andrew-healey commited on
Commit
929f9a2
·
verified ·
1 Parent(s): 32b42a2

Upload folder using huggingface_hub

Browse files
lr6e-4_total_batch_size20480_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr6e-4_total_batch_size20480_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": 20480, "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": 0.0006, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr6e-4_total_batch_size20480_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8513131d2bf732f8f6d97ff82d32c505b2a2f3f9f2247bac315bb09ce0e1f97
3
+ size 964
lr6e-4_total_batch_size20480_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 11.700421 (lr=2.0979e-06) (hash(x)=5113283)
5
+ 1 train 11.688197 (lr=4.1958e-06) (hash(x)=5534142)
6
+ 2 train 11.689584 (lr=6.2937e-06) (hash(x)=6763231)
7
+ 3 train 11.659691 (lr=8.3916e-06) (hash(x)=6328012)
8
+ 4 train 11.645797 (lr=1.0490e-05) (hash(x)=7052161)
9
+ 5 train 11.593186 (lr=1.2587e-05) (hash(x)=5556098)
10
+ 6 train 11.556558 (lr=1.4685e-05) (hash(x)=4696943)
11
+ 7 train 11.504883 (lr=1.6783e-05) (hash(x)=6219138)
12
+ 8 train 11.424795 (lr=1.8881e-05) (hash(x)=6049878)
13
+ 9 train 11.399450 (lr=2.0979e-05) (hash(x)=5491972)
14
+ 10 train 11.321382 (lr=2.3077e-05) (hash(x)=8075458)
15
+ 11 train 11.218025 (lr=2.5175e-05) (hash(x)=5917741)
16
+ 12 train 11.168683 (lr=2.7273e-05) (hash(x)=5750403)
17
+ 13 train 11.092779 (lr=2.9371e-05) (hash(x)=6112458)
18
+ 14 train 11.031763 (lr=3.1469e-05) (hash(x)=5806490)
19
+ 15 train 10.914715 (lr=3.3566e-05) (hash(x)=6568379)
20
+ 16 train 10.824193 (lr=3.5664e-05) (hash(x)=7212403)
21
+ 17 train 10.762385 (lr=3.7762e-05) (hash(x)=7386082)
22
+ 18 train 10.689731 (lr=3.9860e-05) (hash(x)=6359743)
23
+ 19 train 10.633512 (lr=4.1958e-05) (hash(x)=6151674)
24
+ 20 train 10.609541 (lr=4.4056e-05) (hash(x)=7168704)
25
+ 21 train 10.560685 (lr=4.6154e-05) (hash(x)=7859060)
26
+ 22 train 10.487794 (lr=4.8252e-05) (hash(x)=6976239)
27
+ 23 train 10.479760 (lr=5.0350e-05) (hash(x)=6391506)
28
+ 24 train 10.468954 (lr=5.2448e-05) (hash(x)=5245999)
29
+ 25 train 10.421699 (lr=5.4545e-05) (hash(x)=5985675)
30
+ 26 train 10.332866 (lr=5.6643e-05) (hash(x)=4355733)
31
+ 27 train 10.504283 (lr=5.8741e-05) (hash(x)=6984772)
32
+ 28 train 10.353855 (lr=6.0839e-05) (hash(x)=6244096)
33
+ 29 train 10.374558 (lr=6.2937e-05) (hash(x)=6603717)
34
+ 30 train 10.280658 (lr=6.5035e-05) (hash(x)=5669522)
35
+ 31 train 10.258889 (lr=6.7133e-05) (hash(x)=6190579)
36
+ 32 train 10.318835 (lr=6.9231e-05) (hash(x)=4912983)
37
+ 33 train 10.189525 (lr=7.1329e-05) (hash(x)=7276303)
38
+ 34 train 10.252820 (lr=7.3427e-05) (hash(x)=5479947)
39
+ 35 train 10.347001 (lr=7.5524e-05) (hash(x)=6630378)
40
+ 36 train 10.270600 (lr=7.7622e-05) (hash(x)=8097010)
41
+ 37 train 10.325130 (lr=7.9720e-05) (hash(x)=9989744)
42
+ 38 train 10.191603 (lr=8.1818e-05) (hash(x)=6694932)
43
+ 39 train 10.117285 (lr=8.3916e-05) (hash(x)=6317894)
44
+ 40 train 10.018936 (lr=8.6014e-05) (hash(x)=5046327)
45
+ 41 train 10.081988 (lr=8.8112e-05) (hash(x)=7947703)
46
+ 42 train 9.935648 (lr=9.0210e-05) (hash(x)=5916068)
47
+ 43 train 10.025830 (lr=9.2308e-05) (hash(x)=5777498)
48
+ 44 train 9.942770 (lr=9.4406e-05) (hash(x)=5753092)
49
+ 45 train 9.906315 (lr=9.6503e-05) (hash(x)=7723226)
50
+ 46 train 9.879588 (lr=9.8601e-05) (hash(x)=7007016)
51
+ 47 train 9.786048 (lr=1.0070e-04) (hash(x)=6491229)
52
+ 48 train 9.814020 (lr=1.0280e-04) (hash(x)=6390752)
53
+ 49 train 9.911051 (lr=1.0490e-04) (hash(x)=6019389)
54
+ 50 val loss 9.7619
55
+ 50 val perplexity 17359.5156
56
+ 50 train 9.756538 (lr=1.0699e-04) (hash(x)=7662067)
57
+ 51 train 9.734842 (lr=1.0909e-04) (hash(x)=4648609)
58
+ 52 train 9.649823 (lr=1.1119e-04) (hash(x)=4725966)
59
+ 53 train 9.586785 (lr=1.1329e-04) (hash(x)=5558355)
60
+ 54 train 9.607746 (lr=1.1538e-04) (hash(x)=6524933)
61
+ 55 train 9.544440 (lr=1.1748e-04) (hash(x)=6369326)
62
+ 56 train 9.485706 (lr=1.1958e-04) (hash(x)=6022625)
63
+ 57 train 9.400290 (lr=1.2168e-04) (hash(x)=4906853)
64
+ 58 train 9.194029 (lr=1.2378e-04) (hash(x)=5589269)
65
+ 59 train 9.299592 (lr=1.2587e-04) (hash(x)=6437376)
66
+ 60 train 9.329102 (lr=1.2797e-04) (hash(x)=7304892)
67
+ 61 train 9.268149 (lr=1.3007e-04) (hash(x)=7084093)
68
+ 62 train 9.262506 (lr=1.3217e-04) (hash(x)=5908178)
69
+ 63 train 9.017707 (lr=1.3427e-04) (hash(x)=7048804)
70
+ 64 train 9.079869 (lr=1.3636e-04) (hash(x)=6085549)
71
+ 65 train 9.080568 (lr=1.3846e-04) (hash(x)=5752594)
72
+ 66 train 8.854486 (lr=1.4056e-04) (hash(x)=7662769)
73
+ 67 train 8.702577 (lr=1.4266e-04) (hash(x)=6393520)
74
+ 68 train 8.934906 (lr=1.4476e-04) (hash(x)=5942867)
75
+ 69 train 8.912291 (lr=1.4685e-04) (hash(x)=6793550)
76
+ 70 train 8.729600 (lr=1.4895e-04) (hash(x)=7105976)
77
+ 71 train 8.747009 (lr=1.5105e-04) (hash(x)=6581348)
78
+ 72 train 8.656689 (lr=1.5315e-04) (hash(x)=5555598)
79
+ 73 train 8.614600 (lr=1.5524e-04) (hash(x)=6235837)
80
+ 74 train 8.536751 (lr=1.5734e-04) (hash(x)=5591770)
81
+ 75 train 8.758008 (lr=1.5944e-04) (hash(x)=7083021)
82
+ 76 train 8.594731 (lr=1.6154e-04) (hash(x)=7281348)
83
+ 77 train 8.643272 (lr=1.6364e-04) (hash(x)=6155821)
84
+ 78 train 8.458706 (lr=1.6573e-04) (hash(x)=5736241)
85
+ 79 train 8.366180 (lr=1.6783e-04) (hash(x)=5657123)
86
+ 80 train 8.340291 (lr=1.6993e-04) (hash(x)=6452095)
87
+ 81 train 8.261395 (lr=1.7203e-04) (hash(x)=5312267)
88
+ 82 train 8.100721 (lr=1.7413e-04) (hash(x)=5434493)
89
+ 83 train 8.352497 (lr=1.7622e-04) (hash(x)=6830813)
90
+ 84 train 8.178012 (lr=1.7832e-04) (hash(x)=4752771)
91
+ 85 train 8.282215 (lr=1.8042e-04) (hash(x)=6424933)
92
+ 86 train 8.196641 (lr=1.8252e-04) (hash(x)=5786517)
93
+ 87 train 8.275944 (lr=1.8462e-04) (hash(x)=6680023)
94
+ 88 train 8.183519 (lr=1.8671e-04) (hash(x)=6474729)
95
+ 89 train 8.204452 (lr=1.8881e-04) (hash(x)=7523279)
96
+ 90 train 7.996569 (lr=1.9091e-04) (hash(x)=5508768)
97
+ 91 train 8.188123 (lr=1.9301e-04) (hash(x)=6635925)
98
+ 92 train 8.069695 (lr=1.9510e-04) (hash(x)=6663801)
99
+ 93 train 8.044952 (lr=1.9720e-04) (hash(x)=5869441)
100
+ 94 train 8.103032 (lr=1.9930e-04) (hash(x)=6698878)
101
+ 95 train 8.019039 (lr=2.0140e-04) (hash(x)=6349004)
102
+ 96 train 8.017328 (lr=2.0350e-04) (hash(x)=4986137)
103
+ 97 train 7.808747 (lr=2.0559e-04) (hash(x)=6008454)
104
+ 98 train 7.879772 (lr=2.0769e-04) (hash(x)=6550770)
105
+ 99 train 8.014522 (lr=2.0979e-04) (hash(x)=7682741)
106
+ 100 val loss 7.9707
107
+ 100 val perplexity 2894.9226
108
+ 100 train 8.070560 (lr=2.1189e-04) (hash(x)=7892158)
109
+ 101 train 7.940425 (lr=2.1399e-04) (hash(x)=7794028)
110
+ 102 train 7.933344 (lr=2.1608e-04) (hash(x)=7093523)
111
+ 103 train 7.980068 (lr=2.1818e-04) (hash(x)=6166062)
112
+ 104 train 7.918315 (lr=2.2028e-04) (hash(x)=6645781)
113
+ 105 train 7.886864 (lr=2.2238e-04) (hash(x)=4811192)
114
+ 106 train 7.809242 (lr=2.2448e-04) (hash(x)=5970866)
115
+ 107 train 7.818639 (lr=2.2657e-04) (hash(x)=7822680)
116
+ 108 train 7.782276 (lr=2.2867e-04) (hash(x)=6317317)
117
+ 109 train 7.880509 (lr=2.3077e-04) (hash(x)=6431409)
118
+ 110 train 7.896371 (lr=2.3287e-04) (hash(x)=6653337)
119
+ 111 train 7.825218 (lr=2.3497e-04) (hash(x)=5323032)
120
+ 112 train 7.722114 (lr=2.3706e-04) (hash(x)=5125339)
121
+ 113 train 7.667290 (lr=2.3916e-04) (hash(x)=7363286)
122
+ 114 train 8.025681 (lr=2.4126e-04) (hash(x)=7670050)
123
+ 115 train 7.791495 (lr=2.4336e-04) (hash(x)=5616075)
124
+ 116 train 7.844471 (lr=2.4545e-04) (hash(x)=6593764)
125
+ 117 train 7.914949 (lr=2.4755e-04) (hash(x)=5672608)
126
+ 118 train 7.744909 (lr=2.4965e-04) (hash(x)=6828653)
127
+ 119 train 7.746331 (lr=2.5175e-04) (hash(x)=6839867)
128
+ 120 train 7.873523 (lr=2.5385e-04) (hash(x)=8426992)
129
+ 121 train 7.546752 (lr=2.5594e-04) (hash(x)=4916928)
130
+ 122 train 7.755312 (lr=2.5804e-04) (hash(x)=6484050)
131
+ 123 train 7.837285 (lr=2.6014e-04) (hash(x)=8353379)
132
+ 124 train 7.757292 (lr=2.6224e-04) (hash(x)=6552510)
133
+ 125 train 7.735661 (lr=2.6434e-04) (hash(x)=4838871)
134
+ 126 train 7.468061 (lr=2.6643e-04) (hash(x)=3021697)
135
+ 127 train 7.626331 (lr=2.6853e-04) (hash(x)=6795665)
136
+ 128 train 7.649026 (lr=2.7063e-04) (hash(x)=5486935)
137
+ 129 train 7.834298 (lr=2.7273e-04) (hash(x)=6678038)
138
+ 130 train 7.862762 (lr=2.7483e-04) (hash(x)=6657714)
139
+ 131 train 7.731431 (lr=2.7692e-04) (hash(x)=6112215)
140
+ 132 train 7.785623 (lr=2.7902e-04) (hash(x)=8353143)
141
+ 133 train 7.810211 (lr=2.8112e-04) (hash(x)=5915361)
142
+ 134 train 7.724842 (lr=2.8322e-04) (hash(x)=6254885)
143
+ 135 train 7.725986 (lr=2.8531e-04) (hash(x)=5771588)
144
+ 136 train 7.697923 (lr=2.8741e-04) (hash(x)=5561507)
145
+ 137 train 7.665032 (lr=2.8951e-04) (hash(x)=5688829)
146
+ 138 train 7.640872 (lr=2.9161e-04) (hash(x)=5371951)
147
+ 139 train 7.810389 (lr=2.9371e-04) (hash(x)=5443305)
148
+ 140 train 7.785834 (lr=2.9580e-04) (hash(x)=5016757)
149
+ 141 train 7.787242 (lr=2.9790e-04) (hash(x)=6901933)
150
+ 142 train 7.758804 (lr=3.0000e-04) (hash(x)=6019085)
151
+ 143 train 7.603468 (lr=3.0210e-04) (hash(x)=4294425)
152
+ 144 train 7.699917 (lr=3.0420e-04) (hash(x)=6095229)
153
+ 145 train 7.537947 (lr=3.0629e-04) (hash(x)=4363941)
154
+ 146 train 7.617387 (lr=3.0839e-04) (hash(x)=4693798)
155
+ 147 train 8.164754 (lr=3.1049e-04) (hash(x)=5486335)
156
+ 148 train 7.974043 (lr=3.1259e-04) (hash(x)=7907450)
157
+ 149 train 7.805943 (lr=3.1469e-04) (hash(x)=7295165)
158
+ 150 val loss 7.6933
159
+ 150 val perplexity 2193.6753
160
+ 150 train 7.738387 (lr=3.1678e-04) (hash(x)=6814026)
161
+ 151 train 7.818550 (lr=3.1888e-04) (hash(x)=6267424)
162
+ 152 train 7.825542 (lr=3.2098e-04) (hash(x)=8086437)
163
+ 153 train 7.876915 (lr=3.2308e-04) (hash(x)=5270452)
164
+ 154 train 7.717978 (lr=3.2517e-04) (hash(x)=6420820)
165
+ 155 train 7.604784 (lr=3.2727e-04) (hash(x)=6305297)
166
+ 156 train 7.771953 (lr=3.2937e-04) (hash(x)=7889849)
167
+ 157 train 8.124987 (lr=3.3147e-04) (hash(x)=6040077)
168
+ 158 train 8.112553 (lr=3.3357e-04) (hash(x)=5652497)
169
+ 159 train 7.707443 (lr=3.3566e-04) (hash(x)=7130251)
170
+ 160 train 7.572378 (lr=3.3776e-04) (hash(x)=6610177)
171
+ 161 train 7.636499 (lr=3.3986e-04) (hash(x)=6508350)
172
+ 162 train 7.788102 (lr=3.4196e-04) (hash(x)=6994983)
173
+ 163 train 7.744952 (lr=3.4406e-04) (hash(x)=5859576)
174
+ 164 train 7.304287 (lr=3.4615e-04) (hash(x)=5487065)
175
+ 165 train 7.537818 (lr=3.4825e-04) (hash(x)=4892065)
176
+ 166 train 7.439801 (lr=3.5035e-04) (hash(x)=4933674)
177
+ 167 train 7.709087 (lr=3.5245e-04) (hash(x)=7317289)
178
+ 168 train 7.810067 (lr=3.5455e-04) (hash(x)=6811522)
179
+ 169 train 7.639495 (lr=3.5664e-04) (hash(x)=5510218)
180
+ 170 train 7.800025 (lr=3.5874e-04) (hash(x)=7304235)
181
+ 171 train 7.956987 (lr=3.6084e-04) (hash(x)=6752265)
182
+ 172 train 7.660645 (lr=3.6294e-04) (hash(x)=5994476)
183
+ 173 train 7.622958 (lr=3.6503e-04) (hash(x)=6530867)
184
+ 174 train 7.724622 (lr=3.6713e-04) (hash(x)=5746260)
185
+ 175 train 7.680183 (lr=3.6923e-04) (hash(x)=5790078)
186
+ 176 train 7.700298 (lr=3.7133e-04) (hash(x)=6430135)
187
+ 177 train 7.769589 (lr=3.7343e-04) (hash(x)=7614023)
188
+ 178 train 7.755341 (lr=3.7552e-04) (hash(x)=5244861)
189
+ 179 train 7.741224 (lr=3.7762e-04) (hash(x)=5245146)
190
+ 180 train 7.632860 (lr=3.7972e-04) (hash(x)=6415565)
191
+ 181 train 7.684500 (lr=3.8182e-04) (hash(x)=5786972)
192
+ 182 train 7.722946 (lr=3.8392e-04) (hash(x)=6131703)
193
+ 183 train 7.754008 (lr=3.8601e-04) (hash(x)=6346459)
194
+ 184 train 7.784397 (lr=3.8811e-04) (hash(x)=6504131)
195
+ 185 train 7.767685 (lr=3.9021e-04) (hash(x)=5209234)
196
+ 186 train 7.668563 (lr=3.9231e-04) (hash(x)=6163272)
197
+ 187 train 7.676054 (lr=3.9441e-04) (hash(x)=5676768)
198
+ 188 train 7.745142 (lr=3.9650e-04) (hash(x)=6090561)
199
+ 189 train 7.885430 (lr=3.9860e-04) (hash(x)=5863092)
200
+ 190 train 7.799054 (lr=4.0070e-04) (hash(x)=7201312)
201
+ 191 train 7.888859 (lr=4.0280e-04) (hash(x)=7019489)
202
+ 192 train 7.651678 (lr=4.0490e-04) (hash(x)=6225608)
203
+ 193 train 7.722473 (lr=4.0699e-04) (hash(x)=5483019)
204
+ 194 train 7.673765 (lr=4.0909e-04) (hash(x)=5903258)
205
+ 195 train 7.713775 (lr=4.1119e-04) (hash(x)=5968716)
206
+ 196 train 7.700187 (lr=4.1329e-04) (hash(x)=5778017)
207
+ 197 train 7.435637 (lr=4.1538e-04) (hash(x)=7780194)
208
+ 198 train 7.640100 (lr=4.1748e-04) (hash(x)=6140998)
209
+ 199 train 7.632166 (lr=4.1958e-04) (hash(x)=3784321)
210
+ 200 val loss 7.6571
211
+ 200 val perplexity 2115.7114
212
+ 200 train 7.650524 (lr=4.2168e-04) (hash(x)=5860821)
213
+ 201 train 7.712367 (lr=4.2378e-04) (hash(x)=7774109)
214
+ 202 train 7.875457 (lr=4.2587e-04) (hash(x)=7997664)
215
+ 203 train 7.848608 (lr=4.2797e-04) (hash(x)=6576749)
216
+ 204 train 7.761374 (lr=4.3007e-04) (hash(x)=7266655)
217
+ 205 train 7.724911 (lr=4.3217e-04) (hash(x)=7072404)
218
+ 206 train 7.627774 (lr=4.3427e-04) (hash(x)=6626459)
219
+ 207 train 7.647211 (lr=4.3636e-04) (hash(x)=6556025)
220
+ 208 train 7.682404 (lr=4.3846e-04) (hash(x)=5942129)
221
+ 209 train 7.794866 (lr=4.4056e-04) (hash(x)=7325027)
222
+ 210 train 7.780463 (lr=4.4266e-04) (hash(x)=7322467)
223
+ 211 train 7.711524 (lr=4.4476e-04) (hash(x)=6373412)
224
+ 212 train 7.501585 (lr=4.4685e-04) (hash(x)=7332255)
225
+ 213 train 7.783019 (lr=4.4895e-04) (hash(x)=5262868)
226
+ 214 train 7.489666 (lr=4.5105e-04) (hash(x)=5688247)
227
+ 215 train 7.653491 (lr=4.5315e-04) (hash(x)=8055563)
228
+ 216 train 7.696162 (lr=4.5524e-04) (hash(x)=5352405)
229
+ 217 train 7.738225 (lr=4.5734e-04) (hash(x)=5932790)
230
+ 218 train 7.673267 (lr=4.5944e-04) (hash(x)=6249312)
231
+ 219 train 7.632036 (lr=4.6154e-04) (hash(x)=5680154)
232
+ 220 train 7.737161 (lr=4.6364e-04) (hash(x)=8156280)
233
+ 221 train 7.684314 (lr=4.6573e-04) (hash(x)=5914217)
234
+ 222 train 7.527965 (lr=4.6783e-04) (hash(x)=7120215)
235
+ 223 train 7.698344 (lr=4.6993e-04) (hash(x)=6424128)
236
+ 224 train 7.462379 (lr=4.7203e-04) (hash(x)=4390027)
237
+ 225 train 7.670824 (lr=4.7413e-04) (hash(x)=7804089)
238
+ 226 train 7.782909 (lr=4.7622e-04) (hash(x)=7130267)
239
+ 227 train 7.552779 (lr=4.7832e-04) (hash(x)=6490149)
240
+ 228 train 7.619543 (lr=4.8042e-04) (hash(x)=9276378)
241
+ 229 train 7.649869 (lr=4.8252e-04) (hash(x)=6283540)
242
+ 230 train 7.624301 (lr=4.8462e-04) (hash(x)=5781680)
243
+ 231 train 7.572819 (lr=4.8671e-04) (hash(x)=5728337)
244
+ 232 train 7.556737 (lr=4.8881e-04) (hash(x)=6278202)
245
+ 233 train 7.643196 (lr=4.9091e-04) (hash(x)=7400272)
246
+ 234 train 7.610126 (lr=4.9301e-04) (hash(x)=5795128)
247
+ 235 train 7.590264 (lr=4.9510e-04) (hash(x)=6473432)
248
+ 236 train 7.608821 (lr=4.9720e-04) (hash(x)=5201502)
249
+ 237 train 7.847617 (lr=4.9930e-04) (hash(x)=6922646)
250
+ 238 train 7.564544 (lr=5.0140e-04) (hash(x)=5484641)
251
+ 239 train 7.620001 (lr=5.0350e-04) (hash(x)=7277141)
252
+ 240 train 7.646079 (lr=5.0559e-04) (hash(x)=5604614)
253
+ 241 train 7.685642 (lr=5.0769e-04) (hash(x)=7443124)
254
+ 242 train 7.688267 (lr=5.0979e-04) (hash(x)=5601968)
255
+ 243 train 7.573340 (lr=5.1189e-04) (hash(x)=5632735)
256
+ 244 train 7.594320 (lr=5.1399e-04) (hash(x)=5594810)
257
+ 245 train 7.434153 (lr=5.1608e-04) (hash(x)=5881015)
258
+ 246 train 7.561196 (lr=5.1818e-04) (hash(x)=6865361)
259
+ 247 train 7.664344 (lr=5.2028e-04) (hash(x)=5809411)
260
+ 248 train 7.645874 (lr=5.2238e-04) (hash(x)=6293025)
261
+ 249 train 7.589986 (lr=5.2448e-04) (hash(x)=7232964)
262
+ 250 val loss 7.5984
263
+ 250 val perplexity 1995.0034
264
+ 250 train 7.527665 (lr=5.2657e-04) (hash(x)=5646619)
265
+ 251 train 7.642452 (lr=5.2867e-04) (hash(x)=6171829)
266
+ 252 train 7.694654 (lr=5.3077e-04) (hash(x)=7787552)
267
+ 253 train 7.642994 (lr=5.3287e-04) (hash(x)=5884456)
268
+ 254 train 7.670448 (lr=5.3497e-04) (hash(x)=6752162)
269
+ 255 train 7.585383 (lr=5.3706e-04) (hash(x)=5077031)
270
+ 256 train 7.768045 (lr=5.3916e-04) (hash(x)=6883481)
271
+ 257 train 7.629124 (lr=5.4126e-04) (hash(x)=6248895)
272
+ 258 train 7.514213 (lr=5.4336e-04) (hash(x)=6636615)
273
+ 259 train 7.706795 (lr=5.4545e-04) (hash(x)=6918004)
274
+ 260 train 7.621819 (lr=5.4755e-04) (hash(x)=7302874)
275
+ 261 train 7.727994 (lr=5.4965e-04) (hash(x)=6482466)
276
+ 262 train 7.571028 (lr=5.5175e-04) (hash(x)=7298317)
277
+ 263 train 7.595658 (lr=5.5385e-04) (hash(x)=6588573)
278
+ 264 train 7.590421 (lr=5.5594e-04) (hash(x)=6229630)
279
+ 265 train 7.654818 (lr=5.5804e-04) (hash(x)=6136939)
280
+ 266 train 7.409466 (lr=5.6014e-04) (hash(x)=6394645)
281
+ 267 train 7.650968 (lr=5.6224e-04) (hash(x)=7040157)
282
+ 268 train 7.685484 (lr=5.6434e-04) (hash(x)=7041858)
283
+ 269 train 7.590397 (lr=5.6643e-04) (hash(x)=6455595)
284
+ 270 train 7.526274 (lr=5.6853e-04) (hash(x)=6480443)
285
+ 271 train 7.717497 (lr=5.7063e-04) (hash(x)=4122983)
286
+ 272 train 7.519444 (lr=5.7273e-04) (hash(x)=7672190)
287
+ 273 train 7.560659 (lr=5.7483e-04) (hash(x)=4724052)
288
+ 274 train 7.504530 (lr=5.7692e-04) (hash(x)=6654267)
289
+ 275 train 7.497781 (lr=5.7902e-04) (hash(x)=6820797)
290
+ 276 train 7.565026 (lr=5.8112e-04) (hash(x)=6458662)
291
+ 277 train 7.988827 (lr=5.8322e-04) (hash(x)=7477594)
292
+ 278 train 7.661254 (lr=5.8531e-04) (hash(x)=5876574)
293
+ 279 train 7.362312 (lr=5.8741e-04) (hash(x)=6146305)
294
+ 280 train 7.524668 (lr=5.8951e-04) (hash(x)=6162080)
295
+ 281 train 7.482239 (lr=5.9161e-04) (hash(x)=6814831)
296
+ 282 train 7.626757 (lr=5.9371e-04) (hash(x)=7138295)
297
+ 283 train 7.402195 (lr=5.9580e-04) (hash(x)=5548298)
298
+ 284 train 7.490658 (lr=5.9790e-04) (hash(x)=6265078)
299
+ 285 train 7.874024 (lr=6.0000e-04) (hash(x)=7609416)
300
+ 286 train 7.555768 (lr=6.0000e-04) (hash(x)=6909367)
301
+ 287 train 7.676691 (lr=6.0000e-04) (hash(x)=7574342)
302
+ 288 train 7.471145 (lr=5.9999e-04) (hash(x)=8859748)
303
+ 289 train 7.437964 (lr=5.9998e-04) (hash(x)=5906937)
304
+ 290 train 7.475700 (lr=5.9996e-04) (hash(x)=7243024)
305
+ 291 train 7.579866 (lr=5.9993e-04) (hash(x)=8013471)
306
+ 292 train 7.490800 (lr=5.9991e-04) (hash(x)=5659461)
307
+ 293 train 7.555206 (lr=5.9987e-04) (hash(x)=6146213)
308
+ 294 train 7.571939 (lr=5.9983e-04) (hash(x)=6602917)
309
+ 295 train 7.492304 (lr=5.9979e-04) (hash(x)=6709902)
310
+ 296 train 7.608312 (lr=5.9974e-04) (hash(x)=8154341)
311
+ 297 train 7.744592 (lr=5.9968e-04) (hash(x)=6402571)
312
+ 298 train 7.495922 (lr=5.9962e-04) (hash(x)=5605466)
313
+ 299 train 7.456540 (lr=5.9956e-04) (hash(x)=7683614)
314
+ 300 val loss 7.5021
315
+ 300 val perplexity 1811.8381
316
+ 300 train 7.545509 (lr=5.9949e-04) (hash(x)=5421572)
317
+ 301 train 7.393278 (lr=5.9941e-04) (hash(x)=5453026)
318
+ 302 train 7.454529 (lr=5.9933e-04) (hash(x)=6437746)
319
+ 303 train 7.384210 (lr=5.9925e-04) (hash(x)=6260017)
320
+ 304 train 7.468595 (lr=5.9915e-04) (hash(x)=6119053)
321
+ 305 train 7.639496 (lr=5.9906e-04) (hash(x)=6591466)
322
+ 306 train 7.565525 (lr=5.9896e-04) (hash(x)=5652701)
323
+ 307 train 7.521061 (lr=5.9885e-04) (hash(x)=5501472)
324
+ 308 train 7.706266 (lr=5.9874e-04) (hash(x)=6160695)
325
+ 309 train 7.445512 (lr=5.9862e-04) (hash(x)=5871698)
326
+ 310 train 7.426938 (lr=5.9850e-04) (hash(x)=7301184)
327
+ 311 train 7.362667 (lr=5.9837e-04) (hash(x)=4282052)
328
+ 312 train 7.575198 (lr=5.9824e-04) (hash(x)=5679330)
329
+ 313 train 7.427068 (lr=5.9810e-04) (hash(x)=7824660)
330
+ 314 train 7.480555 (lr=5.9795e-04) (hash(x)=4857318)
331
+ 315 train 7.393532 (lr=5.9780e-04) (hash(x)=5954950)
332
+ 316 train 7.460410 (lr=5.9765e-04) (hash(x)=5758178)
333
+ 317 train 7.485731 (lr=5.9749e-04) (hash(x)=5484656)
334
+ 318 train 7.496554 (lr=5.9733e-04) (hash(x)=6818913)
335
+ 319 train 7.690413 (lr=5.9716e-04) (hash(x)=7031328)
336
+ 320 train 7.540304 (lr=5.9698e-04) (hash(x)=6837621)
337
+ 321 train 7.554158 (lr=5.9680e-04) (hash(x)=9157069)
338
+ 322 train 7.795562 (lr=5.9662e-04) (hash(x)=6509536)
339
+ 323 train 7.489796 (lr=5.9643e-04) (hash(x)=5996883)
340
+ 324 train 7.722128 (lr=5.9623e-04) (hash(x)=6850868)
341
+ 325 train 7.660053 (lr=5.9603e-04) (hash(x)=7645447)
342
+ 326 train 7.658932 (lr=5.9583e-04) (hash(x)=6717515)
343
+ 327 train 7.763359 (lr=5.9562e-04) (hash(x)=6336164)
344
+ 328 train 7.496643 (lr=5.9540e-04) (hash(x)=6897445)
345
+ 329 train 7.418449 (lr=5.9518e-04) (hash(x)=5224240)
346
+ 330 train 7.554682 (lr=5.9496e-04) (hash(x)=6612945)
347
+ 331 train 7.998134 (lr=5.9472e-04) (hash(x)=5447801)
348
+ 332 train 7.550564 (lr=5.9449e-04) (hash(x)=7308579)
349
+ 333 train 7.430622 (lr=5.9425e-04) (hash(x)=4155598)
350
+ 334 train 7.454174 (lr=5.9400e-04) (hash(x)=6106773)
351
+ 335 train 7.510879 (lr=5.9375e-04) (hash(x)=7482744)
352
+ 336 train 7.552884 (lr=5.9349e-04) (hash(x)=6310687)
353
+ 337 train 7.433215 (lr=5.9323e-04) (hash(x)=4733414)
354
+ 338 train 7.544614 (lr=5.9296e-04) (hash(x)=6979734)
355
+ 339 train 7.365823 (lr=5.9269e-04) (hash(x)=7806284)
356
+ 340 train 7.376942 (lr=5.9241e-04) (hash(x)=5080536)
357
+ 341 train 7.578179 (lr=5.9213e-04) (hash(x)=7236478)
358
+ 342 train 7.517932 (lr=5.9185e-04) (hash(x)=6619318)
359
+ 343 train 7.446670 (lr=5.9155e-04) (hash(x)=4656255)
360
+ 344 train 7.389146 (lr=5.9126e-04) (hash(x)=5546888)
361
+ 345 train 7.812582 (lr=5.9095e-04) (hash(x)=9298325)
362
+ 346 train 7.525166 (lr=5.9065e-04) (hash(x)=5994232)
363
+ 347 train 7.501871 (lr=5.9033e-04) (hash(x)=6287524)
364
+ 348 train 7.623867 (lr=5.9002e-04) (hash(x)=6200049)
365
+ 349 train 7.513824 (lr=5.8969e-04) (hash(x)=7895970)
366
+ 350 val loss 7.4525
367
+ 350 val perplexity 1724.0956
368
+ 350 train 7.429665 (lr=5.8937e-04) (hash(x)=6051847)
369
+ 351 train 7.600766 (lr=5.8903e-04) (hash(x)=8847992)
370
+ 352 train 7.969036 (lr=5.8869e-04) (hash(x)=5123883)
371
+ 353 train 7.567740 (lr=5.8835e-04) (hash(x)=6338998)
372
+ 354 train 7.541628 (lr=5.8800e-04) (hash(x)=5765866)
373
+ 355 train 7.858485 (lr=5.8765e-04) (hash(x)=7104011)
374
+ 356 train 7.639140 (lr=5.8729e-04) (hash(x)=7438267)
375
+ 357 train 7.351393 (lr=5.8693e-04) (hash(x)=5573214)
376
+ 358 train 7.638958 (lr=5.8656e-04) (hash(x)=6897283)
377
+ 359 train 7.435463 (lr=5.8619e-04) (hash(x)=6882897)
378
+ 360 train 7.536877 (lr=5.8581e-04) (hash(x)=5343615)
379
+ 361 train 7.447063 (lr=5.8543e-04) (hash(x)=5854242)
380
+ 362 train 7.363630 (lr=5.8504e-04) (hash(x)=6317324)
381
+ 363 train 7.303843 (lr=5.8465e-04) (hash(x)=6972494)
382
+ 364 train 7.378379 (lr=5.8425e-04) (hash(x)=5661167)
383
+ 365 train 7.406442 (lr=5.8385e-04) (hash(x)=6796805)
384
+ 366 train 7.236358 (lr=5.8345e-04) (hash(x)=5817209)
385
+ 367 train 7.316934 (lr=5.8303e-04) (hash(x)=6139120)
386
+ 368 train 7.468953 (lr=5.8262e-04) (hash(x)=6121296)
387
+ 369 train 7.448250 (lr=5.8219e-04) (hash(x)=6486796)
388
+ 370 train 7.379686 (lr=5.8177e-04) (hash(x)=5514645)
389
+ 371 train 7.305516 (lr=5.8134e-04) (hash(x)=5578797)
390
+ 372 train 7.298994 (lr=5.8090e-04) (hash(x)=7651738)
391
+ 373 train 7.493082 (lr=5.8046e-04) (hash(x)=6895326)
392
+ 374 train 7.509957 (lr=5.8001e-04) (hash(x)=5490256)
393
+ 375 train 7.409189 (lr=5.7956e-04) (hash(x)=7783160)
394
+ 376 train 7.629080 (lr=5.7911e-04) (hash(x)=5608424)
395
+ 377 train 7.391338 (lr=5.7864e-04) (hash(x)=4867625)
396
+ 378 train 7.285182 (lr=5.7818e-04) (hash(x)=4655472)
397
+ 379 train 7.368536 (lr=5.7771e-04) (hash(x)=6631123)
398
+ 380 train 7.065244 (lr=5.7723e-04) (hash(x)=5751990)
399
+ 381 train 7.326099 (lr=5.7675e-04) (hash(x)=6710965)
400
+ 382 train 7.294798 (lr=5.7627e-04) (hash(x)=5293800)
401
+ 383 train 7.311748 (lr=5.7578e-04) (hash(x)=6294657)
402
+ 384 train 7.457703 (lr=5.7529e-04) (hash(x)=6999538)
403
+ 385 train 7.446897 (lr=5.7479e-04) (hash(x)=5443606)
404
+ 386 train 7.431631 (lr=5.7428e-04) (hash(x)=5650401)
405
+ 387 train 7.419811 (lr=5.7377e-04) (hash(x)=7035770)
406
+ 388 train 7.378371 (lr=5.7326e-04) (hash(x)=6678271)
407
+ 389 train 7.578262 (lr=5.7274e-04) (hash(x)=4899541)
408
+ 390 train 7.334831 (lr=5.7222e-04) (hash(x)=5359260)
409
+ 391 train 7.347326 (lr=5.7169e-04) (hash(x)=6227120)
410
+ 392 train 7.409576 (lr=5.7116e-04) (hash(x)=6643916)
411
+ 393 train 7.482464 (lr=5.7063e-04) (hash(x)=5981648)
412
+ 394 train 7.415128 (lr=5.7008e-04) (hash(x)=6467540)
413
+ 395 train 7.344954 (lr=5.6954e-04) (hash(x)=5558993)
414
+ 396 train 7.518650 (lr=5.6899e-04) (hash(x)=6698959)
415
+ 397 train 7.346379 (lr=5.6843e-04) (hash(x)=7015671)
416
+ 398 train 7.336495 (lr=5.6787e-04) (hash(x)=6800890)
417
+ 399 train 7.492150 (lr=5.6731e-04) (hash(x)=5931234)
418
+ 400 val loss 7.4163
419
+ 400 val perplexity 1662.9097
420
+ 400 train 7.473899 (lr=5.6674e-04) (hash(x)=3995070)
421
+ 401 train 7.405529 (lr=5.6617e-04) (hash(x)=4582159)
422
+ 402 train 7.444661 (lr=5.6559e-04) (hash(x)=7081906)
423
+ 403 train 7.383349 (lr=5.6501e-04) (hash(x)=5029472)
424
+ 404 train 7.406818 (lr=5.6442e-04) (hash(x)=3946232)
425
+ 405 train 7.324516 (lr=5.6383e-04) (hash(x)=5434102)
426
+ 406 train 8.035899 (lr=5.6323e-04) (hash(x)=5340646)
427
+ 407 train 7.208325 (lr=5.6263e-04) (hash(x)=6810638)
428
+ 408 train 7.441220 (lr=5.6202e-04) (hash(x)=6768026)
429
+ 409 train 7.444545 (lr=5.6141e-04) (hash(x)=6521562)
430
+ 410 train 7.406030 (lr=5.6080e-04) (hash(x)=4429434)
431
+ 411 train 7.473647 (lr=5.6018e-04) (hash(x)=7226932)
432
+ 412 train 7.224082 (lr=5.5956e-04) (hash(x)=5065932)
433
+ 413 train 7.402736 (lr=5.5893e-04) (hash(x)=6957849)
434
+ 414 train 7.380407 (lr=5.5830e-04) (hash(x)=6431457)
435
+ 415 train 7.463393 (lr=5.5766e-04) (hash(x)=5334281)
436
+ 416 train 7.471122 (lr=5.5702e-04) (hash(x)=5471747)
437
+ 417 train 7.560372 (lr=5.5638e-04) (hash(x)=5887983)
438
+ 418 train 7.458456 (lr=5.5573e-04) (hash(x)=5075909)
439
+ 419 train 7.284758 (lr=5.5507e-04) (hash(x)=6393576)
440
+ 420 train 7.388631 (lr=5.5441e-04) (hash(x)=6608979)
441
+ 421 train 7.799415 (lr=5.5375e-04) (hash(x)=5057959)
442
+ 422 train 7.733308 (lr=5.5308e-04) (hash(x)=6414858)
443
+ 423 train 7.501235 (lr=5.5241e-04) (hash(x)=8595271)
444
+ 424 train 7.355200 (lr=5.5174e-04) (hash(x)=6500949)
445
+ 425 train 7.530478 (lr=5.5106e-04) (hash(x)=8395167)
446
+ 426 train 7.651070 (lr=5.5037e-04) (hash(x)=6288420)
447
+ 427 train 7.482823 (lr=5.4968e-04) (hash(x)=7634417)
448
+ 428 train 7.365458 (lr=5.4899e-04) (hash(x)=6795561)
449
+ 429 train 7.265526 (lr=5.4829e-04) (hash(x)=5257771)
450
+ 430 train 7.415299 (lr=5.4759e-04) (hash(x)=7506860)
451
+ 431 train 7.477797 (lr=5.4689e-04) (hash(x)=7674238)
452
+ 432 train 7.571830 (lr=5.4618e-04) (hash(x)=5353794)
453
+ 433 train 7.498618 (lr=5.4546e-04) (hash(x)=5300555)
454
+ 434 train 7.222622 (lr=5.4475e-04) (hash(x)=5564419)
455
+ 435 train 7.468519 (lr=5.4402e-04) (hash(x)=6066139)
456
+ 436 train 7.445884 (lr=5.4330e-04) (hash(x)=6863709)
457
+ 437 train 7.445212 (lr=5.4257e-04) (hash(x)=6688212)
458
+ 438 train 7.601372 (lr=5.4183e-04) (hash(x)=7095090)
459
+ 439 train 7.693650 (lr=5.4109e-04) (hash(x)=6642404)
460
+ 440 train 7.403536 (lr=5.4035e-04) (hash(x)=6539227)
461
+ 441 train 7.387924 (lr=5.3960e-04) (hash(x)=4961570)
462
+ 442 train 7.336418 (lr=5.3885e-04) (hash(x)=6606391)
463
+ 443 train 7.152795 (lr=5.3810e-04) (hash(x)=5105190)
464
+ 444 train 7.710122 (lr=5.3734e-04) (hash(x)=5695997)
465
+ 445 train 7.412235 (lr=5.3658e-04) (hash(x)=7155991)
466
+ 446 train 7.391785 (lr=5.3581e-04) (hash(x)=7249727)
467
+ 447 train 7.382384 (lr=5.3504e-04) (hash(x)=5132972)
468
+ 448 train 7.379567 (lr=5.3426e-04) (hash(x)=5381802)
469
+ 449 train 7.402682 (lr=5.3349e-04) (hash(x)=6269946)
470
+ 450 val loss 7.3813
471
+ 450 val perplexity 1605.5961
472
+ 450 train 7.348353 (lr=5.3270e-04) (hash(x)=4976762)
473
+ 451 train 7.441354 (lr=5.3192e-04) (hash(x)=5396133)
474
+ 452 train 7.399624 (lr=5.3113e-04) (hash(x)=6173473)
475
+ 453 train 7.373999 (lr=5.3033e-04) (hash(x)=6394083)
476
+ 454 train 7.434234 (lr=5.2953e-04) (hash(x)=6436030)
477
+ 455 train 7.304087 (lr=5.2873e-04) (hash(x)=5237804)
478
+ 456 train 7.287929 (lr=5.2792e-04) (hash(x)=6540397)
479
+ 457 train 7.531482 (lr=5.2711e-04) (hash(x)=5216638)
480
+ 458 train 7.341828 (lr=5.2630e-04) (hash(x)=5750537)
481
+ 459 train 7.325881 (lr=5.2548e-04) (hash(x)=6735565)
482
+ 460 train 7.355223 (lr=5.2466e-04) (hash(x)=7297488)
483
+ 461 train 7.519494 (lr=5.2384e-04) (hash(x)=5983575)
484
+ 462 train 7.353300 (lr=5.2301e-04) (hash(x)=4808543)
485
+ 463 train 7.451389 (lr=5.2217e-04) (hash(x)=5831619)
486
+ 464 train 7.222446 (lr=5.2134e-04) (hash(x)=6280052)
487
+ 465 train 7.237765 (lr=5.2050e-04) (hash(x)=9109083)
488
+ 466 train 7.515056 (lr=5.1965e-04) (hash(x)=5848604)
489
+ 467 train 7.420091 (lr=5.1881e-04) (hash(x)=6912967)
490
+ 468 train 7.590788 (lr=5.1796e-04) (hash(x)=6615365)
491
+ 469 train 7.629678 (lr=5.1710e-04) (hash(x)=5965674)
492
+ 470 train 7.471103 (lr=5.1624e-04) (hash(x)=6879965)
493
+ 471 train 7.474340 (lr=5.1538e-04) (hash(x)=5822533)
494
+ 472 train 7.322839 (lr=5.1452e-04) (hash(x)=5794462)
495
+ 473 train 7.402553 (lr=5.1365e-04) (hash(x)=5268216)
496
+ 474 train 7.223906 (lr=5.1277e-04) (hash(x)=7058396)
497
+ 475 train 7.320179 (lr=5.1190e-04) (hash(x)=6531900)
498
+ 476 train 7.352033 (lr=5.1102e-04) (hash(x)=5878912)
499
+ 477 train 7.165132 (lr=5.1013e-04) (hash(x)=5426499)
500
+ 478 train 7.382191 (lr=5.0925e-04) (hash(x)=8107004)
501
+ 479 train 8.164145 (lr=5.0836e-04) (hash(x)=7455372)
502
+ 480 train 8.072473 (lr=5.0746e-04) (hash(x)=7727959)
503
+ 481 train 7.992570 (lr=5.0657e-04) (hash(x)=8319119)
504
+ 482 train 7.894061 (lr=5.0567e-04) (hash(x)=7581656)
505
+ 483 train 7.942920 (lr=5.0476e-04) (hash(x)=7592128)
506
+ 484 train 7.911593 (lr=5.0386e-04) (hash(x)=10506633)
507
+ 485 train 7.855330 (lr=5.0295e-04) (hash(x)=8704170)
508
+ 486 train 7.704393 (lr=5.0203e-04) (hash(x)=7461138)
509
+ 487 train 7.667949 (lr=5.0111e-04) (hash(x)=7165037)
510
+ 488 train 7.703639 (lr=5.0019e-04) (hash(x)=4976998)
511
+ 489 train 7.457592 (lr=4.9927e-04) (hash(x)=7678000)
512
+ 490 train 7.521662 (lr=4.9834e-04) (hash(x)=6299409)
513
+ 491 train 7.461071 (lr=4.9741e-04) (hash(x)=7414788)
514
+ 492 train 7.513994 (lr=4.9648e-04) (hash(x)=5063061)
515
+ 493 train 7.405411 (lr=4.9554e-04) (hash(x)=7326216)
516
+ 494 train 7.547427 (lr=4.9460e-04) (hash(x)=7492012)
517
+ 495 train 7.518259 (lr=4.9366e-04) (hash(x)=5857162)
518
+ 496 train 7.316451 (lr=4.9271e-04) (hash(x)=6169001)
519
+ 497 train 9.262093 (lr=4.9176e-04) (hash(x)=14715669)
520
+ 498 train 12.040575 (lr=4.9081e-04) (hash(x)=18256706)
521
+ 499 train 11.780813 (lr=4.8985e-04) (hash(x)=16950560)
522
+ 500 val loss 7.4213
523
+ 500 val perplexity 1671.1719
524
+ 500 train 8.108996 (lr=4.8889e-04) (hash(x)=6113563)
525
+ 501 train 7.123283 (lr=4.8793e-04) (hash(x)=6159125)
526
+ 502 train 7.496033 (lr=4.8697e-04) (hash(x)=5482303)
527
+ 503 train 7.545724 (lr=4.8600e-04) (hash(x)=5908442)
528
+ 504 train 7.393165 (lr=4.8503e-04) (hash(x)=4985045)
529
+ 505 train 7.299108 (lr=4.8405e-04) (hash(x)=6668887)
530
+ 506 train 7.349122 (lr=4.8308e-04) (hash(x)=6123910)
531
+ 507 train 7.432206 (lr=4.8210e-04) (hash(x)=7662975)
532
+ 508 train 7.376202 (lr=4.8111e-04) (hash(x)=7191898)
533
+ 509 train 7.326754 (lr=4.8013e-04) (hash(x)=6669539)
534
+ 510 train 7.402351 (lr=4.7914e-04) (hash(x)=6557481)
535
+ 511 train 7.294858 (lr=4.7815e-04) (hash(x)=5015998)
536
+ 512 train 7.359859 (lr=4.7715e-04) (hash(x)=4964540)
537
+ 513 train 7.199244 (lr=4.7615e-04) (hash(x)=5458386)
538
+ 514 train 7.597562 (lr=4.7515e-04) (hash(x)=6077616)
539
+ 515 train 7.312298 (lr=4.7415e-04) (hash(x)=4849847)
540
+ 516 train 7.373060 (lr=4.7315e-04) (hash(x)=5431422)
541
+ 517 train 7.567559 (lr=4.7214e-04) (hash(x)=5125562)
542
+ 518 train 7.465059 (lr=4.7113e-04) (hash(x)=5587715)
543
+ 519 train 7.523211 (lr=4.7011e-04) (hash(x)=6012047)
544
+ 520 train 7.347685 (lr=4.6909e-04) (hash(x)=5668436)
545
+ 521 train 7.361018 (lr=4.6807e-04) (hash(x)=6079364)
546
+ 522 train 7.491480 (lr=4.6705e-04) (hash(x)=5200504)
547
+ 523 train 7.251234 (lr=4.6603e-04) (hash(x)=6030395)
548
+ 524 train 7.475100 (lr=4.6500e-04) (hash(x)=6855319)
549
+ 525 train 7.252298 (lr=4.6397e-04) (hash(x)=6781762)
550
+ 526 train 7.397111 (lr=4.6294e-04) (hash(x)=6717939)
551
+ 527 train 7.390493 (lr=4.6190e-04) (hash(x)=5238325)
552
+ 528 train 7.305761 (lr=4.6086e-04) (hash(x)=5808069)
553
+ 529 train 7.629593 (lr=4.5982e-04) (hash(x)=8611144)
554
+ 530 train 7.293363 (lr=4.5878e-04) (hash(x)=6028738)
555
+ 531 train 7.320099 (lr=4.5774e-04) (hash(x)=6901605)
556
+ 532 train 7.285803 (lr=4.5669e-04) (hash(x)=6723307)
557
+ 533 train 7.392346 (lr=4.5564e-04) (hash(x)=5499337)
558
+ 534 train 7.317383 (lr=4.5458e-04) (hash(x)=6586372)
559
+ 535 train 7.436609 (lr=4.5353e-04) (hash(x)=7134874)
560
+ 536 train 7.328908 (lr=4.5247e-04) (hash(x)=8415623)
561
+ 537 train 7.356822 (lr=4.5141e-04) (hash(x)=6149038)
562
+ 538 train 7.395509 (lr=4.5035e-04) (hash(x)=6078325)
563
+ 539 train 7.368484 (lr=4.4928e-04) (hash(x)=6609760)
564
+ 540 train 7.260230 (lr=4.4822e-04) (hash(x)=6185253)
565
+ 541 train 7.607190 (lr=4.4715e-04) (hash(x)=6652627)
566
+ 542 train 7.451065 (lr=4.4608e-04) (hash(x)=7758536)
567
+ 543 train 7.432088 (lr=4.4500e-04) (hash(x)=8618579)
568
+ 544 train 7.384881 (lr=4.4393e-04) (hash(x)=6710226)
569
+ 545 train 7.240624 (lr=4.4285e-04) (hash(x)=5896892)
570
+ 546 train 7.258795 (lr=4.4177e-04) (hash(x)=6332995)
571
+ 547 train 7.687635 (lr=4.4069e-04) (hash(x)=8945854)
572
+ 548 train 7.238194 (lr=4.3960e-04) (hash(x)=4329129)
573
+ 549 train 7.280193 (lr=4.3852e-04) (hash(x)=6775103)
574
+ 550 val loss 7.2985
575
+ 550 val perplexity 1478.1306
576
+ 550 train 7.394739 (lr=4.3743e-04) (hash(x)=4465735)
577
+ 551 train 7.258584 (lr=4.3634e-04) (hash(x)=6200024)
578
+ 552 train 7.218898 (lr=4.3524e-04) (hash(x)=5520748)
579
+ 553 train 7.240789 (lr=4.3415e-04) (hash(x)=5238139)
580
+ 554 train 7.508582 (lr=4.3305e-04) (hash(x)=7639817)
581
+ 555 train 7.255368 (lr=4.3195e-04) (hash(x)=5780948)
582
+ 556 train 7.379350 (lr=4.3085e-04) (hash(x)=6733444)
583
+ 557 train 7.327758 (lr=4.2975e-04) (hash(x)=6168407)
584
+ 558 train 7.368332 (lr=4.2864e-04) (hash(x)=6680197)
585
+ 559 train 7.252237 (lr=4.2754e-04) (hash(x)=3775881)
586
+ 560 train 7.098436 (lr=4.2643e-04) (hash(x)=5784096)
587
+ 561 train 7.550383 (lr=4.2532e-04) (hash(x)=6848709)
588
+ 562 train 7.601026 (lr=4.2420e-04) (hash(x)=7449019)
589
+ 563 train 7.701544 (lr=4.2309e-04) (hash(x)=6478113)
590
+ 564 train 7.335200 (lr=4.2197e-04) (hash(x)=5925744)
591
+ 565 train 7.400020 (lr=4.2086e-04) (hash(x)=6483914)
592
+ 566 train 7.276992 (lr=4.1974e-04) (hash(x)=6010847)
593
+ 567 train 7.368370 (lr=4.1861e-04) (hash(x)=5927795)
594
+ 568 train 7.358539 (lr=4.1749e-04) (hash(x)=5896332)
595
+ 569 train 7.321370 (lr=4.1637e-04) (hash(x)=8438033)
596
+ 570 train 7.427846 (lr=4.1524e-04) (hash(x)=6081660)
597
+ 571 train 7.340557 (lr=4.1411e-04) (hash(x)=5684788)
598
+ 572 train 7.233648 (lr=4.1298e-04) (hash(x)=6272055)
599
+ 573 train 7.190294 (lr=4.1185e-04) (hash(x)=5710182)
600
+ 574 train 7.295902 (lr=4.1072e-04) (hash(x)=6219616)
601
+ 575 train 7.259150 (lr=4.0958e-04) (hash(x)=6160063)
602
+ 576 train 7.281532 (lr=4.0845e-04) (hash(x)=5710809)
603
+ 577 train 7.027317 (lr=4.0731e-04) (hash(x)=5955100)
604
+ 578 train 7.036273 (lr=4.0617e-04) (hash(x)=5356088)
605
+ 579 train 7.096804 (lr=4.0503e-04) (hash(x)=6672309)
606
+ 580 train 7.316266 (lr=4.0389e-04) (hash(x)=5841684)
607
+ 581 train 7.673600 (lr=4.0275e-04) (hash(x)=9001564)
608
+ 582 train 8.262786 (lr=4.0160e-04) (hash(x)=6897178)
609
+ 583 train 7.422684 (lr=4.0045e-04) (hash(x)=5454654)
610
+ 584 train 7.163071 (lr=3.9931e-04) (hash(x)=6391454)
611
+ 585 train 7.222925 (lr=3.9816e-04) (hash(x)=5174062)
612
+ 586 train 7.356100 (lr=3.9701e-04) (hash(x)=5197924)
613
+ 587 train 7.333878 (lr=3.9586e-04) (hash(x)=6199203)
614
+ 588 train 7.165671 (lr=3.9470e-04) (hash(x)=5786498)
615
+ 589 train 7.305725 (lr=3.9355e-04) (hash(x)=8352747)
616
+ 590 train 7.589388 (lr=3.9239e-04) (hash(x)=4829700)
617
+ 591 train 7.135007 (lr=3.9124e-04) (hash(x)=4547106)
618
+ 592 train 7.408190 (lr=3.9008e-04) (hash(x)=7686831)
619
+ 593 train 7.281416 (lr=3.8892e-04) (hash(x)=6433099)
620
+ 594 train 7.147138 (lr=3.8776e-04) (hash(x)=4826460)
621
+ 595 train 7.231416 (lr=3.8660e-04) (hash(x)=6424428)
622
+ 596 train 7.110586 (lr=3.8544e-04) (hash(x)=6073605)
623
+ 597 train 7.457087 (lr=3.8428e-04) (hash(x)=5975712)
624
+ 598 train 7.083070 (lr=3.8311e-04) (hash(x)=7030029)
625
+ 599 train 7.378308 (lr=3.8195e-04) (hash(x)=5458532)
626
+ 600 val loss 7.2907
627
+ 600 val perplexity 1466.6221
628
+ 600 train 7.221822 (lr=3.8078e-04) (hash(x)=7823919)
629
+ 601 train 7.660409 (lr=3.7961e-04) (hash(x)=7002774)
630
+ 602 train 7.456349 (lr=3.7844e-04) (hash(x)=7056399)
631
+ 603 train 7.667455 (lr=3.7727e-04) (hash(x)=8206173)
632
+ 604 train 7.484560 (lr=3.7610e-04) (hash(x)=8034846)
633
+ 605 train 7.794947 (lr=3.7493e-04) (hash(x)=7784971)
634
+ 606 train 8.001055 (lr=3.7376e-04) (hash(x)=7118076)
635
+ 607 train 7.238340 (lr=3.7259e-04) (hash(x)=6001843)
636
+ 608 train 7.328465 (lr=3.7142e-04) (hash(x)=5582437)
637
+ 609 train 7.435714 (lr=3.7024e-04) (hash(x)=5698642)
638
+ 610 train 7.306168 (lr=3.6907e-04) (hash(x)=6615301)
639
+ 611 train 7.306557 (lr=3.6789e-04) (hash(x)=6600751)
640
+ 612 train 7.223703 (lr=3.6671e-04) (hash(x)=6880955)
641
+ 613 train 7.293599 (lr=3.6554e-04) (hash(x)=5804610)
642
+ 614 train 7.200863 (lr=3.6436e-04) (hash(x)=6263347)
643
+ 615 train 7.191039 (lr=3.6318e-04) (hash(x)=5162652)
644
+ 616 train 7.244071 (lr=3.6200e-04) (hash(x)=6055710)
645
+ 617 train 7.204926 (lr=3.6082e-04) (hash(x)=5297006)
646
+ 618 train 7.209697 (lr=3.5964e-04) (hash(x)=5923648)
647
+ 619 train 7.236640 (lr=3.5846e-04) (hash(x)=6728323)
648
+ 620 train 7.244182 (lr=3.5728e-04) (hash(x)=6660239)
649
+ 621 train 7.209693 (lr=3.5610e-04) (hash(x)=3964882)
650
+ 622 train 7.266714 (lr=3.5491e-04) (hash(x)=5868942)
651
+ 623 train 7.194851 (lr=3.5373e-04) (hash(x)=6735993)
652
+ 624 train 7.519069 (lr=3.5255e-04) (hash(x)=5926190)
653
+ 625 train 7.199207 (lr=3.5136e-04) (hash(x)=5150316)
654
+ 626 train 7.213357 (lr=3.5018e-04) (hash(x)=7571467)
655
+ 627 train 7.117115 (lr=3.4899e-04) (hash(x)=5220448)
656
+ 628 train 7.148256 (lr=3.4781e-04) (hash(x)=6790397)
657
+ 629 train 7.605984 (lr=3.4662e-04) (hash(x)=6971289)
658
+ 630 train 7.542478 (lr=3.4544e-04) (hash(x)=7123359)
659
+ 631 train 7.333251 (lr=3.4425e-04) (hash(x)=5012069)
660
+ 632 train 7.304112 (lr=3.4306e-04) (hash(x)=6183574)
661
+ 633 train 7.219798 (lr=3.4188e-04) (hash(x)=6576136)
662
+ 634 train 7.243553 (lr=3.4069e-04) (hash(x)=6157346)
663
+ 635 train 7.246114 (lr=3.3950e-04) (hash(x)=6077647)
664
+ 636 train 7.469715 (lr=3.3831e-04) (hash(x)=8369488)
665
+ 637 train 7.258065 (lr=3.3713e-04) (hash(x)=6032381)
666
+ 638 train 7.215828 (lr=3.3594e-04) (hash(x)=5924756)
667
+ 639 train 7.233202 (lr=3.3475e-04) (hash(x)=8410716)
668
+ 640 train 7.225166 (lr=3.3356e-04) (hash(x)=5803440)
669
+ 641 train 7.311962 (lr=3.3238e-04) (hash(x)=6212516)
670
+ 642 train 7.378681 (lr=3.3119e-04) (hash(x)=7165945)
671
+ 643 train 7.109223 (lr=3.3000e-04) (hash(x)=5298545)
672
+ 644 train 7.354257 (lr=3.2881e-04) (hash(x)=6281739)
673
+ 645 train 7.327255 (lr=3.2762e-04) (hash(x)=5779073)
674
+ 646 train 7.199654 (lr=3.2644e-04) (hash(x)=5090107)
675
+ 647 train 7.245661 (lr=3.2525e-04) (hash(x)=6340947)
676
+ 648 train 6.996275 (lr=3.2406e-04) (hash(x)=4741418)
677
+ 649 train 6.933474 (lr=3.2287e-04) (hash(x)=6530441)
678
+ 650 val loss 7.2791
679
+ 650 val perplexity 1449.6439
680
+ 650 train 7.456884 (lr=3.2169e-04) (hash(x)=6396041)
681
+ 651 train 7.136638 (lr=3.2050e-04) (hash(x)=7818007)
682
+ 652 train 7.147539 (lr=3.1931e-04) (hash(x)=6478457)
683
+ 653 train 7.209266 (lr=3.1812e-04) (hash(x)=6277435)
684
+ 654 train 7.184879 (lr=3.1694e-04) (hash(x)=6042369)
685
+ 655 train 7.198368 (lr=3.1575e-04) (hash(x)=6226320)
686
+ 656 train 7.217704 (lr=3.1456e-04) (hash(x)=5826883)
687
+ 657 train 7.348736 (lr=3.1338e-04) (hash(x)=5463751)
688
+ 658 train 7.266311 (lr=3.1219e-04) (hash(x)=6433080)
689
+ 659 train 7.115076 (lr=3.1101e-04) (hash(x)=6579329)
690
+ 660 train 7.121492 (lr=3.0982e-04) (hash(x)=4724703)
691
+ 661 train 7.460013 (lr=3.0864e-04) (hash(x)=6660540)
692
+ 662 train 7.267281 (lr=3.0745e-04) (hash(x)=5918756)
693
+ 663 train 7.206376 (lr=3.0627e-04) (hash(x)=6491272)
694
+ 664 train 7.130589 (lr=3.0509e-04) (hash(x)=5452863)
695
+ 665 train 7.333635 (lr=3.0390e-04) (hash(x)=6766947)
696
+ 666 train 7.698385 (lr=3.0272e-04) (hash(x)=7459504)
697
+ 667 train 7.339298 (lr=3.0154e-04) (hash(x)=3721218)
698
+ 668 train 7.514933 (lr=3.0036e-04) (hash(x)=7929571)
699
+ 669 train 7.195997 (lr=2.9918e-04) (hash(x)=5229898)
700
+ 670 train 7.219360 (lr=2.9800e-04) (hash(x)=6083090)
701
+ 671 train 7.084272 (lr=2.9682e-04) (hash(x)=6054428)
702
+ 672 train 7.240918 (lr=2.9564e-04) (hash(x)=6072563)
703
+ 673 train 7.347969 (lr=2.9446e-04) (hash(x)=6290464)
704
+ 674 train 7.423286 (lr=2.9329e-04) (hash(x)=6314742)
705
+ 675 train 7.103764 (lr=2.9211e-04) (hash(x)=6904498)
706
+ 676 train 7.284342 (lr=2.9093e-04) (hash(x)=5378107)
707
+ 677 train 7.379695 (lr=2.8976e-04) (hash(x)=7765805)
708
+ 678 train 7.209647 (lr=2.8858e-04) (hash(x)=5871280)
709
+ 679 train 7.352811 (lr=2.8741e-04) (hash(x)=6543764)
710
+ 680 train 7.244007 (lr=2.8624e-04) (hash(x)=7369665)
711
+ 681 train 7.327340 (lr=2.8507e-04) (hash(x)=7204955)
712
+ 682 train 7.268192 (lr=2.8390e-04) (hash(x)=5856543)
713
+ 683 train 7.142976 (lr=2.8273e-04) (hash(x)=4423711)
714
+ 684 train 7.114275 (lr=2.8156e-04) (hash(x)=5818153)
715
+ 685 train 7.214947 (lr=2.8039e-04) (hash(x)=5343477)
716
+ 686 train 7.409268 (lr=2.7922e-04) (hash(x)=6581702)
717
+ 687 train 7.236282 (lr=2.7805e-04) (hash(x)=6595856)
718
+ 688 train 7.174536 (lr=2.7689e-04) (hash(x)=5325378)
719
+ 689 train 7.094116 (lr=2.7572e-04) (hash(x)=7229015)
720
+ 690 train 7.262506 (lr=2.7456e-04) (hash(x)=5918186)
721
+ 691 train 7.146507 (lr=2.7340e-04) (hash(x)=5916034)
722
+ 692 train 7.164917 (lr=2.7224e-04) (hash(x)=5280722)
723
+ 693 train 7.570690 (lr=2.7108e-04) (hash(x)=6939521)
724
+ 694 train 7.298468 (lr=2.6992e-04) (hash(x)=5172566)
725
+ 695 train 7.147656 (lr=2.6876e-04) (hash(x)=5492185)
726
+ 696 train 7.028531 (lr=2.6761e-04) (hash(x)=5779011)
727
+ 697 train 7.283820 (lr=2.6645e-04) (hash(x)=6898006)
728
+ 698 train 7.136384 (lr=2.6530e-04) (hash(x)=5779217)
729
+ 699 train 7.003380 (lr=2.6414e-04) (hash(x)=5462008)
730
+ 700 val loss 7.2763
731
+ 700 val perplexity 1445.6913
732
+ 700 train 7.133677 (lr=2.6299e-04) (hash(x)=6243843)
733
+ 701 train 7.201668 (lr=2.6184e-04) (hash(x)=7821902)
734
+ 702 train 7.263968 (lr=2.6069e-04) (hash(x)=5222501)
735
+ 703 train 7.229538 (lr=2.5955e-04) (hash(x)=6608108)
736
+ 704 train 7.353471 (lr=2.5840e-04) (hash(x)=7602879)
737
+ 705 train 7.283667 (lr=2.5725e-04) (hash(x)=6295394)
738
+ 706 train 7.240102 (lr=2.5611e-04) (hash(x)=6579907)
739
+ 707 train 7.131851 (lr=2.5497e-04) (hash(x)=6803449)
740
+ 708 train 7.206299 (lr=2.5383e-04) (hash(x)=5413733)
741
+ 709 train 7.156231 (lr=2.5269e-04) (hash(x)=7444134)
742
+ 710 train 7.320212 (lr=2.5155e-04) (hash(x)=7028652)
743
+ 711 train 7.103862 (lr=2.5042e-04) (hash(x)=5361923)
744
+ 712 train 7.224922 (lr=2.4928e-04) (hash(x)=5478293)
745
+ 713 train 7.201499 (lr=2.4815e-04) (hash(x)=7803039)
746
+ 714 train 7.387799 (lr=2.4702e-04) (hash(x)=6469017)
747
+ 715 train 7.281194 (lr=2.4589e-04) (hash(x)=7042578)
748
+ 716 train 7.329437 (lr=2.4476e-04) (hash(x)=6927957)
749
+ 717 train 7.153640 (lr=2.4363e-04) (hash(x)=6647007)
750
+ 718 train 7.122975 (lr=2.4251e-04) (hash(x)=6230888)
751
+ 719 train 7.378871 (lr=2.4139e-04) (hash(x)=6303658)
752
+ 720 train 7.415677 (lr=2.4026e-04) (hash(x)=6769880)
753
+ 721 train 7.258996 (lr=2.3914e-04) (hash(x)=6158927)
754
+ 722 train 7.277626 (lr=2.3803e-04) (hash(x)=5991631)
755
+ 723 train 7.130712 (lr=2.3691e-04) (hash(x)=6419656)
756
+ 724 train 7.158002 (lr=2.3580e-04) (hash(x)=7739209)
757
+ 725 train 7.116948 (lr=2.3468e-04) (hash(x)=7891271)
758
+ 726 train 7.466895 (lr=2.3357e-04) (hash(x)=7171433)
759
+ 727 train 7.247159 (lr=2.3246e-04) (hash(x)=7099888)
760
+ 728 train 7.271505 (lr=2.3136e-04) (hash(x)=6103730)
761
+ 729 train 7.385959 (lr=2.3025e-04) (hash(x)=6352424)
762
+ 730 train 7.266267 (lr=2.2915e-04) (hash(x)=4893671)
763
+ 731 train 7.211841 (lr=2.2805e-04) (hash(x)=5863145)
764
+ 732 train 7.324223 (lr=2.2695e-04) (hash(x)=5502233)
765
+ 733 train 7.219761 (lr=2.2585e-04) (hash(x)=6844597)
766
+ 734 train 7.230278 (lr=2.2476e-04) (hash(x)=6744526)
767
+ 735 train 7.369040 (lr=2.2366e-04) (hash(x)=6009861)
768
+ 736 train 7.245226 (lr=2.2257e-04) (hash(x)=6152358)
769
+ 737 train 7.159678 (lr=2.2148e-04) (hash(x)=5374957)
770
+ 738 train 7.172276 (lr=2.2040e-04) (hash(x)=6495309)
771
+ 739 train 7.342313 (lr=2.1931e-04) (hash(x)=3923225)
772
+ 740 train 7.127966 (lr=2.1823e-04) (hash(x)=6255243)
773
+ 741 train 7.260824 (lr=2.1715e-04) (hash(x)=5848862)
774
+ 742 train 7.277327 (lr=2.1607e-04) (hash(x)=4535000)
775
+ 743 train 7.176571 (lr=2.1500e-04) (hash(x)=4440634)
776
+ 744 train 7.163837 (lr=2.1392e-04) (hash(x)=5222077)
777
+ 745 train 7.177392 (lr=2.1285e-04) (hash(x)=4788998)
778
+ 746 train 7.183031 (lr=2.1178e-04) (hash(x)=6108350)
779
+ 747 train 7.096934 (lr=2.1072e-04) (hash(x)=7378165)
780
+ 748 train 7.179675 (lr=2.0965e-04) (hash(x)=5475646)
781
+ 749 train 7.192242 (lr=2.0859e-04) (hash(x)=4771243)
782
+ 750 val loss 7.2765
783
+ 750 val perplexity 1445.8718
784
+ 750 train 7.071801 (lr=2.0753e-04) (hash(x)=6668088)
785
+ 751 train 7.283248 (lr=2.0647e-04) (hash(x)=5716089)
786
+ 752 train 7.241159 (lr=2.0542e-04) (hash(x)=6991946)
787
+ 753 train 7.142903 (lr=2.0436e-04) (hash(x)=7991009)
788
+ 754 train 7.257631 (lr=2.0331e-04) (hash(x)=6037018)
789
+ 755 train 7.224583 (lr=2.0226e-04) (hash(x)=5348607)
790
+ 756 train 7.348554 (lr=2.0122e-04) (hash(x)=5190526)
791
+ 757 train 7.239732 (lr=2.0018e-04) (hash(x)=6036146)
792
+ 758 train 7.129354 (lr=1.9914e-04) (hash(x)=5329783)
793
+ 759 train 7.226395 (lr=1.9810e-04) (hash(x)=6658718)
794
+ 760 train 7.139090 (lr=1.9706e-04) (hash(x)=4820939)
795
+ 761 train 7.118741 (lr=1.9603e-04) (hash(x)=6176253)
796
+ 762 train 7.102529 (lr=1.9500e-04) (hash(x)=5316750)
797
+ 763 train 7.180678 (lr=1.9397e-04) (hash(x)=4542895)
798
+ 764 train 7.026456 (lr=1.9295e-04) (hash(x)=5395673)
799
+ 765 train 7.069162 (lr=1.9193e-04) (hash(x)=5549073)
800
+ 766 train 7.290985 (lr=1.9091e-04) (hash(x)=5233723)
801
+ 767 train 7.365542 (lr=1.8989e-04) (hash(x)=5342783)
802
+ 768 train 7.304257 (lr=1.8887e-04) (hash(x)=5552448)
803
+ 769 train 7.426475 (lr=1.8786e-04) (hash(x)=6549546)
804
+ 770 train 7.121472 (lr=1.8685e-04) (hash(x)=3816493)
805
+ 771 train 7.160312 (lr=1.8585e-04) (hash(x)=6170599)
806
+ 772 train 7.608524 (lr=1.8485e-04) (hash(x)=6708759)
807
+ 773 train 7.244777 (lr=1.8385e-04) (hash(x)=8483134)
808
+ 774 train 7.804720 (lr=1.8285e-04) (hash(x)=5583732)
809
+ 775 train 7.228591 (lr=1.8185e-04) (hash(x)=4865036)
810
+ 776 train 7.161191 (lr=1.8086e-04) (hash(x)=7234336)
811
+ 777 train 7.200143 (lr=1.7987e-04) (hash(x)=4433890)
812
+ 778 train 7.512131 (lr=1.7889e-04) (hash(x)=7484636)
813
+ 779 train 7.926641 (lr=1.7790e-04) (hash(x)=6738597)
814
+ 780 train 7.212890 (lr=1.7692e-04) (hash(x)=6752421)
815
+ 781 train 7.392477 (lr=1.7595e-04) (hash(x)=5459620)
816
+ 782 train 7.297678 (lr=1.7497e-04) (hash(x)=6718131)
817
+ 783 train 7.289261 (lr=1.7400e-04) (hash(x)=4604074)
818
+ 784 train 7.152444 (lr=1.7303e-04) (hash(x)=5946124)
819
+ 785 train 7.419847 (lr=1.7207e-04) (hash(x)=8045501)
820
+ 786 train 7.536399 (lr=1.7111e-04) (hash(x)=5712725)
821
+ 787 train 7.256878 (lr=1.7015e-04) (hash(x)=6176464)
822
+ 788 train 7.292347 (lr=1.6919e-04) (hash(x)=6276702)
823
+ 789 train 7.115613 (lr=1.6824e-04) (hash(x)=6127317)
824
+ 790 train 7.273091 (lr=1.6729e-04) (hash(x)=7221374)
825
+ 791 train 7.248471 (lr=1.6634e-04) (hash(x)=5571462)
826
+ 792 train 7.249466 (lr=1.6540e-04) (hash(x)=7518021)
827
+ 793 train 7.248608 (lr=1.6446e-04) (hash(x)=6640755)
828
+ 794 train 7.330340 (lr=1.6352e-04) (hash(x)=6870405)
829
+ 795 train 7.290612 (lr=1.6259e-04) (hash(x)=6147295)
830
+ 796 train 7.187652 (lr=1.6166e-04) (hash(x)=6782583)
831
+ 797 train 7.304230 (lr=1.6073e-04) (hash(x)=6120461)
832
+ 798 train 7.238538 (lr=1.5981e-04) (hash(x)=5893870)
833
+ 799 train 7.175369 (lr=1.5889e-04) (hash(x)=5123928)
834
+ 800 val loss 7.2628
835
+ 800 val perplexity 1426.1842
836
+ 800 train 7.238575 (lr=1.5797e-04) (hash(x)=7037251)
837
+ 801 train 7.240525 (lr=1.5705e-04) (hash(x)=5428813)
838
+ 802 train 7.165471 (lr=1.5614e-04) (hash(x)=5550387)
839
+ 803 train 7.245495 (lr=1.5524e-04) (hash(x)=6448586)
840
+ 804 train 7.384059 (lr=1.5433e-04) (hash(x)=5916934)
841
+ 805 train 7.335072 (lr=1.5343e-04) (hash(x)=8116987)
842
+ 806 train 7.318943 (lr=1.5254e-04) (hash(x)=6210899)
843
+ 807 train 7.430175 (lr=1.5164e-04) (hash(x)=5974342)
844
+ 808 train 7.420037 (lr=1.5075e-04) (hash(x)=7236199)
845
+ 809 train 7.347103 (lr=1.4987e-04) (hash(x)=5261336)
846
+ 810 train 7.184433 (lr=1.4898e-04) (hash(x)=6310720)
847
+ 811 train 7.213087 (lr=1.4810e-04) (hash(x)=5307056)
848
+ 812 train 7.075014 (lr=1.4723e-04) (hash(x)=7356633)
849
+ 813 train 7.190030 (lr=1.4635e-04) (hash(x)=5865119)
850
+ 814 train 7.088152 (lr=1.4548e-04) (hash(x)=5979205)
851
+ 815 train 7.239465 (lr=1.4462e-04) (hash(x)=6672380)
852
+ 816 train 7.222394 (lr=1.4376e-04) (hash(x)=5565030)
853
+ 817 train 7.206617 (lr=1.4290e-04) (hash(x)=6886810)
854
+ 818 train 7.270175 (lr=1.4204e-04) (hash(x)=5926478)
855
+ 819 train 7.487834 (lr=1.4119e-04) (hash(x)=6564071)
856
+ 820 train 7.114408 (lr=1.4035e-04) (hash(x)=5024280)
857
+ 821 train 7.104883 (lr=1.3950e-04) (hash(x)=7705089)
858
+ 822 train 7.086552 (lr=1.3866e-04) (hash(x)=4037371)
859
+ 823 train 7.180008 (lr=1.3783e-04) (hash(x)=5773073)
860
+ 824 train 7.194257 (lr=1.3699e-04) (hash(x)=7158139)
861
+ 825 train 7.046433 (lr=1.3616e-04) (hash(x)=5905976)
862
+ 826 train 7.471286 (lr=1.3534e-04) (hash(x)=13897369)
863
+ 827 train 8.557480 (lr=1.3452e-04) (hash(x)=6936653)
864
+ 828 train 7.230253 (lr=1.3370e-04) (hash(x)=6592246)
865
+ 829 train 7.204882 (lr=1.3289e-04) (hash(x)=6592773)
866
+ 830 train 7.249508 (lr=1.3208e-04) (hash(x)=5461178)
867
+ 831 train 7.038263 (lr=1.3127e-04) (hash(x)=6796551)
868
+ 832 train 7.265030 (lr=1.3047e-04) (hash(x)=4721691)
869
+ 833 train 7.162796 (lr=1.2967e-04) (hash(x)=6151589)
870
+ 834 train 7.321404 (lr=1.2887e-04) (hash(x)=7008995)
871
+ 835 train 7.637057 (lr=1.2808e-04) (hash(x)=7045164)
872
+ 836 train 7.237843 (lr=1.2730e-04) (hash(x)=6614221)
873
+ 837 train 7.277291 (lr=1.2651e-04) (hash(x)=6650506)
874
+ 838 train 7.078917 (lr=1.2574e-04) (hash(x)=6363762)
875
+ 839 train 7.249407 (lr=1.2496e-04) (hash(x)=7839748)
876
+ 840 train 7.154269 (lr=1.2419e-04) (hash(x)=5124077)
877
+ 841 train 7.618376 (lr=1.2342e-04) (hash(x)=6028671)
878
+ 842 train 7.123517 (lr=1.2266e-04) (hash(x)=5427175)
879
+ 843 train 7.441318 (lr=1.2190e-04) (hash(x)=6118408)
880
+ 844 train 7.220417 (lr=1.2115e-04) (hash(x)=6022552)
881
+ 845 train 7.221863 (lr=1.2040e-04) (hash(x)=5757526)
882
+ 846 train 7.683441 (lr=1.1965e-04) (hash(x)=6476549)
883
+ 847 train 7.029605 (lr=1.1891e-04) (hash(x)=6030172)
884
+ 848 train 7.395398 (lr=1.1817e-04) (hash(x)=5917229)
885
+ 849 train 7.497690 (lr=1.1743e-04) (hash(x)=6719882)
886
+ 850 val loss 7.2594
887
+ 850 val perplexity 1421.3531
888
+ 850 train 7.322224 (lr=1.1670e-04) (hash(x)=5441498)
889
+ 851 train 7.221459 (lr=1.1598e-04) (hash(x)=5028050)
890
+ 852 train 7.235754 (lr=1.1525e-04) (hash(x)=8487491)
891
+ 853 train 7.227039 (lr=1.1454e-04) (hash(x)=5682910)
892
+ 854 train 7.414336 (lr=1.1382e-04) (hash(x)=6781362)
893
+ 855 train 7.230700 (lr=1.1311e-04) (hash(x)=5747891)
894
+ 856 train 7.278474 (lr=1.1241e-04) (hash(x)=7627661)
895
+ 857 train 6.966769 (lr=1.1171e-04) (hash(x)=5734208)
896
+ 858 train 7.106152 (lr=1.1101e-04) (hash(x)=6162015)
897
+ 859 train 7.304794 (lr=1.1032e-04) (hash(x)=7024381)
898
+ 860 train 7.513941 (lr=1.0963e-04) (hash(x)=5198498)
899
+ 861 train 7.495351 (lr=1.0894e-04) (hash(x)=7814361)
900
+ 862 train 7.365499 (lr=1.0826e-04) (hash(x)=7394995)
901
+ 863 train 7.259632 (lr=1.0759e-04) (hash(x)=5765893)
902
+ 864 train 7.064799 (lr=1.0692e-04) (hash(x)=4990435)
903
+ 865 train 7.197665 (lr=1.0625e-04) (hash(x)=5572048)
904
+ 866 train 7.365871 (lr=1.0559e-04) (hash(x)=7093293)
905
+ 867 train 7.245916 (lr=1.0493e-04) (hash(x)=5610058)
906
+ 868 train 7.142777 (lr=1.0427e-04) (hash(x)=6771468)
907
+ 869 train 7.187006 (lr=1.0362e-04) (hash(x)=4459483)
908
+ 870 train 7.074708 (lr=1.0298e-04) (hash(x)=4926546)
909
+ 871 train 7.193995 (lr=1.0234e-04) (hash(x)=6758411)
910
+ 872 train 7.149498 (lr=1.0170e-04) (hash(x)=7394277)
911
+ 873 train 7.278939 (lr=1.0107e-04) (hash(x)=6701381)
912
+ 874 train 7.607490 (lr=1.0044e-04) (hash(x)=8120966)
913
+ 875 train 7.140362 (lr=9.9818e-05) (hash(x)=5416400)
914
+ 876 train 7.140898 (lr=9.9199e-05) (hash(x)=6292778)
915
+ 877 train 7.078300 (lr=9.8585e-05) (hash(x)=5919569)
916
+ 878 train 7.399718 (lr=9.7975e-05) (hash(x)=7561740)
917
+ 879 train 7.143291 (lr=9.7370e-05) (hash(x)=5282870)
918
+ 880 train 7.186507 (lr=9.6769e-05) (hash(x)=5473238)
919
+ 881 train 7.086699 (lr=9.6173e-05) (hash(x)=8072181)
920
+ 882 train 7.101425 (lr=9.5581e-05) (hash(x)=6757322)
921
+ 883 train 7.067219 (lr=9.4994e-05) (hash(x)=7996770)
922
+ 884 train 7.166949 (lr=9.4412e-05) (hash(x)=3181021)
923
+ 885 train 7.211201 (lr=9.3833e-05) (hash(x)=5093519)
924
+ 886 train 7.172001 (lr=9.3260e-05) (hash(x)=5391352)
925
+ 887 train 7.211829 (lr=9.2691e-05) (hash(x)=6270917)
926
+ 888 train 7.215734 (lr=9.2127e-05) (hash(x)=5818195)
927
+ 889 train 6.975936 (lr=9.1567e-05) (hash(x)=5852780)
928
+ 890 train 6.901105 (lr=9.1012e-05) (hash(x)=5777928)
929
+ 891 train 7.089032 (lr=9.0461e-05) (hash(x)=5327728)
930
+ 892 train 7.121693 (lr=8.9916e-05) (hash(x)=7651281)
931
+ 893 train 7.317436 (lr=8.9374e-05) (hash(x)=6260765)
932
+ 894 train 7.196193 (lr=8.8838e-05) (hash(x)=6265102)
933
+ 895 train 7.354897 (lr=8.8306e-05) (hash(x)=6212894)
934
+ 896 train 7.137350 (lr=8.7779e-05) (hash(x)=6901967)
935
+ 897 train 7.201009 (lr=8.7256e-05) (hash(x)=6724010)
936
+ 898 train 7.199869 (lr=8.6738e-05) (hash(x)=6299033)
937
+ 899 train 7.367983 (lr=8.6225e-05) (hash(x)=6059934)
938
+ 900 val loss 7.2706
939
+ 900 val perplexity 1437.3446
940
+ 900 train 7.058852 (lr=8.5717e-05) (hash(x)=6029696)
941
+ 901 train 6.981483 (lr=8.5213e-05) (hash(x)=5587382)
942
+ 902 train 7.180746 (lr=8.4714e-05) (hash(x)=5890076)
943
+ 903 train 7.108450 (lr=8.4220e-05) (hash(x)=7274108)
944
+ 904 train 7.201891 (lr=8.3731e-05) (hash(x)=7975528)
945
+ 905 train 7.241421 (lr=8.3246e-05) (hash(x)=6085396)
946
+ 906 train 7.133669 (lr=8.2766e-05) (hash(x)=6724496)
947
+ 907 train 7.130234 (lr=8.2291e-05) (hash(x)=4784666)
948
+ 908 train 7.145245 (lr=8.1821e-05) (hash(x)=7193845)
949
+ 909 train 7.300814 (lr=8.1356e-05) (hash(x)=6288638)
950
+ 910 train 7.286870 (lr=8.0895e-05) (hash(x)=5685285)
951
+ 911 train 7.265322 (lr=8.0439e-05) (hash(x)=7714288)
952
+ 912 train 7.338243 (lr=7.9988e-05) (hash(x)=5983679)
953
+ 913 train 7.293921 (lr=7.9542e-05) (hash(x)=5534748)
954
+ 914 train 7.239863 (lr=7.9101e-05) (hash(x)=6545471)
955
+ 915 train 7.222851 (lr=7.8664e-05) (hash(x)=6361270)
956
+ 916 train 7.217522 (lr=7.8232e-05) (hash(x)=6257744)
957
+ 917 train 7.232538 (lr=7.7806e-05) (hash(x)=4682956)
958
+ 918 train 7.116433 (lr=7.7384e-05) (hash(x)=4521756)
959
+ 919 train 7.261102 (lr=7.6967e-05) (hash(x)=7568935)
960
+ 920 train 7.169036 (lr=7.6555e-05) (hash(x)=7969353)
961
+ 921 train 7.351889 (lr=7.6148e-05) (hash(x)=7973311)
962
+ 922 train 7.057231 (lr=7.5746e-05) (hash(x)=4843937)
963
+ 923 train 7.203022 (lr=7.5348e-05) (hash(x)=6987352)
964
+ 924 train 7.316325 (lr=7.4956e-05) (hash(x)=7179418)
965
+ 925 train 6.981483 (lr=7.4568e-05) (hash(x)=5413904)
966
+ 926 train 7.367393 (lr=7.4186e-05) (hash(x)=6028105)
967
+ 927 train 7.275677 (lr=7.3808e-05) (hash(x)=6732924)
968
+ 928 train 7.167045 (lr=7.3436e-05) (hash(x)=4547651)
969
+ 929 train 7.338727 (lr=7.3068e-05) (hash(x)=6219754)
970
+ 930 train 7.117935 (lr=7.2706e-05) (hash(x)=7465186)
971
+ 931 train 7.254127 (lr=7.2348e-05) (hash(x)=6839070)
972
+ 932 train 7.124338 (lr=7.1995e-05) (hash(x)=5390324)
973
+ 933 train 7.100323 (lr=7.1648e-05) (hash(x)=6090120)
974
+ 934 train 7.719783 (lr=7.1305e-05) (hash(x)=7123181)
975
+ 935 train 7.401228 (lr=7.0967e-05) (hash(x)=7115176)
976
+ 936 train 7.405222 (lr=7.0635e-05) (hash(x)=7016990)
977
+ 937 train 7.167251 (lr=7.0307e-05) (hash(x)=4989299)
978
+ 938 train 7.187303 (lr=6.9984e-05) (hash(x)=6783311)
979
+ 939 train 7.153462 (lr=6.9667e-05) (hash(x)=7497590)
980
+ 940 train 7.370980 (lr=6.9354e-05) (hash(x)=6767514)
981
+ 941 train 7.134829 (lr=6.9047e-05) (hash(x)=3977557)
982
+ 942 train 7.244173 (lr=6.8744e-05) (hash(x)=4806492)
983
+ 943 train 7.177331 (lr=6.8447e-05) (hash(x)=5524165)
984
+ 944 train 7.236280 (lr=6.8155e-05) (hash(x)=4930752)
985
+ 945 train 7.208829 (lr=6.7868e-05) (hash(x)=5124936)
986
+ 946 train 7.091607 (lr=6.7585e-05) (hash(x)=6877351)
987
+ 947 train 7.375976 (lr=6.7308e-05) (hash(x)=5522119)
988
+ 948 train 7.168471 (lr=6.7036e-05) (hash(x)=6183512)
989
+ 949 train 7.099660 (lr=6.6769e-05) (hash(x)=5030865)
990
+ 950 val loss 7.2418
991
+ 950 val perplexity 1396.6221
992
+ 950 train 7.184534 (lr=6.6508e-05) (hash(x)=6619812)
993
+ 951 train 7.195469 (lr=6.6251e-05) (hash(x)=6411783)
994
+ 952 train 7.184252 (lr=6.5999e-05) (hash(x)=6748982)
995
+ 953 train 7.088627 (lr=6.5753e-05) (hash(x)=6581937)
996
+ 954 train 7.188861 (lr=6.5511e-05) (hash(x)=7881022)
997
+ 955 train 7.157289 (lr=6.5275e-05) (hash(x)=5116910)
998
+ 956 train 6.949575 (lr=6.5044e-05) (hash(x)=4880833)
999
+ 957 train 6.997415 (lr=6.4818e-05) (hash(x)=7353459)
1000
+ 958 train 7.074324 (lr=6.4597e-05) (hash(x)=4951302)
1001
+ 959 train 6.977889 (lr=6.4382e-05) (hash(x)=6344535)
1002
+ 960 train 7.595117 (lr=6.4171e-05) (hash(x)=7411763)
1003
+ 961 train 7.153414 (lr=6.3966e-05) (hash(x)=5840169)
1004
+ 962 train 7.257590 (lr=6.3765e-05) (hash(x)=5784788)
1005
+ 963 train 7.068092 (lr=6.3570e-05) (hash(x)=6754290)
1006
+ 964 train 7.341851 (lr=6.3380e-05) (hash(x)=6369963)
1007
+ 965 train 7.421086 (lr=6.3195e-05) (hash(x)=6484573)
1008
+ 966 train 7.015618 (lr=6.3016e-05) (hash(x)=6516338)
1009
+ 967 train 7.039011 (lr=6.2841e-05) (hash(x)=5260125)
1010
+ 968 train 7.170541 (lr=6.2672e-05) (hash(x)=5781624)
1011
+ 969 train 6.965876 (lr=6.2508e-05) (hash(x)=4881179)
1012
+ 970 train 7.078770 (lr=6.2349e-05) (hash(x)=6707821)
1013
+ 971 train 7.228146 (lr=6.2195e-05) (hash(x)=5001634)
1014
+ 972 train 7.092967 (lr=6.2046e-05) (hash(x)=4402949)
1015
+ 973 train 7.117911 (lr=6.1903e-05) (hash(x)=7424810)
1016
+ 974 train 7.009636 (lr=6.1765e-05) (hash(x)=4814803)
1017
+ 975 train 6.924478 (lr=6.1632e-05) (hash(x)=5176431)
1018
+ 976 train 7.129508 (lr=6.1504e-05) (hash(x)=5769393)
1019
+ 977 train 7.071016 (lr=6.1381e-05) (hash(x)=5421624)
1020
+ 978 train 7.266103 (lr=6.1264e-05) (hash(x)=5999773)
1021
+ 979 train 6.982728 (lr=6.1152e-05) (hash(x)=5674439)
1022
+ 980 train 7.075795 (lr=6.1045e-05) (hash(x)=6345351)
1023
+ 981 train 7.013047 (lr=6.0943e-05) (hash(x)=5819157)
1024
+ 982 train 7.228723 (lr=6.0846e-05) (hash(x)=11159355)
1025
+ 983 train 7.189655 (lr=6.0755e-05) (hash(x)=6506856)
1026
+ 984 train 7.098700 (lr=6.0669e-05) (hash(x)=6331023)
1027
+ 985 train 7.012566 (lr=6.0588e-05) (hash(x)=6119124)
1028
+ 986 train 7.222008 (lr=6.0512e-05) (hash(x)=8142528)
1029
+ 987 train 7.398296 (lr=6.0442e-05) (hash(x)=6241669)
1030
+ 988 train 7.019189 (lr=6.0376e-05) (hash(x)=3852991)
1031
+ 989 train 6.764310 (lr=6.0316e-05) (hash(x)=4992756)
1032
+ 990 train 7.469626 (lr=6.0261e-05) (hash(x)=7097145)
1033
+ 991 train 7.224168 (lr=6.0212e-05) (hash(x)=7277927)
1034
+ 992 train 7.309268 (lr=6.0167e-05) (hash(x)=5661400)
1035
+ 993 train 7.233738 (lr=6.0128e-05) (hash(x)=6622097)
1036
+ 994 train 7.360166 (lr=6.0094e-05) (hash(x)=5016804)
1037
+ 995 train 7.464649 (lr=6.0065e-05) (hash(x)=7026582)
1038
+ 996 train 7.222393 (lr=6.0042e-05) (hash(x)=5639136)
1039
+ 997 train 7.373614 (lr=6.0024e-05) (hash(x)=5840987)
1040
+ 998 train 6.998946 (lr=6.0010e-05) (hash(x)=5265410)
1041
+ 999 val loss 7.2390
1042
+ 999 val perplexity 1392.6698
1043
+ 999 train 6.987595 (lr=6.0003e-05) (hash(x)=4351074)
lr6e-4_total_batch_size20480_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8fd1351b9402e5d451275fcc7123642c74a1c85ab783e5d90bee901b44bfbe00
3
+ size 36163010
lr6e-4_total_batch_size20480_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1ed05284f556dc76f4ac1ff8fb0061a10f172a28287b94230c270990fbccdf0
3
+ size 70764358