andrew-healey commited on
Commit
8cc4804
·
verified ·
1 Parent(s): ea95799

Upload folder using huggingface_hub

Browse files
lr3e-3_total_batch_size5120_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr3e-3_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": 0.003, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr3e-3_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
lr3e-3_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=1.0490e-05) (hash(x)=6154740)
5
+ 1 train 11.691827 (lr=2.0979e-05) (hash(x)=6605789)
6
+ 2 train 11.621720 (lr=3.1469e-05) (hash(x)=6696707)
7
+ 3 train 11.607746 (lr=4.1958e-05) (hash(x)=5113283)
8
+ 4 train 11.421369 (lr=5.2448e-05) (hash(x)=5607069)
9
+ 5 train 11.339355 (lr=6.2937e-05) (hash(x)=7007748)
10
+ 6 train 11.180422 (lr=7.3427e-05) (hash(x)=7643299)
11
+ 7 train 11.067919 (lr=8.3916e-05) (hash(x)=5534142)
12
+ 8 train 10.951596 (lr=9.4406e-05) (hash(x)=6539293)
13
+ 9 train 10.786669 (lr=1.0490e-04) (hash(x)=5974274)
14
+ 10 train 10.697496 (lr=1.1538e-04) (hash(x)=7199085)
15
+ 11 train 10.618728 (lr=1.2587e-04) (hash(x)=6763231)
16
+ 12 train 10.429303 (lr=1.3636e-04) (hash(x)=6720398)
17
+ 13 train 10.407563 (lr=1.4685e-04) (hash(x)=6706513)
18
+ 14 train 10.407973 (lr=1.5734e-04) (hash(x)=5727925)
19
+ 15 train 10.267708 (lr=1.6783e-04) (hash(x)=6328012)
20
+ 16 train 10.145553 (lr=1.7832e-04) (hash(x)=5812722)
21
+ 17 train 10.085203 (lr=1.8881e-04) (hash(x)=6183012)
22
+ 18 train 10.137742 (lr=1.9930e-04) (hash(x)=5952539)
23
+ 19 train 10.063640 (lr=2.0979e-04) (hash(x)=7052161)
24
+ 20 train 9.832780 (lr=2.2028e-04) (hash(x)=5810871)
25
+ 21 train 10.021389 (lr=2.3077e-04) (hash(x)=6586466)
26
+ 22 train 9.771036 (lr=2.4126e-04) (hash(x)=5061781)
27
+ 23 train 9.809976 (lr=2.5175e-04) (hash(x)=5556098)
28
+ 24 train 9.816788 (lr=2.6224e-04) (hash(x)=7268535)
29
+ 25 train 9.761164 (lr=2.7273e-04) (hash(x)=8155572)
30
+ 26 train 9.586716 (lr=2.8322e-04) (hash(x)=5019590)
31
+ 27 train 9.377479 (lr=2.9371e-04) (hash(x)=4696943)
32
+ 28 train 9.315301 (lr=3.0420e-04) (hash(x)=5984010)
33
+ 29 train 9.557399 (lr=3.1469e-04) (hash(x)=7649389)
34
+ 30 train 9.164330 (lr=3.2517e-04) (hash(x)=6332817)
35
+ 31 train 9.088143 (lr=3.3566e-04) (hash(x)=6219138)
36
+ 32 train 9.192779 (lr=3.4615e-04) (hash(x)=6895681)
37
+ 33 train 9.186225 (lr=3.5664e-04) (hash(x)=7075794)
38
+ 34 train 8.914955 (lr=3.6713e-04) (hash(x)=5024163)
39
+ 35 train 8.863305 (lr=3.7762e-04) (hash(x)=6049878)
40
+ 36 train 8.897887 (lr=3.8811e-04) (hash(x)=4774943)
41
+ 37 train 8.919571 (lr=3.9860e-04) (hash(x)=6385927)
42
+ 38 train 8.636589 (lr=4.0909e-04) (hash(x)=5070217)
43
+ 39 train 8.484549 (lr=4.1958e-04) (hash(x)=5491972)
44
+ 40 train 8.632951 (lr=4.3007e-04) (hash(x)=5695751)
45
+ 41 train 8.568641 (lr=4.4056e-04) (hash(x)=6061420)
46
+ 42 train 8.505939 (lr=4.5105e-04) (hash(x)=7836710)
47
+ 43 train 8.561627 (lr=4.6154e-04) (hash(x)=8075458)
48
+ 44 train 8.306105 (lr=4.7203e-04) (hash(x)=7190584)
49
+ 45 train 8.509636 (lr=4.8252e-04) (hash(x)=7238769)
50
+ 46 train 8.818381 (lr=4.9301e-04) (hash(x)=4931205)
51
+ 47 train 8.195147 (lr=5.0350e-04) (hash(x)=5917741)
52
+ 48 train 8.269698 (lr=5.1399e-04) (hash(x)=5530785)
53
+ 49 train 8.153329 (lr=5.2448e-04) (hash(x)=5430605)
54
+ 50 val loss 8.1711
55
+ 50 val perplexity 3537.2180
56
+ 50 train 8.233984 (lr=5.3497e-04) (hash(x)=6056774)
57
+ 51 train 7.990573 (lr=5.4545e-04) (hash(x)=5750403)
58
+ 52 train 8.081745 (lr=5.5594e-04) (hash(x)=6187497)
59
+ 53 train 7.982611 (lr=5.6643e-04) (hash(x)=4792640)
60
+ 54 train 8.181334 (lr=5.7692e-04) (hash(x)=6868430)
61
+ 55 train 7.946690 (lr=5.8741e-04) (hash(x)=6112458)
62
+ 56 train 8.044004 (lr=5.9790e-04) (hash(x)=5523601)
63
+ 57 train 7.826697 (lr=6.0839e-04) (hash(x)=5801122)
64
+ 58 train 8.098812 (lr=6.1888e-04) (hash(x)=6826295)
65
+ 59 train 8.363143 (lr=6.2937e-04) (hash(x)=5806490)
66
+ 60 train 8.045688 (lr=6.3986e-04) (hash(x)=6223100)
67
+ 61 train 7.857108 (lr=6.5035e-04) (hash(x)=5754140)
68
+ 62 train 8.028936 (lr=6.6084e-04) (hash(x)=6675367)
69
+ 63 train 7.914736 (lr=6.7133e-04) (hash(x)=6568379)
70
+ 64 train 7.872281 (lr=6.8182e-04) (hash(x)=5648233)
71
+ 65 train 8.284698 (lr=6.9231e-04) (hash(x)=7134932)
72
+ 66 train 7.904185 (lr=7.0280e-04) (hash(x)=5242536)
73
+ 67 train 8.122234 (lr=7.1329e-04) (hash(x)=7212403)
74
+ 68 train 7.938879 (lr=7.2378e-04) (hash(x)=5342967)
75
+ 69 train 8.147705 (lr=7.3427e-04) (hash(x)=7720703)
76
+ 70 train 8.298388 (lr=7.4476e-04) (hash(x)=7351337)
77
+ 71 train 8.144358 (lr=7.5524e-04) (hash(x)=7386082)
78
+ 72 train 7.983346 (lr=7.6573e-04) (hash(x)=6931318)
79
+ 73 train 8.280849 (lr=7.7622e-04) (hash(x)=6848513)
80
+ 74 train 10.241465 (lr=7.8671e-04) (hash(x)=7714274)
81
+ 75 train 8.279107 (lr=7.9720e-04) (hash(x)=6359743)
82
+ 76 train 8.106997 (lr=8.0769e-04) (hash(x)=6801650)
83
+ 77 train 8.149957 (lr=8.1818e-04) (hash(x)=7536538)
84
+ 78 train 8.144439 (lr=8.2867e-04) (hash(x)=6957410)
85
+ 79 train 8.144677 (lr=8.3916e-04) (hash(x)=6151674)
86
+ 80 train 7.963267 (lr=8.4965e-04) (hash(x)=6033844)
87
+ 81 train 7.787290 (lr=8.6014e-04) (hash(x)=6317099)
88
+ 82 train 7.740457 (lr=8.7063e-04) (hash(x)=5357844)
89
+ 83 train 7.960593 (lr=8.8112e-04) (hash(x)=7168704)
90
+ 84 train 8.322322 (lr=8.9161e-04) (hash(x)=7157938)
91
+ 85 train 7.872025 (lr=9.0210e-04) (hash(x)=6736092)
92
+ 86 train 7.770463 (lr=9.1259e-04) (hash(x)=5383012)
93
+ 87 train 8.072406 (lr=9.2308e-04) (hash(x)=7859060)
94
+ 88 train 7.966494 (lr=9.3357e-04) (hash(x)=5869288)
95
+ 89 train 7.979216 (lr=9.4406e-04) (hash(x)=6958223)
96
+ 90 train 7.794937 (lr=9.5455e-04) (hash(x)=6731344)
97
+ 91 train 7.992211 (lr=9.6503e-04) (hash(x)=6976239)
98
+ 92 train 7.765761 (lr=9.7552e-04) (hash(x)=7769441)
99
+ 93 train 7.916177 (lr=9.8601e-04) (hash(x)=6956845)
100
+ 94 train 8.041010 (lr=9.9650e-04) (hash(x)=6638037)
101
+ 95 train 7.594099 (lr=1.0070e-03) (hash(x)=6391506)
102
+ 96 train 8.140839 (lr=1.0175e-03) (hash(x)=7319928)
103
+ 97 train 8.032586 (lr=1.0280e-03) (hash(x)=6113701)
104
+ 98 train 7.730045 (lr=1.0385e-03) (hash(x)=6246453)
105
+ 99 train 7.858302 (lr=1.0490e-03) (hash(x)=5245999)
106
+ 100 val loss 7.9336
107
+ 100 val perplexity 2789.5491
108
+ 100 train 8.010086 (lr=1.0594e-03) (hash(x)=6227256)
109
+ 101 train 7.900714 (lr=1.0699e-03) (hash(x)=5774189)
110
+ 102 train 7.812904 (lr=1.0804e-03) (hash(x)=5588220)
111
+ 103 train 7.742281 (lr=1.0909e-03) (hash(x)=5985675)
112
+ 104 train 7.711298 (lr=1.1014e-03) (hash(x)=6259214)
113
+ 105 train 7.976234 (lr=1.1119e-03) (hash(x)=6451776)
114
+ 106 train 7.601935 (lr=1.1224e-03) (hash(x)=4940406)
115
+ 107 train 7.721811 (lr=1.1329e-03) (hash(x)=4355733)
116
+ 108 train 7.741394 (lr=1.1434e-03) (hash(x)=6186065)
117
+ 109 train 7.959008 (lr=1.1538e-03) (hash(x)=6669188)
118
+ 110 train 9.245372 (lr=1.1643e-03) (hash(x)=7118997)
119
+ 111 train 8.363277 (lr=1.1748e-03) (hash(x)=6984772)
120
+ 112 train 7.827683 (lr=1.1853e-03) (hash(x)=6472877)
121
+ 113 train 8.078576 (lr=1.1958e-03) (hash(x)=7440580)
122
+ 114 train 8.033119 (lr=1.2063e-03) (hash(x)=6378428)
123
+ 115 train 8.179004 (lr=1.2168e-03) (hash(x)=6244096)
124
+ 116 train 8.184355 (lr=1.2273e-03) (hash(x)=9100578)
125
+ 117 train 8.414863 (lr=1.2378e-03) (hash(x)=8381560)
126
+ 118 train 8.180249 (lr=1.2483e-03) (hash(x)=6066072)
127
+ 119 train 7.864053 (lr=1.2587e-03) (hash(x)=6603717)
128
+ 120 train 7.689608 (lr=1.2692e-03) (hash(x)=5138438)
129
+ 121 train 7.650041 (lr=1.2797e-03) (hash(x)=6437516)
130
+ 122 train 7.890066 (lr=1.2902e-03) (hash(x)=5743807)
131
+ 123 train 7.657547 (lr=1.3007e-03) (hash(x)=5669522)
132
+ 124 train 7.944245 (lr=1.3112e-03) (hash(x)=5963623)
133
+ 125 train 7.684509 (lr=1.3217e-03) (hash(x)=5505889)
134
+ 126 train 8.141639 (lr=1.3322e-03) (hash(x)=7607512)
135
+ 127 train 7.905875 (lr=1.3427e-03) (hash(x)=6190579)
136
+ 128 train 8.103085 (lr=1.3531e-03) (hash(x)=7444834)
137
+ 129 train 7.945289 (lr=1.3636e-03) (hash(x)=5153196)
138
+ 130 train 7.926319 (lr=1.3741e-03) (hash(x)=6495326)
139
+ 131 train 7.946622 (lr=1.3846e-03) (hash(x)=4912983)
140
+ 132 train 8.303801 (lr=1.3951e-03) (hash(x)=7094046)
141
+ 133 train 7.751832 (lr=1.4056e-03) (hash(x)=5109125)
142
+ 134 train 7.796208 (lr=1.4161e-03) (hash(x)=7119136)
143
+ 135 train 7.842940 (lr=1.4266e-03) (hash(x)=7276303)
144
+ 136 train 8.096292 (lr=1.4371e-03) (hash(x)=9147123)
145
+ 137 train 8.034124 (lr=1.4476e-03) (hash(x)=6645716)
146
+ 138 train 7.996403 (lr=1.4580e-03) (hash(x)=6347875)
147
+ 139 train 7.765953 (lr=1.4685e-03) (hash(x)=5479947)
148
+ 140 train 7.904454 (lr=1.4790e-03) (hash(x)=5832595)
149
+ 141 train 8.199355 (lr=1.4895e-03) (hash(x)=7850360)
150
+ 142 train 8.590281 (lr=1.5000e-03) (hash(x)=8499683)
151
+ 143 train 8.114748 (lr=1.5105e-03) (hash(x)=6630378)
152
+ 144 train 8.059767 (lr=1.5210e-03) (hash(x)=5340846)
153
+ 145 train 7.880220 (lr=1.5315e-03) (hash(x)=6295358)
154
+ 146 train 7.928468 (lr=1.5420e-03) (hash(x)=5974729)
155
+ 147 train 8.522675 (lr=1.5524e-03) (hash(x)=8097010)
156
+ 148 train 8.151880 (lr=1.5629e-03) (hash(x)=6681250)
157
+ 149 train 7.607543 (lr=1.5734e-03) (hash(x)=3958188)
158
+ 150 val loss 7.8539
159
+ 150 val perplexity 2575.7957
160
+ 150 train 7.811275 (lr=1.5839e-03) (hash(x)=5975539)
161
+ 151 train 9.865631 (lr=1.5944e-03) (hash(x)=9989744)
162
+ 152 train 8.762166 (lr=1.6049e-03) (hash(x)=6748487)
163
+ 153 train 7.909334 (lr=1.6154e-03) (hash(x)=6343832)
164
+ 154 train 8.000429 (lr=1.6259e-03) (hash(x)=5688949)
165
+ 155 train 8.201068 (lr=1.6364e-03) (hash(x)=6694932)
166
+ 156 train 7.870391 (lr=1.6469e-03) (hash(x)=5872835)
167
+ 157 train 7.935297 (lr=1.6573e-03) (hash(x)=6429137)
168
+ 158 train 8.130231 (lr=1.6678e-03) (hash(x)=6999390)
169
+ 159 train 7.985057 (lr=1.6783e-03) (hash(x)=6317894)
170
+ 160 train 7.650081 (lr=1.6888e-03) (hash(x)=5813657)
171
+ 161 train 7.728964 (lr=1.6993e-03) (hash(x)=5165595)
172
+ 162 train 7.672716 (lr=1.7098e-03) (hash(x)=5682633)
173
+ 163 train 8.043719 (lr=1.7203e-03) (hash(x)=5046327)
174
+ 164 train 8.078794 (lr=1.7308e-03) (hash(x)=6195808)
175
+ 165 train 7.942073 (lr=1.7413e-03) (hash(x)=7060363)
176
+ 166 train 7.930969 (lr=1.7517e-03) (hash(x)=5837126)
177
+ 167 train 7.882886 (lr=1.7622e-03) (hash(x)=7947703)
178
+ 168 train 7.473785 (lr=1.7727e-03) (hash(x)=5277865)
179
+ 169 train 7.762346 (lr=1.7832e-03) (hash(x)=5971240)
180
+ 170 train 7.692909 (lr=1.7937e-03) (hash(x)=6590408)
181
+ 171 train 7.849571 (lr=1.8042e-03) (hash(x)=5916068)
182
+ 172 train 7.791912 (lr=1.8147e-03) (hash(x)=6069727)
183
+ 173 train 7.826770 (lr=1.8252e-03) (hash(x)=7426277)
184
+ 174 train 7.992832 (lr=1.8357e-03) (hash(x)=5356513)
185
+ 175 train 7.784763 (lr=1.8462e-03) (hash(x)=5777498)
186
+ 176 train 8.048485 (lr=1.8566e-03) (hash(x)=5989756)
187
+ 177 train 7.738062 (lr=1.8671e-03) (hash(x)=6403331)
188
+ 178 train 7.760424 (lr=1.8776e-03) (hash(x)=7650667)
189
+ 179 train 7.908355 (lr=1.8881e-03) (hash(x)=5753092)
190
+ 180 train 7.733218 (lr=1.8986e-03) (hash(x)=5751457)
191
+ 181 train 7.570724 (lr=1.9091e-03) (hash(x)=5981486)
192
+ 182 train 7.573308 (lr=1.9196e-03) (hash(x)=5918229)
193
+ 183 train 7.777656 (lr=1.9301e-03) (hash(x)=7723226)
194
+ 184 train 7.859167 (lr=1.9406e-03) (hash(x)=6111721)
195
+ 185 train 7.600235 (lr=1.9510e-03) (hash(x)=6069558)
196
+ 186 train 7.887666 (lr=1.9615e-03) (hash(x)=5626336)
197
+ 187 train 8.188937 (lr=1.9720e-03) (hash(x)=7007016)
198
+ 188 train 7.693271 (lr=1.9825e-03) (hash(x)=5883013)
199
+ 189 train 7.518561 (lr=1.9930e-03) (hash(x)=5027958)
200
+ 190 train 7.823518 (lr=2.0035e-03) (hash(x)=5474948)
201
+ 191 train 7.635974 (lr=2.0140e-03) (hash(x)=6491229)
202
+ 192 train 7.733161 (lr=2.0245e-03) (hash(x)=5518341)
203
+ 193 train 7.825987 (lr=2.0350e-03) (hash(x)=7036653)
204
+ 194 train 7.823652 (lr=2.0455e-03) (hash(x)=5527616)
205
+ 195 train 8.153146 (lr=2.0559e-03) (hash(x)=6390752)
206
+ 196 train 7.747933 (lr=2.0664e-03) (hash(x)=6032557)
207
+ 197 train 8.527084 (lr=2.0769e-03) (hash(x)=7758167)
208
+ 198 train 8.280541 (lr=2.0874e-03) (hash(x)=7968070)
209
+ 199 train 7.898184 (lr=2.0979e-03) (hash(x)=6019389)
210
+ 200 val loss 7.8510
211
+ 200 val perplexity 2568.3474
212
+ 200 train 7.936243 (lr=2.1084e-03) (hash(x)=7276744)
213
+ 201 train 7.929254 (lr=2.1189e-03) (hash(x)=6503191)
214
+ 202 train 7.784602 (lr=2.1294e-03) (hash(x)=6006880)
215
+ 203 train 7.987221 (lr=2.1399e-03) (hash(x)=7662067)
216
+ 204 train 7.746620 (lr=2.1503e-03) (hash(x)=6297345)
217
+ 205 train 8.002535 (lr=2.1608e-03) (hash(x)=7901992)
218
+ 206 train 8.062188 (lr=2.1713e-03) (hash(x)=6579950)
219
+ 207 train 7.647624 (lr=2.1818e-03) (hash(x)=4648609)
220
+ 208 train 7.845750 (lr=2.1923e-03) (hash(x)=6903216)
221
+ 209 train 7.698059 (lr=2.2028e-03) (hash(x)=5897288)
222
+ 210 train 7.794614 (lr=2.2133e-03) (hash(x)=7300160)
223
+ 211 train 7.602360 (lr=2.2238e-03) (hash(x)=4725966)
224
+ 212 train 7.431656 (lr=2.2343e-03) (hash(x)=5060060)
225
+ 213 train 8.043680 (lr=2.2448e-03) (hash(x)=6243442)
226
+ 214 train 7.658851 (lr=2.2552e-03) (hash(x)=5893816)
227
+ 215 train 7.693961 (lr=2.2657e-03) (hash(x)=5558355)
228
+ 216 train 7.823012 (lr=2.2762e-03) (hash(x)=4378747)
229
+ 217 train 7.687565 (lr=2.2867e-03) (hash(x)=5400407)
230
+ 218 train 7.977879 (lr=2.2972e-03) (hash(x)=6900554)
231
+ 219 train 7.866753 (lr=2.3077e-03) (hash(x)=6524933)
232
+ 220 train 7.761933 (lr=2.3182e-03) (hash(x)=5409944)
233
+ 221 train 7.889899 (lr=2.3287e-03) (hash(x)=5889724)
234
+ 222 train 7.776840 (lr=2.3392e-03) (hash(x)=4970496)
235
+ 223 train 7.565931 (lr=2.3497e-03) (hash(x)=6369326)
236
+ 224 train 7.656119 (lr=2.3601e-03) (hash(x)=6563975)
237
+ 225 train 7.867293 (lr=2.3706e-03) (hash(x)=5911906)
238
+ 226 train 7.955462 (lr=2.3811e-03) (hash(x)=6111462)
239
+ 227 train 7.936445 (lr=2.3916e-03) (hash(x)=6022625)
240
+ 228 train 7.731579 (lr=2.4021e-03) (hash(x)=5884663)
241
+ 229 train 7.838469 (lr=2.4126e-03) (hash(x)=6584810)
242
+ 230 train 7.800380 (lr=2.4231e-03) (hash(x)=5289998)
243
+ 231 train 7.515210 (lr=2.4336e-03) (hash(x)=4906853)
244
+ 232 train 7.588811 (lr=2.4441e-03) (hash(x)=5862071)
245
+ 233 train 7.235142 (lr=2.4545e-03) (hash(x)=4564564)
246
+ 234 train 7.556524 (lr=2.4650e-03) (hash(x)=5723971)
247
+ 235 train 7.689487 (lr=2.4755e-03) (hash(x)=5589269)
248
+ 236 train 7.661816 (lr=2.4860e-03) (hash(x)=5683368)
249
+ 237 train 7.889834 (lr=2.4965e-03) (hash(x)=6969180)
250
+ 238 train 7.786815 (lr=2.5070e-03) (hash(x)=5341421)
251
+ 239 train 7.544628 (lr=2.5175e-03) (hash(x)=6437376)
252
+ 240 train 8.034918 (lr=2.5280e-03) (hash(x)=7527007)
253
+ 241 train 7.868428 (lr=2.5385e-03) (hash(x)=7116258)
254
+ 242 train 7.894980 (lr=2.5490e-03) (hash(x)=6463930)
255
+ 243 train 7.867659 (lr=2.5594e-03) (hash(x)=7304892)
256
+ 244 train 8.072035 (lr=2.5699e-03) (hash(x)=5881808)
257
+ 245 train 7.793243 (lr=2.5804e-03) (hash(x)=7366721)
258
+ 246 train 7.757831 (lr=2.5909e-03) (hash(x)=6457521)
259
+ 247 train 8.104451 (lr=2.6014e-03) (hash(x)=7084093)
260
+ 248 train 7.788870 (lr=2.6119e-03) (hash(x)=6015300)
261
+ 249 train 7.991997 (lr=2.6224e-03) (hash(x)=7197652)
262
+ 250 val loss 7.7428
263
+ 250 val perplexity 2304.9429
264
+ 250 train 8.061850 (lr=2.6329e-03) (hash(x)=4830828)
265
+ 251 train 7.771574 (lr=2.6434e-03) (hash(x)=5908178)
266
+ 252 train 7.777815 (lr=2.6538e-03) (hash(x)=6299092)
267
+ 253 train 7.741354 (lr=2.6643e-03) (hash(x)=6349974)
268
+ 254 train 7.758904 (lr=2.6748e-03) (hash(x)=5641494)
269
+ 255 train 7.745541 (lr=2.6853e-03) (hash(x)=7048804)
270
+ 256 train 7.972313 (lr=2.6958e-03) (hash(x)=7035244)
271
+ 257 train 7.670757 (lr=2.7063e-03) (hash(x)=7909039)
272
+ 258 train 7.894810 (lr=2.7168e-03) (hash(x)=5470939)
273
+ 259 train 7.742216 (lr=2.7273e-03) (hash(x)=6085549)
274
+ 260 train 8.058550 (lr=2.7378e-03) (hash(x)=5882649)
275
+ 261 train 7.832091 (lr=2.7483e-03) (hash(x)=7463181)
276
+ 262 train 8.156913 (lr=2.7587e-03) (hash(x)=5215574)
277
+ 263 train 7.598020 (lr=2.7692e-03) (hash(x)=5752594)
278
+ 264 train 7.764878 (lr=2.7797e-03) (hash(x)=6416619)
279
+ 265 train 7.784012 (lr=2.7902e-03) (hash(x)=6114601)
280
+ 266 train 7.693649 (lr=2.8007e-03) (hash(x)=5646171)
281
+ 267 train 7.665197 (lr=2.8112e-03) (hash(x)=7662769)
282
+ 268 train 7.945850 (lr=2.8217e-03) (hash(x)=6394376)
283
+ 269 train 7.533813 (lr=2.8322e-03) (hash(x)=7485666)
284
+ 270 train 7.626507 (lr=2.8427e-03) (hash(x)=6636823)
285
+ 271 train 7.513787 (lr=2.8531e-03) (hash(x)=6393520)
286
+ 272 train 7.762655 (lr=2.8636e-03) (hash(x)=5235659)
287
+ 273 train 7.797278 (lr=2.8741e-03) (hash(x)=5586291)
288
+ 274 train 8.032333 (lr=2.8846e-03) (hash(x)=7034674)
289
+ 275 train 8.133555 (lr=2.8951e-03) (hash(x)=5942867)
290
+ 276 train 7.825963 (lr=2.9056e-03) (hash(x)=7812344)
291
+ 277 train 8.128945 (lr=2.9161e-03) (hash(x)=7574027)
292
+ 278 train 8.061168 (lr=2.9266e-03) (hash(x)=5960216)
293
+ 279 train 7.820415 (lr=2.9371e-03) (hash(x)=6793550)
294
+ 280 train 7.842480 (lr=2.9476e-03) (hash(x)=6879965)
295
+ 281 train 7.508279 (lr=2.9580e-03) (hash(x)=6264789)
296
+ 282 train 7.664940 (lr=2.9685e-03) (hash(x)=5175356)
297
+ 283 train 7.884449 (lr=2.9790e-03) (hash(x)=7105976)
298
+ 284 train 7.880244 (lr=2.9895e-03) (hash(x)=7286386)
299
+ 285 train 8.064455 (lr=3.0000e-03) (hash(x)=7244279)
300
+ 286 train 7.590000 (lr=3.0000e-03) (hash(x)=4877247)
301
+ 287 train 7.530113 (lr=3.0000e-03) (hash(x)=6581348)
302
+ 288 train 7.518070 (lr=2.9999e-03) (hash(x)=6397208)
303
+ 289 train 7.654319 (lr=2.9999e-03) (hash(x)=6543825)
304
+ 290 train 7.411352 (lr=2.9998e-03) (hash(x)=6088993)
305
+ 291 train 7.740983 (lr=2.9997e-03) (hash(x)=5555598)
306
+ 292 train 7.819715 (lr=2.9995e-03) (hash(x)=6924240)
307
+ 293 train 7.949368 (lr=2.9994e-03) (hash(x)=7254232)
308
+ 294 train 7.529939 (lr=2.9992e-03) (hash(x)=7689743)
309
+ 295 train 7.532385 (lr=2.9989e-03) (hash(x)=6235837)
310
+ 296 train 7.762885 (lr=2.9987e-03) (hash(x)=5216883)
311
+ 297 train 7.391374 (lr=2.9984e-03) (hash(x)=4738550)
312
+ 298 train 7.585786 (lr=2.9981e-03) (hash(x)=5126918)
313
+ 299 train 7.836045 (lr=2.9978e-03) (hash(x)=5591770)
314
+ 300 val loss 7.6792
315
+ 300 val perplexity 2162.9758
316
+ 300 train 8.613384 (lr=2.9974e-03) (hash(x)=8001573)
317
+ 301 train 8.314347 (lr=2.9971e-03) (hash(x)=8012491)
318
+ 302 train 8.041958 (lr=2.9967e-03) (hash(x)=3997428)
319
+ 303 train 7.369760 (lr=2.9962e-03) (hash(x)=7083021)
320
+ 304 train 7.934669 (lr=2.9958e-03) (hash(x)=7177371)
321
+ 305 train 7.731037 (lr=2.9953e-03) (hash(x)=6456022)
322
+ 306 train 7.825009 (lr=2.9948e-03) (hash(x)=5238861)
323
+ 307 train 7.766739 (lr=2.9942e-03) (hash(x)=7281348)
324
+ 308 train 7.963368 (lr=2.9937e-03) (hash(x)=6848441)
325
+ 309 train 8.020231 (lr=2.9931e-03) (hash(x)=7824794)
326
+ 310 train 8.120831 (lr=2.9925e-03) (hash(x)=7440208)
327
+ 311 train 7.943194 (lr=2.9918e-03) (hash(x)=6155821)
328
+ 312 train 7.682524 (lr=2.9912e-03) (hash(x)=5350461)
329
+ 313 train 7.894571 (lr=2.9905e-03) (hash(x)=7807564)
330
+ 314 train 7.720731 (lr=2.9898e-03) (hash(x)=5992165)
331
+ 315 train 7.664895 (lr=2.9890e-03) (hash(x)=5736241)
332
+ 316 train 7.654248 (lr=2.9883e-03) (hash(x)=6536253)
333
+ 317 train 7.844842 (lr=2.9875e-03) (hash(x)=5688415)
334
+ 318 train 7.500063 (lr=2.9866e-03) (hash(x)=5839705)
335
+ 319 train 7.696191 (lr=2.9858e-03) (hash(x)=5657123)
336
+ 320 train 7.702456 (lr=2.9849e-03) (hash(x)=6354364)
337
+ 321 train 7.734170 (lr=2.9840e-03) (hash(x)=6102985)
338
+ 322 train 7.984921 (lr=2.9831e-03) (hash(x)=7161967)
339
+ 323 train 7.843495 (lr=2.9821e-03) (hash(x)=6452095)
340
+ 324 train 7.595947 (lr=2.9812e-03) (hash(x)=5547396)
341
+ 325 train 7.930247 (lr=2.9802e-03) (hash(x)=5848921)
342
+ 326 train 7.760123 (lr=2.9791e-03) (hash(x)=4890294)
343
+ 327 train 7.472515 (lr=2.9781e-03) (hash(x)=5312267)
344
+ 328 train 7.491535 (lr=2.9770e-03) (hash(x)=5348929)
345
+ 329 train 7.419786 (lr=2.9759e-03) (hash(x)=5025297)
346
+ 330 train 7.473283 (lr=2.9748e-03) (hash(x)=5670052)
347
+ 331 train 7.348184 (lr=2.9736e-03) (hash(x)=5434493)
348
+ 332 train 8.106541 (lr=2.9724e-03) (hash(x)=9122453)
349
+ 333 train 7.853958 (lr=2.9712e-03) (hash(x)=5197474)
350
+ 334 train 7.698234 (lr=2.9700e-03) (hash(x)=5550786)
351
+ 335 train 7.787517 (lr=2.9687e-03) (hash(x)=6830813)
352
+ 336 train 7.578610 (lr=2.9675e-03) (hash(x)=5258435)
353
+ 337 train 7.623184 (lr=2.9662e-03) (hash(x)=5242548)
354
+ 338 train 7.505287 (lr=2.9648e-03) (hash(x)=3754886)
355
+ 339 train 7.450475 (lr=2.9635e-03) (hash(x)=4752771)
356
+ 340 train 7.635046 (lr=2.9621e-03) (hash(x)=5926875)
357
+ 341 train 8.034277 (lr=2.9607e-03) (hash(x)=6691249)
358
+ 342 train 7.771434 (lr=2.9592e-03) (hash(x)=7514623)
359
+ 343 train 7.834600 (lr=2.9578e-03) (hash(x)=6424933)
360
+ 344 train 7.822384 (lr=2.9563e-03) (hash(x)=6347757)
361
+ 345 train 7.743492 (lr=2.9548e-03) (hash(x)=6960120)
362
+ 346 train 7.760602 (lr=2.9532e-03) (hash(x)=7597197)
363
+ 347 train 7.707465 (lr=2.9517e-03) (hash(x)=5786517)
364
+ 348 train 7.709444 (lr=2.9501e-03) (hash(x)=6424363)
365
+ 349 train 7.854973 (lr=2.9485e-03) (hash(x)=7229689)
366
+ 350 val loss 7.7208
367
+ 350 val perplexity 2254.6948
368
+ 350 train 7.926136 (lr=2.9468e-03) (hash(x)=6772284)
369
+ 351 train 8.007377 (lr=2.9452e-03) (hash(x)=6680023)
370
+ 352 train 7.821825 (lr=2.9435e-03) (hash(x)=6218520)
371
+ 353 train 7.811266 (lr=2.9418e-03) (hash(x)=7688393)
372
+ 354 train 7.788617 (lr=2.9400e-03) (hash(x)=6327977)
373
+ 355 train 7.657168 (lr=2.9383e-03) (hash(x)=6474729)
374
+ 356 train 7.713895 (lr=2.9365e-03) (hash(x)=6657266)
375
+ 357 train 7.774554 (lr=2.9347e-03) (hash(x)=5946650)
376
+ 358 train 7.690627 (lr=2.9328e-03) (hash(x)=6734878)
377
+ 359 train 7.738165 (lr=2.9310e-03) (hash(x)=7523279)
378
+ 360 train 7.591099 (lr=2.9291e-03) (hash(x)=6179478)
379
+ 361 train 7.663118 (lr=2.9272e-03) (hash(x)=5531605)
380
+ 362 train 7.662255 (lr=2.9252e-03) (hash(x)=5844980)
381
+ 363 train 7.330174 (lr=2.9233e-03) (hash(x)=5508768)
382
+ 364 train 7.188513 (lr=2.9213e-03) (hash(x)=6337190)
383
+ 365 train 8.705332 (lr=2.9193e-03) (hash(x)=6415982)
384
+ 366 train 7.818138 (lr=2.9172e-03) (hash(x)=6828959)
385
+ 367 train 7.796262 (lr=2.9152e-03) (hash(x)=6635925)
386
+ 368 train 7.537086 (lr=2.9131e-03) (hash(x)=6358540)
387
+ 369 train 7.922810 (lr=2.9110e-03) (hash(x)=5923706)
388
+ 370 train 7.535047 (lr=2.9088e-03) (hash(x)=4843320)
389
+ 371 train 7.835190 (lr=2.9067e-03) (hash(x)=6663801)
390
+ 372 train 7.647429 (lr=2.9045e-03) (hash(x)=5583287)
391
+ 373 train 7.642046 (lr=2.9023e-03) (hash(x)=6113540)
392
+ 374 train 7.756891 (lr=2.9001e-03) (hash(x)=6389936)
393
+ 375 train 7.721598 (lr=2.8978e-03) (hash(x)=5869441)
394
+ 376 train 7.804334 (lr=2.8955e-03) (hash(x)=5148617)
395
+ 377 train 8.048138 (lr=2.8932e-03) (hash(x)=6913454)
396
+ 378 train 7.781920 (lr=2.8909e-03) (hash(x)=5356233)
397
+ 379 train 7.688813 (lr=2.8885e-03) (hash(x)=6698878)
398
+ 380 train 7.664706 (lr=2.8862e-03) (hash(x)=6280665)
399
+ 381 train 7.873877 (lr=2.8838e-03) (hash(x)=6094899)
400
+ 382 train 7.441854 (lr=2.8813e-03) (hash(x)=6249746)
401
+ 383 train 7.761742 (lr=2.8789e-03) (hash(x)=6349004)
402
+ 384 train 7.669205 (lr=2.8764e-03) (hash(x)=7013468)
403
+ 385 train 7.662791 (lr=2.8739e-03) (hash(x)=6045933)
404
+ 386 train 7.946184 (lr=2.8714e-03) (hash(x)=6571735)
405
+ 387 train 7.570157 (lr=2.8689e-03) (hash(x)=4986137)
406
+ 388 train 7.226635 (lr=2.8663e-03) (hash(x)=5244912)
407
+ 389 train 7.622253 (lr=2.8637e-03) (hash(x)=4798229)
408
+ 390 train 7.655063 (lr=2.8611e-03) (hash(x)=5815783)
409
+ 391 train 7.472836 (lr=2.8585e-03) (hash(x)=6008454)
410
+ 392 train 7.728201 (lr=2.8558e-03) (hash(x)=6407333)
411
+ 393 train 7.555892 (lr=2.8531e-03) (hash(x)=5938362)
412
+ 394 train 7.468610 (lr=2.8504e-03) (hash(x)=6077124)
413
+ 395 train 7.718618 (lr=2.8477e-03) (hash(x)=6550770)
414
+ 396 train 7.598958 (lr=2.8449e-03) (hash(x)=6181528)
415
+ 397 train 7.861402 (lr=2.8422e-03) (hash(x)=7055344)
416
+ 398 train 7.996975 (lr=2.8394e-03) (hash(x)=7689348)
417
+ 399 train 7.889503 (lr=2.8365e-03) (hash(x)=7682741)
418
+ 400 val loss 7.7936
419
+ 400 val perplexity 2425.1235
420
+ 400 train 7.714425 (lr=2.8337e-03) (hash(x)=5189401)
421
+ 401 train 7.778482 (lr=2.8308e-03) (hash(x)=7372095)
422
+ 402 train 7.859134 (lr=2.8279e-03) (hash(x)=6838219)
423
+ 403 train 8.362775 (lr=2.8250e-03) (hash(x)=7892158)
424
+ 404 train 8.013268 (lr=2.8221e-03) (hash(x)=5677077)
425
+ 405 train 7.622712 (lr=2.8191e-03) (hash(x)=5929740)
426
+ 406 train 7.711746 (lr=2.8162e-03) (hash(x)=5285055)
427
+ 407 train 7.842814 (lr=2.8131e-03) (hash(x)=7794028)
428
+ 408 train 7.963338 (lr=2.8101e-03) (hash(x)=6640802)
429
+ 409 train 8.038907 (lr=2.8071e-03) (hash(x)=6839923)
430
+ 410 train 7.480287 (lr=2.8040e-03) (hash(x)=4020997)
431
+ 411 train 7.696696 (lr=2.8009e-03) (hash(x)=7093523)
432
+ 412 train 7.965215 (lr=2.7978e-03) (hash(x)=7370045)
433
+ 413 train 7.676276 (lr=2.7947e-03) (hash(x)=5341312)
434
+ 414 train 7.697949 (lr=2.7915e-03) (hash(x)=5589312)
435
+ 415 train 8.140918 (lr=2.7883e-03) (hash(x)=6166062)
436
+ 416 train 8.046467 (lr=2.7851e-03) (hash(x)=8354149)
437
+ 417 train 7.705122 (lr=2.7819e-03) (hash(x)=5992164)
438
+ 418 train 7.695176 (lr=2.7786e-03) (hash(x)=4748657)
439
+ 419 train 7.731619 (lr=2.7754e-03) (hash(x)=6645781)
440
+ 420 train 7.970444 (lr=2.7721e-03) (hash(x)=6689147)
441
+ 421 train 7.796742 (lr=2.7688e-03) (hash(x)=6207724)
442
+ 422 train 7.798504 (lr=2.7654e-03) (hash(x)=6232863)
443
+ 423 train 7.549628 (lr=2.7621e-03) (hash(x)=4811192)
444
+ 424 train 7.585619 (lr=2.7587e-03) (hash(x)=6211761)
445
+ 425 train 7.769132 (lr=2.7553e-03) (hash(x)=6906335)
446
+ 426 train 7.681476 (lr=2.7519e-03) (hash(x)=4922365)
447
+ 427 train 7.733047 (lr=2.7484e-03) (hash(x)=5970866)
448
+ 428 train 7.644405 (lr=2.7450e-03) (hash(x)=6417833)
449
+ 429 train 7.741057 (lr=2.7415e-03) (hash(x)=5238015)
450
+ 430 train 7.697011 (lr=2.7380e-03) (hash(x)=5375451)
451
+ 431 train 7.800653 (lr=2.7344e-03) (hash(x)=7822680)
452
+ 432 train 7.816151 (lr=2.7309e-03) (hash(x)=6173200)
453
+ 433 train 7.658319 (lr=2.7273e-03) (hash(x)=5478019)
454
+ 434 train 7.838753 (lr=2.7237e-03) (hash(x)=5597859)
455
+ 435 train 7.723007 (lr=2.7201e-03) (hash(x)=6317317)
456
+ 436 train 7.882455 (lr=2.7165e-03) (hash(x)=7120467)
457
+ 437 train 7.927950 (lr=2.7128e-03) (hash(x)=5832344)
458
+ 438 train 7.826589 (lr=2.7092e-03) (hash(x)=6813802)
459
+ 439 train 7.901769 (lr=2.7055e-03) (hash(x)=6431409)
460
+ 440 train 8.027299 (lr=2.7018e-03) (hash(x)=5886861)
461
+ 441 train 7.751678 (lr=2.6980e-03) (hash(x)=5900130)
462
+ 442 train 7.979319 (lr=2.6943e-03) (hash(x)=6061563)
463
+ 443 train 7.935695 (lr=2.6905e-03) (hash(x)=6653337)
464
+ 444 train 7.875321 (lr=2.6867e-03) (hash(x)=6575992)
465
+ 445 train 7.803492 (lr=2.6829e-03) (hash(x)=6696784)
466
+ 446 train 7.929328 (lr=2.6791e-03) (hash(x)=6242161)
467
+ 447 train 7.810371 (lr=2.6752e-03) (hash(x)=5323032)
468
+ 448 train 7.638997 (lr=2.6713e-03) (hash(x)=6439367)
469
+ 449 train 7.848484 (lr=2.6674e-03) (hash(x)=5372100)
470
+ 450 val loss 7.8295
471
+ 450 val perplexity 2513.7891
472
+ 450 train 7.891420 (lr=2.6635e-03) (hash(x)=6344849)
473
+ 451 train 7.597983 (lr=2.6596e-03) (hash(x)=5125339)
474
+ 452 train 7.553408 (lr=2.6556e-03) (hash(x)=4876156)
475
+ 453 train 7.682061 (lr=2.6517e-03) (hash(x)=5664463)
476
+ 454 train 7.560354 (lr=2.6477e-03) (hash(x)=6010042)
477
+ 455 train 7.756770 (lr=2.6437e-03) (hash(x)=7363286)
478
+ 456 train 8.239257 (lr=2.6396e-03) (hash(x)=4873268)
479
+ 457 train 8.032410 (lr=2.6356e-03) (hash(x)=7444468)
480
+ 458 train 8.039648 (lr=2.6315e-03) (hash(x)=6698134)
481
+ 459 train 7.879439 (lr=2.6274e-03) (hash(x)=7670050)
482
+ 460 train 7.892153 (lr=2.6233e-03) (hash(x)=6280866)
483
+ 461 train 7.771099 (lr=2.6192e-03) (hash(x)=8189420)
484
+ 462 train 7.876719 (lr=2.6150e-03) (hash(x)=6878292)
485
+ 463 train 7.646762 (lr=2.6109e-03) (hash(x)=5616075)
486
+ 464 train 7.793534 (lr=2.6067e-03) (hash(x)=5839939)
487
+ 465 train 7.680598 (lr=2.6025e-03) (hash(x)=6360924)
488
+ 466 train 8.105989 (lr=2.5983e-03) (hash(x)=6466475)
489
+ 467 train 7.894341 (lr=2.5940e-03) (hash(x)=6593764)
490
+ 468 train 8.313903 (lr=2.5898e-03) (hash(x)=7389698)
491
+ 469 train 7.988275 (lr=2.5855e-03) (hash(x)=6764377)
492
+ 470 train 7.791615 (lr=2.5812e-03) (hash(x)=7293570)
493
+ 471 train 7.823998 (lr=2.5769e-03) (hash(x)=5672608)
494
+ 472 train 7.624279 (lr=2.5726e-03) (hash(x)=5587975)
495
+ 473 train 7.502513 (lr=2.5682e-03) (hash(x)=6985256)
496
+ 474 train 8.078125 (lr=2.5639e-03) (hash(x)=6549231)
497
+ 475 train 7.812413 (lr=2.5595e-03) (hash(x)=6828653)
498
+ 476 train 7.692824 (lr=2.5551e-03) (hash(x)=5478970)
499
+ 477 train 7.714128 (lr=2.5507e-03) (hash(x)=5199845)
500
+ 478 train 7.779286 (lr=2.5462e-03) (hash(x)=6482073)
501
+ 479 train 7.838790 (lr=2.5418e-03) (hash(x)=6839867)
502
+ 480 train 7.843864 (lr=2.5373e-03) (hash(x)=7073454)
503
+ 481 train 8.029387 (lr=2.5328e-03) (hash(x)=6928712)
504
+ 482 train 7.794550 (lr=2.5283e-03) (hash(x)=5962982)
505
+ 483 train 7.915581 (lr=2.5238e-03) (hash(x)=8426992)
506
+ 484 train 7.824721 (lr=2.5193e-03) (hash(x)=6422394)
507
+ 485 train 7.781260 (lr=2.5147e-03) (hash(x)=5330355)
508
+ 486 train 7.311688 (lr=2.5102e-03) (hash(x)=4111491)
509
+ 487 train 7.367527 (lr=2.5056e-03) (hash(x)=4916928)
510
+ 488 train 7.510482 (lr=2.5010e-03) (hash(x)=6610244)
511
+ 489 train 7.939168 (lr=2.4963e-03) (hash(x)=7287928)
512
+ 490 train 7.864048 (lr=2.4917e-03) (hash(x)=6525083)
513
+ 491 train 7.841878 (lr=2.4871e-03) (hash(x)=6484050)
514
+ 492 train 7.930259 (lr=2.4824e-03) (hash(x)=6573424)
515
+ 493 train 7.820003 (lr=2.4777e-03) (hash(x)=5062850)
516
+ 494 train 7.736902 (lr=2.4730e-03) (hash(x)=6215995)
517
+ 495 train 7.928196 (lr=2.4683e-03) (hash(x)=8353379)
518
+ 496 train 7.785374 (lr=2.4636e-03) (hash(x)=6344823)
519
+ 497 train 7.803947 (lr=2.4588e-03) (hash(x)=6805358)
520
+ 498 train 7.662997 (lr=2.4540e-03) (hash(x)=6018070)
521
+ 499 train 7.934598 (lr=2.4493e-03) (hash(x)=6552510)
522
+ 500 val loss 7.7762
523
+ 500 val perplexity 2383.1213
524
+ 500 train 7.693018 (lr=2.4445e-03) (hash(x)=7218108)
525
+ 501 train 7.875204 (lr=2.4397e-03) (hash(x)=5105545)
526
+ 502 train 7.874400 (lr=2.4348e-03) (hash(x)=5997860)
527
+ 503 train 7.673080 (lr=2.4300e-03) (hash(x)=4838871)
528
+ 504 train 7.723816 (lr=2.4251e-03) (hash(x)=7105888)
529
+ 505 train 7.998420 (lr=2.4203e-03) (hash(x)=5945286)
530
+ 506 train 7.171471 (lr=2.4154e-03) (hash(x)=4381720)
531
+ 507 train 6.899977 (lr=2.4105e-03) (hash(x)=3021697)
532
+ 508 train 7.291513 (lr=2.4056e-03) (hash(x)=6826973)
533
+ 509 train 7.887335 (lr=2.4006e-03) (hash(x)=6443070)
534
+ 510 train 7.607554 (lr=2.3957e-03) (hash(x)=5520637)
535
+ 511 train 7.861553 (lr=2.3907e-03) (hash(x)=6795665)
536
+ 512 train 7.661450 (lr=2.3858e-03) (hash(x)=6063891)
537
+ 513 train 8.130020 (lr=2.3808e-03) (hash(x)=7683068)
538
+ 514 train 7.472810 (lr=2.3758e-03) (hash(x)=5526694)
539
+ 515 train 7.511545 (lr=2.3708e-03) (hash(x)=5486935)
540
+ 516 train 7.825277 (lr=2.3657e-03) (hash(x)=6002252)
541
+ 517 train 7.759086 (lr=2.3607e-03) (hash(x)=6115232)
542
+ 518 train 8.092455 (lr=2.3556e-03) (hash(x)=9039136)
543
+ 519 train 7.725471 (lr=2.3506e-03) (hash(x)=6678038)
544
+ 520 train 7.910548 (lr=2.3455e-03) (hash(x)=5383926)
545
+ 521 train 7.895214 (lr=2.3404e-03) (hash(x)=5747629)
546
+ 522 train 7.795856 (lr=2.3353e-03) (hash(x)=7090040)
547
+ 523 train 7.963385 (lr=2.3301e-03) (hash(x)=6657714)
548
+ 524 train 7.588746 (lr=2.3250e-03) (hash(x)=5414034)
549
+ 525 train 7.720796 (lr=2.3198e-03) (hash(x)=6190642)
550
+ 526 train 7.877107 (lr=2.3147e-03) (hash(x)=7431177)
551
+ 527 train 7.801269 (lr=2.3095e-03) (hash(x)=6112215)
552
+ 528 train 7.880668 (lr=2.3043e-03) (hash(x)=6785874)
553
+ 529 train 7.648865 (lr=2.2991e-03) (hash(x)=7314989)
554
+ 530 train 7.811955 (lr=2.2939e-03) (hash(x)=6222614)
555
+ 531 train 7.904590 (lr=2.2887e-03) (hash(x)=8353143)
556
+ 532 train 7.984299 (lr=2.2834e-03) (hash(x)=6752498)
557
+ 533 train 7.792675 (lr=2.2782e-03) (hash(x)=5912570)
558
+ 534 train 7.841045 (lr=2.2729e-03) (hash(x)=5621785)
559
+ 535 train 7.789302 (lr=2.2676e-03) (hash(x)=5915361)
560
+ 536 train 7.751151 (lr=2.2624e-03) (hash(x)=6853672)
561
+ 537 train 7.654902 (lr=2.2571e-03) (hash(x)=6369494)
562
+ 538 train 7.771201 (lr=2.2517e-03) (hash(x)=6039652)
563
+ 539 train 7.918062 (lr=2.2464e-03) (hash(x)=6254885)
564
+ 540 train 7.823614 (lr=2.2411e-03) (hash(x)=5829860)
565
+ 541 train 7.724267 (lr=2.2357e-03) (hash(x)=5425120)
566
+ 542 train 7.525198 (lr=2.2304e-03) (hash(x)=5209746)
567
+ 543 train 8.047946 (lr=2.2250e-03) (hash(x)=5771588)
568
+ 544 train 7.790221 (lr=2.2196e-03) (hash(x)=8298790)
569
+ 545 train 7.824797 (lr=2.2142e-03) (hash(x)=6763967)
570
+ 546 train 7.792260 (lr=2.2088e-03) (hash(x)=4882397)
571
+ 547 train 7.672999 (lr=2.2034e-03) (hash(x)=5561507)
572
+ 548 train 7.638153 (lr=2.1980e-03) (hash(x)=5762755)
573
+ 549 train 7.772575 (lr=2.1926e-03) (hash(x)=7081151)
574
+ 550 val loss 7.8234
575
+ 550 val perplexity 2498.5034
576
+ 550 train 7.741831 (lr=2.1871e-03) (hash(x)=5047520)
577
+ 551 train 7.845212 (lr=2.1817e-03) (hash(x)=5688829)
578
+ 552 train 7.697578 (lr=2.1762e-03) (hash(x)=7008925)
579
+ 553 train 7.785387 (lr=2.1707e-03) (hash(x)=7821554)
580
+ 554 train 7.837257 (lr=2.1653e-03) (hash(x)=6469035)
581
+ 555 train 7.593238 (lr=2.1598e-03) (hash(x)=5371951)
582
+ 556 train 7.884424 (lr=2.1542e-03) (hash(x)=6436050)
583
+ 557 train 7.856150 (lr=2.1487e-03) (hash(x)=8120149)
584
+ 558 train 8.155239 (lr=2.1432e-03) (hash(x)=6189935)
585
+ 559 train 7.856887 (lr=2.1377e-03) (hash(x)=5443305)
586
+ 560 train 7.968640 (lr=2.1321e-03) (hash(x)=6294312)
587
+ 561 train 7.612878 (lr=2.1266e-03) (hash(x)=7066030)
588
+ 562 train 7.968583 (lr=2.1210e-03) (hash(x)=7744989)
589
+ 563 train 8.092594 (lr=2.1154e-03) (hash(x)=5016757)
590
+ 564 train 7.910145 (lr=2.1099e-03) (hash(x)=8001284)
591
+ 565 train 7.867874 (lr=2.1043e-03) (hash(x)=6296163)
592
+ 566 train 7.939247 (lr=2.0987e-03) (hash(x)=6027688)
593
+ 567 train 7.952734 (lr=2.0931e-03) (hash(x)=6901933)
594
+ 568 train 7.776990 (lr=2.0875e-03) (hash(x)=5124305)
595
+ 569 train 7.895994 (lr=2.0818e-03) (hash(x)=8056633)
596
+ 570 train 7.938997 (lr=2.0762e-03) (hash(x)=6677566)
597
+ 571 train 7.974181 (lr=2.0706e-03) (hash(x)=6019085)
598
+ 572 train 7.670002 (lr=2.0649e-03) (hash(x)=5924495)
599
+ 573 train 7.768329 (lr=2.0593e-03) (hash(x)=7003893)
600
+ 574 train 7.769730 (lr=2.0536e-03) (hash(x)=5563075)
601
+ 575 train 7.551907 (lr=2.0479e-03) (hash(x)=4294425)
602
+ 576 train 7.779541 (lr=2.0422e-03) (hash(x)=5677870)
603
+ 577 train 7.685016 (lr=2.0366e-03) (hash(x)=6568540)
604
+ 578 train 7.958889 (lr=2.0309e-03) (hash(x)=7473394)
605
+ 579 train 7.981988 (lr=2.0252e-03) (hash(x)=6095229)
606
+ 580 train 7.958833 (lr=2.0194e-03) (hash(x)=6946168)
607
+ 581 train 7.761086 (lr=2.0137e-03) (hash(x)=5512150)
608
+ 582 train 7.381017 (lr=2.0080e-03) (hash(x)=4380268)
609
+ 583 train 7.435841 (lr=2.0023e-03) (hash(x)=4363941)
610
+ 584 train 7.623767 (lr=1.9965e-03) (hash(x)=4836009)
611
+ 585 train 7.898582 (lr=1.9908e-03) (hash(x)=6457783)
612
+ 586 train 7.764435 (lr=1.9850e-03) (hash(x)=6378784)
613
+ 587 train 7.672204 (lr=1.9793e-03) (hash(x)=4693798)
614
+ 588 train 7.954417 (lr=1.9735e-03) (hash(x)=9193310)
615
+ 589 train 8.570801 (lr=1.9677e-03) (hash(x)=8759806)
616
+ 590 train 8.852253 (lr=1.9620e-03) (hash(x)=9176140)
617
+ 591 train 7.752878 (lr=1.9562e-03) (hash(x)=5486335)
618
+ 592 train 8.013863 (lr=1.9504e-03) (hash(x)=7868436)
619
+ 593 train 7.901806 (lr=1.9446e-03) (hash(x)=6964993)
620
+ 594 train 8.111259 (lr=1.9388e-03) (hash(x)=6741270)
621
+ 595 train 8.410944 (lr=1.9330e-03) (hash(x)=7907450)
622
+ 596 train 8.144588 (lr=1.9272e-03) (hash(x)=6443452)
623
+ 597 train 7.774265 (lr=1.9214e-03) (hash(x)=5236642)
624
+ 598 train 7.877284 (lr=1.9156e-03) (hash(x)=5604795)
625
+ 599 train 7.842079 (lr=1.9097e-03) (hash(x)=7295165)
626
+ 600 val loss 7.7932
627
+ 600 val perplexity 2423.9478
628
+ 600 train 7.845168 (lr=1.9039e-03) (hash(x)=6034549)
629
+ 601 train 7.787958 (lr=1.8981e-03) (hash(x)=4913602)
630
+ 602 train 7.781082 (lr=1.8922e-03) (hash(x)=6271294)
631
+ 603 train 7.978246 (lr=1.8864e-03) (hash(x)=6814026)
632
+ 604 train 7.858347 (lr=1.8805e-03) (hash(x)=6848040)
633
+ 605 train 8.200934 (lr=1.8747e-03) (hash(x)=7042051)
634
+ 606 train 8.075935 (lr=1.8688e-03) (hash(x)=6481002)
635
+ 607 train 7.726669 (lr=1.8629e-03) (hash(x)=6267424)
636
+ 608 train 7.981040 (lr=1.8571e-03) (hash(x)=7665306)
637
+ 609 train 7.864211 (lr=1.8512e-03) (hash(x)=5614727)
638
+ 610 train 7.917119 (lr=1.8453e-03) (hash(x)=7039197)
639
+ 611 train 7.977029 (lr=1.8395e-03) (hash(x)=8086437)
640
+ 612 train 8.047019 (lr=1.8336e-03) (hash(x)=6993846)
641
+ 613 train 8.104912 (lr=1.8277e-03) (hash(x)=5900143)
642
+ 614 train 8.002676 (lr=1.8218e-03) (hash(x)=4841318)
643
+ 615 train 7.971869 (lr=1.8159e-03) (hash(x)=5270452)
644
+ 616 train 7.840648 (lr=1.8100e-03) (hash(x)=5955026)
645
+ 617 train 7.979548 (lr=1.8041e-03) (hash(x)=8617707)
646
+ 618 train 7.827607 (lr=1.7982e-03) (hash(x)=5159401)
647
+ 619 train 7.796496 (lr=1.7923e-03) (hash(x)=6420820)
648
+ 620 train 7.756218 (lr=1.7864e-03) (hash(x)=6628863)
649
+ 621 train 7.529921 (lr=1.7805e-03) (hash(x)=6215930)
650
+ 622 train 7.837785 (lr=1.7746e-03) (hash(x)=5248645)
651
+ 623 train 7.747191 (lr=1.7686e-03) (hash(x)=6305297)
652
+ 624 train 8.248964 (lr=1.7627e-03) (hash(x)=6989704)
653
+ 625 train 7.866917 (lr=1.7568e-03) (hash(x)=4026682)
654
+ 626 train 7.667360 (lr=1.7509e-03) (hash(x)=6367688)
655
+ 627 train 7.817747 (lr=1.7450e-03) (hash(x)=7889849)
656
+ 628 train 7.993845 (lr=1.7390e-03) (hash(x)=7334719)
657
+ 629 train 8.062972 (lr=1.7331e-03) (hash(x)=6141721)
658
+ 630 train 8.772450 (lr=1.7272e-03) (hash(x)=5056572)
659
+ 631 train 8.234645 (lr=1.7212e-03) (hash(x)=6040077)
660
+ 632 train 8.329988 (lr=1.7153e-03) (hash(x)=6866651)
661
+ 633 train 8.576565 (lr=1.7094e-03) (hash(x)=6155449)
662
+ 634 train 8.136023 (lr=1.7034e-03) (hash(x)=6690174)
663
+ 635 train 7.826271 (lr=1.6975e-03) (hash(x)=5652497)
664
+ 636 train 7.869275 (lr=1.6916e-03) (hash(x)=5931050)
665
+ 637 train 7.974321 (lr=1.6856e-03) (hash(x)=6961314)
666
+ 638 train 7.873343 (lr=1.6797e-03) (hash(x)=5973636)
667
+ 639 train 7.924514 (lr=1.6738e-03) (hash(x)=7130251)
668
+ 640 train 7.715878 (lr=1.6678e-03) (hash(x)=7503390)
669
+ 641 train 7.740171 (lr=1.6619e-03) (hash(x)=5927461)
670
+ 642 train 7.863371 (lr=1.6559e-03) (hash(x)=6196741)
671
+ 643 train 7.806585 (lr=1.6500e-03) (hash(x)=6610177)
672
+ 644 train 7.722948 (lr=1.6441e-03) (hash(x)=6635147)
673
+ 645 train 7.967733 (lr=1.6381e-03) (hash(x)=7277580)
674
+ 646 train 7.636018 (lr=1.6322e-03) (hash(x)=5050080)
675
+ 647 train 7.730386 (lr=1.6262e-03) (hash(x)=6508350)
676
+ 648 train 7.896969 (lr=1.6203e-03) (hash(x)=5276338)
677
+ 649 train 8.100254 (lr=1.6144e-03) (hash(x)=6536034)
678
+ 650 val loss 7.8640
679
+ 650 val perplexity 2601.8447
680
+ 650 train 7.907162 (lr=1.6084e-03) (hash(x)=6944772)
681
+ 651 train 7.990850 (lr=1.6025e-03) (hash(x)=6994983)
682
+ 652 train 7.894617 (lr=1.5966e-03) (hash(x)=7172017)
683
+ 653 train 7.975374 (lr=1.5906e-03) (hash(x)=8700721)
684
+ 654 train 7.933091 (lr=1.5847e-03) (hash(x)=6774360)
685
+ 655 train 7.791847 (lr=1.5788e-03) (hash(x)=5859576)
686
+ 656 train 7.432934 (lr=1.5728e-03) (hash(x)=5899275)
687
+ 657 train 7.364005 (lr=1.5669e-03) (hash(x)=5264962)
688
+ 658 train 7.197262 (lr=1.5610e-03) (hash(x)=5679861)
689
+ 659 train 7.616747 (lr=1.5550e-03) (hash(x)=5487065)
690
+ 660 train 7.473057 (lr=1.5491e-03) (hash(x)=4239476)
691
+ 661 train 7.754174 (lr=1.5432e-03) (hash(x)=5731624)
692
+ 662 train 7.857050 (lr=1.5373e-03) (hash(x)=5883465)
693
+ 663 train 7.685098 (lr=1.5314e-03) (hash(x)=4892065)
694
+ 664 train 7.583898 (lr=1.5254e-03) (hash(x)=5858782)
695
+ 665 train 7.819419 (lr=1.5195e-03) (hash(x)=5489496)
696
+ 666 train 7.539626 (lr=1.5136e-03) (hash(x)=4485195)
697
+ 667 train 7.578668 (lr=1.5077e-03) (hash(x)=4933674)
698
+ 668 train 7.822418 (lr=1.5018e-03) (hash(x)=5746292)
699
+ 669 train 7.755854 (lr=1.4959e-03) (hash(x)=7021003)
700
+ 670 train 7.895463 (lr=1.4900e-03) (hash(x)=5876710)
701
+ 671 train 7.928074 (lr=1.4841e-03) (hash(x)=7317289)
702
+ 672 train 8.013015 (lr=1.4782e-03) (hash(x)=5598226)
703
+ 673 train 7.901344 (lr=1.4723e-03) (hash(x)=7869305)
704
+ 674 train 7.996399 (lr=1.4664e-03) (hash(x)=6611408)
705
+ 675 train 8.034737 (lr=1.4605e-03) (hash(x)=6811522)
706
+ 676 train 8.069622 (lr=1.4547e-03) (hash(x)=6704714)
707
+ 677 train 7.957836 (lr=1.4488e-03) (hash(x)=6601423)
708
+ 678 train 8.238927 (lr=1.4429e-03) (hash(x)=6726071)
709
+ 679 train 7.521258 (lr=1.4371e-03) (hash(x)=5510218)
710
+ 680 train 8.352924 (lr=1.4312e-03) (hash(x)=7950952)
711
+ 681 train 7.999514 (lr=1.4253e-03) (hash(x)=7180298)
712
+ 682 train 7.819921 (lr=1.4195e-03) (hash(x)=6068813)
713
+ 683 train 7.907999 (lr=1.4136e-03) (hash(x)=7304235)
714
+ 684 train 8.004745 (lr=1.4078e-03) (hash(x)=7441806)
715
+ 685 train 8.070646 (lr=1.4019e-03) (hash(x)=8111920)
716
+ 686 train 8.553576 (lr=1.3961e-03) (hash(x)=6222783)
717
+ 687 train 7.863027 (lr=1.3903e-03) (hash(x)=6752265)
718
+ 688 train 7.705482 (lr=1.3844e-03) (hash(x)=6147634)
719
+ 689 train 7.745384 (lr=1.3786e-03) (hash(x)=6788720)
720
+ 690 train 7.869769 (lr=1.3728e-03) (hash(x)=6413518)
721
+ 691 train 7.797270 (lr=1.3670e-03) (hash(x)=5994476)
722
+ 692 train 7.699574 (lr=1.3612e-03) (hash(x)=5462082)
723
+ 693 train 7.748358 (lr=1.3554e-03) (hash(x)=5862533)
724
+ 694 train 7.888682 (lr=1.3496e-03) (hash(x)=7132796)
725
+ 695 train 7.793264 (lr=1.3438e-03) (hash(x)=6530867)
726
+ 696 train 7.825474 (lr=1.3380e-03) (hash(x)=7932207)
727
+ 697 train 7.814785 (lr=1.3323e-03) (hash(x)=5914734)
728
+ 698 train 7.963404 (lr=1.3265e-03) (hash(x)=6361594)
729
+ 699 train 7.892280 (lr=1.3207e-03) (hash(x)=5746260)
730
+ 700 val loss 7.8534
731
+ 700 val perplexity 2574.4094
732
+ 700 train 7.896994 (lr=1.3150e-03) (hash(x)=7065743)
733
+ 701 train 7.809541 (lr=1.3092e-03) (hash(x)=6856097)
734
+ 702 train 7.906601 (lr=1.3035e-03) (hash(x)=7588648)
735
+ 703 train 7.789912 (lr=1.2977e-03) (hash(x)=5790078)
736
+ 704 train 7.800962 (lr=1.2920e-03) (hash(x)=6069422)
737
+ 705 train 8.050739 (lr=1.2863e-03) (hash(x)=5805002)
738
+ 706 train 7.692433 (lr=1.2806e-03) (hash(x)=5344711)
739
+ 707 train 7.646271 (lr=1.2748e-03) (hash(x)=6430135)
740
+ 708 train 7.796543 (lr=1.2691e-03) (hash(x)=6317763)
741
+ 709 train 7.716746 (lr=1.2634e-03) (hash(x)=6156715)
742
+ 710 train 8.011200 (lr=1.2578e-03) (hash(x)=6321494)
743
+ 711 train 8.100537 (lr=1.2521e-03) (hash(x)=7614023)
744
+ 712 train 8.022457 (lr=1.2464e-03) (hash(x)=6740380)
745
+ 713 train 7.868315 (lr=1.2407e-03) (hash(x)=4861744)
746
+ 714 train 7.761437 (lr=1.2351e-03) (hash(x)=6542179)
747
+ 715 train 7.834014 (lr=1.2294e-03) (hash(x)=5244861)
748
+ 716 train 7.901673 (lr=1.2238e-03) (hash(x)=7306636)
749
+ 717 train 7.809077 (lr=1.2182e-03) (hash(x)=7163697)
750
+ 718 train 7.977517 (lr=1.2125e-03) (hash(x)=6421642)
751
+ 719 train 7.768769 (lr=1.2069e-03) (hash(x)=5245146)
752
+ 720 train 7.836417 (lr=1.2013e-03) (hash(x)=6046027)
753
+ 721 train 7.683909 (lr=1.1957e-03) (hash(x)=6153866)
754
+ 722 train 7.792342 (lr=1.1901e-03) (hash(x)=5827481)
755
+ 723 train 7.792589 (lr=1.1846e-03) (hash(x)=6415565)
756
+ 724 train 7.871150 (lr=1.1790e-03) (hash(x)=6409570)
757
+ 725 train 8.248946 (lr=1.1734e-03) (hash(x)=7835853)
758
+ 726 train 7.503693 (lr=1.1679e-03) (hash(x)=4827589)
759
+ 727 train 7.647484 (lr=1.1623e-03) (hash(x)=5786972)
760
+ 728 train 7.999010 (lr=1.1568e-03) (hash(x)=6736612)
761
+ 729 train 7.863574 (lr=1.1513e-03) (hash(x)=6733560)
762
+ 730 train 8.019468 (lr=1.1458e-03) (hash(x)=4877208)
763
+ 731 train 7.590273 (lr=1.1402e-03) (hash(x)=6131703)
764
+ 732 train 7.690584 (lr=1.1347e-03) (hash(x)=6533769)
765
+ 733 train 7.590281 (lr=1.1293e-03) (hash(x)=6001331)
766
+ 734 train 8.107752 (lr=1.1238e-03) (hash(x)=10602643)
767
+ 735 train 8.073597 (lr=1.1183e-03) (hash(x)=6346459)
768
+ 736 train 7.780016 (lr=1.1129e-03) (hash(x)=6728215)
769
+ 737 train 8.104657 (lr=1.1074e-03) (hash(x)=8943770)
770
+ 738 train 7.929954 (lr=1.1020e-03) (hash(x)=7141912)
771
+ 739 train 7.886556 (lr=1.0966e-03) (hash(x)=6504131)
772
+ 740 train 7.724180 (lr=1.0912e-03) (hash(x)=6461667)
773
+ 741 train 7.912161 (lr=1.0858e-03) (hash(x)=5869339)
774
+ 742 train 8.213282 (lr=1.0804e-03) (hash(x)=7948065)
775
+ 743 train 7.586617 (lr=1.0750e-03) (hash(x)=5209234)
776
+ 744 train 7.682284 (lr=1.0696e-03) (hash(x)=6372244)
777
+ 745 train 7.845623 (lr=1.0643e-03) (hash(x)=7678937)
778
+ 746 train 7.786153 (lr=1.0589e-03) (hash(x)=6519438)
779
+ 747 train 7.847106 (lr=1.0536e-03) (hash(x)=6163272)
780
+ 748 train 7.898023 (lr=1.0483e-03) (hash(x)=7025209)
781
+ 749 train 7.771480 (lr=1.0429e-03) (hash(x)=5774172)
782
+ 750 val loss 7.7869
783
+ 750 val perplexity 2408.7834
784
+ 750 train 7.885807 (lr=1.0376e-03) (hash(x)=5327301)
785
+ 751 train 7.657256 (lr=1.0324e-03) (hash(x)=5676768)
786
+ 752 train 7.402671 (lr=1.0271e-03) (hash(x)=4823304)
787
+ 753 train 7.513562 (lr=1.0218e-03) (hash(x)=6443895)
788
+ 754 train 8.257156 (lr=1.0166e-03) (hash(x)=8032890)
789
+ 755 train 8.311293 (lr=1.0113e-03) (hash(x)=6090561)
790
+ 756 train 7.689796 (lr=1.0061e-03) (hash(x)=4514262)
791
+ 757 train 8.128029 (lr=1.0009e-03) (hash(x)=7788628)
792
+ 758 train 8.070259 (lr=9.9568e-04) (hash(x)=7239456)
793
+ 759 train 8.083890 (lr=9.9049e-04) (hash(x)=5863092)
794
+ 760 train 7.795516 (lr=9.8531e-04) (hash(x)=5572628)
795
+ 761 train 7.835113 (lr=9.8015e-04) (hash(x)=7141354)
796
+ 762 train 8.108458 (lr=9.7500e-04) (hash(x)=7272638)
797
+ 763 train 7.911903 (lr=9.6986e-04) (hash(x)=7201312)
798
+ 764 train 7.954917 (lr=9.6474e-04) (hash(x)=6015388)
799
+ 765 train 7.998142 (lr=9.5963e-04) (hash(x)=6414278)
800
+ 766 train 7.780207 (lr=9.5453e-04) (hash(x)=7137466)
801
+ 767 train 8.301833 (lr=9.4944e-04) (hash(x)=7019489)
802
+ 768 train 7.870632 (lr=9.4437e-04) (hash(x)=7233453)
803
+ 769 train 7.661254 (lr=9.3932e-04) (hash(x)=7914626)
804
+ 770 train 7.752910 (lr=9.3427e-04) (hash(x)=5764080)
805
+ 771 train 7.802679 (lr=9.2924e-04) (hash(x)=6225608)
806
+ 772 train 7.887588 (lr=9.2423e-04) (hash(x)=8097255)
807
+ 773 train 7.877341 (lr=9.1923e-04) (hash(x)=5998078)
808
+ 774 train 7.979975 (lr=9.1424e-04) (hash(x)=5416254)
809
+ 775 train 7.636493 (lr=9.0927e-04) (hash(x)=5483019)
810
+ 776 train 7.531068 (lr=9.0431e-04) (hash(x)=4702208)
811
+ 777 train 7.838105 (lr=8.9936e-04) (hash(x)=5911642)
812
+ 778 train 7.856088 (lr=8.9443e-04) (hash(x)=6132487)
813
+ 779 train 7.991394 (lr=8.8952e-04) (hash(x)=5903258)
814
+ 780 train 7.900704 (lr=8.8462e-04) (hash(x)=7915382)
815
+ 781 train 7.934622 (lr=8.7973e-04) (hash(x)=5632006)
816
+ 782 train 7.753798 (lr=8.7486e-04) (hash(x)=6518211)
817
+ 783 train 7.818429 (lr=8.7001e-04) (hash(x)=5968716)
818
+ 784 train 7.905376 (lr=8.6516e-04) (hash(x)=7344525)
819
+ 785 train 7.797600 (lr=8.6034e-04) (hash(x)=6401968)
820
+ 786 train 7.729673 (lr=8.5553e-04) (hash(x)=6276127)
821
+ 787 train 7.864790 (lr=8.5073e-04) (hash(x)=5778017)
822
+ 788 train 7.446217 (lr=8.4596e-04) (hash(x)=5387306)
823
+ 789 train 7.433522 (lr=8.4119e-04) (hash(x)=5772567)
824
+ 790 train 7.629452 (lr=8.3644e-04) (hash(x)=6383748)
825
+ 791 train 7.892559 (lr=8.3171e-04) (hash(x)=7780194)
826
+ 792 train 7.759734 (lr=8.2699e-04) (hash(x)=7119030)
827
+ 793 train 7.855388 (lr=8.2229e-04) (hash(x)=6424771)
828
+ 794 train 7.713700 (lr=8.1761e-04) (hash(x)=6540151)
829
+ 795 train 7.736581 (lr=8.1294e-04) (hash(x)=6140998)
830
+ 796 train 7.766694 (lr=8.0829e-04) (hash(x)=6208271)
831
+ 797 train 7.618309 (lr=8.0365e-04) (hash(x)=7859566)
832
+ 798 train 8.209830 (lr=7.9903e-04) (hash(x)=7064477)
833
+ 799 train 7.495323 (lr=7.9443e-04) (hash(x)=3784321)
834
+ 800 val loss 7.7855
835
+ 800 val perplexity 2405.5466
836
+ 800 train 7.730753 (lr=7.8984e-04) (hash(x)=4472758)
837
+ 801 train 7.684057 (lr=7.8527e-04) (hash(x)=5557891)
838
+ 802 train 7.865274 (lr=7.8072e-04) (hash(x)=7969325)
839
+ 803 train 7.885203 (lr=7.7618e-04) (hash(x)=5860821)
840
+ 804 train 7.910656 (lr=7.7167e-04) (hash(x)=6750848)
841
+ 805 train 7.815688 (lr=7.6716e-04) (hash(x)=5674826)
842
+ 806 train 7.623229 (lr=7.6268e-04) (hash(x)=5529163)
843
+ 807 train 7.994808 (lr=7.5821e-04) (hash(x)=7774109)
844
+ 808 train 7.807436 (lr=7.5376e-04) (hash(x)=6762509)
845
+ 809 train 7.866042 (lr=7.4933e-04) (hash(x)=5466449)
846
+ 810 train 7.772157 (lr=7.4491e-04) (hash(x)=7046935)
847
+ 811 train 8.480219 (lr=7.4051e-04) (hash(x)=7997664)
848
+ 812 train 7.922144 (lr=7.3613e-04) (hash(x)=6882594)
849
+ 813 train 8.067657 (lr=7.3177e-04) (hash(x)=7006517)
850
+ 814 train 7.969409 (lr=7.2742e-04) (hash(x)=6832530)
851
+ 815 train 7.831038 (lr=7.2309e-04) (hash(x)=6576749)
852
+ 816 train 7.892548 (lr=7.1878e-04) (hash(x)=8276629)
853
+ 817 train 7.716998 (lr=7.1449e-04) (hash(x)=6896198)
854
+ 818 train 7.922756 (lr=7.1022e-04) (hash(x)=5829252)
855
+ 819 train 8.116790 (lr=7.0596e-04) (hash(x)=7266655)
856
+ 820 train 7.802094 (lr=7.0173e-04) (hash(x)=6015975)
857
+ 821 train 7.792892 (lr=6.9751e-04) (hash(x)=5696696)
858
+ 822 train 7.831595 (lr=6.9331e-04) (hash(x)=5411666)
859
+ 823 train 8.044963 (lr=6.8913e-04) (hash(x)=7072404)
860
+ 824 train 7.637033 (lr=6.8496e-04) (hash(x)=4910095)
861
+ 825 train 7.823920 (lr=6.8082e-04) (hash(x)=6590657)
862
+ 826 train 7.784348 (lr=6.7669e-04) (hash(x)=7665574)
863
+ 827 train 7.805305 (lr=6.7259e-04) (hash(x)=6626459)
864
+ 828 train 7.852653 (lr=6.6850e-04) (hash(x)=6731971)
865
+ 829 train 7.830681 (lr=6.6443e-04) (hash(x)=5644977)
866
+ 830 train 7.811158 (lr=6.6038e-04) (hash(x)=6238654)
867
+ 831 train 7.752955 (lr=6.5635e-04) (hash(x)=6556025)
868
+ 832 train 8.023140 (lr=6.5234e-04) (hash(x)=7175563)
869
+ 833 train 7.908309 (lr=6.4835e-04) (hash(x)=6407492)
870
+ 834 train 7.702672 (lr=6.4437e-04) (hash(x)=4934335)
871
+ 835 train 7.693850 (lr=6.4042e-04) (hash(x)=5942129)
872
+ 836 train 7.752131 (lr=6.3649e-04) (hash(x)=5995643)
873
+ 837 train 7.820917 (lr=6.3257e-04) (hash(x)=6012163)
874
+ 838 train 7.925884 (lr=6.2868e-04) (hash(x)=8464831)
875
+ 839 train 8.110814 (lr=6.2480e-04) (hash(x)=7325027)
876
+ 840 train 7.992098 (lr=6.2095e-04) (hash(x)=6785865)
877
+ 841 train 7.867276 (lr=6.1711e-04) (hash(x)=4425520)
878
+ 842 train 7.713413 (lr=6.1330e-04) (hash(x)=5388267)
879
+ 843 train 7.996577 (lr=6.0950e-04) (hash(x)=7322467)
880
+ 844 train 7.780321 (lr=6.0573e-04) (hash(x)=6681766)
881
+ 845 train 7.795659 (lr=6.0198e-04) (hash(x)=7482800)
882
+ 846 train 7.791105 (lr=5.9824e-04) (hash(x)=5554493)
883
+ 847 train 7.969469 (lr=5.9453e-04) (hash(x)=6373412)
884
+ 848 train 7.459041 (lr=5.9083e-04) (hash(x)=5117517)
885
+ 849 train 7.670384 (lr=5.8716e-04) (hash(x)=6981426)
886
+ 850 val loss 7.7963
887
+ 850 val perplexity 2431.5999
888
+ 850 train 7.664843 (lr=5.8351e-04) (hash(x)=6886188)
889
+ 851 train 7.845689 (lr=5.7988e-04) (hash(x)=7332255)
890
+ 852 train 7.846220 (lr=5.7627e-04) (hash(x)=6172042)
891
+ 853 train 7.973017 (lr=5.7268e-04) (hash(x)=5930894)
892
+ 854 train 8.009789 (lr=5.6911e-04) (hash(x)=7448958)
893
+ 855 train 7.809933 (lr=5.6556e-04) (hash(x)=5262868)
894
+ 856 train 7.708475 (lr=5.6203e-04) (hash(x)=5558427)
895
+ 857 train 7.594190 (lr=5.5853e-04) (hash(x)=5585769)
896
+ 858 train 7.522645 (lr=5.5504e-04) (hash(x)=5838081)
897
+ 859 train 7.679835 (lr=5.5158e-04) (hash(x)=5688247)
898
+ 860 train 7.821201 (lr=5.4813e-04) (hash(x)=5162020)
899
+ 861 train 7.966823 (lr=5.4471e-04) (hash(x)=7462079)
900
+ 862 train 7.796172 (lr=5.4131e-04) (hash(x)=6516108)
901
+ 863 train 7.651901 (lr=5.3793e-04) (hash(x)=8055563)
902
+ 864 train 7.792455 (lr=5.3458e-04) (hash(x)=6271901)
903
+ 865 train 7.843567 (lr=5.3124e-04) (hash(x)=6221701)
904
+ 866 train 7.817536 (lr=5.2793e-04) (hash(x)=5772861)
905
+ 867 train 7.791176 (lr=5.2464e-04) (hash(x)=5352405)
906
+ 868 train 7.846739 (lr=5.2137e-04) (hash(x)=6111630)
907
+ 869 train 7.921049 (lr=5.1812e-04) (hash(x)=6730666)
908
+ 870 train 8.018764 (lr=5.1489e-04) (hash(x)=8698315)
909
+ 871 train 7.735783 (lr=5.1169e-04) (hash(x)=5932790)
910
+ 872 train 8.091751 (lr=5.0851e-04) (hash(x)=8099892)
911
+ 873 train 7.728780 (lr=5.0534e-04) (hash(x)=6223114)
912
+ 874 train 7.768055 (lr=5.0221e-04) (hash(x)=5798363)
913
+ 875 train 7.710142 (lr=4.9909e-04) (hash(x)=6249312)
914
+ 876 train 7.833638 (lr=4.9600e-04) (hash(x)=6929692)
915
+ 877 train 7.950263 (lr=4.9293e-04) (hash(x)=7242827)
916
+ 878 train 7.711322 (lr=4.8988e-04) (hash(x)=6332123)
917
+ 879 train 7.603313 (lr=4.8685e-04) (hash(x)=5680154)
918
+ 880 train 7.949034 (lr=4.8385e-04) (hash(x)=6352331)
919
+ 881 train 7.792794 (lr=4.8087e-04) (hash(x)=6332332)
920
+ 882 train 7.883682 (lr=4.7791e-04) (hash(x)=6180674)
921
+ 883 train 7.806831 (lr=4.7497e-04) (hash(x)=8156280)
922
+ 884 train 7.873969 (lr=4.7206e-04) (hash(x)=5874247)
923
+ 885 train 7.799663 (lr=4.6917e-04) (hash(x)=6659781)
924
+ 886 train 7.909688 (lr=4.6630e-04) (hash(x)=5780147)
925
+ 887 train 7.726494 (lr=4.6346e-04) (hash(x)=5914217)
926
+ 888 train 7.621320 (lr=4.6063e-04) (hash(x)=5411762)
927
+ 889 train 7.830652 (lr=4.5783e-04) (hash(x)=6714110)
928
+ 890 train 7.568984 (lr=4.5506e-04) (hash(x)=5999685)
929
+ 891 train 7.750476 (lr=4.5231e-04) (hash(x)=7120215)
930
+ 892 train 7.978051 (lr=4.4958e-04) (hash(x)=6546587)
931
+ 893 train 7.762921 (lr=4.4687e-04) (hash(x)=6593413)
932
+ 894 train 7.854074 (lr=4.4419e-04) (hash(x)=6369927)
933
+ 895 train 7.783967 (lr=4.4153e-04) (hash(x)=6424128)
934
+ 896 train 7.732890 (lr=4.3889e-04) (hash(x)=5736158)
935
+ 897 train 7.544328 (lr=4.3628e-04) (hash(x)=5903043)
936
+ 898 train 7.632066 (lr=4.3369e-04) (hash(x)=4419128)
937
+ 899 train 7.432707 (lr=4.3113e-04) (hash(x)=4390027)
938
+ 900 val loss 7.7722
939
+ 900 val perplexity 2373.6902
940
+ 900 train 7.711281 (lr=4.2858e-04) (hash(x)=6728135)
941
+ 901 train 7.819985 (lr=4.2607e-04) (hash(x)=6945760)
942
+ 902 train 7.805962 (lr=4.2357e-04) (hash(x)=6081534)
943
+ 903 train 7.852594 (lr=4.2110e-04) (hash(x)=7804089)
944
+ 904 train 7.816335 (lr=4.1865e-04) (hash(x)=6225832)
945
+ 905 train 7.880722 (lr=4.1623e-04) (hash(x)=6273417)
946
+ 906 train 7.941810 (lr=4.1383e-04) (hash(x)=7775633)
947
+ 907 train 7.882707 (lr=4.1146e-04) (hash(x)=7130267)
948
+ 908 train 7.697666 (lr=4.0910e-04) (hash(x)=6554076)
949
+ 909 train 7.793682 (lr=4.0678e-04) (hash(x)=6140697)
950
+ 910 train 7.648546 (lr=4.0447e-04) (hash(x)=6128181)
951
+ 911 train 7.749491 (lr=4.0220e-04) (hash(x)=6490149)
952
+ 912 train 7.400870 (lr=3.9994e-04) (hash(x)=5422426)
953
+ 913 train 7.749549 (lr=3.9771e-04) (hash(x)=4847733)
954
+ 914 train 7.777308 (lr=3.9550e-04) (hash(x)=6349747)
955
+ 915 train 8.021307 (lr=3.9332e-04) (hash(x)=9276378)
956
+ 916 train 7.683119 (lr=3.9116e-04) (hash(x)=6869524)
957
+ 917 train 7.727093 (lr=3.8903e-04) (hash(x)=7082053)
958
+ 918 train 7.772127 (lr=3.8692e-04) (hash(x)=7799351)
959
+ 919 train 7.921618 (lr=3.8484e-04) (hash(x)=6283540)
960
+ 920 train 7.733039 (lr=3.8277e-04) (hash(x)=6660860)
961
+ 921 train 7.666382 (lr=3.8074e-04) (hash(x)=5200956)
962
+ 922 train 7.778658 (lr=3.7873e-04) (hash(x)=5653241)
963
+ 923 train 7.879830 (lr=3.7674e-04) (hash(x)=5781680)
964
+ 924 train 7.892780 (lr=3.7478e-04) (hash(x)=6146666)
965
+ 925 train 7.736541 (lr=3.7284e-04) (hash(x)=8652907)
966
+ 926 train 7.516287 (lr=3.7093e-04) (hash(x)=5311546)
967
+ 927 train 7.686447 (lr=3.6904e-04) (hash(x)=5728337)
968
+ 928 train 7.672038 (lr=3.6718e-04) (hash(x)=5413240)
969
+ 929 train 7.733774 (lr=3.6534e-04) (hash(x)=5832401)
970
+ 930 train 7.487809 (lr=3.6353e-04) (hash(x)=7420230)
971
+ 931 train 7.817134 (lr=3.6174e-04) (hash(x)=6278202)
972
+ 932 train 7.750212 (lr=3.5998e-04) (hash(x)=6189873)
973
+ 933 train 7.872146 (lr=3.5824e-04) (hash(x)=7403868)
974
+ 934 train 7.714795 (lr=3.5653e-04) (hash(x)=6867459)
975
+ 935 train 7.896810 (lr=3.5484e-04) (hash(x)=7400272)
976
+ 936 train 7.889824 (lr=3.5317e-04) (hash(x)=7936275)
977
+ 937 train 7.679620 (lr=3.5154e-04) (hash(x)=5943413)
978
+ 938 train 7.863369 (lr=3.4992e-04) (hash(x)=5442166)
979
+ 939 train 7.729724 (lr=3.4833e-04) (hash(x)=5795128)
980
+ 940 train 7.979134 (lr=3.4677e-04) (hash(x)=6831680)
981
+ 941 train 7.651867 (lr=3.4523e-04) (hash(x)=5519980)
982
+ 942 train 7.714245 (lr=3.4372e-04) (hash(x)=6765461)
983
+ 943 train 7.699864 (lr=3.4224e-04) (hash(x)=6473432)
984
+ 944 train 7.635741 (lr=3.4077e-04) (hash(x)=6490395)
985
+ 945 train 7.807097 (lr=3.3934e-04) (hash(x)=5484669)
986
+ 946 train 7.840549 (lr=3.3793e-04) (hash(x)=7278970)
987
+ 947 train 7.782706 (lr=3.3654e-04) (hash(x)=5201502)
988
+ 948 train 7.972998 (lr=3.3518e-04) (hash(x)=7277915)
989
+ 949 train 7.747558 (lr=3.3385e-04) (hash(x)=7227057)
990
+ 950 val loss 7.7537
991
+ 950 val perplexity 2330.0776
992
+ 950 train 7.940173 (lr=3.3254e-04) (hash(x)=8178996)
993
+ 951 train 8.043763 (lr=3.3125e-04) (hash(x)=6922646)
994
+ 952 train 7.641971 (lr=3.3000e-04) (hash(x)=5941772)
995
+ 953 train 7.617591 (lr=3.2876e-04) (hash(x)=6056673)
996
+ 954 train 7.979010 (lr=3.2756e-04) (hash(x)=6912178)
997
+ 955 train 7.532602 (lr=3.2638e-04) (hash(x)=5484641)
998
+ 956 train 7.833731 (lr=3.2522e-04) (hash(x)=5121117)
999
+ 957 train 7.863537 (lr=3.2409e-04) (hash(x)=7100865)
1000
+ 958 train 7.618775 (lr=3.2299e-04) (hash(x)=5334280)
1001
+ 959 train 7.685956 (lr=3.2191e-04) (hash(x)=7277141)
1002
+ 960 train 7.693188 (lr=3.2085e-04) (hash(x)=6401960)
1003
+ 961 train 7.730649 (lr=3.1983e-04) (hash(x)=6407603)
1004
+ 962 train 7.868009 (lr=3.1883e-04) (hash(x)=5786089)
1005
+ 963 train 7.761530 (lr=3.1785e-04) (hash(x)=5604614)
1006
+ 964 train 7.771940 (lr=3.1690e-04) (hash(x)=5454287)
1007
+ 965 train 7.881487 (lr=3.1598e-04) (hash(x)=6200612)
1008
+ 966 train 7.835196 (lr=3.1508e-04) (hash(x)=5691759)
1009
+ 967 train 7.826319 (lr=3.1421e-04) (hash(x)=7443124)
1010
+ 968 train 7.827312 (lr=3.1336e-04) (hash(x)=6711366)
1011
+ 969 train 7.750677 (lr=3.1254e-04) (hash(x)=6728764)
1012
+ 970 train 8.010289 (lr=3.1174e-04) (hash(x)=6870240)
1013
+ 971 train 7.706845 (lr=3.1098e-04) (hash(x)=5601968)
1014
+ 972 train 7.573323 (lr=3.1023e-04) (hash(x)=5311504)
1015
+ 973 train 7.772033 (lr=3.0952e-04) (hash(x)=6877761)
1016
+ 974 train 7.867342 (lr=3.0882e-04) (hash(x)=6457033)
1017
+ 975 train 7.776309 (lr=3.0816e-04) (hash(x)=5632735)
1018
+ 976 train 7.840657 (lr=3.0752e-04) (hash(x)=6024880)
1019
+ 977 train 7.626174 (lr=3.0691e-04) (hash(x)=5658845)
1020
+ 978 train 7.524320 (lr=3.0632e-04) (hash(x)=6089461)
1021
+ 979 train 7.644247 (lr=3.0576e-04) (hash(x)=5594810)
1022
+ 980 train 7.434683 (lr=3.0522e-04) (hash(x)=4667026)
1023
+ 981 train 7.280829 (lr=3.0471e-04) (hash(x)=4738523)
1024
+ 982 train 7.571605 (lr=3.0423e-04) (hash(x)=7451876)
1025
+ 983 train 7.827176 (lr=3.0377e-04) (hash(x)=5881015)
1026
+ 984 train 7.780199 (lr=3.0334e-04) (hash(x)=6255031)
1027
+ 985 train 7.585758 (lr=3.0294e-04) (hash(x)=5263325)
1028
+ 986 train 7.647796 (lr=3.0256e-04) (hash(x)=6654585)
1029
+ 987 train 7.760674 (lr=3.0221e-04) (hash(x)=6865361)
1030
+ 988 train 7.822405 (lr=3.0188e-04) (hash(x)=6978986)
1031
+ 989 train 7.861642 (lr=3.0158e-04) (hash(x)=5644619)
1032
+ 990 train 7.727475 (lr=3.0131e-04) (hash(x)=6551023)
1033
+ 991 train 7.712071 (lr=3.0106e-04) (hash(x)=5809411)
1034
+ 992 train 7.773301 (lr=3.0084e-04) (hash(x)=5501837)
1035
+ 993 train 7.789573 (lr=3.0064e-04) (hash(x)=7044374)
1036
+ 994 train 7.738503 (lr=3.0047e-04) (hash(x)=6526130)
1037
+ 995 train 7.864846 (lr=3.0033e-04) (hash(x)=6293025)
1038
+ 996 train 7.833774 (lr=3.0021e-04) (hash(x)=6295768)
1039
+ 997 train 7.686445 (lr=3.0012e-04) (hash(x)=5189023)
1040
+ 998 train 7.493783 (lr=3.0005e-04) (hash(x)=6255467)
1041
+ 999 val loss 7.7441
1042
+ 999 val perplexity 2307.9685
1043
+ 999 train 7.845642 (lr=3.0001e-04) (hash(x)=7232964)
lr3e-3_total_batch_size5120_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:969b577b1776ae60712bd3189d67d44806366aad91a9a0c127eec0b5fb3d34eb
3
+ size 36163010
lr3e-3_total_batch_size5120_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fcf9447b144cfb106879f9c9f7a3b16e50d35a66c3ee875808279e5fc57e38c
3
+ size 70764358