 7.691179 (lr=9.2621e-05) (hash(x)=6012163)
439
+ 419 train 7.934721 (lr=9.2512e-05) (hash(x)=7325027)
440
+ 420 train 7.861867 (lr=9.2402e-05) (hash(x)=4425520)
441
+ 421 train 7.736204 (lr=9.2292e-05) (hash(x)=7322467)
442
+ 422 train 7.683742 (lr=9.2181e-05) (hash(x)=7482800)
443
+ 423 train 7.769143 (lr=9.2069e-05) (hash(x)=6373412)
444
+ 424 train 7.429379 (lr=9.1956e-05) (hash(x)=6981426)
445
+ 425 train 7.612888 (lr=9.1843e-05) (hash(x)=7332255)
446
+ 426 train 7.804729 (lr=9.1729e-05) (hash(x)=5930894)
447
+ 427 train 7.813303 (lr=9.1614e-05) (hash(x)=5262868)
448
+ 428 train 7.553454 (lr=9.1499e-05) (hash(x)=5585769)
449
+ 429 train 7.464406 (lr=9.1382e-05) (hash(x)=5688247)
450
+ 430 train 7.767093 (lr=9.1266e-05) (hash(x)=7462079)
451
+ 431 train 7.604372 (lr=9.1148e-05) (hash(x)=8055563)
452
+ 432 train 7.691264 (lr=9.1030e-05) (hash(x)=6221701)
453
+ 433 train 7.702159 (lr=9.0911e-05) (hash(x)=5352405)
454
+ 434 train 7.754059 (lr=9.0791e-05) (hash(x)=6730666)
455
+ 435 train 7.766275 (lr=9.0671e-05) (hash(x)=5932790)
456
+ 436 train 7.800768 (lr=9.0550e-05) (hash(x)=6223114)
457
+ 437 train 7.628172 (lr=9.0428e-05) (hash(x)=6249312)
458
+ 438 train 7.776098 (lr=9.0306e-05) (hash(x)=7242827)
459
+ 439 train 7.543546 (lr=9.0182e-05) (hash(x)=5680154)
460
+ 440 train 7.795954 (lr=9.0059e-05) (hash(x)=6332332)
461
+ 441 train 7.731644 (lr=8.9934e-05) (hash(x)=8156280)
462
+ 442 train 7.718644 (lr=8.9809e-05) (hash(x)=6659781)
463
+ 443 train 7.692819 (lr=8.9683e-05) (hash(x)=5914217)
464
+ 444 train 7.604506 (lr=8.9557e-05) (hash(x)=6714110)
465
+ 445 train 7.530135 (lr=8.9430e-05) (hash(x)=7120215)
466
+ 446 train 7.770674 (lr=8.9302e-05) (hash(x)=6593413)
467
+ 447 train 7.710507 (lr=8.9173e-05) (hash(x)=6424128)
468
+ 448 train 7.525712 (lr=8.9044e-05) (hash(x)=5903043)
469
+ 449 train 7.437259 (lr=8.8914e-05) (hash(x)=4390027)
470
+ 450 val loss 7.6538
471
+ 450 val perplexity 2108.5488
472
+ 450 train 7.666108 (lr=8.8784e-05) (hash(x)=6945760)
473
+ 451 train 7.721063 (lr=8.8653e-05) (hash(x)=7804089)
474
+ 452 train 7.779444 (lr=8.8521e-05) (hash(x)=6273417)
475
+ 453 train 7.831733 (lr=8.8388e-05) (hash(x)=7130267)
476
+ 454 train 7.636298 (lr=8.8255e-05) (hash(x)=6140697)
477
+ 455 train 7.566982 (lr=8.8122e-05) (hash(x)=6490149)
478
+ 456 train 7.453558 (lr=8.7987e-05) (hash(x)=4847733)
479
+ 457 train 7.814668 (lr=8.7852e-05) (hash(x)=9276378)
480
+ 458 train 7.584894 (lr=8.7717e-05) (hash(x)=7082053)
481
+ 459 train 7.739535 (lr=8.7580e-05) (hash(x)=6283540)
482
+ 460 train 7.581912 (lr=8.7444e-05) (hash(x)=5200956)
483
+ 461 train 7.716771 (lr=8.7306e-05) (hash(x)=5781680)
484
+ 462 train 7.701544 (lr=8.7168e-05) (hash(x)=8652907)
485
+ 463 train 7.488564 (lr=8.7029e-05) (hash(x)=5728337)
486
+ 464 train 7.597054 (lr=8.6890e-05) (hash(x)=5832401)
487
+ 465 train 7.557499 (lr=8.6750e-05) (hash(x)=6278202)
488
+ 466 train 7.686906 (lr=8.6609e-05) (hash(x)=7403868)
489
+ 467 train 7.689198 (lr=8.6468e-05) (hash(x)=7400272)
490
+ 468 train 7.656137 (lr=8.6326e-05) (hash(x)=5943413)
491
+ 469 train 7.662551 (lr=8.6184e-05) (hash(x)=5795128)
492
+ 470 train 7.689536 (lr=8.6041e-05) (hash(x)=5519980)
493
+ 471 train 7.564383 (lr=8.5897e-05) (hash(x)=6473432)
494
+ 472 train 7.585644 (lr=8.5753e-05) (hash(x)=5484669)
495
+ 473 train 7.686556 (lr=8.5608e-05) (hash(x)=5201502)
496
+ 474 train 7.767343 (lr=8.5462e-05) (hash(x)=7227057)
497
+ 475 train 7.912324 (lr=8.5316e-05) (hash(x)=6922646)
498
+ 476 train 7.503634 (lr=8.5170e-05) (hash(x)=6056673)
499
+ 477 train 7.645999 (lr=8.5022e-05) (hash(x)=5484641)
500
+ 478 train 7.756403 (lr=8.4875e-05) (hash(x)=7100865)
501
+ 479 train 7.513305 (lr=8.4726e-05) (hash(x)=7277141)
502
+ 480 train 7.594075 (lr=8.4577e-05) (hash(x)=6407603)
503
+ 481 train 7.696346 (lr=8.4428e-05) (hash(x)=5604614)
504
+ 482 train 7.704046 (lr=8.4278e-05) (hash(x)=6200612)
505
+ 483 train 7.708209 (lr=8.4127e-05) (hash(x)=7443124)
506
+ 484 train 7.666949 (lr=8.3976e-05) (hash(x)=6728764)
507
+ 485 train 7.768128 (lr=8.3824e-05) (hash(x)=5601968)
508
+ 486 train 7.527875 (lr=8.3672e-05) (hash(x)=6877761)
509
+ 487 train 7.716736 (lr=8.3519e-05) (hash(x)=5632735)
510
+ 488 train 7.713551 (lr=8.3366e-05) (hash(x)=5658845)
511
+ 489 train 7.498044 (lr=8.3212e-05) (hash(x)=5594810)
512
+ 490 train 7.268214 (lr=8.3057e-05) (hash(x)=4738523)
513
+ 491 train 7.632048 (lr=8.2902e-05) (hash(x)=5881015)
514
+ 492 train 7.617334 (lr=8.2746e-05) (hash(x)=5263325)
515
+ 493 train 7.590922 (lr=8.2590e-05) (hash(x)=6865361)
516
+ 494 train 7.765965 (lr=8.2434e-05) (hash(x)=5644619)
517
+ 495 train 7.622214 (lr=8.2276e-05) (hash(x)=5809411)
518
+ 496 train 7.667578 (lr=8.2119e-05) (hash(x)=7044374)
519
+ 497 train 7.711046 (lr=8.1960e-05) (hash(x)=6293025)
520
+ 498 train 7.657429 (lr=8.1801e-05) (hash(x)=5189023)
521
+ 499 train 7.577607 (lr=8.1642e-05) (hash(x)=7232964)
522
+ 500 val loss 7.6342
523
+ 500 val perplexity 2067.7461
524
+ 500 train 7.604221 (lr=8.1482e-05) (hash(x)=5806257)
525
+ 501 train 7.511958 (lr=8.1322e-05) (hash(x)=5646619)
526
+ 502 train 7.674515 (lr=8.1161e-05) (hash(x)=6966594)
527
+ 503 train 7.697623 (lr=8.1000e-05) (hash(x)=6171829)
528
+ 504 train 7.621675 (lr=8.0838e-05) (hash(x)=5945953)
529
+ 505 train 7.853555 (lr=8.0676e-05) (hash(x)=7787552)
530
+ 506 train 7.767939 (lr=8.0513e-05) (hash(x)=7184332)
531
+ 507 train 7.610711 (lr=8.0349e-05) (hash(x)=5884456)
532
+ 508 train 7.731466 (lr=8.0186e-05) (hash(x)=6443281)
533
+ 509 train 7.711513 (lr=8.0021e-05) (hash(x)=6752162)
534
+ 510 train 7.697983 (lr=7.9856e-05) (hash(x)=7140204)
535
+ 511 train 7.533455 (lr=7.9691e-05) (hash(x)=5077031)
536
+ 512 train 7.783298 (lr=7.9525e-05) (hash(x)=6490444)
537
+ 513 train 7.817858 (lr=7.9359e-05) (hash(x)=6883481)
538
+ 514 train 7.665957 (lr=7.9192e-05) (hash(x)=7381497)
539
+ 515 train 7.700274 (lr=7.9025e-05) (hash(x)=6248895)
540
+ 516 train 7.593194 (lr=7.8858e-05) (hash(x)=6655640)
541
+ 517 train 7.509380 (lr=7.8689e-05) (hash(x)=6636615)
542
+ 518 train 7.836347 (lr=7.8521e-05) (hash(x)=5452148)
543
+ 519 train 7.655565 (lr=7.8352e-05) (hash(x)=6918004)
544
+ 520 train 7.647577 (lr=7.8182e-05) (hash(x)=6547295)
545
+ 521 train 7.682449 (lr=7.8012e-05) (hash(x)=7302874)
546
+ 522 train 7.843414 (lr=7.7842e-05) (hash(x)=5637695)
547
+ 523 train 7.696862 (lr=7.7671e-05) (hash(x)=6482466)
548
+ 524 train 7.644479 (lr=7.7500e-05) (hash(x)=7612126)
549
+ 525 train 7.585289 (lr=7.7328e-05) (hash(x)=7298317)
550
+ 526 train 7.735364 (lr=7.7156e-05) (hash(x)=5635777)
551
+ 527 train 7.586355 (lr=7.6984e-05) (hash(x)=6588573)
552
+ 528 train 7.776093 (lr=7.6811e-05) (hash(x)=6634173)
553
+ 529 train 7.548323 (lr=7.6637e-05) (hash(x)=6229630)
554
+ 530 train 7.830211 (lr=7.6463e-05) (hash(x)=6867600)
555
+ 531 train 7.612793 (lr=7.6289e-05) (hash(x)=6136939)
556
+ 532 train 7.463357 (lr=7.6115e-05) (hash(x)=6014704)
557
+ 533 train 7.484867 (lr=7.5940e-05) (hash(x)=6394645)
558
+ 534 train 7.821832 (lr=7.5764e-05) (hash(x)=6548145)
559
+ 535 train 7.612110 (lr=7.5588e-05) (hash(x)=7040157)
560
+ 536 train 7.786055 (lr=7.5412e-05) (hash(x)=5642718)
561
+ 537 train 7.720269 (lr=7.5235e-05) (hash(x)=7041858)
562
+ 538 train 7.673180 (lr=7.5058e-05) (hash(x)=6392891)
563
+ 539 train 7.687346 (lr=7.4881e-05) (hash(x)=6455595)
564
+ 540 train 7.651456 (lr=7.4703e-05) (hash(x)=6245990)
565
+ 541 train 7.542789 (lr=7.4525e-05) (hash(x)=6480443)
566
+ 542 train 7.700452 (lr=7.4346e-05) (hash(x)=7680751)
567
+ 543 train 7.809138 (lr=7.4167e-05) (hash(x)=4122983)
568
+ 544 train 7.525611 (lr=7.3988e-05) (hash(x)=5996900)
569
+ 545 train 7.663051 (lr=7.3808e-05) (hash(x)=7672190)
570
+ 546 train 7.552675 (lr=7.3628e-05) (hash(x)=8149488)
571
+ 547 train 7.692812 (lr=7.3448e-05) (hash(x)=4724052)
572
+ 548 train 7.614067 (lr=7.3267e-05) (hash(x)=6812761)
573
+ 549 train 7.547799 (lr=7.3086e-05) (hash(x)=6654267)
574
+ 550 val loss 7.5978
575
+ 550 val perplexity 1993.7491
576
+ 550 train 7.581007 (lr=7.2904e-05) (hash(x)=6756431)
577
+ 551 train 7.569304 (lr=7.2723e-05) (hash(x)=6820797)
578
+ 552 train 7.670911 (lr=7.2540e-05) (hash(x)=5645711)
579
+ 553 train 7.620242 (lr=7.2358e-05) (hash(x)=6458662)
580
+ 554 train 7.634208 (lr=7.2175e-05) (hash(x)=5300771)
581
+ 555 train 8.420155 (lr=7.1992e-05) (hash(x)=7477594)
582
+ 556 train 7.725153 (lr=7.1808e-05) (hash(x)=5157748)
583
+ 557 train 7.665180 (lr=7.1624e-05) (hash(x)=5876574)
584
+ 558 train 7.438022 (lr=7.1440e-05) (hash(x)=6821258)
585
+ 559 train 7.416778 (lr=7.1256e-05) (hash(x)=6146305)
586
+ 560 train 7.691299 (lr=7.1071e-05) (hash(x)=5278517)
587
+ 561 train 7.515166 (lr=7.0886e-05) (hash(x)=6162080)
588
+ 562 train 7.483430 (lr=7.0701e-05) (hash(x)=5908198)
589
+ 563 train 7.627465 (lr=7.0515e-05) (hash(x)=6814831)
590
+ 564 train 7.675510 (lr=7.0329e-05) (hash(x)=5632357)
591
+ 565 train 7.706075 (lr=7.0143e-05) (hash(x)=7138295)
592
+ 566 train 7.324525 (lr=6.9956e-05) (hash(x)=4678562)
593
+ 567 train 7.589485 (lr=6.9769e-05) (hash(x)=5548298)
594
+ 568 train 7.452751 (lr=6.9582e-05) (hash(x)=6495991)
595
+ 569 train 7.659283 (lr=6.9394e-05) (hash(x)=6265078)
596
+ 570 train 7.857799 (lr=6.9207e-05) (hash(x)=5746797)
597
+ 571 train 8.023754 (lr=6.9019e-05) (hash(x)=7609416)
598
+ 572 train 7.759391 (lr=6.8830e-05) (hash(x)=5593869)
599
+ 573 train 7.514686 (lr=6.8642e-05) (hash(x)=6909367)
600
+ 574 train 7.552631 (lr=6.8453e-05) (hash(x)=4756034)
601
+ 575 train 7.905088 (lr=6.8264e-05) (hash(x)=7574342)
602
+ 576 train 7.347488 (lr=6.8075e-05) (hash(x)=4502769)
603
+ 577 train 7.770018 (lr=6.7885e-05) (hash(x)=8859748)
604
+ 578 train 7.527095 (lr=6.7695e-05) (hash(x)=5683686)
605
+ 579 train 7.508165 (lr=6.7505e-05) (hash(x)=5906937)
606
+ 580 train 7.638266 (lr=6.7315e-05) (hash(x)=6051136)
607
+ 581 train 7.415813 (lr=6.7124e-05) (hash(x)=7243024)
608
+ 582 train 7.450380 (lr=6.6933e-05) (hash(x)=5382761)
609
+ 583 train 7.858344 (lr=6.6742e-05) (hash(x)=8013471)
610
+ 584 train 7.496560 (lr=6.6551e-05) (hash(x)=5165270)
611
+ 585 train 7.624969 (lr=6.6360e-05) (hash(x)=5659461)
612
+ 586 train 7.658638 (lr=6.6168e-05) (hash(x)=5767800)
613
+ 587 train 7.519105 (lr=6.5976e-05) (hash(x)=6146213)
614
+ 588 train 7.859895 (lr=6.5784e-05) (hash(x)=7238407)
615
+ 589 train 7.418145 (lr=6.5592e-05) (hash(x)=6602917)
616
+ 590 train 7.626936 (lr=6.5399e-05) (hash(x)=6287715)
617
+ 591 train 7.515945 (lr=6.5206e-05) (hash(x)=6709902)
618
+ 592 train 7.636279 (lr=6.5013e-05) (hash(x)=6902255)
619
+ 593 train 7.716199 (lr=6.4820e-05) (hash(x)=8154341)
620
+ 594 train 7.952908 (lr=6.4627e-05) (hash(x)=5220688)
621
+ 595 train 7.713269 (lr=6.4433e-05) (hash(x)=6402571)
622
+ 596 train 7.512260 (lr=6.4240e-05) (hash(x)=5895167)
623
+ 597 train 7.653013 (lr=6.4046e-05) (hash(x)=5605466)
624
+ 598 train 7.576892 (lr=6.3852e-05) (hash(x)=6404902)
625
+ 599 train 7.510896 (lr=6.3658e-05) (hash(x)=7683614)
626
+ 600 val loss 7.5828
627
+ 600 val perplexity 1964.1759
628
+ 600 train 7.673580 (lr=6.3463e-05) (hash(x)=6103542)
629
+ 601 train 7.578035 (lr=6.3269e-05) (hash(x)=5421572)
630
+ 602 train 7.477685 (lr=6.3074e-05) (hash(x)=5808144)
631
+ 603 train 7.465957 (lr=6.2879e-05) (hash(x)=5453026)
632
+ 604 train 7.486405 (lr=6.2684e-05) (hash(x)=6337335)
633
+ 605 train 7.598549 (lr=6.2489e-05) (hash(x)=6437746)
634
+ 606 train 7.574255 (lr=6.2294e-05) (hash(x)=8782554)
635
+ 607 train 7.327761 (lr=6.2098e-05) (hash(x)=6260017)
636
+ 608 train 7.508539 (lr=6.1903e-05) (hash(x)=6917394)
637
+ 609 train 7.602421 (lr=6.1707e-05) (hash(x)=6119053)
638
+ 610 train 7.702817 (lr=6.1511e-05) (hash(x)=6089308)
639
+ 611 train 7.738514 (lr=6.1315e-05) (hash(x)=6591466)
640
+ 612 train 7.662880 (lr=6.1119e-05) (hash(x)=5858271)
641
+ 613 train 7.549045 (lr=6.0923e-05) (hash(x)=5652701)
642
+ 614 train 7.612792 (lr=6.0726e-05) (hash(x)=7058229)
643
+ 615 train 7.602983 (lr=6.0530e-05) (hash(x)=5501472)
644
+ 616 train 7.712741 (lr=6.0333e-05) (hash(x)=6232515)
645
+ 617 train 7.902630 (lr=6.0137e-05) (hash(x)=6160695)
646
+ 618 train 7.610404 (lr=5.9940e-05) (hash(x)=6020418)
647
+ 619 train 7.458741 (lr=5.9743e-05) (hash(x)=5871698)
648
+ 620 train 7.486407 (lr=5.9546e-05) (hash(x)=6796897)
649
+ 621 train 7.544331 (lr=5.9349e-05) (hash(x)=7301184)
650
+ 622 train 7.507976 (lr=5.9152e-05) (hash(x)=5748484)
651
+ 623 train 7.388951 (lr=5.8955e-05) (hash(x)=4282052)
652
+ 624 train 7.545971 (lr=5.8758e-05) (hash(x)=4758217)
653
+ 625 train 7.757703 (lr=5.8560e-05) (hash(x)=5679330)
654
+ 626 train 7.435371 (lr=5.8363e-05) (hash(x)=6726349)
655
+ 627 train 7.598531 (lr=5.8165e-05) (hash(x)=7824660)
656
+ 628 train 7.476168 (lr=5.7968e-05) (hash(x)=6255315)
657
+ 629 train 7.613735 (lr=5.7770e-05) (hash(x)=4857318)
658
+ 630 train 7.581805 (lr=5.7573e-05) (hash(x)=6974405)
659
+ 631 train 7.376272 (lr=5.7375e-05) (hash(x)=5954950)
660
+ 632 train 7.628785 (lr=5.7177e-05) (hash(x)=5981982)
661
+ 633 train 7.492799 (lr=5.6979e-05) (hash(x)=5758178)
662
+ 634 train 7.725732 (lr=5.6782e-05) (hash(x)=7816503)
663
+ 635 train 7.436110 (lr=5.6584e-05) (hash(x)=5484656)
664
+ 636 train 7.560784 (lr=5.6386e-05) (hash(x)=5513418)
665
+ 637 train 7.618986 (lr=5.6188e-05) (hash(x)=6818913)
666
+ 638 train 7.848207 (lr=5.5990e-05) (hash(x)=7444913)
667
+ 639 train 7.693310 (lr=5.5792e-05) (hash(x)=7031328)
668
+ 640 train 7.546820 (lr=5.5594e-05) (hash(x)=5202673)
669
+ 641 train 7.702653 (lr=5.5396e-05) (hash(x)=6837621)
670
+ 642 train 7.436080 (lr=5.5198e-05) (hash(x)=6590985)
671
+ 643 train 7.834605 (lr=5.5000e-05) (hash(x)=9157069)
672
+ 644 train 8.068957 (lr=5.4802e-05) (hash(x)=5123556)
673
+ 645 train 7.667952 (lr=5.4604e-05) (hash(x)=6509536)
674
+ 646 train 7.817702 (lr=5.4406e-05) (hash(x)=6252013)
675
+ 647 train 7.330516 (lr=5.4208e-05) (hash(x)=5996883)
676
+ 648 train 7.532339 (lr=5.4010e-05) (hash(x)=5942379)
677
+ 649 train 8.084683 (lr=5.3812e-05) (hash(x)=6850868)
678
+ 650 val loss 7.5635
679
+ 650 val perplexity 1926.4976
680
+ 650 train 7.701354 (lr=5.3614e-05) (hash(x)=8017172)
681
+ 651 train 7.821044 (lr=5.3416e-05) (hash(x)=7645447)
682
+ 652 train 7.689682 (lr=5.3218e-05) (hash(x)=8751332)
683
+ 653 train 7.828826 (lr=5.3021e-05) (hash(x)=6717515)
684
+ 654 train 7.952513 (lr=5.2823e-05) (hash(x)=6521028)
685
+ 655 train 7.809446 (lr=5.2625e-05) (hash(x)=6336164)
686
+ 656 train 7.586422 (lr=5.2427e-05) (hash(x)=6289337)
687
+ 657 train 7.615901 (lr=5.2230e-05) (hash(x)=6897445)
688
+ 658 train 7.732130 (lr=5.2032e-05) (hash(x)=7037289)
689
+ 659 train 7.288346 (lr=5.1835e-05) (hash(x)=5224240)
690
+ 660 train 7.626452 (lr=5.1637e-05) (hash(x)=6834715)
691
+ 661 train 7.698432 (lr=5.1440e-05) (hash(x)=6612945)
692
+ 662 train 8.323870 (lr=5.1242e-05) (hash(x)=8320588)
693
+ 663 train 7.959212 (lr=5.1045e-05) (hash(x)=5447801)
694
+ 664 train 7.603154 (lr=5.0848e-05) (hash(x)=6656376)
695
+ 665 train 7.720012 (lr=5.0651e-05) (hash(x)=7308579)
696
+ 666 train 7.732092 (lr=5.0454e-05) (hash(x)=5344256)
697
+ 667 train 7.317959 (lr=5.0257e-05) (hash(x)=4155598)
698
+ 668 train 7.536976 (lr=5.0060e-05) (hash(x)=5177976)
699
+ 669 train 7.522072 (lr=4.9863e-05) (hash(x)=6106773)
700
+ 670 train 7.505962 (lr=4.9667e-05) (hash(x)=8334761)
701
+ 671 train 7.661094 (lr=4.9470e-05) (hash(x)=7482744)
702
+ 672 train 7.945910 (lr=4.9274e-05) (hash(x)=6435528)
703
+ 673 train 7.354980 (lr=4.9077e-05) (hash(x)=6310687)
704
+ 674 train 7.367802 (lr=4.8881e-05) (hash(x)=6917344)
705
+ 675 train 7.678216 (lr=4.8685e-05) (hash(x)=4733414)
706
+ 676 train 7.439636 (lr=4.8489e-05) (hash(x)=5146340)
707
+ 677 train 7.894372 (lr=4.8293e-05) (hash(x)=6979734)
708
+ 678 train 7.399564 (lr=4.8097e-05) (hash(x)=5847143)
709
+ 679 train 7.486138 (lr=4.7902e-05) (hash(x)=7806284)
710
+ 680 train 7.366781 (lr=4.7706e-05) (hash(x)=6367699)
711
+ 681 train 7.547650 (lr=4.7511e-05) (hash(x)=5080536)
712
+ 682 train 7.731310 (lr=4.7316e-05) (hash(x)=7218763)
713
+ 683 train 7.627647 (lr=4.7121e-05) (hash(x)=7236478)
714
+ 684 train 7.586207 (lr=4.6926e-05) (hash(x)=5497206)
715
+ 685 train 7.659680 (lr=4.6731e-05) (hash(x)=6619318)
716
+ 686 train 7.595905 (lr=4.6537e-05) (hash(x)=5916845)
717
+ 687 train 7.429915 (lr=4.6342e-05) (hash(x)=4656255)
718
+ 688 train 7.439491 (lr=4.6148e-05) (hash(x)=5788889)
719
+ 689 train 7.471800 (lr=4.5954e-05) (hash(x)=5546888)
720
+ 690 train 7.644482 (lr=4.5760e-05) (hash(x)=8072073)
721
+ 691 train 8.184129 (lr=4.5567e-05) (hash(x)=9298325)
722
+ 692 train 7.674432 (lr=4.5373e-05) (hash(x)=5553198)
723
+ 693 train 7.539440 (lr=4.5180e-05) (hash(x)=5994232)
724
+ 694 train 7.578222 (lr=4.4987e-05) (hash(x)=7133014)
725
+ 695 train 7.602300 (lr=4.4794e-05) (hash(x)=6287524)
726
+ 696 train 7.478854 (lr=4.4601e-05) (hash(x)=4968095)
727
+ 697 train 7.919140 (lr=4.4408e-05) (hash(x)=6200049)
728
+ 698 train 7.578990 (lr=4.4216e-05) (hash(x)=6569299)
729
+ 699 train 7.655171 (lr=4.4024e-05) (hash(x)=7895970)
730
+ 700 val loss 7.5421
731
+ 700 val perplexity 1885.8561
732
+ 700 train 7.649476 (lr=4.3832e-05) (hash(x)=5286600)
733
+ 701 train 7.434981 (lr=4.3640e-05) (hash(x)=6051847)
734
+ 702 train 7.474664 (lr=4.3449e-05) (hash(x)=5156899)
735
+ 703 train 7.943992 (lr=4.3258e-05) (hash(x)=8847992)
736
+ 704 train 7.987175 (lr=4.3067e-05) (hash(x)=5343653)
737
+ 705 train 8.064356 (lr=4.2876e-05) (hash(x)=5123883)
738
+ 706 train 7.697535 (lr=4.2685e-05) (hash(x)=6462379)
739
+ 707 train 7.573382 (lr=4.2495e-05) (hash(x)=6338998)
740
+ 708 train 7.735487 (lr=4.2305e-05) (hash(x)=5803128)
741
+ 709 train 7.505132 (lr=4.2115e-05) (hash(x)=5765866)
742
+ 710 train 8.120043 (lr=4.1925e-05) (hash(x)=7297901)
743
+ 711 train 7.863288 (lr=4.1736e-05) (hash(x)=7104011)
744
+ 712 train 7.469209 (lr=4.1547e-05) (hash(x)=6511383)
745
+ 713 train 7.996648 (lr=4.1358e-05) (hash(x)=7438267)
746
+ 714 train 7.453244 (lr=4.1170e-05) (hash(x)=5600336)
747
+ 715 train 7.452913 (lr=4.0981e-05) (hash(x)=5573214)
748
+ 716 train 7.562665 (lr=4.0793e-05) (hash(x)=5751679)
749
+ 717 train 7.951883 (lr=4.0606e-05) (hash(x)=6897283)
750
+ 718 train 7.560116 (lr=4.0418e-05) (hash(x)=6011576)
751
+ 719 train 7.497442 (lr=4.0231e-05) (hash(x)=6882897)
752
+ 720 train 7.610111 (lr=4.0044e-05) (hash(x)=7223798)
753
+ 721 train 7.707412 (lr=3.9857e-05) (hash(x)=5343615)
754
+ 722 train 7.608165 (lr=3.9671e-05) (hash(x)=5500678)
755
+ 723 train 7.487851 (lr=3.9485e-05) (hash(x)=5854242)
756
+ 724 train 7.547425 (lr=3.9299e-05) (hash(x)=6087004)
757
+ 725 train 7.376379 (lr=3.9114e-05) (hash(x)=6317324)
758
+ 726 train 7.385437 (lr=3.8929e-05) (hash(x)=5502743)
759
+ 727 train 7.432744 (lr=3.8744e-05) (hash(x)=6972494)
760
+ 728 train 7.649336 (lr=3.8560e-05) (hash(x)=5138900)
761
+ 729 train 7.351365 (lr=3.8376e-05) (hash(x)=5661167)
762
+ 730 train 7.393908 (lr=3.8192e-05) (hash(x)=5204562)
763
+ 731 train 7.593745 (lr=3.8008e-05) (hash(x)=6796805)
764
+ 732 train 7.319299 (lr=3.7825e-05) (hash(x)=4915685)
765
+ 733 train 7.356728 (lr=3.7642e-05) (hash(x)=5817209)
766
+ 734 train 7.510828 (lr=3.7460e-05) (hash(x)=5270999)
767
+ 735 train 7.280176 (lr=3.7277e-05) (hash(x)=6139120)
768
+ 736 train 7.549130 (lr=3.7096e-05) (hash(x)=6348780)
769
+ 737 train 7.591702 (lr=3.6914e-05) (hash(x)=6121296)
770
+ 738 train 7.489943 (lr=3.6733e-05) (hash(x)=5527309)
771
+ 739 train 7.595429 (lr=3.6552e-05) (hash(x)=6486796)
772
+ 740 train 7.565967 (lr=3.6372e-05) (hash(x)=5859868)
773
+ 741 train 7.372787 (lr=3.6192e-05) (hash(x)=5514645)
774
+ 742 train 7.541854 (lr=3.6012e-05) (hash(x)=5740298)
775
+ 743 train 7.251213 (lr=3.5833e-05) (hash(x)=5578797)
776
+ 744 train 7.407275 (lr=3.5654e-05) (hash(x)=3634442)
777
+ 745 train 7.400749 (lr=3.5475e-05) (hash(x)=7651738)
778
+ 746 train 7.615992 (lr=3.5297e-05) (hash(x)=7118416)
779
+ 747 train 7.662997 (lr=3.5119e-05) (hash(x)=6895326)
780
+ 748 train 7.671333 (lr=3.4942e-05) (hash(x)=8607990)
781
+ 749 train 7.560277 (lr=3.4765e-05) (hash(x)=5490256)
782
+ 750 val loss 7.5285
783
+ 750 val perplexity 1860.3459
784
+ 750 train 7.499443 (lr=3.4588e-05) (hash(x)=6918464)
785
+ 751 train 7.474308 (lr=3.4412e-05) (hash(x)=7783160)
786
+ 752 train 7.494863 (lr=3.4236e-05) (hash(x)=4910007)
787
+ 753 train 7.902472 (lr=3.4060e-05) (hash(x)=5608424)
788
+ 754 train 7.686212 (lr=3.3885e-05) (hash(x)=5176743)
789
+ 755 train 7.244814 (lr=3.3711e-05) (hash(x)=4867625)
790
+ 756 train 7.361460 (lr=3.3537e-05) (hash(x)=4859474)
791
+ 757 train 7.342831 (lr=3.3363e-05) (hash(x)=4655472)
792
+ 758 train 7.362894 (lr=3.3189e-05) (hash(x)=4985126)
793
+ 759 train 7.486481 (lr=3.3016e-05) (hash(x)=6631123)
794
+ 760 train 7.258254 (lr=3.2844e-05) (hash(x)=3652957)
795
+ 761 train 6.904194 (lr=3.2672e-05) (hash(x)=5751990)
796
+ 762 train 7.301682 (lr=3.2500e-05) (hash(x)=6798916)
797
+ 763 train 7.527537 (lr=3.2329e-05) (hash(x)=6710965)
798
+ 764 train 7.436090 (lr=3.2158e-05) (hash(x)=5490166)
799
+ 765 train 7.301960 (lr=3.1988e-05) (hash(x)=5293800)
800
+ 766 train 7.307724 (lr=3.1818e-05) (hash(x)=5884835)
801
+ 767 train 7.490651 (lr=3.1648e-05) (hash(x)=6294657)
802
+ 768 train 7.662206 (lr=3.1479e-05) (hash(x)=7085154)
803
+ 769 train 7.454737 (lr=3.1311e-05) (hash(x)=6999538)
804
+ 770 train 7.626788 (lr=3.1142e-05) (hash(x)=5244581)
805
+ 771 train 7.457083 (lr=3.0975e-05) (hash(x)=5443606)
806
+ 772 train 7.585644 (lr=3.0808e-05) (hash(x)=7028408)
807
+ 773 train 7.454031 (lr=3.0641e-05) (hash(x)=5650401)
808
+ 774 train 7.538034 (lr=3.0475e-05) (hash(x)=6770125)
809
+ 775 train 7.506220 (lr=3.0309e-05) (hash(x)=7035770)
810
+ 776 train 7.360514 (lr=3.0144e-05) (hash(x)=6405982)
811
+ 777 train 7.576938 (lr=2.9979e-05) (hash(x)=6678271)
812
+ 778 train 7.474920 (lr=2.9814e-05) (hash(x)=6088760)
813
+ 779 train 7.878087 (lr=2.9651e-05) (hash(x)=4899541)
814
+ 780 train 7.381652 (lr=2.9487e-05) (hash(x)=5154342)
815
+ 781 train 7.450532 (lr=2.9324e-05) (hash(x)=5359260)
816
+ 782 train 7.276431 (lr=2.9162e-05) (hash(x)=5513509)
817
+ 783 train 7.547409 (lr=2.9000e-05) (hash(x)=6227120)
818
+ 784 train 7.462834 (lr=2.8839e-05) (hash(x)=5165184)
819
+ 785 train 7.525308 (lr=2.8678e-05) (hash(x)=6643916)
820
+ 786 train 7.538626 (lr=2.8518e-05) (hash(x)=7048323)
821
+ 787 train 7.549625 (lr=2.8358e-05) (hash(x)=5981648)
822
+ 788 train 7.409541 (lr=2.8199e-05) (hash(x)=4827872)
823
+ 789 train 7.600919 (lr=2.8040e-05) (hash(x)=6467540)
824
+ 790 train 7.468550 (lr=2.7881e-05) (hash(x)=6542722)
825
+ 791 train 7.412214 (lr=2.7724e-05) (hash(x)=5558993)
826
+ 792 train 7.625873 (lr=2.7566e-05) (hash(x)=5310679)
827
+ 793 train 7.614507 (lr=2.7410e-05) (hash(x)=6698959)
828
+ 794 train 7.320004 (lr=2.7254e-05) (hash(x)=6858783)
829
+ 795 train 7.536436 (lr=2.7098e-05) (hash(x)=7015671)
830
+ 796 train 7.450267 (lr=2.6943e-05) (hash(x)=4483663)
831
+ 797 train 7.415480 (lr=2.6788e-05) (hash(x)=6800890)
832
+ 798 train 7.439596 (lr=2.6634e-05) (hash(x)=6730827)
833
+ 799 train 7.770207 (lr=2.6481e-05) (hash(x)=5931234)
834
+ 800 val loss 7.5067
835
+ 800 val perplexity 1820.1267
836
+ 800 train 7.566363 (lr=2.6328e-05) (hash(x)=4866660)
837
+ 801 train 7.511165 (lr=2.6176e-05) (hash(x)=3995070)
838
+ 802 train 7.588165 (lr=2.6024e-05) (hash(x)=4488617)
839
+ 803 train 7.374712 (lr=2.5873e-05) (hash(x)=4582159)
840
+ 804 train 7.550923 (lr=2.5722e-05) (hash(x)=5164570)
841
+ 805 train 7.630752 (lr=2.5572e-05) (hash(x)=7081906)
842
+ 806 train 7.414999 (lr=2.5423e-05) (hash(x)=5480570)
843
+ 807 train 7.515975 (lr=2.5274e-05) (hash(x)=5029472)
844
+ 808 train 7.503846 (lr=2.5125e-05) (hash(x)=8207798)
845
+ 809 train 7.397047 (lr=2.4978e-05) (hash(x)=3946232)
846
+ 810 train 7.329582 (lr=2.4830e-05) (hash(x)=5652175)
847
+ 811 train 7.418349 (lr=2.4684e-05) (hash(x)=5434102)
848
+ 812 train 8.598052 (lr=2.4538e-05) (hash(x)=6558885)
849
+ 813 train 7.662760 (lr=2.4392e-05) (hash(x)=5340646)
850
+ 814 train 7.212672 (lr=2.4247e-05) (hash(x)=6403613)
851
+ 815 train 7.261681 (lr=2.4103e-05) (hash(x)=6810638)
852
+ 816 train 7.191861 (lr=2.3959e-05) (hash(x)=4382313)
853
+ 817 train 7.828535 (lr=2.3816e-05) (hash(x)=6768026)
854
+ 818 train 7.484043 (lr=2.3674e-05) (hash(x)=6423059)
855
+ 819 train 7.539890 (lr=2.3532e-05) (hash(x)=6521562)
856
+ 820 train 7.492550 (lr=2.3391e-05) (hash(x)=5653405)
857
+ 821 train 7.506424 (lr=2.3250e-05) (hash(x)=4429434)
858
+ 822 train 7.619540 (lr=2.3110e-05) (hash(x)=6885745)
859
+ 823 train 7.512046 (lr=2.2971e-05) (hash(x)=7226932)
860
+ 824 train 7.276736 (lr=2.2832e-05) (hash(x)=5371082)
861
+ 825 train 7.385677 (lr=2.2694e-05) (hash(x)=5065932)
862
+ 826 train 7.425768 (lr=2.2556e-05) (hash(x)=5384078)
863
+ 827 train 7.552860 (lr=2.2420e-05) (hash(x)=6957849)
864
+ 828 train 7.653790 (lr=2.2283e-05) (hash(x)=6049310)
865
+ 829 train 7.328592 (lr=2.2148e-05) (hash(x)=6431457)
866
+ 830 train 7.667130 (lr=2.2013e-05) (hash(x)=6479516)
867
+ 831 train 7.457409 (lr=2.1878e-05) (hash(x)=5334281)
868
+ 832 train 7.389798 (lr=2.1745e-05) (hash(x)=6811166)
869
+ 833 train 7.769713 (lr=2.1612e-05) (hash(x)=5471747)
870
+ 834 train 7.652188 (lr=2.1479e-05) (hash(x)=6733641)
871
+ 835 train 7.665160 (lr=2.1347e-05) (hash(x)=5887983)
872
+ 836 train 7.673751 (lr=2.1216e-05) (hash(x)=5366857)
873
+ 837 train 7.466955 (lr=2.1086e-05) (hash(x)=5075909)
874
+ 838 train 7.351113 (lr=2.0956e-05) (hash(x)=4961569)
875
+ 839 train 7.391370 (lr=2.0827e-05) (hash(x)=6393576)
876
+ 840 train 7.511384 (lr=2.0698e-05) (hash(x)=5520210)
877
+ 841 train 7.461373 (lr=2.0570e-05) (hash(x)=6608979)
878
+ 842 train 7.790101 (lr=2.0443e-05) (hash(x)=6223763)
879
+ 843 train 7.976683 (lr=2.0317e-05) (hash(x)=5057959)
880
+ 844 train 7.811944 (lr=2.0191e-05) (hash(x)=8268189)
881
+ 845 train 7.900192 (lr=2.0066e-05) (hash(x)=6414858)
882
+ 846 train 7.415768 (lr=1.9941e-05) (hash(x)=6468918)
883
+ 847 train 7.793662 (lr=1.9818e-05) (hash(x)=8595271)
884
+ 848 train 7.463602 (lr=1.9694e-05) (hash(x)=5833443)
885
+ 849 train 7.461945 (lr=1.9572e-05) (hash(x)=6500949)
886
+ 850 val loss 7.4960
887
+ 850 val perplexity 1800.8192
888
+ 850 train 7.441362 (lr=1.9450e-05) (hash(x)=5866207)
889
+ 851 train 7.809490 (lr=1.9329e-05) (hash(x)=8395167)
890
+ 852 train 7.822474 (lr=1.9209e-05) (hash(x)=8155803)
891
+ 853 train 7.702494 (lr=1.9089e-05) (hash(x)=6288420)
892
+ 854 train 7.661083 (lr=1.8970e-05) (hash(x)=6770885)
893
+ 855 train 7.479127 (lr=1.8852e-05) (hash(x)=7634417)
894
+ 856 train 7.401660 (lr=1.8734e-05) (hash(x)=6102598)
895
+ 857 train 7.487947 (lr=1.8618e-05) (hash(x)=6795561)
896
+ 858 train 7.428987 (lr=1.8501e-05) (hash(x)=6178545)
897
+ 859 train 7.279370 (lr=1.8386e-05) (hash(x)=5257771)
898
+ 860 train 7.226755 (lr=1.8271e-05) (hash(x)=5794385)
899
+ 861 train 7.770595 (lr=1.8157e-05) (hash(x)=7506860)
900
+ 862 train 7.579564 (lr=1.8044e-05) (hash(x)=4617636)
901
+ 863 train 7.545092 (lr=1.7931e-05) (hash(x)=7674238)
902
+ 864 train 8.023362 (lr=1.7819e-05) (hash(x)=7502317)
903
+ 865 train 7.372228 (lr=1.7708e-05) (hash(x)=5353794)
904
+ 866 train 7.906255 (lr=1.7598e-05) (hash(x)=8474964)
905
+ 867 train 7.324858 (lr=1.7488e-05) (hash(x)=5300555)
906
+ 868 train 7.202689 (lr=1.7379e-05) (hash(x)=6638039)
907
+ 869 train 7.376221 (lr=1.7271e-05) (hash(x)=5564419)
908
+ 870 train 7.725939 (lr=1.7163e-05) (hash(x)=6466311)
909
+ 871 train 7.428892 (lr=1.7056e-05) (hash(x)=6066139)
910
+ 872 train 7.577918 (lr=1.6950e-05) (hash(x)=6790356)
911
+ 873 train 7.466569 (lr=1.6845e-05) (hash(x)=6863709)
912
+ 874 train 7.622839 (lr=1.6740e-05) (hash(x)=6872703)
913
+ 875 train 7.422786 (lr=1.6636e-05) (hash(x)=6688212)
914
+ 876 train 7.627353 (lr=1.6533e-05) (hash(x)=4838260)
915
+ 877 train 7.753117 (lr=1.6431e-05) (hash(x)=7095090)
916
+ 878 train 7.931962 (lr=1.6329e-05) (hash(x)=6775236)
917
+ 879 train 7.579637 (lr=1.6228e-05) (hash(x)=6642404)
918
+ 880 train 7.413118 (lr=1.6128e-05) (hash(x)=6365570)
919
+ 881 train 7.552148 (lr=1.6029e-05) (hash(x)=6539227)
920
+ 882 train 7.508293 (lr=1.5930e-05) (hash(x)=5854007)
921
+ 883 train 7.453975 (lr=1.5832e-05) (hash(x)=4961570)
922
+ 884 train 7.250690 (lr=1.5735e-05) (hash(x)=3961674)
923
+ 885 train 7.567011 (lr=1.5639e-05) (hash(x)=6606391)
924
+ 886 train 7.019990 (lr=1.5543e-05) (hash(x)=6179396)
925
+ 887 train 7.426561 (lr=1.5449e-05) (hash(x)=5105190)
926
+ 888 train 7.890964 (lr=1.5354e-05) (hash(x)=5970936)
927
+ 889 train 7.700416 (lr=1.5261e-05) (hash(x)=5695997)
928
+ 890 train 7.427362 (lr=1.5169e-05) (hash(x)=6171397)
929
+ 891 train 7.587002 (lr=1.5077e-05) (hash(x)=7155991)
930
+ 892 train 7.333276 (lr=1.4986e-05) (hash(x)=6739404)
931
+ 893 train 7.643411 (lr=1.4896e-05) (hash(x)=7249727)
932
+ 894 train 7.569254 (lr=1.4806e-05) (hash(x)=6782303)
933
+ 895 train 7.370490 (lr=1.4718e-05) (hash(x)=5132972)
934
+ 896 train 7.566735 (lr=1.4630e-05) (hash(x)=5962092)
935
+ 897 train 7.343990 (lr=1.4543e-05) (hash(x)=5381802)
936
+ 898 train 7.379108 (lr=1.4456e-05) (hash(x)=6546040)
937
+ 899 train 7.614974 (lr=1.4371e-05) (hash(x)=6269946)
938
+ 900 val loss 7.4845
939
+ 900 val perplexity 1780.1853
940
+ 900 train 7.518080 (lr=1.4286e-05) (hash(x)=5427088)
941
+ 901 train 7.408569 (lr=1.4202e-05) (hash(x)=4976762)
942
+ 902 train 7.536988 (lr=1.4119e-05) (hash(x)=5758632)
943
+ 903 train 7.545311 (lr=1.4037e-05) (hash(x)=5396133)
944
+ 904 train 7.586462 (lr=1.3955e-05) (hash(x)=6805706)
945
+ 905 train 7.417842 (lr=1.3874e-05) (hash(x)=6173473)
946
+ 906 train 7.614098 (lr=1.3794e-05) (hash(x)=6565677)
947
+ 907 train 7.362490 (lr=1.3715e-05) (hash(x)=6394083)
948
+ 908 train 7.521427 (lr=1.3637e-05) (hash(x)=6445589)
949
+ 909 train 7.541846 (lr=1.3559e-05) (hash(x)=6436030)
950
+ 910 train 7.422109 (lr=1.3482e-05) (hash(x)=7220327)
951
+ 911 train 7.407835 (lr=1.3407e-05) (hash(x)=5237804)
952
+ 912 train 7.362480 (lr=1.3331e-05) (hash(x)=4906982)
953
+ 913 train 7.412511 (lr=1.3257e-05) (hash(x)=6540397)
954
+ 914 train 7.627008 (lr=1.3183e-05) (hash(x)=5799609)
955
+ 915 train 7.672608 (lr=1.3111e-05) (hash(x)=5216638)
956
+ 916 train 7.242054 (lr=1.3039e-05) (hash(x)=5646696)
957
+ 917 train 7.679734 (lr=1.2968e-05) (hash(x)=5750537)
958
+ 918 train 7.431813 (lr=1.2897e-05) (hash(x)=6625042)
959
+ 919 train 7.462953 (lr=1.2828e-05) (hash(x)=6735565)
960
+ 920 train 7.414454 (lr=1.2759e-05) (hash(x)=6600526)
961
+ 921 train 7.516166 (lr=1.2691e-05) (hash(x)=7297488)
962
+ 922 train 7.594961 (lr=1.2624e-05) (hash(x)=8039568)
963
+ 923 train 7.804259 (lr=1.2558e-05) (hash(x)=5983575)
964
+ 924 train 7.511978 (lr=1.2493e-05) (hash(x)=5322135)
965
+ 925 train 7.457790 (lr=1.2428e-05) (hash(x)=4808543)
966
+ 926 train 7.530905 (lr=1.2364e-05) (hash(x)=6387352)
967
+ 927 train 7.608785 (lr=1.2301e-05) (hash(x)=5831619)
968
+ 928 train 7.357819 (lr=1.2239e-05) (hash(x)=5418207)
969
+ 929 train 7.342892 (lr=1.2178e-05) (hash(x)=6280052)
970
+ 930 train 7.435105 (lr=1.2118e-05) (hash(x)=5668790)
971
+ 931 train 7.335024 (lr=1.2058e-05) (hash(x)=9109083)
972
+ 932 train 8.112984 (lr=1.1999e-05) (hash(x)=7957194)
973
+ 933 train 7.182177 (lr=1.1941e-05) (hash(x)=5848604)
974
+ 934 train 7.443889 (lr=1.1884e-05) (hash(x)=6466898)
975
+ 935 train 7.651526 (lr=1.1828e-05) (hash(x)=6912967)
976
+ 936 train 7.759449 (lr=1.1772e-05) (hash(x)=6802045)
977
+ 937 train 7.676170 (lr=1.1718e-05) (hash(x)=6615365)
978
+ 938 train 7.849210 (lr=1.1664e-05) (hash(x)=5956271)
979
+ 939 train 7.719796 (lr=1.1611e-05) (hash(x)=5965674)
980
+ 940 train 7.558442 (lr=1.1559e-05) (hash(x)=5925388)
981
+ 941 train 7.711016 (lr=1.1508e-05) (hash(x)=6879965)
982
+ 942 train 7.847147 (lr=1.1457e-05) (hash(x)=6850068)
983
+ 943 train 7.360738 (lr=1.1408e-05) (hash(x)=5822533)
984
+ 944 train 7.387764 (lr=1.1359e-05) (hash(x)=5175909)
985
+ 945 train 7.514094 (lr=1.1311e-05) (hash(x)=5794462)
986
+ 946 train 7.446325 (lr=1.1264e-05) (hash(x)=5451297)
987
+ 947 train 7.654750 (lr=1.1218e-05) (hash(x)=5268216)
988
+ 948 train 7.382711 (lr=1.1173e-05) (hash(x)=5828351)
989
+ 949 train 7.319025 (lr=1.1128e-05) (hash(x)=7058396)
990
+ 950 val loss 7.4760
991
+ 950 val perplexity 1765.1952
992
+ 950 train 7.331544 (lr=1.1085e-05) (hash(x)=6202824)
993
+ 951 train 7.615052 (lr=1.1042e-05) (hash(x)=6531900)
994
+ 952 train 7.483186 (lr=1.1000e-05) (hash(x)=6989972)
995
+ 953 train 7.487875 (lr=1.0959e-05) (hash(x)=5878912)
996
+ 954 train 7.285697 (lr=1.0919e-05) (hash(x)=7862929)
997
+ 955 train 7.305834 (lr=1.0879e-05) (hash(x)=5426499)
998
+ 956 train 7.307633 (lr=1.0841e-05) (hash(x)=5926459)
999
+ 957 train 7.739917 (lr=1.0803e-05) (hash(x)=8107004)
1000
+ 958 train 8.225524 (lr=1.0766e-05) (hash(x)=8057219)
1001
+ 959 train 8.225957 (lr=1.0730e-05) (hash(x)=7455372)
1002
+ 960 train 8.154076 (lr=1.0695e-05) (hash(x)=7083241)
1003
+ 961 train 8.191495 (lr=1.0661e-05) (hash(x)=7727959)
1004
+ 962 train 8.145690 (lr=1.0628e-05) (hash(x)=8069628)
1005
+ 963 train 8.165591 (lr=1.0595e-05) (hash(x)=8319119)
1006
+ 964 train 8.110381 (lr=1.0563e-05) (hash(x)=8080739)
1007
+ 965 train 8.121663 (lr=1.0533e-05) (hash(x)=7581656)
1008
+ 966 train 8.160464 (lr=1.0503e-05) (hash(x)=7644116)
1009
+ 967 train 8.227348 (lr=1.0474e-05) (hash(x)=7592128)
1010
+ 968 train 8.112431 (lr=1.0445e-05) (hash(x)=7555224)
1011
+ 969 train 8.325616 (lr=1.0418e-05) (hash(x)=10506633)
1012
+ 970 train 8.113401 (lr=1.0391e-05) (hash(x)=7171723)
1013
+ 971 train 8.276295 (lr=1.0366e-05) (hash(x)=8704170)
1014
+ 972 train 8.171270 (lr=1.0341e-05) (hash(x)=6236578)
1015
+ 973 train 7.578439 (lr=1.0317e-05) (hash(x)=7461138)
1016
+ 974 train 7.582643 (lr=1.0294e-05) (hash(x)=6802094)
1017
+ 975 train 7.687768 (lr=1.0272e-05) (hash(x)=7165037)
1018
+ 976 train 7.962582 (lr=1.0251e-05) (hash(x)=7619348)
1019
+ 977 train 7.602947 (lr=1.0230e-05) (hash(x)=4976998)
1020
+ 978 train 7.525236 (lr=1.0211e-05) (hash(x)=7270148)
1021
+ 979 train 7.372386 (lr=1.0192e-05) (hash(x)=7678000)
1022
+ 980 train 7.524564 (lr=1.0174e-05) (hash(x)=5853099)
1023
+ 981 train 7.466434 (lr=1.0157e-05) (hash(x)=6299409)
1024
+ 982 train 7.458116 (lr=1.0141e-05) (hash(x)=5422994)
1025
+ 983 train 7.372293 (lr=1.0126e-05) (hash(x)=7414788)
1026
+ 984 train 7.663271 (lr=1.0111e-05) (hash(x)=5094219)
1027
+ 985 train 7.358177 (lr=1.0098e-05) (hash(x)=5063061)
1028
+ 986 train 7.395866 (lr=1.0085e-05) (hash(x)=5964965)
1029
+ 987 train 7.520918 (lr=1.0074e-05) (hash(x)=7326216)
1030
+ 988 train 7.642244 (lr=1.0063e-05) (hash(x)=7677783)
1031
+ 989 train 7.547838 (lr=1.0053e-05) (hash(x)=7492012)
1032
+ 990 train 7.513676 (lr=1.0044e-05) (hash(x)=5898822)
1033
+ 991 train 7.609740 (lr=1.0035e-05) (hash(x)=5857162)
1034
+ 992 train 7.313636 (lr=1.0028e-05) (hash(x)=5675101)
1035
+ 993 train 7.396257 (lr=1.0021e-05) (hash(x)=6169001)
1036
+ 994 train 7.530569 (lr=1.0016e-05) (hash(x)=6691027)
1037
+ 995 train 10.939221 (lr=1.0011e-05) (hash(x)=14715669)
1038
+ 996 train 12.059456 (lr=1.0007e-05) (hash(x)=14724033)
1039
+ 997 train 12.056811 (lr=1.0004e-05) (hash(x)=18256706)
1040
+ 998 train 12.625253 (lr=1.0002e-05) (hash(x)=16949966)
1041
+ 999 val loss 7.4720
1042
+ 999 val perplexity 1758.0585
1043
+ 999 train 11.496600 (lr=1.0000e-05) (hash(x)=16950560)
lr1e-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 sha256:8776e353de6baec901709993a85b5f891aaefb9fb90780baa65e731832cd61fb
3
+ size 36163010
lr1e-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 sha256:8fb232713058bd3148b396999d7b89323f199d5feaa2084d5f03f7a11092b07a
3
+ size 70764358