andrew-healey commited on
Commit
ca99533
·
verified ·
1 Parent(s): 8a337aa

Upload folder using huggingface_hub

Browse files
lr8e-4_total_batch_size20480_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr8e-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": 0.0008, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr8e-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
lr8e-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=2.7972e-06) (hash(x)=5113283)
5
+ 1 train 11.681684 (lr=5.5944e-06) (hash(x)=5534142)
6
+ 2 train 11.678484 (lr=8.3916e-06) (hash(x)=6763231)
7
+ 3 train 11.641376 (lr=1.1189e-05) (hash(x)=6328012)
8
+ 4 train 11.611877 (lr=1.3986e-05) (hash(x)=7052161)
9
+ 5 train 11.566850 (lr=1.6783e-05) (hash(x)=5556098)
10
+ 6 train 11.518194 (lr=1.9580e-05) (hash(x)=4696943)
11
+ 7 train 11.435740 (lr=2.2378e-05) (hash(x)=6219138)
12
+ 8 train 11.353488 (lr=2.5175e-05) (hash(x)=6049878)
13
+ 9 train 11.301502 (lr=2.7972e-05) (hash(x)=5491972)
14
+ 10 train 11.206249 (lr=3.0769e-05) (hash(x)=8075458)
15
+ 11 train 11.080152 (lr=3.3566e-05) (hash(x)=5917741)
16
+ 12 train 11.015227 (lr=3.6364e-05) (hash(x)=5750403)
17
+ 13 train 10.945143 (lr=3.9161e-05) (hash(x)=6112458)
18
+ 14 train 10.858475 (lr=4.1958e-05) (hash(x)=5806490)
19
+ 15 train 10.748863 (lr=4.4755e-05) (hash(x)=6568379)
20
+ 16 train 10.646448 (lr=4.7552e-05) (hash(x)=7212403)
21
+ 17 train 10.580310 (lr=5.0350e-05) (hash(x)=7386082)
22
+ 18 train 10.544984 (lr=5.3147e-05) (hash(x)=6359743)
23
+ 19 train 10.486224 (lr=5.5944e-05) (hash(x)=6151674)
24
+ 20 train 10.478385 (lr=5.8741e-05) (hash(x)=7168704)
25
+ 21 train 10.442904 (lr=6.1538e-05) (hash(x)=7859060)
26
+ 22 train 10.375811 (lr=6.4336e-05) (hash(x)=6976239)
27
+ 23 train 10.381527 (lr=6.7133e-05) (hash(x)=6391506)
28
+ 24 train 10.379046 (lr=6.9930e-05) (hash(x)=5245999)
29
+ 25 train 10.336301 (lr=7.2727e-05) (hash(x)=5985675)
30
+ 26 train 10.250107 (lr=7.5524e-05) (hash(x)=4355733)
31
+ 27 train 10.433285 (lr=7.8322e-05) (hash(x)=6984772)
32
+ 28 train 10.267898 (lr=8.1119e-05) (hash(x)=6244096)
33
+ 29 train 10.289906 (lr=8.3916e-05) (hash(x)=6603717)
34
+ 30 train 10.189299 (lr=8.6713e-05) (hash(x)=5669522)
35
+ 31 train 10.165399 (lr=8.9510e-05) (hash(x)=6190579)
36
+ 32 train 10.219600 (lr=9.2308e-05) (hash(x)=4912983)
37
+ 33 train 10.084324 (lr=9.5105e-05) (hash(x)=7276303)
38
+ 34 train 10.141816 (lr=9.7902e-05) (hash(x)=5479947)
39
+ 35 train 10.234128 (lr=1.0070e-04) (hash(x)=6630378)
40
+ 36 train 10.146675 (lr=1.0350e-04) (hash(x)=8097010)
41
+ 37 train 10.198082 (lr=1.0629e-04) (hash(x)=9989744)
42
+ 38 train 10.058139 (lr=1.0909e-04) (hash(x)=6694932)
43
+ 39 train 9.965734 (lr=1.1189e-04) (hash(x)=6317894)
44
+ 40 train 9.854935 (lr=1.1469e-04) (hash(x)=5046327)
45
+ 41 train 9.911474 (lr=1.1748e-04) (hash(x)=7947703)
46
+ 42 train 9.749743 (lr=1.2028e-04) (hash(x)=5916068)
47
+ 43 train 9.834318 (lr=1.2308e-04) (hash(x)=5777498)
48
+ 44 train 9.737040 (lr=1.2587e-04) (hash(x)=5753092)
49
+ 45 train 9.698650 (lr=1.2867e-04) (hash(x)=7723226)
50
+ 46 train 9.652713 (lr=1.3147e-04) (hash(x)=7007016)
51
+ 47 train 9.553020 (lr=1.3427e-04) (hash(x)=6491229)
52
+ 48 train 9.561281 (lr=1.3706e-04) (hash(x)=6390752)
53
+ 49 train 9.660604 (lr=1.3986e-04) (hash(x)=6019389)
54
+ 50 val loss 9.4925
55
+ 50 val perplexity 13259.3516
56
+ 50 train 9.492965 (lr=1.4266e-04) (hash(x)=7662067)
57
+ 51 train 9.459835 (lr=1.4545e-04) (hash(x)=4648609)
58
+ 52 train 9.352439 (lr=1.4825e-04) (hash(x)=4725966)
59
+ 53 train 9.276791 (lr=1.5105e-04) (hash(x)=5558355)
60
+ 54 train 9.295496 (lr=1.5385e-04) (hash(x)=6524933)
61
+ 55 train 9.215850 (lr=1.5664e-04) (hash(x)=6369326)
62
+ 56 train 9.152513 (lr=1.5944e-04) (hash(x)=6022625)
63
+ 57 train 9.048554 (lr=1.6224e-04) (hash(x)=4906853)
64
+ 58 train 8.830503 (lr=1.6503e-04) (hash(x)=5589269)
65
+ 59 train 8.943124 (lr=1.6783e-04) (hash(x)=6437376)
66
+ 60 train 8.991794 (lr=1.7063e-04) (hash(x)=7304892)
67
+ 61 train 8.924947 (lr=1.7343e-04) (hash(x)=7084093)
68
+ 62 train 8.917912 (lr=1.7622e-04) (hash(x)=5908178)
69
+ 63 train 8.676782 (lr=1.7902e-04) (hash(x)=7048804)
70
+ 64 train 8.735821 (lr=1.8182e-04) (hash(x)=6085549)
71
+ 65 train 8.755760 (lr=1.8462e-04) (hash(x)=5752594)
72
+ 66 train 8.520637 (lr=1.8741e-04) (hash(x)=7662769)
73
+ 67 train 8.382345 (lr=1.9021e-04) (hash(x)=6393520)
74
+ 68 train 8.623681 (lr=1.9301e-04) (hash(x)=5942867)
75
+ 69 train 8.613140 (lr=1.9580e-04) (hash(x)=6793550)
76
+ 70 train 8.412848 (lr=1.9860e-04) (hash(x)=7105976)
77
+ 71 train 8.442006 (lr=2.0140e-04) (hash(x)=6581348)
78
+ 72 train 8.330273 (lr=2.0420e-04) (hash(x)=5555598)
79
+ 73 train 8.318186 (lr=2.0699e-04) (hash(x)=6235837)
80
+ 74 train 8.239014 (lr=2.0979e-04) (hash(x)=5591770)
81
+ 75 train 8.499212 (lr=2.1259e-04) (hash(x)=7083021)
82
+ 76 train 8.323892 (lr=2.1538e-04) (hash(x)=7281348)
83
+ 77 train 8.397629 (lr=2.1818e-04) (hash(x)=6155821)
84
+ 78 train 8.195867 (lr=2.2098e-04) (hash(x)=5736241)
85
+ 79 train 8.115394 (lr=2.2378e-04) (hash(x)=5657123)
86
+ 80 train 8.125010 (lr=2.2657e-04) (hash(x)=6452095)
87
+ 81 train 8.013991 (lr=2.2937e-04) (hash(x)=5312267)
88
+ 82 train 7.805916 (lr=2.3217e-04) (hash(x)=5434493)
89
+ 83 train 8.116276 (lr=2.3497e-04) (hash(x)=6830813)
90
+ 84 train 7.874906 (lr=2.3776e-04) (hash(x)=4752771)
91
+ 85 train 8.060129 (lr=2.4056e-04) (hash(x)=6424933)
92
+ 86 train 7.990145 (lr=2.4336e-04) (hash(x)=5786517)
93
+ 87 train 8.072040 (lr=2.4615e-04) (hash(x)=6680023)
94
+ 88 train 8.003986 (lr=2.4895e-04) (hash(x)=6474729)
95
+ 89 train 7.989612 (lr=2.5175e-04) (hash(x)=7523279)
96
+ 90 train 7.787131 (lr=2.5455e-04) (hash(x)=5508768)
97
+ 91 train 8.018957 (lr=2.5734e-04) (hash(x)=6635925)
98
+ 92 train 7.863836 (lr=2.6014e-04) (hash(x)=6663801)
99
+ 93 train 7.837631 (lr=2.6294e-04) (hash(x)=5869441)
100
+ 94 train 7.926920 (lr=2.6573e-04) (hash(x)=6698878)
101
+ 95 train 7.816452 (lr=2.6853e-04) (hash(x)=6349004)
102
+ 96 train 7.832438 (lr=2.7133e-04) (hash(x)=4986137)
103
+ 97 train 7.609951 (lr=2.7413e-04) (hash(x)=6008454)
104
+ 98 train 7.686372 (lr=2.7692e-04) (hash(x)=6550770)
105
+ 99 train 7.867159 (lr=2.7972e-04) (hash(x)=7682741)
106
+ 100 val loss 7.7968
107
+ 100 val perplexity 2432.7375
108
+ 100 train 7.915645 (lr=2.8252e-04) (hash(x)=7892158)
109
+ 101 train 7.765061 (lr=2.8531e-04) (hash(x)=7794028)
110
+ 102 train 7.765311 (lr=2.8811e-04) (hash(x)=7093523)
111
+ 103 train 7.833509 (lr=2.9091e-04) (hash(x)=6166062)
112
+ 104 train 7.775701 (lr=2.9371e-04) (hash(x)=6645781)
113
+ 105 train 7.729744 (lr=2.9650e-04) (hash(x)=4811192)
114
+ 106 train 7.647799 (lr=2.9930e-04) (hash(x)=5970866)
115
+ 107 train 7.664854 (lr=3.0210e-04) (hash(x)=7822680)
116
+ 108 train 7.642365 (lr=3.0490e-04) (hash(x)=6317317)
117
+ 109 train 7.768034 (lr=3.0769e-04) (hash(x)=6431409)
118
+ 110 train 7.770483 (lr=3.1049e-04) (hash(x)=6653337)
119
+ 111 train 7.728636 (lr=3.1329e-04) (hash(x)=5323032)
120
+ 112 train 7.606389 (lr=3.1608e-04) (hash(x)=5125339)
121
+ 113 train 7.523371 (lr=3.1888e-04) (hash(x)=7363286)
122
+ 114 train 7.965549 (lr=3.2168e-04) (hash(x)=7670050)
123
+ 115 train 7.700939 (lr=3.2448e-04) (hash(x)=5616075)
124
+ 116 train 7.764788 (lr=3.2727e-04) (hash(x)=6593764)
125
+ 117 train 7.843567 (lr=3.3007e-04) (hash(x)=5672608)
126
+ 118 train 7.682357 (lr=3.3287e-04) (hash(x)=6828653)
127
+ 119 train 7.665688 (lr=3.3566e-04) (hash(x)=6839867)
128
+ 120 train 7.816647 (lr=3.3846e-04) (hash(x)=8426992)
129
+ 121 train 7.461337 (lr=3.4126e-04) (hash(x)=4916928)
130
+ 122 train 7.665308 (lr=3.4406e-04) (hash(x)=6484050)
131
+ 123 train 7.777134 (lr=3.4685e-04) (hash(x)=8353379)
132
+ 124 train 7.697796 (lr=3.4965e-04) (hash(x)=6552510)
133
+ 125 train 7.674901 (lr=3.5245e-04) (hash(x)=4838871)
134
+ 126 train 7.445438 (lr=3.5524e-04) (hash(x)=3021697)
135
+ 127 train 7.590216 (lr=3.5804e-04) (hash(x)=6795665)
136
+ 128 train 7.591175 (lr=3.6084e-04) (hash(x)=5486935)
137
+ 129 train 7.780981 (lr=3.6364e-04) (hash(x)=6678038)
138
+ 130 train 7.820096 (lr=3.6643e-04) (hash(x)=6657714)
139
+ 131 train 7.696929 (lr=3.6923e-04) (hash(x)=6112215)
140
+ 132 train 7.743428 (lr=3.7203e-04) (hash(x)=8353143)
141
+ 133 train 7.805314 (lr=3.7483e-04) (hash(x)=5915361)
142
+ 134 train 7.683172 (lr=3.7762e-04) (hash(x)=6254885)
143
+ 135 train 7.697731 (lr=3.8042e-04) (hash(x)=5771588)
144
+ 136 train 7.677138 (lr=3.8322e-04) (hash(x)=5561507)
145
+ 137 train 7.603652 (lr=3.8601e-04) (hash(x)=5688829)
146
+ 138 train 7.632978 (lr=3.8881e-04) (hash(x)=5371951)
147
+ 139 train 7.792481 (lr=3.9161e-04) (hash(x)=5443305)
148
+ 140 train 7.744305 (lr=3.9441e-04) (hash(x)=5016757)
149
+ 141 train 7.765686 (lr=3.9720e-04) (hash(x)=6901933)
150
+ 142 train 7.733579 (lr=4.0000e-04) (hash(x)=6019085)
151
+ 143 train 7.565797 (lr=4.0280e-04) (hash(x)=4294425)
152
+ 144 train 7.669558 (lr=4.0559e-04) (hash(x)=6095229)
153
+ 145 train 7.480578 (lr=4.0839e-04) (hash(x)=4363941)
154
+ 146 train 7.574804 (lr=4.1119e-04) (hash(x)=4693798)
155
+ 147 train 8.153797 (lr=4.1399e-04) (hash(x)=5486335)
156
+ 148 train 7.948503 (lr=4.1678e-04) (hash(x)=7907450)
157
+ 149 train 7.767869 (lr=4.1958e-04) (hash(x)=7295165)
158
+ 150 val loss 7.6513
159
+ 150 val perplexity 2103.3701
160
+ 150 train 7.700344 (lr=4.2238e-04) (hash(x)=6814026)
161
+ 151 train 7.770445 (lr=4.2517e-04) (hash(x)=6267424)
162
+ 152 train 7.812463 (lr=4.2797e-04) (hash(x)=8086437)
163
+ 153 train 7.858822 (lr=4.3077e-04) (hash(x)=5270452)
164
+ 154 train 7.677064 (lr=4.3357e-04) (hash(x)=6420820)
165
+ 155 train 7.609263 (lr=4.3636e-04) (hash(x)=6305297)
166
+ 156 train 7.735531 (lr=4.3916e-04) (hash(x)=7889849)
167
+ 157 train 8.107779 (lr=4.4196e-04) (hash(x)=6040077)
168
+ 158 train 8.103155 (lr=4.4476e-04) (hash(x)=5652497)
169
+ 159 train 7.694250 (lr=4.4755e-04) (hash(x)=7130251)
170
+ 160 train 7.556124 (lr=4.5035e-04) (hash(x)=6610177)
171
+ 161 train 7.611662 (lr=4.5315e-04) (hash(x)=6508350)
172
+ 162 train 7.772374 (lr=4.5594e-04) (hash(x)=6994983)
173
+ 163 train 7.710647 (lr=4.5874e-04) (hash(x)=5859576)
174
+ 164 train 7.285665 (lr=4.6154e-04) (hash(x)=5487065)
175
+ 165 train 7.530921 (lr=4.6434e-04) (hash(x)=4892065)
176
+ 166 train 7.420722 (lr=4.6713e-04) (hash(x)=4933674)
177
+ 167 train 7.683956 (lr=4.6993e-04) (hash(x)=7317289)
178
+ 168 train 7.797112 (lr=4.7273e-04) (hash(x)=6811522)
179
+ 169 train 7.626160 (lr=4.7552e-04) (hash(x)=5510218)
180
+ 170 train 7.779285 (lr=4.7832e-04) (hash(x)=7304235)
181
+ 171 train 7.927741 (lr=4.8112e-04) (hash(x)=6752265)
182
+ 172 train 7.617755 (lr=4.8392e-04) (hash(x)=5994476)
183
+ 173 train 7.577454 (lr=4.8671e-04) (hash(x)=6530867)
184
+ 174 train 7.695047 (lr=4.8951e-04) (hash(x)=5746260)
185
+ 175 train 7.642545 (lr=4.9231e-04) (hash(x)=5790078)
186
+ 176 train 7.674311 (lr=4.9510e-04) (hash(x)=6430135)
187
+ 177 train 7.739342 (lr=4.9790e-04) (hash(x)=7614023)
188
+ 178 train 7.735826 (lr=5.0070e-04) (hash(x)=5244861)
189
+ 179 train 7.711168 (lr=5.0350e-04) (hash(x)=5245146)
190
+ 180 train 7.591269 (lr=5.0629e-04) (hash(x)=6415565)
191
+ 181 train 7.634667 (lr=5.0909e-04) (hash(x)=5786972)
192
+ 182 train 7.662610 (lr=5.1189e-04) (hash(x)=6131703)
193
+ 183 train 7.680529 (lr=5.1469e-04) (hash(x)=6346459)
194
+ 184 train 7.716157 (lr=5.1748e-04) (hash(x)=6504131)
195
+ 185 train 7.715413 (lr=5.2028e-04) (hash(x)=5209234)
196
+ 186 train 7.617397 (lr=5.2308e-04) (hash(x)=6163272)
197
+ 187 train 7.621024 (lr=5.2587e-04) (hash(x)=5676768)
198
+ 188 train 7.687356 (lr=5.2867e-04) (hash(x)=6090561)
199
+ 189 train 7.834733 (lr=5.3147e-04) (hash(x)=5863092)
200
+ 190 train 7.731881 (lr=5.3427e-04) (hash(x)=7201312)
201
+ 191 train 7.812061 (lr=5.3706e-04) (hash(x)=7019489)
202
+ 192 train 7.575806 (lr=5.3986e-04) (hash(x)=6225608)
203
+ 193 train 7.668107 (lr=5.4266e-04) (hash(x)=5483019)
204
+ 194 train 7.598811 (lr=5.4545e-04) (hash(x)=5903258)
205
+ 195 train 7.643778 (lr=5.4825e-04) (hash(x)=5968716)
206
+ 196 train 7.647646 (lr=5.5105e-04) (hash(x)=5778017)
207
+ 197 train 7.360246 (lr=5.5385e-04) (hash(x)=7780194)
208
+ 198 train 7.568745 (lr=5.5664e-04) (hash(x)=6140998)
209
+ 199 train 7.570322 (lr=5.5944e-04) (hash(x)=3784321)
210
+ 200 val loss 7.5877
211
+ 200 val perplexity 1973.8446
212
+ 200 train 7.583595 (lr=5.6224e-04) (hash(x)=5860821)
213
+ 201 train 7.637559 (lr=5.6503e-04) (hash(x)=7774109)
214
+ 202 train 7.830969 (lr=5.6783e-04) (hash(x)=7997664)
215
+ 203 train 7.773515 (lr=5.7063e-04) (hash(x)=6576749)
216
+ 204 train 7.706108 (lr=5.7343e-04) (hash(x)=7266655)
217
+ 205 train 7.663565 (lr=5.7622e-04) (hash(x)=7072404)
218
+ 206 train 7.569844 (lr=5.7902e-04) (hash(x)=6626459)
219
+ 207 train 7.604104 (lr=5.8182e-04) (hash(x)=6556025)
220
+ 208 train 7.637411 (lr=5.8462e-04) (hash(x)=5942129)
221
+ 209 train 7.724604 (lr=5.8741e-04) (hash(x)=7325027)
222
+ 210 train 7.707909 (lr=5.9021e-04) (hash(x)=7322467)
223
+ 211 train 7.640356 (lr=5.9301e-04) (hash(x)=6373412)
224
+ 212 train 7.434405 (lr=5.9580e-04) (hash(x)=7332255)
225
+ 213 train 7.718988 (lr=5.9860e-04) (hash(x)=5262868)
226
+ 214 train 7.412348 (lr=6.0140e-04) (hash(x)=5688247)
227
+ 215 train 7.570026 (lr=6.0420e-04) (hash(x)=8055563)
228
+ 216 train 7.613070 (lr=6.0699e-04) (hash(x)=5352405)
229
+ 217 train 7.673370 (lr=6.0979e-04) (hash(x)=5932790)
230
+ 218 train 7.635149 (lr=6.1259e-04) (hash(x)=6249312)
231
+ 219 train 7.582789 (lr=6.1538e-04) (hash(x)=5680154)
232
+ 220 train 7.675784 (lr=6.1818e-04) (hash(x)=8156280)
233
+ 221 train 7.618041 (lr=6.2098e-04) (hash(x)=5914217)
234
+ 222 train 7.466636 (lr=6.2378e-04) (hash(x)=7120215)
235
+ 223 train 7.654107 (lr=6.2657e-04) (hash(x)=6424128)
236
+ 224 train 7.414338 (lr=6.2937e-04) (hash(x)=4390027)
237
+ 225 train 7.621014 (lr=6.3217e-04) (hash(x)=7804089)
238
+ 226 train 7.697839 (lr=6.3497e-04) (hash(x)=7130267)
239
+ 227 train 7.485427 (lr=6.3776e-04) (hash(x)=6490149)
240
+ 228 train 7.544992 (lr=6.4056e-04) (hash(x)=9276378)
241
+ 229 train 7.591858 (lr=6.4336e-04) (hash(x)=6283540)
242
+ 230 train 7.581809 (lr=6.4615e-04) (hash(x)=5781680)
243
+ 231 train 7.503127 (lr=6.4895e-04) (hash(x)=5728337)
244
+ 232 train 7.481125 (lr=6.5175e-04) (hash(x)=6278202)
245
+ 233 train 7.564575 (lr=6.5455e-04) (hash(x)=7400272)
246
+ 234 train 7.560659 (lr=6.5734e-04) (hash(x)=5795128)
247
+ 235 train 7.514776 (lr=6.6014e-04) (hash(x)=6473432)
248
+ 236 train 7.523838 (lr=6.6294e-04) (hash(x)=5201502)
249
+ 237 train 7.765454 (lr=6.6573e-04) (hash(x)=6922646)
250
+ 238 train 7.456151 (lr=6.6853e-04) (hash(x)=5484641)
251
+ 239 train 7.523126 (lr=6.7133e-04) (hash(x)=7277141)
252
+ 240 train 7.531984 (lr=6.7413e-04) (hash(x)=5604614)
253
+ 241 train 7.585181 (lr=6.7692e-04) (hash(x)=7443124)
254
+ 242 train 7.595470 (lr=6.7972e-04) (hash(x)=5601968)
255
+ 243 train 7.477440 (lr=6.8252e-04) (hash(x)=5632735)
256
+ 244 train 7.481354 (lr=6.8531e-04) (hash(x)=5594810)
257
+ 245 train 7.361511 (lr=6.8811e-04) (hash(x)=5881015)
258
+ 246 train 7.486296 (lr=6.9091e-04) (hash(x)=6865361)
259
+ 247 train 7.583138 (lr=6.9371e-04) (hash(x)=5809411)
260
+ 248 train 7.565957 (lr=6.9650e-04) (hash(x)=6293025)
261
+ 249 train 7.509156 (lr=6.9930e-04) (hash(x)=7232964)
262
+ 250 val loss 7.5298
263
+ 250 val perplexity 1862.6520
264
+ 250 train 7.458787 (lr=7.0210e-04) (hash(x)=5646619)
265
+ 251 train 7.580256 (lr=7.0490e-04) (hash(x)=6171829)
266
+ 252 train 7.622648 (lr=7.0769e-04) (hash(x)=7787552)
267
+ 253 train 7.579946 (lr=7.1049e-04) (hash(x)=5884456)
268
+ 254 train 7.597624 (lr=7.1329e-04) (hash(x)=6752162)
269
+ 255 train 7.520886 (lr=7.1608e-04) (hash(x)=5077031)
270
+ 256 train 7.684924 (lr=7.1888e-04) (hash(x)=6883481)
271
+ 257 train 7.574386 (lr=7.2168e-04) (hash(x)=6248895)
272
+ 258 train 7.432758 (lr=7.2448e-04) (hash(x)=6636615)
273
+ 259 train 7.632457 (lr=7.2727e-04) (hash(x)=6918004)
274
+ 260 train 7.572240 (lr=7.3007e-04) (hash(x)=7302874)
275
+ 261 train 7.671212 (lr=7.3287e-04) (hash(x)=6482466)
276
+ 262 train 7.493440 (lr=7.3566e-04) (hash(x)=7298317)
277
+ 263 train 7.551752 (lr=7.3846e-04) (hash(x)=6588573)
278
+ 264 train 7.544188 (lr=7.4126e-04) (hash(x)=6229630)
279
+ 265 train 7.626081 (lr=7.4406e-04) (hash(x)=6136939)
280
+ 266 train 7.405597 (lr=7.4685e-04) (hash(x)=6394645)
281
+ 267 train 7.617774 (lr=7.4965e-04) (hash(x)=7040157)
282
+ 268 train 7.668215 (lr=7.5245e-04) (hash(x)=7041858)
283
+ 269 train 7.561538 (lr=7.5524e-04) (hash(x)=6455595)
284
+ 270 train 7.518384 (lr=7.5804e-04) (hash(x)=6480443)
285
+ 271 train 7.674230 (lr=7.6084e-04) (hash(x)=4122983)
286
+ 272 train 7.484338 (lr=7.6364e-04) (hash(x)=7672190)
287
+ 273 train 7.518104 (lr=7.6643e-04) (hash(x)=4724052)
288
+ 274 train 7.483405 (lr=7.6923e-04) (hash(x)=6654267)
289
+ 275 train 7.459590 (lr=7.7203e-04) (hash(x)=6820797)
290
+ 276 train 7.519829 (lr=7.7483e-04) (hash(x)=6458662)
291
+ 277 train 7.913469 (lr=7.7762e-04) (hash(x)=7477594)
292
+ 278 train 7.574753 (lr=7.8042e-04) (hash(x)=5876574)
293
+ 279 train 7.319147 (lr=7.8322e-04) (hash(x)=6146305)
294
+ 280 train 7.478566 (lr=7.8601e-04) (hash(x)=6162080)
295
+ 281 train 7.415203 (lr=7.8881e-04) (hash(x)=6814831)
296
+ 282 train 7.569870 (lr=7.9161e-04) (hash(x)=7138295)
297
+ 283 train 7.332067 (lr=7.9441e-04) (hash(x)=5548298)
298
+ 284 train 7.444189 (lr=7.9720e-04) (hash(x)=6265078)
299
+ 285 train 7.824429 (lr=8.0000e-04) (hash(x)=7609416)
300
+ 286 train 7.521755 (lr=8.0000e-04) (hash(x)=6909367)
301
+ 287 train 7.610746 (lr=8.0000e-04) (hash(x)=7574342)
302
+ 288 train 7.426715 (lr=7.9999e-04) (hash(x)=8859748)
303
+ 289 train 7.413340 (lr=7.9997e-04) (hash(x)=5906937)
304
+ 290 train 7.433335 (lr=7.9994e-04) (hash(x)=7243024)
305
+ 291 train 7.528471 (lr=7.9991e-04) (hash(x)=8013471)
306
+ 292 train 7.442040 (lr=7.9987e-04) (hash(x)=5659461)
307
+ 293 train 7.478229 (lr=7.9983e-04) (hash(x)=6146213)
308
+ 294 train 7.527931 (lr=7.9978e-04) (hash(x)=6602917)
309
+ 295 train 7.451355 (lr=7.9972e-04) (hash(x)=6709902)
310
+ 296 train 7.537931 (lr=7.9965e-04) (hash(x)=8154341)
311
+ 297 train 7.685238 (lr=7.9958e-04) (hash(x)=6402571)
312
+ 298 train 7.450487 (lr=7.9950e-04) (hash(x)=5605466)
313
+ 299 train 7.420865 (lr=7.9941e-04) (hash(x)=7683614)
314
+ 300 val loss 7.4656
315
+ 300 val perplexity 1746.8851
316
+ 300 train 7.512493 (lr=7.9932e-04) (hash(x)=5421572)
317
+ 301 train 7.361656 (lr=7.9922e-04) (hash(x)=5453026)
318
+ 302 train 7.432813 (lr=7.9911e-04) (hash(x)=6437746)
319
+ 303 train 7.327455 (lr=7.9899e-04) (hash(x)=6260017)
320
+ 304 train 7.419444 (lr=7.9887e-04) (hash(x)=6119053)
321
+ 305 train 7.594572 (lr=7.9874e-04) (hash(x)=6591466)
322
+ 306 train 7.523496 (lr=7.9861e-04) (hash(x)=5652701)
323
+ 307 train 7.493474 (lr=7.9846e-04) (hash(x)=5501472)
324
+ 308 train 7.669855 (lr=7.9831e-04) (hash(x)=6160695)
325
+ 309 train 7.423739 (lr=7.9816e-04) (hash(x)=5871698)
326
+ 310 train 7.395344 (lr=7.9799e-04) (hash(x)=7301184)
327
+ 311 train 7.323932 (lr=7.9782e-04) (hash(x)=4282052)
328
+ 312 train 7.548635 (lr=7.9765e-04) (hash(x)=5679330)
329
+ 313 train 7.373698 (lr=7.9746e-04) (hash(x)=7824660)
330
+ 314 train 7.421705 (lr=7.9727e-04) (hash(x)=4857318)
331
+ 315 train 7.341324 (lr=7.9707e-04) (hash(x)=5954950)
332
+ 316 train 7.430247 (lr=7.9687e-04) (hash(x)=5758178)
333
+ 317 train 7.454362 (lr=7.9666e-04) (hash(x)=5484656)
334
+ 318 train 7.485174 (lr=7.9644e-04) (hash(x)=6818913)
335
+ 319 train 7.694899 (lr=7.9621e-04) (hash(x)=7031328)
336
+ 320 train 7.536449 (lr=7.9598e-04) (hash(x)=6837621)
337
+ 321 train 7.568770 (lr=7.9574e-04) (hash(x)=9157069)
338
+ 322 train 7.798380 (lr=7.9549e-04) (hash(x)=6509536)
339
+ 323 train 7.486732 (lr=7.9524e-04) (hash(x)=5996883)
340
+ 324 train 7.765191 (lr=7.9498e-04) (hash(x)=6850868)
341
+ 325 train 7.702994 (lr=7.9471e-04) (hash(x)=7645447)
342
+ 326 train 7.716320 (lr=7.9444e-04) (hash(x)=6717515)
343
+ 327 train 7.837692 (lr=7.9416e-04) (hash(x)=6336164)
344
+ 328 train 7.528562 (lr=7.9387e-04) (hash(x)=6897445)
345
+ 329 train 7.451011 (lr=7.9358e-04) (hash(x)=5224240)
346
+ 330 train 7.607484 (lr=7.9327e-04) (hash(x)=6612945)
347
+ 331 train 8.065563 (lr=7.9297e-04) (hash(x)=5447801)
348
+ 332 train 7.589366 (lr=7.9265e-04) (hash(x)=7308579)
349
+ 333 train 7.447890 (lr=7.9233e-04) (hash(x)=4155598)
350
+ 334 train 7.481609 (lr=7.9200e-04) (hash(x)=6106773)
351
+ 335 train 7.545780 (lr=7.9167e-04) (hash(x)=7482744)
352
+ 336 train 7.639565 (lr=7.9132e-04) (hash(x)=6310687)
353
+ 337 train 7.536043 (lr=7.9097e-04) (hash(x)=4733414)
354
+ 338 train 7.665926 (lr=7.9062e-04) (hash(x)=6979734)
355
+ 339 train 7.476223 (lr=7.9026e-04) (hash(x)=7806284)
356
+ 340 train 7.476771 (lr=7.8989e-04) (hash(x)=5080536)
357
+ 341 train 7.685116 (lr=7.8951e-04) (hash(x)=7236478)
358
+ 342 train 7.642132 (lr=7.8913e-04) (hash(x)=6619318)
359
+ 343 train 7.555558 (lr=7.8874e-04) (hash(x)=4656255)
360
+ 344 train 7.524395 (lr=7.8834e-04) (hash(x)=5546888)
361
+ 345 train 7.925096 (lr=7.8794e-04) (hash(x)=9298325)
362
+ 346 train 7.631393 (lr=7.8753e-04) (hash(x)=5994232)
363
+ 347 train 7.598736 (lr=7.8711e-04) (hash(x)=6287524)
364
+ 348 train 7.688113 (lr=7.8669e-04) (hash(x)=6200049)
365
+ 349 train 7.612113 (lr=7.8626e-04) (hash(x)=7895970)
366
+ 350 val loss 7.5586
367
+ 350 val perplexity 1917.2172
368
+ 350 train 7.530797 (lr=7.8582e-04) (hash(x)=6051847)
369
+ 351 train 7.674979 (lr=7.8538e-04) (hash(x)=8847992)
370
+ 352 train 7.964968 (lr=7.8493e-04) (hash(x)=5123883)
371
+ 353 train 7.611274 (lr=7.8447e-04) (hash(x)=6338998)
372
+ 354 train 7.614423 (lr=7.8401e-04) (hash(x)=5765866)
373
+ 355 train 7.930711 (lr=7.8354e-04) (hash(x)=7104011)
374
+ 356 train 7.709740 (lr=7.8306e-04) (hash(x)=7438267)
375
+ 357 train 7.461123 (lr=7.8258e-04) (hash(x)=5573214)
376
+ 358 train 7.723330 (lr=7.8209e-04) (hash(x)=6897283)
377
+ 359 train 7.515640 (lr=7.8159e-04) (hash(x)=6882897)
378
+ 360 train 7.591517 (lr=7.8109e-04) (hash(x)=5343615)
379
+ 361 train 7.522196 (lr=7.8058e-04) (hash(x)=5854242)
380
+ 362 train 7.478942 (lr=7.8006e-04) (hash(x)=6317324)
381
+ 363 train 7.420895 (lr=7.7954e-04) (hash(x)=6972494)
382
+ 364 train 7.481015 (lr=7.7901e-04) (hash(x)=5661167)
383
+ 365 train 7.510581 (lr=7.7847e-04) (hash(x)=6796805)
384
+ 366 train 7.350806 (lr=7.7793e-04) (hash(x)=5817209)
385
+ 367 train 7.402586 (lr=7.7738e-04) (hash(x)=6139120)
386
+ 368 train 7.553034 (lr=7.7682e-04) (hash(x)=6121296)
387
+ 369 train 7.542743 (lr=7.7626e-04) (hash(x)=6486796)
388
+ 370 train 7.488333 (lr=7.7569e-04) (hash(x)=5514645)
389
+ 371 train 7.427167 (lr=7.7511e-04) (hash(x)=5578797)
390
+ 372 train 7.432696 (lr=7.7453e-04) (hash(x)=7651738)
391
+ 373 train 7.604403 (lr=7.7394e-04) (hash(x)=6895326)
392
+ 374 train 7.617215 (lr=7.7335e-04) (hash(x)=5490256)
393
+ 375 train 7.486062 (lr=7.7275e-04) (hash(x)=7783160)
394
+ 376 train 7.687791 (lr=7.7214e-04) (hash(x)=5608424)
395
+ 377 train 7.472507 (lr=7.7153e-04) (hash(x)=4867625)
396
+ 378 train 7.373025 (lr=7.7091e-04) (hash(x)=4655472)
397
+ 379 train 7.438614 (lr=7.7028e-04) (hash(x)=6631123)
398
+ 380 train 7.116121 (lr=7.6964e-04) (hash(x)=5751990)
399
+ 381 train 7.422595 (lr=7.6901e-04) (hash(x)=6710965)
400
+ 382 train 7.385410 (lr=7.6836e-04) (hash(x)=5293800)
401
+ 383 train 7.429354 (lr=7.6771e-04) (hash(x)=6294657)
402
+ 384 train 7.572870 (lr=7.6705e-04) (hash(x)=6999538)
403
+ 385 train 7.542579 (lr=7.6638e-04) (hash(x)=5443606)
404
+ 386 train 7.546618 (lr=7.6571e-04) (hash(x)=5650401)
405
+ 387 train 7.528281 (lr=7.6503e-04) (hash(x)=7035770)
406
+ 388 train 7.473723 (lr=7.6435e-04) (hash(x)=6678271)
407
+ 389 train 7.648479 (lr=7.6366e-04) (hash(x)=4899541)
408
+ 390 train 7.420696 (lr=7.6296e-04) (hash(x)=5359260)
409
+ 391 train 7.432303 (lr=7.6226e-04) (hash(x)=6227120)
410
+ 392 train 7.469637 (lr=7.6155e-04) (hash(x)=6643916)
411
+ 393 train 7.549084 (lr=7.6083e-04) (hash(x)=5981648)
412
+ 394 train 7.523375 (lr=7.6011e-04) (hash(x)=6467540)
413
+ 395 train 7.459850 (lr=7.5938e-04) (hash(x)=5558993)
414
+ 396 train 7.618283 (lr=7.5865e-04) (hash(x)=6698959)
415
+ 397 train 7.446742 (lr=7.5791e-04) (hash(x)=7015671)
416
+ 398 train 7.438100 (lr=7.5716e-04) (hash(x)=6800890)
417
+ 399 train 7.595595 (lr=7.5641e-04) (hash(x)=5931234)
418
+ 400 val loss 7.5235
419
+ 400 val perplexity 1851.0627
420
+ 400 train 7.490206 (lr=7.5565e-04) (hash(x)=3995070)
421
+ 401 train 7.409871 (lr=7.5489e-04) (hash(x)=4582159)
422
+ 402 train 7.493560 (lr=7.5412e-04) (hash(x)=7081906)
423
+ 403 train 7.530330 (lr=7.5334e-04) (hash(x)=5029472)
424
+ 404 train 7.509226 (lr=7.5256e-04) (hash(x)=3946232)
425
+ 405 train 7.433203 (lr=7.5177e-04) (hash(x)=5434102)
426
+ 406 train 8.128281 (lr=7.5097e-04) (hash(x)=5340646)
427
+ 407 train 7.389859 (lr=7.5017e-04) (hash(x)=6810638)
428
+ 408 train 7.596622 (lr=7.4937e-04) (hash(x)=6768026)
429
+ 409 train 7.627458 (lr=7.4855e-04) (hash(x)=6521562)
430
+ 410 train 7.601590 (lr=7.4773e-04) (hash(x)=4429434)
431
+ 411 train 7.619032 (lr=7.4691e-04) (hash(x)=7226932)
432
+ 412 train 7.411312 (lr=7.4608e-04) (hash(x)=5065932)
433
+ 413 train 7.577682 (lr=7.4524e-04) (hash(x)=6957849)
434
+ 414 train 7.587616 (lr=7.4440e-04) (hash(x)=6431457)
435
+ 415 train 7.637734 (lr=7.4355e-04) (hash(x)=5334281)
436
+ 416 train 7.597422 (lr=7.4270e-04) (hash(x)=5471747)
437
+ 417 train 7.712931 (lr=7.4184e-04) (hash(x)=5887983)
438
+ 418 train 7.629362 (lr=7.4097e-04) (hash(x)=5075909)
439
+ 419 train 7.435196 (lr=7.4010e-04) (hash(x)=6393576)
440
+ 420 train 7.549894 (lr=7.3922e-04) (hash(x)=6608979)
441
+ 421 train 7.900916 (lr=7.3834e-04) (hash(x)=5057959)
442
+ 422 train 7.866672 (lr=7.3745e-04) (hash(x)=6414858)
443
+ 423 train 7.655729 (lr=7.3655e-04) (hash(x)=8595271)
444
+ 424 train 7.548162 (lr=7.3565e-04) (hash(x)=6500949)
445
+ 425 train 7.699601 (lr=7.3474e-04) (hash(x)=8395167)
446
+ 426 train 7.825996 (lr=7.3383e-04) (hash(x)=6288420)
447
+ 427 train 7.614532 (lr=7.3291e-04) (hash(x)=7634417)
448
+ 428 train 7.522439 (lr=7.3199e-04) (hash(x)=6795561)
449
+ 429 train 7.447693 (lr=7.3106e-04) (hash(x)=5257771)
450
+ 430 train 7.599300 (lr=7.3012e-04) (hash(x)=7506860)
451
+ 431 train 7.644774 (lr=7.2918e-04) (hash(x)=7674238)
452
+ 432 train 7.737776 (lr=7.2824e-04) (hash(x)=5353794)
453
+ 433 train 7.663313 (lr=7.2729e-04) (hash(x)=5300555)
454
+ 434 train 7.362828 (lr=7.2633e-04) (hash(x)=5564419)
455
+ 435 train 7.634722 (lr=7.2537e-04) (hash(x)=6066139)
456
+ 436 train 7.594750 (lr=7.2440e-04) (hash(x)=6863709)
457
+ 437 train 7.580345 (lr=7.2342e-04) (hash(x)=6688212)
458
+ 438 train 7.760254 (lr=7.2244e-04) (hash(x)=7095090)
459
+ 439 train 7.839207 (lr=7.2146e-04) (hash(x)=6642404)
460
+ 440 train 7.570358 (lr=7.2047e-04) (hash(x)=6539227)
461
+ 441 train 7.539854 (lr=7.1947e-04) (hash(x)=4961570)
462
+ 442 train 7.491604 (lr=7.1847e-04) (hash(x)=6606391)
463
+ 443 train 7.312110 (lr=7.1747e-04) (hash(x)=5105190)
464
+ 444 train 7.828205 (lr=7.1645e-04) (hash(x)=5695997)
465
+ 445 train 7.566549 (lr=7.1544e-04) (hash(x)=7155991)
466
+ 446 train 7.535670 (lr=7.1441e-04) (hash(x)=7249727)
467
+ 447 train 7.521611 (lr=7.1339e-04) (hash(x)=5132972)
468
+ 448 train 7.514251 (lr=7.1235e-04) (hash(x)=5381802)
469
+ 449 train 7.533920 (lr=7.1131e-04) (hash(x)=6269946)
470
+ 450 val loss 7.5288
471
+ 450 val perplexity 1860.9075
472
+ 450 train 7.507209 (lr=7.1027e-04) (hash(x)=4976762)
473
+ 451 train 7.609783 (lr=7.0922e-04) (hash(x)=5396133)
474
+ 452 train 7.585193 (lr=7.0817e-04) (hash(x)=6173473)
475
+ 453 train 7.560557 (lr=7.0711e-04) (hash(x)=6394083)
476
+ 454 train 7.627398 (lr=7.0604e-04) (hash(x)=6436030)
477
+ 455 train 7.497279 (lr=7.0497e-04) (hash(x)=5237804)
478
+ 456 train 7.469097 (lr=7.0390e-04) (hash(x)=6540397)
479
+ 457 train 7.729666 (lr=7.0282e-04) (hash(x)=5216638)
480
+ 458 train 7.556839 (lr=7.0173e-04) (hash(x)=5750537)
481
+ 459 train 7.520763 (lr=7.0064e-04) (hash(x)=6735565)
482
+ 460 train 7.555656 (lr=6.9955e-04) (hash(x)=7297488)
483
+ 461 train 7.712149 (lr=6.9845e-04) (hash(x)=5983575)
484
+ 462 train 7.549114 (lr=6.9734e-04) (hash(x)=4808543)
485
+ 463 train 7.648498 (lr=6.9623e-04) (hash(x)=5831619)
486
+ 464 train 7.428707 (lr=6.9512e-04) (hash(x)=6280052)
487
+ 465 train 7.443779 (lr=6.9400e-04) (hash(x)=9109083)
488
+ 466 train 7.662139 (lr=6.9287e-04) (hash(x)=5848604)
489
+ 467 train 7.630549 (lr=6.9174e-04) (hash(x)=6912967)
490
+ 468 train 7.757546 (lr=6.9061e-04) (hash(x)=6615365)
491
+ 469 train 7.805801 (lr=6.8947e-04) (hash(x)=5965674)
492
+ 470 train 7.660726 (lr=6.8832e-04) (hash(x)=6879965)
493
+ 471 train 7.669535 (lr=6.8717e-04) (hash(x)=5822533)
494
+ 472 train 7.536750 (lr=6.8602e-04) (hash(x)=5794462)
495
+ 473 train 7.621186 (lr=6.8486e-04) (hash(x)=5268216)
496
+ 474 train 7.448437 (lr=6.8370e-04) (hash(x)=7058396)
497
+ 475 train 7.541173 (lr=6.8253e-04) (hash(x)=6531900)
498
+ 476 train 7.559078 (lr=6.8136e-04) (hash(x)=5878912)
499
+ 477 train 7.392334 (lr=6.8018e-04) (hash(x)=5426499)
500
+ 478 train 7.582002 (lr=6.7900e-04) (hash(x)=8107004)
501
+ 479 train 8.220468 (lr=6.7781e-04) (hash(x)=7455372)
502
+ 480 train 8.141043 (lr=6.7662e-04) (hash(x)=7727959)
503
+ 481 train 8.083649 (lr=6.7542e-04) (hash(x)=8319119)
504
+ 482 train 8.038240 (lr=6.7422e-04) (hash(x)=7581656)
505
+ 483 train 8.099391 (lr=6.7302e-04) (hash(x)=7592128)
506
+ 484 train 8.039245 (lr=6.7181e-04) (hash(x)=10506633)
507
+ 485 train 7.986907 (lr=6.7059e-04) (hash(x)=8704170)
508
+ 486 train 7.828151 (lr=6.6937e-04) (hash(x)=7461138)
509
+ 487 train 7.734338 (lr=6.6815e-04) (hash(x)=7165037)
510
+ 488 train 7.800331 (lr=6.6692e-04) (hash(x)=4976998)
511
+ 489 train 7.532074 (lr=6.6569e-04) (hash(x)=7678000)
512
+ 490 train 7.603909 (lr=6.6446e-04) (hash(x)=6299409)
513
+ 491 train 7.575032 (lr=6.6322e-04) (hash(x)=7414788)
514
+ 492 train 7.632190 (lr=6.6197e-04) (hash(x)=5063061)
515
+ 493 train 7.599571 (lr=6.6072e-04) (hash(x)=7326216)
516
+ 494 train 7.760934 (lr=6.5947e-04) (hash(x)=7492012)
517
+ 495 train 7.706675 (lr=6.5821e-04) (hash(x)=5857162)
518
+ 496 train 7.509042 (lr=6.5695e-04) (hash(x)=6169001)
519
+ 497 train 9.273797 (lr=6.5568e-04) (hash(x)=14715669)
520
+ 498 train 11.714999 (lr=6.5441e-04) (hash(x)=18256706)
521
+ 499 train 11.415065 (lr=6.5314e-04) (hash(x)=16950560)
522
+ 500 val loss 7.7813
523
+ 500 val perplexity 2395.4487
524
+ 500 train 8.328017 (lr=6.5186e-04) (hash(x)=6113563)
525
+ 501 train 7.641392 (lr=6.5058e-04) (hash(x)=6159125)
526
+ 502 train 8.087125 (lr=6.4929e-04) (hash(x)=5482303)
527
+ 503 train 8.185935 (lr=6.4800e-04) (hash(x)=5908442)
528
+ 504 train 8.154387 (lr=6.4670e-04) (hash(x)=4985045)
529
+ 505 train 8.127583 (lr=6.4540e-04) (hash(x)=6668887)
530
+ 506 train 8.105291 (lr=6.4410e-04) (hash(x)=6123910)
531
+ 507 train 8.051686 (lr=6.4280e-04) (hash(x)=7662975)
532
+ 508 train 7.922078 (lr=6.4148e-04) (hash(x)=7191898)
533
+ 509 train 7.802843 (lr=6.4017e-04) (hash(x)=6669539)
534
+ 510 train 7.750798 (lr=6.3885e-04) (hash(x)=6557481)
535
+ 511 train 7.633525 (lr=6.3753e-04) (hash(x)=5015998)
536
+ 512 train 7.645160 (lr=6.3620e-04) (hash(x)=4964540)
537
+ 513 train 7.555276 (lr=6.3487e-04) (hash(x)=5458386)
538
+ 514 train 7.891742 (lr=6.3354e-04) (hash(x)=6077616)
539
+ 515 train 7.714035 (lr=6.3220e-04) (hash(x)=4849847)
540
+ 516 train 7.798770 (lr=6.3086e-04) (hash(x)=5431422)
541
+ 517 train 8.014119 (lr=6.2952e-04) (hash(x)=5125562)
542
+ 518 train 7.950609 (lr=6.2817e-04) (hash(x)=5587715)
543
+ 519 train 7.987746 (lr=6.2681e-04) (hash(x)=6012047)
544
+ 520 train 7.827288 (lr=6.2546e-04) (hash(x)=5668436)
545
+ 521 train 7.852293 (lr=6.2410e-04) (hash(x)=6079364)
546
+ 522 train 7.991223 (lr=6.2274e-04) (hash(x)=5200504)
547
+ 523 train 7.647756 (lr=6.2137e-04) (hash(x)=6030395)
548
+ 524 train 7.885254 (lr=6.2000e-04) (hash(x)=6855319)
549
+ 525 train 7.610994 (lr=6.1863e-04) (hash(x)=6781762)
550
+ 526 train 7.726836 (lr=6.1725e-04) (hash(x)=6717939)
551
+ 527 train 7.699976 (lr=6.1587e-04) (hash(x)=5238325)
552
+ 528 train 7.576651 (lr=6.1449e-04) (hash(x)=5808069)
553
+ 529 train 7.897998 (lr=6.1310e-04) (hash(x)=8611144)
554
+ 530 train 7.580552 (lr=6.1171e-04) (hash(x)=6028738)
555
+ 531 train 7.617995 (lr=6.1031e-04) (hash(x)=6901605)
556
+ 532 train 7.573563 (lr=6.0892e-04) (hash(x)=6723307)
557
+ 533 train 7.658122 (lr=6.0752e-04) (hash(x)=5499337)
558
+ 534 train 7.592698 (lr=6.0611e-04) (hash(x)=6586372)
559
+ 535 train 7.672626 (lr=6.0471e-04) (hash(x)=7134874)
560
+ 536 train 7.607370 (lr=6.0330e-04) (hash(x)=8415623)
561
+ 537 train 7.615800 (lr=6.0188e-04) (hash(x)=6149038)
562
+ 538 train 7.648949 (lr=6.0047e-04) (hash(x)=6078325)
563
+ 539 train 7.592303 (lr=5.9905e-04) (hash(x)=6609760)
564
+ 540 train 7.500653 (lr=5.9762e-04) (hash(x)=6185253)
565
+ 541 train 7.835172 (lr=5.9620e-04) (hash(x)=6652627)
566
+ 542 train 7.676136 (lr=5.9477e-04) (hash(x)=7758536)
567
+ 543 train 7.653555 (lr=5.9334e-04) (hash(x)=8618579)
568
+ 544 train 7.600768 (lr=5.9190e-04) (hash(x)=6710226)
569
+ 545 train 7.482594 (lr=5.9047e-04) (hash(x)=5896892)
570
+ 546 train 7.494483 (lr=5.8903e-04) (hash(x)=6332995)
571
+ 547 train 7.870669 (lr=5.8758e-04) (hash(x)=8945854)
572
+ 548 train 7.443300 (lr=5.8614e-04) (hash(x)=4329129)
573
+ 549 train 7.519009 (lr=5.8469e-04) (hash(x)=6775103)
574
+ 550 val loss 7.5365
575
+ 550 val perplexity 1875.2068
576
+ 550 train 7.611613 (lr=5.8323e-04) (hash(x)=4465735)
577
+ 551 train 7.491980 (lr=5.8178e-04) (hash(x)=6200024)
578
+ 552 train 7.461643 (lr=5.8032e-04) (hash(x)=5520748)
579
+ 553 train 7.467099 (lr=5.7886e-04) (hash(x)=5238139)
580
+ 554 train 7.710800 (lr=5.7740e-04) (hash(x)=7639817)
581
+ 555 train 7.464115 (lr=5.7593e-04) (hash(x)=5780948)
582
+ 556 train 7.593911 (lr=5.7447e-04) (hash(x)=6733444)
583
+ 557 train 7.549597 (lr=5.7300e-04) (hash(x)=6168407)
584
+ 558 train 7.564382 (lr=5.7152e-04) (hash(x)=6680197)
585
+ 559 train 7.493705 (lr=5.7005e-04) (hash(x)=3775881)
586
+ 560 train 7.339830 (lr=5.6857e-04) (hash(x)=5784096)
587
+ 561 train 7.741882 (lr=5.6709e-04) (hash(x)=6848709)
588
+ 562 train 7.789749 (lr=5.6560e-04) (hash(x)=7449019)
589
+ 563 train 7.883511 (lr=5.6412e-04) (hash(x)=6478113)
590
+ 564 train 7.546853 (lr=5.6263e-04) (hash(x)=5925744)
591
+ 565 train 7.581130 (lr=5.6114e-04) (hash(x)=6483914)
592
+ 566 train 7.491834 (lr=5.5965e-04) (hash(x)=6010847)
593
+ 567 train 7.582766 (lr=5.5815e-04) (hash(x)=5927795)
594
+ 568 train 7.566588 (lr=5.5666e-04) (hash(x)=5896332)
595
+ 569 train 7.526299 (lr=5.5516e-04) (hash(x)=8438033)
596
+ 570 train 7.615861 (lr=5.5365e-04) (hash(x)=6081660)
597
+ 571 train 7.541177 (lr=5.5215e-04) (hash(x)=5684788)
598
+ 572 train 7.454782 (lr=5.5064e-04) (hash(x)=6272055)
599
+ 573 train 7.405884 (lr=5.4913e-04) (hash(x)=5710182)
600
+ 574 train 7.521661 (lr=5.4762e-04) (hash(x)=6219616)
601
+ 575 train 7.491307 (lr=5.4611e-04) (hash(x)=6160063)
602
+ 576 train 7.489717 (lr=5.4460e-04) (hash(x)=5710809)
603
+ 577 train 7.269548 (lr=5.4308e-04) (hash(x)=5955100)
604
+ 578 train 7.285906 (lr=5.4156e-04) (hash(x)=5356088)
605
+ 579 train 7.332757 (lr=5.4004e-04) (hash(x)=6672309)
606
+ 580 train 7.536458 (lr=5.3852e-04) (hash(x)=5841684)
607
+ 581 train 7.824654 (lr=5.3699e-04) (hash(x)=9001564)
608
+ 582 train 8.361393 (lr=5.3547e-04) (hash(x)=6897178)
609
+ 583 train 7.616403 (lr=5.3394e-04) (hash(x)=5454654)
610
+ 584 train 7.390914 (lr=5.3241e-04) (hash(x)=6391454)
611
+ 585 train 7.461087 (lr=5.3088e-04) (hash(x)=5174062)
612
+ 586 train 7.571717 (lr=5.2934e-04) (hash(x)=5197924)
613
+ 587 train 7.547468 (lr=5.2781e-04) (hash(x)=6199203)
614
+ 588 train 7.378873 (lr=5.2627e-04) (hash(x)=5786498)
615
+ 589 train 7.510535 (lr=5.2473e-04) (hash(x)=8352747)
616
+ 590 train 7.793103 (lr=5.2319e-04) (hash(x)=4829700)
617
+ 591 train 7.354563 (lr=5.2165e-04) (hash(x)=4547106)
618
+ 592 train 7.644701 (lr=5.2011e-04) (hash(x)=7686831)
619
+ 593 train 7.491415 (lr=5.1856e-04) (hash(x)=6433099)
620
+ 594 train 7.385810 (lr=5.1702e-04) (hash(x)=4826460)
621
+ 595 train 7.450578 (lr=5.1547e-04) (hash(x)=6424428)
622
+ 596 train 7.351038 (lr=5.1392e-04) (hash(x)=6073605)
623
+ 597 train 7.663077 (lr=5.1237e-04) (hash(x)=5975712)
624
+ 598 train 7.299032 (lr=5.1082e-04) (hash(x)=7030029)
625
+ 599 train 7.589469 (lr=5.0926e-04) (hash(x)=5458532)
626
+ 600 val loss 7.5102
627
+ 600 val perplexity 1826.5492
628
+ 600 train 7.435382 (lr=5.0771e-04) (hash(x)=7823919)
629
+ 601 train 7.859172 (lr=5.0615e-04) (hash(x)=7002774)
630
+ 602 train 7.673325 (lr=5.0459e-04) (hash(x)=7056399)
631
+ 603 train 7.851257 (lr=5.0303e-04) (hash(x)=8206173)
632
+ 604 train 7.692409 (lr=5.0147e-04) (hash(x)=8034846)
633
+ 605 train 7.947862 (lr=4.9991e-04) (hash(x)=7784971)
634
+ 606 train 8.167578 (lr=4.9835e-04) (hash(x)=7118076)
635
+ 607 train 7.446858 (lr=4.9679e-04) (hash(x)=6001843)
636
+ 608 train 7.565948 (lr=4.9522e-04) (hash(x)=5582437)
637
+ 609 train 7.637152 (lr=4.9366e-04) (hash(x)=5698642)
638
+ 610 train 7.523649 (lr=4.9209e-04) (hash(x)=6615301)
639
+ 611 train 7.526775 (lr=4.9052e-04) (hash(x)=6600751)
640
+ 612 train 7.439714 (lr=4.8895e-04) (hash(x)=6880955)
641
+ 613 train 7.508824 (lr=4.8738e-04) (hash(x)=5804610)
642
+ 614 train 7.432797 (lr=4.8581e-04) (hash(x)=6263347)
643
+ 615 train 7.419387 (lr=4.8424e-04) (hash(x)=5162652)
644
+ 616 train 7.448242 (lr=4.8267e-04) (hash(x)=6055710)
645
+ 617 train 7.406542 (lr=4.8109e-04) (hash(x)=5297006)
646
+ 618 train 7.422055 (lr=4.7952e-04) (hash(x)=5923648)
647
+ 619 train 7.444188 (lr=4.7795e-04) (hash(x)=6728323)
648
+ 620 train 7.439333 (lr=4.7637e-04) (hash(x)=6660239)
649
+ 621 train 7.412280 (lr=4.7479e-04) (hash(x)=3964882)
650
+ 622 train 7.465203 (lr=4.7322e-04) (hash(x)=5868942)
651
+ 623 train 7.384303 (lr=4.7164e-04) (hash(x)=6735993)
652
+ 624 train 7.698549 (lr=4.7006e-04) (hash(x)=5926190)
653
+ 625 train 7.420721 (lr=4.6848e-04) (hash(x)=5150316)
654
+ 626 train 7.440015 (lr=4.6690e-04) (hash(x)=7571467)
655
+ 627 train 7.324441 (lr=4.6532e-04) (hash(x)=5220448)
656
+ 628 train 7.346407 (lr=4.6374e-04) (hash(x)=6790397)
657
+ 629 train 7.735476 (lr=4.6216e-04) (hash(x)=6971289)
658
+ 630 train 7.705855 (lr=4.6058e-04) (hash(x)=7123359)
659
+ 631 train 7.558905 (lr=4.5900e-04) (hash(x)=5012069)
660
+ 632 train 7.522592 (lr=4.5742e-04) (hash(x)=6183574)
661
+ 633 train 7.444564 (lr=4.5583e-04) (hash(x)=6576136)
662
+ 634 train 7.465242 (lr=4.5425e-04) (hash(x)=6157346)
663
+ 635 train 7.448530 (lr=4.5267e-04) (hash(x)=6077647)
664
+ 636 train 7.659685 (lr=4.5109e-04) (hash(x)=8369488)
665
+ 637 train 7.458954 (lr=4.4950e-04) (hash(x)=6032381)
666
+ 638 train 7.400615 (lr=4.4792e-04) (hash(x)=5924756)
667
+ 639 train 7.424606 (lr=4.4634e-04) (hash(x)=8410716)
668
+ 640 train 7.423554 (lr=4.4475e-04) (hash(x)=5803440)
669
+ 641 train 7.488464 (lr=4.4317e-04) (hash(x)=6212516)
670
+ 642 train 7.567594 (lr=4.4158e-04) (hash(x)=7165945)
671
+ 643 train 7.294514 (lr=4.4000e-04) (hash(x)=5298545)
672
+ 644 train 7.537505 (lr=4.3842e-04) (hash(x)=6281739)
673
+ 645 train 7.522266 (lr=4.3683e-04) (hash(x)=5779073)
674
+ 646 train 7.371059 (lr=4.3525e-04) (hash(x)=5090107)
675
+ 647 train 7.430760 (lr=4.3366e-04) (hash(x)=6340947)
676
+ 648 train 7.195107 (lr=4.3208e-04) (hash(x)=4741418)
677
+ 649 train 7.134627 (lr=4.3050e-04) (hash(x)=6530441)
678
+ 650 val loss 7.4764
679
+ 650 val perplexity 1765.8038
680
+ 650 train 7.673190 (lr=4.2891e-04) (hash(x)=6396041)
681
+ 651 train 7.338861 (lr=4.2733e-04) (hash(x)=7818007)
682
+ 652 train 7.347977 (lr=4.2575e-04) (hash(x)=6478457)
683
+ 653 train 7.400683 (lr=4.2417e-04) (hash(x)=6277435)
684
+ 654 train 7.380612 (lr=4.2258e-04) (hash(x)=6042369)
685
+ 655 train 7.385637 (lr=4.2100e-04) (hash(x)=6226320)
686
+ 656 train 7.399297 (lr=4.1942e-04) (hash(x)=5826883)
687
+ 657 train 7.512184 (lr=4.1784e-04) (hash(x)=5463751)
688
+ 658 train 7.449705 (lr=4.1626e-04) (hash(x)=6433080)
689
+ 659 train 7.308726 (lr=4.1468e-04) (hash(x)=6579329)
690
+ 660 train 7.315043 (lr=4.1310e-04) (hash(x)=4724703)
691
+ 661 train 7.647187 (lr=4.1152e-04) (hash(x)=6660540)
692
+ 662 train 7.468054 (lr=4.0994e-04) (hash(x)=5918756)
693
+ 663 train 7.412420 (lr=4.0836e-04) (hash(x)=6491272)
694
+ 664 train 7.311406 (lr=4.0678e-04) (hash(x)=5452863)
695
+ 665 train 7.500142 (lr=4.0521e-04) (hash(x)=6766947)
696
+ 666 train 7.840804 (lr=4.0363e-04) (hash(x)=7459504)
697
+ 667 train 7.518824 (lr=4.0205e-04) (hash(x)=3721218)
698
+ 668 train 7.701254 (lr=4.0048e-04) (hash(x)=7929571)
699
+ 669 train 7.397481 (lr=3.9891e-04) (hash(x)=5229898)
700
+ 670 train 7.408711 (lr=3.9733e-04) (hash(x)=6083090)
701
+ 671 train 7.272428 (lr=3.9576e-04) (hash(x)=6054428)
702
+ 672 train 7.423564 (lr=3.9419e-04) (hash(x)=6072563)
703
+ 673 train 7.562410 (lr=3.9262e-04) (hash(x)=6290464)
704
+ 674 train 7.633403 (lr=3.9105e-04) (hash(x)=6314742)
705
+ 675 train 7.323143 (lr=3.8948e-04) (hash(x)=6904498)
706
+ 676 train 7.449374 (lr=3.8791e-04) (hash(x)=5378107)
707
+ 677 train 7.554780 (lr=3.8634e-04) (hash(x)=7765805)
708
+ 678 train 7.392049 (lr=3.8478e-04) (hash(x)=5871280)
709
+ 679 train 7.531803 (lr=3.8321e-04) (hash(x)=6543764)
710
+ 680 train 7.449203 (lr=3.8165e-04) (hash(x)=7369665)
711
+ 681 train 7.504692 (lr=3.8009e-04) (hash(x)=7204955)
712
+ 682 train 7.451034 (lr=3.7853e-04) (hash(x)=5856543)
713
+ 683 train 7.331480 (lr=3.7697e-04) (hash(x)=4423711)
714
+ 684 train 7.296960 (lr=3.7541e-04) (hash(x)=5818153)
715
+ 685 train 7.393583 (lr=3.7385e-04) (hash(x)=5343477)
716
+ 686 train 7.572783 (lr=3.7229e-04) (hash(x)=6581702)
717
+ 687 train 7.421894 (lr=3.7074e-04) (hash(x)=6595856)
718
+ 688 train 7.354374 (lr=3.6918e-04) (hash(x)=5325378)
719
+ 689 train 7.269637 (lr=3.6763e-04) (hash(x)=7229015)
720
+ 690 train 7.451819 (lr=3.6608e-04) (hash(x)=5918186)
721
+ 691 train 7.335814 (lr=3.6453e-04) (hash(x)=5916034)
722
+ 692 train 7.318521 (lr=3.6298e-04) (hash(x)=5280722)
723
+ 693 train 7.753035 (lr=3.6144e-04) (hash(x)=6939521)
724
+ 694 train 7.458052 (lr=3.5989e-04) (hash(x)=5172566)
725
+ 695 train 7.333789 (lr=3.5835e-04) (hash(x)=5492185)
726
+ 696 train 7.203653 (lr=3.5681e-04) (hash(x)=5779011)
727
+ 697 train 7.469590 (lr=3.5527e-04) (hash(x)=6898006)
728
+ 698 train 7.329360 (lr=3.5373e-04) (hash(x)=5779217)
729
+ 699 train 7.208159 (lr=3.5219e-04) (hash(x)=5462008)
730
+ 700 val loss 7.4643
731
+ 700 val perplexity 1744.7040
732
+ 700 train 7.330283 (lr=3.5066e-04) (hash(x)=6243843)
733
+ 701 train 7.391644 (lr=3.4912e-04) (hash(x)=7821902)
734
+ 702 train 7.446388 (lr=3.4759e-04) (hash(x)=5222501)
735
+ 703 train 7.398406 (lr=3.4606e-04) (hash(x)=6608108)
736
+ 704 train 7.506873 (lr=3.4453e-04) (hash(x)=7602879)
737
+ 705 train 7.436843 (lr=3.4301e-04) (hash(x)=6295394)
738
+ 706 train 7.400076 (lr=3.4148e-04) (hash(x)=6579907)
739
+ 707 train 7.300834 (lr=3.3996e-04) (hash(x)=6803449)
740
+ 708 train 7.372502 (lr=3.3844e-04) (hash(x)=5413733)
741
+ 709 train 7.341568 (lr=3.3692e-04) (hash(x)=7444134)
742
+ 710 train 7.481487 (lr=3.3540e-04) (hash(x)=7028652)
743
+ 711 train 7.280587 (lr=3.3389e-04) (hash(x)=5361923)
744
+ 712 train 7.396523 (lr=3.3238e-04) (hash(x)=5478293)
745
+ 713 train 7.371061 (lr=3.3087e-04) (hash(x)=7803039)
746
+ 714 train 7.550688 (lr=3.2936e-04) (hash(x)=6469017)
747
+ 715 train 7.429708 (lr=3.2785e-04) (hash(x)=7042578)
748
+ 716 train 7.504142 (lr=3.2635e-04) (hash(x)=6927957)
749
+ 717 train 7.310122 (lr=3.2484e-04) (hash(x)=6647007)
750
+ 718 train 7.280912 (lr=3.2334e-04) (hash(x)=6230888)
751
+ 719 train 7.544070 (lr=3.2185e-04) (hash(x)=6303658)
752
+ 720 train 7.572004 (lr=3.2035e-04) (hash(x)=6769880)
753
+ 721 train 7.417339 (lr=3.1886e-04) (hash(x)=6158927)
754
+ 722 train 7.422073 (lr=3.1737e-04) (hash(x)=5991631)
755
+ 723 train 7.275856 (lr=3.1588e-04) (hash(x)=6419656)
756
+ 724 train 7.283467 (lr=3.1440e-04) (hash(x)=7739209)
757
+ 725 train 7.214581 (lr=3.1291e-04) (hash(x)=7891271)
758
+ 726 train 7.549651 (lr=3.1143e-04) (hash(x)=7171433)
759
+ 727 train 7.340818 (lr=3.0995e-04) (hash(x)=7099888)
760
+ 728 train 7.368183 (lr=3.0848e-04) (hash(x)=6103730)
761
+ 729 train 7.489967 (lr=3.0700e-04) (hash(x)=6352424)
762
+ 730 train 7.355912 (lr=3.0553e-04) (hash(x)=4893671)
763
+ 731 train 7.315920 (lr=3.0407e-04) (hash(x)=5863145)
764
+ 732 train 7.425392 (lr=3.0260e-04) (hash(x)=5502233)
765
+ 733 train 7.320953 (lr=3.0114e-04) (hash(x)=6844597)
766
+ 734 train 7.352549 (lr=2.9968e-04) (hash(x)=6744526)
767
+ 735 train 7.487610 (lr=2.9822e-04) (hash(x)=6009861)
768
+ 736 train 7.380800 (lr=2.9677e-04) (hash(x)=6152358)
769
+ 737 train 7.282702 (lr=2.9531e-04) (hash(x)=5374957)
770
+ 738 train 7.310039 (lr=2.9386e-04) (hash(x)=6495309)
771
+ 739 train 7.460121 (lr=2.9242e-04) (hash(x)=3923225)
772
+ 740 train 7.237490 (lr=2.9097e-04) (hash(x)=6255243)
773
+ 741 train 7.406158 (lr=2.8953e-04) (hash(x)=5848862)
774
+ 742 train 7.408817 (lr=2.8810e-04) (hash(x)=4535000)
775
+ 743 train 7.291728 (lr=2.8666e-04) (hash(x)=4440634)
776
+ 744 train 7.289561 (lr=2.8523e-04) (hash(x)=5222077)
777
+ 745 train 7.292668 (lr=2.8380e-04) (hash(x)=4788998)
778
+ 746 train 7.310324 (lr=2.8238e-04) (hash(x)=6108350)
779
+ 747 train 7.242853 (lr=2.8095e-04) (hash(x)=7378165)
780
+ 748 train 7.301981 (lr=2.7953e-04) (hash(x)=5475646)
781
+ 749 train 7.322441 (lr=2.7812e-04) (hash(x)=4771243)
782
+ 750 val loss 7.4106
783
+ 750 val perplexity 1653.4775
784
+ 750 train 7.200758 (lr=2.7670e-04) (hash(x)=6668088)
785
+ 751 train 7.413173 (lr=2.7529e-04) (hash(x)=5716089)
786
+ 752 train 7.366676 (lr=2.7389e-04) (hash(x)=6991946)
787
+ 753 train 7.264510 (lr=2.7248e-04) (hash(x)=7991009)
788
+ 754 train 7.388874 (lr=2.7108e-04) (hash(x)=6037018)
789
+ 755 train 7.358085 (lr=2.6969e-04) (hash(x)=5348607)
790
+ 756 train 7.480811 (lr=2.6829e-04) (hash(x)=5190526)
791
+ 757 train 7.375781 (lr=2.6690e-04) (hash(x)=6036146)
792
+ 758 train 7.267815 (lr=2.6551e-04) (hash(x)=5329783)
793
+ 759 train 7.354149 (lr=2.6413e-04) (hash(x)=6658718)
794
+ 760 train 7.236174 (lr=2.6275e-04) (hash(x)=4820939)
795
+ 761 train 7.255917 (lr=2.6137e-04) (hash(x)=6176253)
796
+ 762 train 7.239396 (lr=2.6000e-04) (hash(x)=5316750)
797
+ 763 train 7.320350 (lr=2.5863e-04) (hash(x)=4542895)
798
+ 764 train 7.109778 (lr=2.5726e-04) (hash(x)=5395673)
799
+ 765 train 7.210369 (lr=2.5590e-04) (hash(x)=5549073)
800
+ 766 train 7.447423 (lr=2.5454e-04) (hash(x)=5233723)
801
+ 767 train 7.503806 (lr=2.5319e-04) (hash(x)=5342783)
802
+ 768 train 7.444317 (lr=2.5183e-04) (hash(x)=5552448)
803
+ 769 train 7.579399 (lr=2.5048e-04) (hash(x)=6549546)
804
+ 770 train 7.244279 (lr=2.4914e-04) (hash(x)=3816493)
805
+ 771 train 7.288412 (lr=2.4780e-04) (hash(x)=6170599)
806
+ 772 train 7.754220 (lr=2.4646e-04) (hash(x)=6708759)
807
+ 773 train 7.374780 (lr=2.4513e-04) (hash(x)=8483134)
808
+ 774 train 7.959329 (lr=2.4380e-04) (hash(x)=5583732)
809
+ 775 train 7.355072 (lr=2.4247e-04) (hash(x)=4865036)
810
+ 776 train 7.304970 (lr=2.4115e-04) (hash(x)=7234336)
811
+ 777 train 7.296145 (lr=2.3983e-04) (hash(x)=4433890)
812
+ 778 train 7.624659 (lr=2.3852e-04) (hash(x)=7484636)
813
+ 779 train 8.136867 (lr=2.3720e-04) (hash(x)=6738597)
814
+ 780 train 7.316218 (lr=2.3590e-04) (hash(x)=6752421)
815
+ 781 train 7.529748 (lr=2.3460e-04) (hash(x)=5459620)
816
+ 782 train 7.429182 (lr=2.3330e-04) (hash(x)=6718131)
817
+ 783 train 7.411053 (lr=2.3200e-04) (hash(x)=4604074)
818
+ 784 train 7.276110 (lr=2.3071e-04) (hash(x)=5946124)
819
+ 785 train 7.552727 (lr=2.2942e-04) (hash(x)=8045501)
820
+ 786 train 7.652944 (lr=2.2814e-04) (hash(x)=5712725)
821
+ 787 train 7.375810 (lr=2.2686e-04) (hash(x)=6176464)
822
+ 788 train 7.444021 (lr=2.2559e-04) (hash(x)=6276702)
823
+ 789 train 7.253860 (lr=2.2432e-04) (hash(x)=6127317)
824
+ 790 train 7.403112 (lr=2.2305e-04) (hash(x)=7221374)
825
+ 791 train 7.368992 (lr=2.2179e-04) (hash(x)=5571462)
826
+ 792 train 7.372131 (lr=2.2053e-04) (hash(x)=7518021)
827
+ 793 train 7.380013 (lr=2.1928e-04) (hash(x)=6640755)
828
+ 794 train 7.451841 (lr=2.1803e-04) (hash(x)=6870405)
829
+ 795 train 7.450054 (lr=2.1678e-04) (hash(x)=6147295)
830
+ 796 train 7.324287 (lr=2.1554e-04) (hash(x)=6782583)
831
+ 797 train 7.428697 (lr=2.1431e-04) (hash(x)=6120461)
832
+ 798 train 7.365417 (lr=2.1308e-04) (hash(x)=5893870)
833
+ 799 train 7.297222 (lr=2.1185e-04) (hash(x)=5123928)
834
+ 800 val loss 7.3868
835
+ 800 val perplexity 1614.4825
836
+ 800 train 7.367552 (lr=2.1063e-04) (hash(x)=7037251)
837
+ 801 train 7.364194 (lr=2.0941e-04) (hash(x)=5428813)
838
+ 802 train 7.293941 (lr=2.0819e-04) (hash(x)=5550387)
839
+ 803 train 7.382787 (lr=2.0698e-04) (hash(x)=6448586)
840
+ 804 train 7.525928 (lr=2.0578e-04) (hash(x)=5916934)
841
+ 805 train 7.477815 (lr=2.0458e-04) (hash(x)=8116987)
842
+ 806 train 7.459467 (lr=2.0338e-04) (hash(x)=6210899)
843
+ 807 train 7.526366 (lr=2.0219e-04) (hash(x)=5974342)
844
+ 808 train 7.548513 (lr=2.0100e-04) (hash(x)=7236199)
845
+ 809 train 7.490525 (lr=1.9982e-04) (hash(x)=5261336)
846
+ 810 train 7.321994 (lr=1.9864e-04) (hash(x)=6310720)
847
+ 811 train 7.351141 (lr=1.9747e-04) (hash(x)=5307056)
848
+ 812 train 7.204102 (lr=1.9630e-04) (hash(x)=7356633)
849
+ 813 train 7.315934 (lr=1.9514e-04) (hash(x)=5865119)
850
+ 814 train 7.221696 (lr=1.9398e-04) (hash(x)=5979205)
851
+ 815 train 7.363548 (lr=1.9283e-04) (hash(x)=6672380)
852
+ 816 train 7.363645 (lr=1.9168e-04) (hash(x)=5565030)
853
+ 817 train 7.338445 (lr=1.9053e-04) (hash(x)=6886810)
854
+ 818 train 7.379807 (lr=1.8939e-04) (hash(x)=5926478)
855
+ 819 train 7.575223 (lr=1.8826e-04) (hash(x)=6564071)
856
+ 820 train 7.236511 (lr=1.8713e-04) (hash(x)=5024280)
857
+ 821 train 7.235013 (lr=1.8600e-04) (hash(x)=7705089)
858
+ 822 train 7.209301 (lr=1.8488e-04) (hash(x)=4037371)
859
+ 823 train 7.302983 (lr=1.8377e-04) (hash(x)=5773073)
860
+ 824 train 7.315555 (lr=1.8266e-04) (hash(x)=7158139)
861
+ 825 train 7.161967 (lr=1.8155e-04) (hash(x)=5905976)
862
+ 826 train 7.592534 (lr=1.8045e-04) (hash(x)=13897369)
863
+ 827 train 8.607789 (lr=1.7936e-04) (hash(x)=6936653)
864
+ 828 train 7.366424 (lr=1.7827e-04) (hash(x)=6592246)
865
+ 829 train 7.314444 (lr=1.7718e-04) (hash(x)=6592773)
866
+ 830 train 7.387743 (lr=1.7610e-04) (hash(x)=5461178)
867
+ 831 train 7.169720 (lr=1.7503e-04) (hash(x)=6796551)
868
+ 832 train 7.379313 (lr=1.7396e-04) (hash(x)=4721691)
869
+ 833 train 7.304832 (lr=1.7289e-04) (hash(x)=6151589)
870
+ 834 train 7.435142 (lr=1.7183e-04) (hash(x)=7008995)
871
+ 835 train 7.765347 (lr=1.7078e-04) (hash(x)=7045164)
872
+ 836 train 7.352397 (lr=1.6973e-04) (hash(x)=6614221)
873
+ 837 train 7.389338 (lr=1.6869e-04) (hash(x)=6650506)
874
+ 838 train 7.202075 (lr=1.6765e-04) (hash(x)=6363762)
875
+ 839 train 7.357403 (lr=1.6661e-04) (hash(x)=7839748)
876
+ 840 train 7.275794 (lr=1.6559e-04) (hash(x)=5124077)
877
+ 841 train 7.775998 (lr=1.6456e-04) (hash(x)=6028671)
878
+ 842 train 7.230673 (lr=1.6355e-04) (hash(x)=5427175)
879
+ 843 train 7.538814 (lr=1.6253e-04) (hash(x)=6118408)
880
+ 844 train 7.348313 (lr=1.6153e-04) (hash(x)=6022552)
881
+ 845 train 7.335935 (lr=1.6053e-04) (hash(x)=5757526)
882
+ 846 train 7.794827 (lr=1.5953e-04) (hash(x)=6476549)
883
+ 847 train 7.155478 (lr=1.5854e-04) (hash(x)=6030172)
884
+ 848 train 7.501359 (lr=1.5756e-04) (hash(x)=5917229)
885
+ 849 train 7.556592 (lr=1.5658e-04) (hash(x)=6719882)
886
+ 850 val loss 7.3924
887
+ 850 val perplexity 1623.5242
888
+ 850 train 7.430698 (lr=1.5560e-04) (hash(x)=5441498)
889
+ 851 train 7.318513 (lr=1.5463e-04) (hash(x)=5028050)
890
+ 852 train 7.352950 (lr=1.5367e-04) (hash(x)=8487491)
891
+ 853 train 7.368348 (lr=1.5271e-04) (hash(x)=5682910)
892
+ 854 train 7.580895 (lr=1.5176e-04) (hash(x)=6781362)
893
+ 855 train 7.361157 (lr=1.5082e-04) (hash(x)=5747891)
894
+ 856 train 7.408920 (lr=1.4988e-04) (hash(x)=7627661)
895
+ 857 train 7.061060 (lr=1.4894e-04) (hash(x)=5734208)
896
+ 858 train 7.204166 (lr=1.4801e-04) (hash(x)=6162015)
897
+ 859 train 7.439414 (lr=1.4709e-04) (hash(x)=7024381)
898
+ 860 train 7.621377 (lr=1.4617e-04) (hash(x)=5198498)
899
+ 861 train 7.611856 (lr=1.4526e-04) (hash(x)=7814361)
900
+ 862 train 7.476402 (lr=1.4435e-04) (hash(x)=7394995)
901
+ 863 train 7.366238 (lr=1.4345e-04) (hash(x)=5765893)
902
+ 864 train 7.186849 (lr=1.4255e-04) (hash(x)=4990435)
903
+ 865 train 7.316489 (lr=1.4166e-04) (hash(x)=5572048)
904
+ 866 train 7.485728 (lr=1.4078e-04) (hash(x)=7093293)
905
+ 867 train 7.377780 (lr=1.3990e-04) (hash(x)=5610058)
906
+ 868 train 7.266065 (lr=1.3903e-04) (hash(x)=6771468)
907
+ 869 train 7.299858 (lr=1.3816e-04) (hash(x)=4459483)
908
+ 870 train 7.221372 (lr=1.3730e-04) (hash(x)=4926546)
909
+ 871 train 7.315898 (lr=1.3645e-04) (hash(x)=6758411)
910
+ 872 train 7.269675 (lr=1.3560e-04) (hash(x)=7394277)
911
+ 873 train 7.388486 (lr=1.3476e-04) (hash(x)=6701381)
912
+ 874 train 7.748867 (lr=1.3392e-04) (hash(x)=8120966)
913
+ 875 train 7.275156 (lr=1.3309e-04) (hash(x)=5416400)
914
+ 876 train 7.265032 (lr=1.3227e-04) (hash(x)=6292778)
915
+ 877 train 7.195174 (lr=1.3145e-04) (hash(x)=5919569)
916
+ 878 train 7.501742 (lr=1.3063e-04) (hash(x)=7561740)
917
+ 879 train 7.256497 (lr=1.2983e-04) (hash(x)=5282870)
918
+ 880 train 7.302969 (lr=1.2903e-04) (hash(x)=5473238)
919
+ 881 train 7.198829 (lr=1.2823e-04) (hash(x)=8072181)
920
+ 882 train 7.207558 (lr=1.2744e-04) (hash(x)=6757322)
921
+ 883 train 7.180265 (lr=1.2666e-04) (hash(x)=7996770)
922
+ 884 train 7.242212 (lr=1.2588e-04) (hash(x)=3181021)
923
+ 885 train 7.288494 (lr=1.2511e-04) (hash(x)=5093519)
924
+ 886 train 7.265769 (lr=1.2435e-04) (hash(x)=5391352)
925
+ 887 train 7.310280 (lr=1.2359e-04) (hash(x)=6270917)
926
+ 888 train 7.291438 (lr=1.2284e-04) (hash(x)=5818195)
927
+ 889 train 7.043289 (lr=1.2209e-04) (hash(x)=5852780)
928
+ 890 train 6.959786 (lr=1.2135e-04) (hash(x)=5777928)
929
+ 891 train 7.165984 (lr=1.2062e-04) (hash(x)=5327728)
930
+ 892 train 7.211112 (lr=1.1989e-04) (hash(x)=7651281)
931
+ 893 train 7.404027 (lr=1.1917e-04) (hash(x)=6260765)
932
+ 894 train 7.306460 (lr=1.1845e-04) (hash(x)=6265102)
933
+ 895 train 7.483799 (lr=1.1774e-04) (hash(x)=6212894)
934
+ 896 train 7.234100 (lr=1.1704e-04) (hash(x)=6901967)
935
+ 897 train 7.310962 (lr=1.1634e-04) (hash(x)=6724010)
936
+ 898 train 7.295437 (lr=1.1565e-04) (hash(x)=6299033)
937
+ 899 train 7.468415 (lr=1.1497e-04) (hash(x)=6059934)
938
+ 900 val loss 7.3844
939
+ 900 val perplexity 1610.7269
940
+ 900 train 7.173821 (lr=1.1429e-04) (hash(x)=6029696)
941
+ 901 train 7.090920 (lr=1.1362e-04) (hash(x)=5587382)
942
+ 902 train 7.309878 (lr=1.1295e-04) (hash(x)=5890076)
943
+ 903 train 7.230813 (lr=1.1229e-04) (hash(x)=7274108)
944
+ 904 train 7.307291 (lr=1.1164e-04) (hash(x)=7975528)
945
+ 905 train 7.338127 (lr=1.1099e-04) (hash(x)=6085396)
946
+ 906 train 7.234980 (lr=1.1036e-04) (hash(x)=6724496)
947
+ 907 train 7.233382 (lr=1.0972e-04) (hash(x)=4784666)
948
+ 908 train 7.240129 (lr=1.0909e-04) (hash(x)=7193845)
949
+ 909 train 7.410401 (lr=1.0847e-04) (hash(x)=6288638)
950
+ 910 train 7.360919 (lr=1.0786e-04) (hash(x)=5685285)
951
+ 911 train 7.336857 (lr=1.0725e-04) (hash(x)=7714288)
952
+ 912 train 7.423934 (lr=1.0665e-04) (hash(x)=5983679)
953
+ 913 train 7.372122 (lr=1.0606e-04) (hash(x)=5534748)
954
+ 914 train 7.324170 (lr=1.0547e-04) (hash(x)=6545471)
955
+ 915 train 7.319057 (lr=1.0489e-04) (hash(x)=6361270)
956
+ 916 train 7.318522 (lr=1.0431e-04) (hash(x)=6257744)
957
+ 917 train 7.329871 (lr=1.0374e-04) (hash(x)=4682956)
958
+ 918 train 7.214496 (lr=1.0318e-04) (hash(x)=4521756)
959
+ 919 train 7.367909 (lr=1.0262e-04) (hash(x)=7568935)
960
+ 920 train 7.282677 (lr=1.0207e-04) (hash(x)=7969353)
961
+ 921 train 7.454181 (lr=1.0153e-04) (hash(x)=7973311)
962
+ 922 train 7.171074 (lr=1.0099e-04) (hash(x)=4843937)
963
+ 923 train 7.308990 (lr=1.0046e-04) (hash(x)=6987352)
964
+ 924 train 7.402907 (lr=9.9941e-05) (hash(x)=7179418)
965
+ 925 train 7.096810 (lr=9.9425e-05) (hash(x)=5413904)
966
+ 926 train 7.454630 (lr=9.8915e-05) (hash(x)=6028105)
967
+ 927 train 7.395092 (lr=9.8411e-05) (hash(x)=6732924)
968
+ 928 train 7.301415 (lr=9.7915e-05) (hash(x)=4547651)
969
+ 929 train 7.445501 (lr=9.7424e-05) (hash(x)=6219754)
970
+ 930 train 7.239707 (lr=9.6941e-05) (hash(x)=7465186)
971
+ 931 train 7.370588 (lr=9.6464e-05) (hash(x)=6839070)
972
+ 932 train 7.258953 (lr=9.5994e-05) (hash(x)=5390324)
973
+ 933 train 7.227988 (lr=9.5530e-05) (hash(x)=6090120)
974
+ 934 train 7.827703 (lr=9.5073e-05) (hash(x)=7123181)
975
+ 935 train 7.514692 (lr=9.4623e-05) (hash(x)=7115176)
976
+ 936 train 7.521098 (lr=9.4180e-05) (hash(x)=7016990)
977
+ 937 train 7.282057 (lr=9.3743e-05) (hash(x)=4989299)
978
+ 938 train 7.311768 (lr=9.3313e-05) (hash(x)=6783311)
979
+ 939 train 7.276971 (lr=9.2889e-05) (hash(x)=7497590)
980
+ 940 train 7.496283 (lr=9.2473e-05) (hash(x)=6767514)
981
+ 941 train 7.257604 (lr=9.2063e-05) (hash(x)=3977557)
982
+ 942 train 7.365760 (lr=9.1659e-05) (hash(x)=4806492)
983
+ 943 train 7.307173 (lr=9.1263e-05) (hash(x)=5524165)
984
+ 944 train 7.354510 (lr=9.0873e-05) (hash(x)=4930752)
985
+ 945 train 7.328971 (lr=9.0490e-05) (hash(x)=5124936)
986
+ 946 train 7.207413 (lr=9.0114e-05) (hash(x)=6877351)
987
+ 947 train 7.497658 (lr=8.9744e-05) (hash(x)=5522119)
988
+ 948 train 7.296420 (lr=8.9382e-05) (hash(x)=6183512)
989
+ 949 train 7.211228 (lr=8.9026e-05) (hash(x)=5030865)
990
+ 950 val loss 7.3650
991
+ 950 val perplexity 1579.6495
992
+ 950 train 7.316410 (lr=8.8677e-05) (hash(x)=6619812)
993
+ 951 train 7.308295 (lr=8.8335e-05) (hash(x)=6411783)
994
+ 952 train 7.309752 (lr=8.7999e-05) (hash(x)=6748982)
995
+ 953 train 7.211441 (lr=8.7670e-05) (hash(x)=6581937)
996
+ 954 train 7.318934 (lr=8.7349e-05) (hash(x)=7881022)
997
+ 955 train 7.280806 (lr=8.7034e-05) (hash(x)=5116910)
998
+ 956 train 7.076656 (lr=8.6725e-05) (hash(x)=4880833)
999
+ 957 train 7.137047 (lr=8.6424e-05) (hash(x)=7353459)
1000
+ 958 train 7.185609 (lr=8.6130e-05) (hash(x)=4951302)
1001
+ 959 train 7.101201 (lr=8.5842e-05) (hash(x)=6344535)
1002
+ 960 train 7.729256 (lr=8.5561e-05) (hash(x)=7411763)
1003
+ 961 train 7.278145 (lr=8.5287e-05) (hash(x)=5840169)
1004
+ 962 train 7.370535 (lr=8.5020e-05) (hash(x)=5784788)
1005
+ 963 train 7.179125 (lr=8.4760e-05) (hash(x)=6754290)
1006
+ 964 train 7.468663 (lr=8.4507e-05) (hash(x)=6369963)
1007
+ 965 train 7.527931 (lr=8.4260e-05) (hash(x)=6484573)
1008
+ 966 train 7.120367 (lr=8.4021e-05) (hash(x)=6516338)
1009
+ 967 train 7.165777 (lr=8.3788e-05) (hash(x)=5260125)
1010
+ 968 train 7.280411 (lr=8.3563e-05) (hash(x)=5781624)
1011
+ 969 train 7.096311 (lr=8.3344e-05) (hash(x)=4881179)
1012
+ 970 train 7.210111 (lr=8.3132e-05) (hash(x)=6707821)
1013
+ 971 train 7.345204 (lr=8.2927e-05) (hash(x)=5001634)
1014
+ 972 train 7.223913 (lr=8.2729e-05) (hash(x)=4402949)
1015
+ 973 train 7.243615 (lr=8.2537e-05) (hash(x)=7424810)
1016
+ 974 train 7.101247 (lr=8.2353e-05) (hash(x)=4814803)
1017
+ 975 train 7.021827 (lr=8.2176e-05) (hash(x)=5176431)
1018
+ 976 train 7.219872 (lr=8.2005e-05) (hash(x)=5769393)
1019
+ 977 train 7.159587 (lr=8.1842e-05) (hash(x)=5421624)
1020
+ 978 train 7.379766 (lr=8.1685e-05) (hash(x)=5999773)
1021
+ 979 train 7.101814 (lr=8.1536e-05) (hash(x)=5674439)
1022
+ 980 train 7.188630 (lr=8.1393e-05) (hash(x)=6345351)
1023
+ 981 train 7.132586 (lr=8.1257e-05) (hash(x)=5819157)
1024
+ 982 train 7.346567 (lr=8.1128e-05) (hash(x)=11159355)
1025
+ 983 train 7.314591 (lr=8.1007e-05) (hash(x)=6506856)
1026
+ 984 train 7.233915 (lr=8.0892e-05) (hash(x)=6331023)
1027
+ 985 train 7.131012 (lr=8.0784e-05) (hash(x)=6119124)
1028
+ 986 train 7.341985 (lr=8.0683e-05) (hash(x)=8142528)
1029
+ 987 train 7.499068 (lr=8.0589e-05) (hash(x)=6241669)
1030
+ 988 train 7.133759 (lr=8.0502e-05) (hash(x)=3852991)
1031
+ 989 train 6.899403 (lr=8.0422e-05) (hash(x)=4992756)
1032
+ 990 train 7.569063 (lr=8.0348e-05) (hash(x)=7097145)
1033
+ 991 train 7.367691 (lr=8.0282e-05) (hash(x)=7277927)
1034
+ 992 train 7.425050 (lr=8.0223e-05) (hash(x)=5661400)
1035
+ 993 train 7.348803 (lr=8.0171e-05) (hash(x)=6622097)
1036
+ 994 train 7.481235 (lr=8.0125e-05) (hash(x)=5016804)
1037
+ 995 train 7.551160 (lr=8.0087e-05) (hash(x)=7026582)
1038
+ 996 train 7.336741 (lr=8.0056e-05) (hash(x)=5639136)
1039
+ 997 train 7.495724 (lr=8.0031e-05) (hash(x)=5840987)
1040
+ 998 train 7.117863 (lr=8.0014e-05) (hash(x)=5265410)
1041
+ 999 val loss 7.3536
1042
+ 999 val perplexity 1561.8075
1043
+ 999 train 7.113688 (lr=8.0003e-05) (hash(x)=4351074)
lr8e-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 sha256:5ed5419c6654024e16dcbb1cdcd4529170c528633973fdd349cce7e08d78a2ef
3
+ size 36163010
lr8e-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 sha256:2db5a15475d530d20c7d4d253c0e4b6fba01662ad68e581e964ad96b418ac555
3
+ size 70764358