andrew-healey commited on
Commit
22ecfe3
·
verified ·
1 Parent(s): 00b98b6

Upload folder using huggingface_hub

Browse files
lr8e-4_total_batch_size5120_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr8e-4_total_batch_size5120_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_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 128, "batch_size": null, "total_batch_size": 5120, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0008, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr8e-4_total_batch_size5120_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr8e-4_total_batch_size5120_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 11.683437 (lr=2.7972e-06) (hash(x)=6154740)
5
+ 1 train 11.706587 (lr=5.5944e-06) (hash(x)=6605789)
6
+ 2 train 11.671456 (lr=8.3916e-06) (hash(x)=6696707)
7
+ 3 train 11.689585 (lr=1.1189e-05) (hash(x)=5113283)
8
+ 4 train 11.593265 (lr=1.3986e-05) (hash(x)=5607069)
9
+ 5 train 11.581829 (lr=1.6783e-05) (hash(x)=7007748)
10
+ 6 train 11.563037 (lr=1.9580e-05) (hash(x)=7643299)
11
+ 7 train 11.512960 (lr=2.2378e-05) (hash(x)=5534142)
12
+ 8 train 11.453259 (lr=2.5175e-05) (hash(x)=6539293)
13
+ 9 train 11.405453 (lr=2.7972e-05) (hash(x)=5974274)
14
+ 10 train 11.368162 (lr=3.0769e-05) (hash(x)=7199085)
15
+ 11 train 11.290998 (lr=3.3566e-05) (hash(x)=6763231)
16
+ 12 train 11.141034 (lr=3.6364e-05) (hash(x)=6720398)
17
+ 13 train 11.126776 (lr=3.9161e-05) (hash(x)=6706513)
18
+ 14 train 11.101662 (lr=4.1958e-05) (hash(x)=5727925)
19
+ 15 train 10.959509 (lr=4.4755e-05) (hash(x)=6328012)
20
+ 16 train 10.821723 (lr=4.7552e-05) (hash(x)=5812722)
21
+ 17 train 10.729154 (lr=5.0350e-05) (hash(x)=6183012)
22
+ 18 train 10.736969 (lr=5.3147e-05) (hash(x)=5952539)
23
+ 19 train 10.671859 (lr=5.5944e-05) (hash(x)=7052161)
24
+ 20 train 10.457150 (lr=5.8741e-05) (hash(x)=5810871)
25
+ 21 train 10.596243 (lr=6.1538e-05) (hash(x)=6586466)
26
+ 22 train 10.421978 (lr=6.4336e-05) (hash(x)=5061781)
27
+ 23 train 10.460141 (lr=6.7133e-05) (hash(x)=5556098)
28
+ 24 train 10.522156 (lr=6.9930e-05) (hash(x)=7268535)
29
+ 25 train 10.430445 (lr=7.2727e-05) (hash(x)=8155572)
30
+ 26 train 10.319763 (lr=7.5524e-05) (hash(x)=5019590)
31
+ 27 train 10.214679 (lr=7.8322e-05) (hash(x)=4696943)
32
+ 28 train 10.181250 (lr=8.1119e-05) (hash(x)=5984010)
33
+ 29 train 10.406229 (lr=8.3916e-05) (hash(x)=7649389)
34
+ 30 train 10.103769 (lr=8.6713e-05) (hash(x)=6332817)
35
+ 31 train 10.058916 (lr=8.9510e-05) (hash(x)=6219138)
36
+ 32 train 10.193588 (lr=9.2308e-05) (hash(x)=6895681)
37
+ 33 train 10.142196 (lr=9.5105e-05) (hash(x)=7075794)
38
+ 34 train 10.012372 (lr=9.7902e-05) (hash(x)=5024163)
39
+ 35 train 10.047455 (lr=1.0070e-04) (hash(x)=6049878)
40
+ 36 train 10.135862 (lr=1.0350e-04) (hash(x)=4774943)
41
+ 37 train 10.109038 (lr=1.0629e-04) (hash(x)=6385927)
42
+ 38 train 9.919786 (lr=1.0909e-04) (hash(x)=5070217)
43
+ 39 train 9.870679 (lr=1.1189e-04) (hash(x)=5491972)
44
+ 40 train 9.924762 (lr=1.1469e-04) (hash(x)=5695751)
45
+ 41 train 9.837000 (lr=1.1748e-04) (hash(x)=6061420)
46
+ 42 train 9.767004 (lr=1.2028e-04) (hash(x)=7836710)
47
+ 43 train 9.858080 (lr=1.2308e-04) (hash(x)=8075458)
48
+ 44 train 9.663966 (lr=1.2587e-04) (hash(x)=7190584)
49
+ 45 train 9.761713 (lr=1.2867e-04) (hash(x)=7238769)
50
+ 46 train 9.859268 (lr=1.3147e-04) (hash(x)=4931205)
51
+ 47 train 9.613365 (lr=1.3427e-04) (hash(x)=5917741)
52
+ 48 train 9.546713 (lr=1.3706e-04) (hash(x)=5530785)
53
+ 49 train 9.570498 (lr=1.3986e-04) (hash(x)=5430605)
54
+ 50 val loss 9.5134
55
+ 50 val perplexity 13539.3975
56
+ 50 train 9.537070 (lr=1.4266e-04) (hash(x)=6056774)
57
+ 51 train 9.373644 (lr=1.4545e-04) (hash(x)=5750403)
58
+ 52 train 9.383115 (lr=1.4825e-04) (hash(x)=6187497)
59
+ 53 train 9.337363 (lr=1.5105e-04) (hash(x)=4792640)
60
+ 54 train 9.459874 (lr=1.5385e-04) (hash(x)=6868430)
61
+ 55 train 9.268411 (lr=1.5664e-04) (hash(x)=6112458)
62
+ 56 train 9.198289 (lr=1.5944e-04) (hash(x)=5523601)
63
+ 57 train 9.066706 (lr=1.6224e-04) (hash(x)=5801122)
64
+ 58 train 9.187593 (lr=1.6503e-04) (hash(x)=6826295)
65
+ 59 train 9.329735 (lr=1.6783e-04) (hash(x)=5806490)
66
+ 60 train 9.102900 (lr=1.7063e-04) (hash(x)=6223100)
67
+ 61 train 8.975108 (lr=1.7343e-04) (hash(x)=5754140)
68
+ 62 train 8.981058 (lr=1.7622e-04) (hash(x)=6675367)
69
+ 63 train 8.861685 (lr=1.7902e-04) (hash(x)=6568379)
70
+ 64 train 8.709698 (lr=1.8182e-04) (hash(x)=5648233)
71
+ 65 train 8.934361 (lr=1.8462e-04) (hash(x)=7134932)
72
+ 66 train 8.788771 (lr=1.8741e-04) (hash(x)=5242536)
73
+ 67 train 8.860897 (lr=1.9021e-04) (hash(x)=7212403)
74
+ 68 train 8.672624 (lr=1.9301e-04) (hash(x)=5342967)
75
+ 69 train 8.761984 (lr=1.9580e-04) (hash(x)=7720703)
76
+ 70 train 8.777204 (lr=1.9860e-04) (hash(x)=7351337)
77
+ 71 train 8.735043 (lr=2.0140e-04) (hash(x)=7386082)
78
+ 72 train 8.539919 (lr=2.0420e-04) (hash(x)=6931318)
79
+ 73 train 8.566191 (lr=2.0699e-04) (hash(x)=6848513)
80
+ 74 train 8.536734 (lr=2.0979e-04) (hash(x)=7714274)
81
+ 75 train 8.625558 (lr=2.1259e-04) (hash(x)=6359743)
82
+ 76 train 8.429007 (lr=2.1538e-04) (hash(x)=6801650)
83
+ 77 train 8.423543 (lr=2.1818e-04) (hash(x)=7536538)
84
+ 78 train 8.477844 (lr=2.2098e-04) (hash(x)=6957410)
85
+ 79 train 8.486066 (lr=2.2378e-04) (hash(x)=6151674)
86
+ 80 train 8.371601 (lr=2.2657e-04) (hash(x)=6033844)
87
+ 81 train 8.187799 (lr=2.2937e-04) (hash(x)=6317099)
88
+ 82 train 8.147289 (lr=2.3217e-04) (hash(x)=5357844)
89
+ 83 train 8.262385 (lr=2.3497e-04) (hash(x)=7168704)
90
+ 84 train 8.527387 (lr=2.3776e-04) (hash(x)=7157938)
91
+ 85 train 8.209927 (lr=2.4056e-04) (hash(x)=6736092)
92
+ 86 train 8.128933 (lr=2.4336e-04) (hash(x)=5383012)
93
+ 87 train 8.252179 (lr=2.4615e-04) (hash(x)=7859060)
94
+ 88 train 8.221327 (lr=2.4895e-04) (hash(x)=5869288)
95
+ 89 train 8.190435 (lr=2.5175e-04) (hash(x)=6958223)
96
+ 90 train 8.063054 (lr=2.5455e-04) (hash(x)=6731344)
97
+ 91 train 8.134937 (lr=2.5734e-04) (hash(x)=6976239)
98
+ 92 train 7.988826 (lr=2.6014e-04) (hash(x)=7769441)
99
+ 93 train 8.101367 (lr=2.6294e-04) (hash(x)=6956845)
100
+ 94 train 8.204744 (lr=2.6573e-04) (hash(x)=6638037)
101
+ 95 train 7.841677 (lr=2.6853e-04) (hash(x)=6391506)
102
+ 96 train 8.242569 (lr=2.7133e-04) (hash(x)=7319928)
103
+ 97 train 8.173428 (lr=2.7413e-04) (hash(x)=6113701)
104
+ 98 train 7.852560 (lr=2.7692e-04) (hash(x)=6246453)
105
+ 99 train 7.953240 (lr=2.7972e-04) (hash(x)=5245999)
106
+ 100 val loss 8.0175
107
+ 100 val perplexity 3033.5171
108
+ 100 train 8.085306 (lr=2.8252e-04) (hash(x)=6227256)
109
+ 101 train 7.943653 (lr=2.8531e-04) (hash(x)=5774189)
110
+ 102 train 7.873219 (lr=2.8811e-04) (hash(x)=5588220)
111
+ 103 train 7.866357 (lr=2.9091e-04) (hash(x)=5985675)
112
+ 104 train 7.791639 (lr=2.9371e-04) (hash(x)=6259214)
113
+ 105 train 7.990948 (lr=2.9650e-04) (hash(x)=6451776)
114
+ 106 train 7.805307 (lr=2.9930e-04) (hash(x)=4940406)
115
+ 107 train 7.799399 (lr=3.0210e-04) (hash(x)=4355733)
116
+ 108 train 7.798152 (lr=3.0490e-04) (hash(x)=6186065)
117
+ 109 train 7.951399 (lr=3.0769e-04) (hash(x)=6669188)
118
+ 110 train 9.139694 (lr=3.1049e-04) (hash(x)=7118997)
119
+ 111 train 8.290688 (lr=3.1329e-04) (hash(x)=6984772)
120
+ 112 train 7.793800 (lr=3.1608e-04) (hash(x)=6472877)
121
+ 113 train 8.016268 (lr=3.1888e-04) (hash(x)=7440580)
122
+ 114 train 7.975979 (lr=3.2168e-04) (hash(x)=6378428)
123
+ 115 train 8.094499 (lr=3.2448e-04) (hash(x)=6244096)
124
+ 116 train 8.125318 (lr=3.2727e-04) (hash(x)=9100578)
125
+ 117 train 8.328566 (lr=3.3007e-04) (hash(x)=8381560)
126
+ 118 train 8.129729 (lr=3.3287e-04) (hash(x)=6066072)
127
+ 119 train 7.806450 (lr=3.3566e-04) (hash(x)=6603717)
128
+ 120 train 7.665348 (lr=3.3846e-04) (hash(x)=5138438)
129
+ 121 train 7.591730 (lr=3.4126e-04) (hash(x)=6437516)
130
+ 122 train 7.834575 (lr=3.4406e-04) (hash(x)=5743807)
131
+ 123 train 7.636527 (lr=3.4685e-04) (hash(x)=5669522)
132
+ 124 train 7.941388 (lr=3.4965e-04) (hash(x)=5963623)
133
+ 125 train 7.651243 (lr=3.5245e-04) (hash(x)=5505889)
134
+ 126 train 8.112898 (lr=3.5524e-04) (hash(x)=7607512)
135
+ 127 train 7.928976 (lr=3.5804e-04) (hash(x)=6190579)
136
+ 128 train 8.152842 (lr=3.6084e-04) (hash(x)=7444834)
137
+ 129 train 7.983224 (lr=3.6364e-04) (hash(x)=5153196)
138
+ 130 train 7.893080 (lr=3.6643e-04) (hash(x)=6495326)
139
+ 131 train 7.972165 (lr=3.6923e-04) (hash(x)=4912983)
140
+ 132 train 8.351508 (lr=3.7203e-04) (hash(x)=7094046)
141
+ 133 train 7.797325 (lr=3.7483e-04) (hash(x)=5109125)
142
+ 134 train 7.851958 (lr=3.7762e-04) (hash(x)=7119136)
143
+ 135 train 7.877728 (lr=3.8042e-04) (hash(x)=7276303)
144
+ 136 train 8.053335 (lr=3.8322e-04) (hash(x)=9147123)
145
+ 137 train 8.058704 (lr=3.8601e-04) (hash(x)=6645716)
146
+ 138 train 8.016753 (lr=3.8881e-04) (hash(x)=6347875)
147
+ 139 train 7.787608 (lr=3.9161e-04) (hash(x)=5479947)
148
+ 140 train 7.887794 (lr=3.9441e-04) (hash(x)=5832595)
149
+ 141 train 8.230128 (lr=3.9720e-04) (hash(x)=7850360)
150
+ 142 train 8.606165 (lr=4.0000e-04) (hash(x)=8499683)
151
+ 143 train 8.142312 (lr=4.0280e-04) (hash(x)=6630378)
152
+ 144 train 8.066635 (lr=4.0559e-04) (hash(x)=5340846)
153
+ 145 train 7.885839 (lr=4.0839e-04) (hash(x)=6295358)
154
+ 146 train 7.905224 (lr=4.1119e-04) (hash(x)=5974729)
155
+ 147 train 8.480494 (lr=4.1399e-04) (hash(x)=8097010)
156
+ 148 train 8.192301 (lr=4.1678e-04) (hash(x)=6681250)
157
+ 149 train 7.593072 (lr=4.1958e-04) (hash(x)=3958188)
158
+ 150 val loss 7.8516
159
+ 150 val perplexity 2569.7183
160
+ 150 train 7.797843 (lr=4.2238e-04) (hash(x)=5975539)
161
+ 151 train 9.771262 (lr=4.2517e-04) (hash(x)=9989744)
162
+ 152 train 8.807684 (lr=4.2797e-04) (hash(x)=6748487)
163
+ 153 train 7.780153 (lr=4.3077e-04) (hash(x)=6343832)
164
+ 154 train 7.872688 (lr=4.3357e-04) (hash(x)=5688949)
165
+ 155 train 8.181031 (lr=4.3636e-04) (hash(x)=6694932)
166
+ 156 train 7.850771 (lr=4.3916e-04) (hash(x)=5872835)
167
+ 157 train 7.880905 (lr=4.4196e-04) (hash(x)=6429137)
168
+ 158 train 8.196158 (lr=4.4476e-04) (hash(x)=6999390)
169
+ 159 train 9.718822 (lr=4.4755e-04) (hash(x)=6317894)
170
+ 160 train 11.017572 (lr=4.5035e-04) (hash(x)=5813657)
171
+ 161 train 10.886911 (lr=4.5315e-04) (hash(x)=5165595)
172
+ 162 train 10.911617 (lr=4.5594e-04) (hash(x)=5682633)
173
+ 163 train 8.788431 (lr=4.5874e-04) (hash(x)=5046327)
174
+ 164 train 8.188199 (lr=4.6154e-04) (hash(x)=6195808)
175
+ 165 train 8.072556 (lr=4.6434e-04) (hash(x)=7060363)
176
+ 166 train 8.033092 (lr=4.6713e-04) (hash(x)=5837126)
177
+ 167 train 7.952327 (lr=4.6993e-04) (hash(x)=7947703)
178
+ 168 train 7.552557 (lr=4.7273e-04) (hash(x)=5277865)
179
+ 169 train 7.848337 (lr=4.7552e-04) (hash(x)=5971240)
180
+ 170 train 7.774890 (lr=4.7832e-04) (hash(x)=6590408)
181
+ 171 train 7.977017 (lr=4.8112e-04) (hash(x)=5916068)
182
+ 172 train 7.872134 (lr=4.8392e-04) (hash(x)=6069727)
183
+ 173 train 7.882347 (lr=4.8671e-04) (hash(x)=7426277)
184
+ 174 train 8.041883 (lr=4.8951e-04) (hash(x)=5356513)
185
+ 175 train 7.822970 (lr=4.9231e-04) (hash(x)=5777498)
186
+ 176 train 8.035238 (lr=4.9510e-04) (hash(x)=5989756)
187
+ 177 train 7.829080 (lr=4.9790e-04) (hash(x)=6403331)
188
+ 178 train 7.892743 (lr=5.0070e-04) (hash(x)=7650667)
189
+ 179 train 8.062049 (lr=5.0350e-04) (hash(x)=5753092)
190
+ 180 train 7.838658 (lr=5.0629e-04) (hash(x)=5751457)
191
+ 181 train 7.681620 (lr=5.0909e-04) (hash(x)=5981486)
192
+ 182 train 7.679676 (lr=5.1189e-04) (hash(x)=5918229)
193
+ 183 train 7.911967 (lr=5.1469e-04) (hash(x)=7723226)
194
+ 184 train 7.988165 (lr=5.1748e-04) (hash(x)=6111721)
195
+ 185 train 7.644462 (lr=5.2028e-04) (hash(x)=6069558)
196
+ 186 train 7.932916 (lr=5.2308e-04) (hash(x)=5626336)
197
+ 187 train 8.131742 (lr=5.2587e-04) (hash(x)=7007016)
198
+ 188 train 7.666480 (lr=5.2867e-04) (hash(x)=5883013)
199
+ 189 train 7.528714 (lr=5.3147e-04) (hash(x)=5027958)
200
+ 190 train 7.863471 (lr=5.3427e-04) (hash(x)=5474948)
201
+ 191 train 7.719158 (lr=5.3706e-04) (hash(x)=6491229)
202
+ 192 train 7.790199 (lr=5.3986e-04) (hash(x)=5518341)
203
+ 193 train 7.872255 (lr=5.4266e-04) (hash(x)=7036653)
204
+ 194 train 7.868250 (lr=5.4545e-04) (hash(x)=5527616)
205
+ 195 train 8.214939 (lr=5.4825e-04) (hash(x)=6390752)
206
+ 196 train 7.849031 (lr=5.5105e-04) (hash(x)=6032557)
207
+ 197 train 8.587428 (lr=5.5385e-04) (hash(x)=7758167)
208
+ 198 train 8.413548 (lr=5.5664e-04) (hash(x)=7968070)
209
+ 199 train 8.000937 (lr=5.5944e-04) (hash(x)=6019389)
210
+ 200 val loss 7.8950
211
+ 200 val perplexity 2683.7144
212
+ 200 train 7.971140 (lr=5.6224e-04) (hash(x)=7276744)
213
+ 201 train 8.008094 (lr=5.6503e-04) (hash(x)=6503191)
214
+ 202 train 7.825707 (lr=5.6783e-04) (hash(x)=6006880)
215
+ 203 train 8.021129 (lr=5.7063e-04) (hash(x)=7662067)
216
+ 204 train 7.789588 (lr=5.7343e-04) (hash(x)=6297345)
217
+ 205 train 8.082823 (lr=5.7622e-04) (hash(x)=7901992)
218
+ 206 train 8.142193 (lr=5.7902e-04) (hash(x)=6579950)
219
+ 207 train 7.692109 (lr=5.8182e-04) (hash(x)=4648609)
220
+ 208 train 7.909349 (lr=5.8462e-04) (hash(x)=6903216)
221
+ 209 train 7.732367 (lr=5.8741e-04) (hash(x)=5897288)
222
+ 210 train 7.839914 (lr=5.9021e-04) (hash(x)=7300160)
223
+ 211 train 7.723544 (lr=5.9301e-04) (hash(x)=4725966)
224
+ 212 train 7.557694 (lr=5.9580e-04) (hash(x)=5060060)
225
+ 213 train 8.033115 (lr=5.9860e-04) (hash(x)=6243442)
226
+ 214 train 7.712458 (lr=6.0140e-04) (hash(x)=5893816)
227
+ 215 train 7.746438 (lr=6.0420e-04) (hash(x)=5558355)
228
+ 216 train 7.865259 (lr=6.0699e-04) (hash(x)=4378747)
229
+ 217 train 7.716820 (lr=6.0979e-04) (hash(x)=5400407)
230
+ 218 train 8.024069 (lr=6.1259e-04) (hash(x)=6900554)
231
+ 219 train 7.917788 (lr=6.1538e-04) (hash(x)=6524933)
232
+ 220 train 7.812368 (lr=6.1818e-04) (hash(x)=5409944)
233
+ 221 train 7.895364 (lr=6.2098e-04) (hash(x)=5889724)
234
+ 222 train 7.826368 (lr=6.2378e-04) (hash(x)=4970496)
235
+ 223 train 7.626686 (lr=6.2657e-04) (hash(x)=6369326)
236
+ 224 train 7.733120 (lr=6.2937e-04) (hash(x)=6563975)
237
+ 225 train 7.897446 (lr=6.3217e-04) (hash(x)=5911906)
238
+ 226 train 7.953748 (lr=6.3497e-04) (hash(x)=6111462)
239
+ 227 train 7.958645 (lr=6.3776e-04) (hash(x)=6022625)
240
+ 228 train 7.768284 (lr=6.4056e-04) (hash(x)=5884663)
241
+ 229 train 7.890339 (lr=6.4336e-04) (hash(x)=6584810)
242
+ 230 train 7.842656 (lr=6.4615e-04) (hash(x)=5289998)
243
+ 231 train 7.514932 (lr=6.4895e-04) (hash(x)=4906853)
244
+ 232 train 7.615761 (lr=6.5175e-04) (hash(x)=5862071)
245
+ 233 train 7.258059 (lr=6.5455e-04) (hash(x)=4564564)
246
+ 234 train 7.601528 (lr=6.5734e-04) (hash(x)=5723971)
247
+ 235 train 7.698314 (lr=6.6014e-04) (hash(x)=5589269)
248
+ 236 train 7.696378 (lr=6.6294e-04) (hash(x)=5683368)
249
+ 237 train 7.960526 (lr=6.6573e-04) (hash(x)=6969180)
250
+ 238 train 7.832767 (lr=6.6853e-04) (hash(x)=5341421)
251
+ 239 train 7.587717 (lr=6.7133e-04) (hash(x)=6437376)
252
+ 240 train 8.080617 (lr=6.7413e-04) (hash(x)=7527007)
253
+ 241 train 7.935441 (lr=6.7692e-04) (hash(x)=7116258)
254
+ 242 train 7.905227 (lr=6.7972e-04) (hash(x)=6463930)
255
+ 243 train 7.925653 (lr=6.8252e-04) (hash(x)=7304892)
256
+ 244 train 8.080555 (lr=6.8531e-04) (hash(x)=5881808)
257
+ 245 train 7.851917 (lr=6.8811e-04) (hash(x)=7366721)
258
+ 246 train 7.762708 (lr=6.9091e-04) (hash(x)=6457521)
259
+ 247 train 8.155014 (lr=6.9371e-04) (hash(x)=7084093)
260
+ 248 train 7.817368 (lr=6.9650e-04) (hash(x)=6015300)
261
+ 249 train 8.036617 (lr=6.9930e-04) (hash(x)=7197652)
262
+ 250 val loss 7.7806
263
+ 250 val perplexity 2393.6228
264
+ 250 train 8.101478 (lr=7.0210e-04) (hash(x)=4830828)
265
+ 251 train 7.793128 (lr=7.0490e-04) (hash(x)=5908178)
266
+ 252 train 7.799882 (lr=7.0769e-04) (hash(x)=6299092)
267
+ 253 train 7.793116 (lr=7.1049e-04) (hash(x)=6349974)
268
+ 254 train 7.800604 (lr=7.1329e-04) (hash(x)=5641494)
269
+ 255 train 7.787890 (lr=7.1608e-04) (hash(x)=7048804)
270
+ 256 train 8.008862 (lr=7.1888e-04) (hash(x)=7035244)
271
+ 257 train 7.693189 (lr=7.2168e-04) (hash(x)=7909039)
272
+ 258 train 7.960866 (lr=7.2448e-04) (hash(x)=5470939)
273
+ 259 train 7.780613 (lr=7.2727e-04) (hash(x)=6085549)
274
+ 260 train 8.054434 (lr=7.3007e-04) (hash(x)=5882649)
275
+ 261 train 7.851082 (lr=7.3287e-04) (hash(x)=7463181)
276
+ 262 train 8.190361 (lr=7.3566e-04) (hash(x)=5215574)
277
+ 263 train 7.644195 (lr=7.3846e-04) (hash(x)=5752594)
278
+ 264 train 7.811648 (lr=7.4126e-04) (hash(x)=6416619)
279
+ 265 train 7.815011 (lr=7.4406e-04) (hash(x)=6114601)
280
+ 266 train 7.758576 (lr=7.4685e-04) (hash(x)=5646171)
281
+ 267 train 7.679123 (lr=7.4965e-04) (hash(x)=7662769)
282
+ 268 train 7.937560 (lr=7.5245e-04) (hash(x)=6394376)
283
+ 269 train 7.557564 (lr=7.5524e-04) (hash(x)=7485666)
284
+ 270 train 7.645489 (lr=7.5804e-04) (hash(x)=6636823)
285
+ 271 train 7.536088 (lr=7.6084e-04) (hash(x)=6393520)
286
+ 272 train 7.761819 (lr=7.6364e-04) (hash(x)=5235659)
287
+ 273 train 7.819428 (lr=7.6643e-04) (hash(x)=5586291)
288
+ 274 train 8.040734 (lr=7.6923e-04) (hash(x)=7034674)
289
+ 275 train 8.158289 (lr=7.7203e-04) (hash(x)=5942867)
290
+ 276 train 7.840230 (lr=7.7483e-04) (hash(x)=7812344)
291
+ 277 train 8.092177 (lr=7.7762e-04) (hash(x)=7574027)
292
+ 278 train 8.046166 (lr=7.8042e-04) (hash(x)=5960216)
293
+ 279 train 7.835858 (lr=7.8322e-04) (hash(x)=6793550)
294
+ 280 train 7.810153 (lr=7.8601e-04) (hash(x)=6879965)
295
+ 281 train 7.487259 (lr=7.8881e-04) (hash(x)=6264789)
296
+ 282 train 7.673707 (lr=7.9161e-04) (hash(x)=5175356)
297
+ 283 train 7.897304 (lr=7.9441e-04) (hash(x)=7105976)
298
+ 284 train 7.885385 (lr=7.9720e-04) (hash(x)=7286386)
299
+ 285 train 8.053342 (lr=8.0000e-04) (hash(x)=7244279)
300
+ 286 train 7.616718 (lr=8.0000e-04) (hash(x)=4877247)
301
+ 287 train 7.571765 (lr=8.0000e-04) (hash(x)=6581348)
302
+ 288 train 7.580690 (lr=7.9999e-04) (hash(x)=6397208)
303
+ 289 train 7.649809 (lr=7.9997e-04) (hash(x)=6543825)
304
+ 290 train 7.424378 (lr=7.9994e-04) (hash(x)=6088993)
305
+ 291 train 7.781647 (lr=7.9991e-04) (hash(x)=5555598)
306
+ 292 train 7.850276 (lr=7.9987e-04) (hash(x)=6924240)
307
+ 293 train 7.990934 (lr=7.9983e-04) (hash(x)=7254232)
308
+ 294 train 7.537994 (lr=7.9978e-04) (hash(x)=7689743)
309
+ 295 train 7.539412 (lr=7.9972e-04) (hash(x)=6235837)
310
+ 296 train 7.812680 (lr=7.9965e-04) (hash(x)=5216883)
311
+ 297 train 7.444831 (lr=7.9958e-04) (hash(x)=4738550)
312
+ 298 train 7.586594 (lr=7.9950e-04) (hash(x)=5126918)
313
+ 299 train 7.822454 (lr=7.9941e-04) (hash(x)=5591770)
314
+ 300 val loss 7.7078
315
+ 300 val perplexity 2225.5979
316
+ 300 train 8.687640 (lr=7.9932e-04) (hash(x)=8001573)
317
+ 301 train 8.380744 (lr=7.9922e-04) (hash(x)=8012491)
318
+ 302 train 8.011705 (lr=7.9911e-04) (hash(x)=3997428)
319
+ 303 train 7.326485 (lr=7.9899e-04) (hash(x)=7083021)
320
+ 304 train 7.927269 (lr=7.9887e-04) (hash(x)=7177371)
321
+ 305 train 7.746573 (lr=7.9874e-04) (hash(x)=6456022)
322
+ 306 train 7.842574 (lr=7.9861e-04) (hash(x)=5238861)
323
+ 307 train 7.776828 (lr=7.9846e-04) (hash(x)=7281348)
324
+ 308 train 7.940588 (lr=7.9831e-04) (hash(x)=6848441)
325
+ 309 train 7.990346 (lr=7.9816e-04) (hash(x)=7824794)
326
+ 310 train 8.146921 (lr=7.9799e-04) (hash(x)=7440208)
327
+ 311 train 7.916420 (lr=7.9782e-04) (hash(x)=6155821)
328
+ 312 train 7.647549 (lr=7.9765e-04) (hash(x)=5350461)
329
+ 313 train 7.856035 (lr=7.9746e-04) (hash(x)=7807564)
330
+ 314 train 7.721174 (lr=7.9727e-04) (hash(x)=5992165)
331
+ 315 train 7.669768 (lr=7.9707e-04) (hash(x)=5736241)
332
+ 316 train 7.646352 (lr=7.9687e-04) (hash(x)=6536253)
333
+ 317 train 7.848963 (lr=7.9666e-04) (hash(x)=5688415)
334
+ 318 train 7.497478 (lr=7.9644e-04) (hash(x)=5839705)
335
+ 319 train 7.707751 (lr=7.9621e-04) (hash(x)=5657123)
336
+ 320 train 7.677433 (lr=7.9598e-04) (hash(x)=6354364)
337
+ 321 train 7.724725 (lr=7.9574e-04) (hash(x)=6102985)
338
+ 322 train 7.975651 (lr=7.9549e-04) (hash(x)=7161967)
339
+ 323 train 7.857307 (lr=7.9524e-04) (hash(x)=6452095)
340
+ 324 train 7.586239 (lr=7.9498e-04) (hash(x)=5547396)
341
+ 325 train 7.915174 (lr=7.9471e-04) (hash(x)=5848921)
342
+ 326 train 7.743857 (lr=7.9444e-04) (hash(x)=4890294)
343
+ 327 train 7.458940 (lr=7.9416e-04) (hash(x)=5312267)
344
+ 328 train 7.456458 (lr=7.9387e-04) (hash(x)=5348929)
345
+ 329 train 7.372284 (lr=7.9358e-04) (hash(x)=5025297)
346
+ 330 train 7.381709 (lr=7.9327e-04) (hash(x)=5670052)
347
+ 331 train 7.252681 (lr=7.9297e-04) (hash(x)=5434493)
348
+ 332 train 8.091735 (lr=7.9265e-04) (hash(x)=9122453)
349
+ 333 train 7.791982 (lr=7.9233e-04) (hash(x)=5197474)
350
+ 334 train 7.628263 (lr=7.9200e-04) (hash(x)=5550786)
351
+ 335 train 7.680819 (lr=7.9167e-04) (hash(x)=6830813)
352
+ 336 train 7.546203 (lr=7.9132e-04) (hash(x)=5258435)
353
+ 337 train 7.556462 (lr=7.9097e-04) (hash(x)=5242548)
354
+ 338 train 7.452109 (lr=7.9062e-04) (hash(x)=3754886)
355
+ 339 train 7.340514 (lr=7.9026e-04) (hash(x)=4752771)
356
+ 340 train 7.495679 (lr=7.8989e-04) (hash(x)=5926875)
357
+ 341 train 8.001736 (lr=7.8951e-04) (hash(x)=6691249)
358
+ 342 train 7.696391 (lr=7.8913e-04) (hash(x)=7514623)
359
+ 343 train 7.789830 (lr=7.8874e-04) (hash(x)=6424933)
360
+ 344 train 7.738369 (lr=7.8834e-04) (hash(x)=6347757)
361
+ 345 train 7.681736 (lr=7.8794e-04) (hash(x)=6960120)
362
+ 346 train 7.731425 (lr=7.8753e-04) (hash(x)=7597197)
363
+ 347 train 7.627442 (lr=7.8711e-04) (hash(x)=5786517)
364
+ 348 train 7.681754 (lr=7.8669e-04) (hash(x)=6424363)
365
+ 349 train 7.764162 (lr=7.8626e-04) (hash(x)=7229689)
366
+ 350 val loss 7.6513
367
+ 350 val perplexity 2103.4543
368
+ 350 train 7.829522 (lr=7.8582e-04) (hash(x)=6772284)
369
+ 351 train 7.909485 (lr=7.8538e-04) (hash(x)=6680023)
370
+ 352 train 7.760623 (lr=7.8493e-04) (hash(x)=6218520)
371
+ 353 train 7.772234 (lr=7.8447e-04) (hash(x)=7688393)
372
+ 354 train 7.699875 (lr=7.8401e-04) (hash(x)=6327977)
373
+ 355 train 7.638770 (lr=7.8354e-04) (hash(x)=6474729)
374
+ 356 train 7.694256 (lr=7.8306e-04) (hash(x)=6657266)
375
+ 357 train 7.735797 (lr=7.8258e-04) (hash(x)=5946650)
376
+ 358 train 7.620255 (lr=7.8209e-04) (hash(x)=6734878)
377
+ 359 train 7.730633 (lr=7.8159e-04) (hash(x)=7523279)
378
+ 360 train 7.560632 (lr=7.8109e-04) (hash(x)=6179478)
379
+ 361 train 7.651139 (lr=7.8058e-04) (hash(x)=5531605)
380
+ 362 train 7.657497 (lr=7.8006e-04) (hash(x)=5844980)
381
+ 363 train 7.296112 (lr=7.7954e-04) (hash(x)=5508768)
382
+ 364 train 7.163783 (lr=7.7901e-04) (hash(x)=6337190)
383
+ 365 train 8.789912 (lr=7.7847e-04) (hash(x)=6415982)
384
+ 366 train 7.774729 (lr=7.7793e-04) (hash(x)=6828959)
385
+ 367 train 7.752129 (lr=7.7738e-04) (hash(x)=6635925)
386
+ 368 train 7.499676 (lr=7.7682e-04) (hash(x)=6358540)
387
+ 369 train 7.863929 (lr=7.7626e-04) (hash(x)=5923706)
388
+ 370 train 7.553805 (lr=7.7569e-04) (hash(x)=4843320)
389
+ 371 train 7.740863 (lr=7.7511e-04) (hash(x)=6663801)
390
+ 372 train 7.600640 (lr=7.7453e-04) (hash(x)=5583287)
391
+ 373 train 7.567753 (lr=7.7394e-04) (hash(x)=6113540)
392
+ 374 train 7.691015 (lr=7.7335e-04) (hash(x)=6389936)
393
+ 375 train 7.630345 (lr=7.7275e-04) (hash(x)=5869441)
394
+ 376 train 7.695819 (lr=7.7214e-04) (hash(x)=5148617)
395
+ 377 train 7.947289 (lr=7.7153e-04) (hash(x)=6913454)
396
+ 378 train 7.665747 (lr=7.7091e-04) (hash(x)=5356233)
397
+ 379 train 7.623816 (lr=7.7028e-04) (hash(x)=6698878)
398
+ 380 train 7.575048 (lr=7.6964e-04) (hash(x)=6280665)
399
+ 381 train 7.801994 (lr=7.6901e-04) (hash(x)=6094899)
400
+ 382 train 7.351874 (lr=7.6836e-04) (hash(x)=6249746)
401
+ 383 train 7.678961 (lr=7.6771e-04) (hash(x)=6349004)
402
+ 384 train 7.602983 (lr=7.6705e-04) (hash(x)=7013468)
403
+ 385 train 7.568041 (lr=7.6638e-04) (hash(x)=6045933)
404
+ 386 train 7.894169 (lr=7.6571e-04) (hash(x)=6571735)
405
+ 387 train 7.539466 (lr=7.6503e-04) (hash(x)=4986137)
406
+ 388 train 7.187318 (lr=7.6435e-04) (hash(x)=5244912)
407
+ 389 train 7.542772 (lr=7.6366e-04) (hash(x)=4798229)
408
+ 390 train 7.571453 (lr=7.6296e-04) (hash(x)=5815783)
409
+ 391 train 7.351474 (lr=7.6226e-04) (hash(x)=6008454)
410
+ 392 train 7.563411 (lr=7.6155e-04) (hash(x)=6407333)
411
+ 393 train 7.459795 (lr=7.6083e-04) (hash(x)=5938362)
412
+ 394 train 7.308470 (lr=7.6011e-04) (hash(x)=6077124)
413
+ 395 train 7.630810 (lr=7.5938e-04) (hash(x)=6550770)
414
+ 396 train 7.476097 (lr=7.5865e-04) (hash(x)=6181528)
415
+ 397 train 7.724872 (lr=7.5791e-04) (hash(x)=7055344)
416
+ 398 train 7.829448 (lr=7.5716e-04) (hash(x)=7689348)
417
+ 399 train 7.761893 (lr=7.5641e-04) (hash(x)=7682741)
418
+ 400 val loss 7.6549
419
+ 400 val perplexity 2111.0076
420
+ 400 train 7.613126 (lr=7.5565e-04) (hash(x)=5189401)
421
+ 401 train 7.569535 (lr=7.5489e-04) (hash(x)=7372095)
422
+ 402 train 7.685690 (lr=7.5412e-04) (hash(x)=6838219)
423
+ 403 train 8.185756 (lr=7.5334e-04) (hash(x)=7892158)
424
+ 404 train 7.810161 (lr=7.5256e-04) (hash(x)=5677077)
425
+ 405 train 7.425462 (lr=7.5177e-04) (hash(x)=5929740)
426
+ 406 train 7.576531 (lr=7.5097e-04) (hash(x)=5285055)
427
+ 407 train 7.686718 (lr=7.5017e-04) (hash(x)=7794028)
428
+ 408 train 7.759375 (lr=7.4937e-04) (hash(x)=6640802)
429
+ 409 train 7.837638 (lr=7.4855e-04) (hash(x)=6839923)
430
+ 410 train 7.312483 (lr=7.4773e-04) (hash(x)=4020997)
431
+ 411 train 7.536624 (lr=7.4691e-04) (hash(x)=7093523)
432
+ 412 train 7.830489 (lr=7.4608e-04) (hash(x)=7370045)
433
+ 413 train 7.545541 (lr=7.4524e-04) (hash(x)=5341312)
434
+ 414 train 7.569070 (lr=7.4440e-04) (hash(x)=5589312)
435
+ 415 train 8.036122 (lr=7.4355e-04) (hash(x)=6166062)
436
+ 416 train 7.920056 (lr=7.4270e-04) (hash(x)=8354149)
437
+ 417 train 7.549467 (lr=7.4184e-04) (hash(x)=5992164)
438
+ 418 train 7.597467 (lr=7.4097e-04) (hash(x)=4748657)
439
+ 419 train 7.603067 (lr=7.4010e-04) (hash(x)=6645781)
440
+ 420 train 7.815924 (lr=7.3922e-04) (hash(x)=6689147)
441
+ 421 train 7.701595 (lr=7.3834e-04) (hash(x)=6207724)
442
+ 422 train 7.614268 (lr=7.3745e-04) (hash(x)=6232863)
443
+ 423 train 7.365573 (lr=7.3655e-04) (hash(x)=4811192)
444
+ 424 train 7.398789 (lr=7.3565e-04) (hash(x)=6211761)
445
+ 425 train 7.597176 (lr=7.3474e-04) (hash(x)=6906335)
446
+ 426 train 7.527941 (lr=7.3383e-04) (hash(x)=4922365)
447
+ 427 train 7.577005 (lr=7.3291e-04) (hash(x)=5970866)
448
+ 428 train 7.501617 (lr=7.3199e-04) (hash(x)=6417833)
449
+ 429 train 7.586618 (lr=7.3106e-04) (hash(x)=5238015)
450
+ 430 train 7.493934 (lr=7.3012e-04) (hash(x)=5375451)
451
+ 431 train 7.615366 (lr=7.2918e-04) (hash(x)=7822680)
452
+ 432 train 7.638727 (lr=7.2824e-04) (hash(x)=6173200)
453
+ 433 train 7.393836 (lr=7.2729e-04) (hash(x)=5478019)
454
+ 434 train 7.623751 (lr=7.2633e-04) (hash(x)=5597859)
455
+ 435 train 7.487439 (lr=7.2537e-04) (hash(x)=6317317)
456
+ 436 train 7.694211 (lr=7.2440e-04) (hash(x)=7120467)
457
+ 437 train 7.773491 (lr=7.2342e-04) (hash(x)=5832344)
458
+ 438 train 7.646698 (lr=7.2244e-04) (hash(x)=6813802)
459
+ 439 train 7.708512 (lr=7.2146e-04) (hash(x)=6431409)
460
+ 440 train 7.771479 (lr=7.2047e-04) (hash(x)=5886861)
461
+ 441 train 7.484012 (lr=7.1947e-04) (hash(x)=5900130)
462
+ 442 train 7.726596 (lr=7.1847e-04) (hash(x)=6061563)
463
+ 443 train 7.713188 (lr=7.1747e-04) (hash(x)=6653337)
464
+ 444 train 7.632481 (lr=7.1645e-04) (hash(x)=6575992)
465
+ 445 train 7.624427 (lr=7.1544e-04) (hash(x)=6696784)
466
+ 446 train 7.764124 (lr=7.1441e-04) (hash(x)=6242161)
467
+ 447 train 7.576370 (lr=7.1339e-04) (hash(x)=5323032)
468
+ 448 train 7.370094 (lr=7.1235e-04) (hash(x)=6439367)
469
+ 449 train 7.638541 (lr=7.1131e-04) (hash(x)=5372100)
470
+ 450 val loss 7.6155
471
+ 450 val perplexity 2029.3579
472
+ 450 train 7.672157 (lr=7.1027e-04) (hash(x)=6344849)
473
+ 451 train 7.301679 (lr=7.0922e-04) (hash(x)=5125339)
474
+ 452 train 7.287890 (lr=7.0817e-04) (hash(x)=4876156)
475
+ 453 train 7.423781 (lr=7.0711e-04) (hash(x)=5664463)
476
+ 454 train 7.298251 (lr=7.0604e-04) (hash(x)=6010042)
477
+ 455 train 7.529947 (lr=7.0497e-04) (hash(x)=7363286)
478
+ 456 train 8.159866 (lr=7.0390e-04) (hash(x)=4873268)
479
+ 457 train 7.874198 (lr=7.0282e-04) (hash(x)=7444468)
480
+ 458 train 7.886745 (lr=7.0173e-04) (hash(x)=6698134)
481
+ 459 train 7.681684 (lr=7.0064e-04) (hash(x)=7670050)
482
+ 460 train 7.685728 (lr=6.9955e-04) (hash(x)=6280866)
483
+ 461 train 7.565620 (lr=6.9845e-04) (hash(x)=8189420)
484
+ 462 train 7.697933 (lr=6.9734e-04) (hash(x)=6878292)
485
+ 463 train 7.418419 (lr=6.9623e-04) (hash(x)=5616075)
486
+ 464 train 7.605835 (lr=6.9512e-04) (hash(x)=5839939)
487
+ 465 train 7.464160 (lr=6.9400e-04) (hash(x)=6360924)
488
+ 466 train 7.939818 (lr=6.9287e-04) (hash(x)=6466475)
489
+ 467 train 7.668953 (lr=6.9174e-04) (hash(x)=6593764)
490
+ 468 train 8.157442 (lr=6.9061e-04) (hash(x)=7389698)
491
+ 469 train 7.796253 (lr=6.8947e-04) (hash(x)=6764377)
492
+ 470 train 7.525248 (lr=6.8832e-04) (hash(x)=7293570)
493
+ 471 train 7.565753 (lr=6.8717e-04) (hash(x)=5672608)
494
+ 472 train 7.397666 (lr=6.8602e-04) (hash(x)=5587975)
495
+ 473 train 7.322899 (lr=6.8486e-04) (hash(x)=6985256)
496
+ 474 train 7.898012 (lr=6.8370e-04) (hash(x)=6549231)
497
+ 475 train 7.594752 (lr=6.8253e-04) (hash(x)=6828653)
498
+ 476 train 7.436860 (lr=6.8136e-04) (hash(x)=5478970)
499
+ 477 train 7.442824 (lr=6.8018e-04) (hash(x)=5199845)
500
+ 478 train 7.532060 (lr=6.7900e-04) (hash(x)=6482073)
501
+ 479 train 7.639960 (lr=6.7781e-04) (hash(x)=6839867)
502
+ 480 train 7.624708 (lr=6.7662e-04) (hash(x)=7073454)
503
+ 481 train 7.897598 (lr=6.7542e-04) (hash(x)=6928712)
504
+ 482 train 7.588937 (lr=6.7422e-04) (hash(x)=5962982)
505
+ 483 train 7.707046 (lr=6.7302e-04) (hash(x)=8426992)
506
+ 484 train 7.629753 (lr=6.7181e-04) (hash(x)=6422394)
507
+ 485 train 7.578392 (lr=6.7059e-04) (hash(x)=5330355)
508
+ 486 train 7.033719 (lr=6.6937e-04) (hash(x)=4111491)
509
+ 487 train 7.064920 (lr=6.6815e-04) (hash(x)=4916928)
510
+ 488 train 7.224048 (lr=6.6692e-04) (hash(x)=6610244)
511
+ 489 train 7.760235 (lr=6.6569e-04) (hash(x)=7287928)
512
+ 490 train 7.603540 (lr=6.6446e-04) (hash(x)=6525083)
513
+ 491 train 7.581148 (lr=6.6322e-04) (hash(x)=6484050)
514
+ 492 train 7.687777 (lr=6.6197e-04) (hash(x)=6573424)
515
+ 493 train 7.580771 (lr=6.6072e-04) (hash(x)=5062850)
516
+ 494 train 7.559315 (lr=6.5947e-04) (hash(x)=6215995)
517
+ 495 train 7.679546 (lr=6.5821e-04) (hash(x)=8353379)
518
+ 496 train 7.552433 (lr=6.5695e-04) (hash(x)=6344823)
519
+ 497 train 7.513611 (lr=6.5568e-04) (hash(x)=6805358)
520
+ 498 train 7.401032 (lr=6.5441e-04) (hash(x)=6018070)
521
+ 499 train 7.702428 (lr=6.5314e-04) (hash(x)=6552510)
522
+ 500 val loss 7.5545
523
+ 500 val perplexity 1909.3619
524
+ 500 train 7.470752 (lr=6.5186e-04) (hash(x)=7218108)
525
+ 501 train 7.678087 (lr=6.5058e-04) (hash(x)=5105545)
526
+ 502 train 7.665430 (lr=6.4929e-04) (hash(x)=5997860)
527
+ 503 train 7.451200 (lr=6.4800e-04) (hash(x)=4838871)
528
+ 504 train 7.502067 (lr=6.4670e-04) (hash(x)=7105888)
529
+ 505 train 7.804259 (lr=6.4540e-04) (hash(x)=5945286)
530
+ 506 train 6.956537 (lr=6.4410e-04) (hash(x)=4381720)
531
+ 507 train 6.653010 (lr=6.4280e-04) (hash(x)=3021697)
532
+ 508 train 7.066971 (lr=6.4148e-04) (hash(x)=6826973)
533
+ 509 train 7.671076 (lr=6.4017e-04) (hash(x)=6443070)
534
+ 510 train 7.353661 (lr=6.3885e-04) (hash(x)=5520637)
535
+ 511 train 7.648328 (lr=6.3753e-04) (hash(x)=6795665)
536
+ 512 train 7.391486 (lr=6.3620e-04) (hash(x)=6063891)
537
+ 513 train 7.955371 (lr=6.3487e-04) (hash(x)=7683068)
538
+ 514 train 7.212945 (lr=6.3354e-04) (hash(x)=5526694)
539
+ 515 train 7.249269 (lr=6.3220e-04) (hash(x)=5486935)
540
+ 516 train 7.588069 (lr=6.3086e-04) (hash(x)=6002252)
541
+ 517 train 7.561894 (lr=6.2952e-04) (hash(x)=6115232)
542
+ 518 train 7.859114 (lr=6.2817e-04) (hash(x)=9039136)
543
+ 519 train 7.505990 (lr=6.2681e-04) (hash(x)=6678038)
544
+ 520 train 7.664935 (lr=6.2546e-04) (hash(x)=5383926)
545
+ 521 train 7.705895 (lr=6.2410e-04) (hash(x)=5747629)
546
+ 522 train 7.605730 (lr=6.2274e-04) (hash(x)=7090040)
547
+ 523 train 7.807600 (lr=6.2137e-04) (hash(x)=6657714)
548
+ 524 train 7.510019 (lr=6.2000e-04) (hash(x)=5414034)
549
+ 525 train 7.584471 (lr=6.1863e-04) (hash(x)=6190642)
550
+ 526 train 7.718000 (lr=6.1725e-04) (hash(x)=7431177)
551
+ 527 train 7.567379 (lr=6.1587e-04) (hash(x)=6112215)
552
+ 528 train 7.674254 (lr=6.1449e-04) (hash(x)=6785874)
553
+ 529 train 7.431906 (lr=6.1310e-04) (hash(x)=7314989)
554
+ 530 train 7.643406 (lr=6.1171e-04) (hash(x)=6222614)
555
+ 531 train 7.753233 (lr=6.1031e-04) (hash(x)=8353143)
556
+ 532 train 7.860522 (lr=6.0892e-04) (hash(x)=6752498)
557
+ 533 train 7.574724 (lr=6.0752e-04) (hash(x)=5912570)
558
+ 534 train 7.728037 (lr=6.0611e-04) (hash(x)=5621785)
559
+ 535 train 7.619519 (lr=6.0471e-04) (hash(x)=5915361)
560
+ 536 train 7.444937 (lr=6.0330e-04) (hash(x)=6853672)
561
+ 537 train 7.431011 (lr=6.0188e-04) (hash(x)=6369494)
562
+ 538 train 7.576741 (lr=6.0047e-04) (hash(x)=6039652)
563
+ 539 train 7.726073 (lr=5.9905e-04) (hash(x)=6254885)
564
+ 540 train 7.572514 (lr=5.9762e-04) (hash(x)=5829860)
565
+ 541 train 7.514441 (lr=5.9620e-04) (hash(x)=5425120)
566
+ 542 train 7.263112 (lr=5.9477e-04) (hash(x)=5209746)
567
+ 543 train 7.829821 (lr=5.9334e-04) (hash(x)=5771588)
568
+ 544 train 7.535992 (lr=5.9190e-04) (hash(x)=8298790)
569
+ 545 train 7.617200 (lr=5.9047e-04) (hash(x)=6763967)
570
+ 546 train 7.615287 (lr=5.8903e-04) (hash(x)=4882397)
571
+ 547 train 7.353822 (lr=5.8758e-04) (hash(x)=5561507)
572
+ 548 train 7.372556 (lr=5.8614e-04) (hash(x)=5762755)
573
+ 549 train 7.491364 (lr=5.8469e-04) (hash(x)=7081151)
574
+ 550 val loss 7.5244
575
+ 550 val perplexity 1852.7688
576
+ 550 train 7.383849 (lr=5.8323e-04) (hash(x)=5047520)
577
+ 551 train 7.550940 (lr=5.8178e-04) (hash(x)=5688829)
578
+ 552 train 7.423963 (lr=5.8032e-04) (hash(x)=7008925)
579
+ 553 train 7.531371 (lr=5.7886e-04) (hash(x)=7821554)
580
+ 554 train 7.576802 (lr=5.7740e-04) (hash(x)=6469035)
581
+ 555 train 7.392823 (lr=5.7593e-04) (hash(x)=5371951)
582
+ 556 train 7.696354 (lr=5.7447e-04) (hash(x)=6436050)
583
+ 557 train 7.610435 (lr=5.7300e-04) (hash(x)=8120149)
584
+ 558 train 7.846079 (lr=5.7152e-04) (hash(x)=6189935)
585
+ 559 train 7.533603 (lr=5.7005e-04) (hash(x)=5443305)
586
+ 560 train 7.647222 (lr=5.6857e-04) (hash(x)=6294312)
587
+ 561 train 7.273014 (lr=5.6709e-04) (hash(x)=7066030)
588
+ 562 train 7.711017 (lr=5.6560e-04) (hash(x)=7744989)
589
+ 563 train 7.851597 (lr=5.6412e-04) (hash(x)=5016757)
590
+ 564 train 7.682428 (lr=5.6263e-04) (hash(x)=8001284)
591
+ 565 train 7.569969 (lr=5.6114e-04) (hash(x)=6296163)
592
+ 566 train 7.678930 (lr=5.5965e-04) (hash(x)=6027688)
593
+ 567 train 7.596194 (lr=5.5815e-04) (hash(x)=6901933)
594
+ 568 train 7.447910 (lr=5.5666e-04) (hash(x)=5124305)
595
+ 569 train 7.595589 (lr=5.5516e-04) (hash(x)=8056633)
596
+ 570 train 7.635362 (lr=5.5365e-04) (hash(x)=6677566)
597
+ 571 train 7.675000 (lr=5.5215e-04) (hash(x)=6019085)
598
+ 572 train 7.383440 (lr=5.5064e-04) (hash(x)=5924495)
599
+ 573 train 7.518988 (lr=5.4913e-04) (hash(x)=7003893)
600
+ 574 train 7.481855 (lr=5.4762e-04) (hash(x)=5563075)
601
+ 575 train 7.299489 (lr=5.4611e-04) (hash(x)=4294425)
602
+ 576 train 7.488084 (lr=5.4460e-04) (hash(x)=5677870)
603
+ 577 train 7.401842 (lr=5.4308e-04) (hash(x)=6568540)
604
+ 578 train 7.599443 (lr=5.4156e-04) (hash(x)=7473394)
605
+ 579 train 7.659194 (lr=5.4004e-04) (hash(x)=6095229)
606
+ 580 train 7.641590 (lr=5.3852e-04) (hash(x)=6946168)
607
+ 581 train 7.437044 (lr=5.3699e-04) (hash(x)=5512150)
608
+ 582 train 7.080405 (lr=5.3547e-04) (hash(x)=4380268)
609
+ 583 train 7.133731 (lr=5.3394e-04) (hash(x)=4363941)
610
+ 584 train 7.321065 (lr=5.3241e-04) (hash(x)=4836009)
611
+ 585 train 7.609378 (lr=5.3088e-04) (hash(x)=6457783)
612
+ 586 train 7.470002 (lr=5.2934e-04) (hash(x)=6378784)
613
+ 587 train 7.357757 (lr=5.2781e-04) (hash(x)=4693798)
614
+ 588 train 7.719779 (lr=5.2627e-04) (hash(x)=9193310)
615
+ 589 train 8.393578 (lr=5.2473e-04) (hash(x)=8759806)
616
+ 590 train 8.701630 (lr=5.2319e-04) (hash(x)=9176140)
617
+ 591 train 7.473298 (lr=5.2165e-04) (hash(x)=5486335)
618
+ 592 train 7.722656 (lr=5.2011e-04) (hash(x)=7868436)
619
+ 593 train 7.582102 (lr=5.1856e-04) (hash(x)=6964993)
620
+ 594 train 7.779227 (lr=5.1702e-04) (hash(x)=6741270)
621
+ 595 train 8.008024 (lr=5.1547e-04) (hash(x)=7907450)
622
+ 596 train 7.877439 (lr=5.1392e-04) (hash(x)=6443452)
623
+ 597 train 7.518343 (lr=5.1237e-04) (hash(x)=5236642)
624
+ 598 train 7.646244 (lr=5.1082e-04) (hash(x)=5604795)
625
+ 599 train 7.578408 (lr=5.0926e-04) (hash(x)=7295165)
626
+ 600 val loss 7.5285
627
+ 600 val perplexity 1860.3167
628
+ 600 train 7.549865 (lr=5.0771e-04) (hash(x)=6034549)
629
+ 601 train 7.589498 (lr=5.0615e-04) (hash(x)=4913602)
630
+ 602 train 7.550111 (lr=5.0459e-04) (hash(x)=6271294)
631
+ 603 train 7.676216 (lr=5.0303e-04) (hash(x)=6814026)
632
+ 604 train 7.530697 (lr=5.0147e-04) (hash(x)=6848040)
633
+ 605 train 7.883473 (lr=4.9991e-04) (hash(x)=7042051)
634
+ 606 train 7.745343 (lr=4.9835e-04) (hash(x)=6481002)
635
+ 607 train 7.381203 (lr=4.9679e-04) (hash(x)=6267424)
636
+ 608 train 7.714677 (lr=4.9522e-04) (hash(x)=7665306)
637
+ 609 train 7.625935 (lr=4.9366e-04) (hash(x)=5614727)
638
+ 610 train 7.624576 (lr=4.9209e-04) (hash(x)=7039197)
639
+ 611 train 7.672980 (lr=4.9052e-04) (hash(x)=8086437)
640
+ 612 train 7.752872 (lr=4.8895e-04) (hash(x)=6993846)
641
+ 613 train 7.798313 (lr=4.8738e-04) (hash(x)=5900143)
642
+ 614 train 7.668924 (lr=4.8581e-04) (hash(x)=4841318)
643
+ 615 train 7.625908 (lr=4.8424e-04) (hash(x)=5270452)
644
+ 616 train 7.544415 (lr=4.8267e-04) (hash(x)=5955026)
645
+ 617 train 7.640775 (lr=4.8109e-04) (hash(x)=8617707)
646
+ 618 train 7.487264 (lr=4.7952e-04) (hash(x)=5159401)
647
+ 619 train 7.475078 (lr=4.7795e-04) (hash(x)=6420820)
648
+ 620 train 7.449168 (lr=4.7637e-04) (hash(x)=6628863)
649
+ 621 train 7.242607 (lr=4.7479e-04) (hash(x)=6215930)
650
+ 622 train 7.580864 (lr=4.7322e-04) (hash(x)=5248645)
651
+ 623 train 7.461339 (lr=4.7164e-04) (hash(x)=6305297)
652
+ 624 train 8.029337 (lr=4.7006e-04) (hash(x)=6989704)
653
+ 625 train 7.593973 (lr=4.6848e-04) (hash(x)=4026682)
654
+ 626 train 7.484196 (lr=4.6690e-04) (hash(x)=6367688)
655
+ 627 train 7.529243 (lr=4.6532e-04) (hash(x)=7889849)
656
+ 628 train 7.694878 (lr=4.6374e-04) (hash(x)=7334719)
657
+ 629 train 7.830132 (lr=4.6216e-04) (hash(x)=6141721)
658
+ 630 train 8.669253 (lr=4.6058e-04) (hash(x)=5056572)
659
+ 631 train 8.044258 (lr=4.5900e-04) (hash(x)=6040077)
660
+ 632 train 8.100708 (lr=4.5742e-04) (hash(x)=6866651)
661
+ 633 train 8.385164 (lr=4.5583e-04) (hash(x)=6155449)
662
+ 634 train 7.893442 (lr=4.5425e-04) (hash(x)=6690174)
663
+ 635 train 7.601172 (lr=4.5267e-04) (hash(x)=5652497)
664
+ 636 train 7.621835 (lr=4.5109e-04) (hash(x)=5931050)
665
+ 637 train 7.746641 (lr=4.4950e-04) (hash(x)=6961314)
666
+ 638 train 7.536393 (lr=4.4792e-04) (hash(x)=5973636)
667
+ 639 train 7.555904 (lr=4.4634e-04) (hash(x)=7130251)
668
+ 640 train 7.309475 (lr=4.4475e-04) (hash(x)=7503390)
669
+ 641 train 7.336823 (lr=4.4317e-04) (hash(x)=5927461)
670
+ 642 train 7.559348 (lr=4.4158e-04) (hash(x)=6196741)
671
+ 643 train 7.467366 (lr=4.4000e-04) (hash(x)=6610177)
672
+ 644 train 7.452167 (lr=4.3842e-04) (hash(x)=6635147)
673
+ 645 train 7.679495 (lr=4.3683e-04) (hash(x)=7277580)
674
+ 646 train 7.317497 (lr=4.3525e-04) (hash(x)=5050080)
675
+ 647 train 7.393287 (lr=4.3366e-04) (hash(x)=6508350)
676
+ 648 train 7.633282 (lr=4.3208e-04) (hash(x)=5276338)
677
+ 649 train 7.786057 (lr=4.3050e-04) (hash(x)=6536034)
678
+ 650 val loss 7.5267
679
+ 650 val perplexity 1856.9524
680
+ 650 train 7.578208 (lr=4.2891e-04) (hash(x)=6944772)
681
+ 651 train 7.613605 (lr=4.2733e-04) (hash(x)=6994983)
682
+ 652 train 7.545150 (lr=4.2575e-04) (hash(x)=7172017)
683
+ 653 train 7.639430 (lr=4.2417e-04) (hash(x)=8700721)
684
+ 654 train 7.615660 (lr=4.2258e-04) (hash(x)=6774360)
685
+ 655 train 7.444989 (lr=4.2100e-04) (hash(x)=5859576)
686
+ 656 train 7.136799 (lr=4.1942e-04) (hash(x)=5899275)
687
+ 657 train 7.056261 (lr=4.1784e-04) (hash(x)=5264962)
688
+ 658 train 6.897883 (lr=4.1626e-04) (hash(x)=5679861)
689
+ 659 train 7.332005 (lr=4.1468e-04) (hash(x)=5487065)
690
+ 660 train 7.112918 (lr=4.1310e-04) (hash(x)=4239476)
691
+ 661 train 7.425945 (lr=4.1152e-04) (hash(x)=5731624)
692
+ 662 train 7.527098 (lr=4.0994e-04) (hash(x)=5883465)
693
+ 663 train 7.285338 (lr=4.0836e-04) (hash(x)=4892065)
694
+ 664 train 7.148315 (lr=4.0678e-04) (hash(x)=5858782)
695
+ 665 train 7.447224 (lr=4.0521e-04) (hash(x)=5489496)
696
+ 666 train 7.110409 (lr=4.0363e-04) (hash(x)=4485195)
697
+ 667 train 7.142278 (lr=4.0205e-04) (hash(x)=4933674)
698
+ 668 train 7.421928 (lr=4.0048e-04) (hash(x)=5746292)
699
+ 669 train 7.431533 (lr=3.9891e-04) (hash(x)=7021003)
700
+ 670 train 7.598784 (lr=3.9733e-04) (hash(x)=5876710)
701
+ 671 train 7.660367 (lr=3.9576e-04) (hash(x)=7317289)
702
+ 672 train 7.760567 (lr=3.9419e-04) (hash(x)=5598226)
703
+ 673 train 7.562491 (lr=3.9262e-04) (hash(x)=7869305)
704
+ 674 train 7.611700 (lr=3.9105e-04) (hash(x)=6611408)
705
+ 675 train 7.587976 (lr=3.8948e-04) (hash(x)=6811522)
706
+ 676 train 7.545211 (lr=3.8791e-04) (hash(x)=6704714)
707
+ 677 train 7.430015 (lr=3.8634e-04) (hash(x)=6601423)
708
+ 678 train 7.661256 (lr=3.8478e-04) (hash(x)=6726071)
709
+ 679 train 7.163018 (lr=3.8321e-04) (hash(x)=5510218)
710
+ 680 train 7.934966 (lr=3.8165e-04) (hash(x)=7950952)
711
+ 681 train 7.630328 (lr=3.8009e-04) (hash(x)=7180298)
712
+ 682 train 7.424460 (lr=3.7853e-04) (hash(x)=6068813)
713
+ 683 train 7.567736 (lr=3.7697e-04) (hash(x)=7304235)
714
+ 684 train 7.623631 (lr=3.7541e-04) (hash(x)=7441806)
715
+ 685 train 7.726415 (lr=3.7385e-04) (hash(x)=8111920)
716
+ 686 train 8.329744 (lr=3.7229e-04) (hash(x)=6222783)
717
+ 687 train 7.561632 (lr=3.7074e-04) (hash(x)=6752265)
718
+ 688 train 7.365996 (lr=3.6918e-04) (hash(x)=6147634)
719
+ 689 train 7.399496 (lr=3.6763e-04) (hash(x)=6788720)
720
+ 690 train 7.550434 (lr=3.6608e-04) (hash(x)=6413518)
721
+ 691 train 7.435906 (lr=3.6453e-04) (hash(x)=5994476)
722
+ 692 train 7.319509 (lr=3.6298e-04) (hash(x)=5462082)
723
+ 693 train 7.369944 (lr=3.6144e-04) (hash(x)=5862533)
724
+ 694 train 7.546873 (lr=3.5989e-04) (hash(x)=7132796)
725
+ 695 train 7.428020 (lr=3.5835e-04) (hash(x)=6530867)
726
+ 696 train 7.503792 (lr=3.5681e-04) (hash(x)=7932207)
727
+ 697 train 7.468715 (lr=3.5527e-04) (hash(x)=5914734)
728
+ 698 train 7.591680 (lr=3.5373e-04) (hash(x)=6361594)
729
+ 699 train 7.556170 (lr=3.5219e-04) (hash(x)=5746260)
730
+ 700 val loss 7.4791
731
+ 700 val perplexity 1770.6807
732
+ 700 train 7.495326 (lr=3.5066e-04) (hash(x)=7065743)
733
+ 701 train 7.439205 (lr=3.4912e-04) (hash(x)=6856097)
734
+ 702 train 7.512464 (lr=3.4759e-04) (hash(x)=7588648)
735
+ 703 train 7.430010 (lr=3.4606e-04) (hash(x)=5790078)
736
+ 704 train 7.448999 (lr=3.4453e-04) (hash(x)=6069422)
737
+ 705 train 7.783248 (lr=3.4301e-04) (hash(x)=5805002)
738
+ 706 train 7.512825 (lr=3.4148e-04) (hash(x)=5344711)
739
+ 707 train 7.321788 (lr=3.3996e-04) (hash(x)=6430135)
740
+ 708 train 7.464485 (lr=3.3844e-04) (hash(x)=6317763)
741
+ 709 train 7.345987 (lr=3.3692e-04) (hash(x)=6156715)
742
+ 710 train 7.713501 (lr=3.3540e-04) (hash(x)=6321494)
743
+ 711 train 7.757513 (lr=3.3389e-04) (hash(x)=7614023)
744
+ 712 train 7.727546 (lr=3.3238e-04) (hash(x)=6740380)
745
+ 713 train 7.563003 (lr=3.3087e-04) (hash(x)=4861744)
746
+ 714 train 7.425184 (lr=3.2936e-04) (hash(x)=6542179)
747
+ 715 train 7.493335 (lr=3.2785e-04) (hash(x)=5244861)
748
+ 716 train 7.634149 (lr=3.2635e-04) (hash(x)=7306636)
749
+ 717 train 7.458386 (lr=3.2484e-04) (hash(x)=7163697)
750
+ 718 train 7.653857 (lr=3.2334e-04) (hash(x)=6421642)
751
+ 719 train 7.434999 (lr=3.2185e-04) (hash(x)=5245146)
752
+ 720 train 7.539366 (lr=3.2035e-04) (hash(x)=6046027)
753
+ 721 train 7.346102 (lr=3.1886e-04) (hash(x)=6153866)
754
+ 722 train 7.503121 (lr=3.1737e-04) (hash(x)=5827481)
755
+ 723 train 7.448753 (lr=3.1588e-04) (hash(x)=6415565)
756
+ 724 train 7.597524 (lr=3.1440e-04) (hash(x)=6409570)
757
+ 725 train 8.010787 (lr=3.1291e-04) (hash(x)=7835853)
758
+ 726 train 7.132950 (lr=3.1143e-04) (hash(x)=4827589)
759
+ 727 train 7.266757 (lr=3.0995e-04) (hash(x)=5786972)
760
+ 728 train 7.684786 (lr=3.0848e-04) (hash(x)=6736612)
761
+ 729 train 7.519439 (lr=3.0700e-04) (hash(x)=6733560)
762
+ 730 train 7.683266 (lr=3.0553e-04) (hash(x)=4877208)
763
+ 731 train 7.194800 (lr=3.0407e-04) (hash(x)=6131703)
764
+ 732 train 7.320161 (lr=3.0260e-04) (hash(x)=6533769)
765
+ 733 train 7.267010 (lr=3.0114e-04) (hash(x)=6001331)
766
+ 734 train 7.847549 (lr=2.9968e-04) (hash(x)=10602643)
767
+ 735 train 7.749301 (lr=2.9822e-04) (hash(x)=6346459)
768
+ 736 train 7.465252 (lr=2.9677e-04) (hash(x)=6728215)
769
+ 737 train 7.756523 (lr=2.9531e-04) (hash(x)=8943770)
770
+ 738 train 7.576502 (lr=2.9386e-04) (hash(x)=7141912)
771
+ 739 train 7.534949 (lr=2.9242e-04) (hash(x)=6504131)
772
+ 740 train 7.363389 (lr=2.9097e-04) (hash(x)=6461667)
773
+ 741 train 7.570962 (lr=2.8953e-04) (hash(x)=5869339)
774
+ 742 train 7.959754 (lr=2.8810e-04) (hash(x)=7948065)
775
+ 743 train 7.249936 (lr=2.8666e-04) (hash(x)=5209234)
776
+ 744 train 7.368248 (lr=2.8523e-04) (hash(x)=6372244)
777
+ 745 train 7.554463 (lr=2.8380e-04) (hash(x)=7678937)
778
+ 746 train 7.463893 (lr=2.8238e-04) (hash(x)=6519438)
779
+ 747 train 7.518559 (lr=2.8095e-04) (hash(x)=6163272)
780
+ 748 train 7.593203 (lr=2.7953e-04) (hash(x)=7025209)
781
+ 749 train 7.435372 (lr=2.7812e-04) (hash(x)=5774172)
782
+ 750 val loss 7.4711
783
+ 750 val perplexity 1756.4613
784
+ 750 train 7.569447 (lr=2.7670e-04) (hash(x)=5327301)
785
+ 751 train 7.386265 (lr=2.7529e-04) (hash(x)=5676768)
786
+ 752 train 7.076365 (lr=2.7389e-04) (hash(x)=4823304)
787
+ 753 train 7.185856 (lr=2.7248e-04) (hash(x)=6443895)
788
+ 754 train 7.975614 (lr=2.7108e-04) (hash(x)=8032890)
789
+ 755 train 8.015595 (lr=2.6969e-04) (hash(x)=6090561)
790
+ 756 train 7.409878 (lr=2.6829e-04) (hash(x)=4514262)
791
+ 757 train 7.832635 (lr=2.6690e-04) (hash(x)=7788628)
792
+ 758 train 7.783027 (lr=2.6551e-04) (hash(x)=7239456)
793
+ 759 train 7.815177 (lr=2.6413e-04) (hash(x)=5863092)
794
+ 760 train 7.496313 (lr=2.6275e-04) (hash(x)=5572628)
795
+ 761 train 7.577047 (lr=2.6137e-04) (hash(x)=7141354)
796
+ 762 train 7.772326 (lr=2.6000e-04) (hash(x)=7272638)
797
+ 763 train 7.620050 (lr=2.5863e-04) (hash(x)=7201312)
798
+ 764 train 7.654937 (lr=2.5726e-04) (hash(x)=6015388)
799
+ 765 train 7.691962 (lr=2.5590e-04) (hash(x)=6414278)
800
+ 766 train 7.471591 (lr=2.5454e-04) (hash(x)=7137466)
801
+ 767 train 8.061747 (lr=2.5319e-04) (hash(x)=7019489)
802
+ 768 train 7.592669 (lr=2.5183e-04) (hash(x)=7233453)
803
+ 769 train 7.363982 (lr=2.5048e-04) (hash(x)=7914626)
804
+ 770 train 7.441041 (lr=2.4914e-04) (hash(x)=5764080)
805
+ 771 train 7.462357 (lr=2.4780e-04) (hash(x)=6225608)
806
+ 772 train 7.541595 (lr=2.4646e-04) (hash(x)=8097255)
807
+ 773 train 7.535522 (lr=2.4513e-04) (hash(x)=5998078)
808
+ 774 train 7.663992 (lr=2.4380e-04) (hash(x)=5416254)
809
+ 775 train 7.383049 (lr=2.4247e-04) (hash(x)=5483019)
810
+ 776 train 7.178077 (lr=2.4115e-04) (hash(x)=4702208)
811
+ 777 train 7.507473 (lr=2.3983e-04) (hash(x)=5911642)
812
+ 778 train 7.538190 (lr=2.3852e-04) (hash(x)=6132487)
813
+ 779 train 7.671747 (lr=2.3720e-04) (hash(x)=5903258)
814
+ 780 train 7.586809 (lr=2.3590e-04) (hash(x)=7915382)
815
+ 781 train 7.582770 (lr=2.3460e-04) (hash(x)=5632006)
816
+ 782 train 7.409109 (lr=2.3330e-04) (hash(x)=6518211)
817
+ 783 train 7.500209 (lr=2.3200e-04) (hash(x)=5968716)
818
+ 784 train 7.592290 (lr=2.3071e-04) (hash(x)=7344525)
819
+ 785 train 7.497097 (lr=2.2942e-04) (hash(x)=6401968)
820
+ 786 train 7.425156 (lr=2.2814e-04) (hash(x)=6276127)
821
+ 787 train 7.542342 (lr=2.2686e-04) (hash(x)=5778017)
822
+ 788 train 7.079448 (lr=2.2559e-04) (hash(x)=5387306)
823
+ 789 train 7.043993 (lr=2.2432e-04) (hash(x)=5772567)
824
+ 790 train 7.265402 (lr=2.2305e-04) (hash(x)=6383748)
825
+ 791 train 7.579001 (lr=2.2179e-04) (hash(x)=7780194)
826
+ 792 train 7.436978 (lr=2.2053e-04) (hash(x)=7119030)
827
+ 793 train 7.548909 (lr=2.1928e-04) (hash(x)=6424771)
828
+ 794 train 7.377677 (lr=2.1803e-04) (hash(x)=6540151)
829
+ 795 train 7.376254 (lr=2.1678e-04) (hash(x)=6140998)
830
+ 796 train 7.474971 (lr=2.1554e-04) (hash(x)=6208271)
831
+ 797 train 7.310043 (lr=2.1431e-04) (hash(x)=7859566)
832
+ 798 train 7.858378 (lr=2.1308e-04) (hash(x)=7064477)
833
+ 799 train 7.126901 (lr=2.1185e-04) (hash(x)=3784321)
834
+ 800 val loss 7.4483
835
+ 800 val perplexity 1716.9646
836
+ 800 train 7.421199 (lr=2.1063e-04) (hash(x)=4472758)
837
+ 801 train 7.359682 (lr=2.0941e-04) (hash(x)=5557891)
838
+ 802 train 7.530683 (lr=2.0819e-04) (hash(x)=7969325)
839
+ 803 train 7.530565 (lr=2.0698e-04) (hash(x)=5860821)
840
+ 804 train 7.546015 (lr=2.0578e-04) (hash(x)=6750848)
841
+ 805 train 7.488190 (lr=2.0458e-04) (hash(x)=5674826)
842
+ 806 train 7.278167 (lr=2.0338e-04) (hash(x)=5529163)
843
+ 807 train 7.691673 (lr=2.0219e-04) (hash(x)=7774109)
844
+ 808 train 7.488771 (lr=2.0100e-04) (hash(x)=6762509)
845
+ 809 train 7.575022 (lr=1.9982e-04) (hash(x)=5466449)
846
+ 810 train 7.445441 (lr=1.9864e-04) (hash(x)=7046935)
847
+ 811 train 8.301371 (lr=1.9747e-04) (hash(x)=7997664)
848
+ 812 train 7.602886 (lr=1.9630e-04) (hash(x)=6882594)
849
+ 813 train 7.765966 (lr=1.9514e-04) (hash(x)=7006517)
850
+ 814 train 7.606372 (lr=1.9398e-04) (hash(x)=6832530)
851
+ 815 train 7.468359 (lr=1.9283e-04) (hash(x)=6576749)
852
+ 816 train 7.582515 (lr=1.9168e-04) (hash(x)=8276629)
853
+ 817 train 7.364508 (lr=1.9053e-04) (hash(x)=6896198)
854
+ 818 train 7.592086 (lr=1.8939e-04) (hash(x)=5829252)
855
+ 819 train 7.723898 (lr=1.8826e-04) (hash(x)=7266655)
856
+ 820 train 7.456199 (lr=1.8713e-04) (hash(x)=6015975)
857
+ 821 train 7.446023 (lr=1.8600e-04) (hash(x)=5696696)
858
+ 822 train 7.469908 (lr=1.8488e-04) (hash(x)=5411666)
859
+ 823 train 7.696084 (lr=1.8377e-04) (hash(x)=7072404)
860
+ 824 train 7.337298 (lr=1.8266e-04) (hash(x)=4910095)
861
+ 825 train 7.485096 (lr=1.8155e-04) (hash(x)=6590657)
862
+ 826 train 7.421481 (lr=1.8045e-04) (hash(x)=7665574)
863
+ 827 train 7.468274 (lr=1.7936e-04) (hash(x)=6626459)
864
+ 828 train 7.508063 (lr=1.7827e-04) (hash(x)=6731971)
865
+ 829 train 7.491275 (lr=1.7718e-04) (hash(x)=5644977)
866
+ 830 train 7.458401 (lr=1.7610e-04) (hash(x)=6238654)
867
+ 831 train 7.354673 (lr=1.7503e-04) (hash(x)=6556025)
868
+ 832 train 7.675327 (lr=1.7396e-04) (hash(x)=7175563)
869
+ 833 train 7.553465 (lr=1.7289e-04) (hash(x)=6407492)
870
+ 834 train 7.382125 (lr=1.7183e-04) (hash(x)=4934335)
871
+ 835 train 7.398152 (lr=1.7078e-04) (hash(x)=5942129)
872
+ 836 train 7.428932 (lr=1.6973e-04) (hash(x)=5995643)
873
+ 837 train 7.476458 (lr=1.6869e-04) (hash(x)=6012163)
874
+ 838 train 7.603671 (lr=1.6765e-04) (hash(x)=8464831)
875
+ 839 train 7.831793 (lr=1.6661e-04) (hash(x)=7325027)
876
+ 840 train 7.702579 (lr=1.6559e-04) (hash(x)=6785865)
877
+ 841 train 7.612064 (lr=1.6456e-04) (hash(x)=4425520)
878
+ 842 train 7.388421 (lr=1.6355e-04) (hash(x)=5388267)
879
+ 843 train 7.638680 (lr=1.6253e-04) (hash(x)=7322467)
880
+ 844 train 7.435736 (lr=1.6153e-04) (hash(x)=6681766)
881
+ 845 train 7.500136 (lr=1.6053e-04) (hash(x)=7482800)
882
+ 846 train 7.464111 (lr=1.5953e-04) (hash(x)=5554493)
883
+ 847 train 7.653424 (lr=1.5854e-04) (hash(x)=6373412)
884
+ 848 train 7.085855 (lr=1.5756e-04) (hash(x)=5117517)
885
+ 849 train 7.330677 (lr=1.5658e-04) (hash(x)=6981426)
886
+ 850 val loss 7.4498
887
+ 850 val perplexity 1719.4774
888
+ 850 train 7.297373 (lr=1.5560e-04) (hash(x)=6886188)
889
+ 851 train 7.487570 (lr=1.5463e-04) (hash(x)=7332255)
890
+ 852 train 7.524590 (lr=1.5367e-04) (hash(x)=6172042)
891
+ 853 train 7.680242 (lr=1.5271e-04) (hash(x)=5930894)
892
+ 854 train 7.713052 (lr=1.5176e-04) (hash(x)=7448958)
893
+ 855 train 7.523213 (lr=1.5082e-04) (hash(x)=5262868)
894
+ 856 train 7.376454 (lr=1.4988e-04) (hash(x)=5558427)
895
+ 857 train 7.250167 (lr=1.4894e-04) (hash(x)=5585769)
896
+ 858 train 7.159152 (lr=1.4801e-04) (hash(x)=5838081)
897
+ 859 train 7.315042 (lr=1.4709e-04) (hash(x)=5688247)
898
+ 860 train 7.451667 (lr=1.4617e-04) (hash(x)=5162020)
899
+ 861 train 7.632195 (lr=1.4526e-04) (hash(x)=7462079)
900
+ 862 train 7.447641 (lr=1.4435e-04) (hash(x)=6516108)
901
+ 863 train 7.246408 (lr=1.4345e-04) (hash(x)=8055563)
902
+ 864 train 7.468148 (lr=1.4255e-04) (hash(x)=6271901)
903
+ 865 train 7.500250 (lr=1.4166e-04) (hash(x)=6221701)
904
+ 866 train 7.509865 (lr=1.4078e-04) (hash(x)=5772861)
905
+ 867 train 7.505138 (lr=1.3990e-04) (hash(x)=5352405)
906
+ 868 train 7.506035 (lr=1.3903e-04) (hash(x)=6111630)
907
+ 869 train 7.540511 (lr=1.3816e-04) (hash(x)=6730666)
908
+ 870 train 7.729849 (lr=1.3730e-04) (hash(x)=8698315)
909
+ 871 train 7.404297 (lr=1.3645e-04) (hash(x)=5932790)
910
+ 872 train 7.785899 (lr=1.3560e-04) (hash(x)=8099892)
911
+ 873 train 7.407391 (lr=1.3476e-04) (hash(x)=6223114)
912
+ 874 train 7.567327 (lr=1.3392e-04) (hash(x)=5798363)
913
+ 875 train 7.373255 (lr=1.3309e-04) (hash(x)=6249312)
914
+ 876 train 7.506389 (lr=1.3227e-04) (hash(x)=6929692)
915
+ 877 train 7.612144 (lr=1.3145e-04) (hash(x)=7242827)
916
+ 878 train 7.389362 (lr=1.3063e-04) (hash(x)=6332123)
917
+ 879 train 7.294970 (lr=1.2983e-04) (hash(x)=5680154)
918
+ 880 train 7.667989 (lr=1.2903e-04) (hash(x)=6352331)
919
+ 881 train 7.482324 (lr=1.2823e-04) (hash(x)=6332332)
920
+ 882 train 7.527276 (lr=1.2744e-04) (hash(x)=6180674)
921
+ 883 train 7.486145 (lr=1.2666e-04) (hash(x)=8156280)
922
+ 884 train 7.518857 (lr=1.2588e-04) (hash(x)=5874247)
923
+ 885 train 7.437901 (lr=1.2511e-04) (hash(x)=6659781)
924
+ 886 train 7.559820 (lr=1.2435e-04) (hash(x)=5780147)
925
+ 887 train 7.379938 (lr=1.2359e-04) (hash(x)=5914217)
926
+ 888 train 7.246047 (lr=1.2284e-04) (hash(x)=5411762)
927
+ 889 train 7.543199 (lr=1.2209e-04) (hash(x)=6714110)
928
+ 890 train 7.189143 (lr=1.2135e-04) (hash(x)=5999685)
929
+ 891 train 7.384521 (lr=1.2062e-04) (hash(x)=7120215)
930
+ 892 train 7.655559 (lr=1.1989e-04) (hash(x)=6546587)
931
+ 893 train 7.437112 (lr=1.1917e-04) (hash(x)=6593413)
932
+ 894 train 7.529776 (lr=1.1845e-04) (hash(x)=6369927)
933
+ 895 train 7.441356 (lr=1.1774e-04) (hash(x)=6424128)
934
+ 896 train 7.418300 (lr=1.1704e-04) (hash(x)=5736158)
935
+ 897 train 7.233034 (lr=1.1634e-04) (hash(x)=5903043)
936
+ 898 train 7.323412 (lr=1.1565e-04) (hash(x)=4419128)
937
+ 899 train 7.160180 (lr=1.1497e-04) (hash(x)=4390027)
938
+ 900 val loss 7.4288
939
+ 900 val perplexity 1683.8197
940
+ 900 train 7.362626 (lr=1.1429e-04) (hash(x)=6728135)
941
+ 901 train 7.513760 (lr=1.1362e-04) (hash(x)=6945760)
942
+ 902 train 7.471305 (lr=1.1295e-04) (hash(x)=6081534)
943
+ 903 train 7.517815 (lr=1.1229e-04) (hash(x)=7804089)
944
+ 904 train 7.571086 (lr=1.1164e-04) (hash(x)=6225832)
945
+ 905 train 7.610134 (lr=1.1099e-04) (hash(x)=6273417)
946
+ 906 train 7.659355 (lr=1.1036e-04) (hash(x)=7775633)
947
+ 907 train 7.620094 (lr=1.0972e-04) (hash(x)=7130267)
948
+ 908 train 7.416354 (lr=1.0909e-04) (hash(x)=6554076)
949
+ 909 train 7.513559 (lr=1.0847e-04) (hash(x)=6140697)
950
+ 910 train 7.335847 (lr=1.0786e-04) (hash(x)=6128181)
951
+ 911 train 7.436421 (lr=1.0725e-04) (hash(x)=6490149)
952
+ 912 train 7.031314 (lr=1.0665e-04) (hash(x)=5422426)
953
+ 913 train 7.456882 (lr=1.0606e-04) (hash(x)=4847733)
954
+ 914 train 7.447791 (lr=1.0547e-04) (hash(x)=6349747)
955
+ 915 train 7.729845 (lr=1.0489e-04) (hash(x)=9276378)
956
+ 916 train 7.343298 (lr=1.0431e-04) (hash(x)=6869524)
957
+ 917 train 7.381476 (lr=1.0374e-04) (hash(x)=7082053)
958
+ 918 train 7.421865 (lr=1.0318e-04) (hash(x)=7799351)
959
+ 919 train 7.616323 (lr=1.0262e-04) (hash(x)=6283540)
960
+ 920 train 7.387016 (lr=1.0207e-04) (hash(x)=6660860)
961
+ 921 train 7.328838 (lr=1.0153e-04) (hash(x)=5200956)
962
+ 922 train 7.414082 (lr=1.0099e-04) (hash(x)=5653241)
963
+ 923 train 7.583707 (lr=1.0046e-04) (hash(x)=5781680)
964
+ 924 train 7.558781 (lr=9.9941e-05) (hash(x)=6146666)
965
+ 925 train 7.397439 (lr=9.9425e-05) (hash(x)=8652907)
966
+ 926 train 7.192469 (lr=9.8915e-05) (hash(x)=5311546)
967
+ 927 train 7.351357 (lr=9.8411e-05) (hash(x)=5728337)
968
+ 928 train 7.318535 (lr=9.7915e-05) (hash(x)=5413240)
969
+ 929 train 7.430328 (lr=9.7424e-05) (hash(x)=5832401)
970
+ 930 train 7.211864 (lr=9.6941e-05) (hash(x)=7420230)
971
+ 931 train 7.519371 (lr=9.6464e-05) (hash(x)=6278202)
972
+ 932 train 7.390079 (lr=9.5994e-05) (hash(x)=6189873)
973
+ 933 train 7.499774 (lr=9.5530e-05) (hash(x)=7403868)
974
+ 934 train 7.363681 (lr=9.5073e-05) (hash(x)=6867459)
975
+ 935 train 7.564868 (lr=9.4623e-05) (hash(x)=7400272)
976
+ 936 train 7.616142 (lr=9.4180e-05) (hash(x)=7936275)
977
+ 937 train 7.311508 (lr=9.3743e-05) (hash(x)=5943413)
978
+ 938 train 7.528198 (lr=9.3313e-05) (hash(x)=5442166)
979
+ 939 train 7.419097 (lr=9.2889e-05) (hash(x)=5795128)
980
+ 940 train 7.617681 (lr=9.2473e-05) (hash(x)=6831680)
981
+ 941 train 7.303300 (lr=9.2063e-05) (hash(x)=5519980)
982
+ 942 train 7.332280 (lr=9.1659e-05) (hash(x)=6765461)
983
+ 943 train 7.305336 (lr=9.1263e-05) (hash(x)=6473432)
984
+ 944 train 7.273232 (lr=9.0873e-05) (hash(x)=6490395)
985
+ 945 train 7.485599 (lr=9.0490e-05) (hash(x)=5484669)
986
+ 946 train 7.463054 (lr=9.0114e-05) (hash(x)=7278970)
987
+ 947 train 7.448289 (lr=8.9744e-05) (hash(x)=5201502)
988
+ 948 train 7.675272 (lr=8.9382e-05) (hash(x)=7277915)
989
+ 949 train 7.396924 (lr=8.9026e-05) (hash(x)=7227057)
990
+ 950 val loss 7.4159
991
+ 950 val perplexity 1662.2073
992
+ 950 train 7.626300 (lr=8.8677e-05) (hash(x)=8178996)
993
+ 951 train 7.771703 (lr=8.8335e-05) (hash(x)=6922646)
994
+ 952 train 7.278928 (lr=8.7999e-05) (hash(x)=5941772)
995
+ 953 train 7.254891 (lr=8.7670e-05) (hash(x)=6056673)
996
+ 954 train 7.648520 (lr=8.7349e-05) (hash(x)=6912178)
997
+ 955 train 7.186590 (lr=8.7034e-05) (hash(x)=5484641)
998
+ 956 train 7.525898 (lr=8.6725e-05) (hash(x)=5121117)
999
+ 957 train 7.557370 (lr=8.6424e-05) (hash(x)=7100865)
1000
+ 958 train 7.256378 (lr=8.6130e-05) (hash(x)=5334280)
1001
+ 959 train 7.362887 (lr=8.5842e-05) (hash(x)=7277141)
1002
+ 960 train 7.307778 (lr=8.5561e-05) (hash(x)=6401960)
1003
+ 961 train 7.387640 (lr=8.5287e-05) (hash(x)=6407603)
1004
+ 962 train 7.527298 (lr=8.5020e-05) (hash(x)=5786089)
1005
+ 963 train 7.405932 (lr=8.4760e-05) (hash(x)=5604614)
1006
+ 964 train 7.417370 (lr=8.4507e-05) (hash(x)=5454287)
1007
+ 965 train 7.542350 (lr=8.4260e-05) (hash(x)=6200612)
1008
+ 966 train 7.499067 (lr=8.4021e-05) (hash(x)=5691759)
1009
+ 967 train 7.459194 (lr=8.3788e-05) (hash(x)=7443124)
1010
+ 968 train 7.500760 (lr=8.3563e-05) (hash(x)=6711366)
1011
+ 969 train 7.407757 (lr=8.3344e-05) (hash(x)=6728764)
1012
+ 970 train 7.693987 (lr=8.3132e-05) (hash(x)=6870240)
1013
+ 971 train 7.399839 (lr=8.2927e-05) (hash(x)=5601968)
1014
+ 972 train 7.162122 (lr=8.2729e-05) (hash(x)=5311504)
1015
+ 973 train 7.457762 (lr=8.2537e-05) (hash(x)=6877761)
1016
+ 974 train 7.511388 (lr=8.2353e-05) (hash(x)=6457033)
1017
+ 975 train 7.453982 (lr=8.2176e-05) (hash(x)=5632735)
1018
+ 976 train 7.637141 (lr=8.2005e-05) (hash(x)=6024880)
1019
+ 977 train 7.358840 (lr=8.1842e-05) (hash(x)=5658845)
1020
+ 978 train 7.237310 (lr=8.1685e-05) (hash(x)=6089461)
1021
+ 979 train 7.310323 (lr=8.1536e-05) (hash(x)=5594810)
1022
+ 980 train 7.092809 (lr=8.1393e-05) (hash(x)=4667026)
1023
+ 981 train 6.967785 (lr=8.1257e-05) (hash(x)=4738523)
1024
+ 982 train 7.273063 (lr=8.1128e-05) (hash(x)=7451876)
1025
+ 983 train 7.529049 (lr=8.1007e-05) (hash(x)=5881015)
1026
+ 984 train 7.547577 (lr=8.0892e-05) (hash(x)=6255031)
1027
+ 985 train 7.360570 (lr=8.0784e-05) (hash(x)=5263325)
1028
+ 986 train 7.339565 (lr=8.0683e-05) (hash(x)=6654585)
1029
+ 987 train 7.397566 (lr=8.0589e-05) (hash(x)=6865361)
1030
+ 988 train 7.496890 (lr=8.0502e-05) (hash(x)=6978986)
1031
+ 989 train 7.604034 (lr=8.0422e-05) (hash(x)=5644619)
1032
+ 990 train 7.427780 (lr=8.0348e-05) (hash(x)=6551023)
1033
+ 991 train 7.371108 (lr=8.0282e-05) (hash(x)=5809411)
1034
+ 992 train 7.454008 (lr=8.0223e-05) (hash(x)=5501837)
1035
+ 993 train 7.474544 (lr=8.0171e-05) (hash(x)=7044374)
1036
+ 994 train 7.448763 (lr=8.0125e-05) (hash(x)=6526130)
1037
+ 995 train 7.531295 (lr=8.0087e-05) (hash(x)=6293025)
1038
+ 996 train 7.533747 (lr=8.0056e-05) (hash(x)=6295768)
1039
+ 997 train 7.332946 (lr=8.0031e-05) (hash(x)=5189023)
1040
+ 998 train 7.145046 (lr=8.0014e-05) (hash(x)=6255467)
1041
+ 999 val loss 7.4143
1042
+ 999 val perplexity 1659.6246
1043
+ 999 train 7.539111 (lr=8.0003e-05) (hash(x)=7232964)
lr8e-4_total_batch_size5120_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd75dc9859521f2d4e2bc7b64b7cf39fcbee9a8f70ec03967dbf990202f95917
3
+ size 36163010
lr8e-4_total_batch_size5120_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9432372977e77a38c802c3205c55f7c21c753814ba15fa0c7c56c0ae1883f81
3
+ size 70764358