andrew-healey commited on
Commit
5fdf209
·
verified ·
1 Parent(s): 588dcfc

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size5120_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size5120_seq_len512", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 512, "batch_size": null, "total_batch_size": 5120, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0001, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1e-4_total_batch_size5120_seq_len512/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr1e-4_total_batch_size5120_seq_len512/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7800
3
+ 0 val perplexity 130608.6406
4
+ 0 train 11.762018 (lr=3.4965e-07) (hash(x)=22886834)
5
+ 1 train 11.772878 (lr=6.9930e-07) (hash(x)=26375038)
6
+ 2 train 11.735553 (lr=1.0490e-06) (hash(x)=30598777)
7
+ 3 train 11.784902 (lr=1.3986e-06) (hash(x)=27234506)
8
+ 4 train 11.726120 (lr=1.7483e-06) (hash(x)=27767880)
9
+ 5 train 11.781272 (lr=2.0979e-06) (hash(x)=23702020)
10
+ 6 train 11.744897 (lr=2.4476e-06) (hash(x)=31986844)
11
+ 7 train 11.758860 (lr=2.7972e-06) (hash(x)=20782690)
12
+ 8 train 11.758535 (lr=3.1469e-06) (hash(x)=25201599)
13
+ 9 train 11.797496 (lr=3.4965e-06) (hash(x)=23094976)
14
+ 10 train 11.773027 (lr=3.8462e-06) (hash(x)=23841096)
15
+ 11 train 11.744201 (lr=4.1958e-06) (hash(x)=26532095)
16
+ 12 train 11.771453 (lr=4.5455e-06) (hash(x)=24432298)
17
+ 13 train 11.706306 (lr=4.8951e-06) (hash(x)=27151649)
18
+ 14 train 11.711528 (lr=5.2448e-06) (hash(x)=24596846)
19
+ 15 train 11.736303 (lr=5.5944e-06) (hash(x)=23890908)
20
+ 16 train 11.721211 (lr=5.9441e-06) (hash(x)=28913955)
21
+ 17 train 11.743484 (lr=6.2937e-06) (hash(x)=25588236)
22
+ 18 train 11.715164 (lr=6.6434e-06) (hash(x)=23770034)
23
+ 19 train 11.749471 (lr=6.9930e-06) (hash(x)=24011372)
24
+ 20 train 11.657885 (lr=7.3427e-06) (hash(x)=25441898)
25
+ 21 train 11.650809 (lr=7.6923e-06) (hash(x)=28375581)
26
+ 22 train 11.681827 (lr=8.0420e-06) (hash(x)=24046679)
27
+ 23 train 11.661906 (lr=8.3916e-06) (hash(x)=24611628)
28
+ 24 train 11.662688 (lr=8.7413e-06) (hash(x)=26169030)
29
+ 25 train 11.630103 (lr=9.0909e-06) (hash(x)=30298407)
30
+ 26 train 11.633911 (lr=9.4406e-06) (hash(x)=23711112)
31
+ 27 train 11.600864 (lr=9.7902e-06) (hash(x)=19245352)
32
+ 28 train 11.580851 (lr=1.0140e-05) (hash(x)=21529136)
33
+ 29 train 11.576693 (lr=1.0490e-05) (hash(x)=28936608)
34
+ 30 train 11.552907 (lr=1.0839e-05) (hash(x)=24339013)
35
+ 31 train 11.528208 (lr=1.1189e-05) (hash(x)=25767553)
36
+ 32 train 11.514632 (lr=1.1538e-05) (hash(x)=26439905)
37
+ 33 train 11.522805 (lr=1.1888e-05) (hash(x)=31093473)
38
+ 34 train 11.437497 (lr=1.2238e-05) (hash(x)=25450374)
39
+ 35 train 11.509692 (lr=1.2587e-05) (hash(x)=24809873)
40
+ 36 train 11.482800 (lr=1.2937e-05) (hash(x)=23253252)
41
+ 37 train 11.508191 (lr=1.3287e-05) (hash(x)=27852919)
42
+ 38 train 11.422361 (lr=1.3636e-05) (hash(x)=23327497)
43
+ 39 train 11.440041 (lr=1.3986e-05) (hash(x)=22512166)
44
+ 40 train 11.441393 (lr=1.4336e-05) (hash(x)=22859419)
45
+ 41 train 11.398632 (lr=1.4685e-05) (hash(x)=27620338)
46
+ 42 train 11.363798 (lr=1.5035e-05) (hash(x)=26397837)
47
+ 43 train 11.393616 (lr=1.5385e-05) (hash(x)=28092148)
48
+ 44 train 11.348220 (lr=1.5734e-05) (hash(x)=24662703)
49
+ 45 train 11.339126 (lr=1.6084e-05) (hash(x)=27938767)
50
+ 46 train 11.360202 (lr=1.6434e-05) (hash(x)=26037988)
51
+ 47 train 11.276166 (lr=1.6783e-05) (hash(x)=24732833)
52
+ 48 train 11.238583 (lr=1.7133e-05) (hash(x)=25259526)
53
+ 49 train 11.252920 (lr=1.7483e-05) (hash(x)=23200230)
54
+ 50 val loss 11.2281
55
+ 50 val perplexity 75211.6719
56
+ 50 train 11.260806 (lr=1.7832e-05) (hash(x)=26721357)
57
+ 51 train 11.232753 (lr=1.8182e-05) (hash(x)=22694718)
58
+ 52 train 11.165962 (lr=1.8531e-05) (hash(x)=28066766)
59
+ 53 train 11.141028 (lr=1.8881e-05) (hash(x)=23125151)
60
+ 54 train 11.162821 (lr=1.9231e-05) (hash(x)=27193725)
61
+ 55 train 11.090745 (lr=1.9580e-05) (hash(x)=25129410)
62
+ 56 train 11.083343 (lr=1.9930e-05) (hash(x)=24263988)
63
+ 57 train 10.970807 (lr=2.0280e-05) (hash(x)=23059154)
64
+ 58 train 11.036725 (lr=2.0629e-05) (hash(x)=26063864)
65
+ 59 train 11.067615 (lr=2.0979e-05) (hash(x)=27858570)
66
+ 60 train 10.916255 (lr=2.1329e-05) (hash(x)=23874620)
67
+ 61 train 10.912874 (lr=2.1678e-05) (hash(x)=22402617)
68
+ 62 train 10.893953 (lr=2.2028e-05) (hash(x)=23600822)
69
+ 63 train 10.727982 (lr=2.2378e-05) (hash(x)=26582391)
70
+ 64 train 10.687499 (lr=2.2727e-05) (hash(x)=23225283)
71
+ 65 train 10.842873 (lr=2.3077e-05) (hash(x)=26075451)
72
+ 66 train 10.820406 (lr=2.3427e-05) (hash(x)=24723419)
73
+ 67 train 10.720102 (lr=2.3776e-05) (hash(x)=27279806)
74
+ 68 train 10.630625 (lr=2.4126e-05) (hash(x)=25870391)
75
+ 69 train 10.676785 (lr=2.4476e-05) (hash(x)=26188136)
76
+ 70 train 10.614663 (lr=2.4825e-05) (hash(x)=30373443)
77
+ 71 train 10.690965 (lr=2.5175e-05) (hash(x)=26472336)
78
+ 72 train 10.593505 (lr=2.5524e-05) (hash(x)=26651572)
79
+ 73 train 10.562083 (lr=2.5874e-05) (hash(x)=26376212)
80
+ 74 train 10.544047 (lr=2.6224e-05) (hash(x)=26733350)
81
+ 75 train 10.636673 (lr=2.6573e-05) (hash(x)=28301589)
82
+ 76 train 10.549991 (lr=2.6923e-05) (hash(x)=27599559)
83
+ 77 train 10.432754 (lr=2.7273e-05) (hash(x)=28035221)
84
+ 78 train 10.493763 (lr=2.7622e-05) (hash(x)=25016783)
85
+ 79 train 10.528786 (lr=2.7972e-05) (hash(x)=27654289)
86
+ 80 train 10.554334 (lr=2.8322e-05) (hash(x)=24597558)
87
+ 81 train 10.510472 (lr=2.8671e-05) (hash(x)=21560904)
88
+ 82 train 10.349799 (lr=2.9021e-05) (hash(x)=21983837)
89
+ 83 train 10.391974 (lr=2.9371e-05) (hash(x)=24995715)
90
+ 84 train 10.429926 (lr=2.9720e-05) (hash(x)=29876413)
91
+ 85 train 10.378074 (lr=3.0070e-05) (hash(x)=23792508)
92
+ 86 train 10.313098 (lr=3.0420e-05) (hash(x)=25509120)
93
+ 87 train 10.453481 (lr=3.0769e-05) (hash(x)=26559876)
94
+ 88 train 10.381368 (lr=3.1119e-05) (hash(x)=23569647)
95
+ 89 train 10.400917 (lr=3.1469e-05) (hash(x)=25758852)
96
+ 90 train 10.305238 (lr=3.1818e-05) (hash(x)=25706298)
97
+ 91 train 10.182747 (lr=3.2168e-05) (hash(x)=28364895)
98
+ 92 train 10.245559 (lr=3.2517e-05) (hash(x)=25304663)
99
+ 93 train 10.323430 (lr=3.2867e-05) (hash(x)=25269299)
100
+ 94 train 10.472919 (lr=3.3217e-05) (hash(x)=25870566)
101
+ 95 train 10.285725 (lr=3.3566e-05) (hash(x)=21770329)
102
+ 96 train 10.470305 (lr=3.3916e-05) (hash(x)=27595900)
103
+ 97 train 10.479575 (lr=3.4266e-05) (hash(x)=24785397)
104
+ 98 train 10.254763 (lr=3.4615e-05) (hash(x)=21521480)
105
+ 99 train 10.292060 (lr=3.4965e-05) (hash(x)=24628606)
106
+ 100 val loss 10.2814
107
+ 100 val perplexity 29183.6934
108
+ 100 train 10.317454 (lr=3.5315e-05) (hash(x)=24670150)
109
+ 101 train 10.309859 (lr=3.5664e-05) (hash(x)=23181910)
110
+ 102 train 10.183026 (lr=3.6014e-05) (hash(x)=22714991)
111
+ 103 train 10.288335 (lr=3.6364e-05) (hash(x)=22723459)
112
+ 104 train 10.221931 (lr=3.6713e-05) (hash(x)=21524316)
113
+ 105 train 10.208892 (lr=3.7063e-05) (hash(x)=25506632)
114
+ 106 train 10.047304 (lr=3.7413e-05) (hash(x)=21675672)
115
+ 107 train 10.167673 (lr=3.7762e-05) (hash(x)=22897919)
116
+ 108 train 10.187581 (lr=3.8112e-05) (hash(x)=23321631)
117
+ 109 train 10.213958 (lr=3.8462e-05) (hash(x)=26546719)
118
+ 110 train 10.697638 (lr=3.8811e-05) (hash(x)=31962348)
119
+ 111 train 10.209302 (lr=3.9161e-05) (hash(x)=30338342)
120
+ 112 train 10.178423 (lr=3.9510e-05) (hash(x)=23724471)
121
+ 113 train 10.162774 (lr=3.9860e-05) (hash(x)=29175888)
122
+ 114 train 10.253581 (lr=4.0210e-05) (hash(x)=23256716)
123
+ 115 train 10.210868 (lr=4.0559e-05) (hash(x)=27063280)
124
+ 116 train 10.166686 (lr=4.0909e-05) (hash(x)=31057659)
125
+ 117 train 10.182264 (lr=4.1259e-05) (hash(x)=32915097)
126
+ 118 train 10.137509 (lr=4.1608e-05) (hash(x)=28842717)
127
+ 119 train 10.130731 (lr=4.1958e-05) (hash(x)=25678059)
128
+ 120 train 10.131679 (lr=4.2308e-05) (hash(x)=21593510)
129
+ 121 train 10.016986 (lr=4.2657e-05) (hash(x)=20083773)
130
+ 122 train 10.050262 (lr=4.3007e-05) (hash(x)=23002820)
131
+ 123 train 10.032412 (lr=4.3357e-05) (hash(x)=21853028)
132
+ 124 train 10.062265 (lr=4.3706e-05) (hash(x)=26985625)
133
+ 125 train 10.012515 (lr=4.4056e-05) (hash(x)=21808483)
134
+ 126 train 10.122076 (lr=4.4406e-05) (hash(x)=28873251)
135
+ 127 train 10.074570 (lr=4.4755e-05) (hash(x)=26109335)
136
+ 128 train 10.080077 (lr=4.5105e-05) (hash(x)=26334674)
137
+ 129 train 10.049609 (lr=4.5455e-05) (hash(x)=24916754)
138
+ 130 train 10.095092 (lr=4.5804e-05) (hash(x)=25449624)
139
+ 131 train 10.160995 (lr=4.6154e-05) (hash(x)=25334848)
140
+ 132 train 10.186499 (lr=4.6503e-05) (hash(x)=27484863)
141
+ 133 train 10.032508 (lr=4.6853e-05) (hash(x)=24917705)
142
+ 134 train 9.876437 (lr=4.7203e-05) (hash(x)=25143449)
143
+ 135 train 9.926795 (lr=4.7552e-05) (hash(x)=25044885)
144
+ 136 train 10.132649 (lr=4.7902e-05) (hash(x)=27821028)
145
+ 137 train 10.067077 (lr=4.8252e-05) (hash(x)=28747022)
146
+ 138 train 10.020879 (lr=4.8601e-05) (hash(x)=27182888)
147
+ 139 train 9.902999 (lr=4.8951e-05) (hash(x)=23678349)
148
+ 140 train 10.043532 (lr=4.9301e-05) (hash(x)=23593235)
149
+ 141 train 10.135101 (lr=4.9650e-05) (hash(x)=28529813)
150
+ 142 train 10.196104 (lr=5.0000e-05) (hash(x)=32074661)
151
+ 143 train 9.922831 (lr=5.0350e-05) (hash(x)=28870690)
152
+ 144 train 10.000449 (lr=5.0699e-05) (hash(x)=27307705)
153
+ 145 train 9.968241 (lr=5.1049e-05) (hash(x)=25044834)
154
+ 146 train 10.031214 (lr=5.1399e-05) (hash(x)=23712023)
155
+ 147 train 10.147308 (lr=5.1748e-05) (hash(x)=32982615)
156
+ 148 train 10.027087 (lr=5.2098e-05) (hash(x)=30113660)
157
+ 149 train 9.871798 (lr=5.2448e-05) (hash(x)=20970960)
158
+ 150 val loss 9.9006
159
+ 150 val perplexity 19943.0254
160
+ 150 train 9.873482 (lr=5.2797e-05) (hash(x)=23132684)
161
+ 151 train 10.466467 (lr=5.3147e-05) (hash(x)=35279941)
162
+ 152 train 10.178487 (lr=5.3497e-05) (hash(x)=31227444)
163
+ 153 train 9.845192 (lr=5.3846e-05) (hash(x)=25529472)
164
+ 154 train 9.846618 (lr=5.4196e-05) (hash(x)=24350409)
165
+ 155 train 10.059424 (lr=5.4545e-05) (hash(x)=26400041)
166
+ 156 train 9.843828 (lr=5.4895e-05) (hash(x)=25262621)
167
+ 157 train 9.832858 (lr=5.5245e-05) (hash(x)=24656138)
168
+ 158 train 9.858442 (lr=5.5594e-05) (hash(x)=26803414)
169
+ 159 train 9.818269 (lr=5.5944e-05) (hash(x)=25015923)
170
+ 160 train 9.722587 (lr=5.6294e-05) (hash(x)=23581172)
171
+ 161 train 9.787653 (lr=5.6643e-05) (hash(x)=22924885)
172
+ 162 train 9.738898 (lr=5.6993e-05) (hash(x)=23414296)
173
+ 163 train 9.952288 (lr=5.7343e-05) (hash(x)=24853586)
174
+ 164 train 9.950456 (lr=5.7692e-05) (hash(x)=25000130)
175
+ 165 train 9.819488 (lr=5.8042e-05) (hash(x)=27004780)
176
+ 166 train 9.778687 (lr=5.8392e-05) (hash(x)=26148573)
177
+ 167 train 9.749748 (lr=5.8741e-05) (hash(x)=26740855)
178
+ 168 train 9.554173 (lr=5.9091e-05) (hash(x)=20965419)
179
+ 169 train 9.678412 (lr=5.9441e-05) (hash(x)=23950114)
180
+ 170 train 9.679404 (lr=5.9790e-05) (hash(x)=24951982)
181
+ 171 train 9.746234 (lr=6.0140e-05) (hash(x)=24584116)
182
+ 172 train 9.834239 (lr=6.0490e-05) (hash(x)=24378759)
183
+ 173 train 9.671216 (lr=6.0839e-05) (hash(x)=25718516)
184
+ 174 train 9.746860 (lr=6.1189e-05) (hash(x)=28424396)
185
+ 175 train 9.654621 (lr=6.1538e-05) (hash(x)=22262151)
186
+ 176 train 9.804202 (lr=6.1888e-05) (hash(x)=26438412)
187
+ 177 train 9.667786 (lr=6.2238e-05) (hash(x)=23025303)
188
+ 178 train 9.562965 (lr=6.2587e-05) (hash(x)=24190770)
189
+ 179 train 9.627902 (lr=6.2937e-05) (hash(x)=26627860)
190
+ 180 train 9.670962 (lr=6.3287e-05) (hash(x)=23663439)
191
+ 181 train 9.473599 (lr=6.3636e-05) (hash(x)=23700532)
192
+ 182 train 9.542630 (lr=6.3986e-05) (hash(x)=23075676)
193
+ 183 train 9.576221 (lr=6.4336e-05) (hash(x)=26621834)
194
+ 184 train 9.574515 (lr=6.4685e-05) (hash(x)=29426269)
195
+ 185 train 9.568290 (lr=6.5035e-05) (hash(x)=21821465)
196
+ 186 train 9.596854 (lr=6.5385e-05) (hash(x)=26506130)
197
+ 187 train 9.712651 (lr=6.5734e-05) (hash(x)=26930630)
198
+ 188 train 9.571115 (lr=6.6084e-05) (hash(x)=22993793)
199
+ 189 train 9.576937 (lr=6.6434e-05) (hash(x)=19557946)
200
+ 190 train 9.678843 (lr=6.6783e-05) (hash(x)=23572891)
201
+ 191 train 9.477477 (lr=6.7133e-05) (hash(x)=23234741)
202
+ 192 train 9.579402 (lr=6.7483e-05) (hash(x)=25547951)
203
+ 193 train 9.576865 (lr=6.7832e-05) (hash(x)=26713563)
204
+ 194 train 9.561234 (lr=6.8182e-05) (hash(x)=25913622)
205
+ 195 train 9.616476 (lr=6.8531e-05) (hash(x)=28102443)
206
+ 196 train 9.552672 (lr=6.8881e-05) (hash(x)=23093351)
207
+ 197 train 9.831481 (lr=6.9231e-05) (hash(x)=31689122)
208
+ 198 train 9.769132 (lr=6.9580e-05) (hash(x)=32252517)
209
+ 199 train 9.546087 (lr=6.9930e-05) (hash(x)=25470563)
210
+ 200 val loss 9.5351
211
+ 200 val perplexity 13836.7422
212
+ 200 train 9.618517 (lr=7.0280e-05) (hash(x)=25597614)
213
+ 201 train 9.605309 (lr=7.0629e-05) (hash(x)=23757479)
214
+ 202 train 9.427135 (lr=7.0979e-05) (hash(x)=25140048)
215
+ 203 train 9.525558 (lr=7.1329e-05) (hash(x)=28282861)
216
+ 204 train 9.444001 (lr=7.1678e-05) (hash(x)=24754885)
217
+ 205 train 9.600517 (lr=7.2028e-05) (hash(x)=26731964)
218
+ 206 train 9.637472 (lr=7.2378e-05) (hash(x)=26660561)
219
+ 207 train 9.476781 (lr=7.2727e-05) (hash(x)=21799102)
220
+ 208 train 9.475918 (lr=7.3077e-05) (hash(x)=27025986)
221
+ 209 train 9.398620 (lr=7.3427e-05) (hash(x)=24672077)
222
+ 210 train 9.441489 (lr=7.3776e-05) (hash(x)=25322984)
223
+ 211 train 9.399508 (lr=7.4126e-05) (hash(x)=23471769)
224
+ 212 train 9.173469 (lr=7.4476e-05) (hash(x)=20766491)
225
+ 213 train 9.673152 (lr=7.4825e-05) (hash(x)=24058931)
226
+ 214 train 9.374883 (lr=7.5175e-05) (hash(x)=23872843)
227
+ 215 train 9.390122 (lr=7.5524e-05) (hash(x)=23275479)
228
+ 216 train 9.504784 (lr=7.5874e-05) (hash(x)=24914695)
229
+ 217 train 9.405884 (lr=7.6224e-05) (hash(x)=24364396)
230
+ 218 train 9.381534 (lr=7.6573e-05) (hash(x)=27986474)
231
+ 219 train 9.468410 (lr=7.6923e-05) (hash(x)=24473581)
232
+ 220 train 9.401658 (lr=7.7273e-05) (hash(x)=22974689)
233
+ 221 train 9.431918 (lr=7.7622e-05) (hash(x)=23774644)
234
+ 222 train 9.477564 (lr=7.7972e-05) (hash(x)=23245327)
235
+ 223 train 9.316342 (lr=7.8322e-05) (hash(x)=22091862)
236
+ 224 train 9.241957 (lr=7.8671e-05) (hash(x)=24362839)
237
+ 225 train 9.181778 (lr=7.9021e-05) (hash(x)=25482303)
238
+ 226 train 9.397807 (lr=7.9371e-05) (hash(x)=24911853)
239
+ 227 train 9.349014 (lr=7.9720e-05) (hash(x)=26018202)
240
+ 228 train 9.258036 (lr=8.0070e-05) (hash(x)=26124495)
241
+ 229 train 9.324333 (lr=8.0420e-05) (hash(x)=24560096)
242
+ 230 train 9.273483 (lr=8.0769e-05) (hash(x)=24695331)
243
+ 231 train 9.225048 (lr=8.1119e-05) (hash(x)=17430373)
244
+ 232 train 9.184372 (lr=8.1469e-05) (hash(x)=21813345)
245
+ 233 train 8.911993 (lr=8.1818e-05) (hash(x)=20098681)
246
+ 234 train 9.054662 (lr=8.2168e-05) (hash(x)=25095928)
247
+ 235 train 9.157641 (lr=8.2517e-05) (hash(x)=24078083)
248
+ 236 train 9.229322 (lr=8.2867e-05) (hash(x)=22901505)
249
+ 237 train 9.181264 (lr=8.3217e-05) (hash(x)=26595592)
250
+ 238 train 9.166348 (lr=8.3566e-05) (hash(x)=27663196)
251
+ 239 train 9.078361 (lr=8.3916e-05) (hash(x)=22954861)
252
+ 240 train 9.248569 (lr=8.4266e-05) (hash(x)=30159234)
253
+ 241 train 9.207042 (lr=8.4615e-05) (hash(x)=29294271)
254
+ 242 train 9.203348 (lr=8.4965e-05) (hash(x)=23728322)
255
+ 243 train 9.217865 (lr=8.5315e-05) (hash(x)=28695016)
256
+ 244 train 9.260241 (lr=8.5664e-05) (hash(x)=26702728)
257
+ 245 train 9.089677 (lr=8.6014e-05) (hash(x)=27100115)
258
+ 246 train 9.083124 (lr=8.6364e-05) (hash(x)=25082752)
259
+ 247 train 9.344508 (lr=8.6713e-05) (hash(x)=26671799)
260
+ 248 train 9.203792 (lr=8.7063e-05) (hash(x)=23718946)
261
+ 249 train 9.210631 (lr=8.7413e-05) (hash(x)=28137394)
262
+ 250 val loss 9.0981
263
+ 250 val perplexity 8938.1631
264
+ 250 train 9.199101 (lr=8.7762e-05) (hash(x)=23893495)
265
+ 251 train 9.124572 (lr=8.8112e-05) (hash(x)=23166092)
266
+ 252 train 9.006707 (lr=8.8462e-05) (hash(x)=25907665)
267
+ 253 train 8.958957 (lr=8.8811e-05) (hash(x)=25885986)
268
+ 254 train 9.030492 (lr=8.9161e-05) (hash(x)=25262712)
269
+ 255 train 8.991970 (lr=8.9510e-05) (hash(x)=26924723)
270
+ 256 train 9.134742 (lr=8.9860e-05) (hash(x)=28744736)
271
+ 257 train 9.001348 (lr=9.0210e-05) (hash(x)=26140590)
272
+ 258 train 9.111969 (lr=9.0559e-05) (hash(x)=25780449)
273
+ 259 train 8.993054 (lr=9.0909e-05) (hash(x)=25138659)
274
+ 260 train 9.091582 (lr=9.1259e-05) (hash(x)=27244046)
275
+ 261 train 9.057082 (lr=9.1608e-05) (hash(x)=27224685)
276
+ 262 train 9.108034 (lr=9.1958e-05) (hash(x)=28274477)
277
+ 263 train 8.890762 (lr=9.2308e-05) (hash(x)=23557495)
278
+ 264 train 8.957961 (lr=9.2657e-05) (hash(x)=24680596)
279
+ 265 train 8.879679 (lr=9.3007e-05) (hash(x)=23928957)
280
+ 266 train 8.808733 (lr=9.3357e-05) (hash(x)=23761390)
281
+ 267 train 8.893611 (lr=9.3706e-05) (hash(x)=25288123)
282
+ 268 train 8.946423 (lr=9.4056e-05) (hash(x)=28705502)
283
+ 269 train 8.662457 (lr=9.4406e-05) (hash(x)=23246294)
284
+ 270 train 8.633026 (lr=9.4755e-05) (hash(x)=28639079)
285
+ 271 train 8.576091 (lr=9.5105e-05) (hash(x)=27804380)
286
+ 272 train 8.795760 (lr=9.5455e-05) (hash(x)=24172235)
287
+ 273 train 8.858045 (lr=9.5804e-05) (hash(x)=23089140)
288
+ 274 train 9.018496 (lr=9.6154e-05) (hash(x)=27163701)
289
+ 275 train 9.041874 (lr=9.6503e-05) (hash(x)=26993263)
290
+ 276 train 8.812342 (lr=9.6853e-05) (hash(x)=28224233)
291
+ 277 train 9.093001 (lr=9.7203e-05) (hash(x)=27397203)
292
+ 278 train 8.876428 (lr=9.7552e-05) (hash(x)=27923882)
293
+ 279 train 8.823366 (lr=9.7902e-05) (hash(x)=26654908)
294
+ 280 train 8.833075 (lr=9.8252e-05) (hash(x)=24213147)
295
+ 281 train 8.613605 (lr=9.8601e-05) (hash(x)=21965022)
296
+ 282 train 8.658870 (lr=9.8951e-05) (hash(x)=25465685)
297
+ 283 train 8.760108 (lr=9.9301e-05) (hash(x)=27347722)
298
+ 284 train 8.754081 (lr=9.9650e-05) (hash(x)=26732050)
299
+ 285 train 8.840265 (lr=1.0000e-04) (hash(x)=28314127)
300
+ 286 train 8.686188 (lr=1.0000e-04) (hash(x)=21471186)
301
+ 287 train 8.637486 (lr=1.0000e-04) (hash(x)=23627518)
302
+ 288 train 8.612662 (lr=9.9998e-05) (hash(x)=20870353)
303
+ 289 train 8.571550 (lr=9.9996e-05) (hash(x)=25024764)
304
+ 290 train 8.409103 (lr=9.9993e-05) (hash(x)=20683822)
305
+ 291 train 8.684848 (lr=9.9989e-05) (hash(x)=21768671)
306
+ 292 train 8.598800 (lr=9.9984e-05) (hash(x)=25557309)
307
+ 293 train 8.766100 (lr=9.9979e-05) (hash(x)=25076667)
308
+ 294 train 8.475862 (lr=9.9972e-05) (hash(x)=23765822)
309
+ 295 train 8.502644 (lr=9.9965e-05) (hash(x)=21889990)
310
+ 296 train 8.543883 (lr=9.9956e-05) (hash(x)=26339893)
311
+ 297 train 8.366486 (lr=9.9947e-05) (hash(x)=20932794)
312
+ 298 train 8.450080 (lr=9.9937e-05) (hash(x)=21750070)
313
+ 299 train 8.592353 (lr=9.9926e-05) (hash(x)=23665838)
314
+ 300 val loss 8.5157
315
+ 300 val perplexity 4992.5649
316
+ 300 train 9.151396 (lr=9.9915e-05) (hash(x)=32888061)
317
+ 301 train 8.950865 (lr=9.9902e-05) (hash(x)=30223582)
318
+ 302 train 8.676606 (lr=9.9889e-05) (hash(x)=26908418)
319
+ 303 train 8.232124 (lr=9.9874e-05) (hash(x)=22528001)
320
+ 304 train 8.632669 (lr=9.9859e-05) (hash(x)=27452187)
321
+ 305 train 8.514595 (lr=9.9843e-05) (hash(x)=25181641)
322
+ 306 train 8.510798 (lr=9.9826e-05) (hash(x)=25546593)
323
+ 307 train 8.406247 (lr=9.9808e-05) (hash(x)=22487328)
324
+ 308 train 8.575992 (lr=9.9789e-05) (hash(x)=27804274)
325
+ 309 train 8.625447 (lr=9.9770e-05) (hash(x)=26544630)
326
+ 310 train 8.757365 (lr=9.9749e-05) (hash(x)=27738934)
327
+ 311 train 8.524184 (lr=9.9728e-05) (hash(x)=29248942)
328
+ 312 train 8.379094 (lr=9.9706e-05) (hash(x)=25103452)
329
+ 313 train 8.572188 (lr=9.9683e-05) (hash(x)=25052066)
330
+ 314 train 8.466971 (lr=9.9659e-05) (hash(x)=24481302)
331
+ 315 train 8.372562 (lr=9.9634e-05) (hash(x)=23543273)
332
+ 316 train 8.368678 (lr=9.9609e-05) (hash(x)=25608244)
333
+ 317 train 8.534225 (lr=9.9582e-05) (hash(x)=27451288)
334
+ 318 train 8.209323 (lr=9.9555e-05) (hash(x)=22806491)
335
+ 319 train 8.381476 (lr=9.9526e-05) (hash(x)=25533417)
336
+ 320 train 8.347603 (lr=9.9497e-05) (hash(x)=24557997)
337
+ 321 train 8.357434 (lr=9.9467e-05) (hash(x)=24432899)
338
+ 322 train 8.460213 (lr=9.9437e-05) (hash(x)=27583287)
339
+ 323 train 8.298083 (lr=9.9405e-05) (hash(x)=25552036)
340
+ 324 train 8.185834 (lr=9.9372e-05) (hash(x)=24201868)
341
+ 325 train 8.490564 (lr=9.9339e-05) (hash(x)=28149782)
342
+ 326 train 8.326546 (lr=9.9305e-05) (hash(x)=25529698)
343
+ 327 train 8.124458 (lr=9.9270e-05) (hash(x)=20612533)
344
+ 328 train 8.161023 (lr=9.9234e-05) (hash(x)=20699000)
345
+ 329 train 8.095327 (lr=9.9197e-05) (hash(x)=19774173)
346
+ 330 train 8.179114 (lr=9.9159e-05) (hash(x)=21681646)
347
+ 331 train 8.048678 (lr=9.9121e-05) (hash(x)=20216795)
348
+ 332 train 8.509418 (lr=9.9081e-05) (hash(x)=27697998)
349
+ 333 train 8.360950 (lr=9.9041e-05) (hash(x)=25896435)
350
+ 334 train 8.244636 (lr=9.9000e-05) (hash(x)=21585310)
351
+ 335 train 8.253590 (lr=9.8958e-05) (hash(x)=24677740)
352
+ 336 train 8.240166 (lr=9.8915e-05) (hash(x)=22027900)
353
+ 337 train 8.173353 (lr=9.8872e-05) (hash(x)=21835643)
354
+ 338 train 8.081265 (lr=9.8827e-05) (hash(x)=20558462)
355
+ 339 train 8.106919 (lr=9.8782e-05) (hash(x)=16722715)
356
+ 340 train 8.177634 (lr=9.8736e-05) (hash(x)=20972655)
357
+ 341 train 8.441692 (lr=9.8689e-05) (hash(x)=26303975)
358
+ 342 train 8.206023 (lr=9.8641e-05) (hash(x)=22938170)
359
+ 343 train 8.219611 (lr=9.8592e-05) (hash(x)=25347203)
360
+ 344 train 8.271708 (lr=9.8543e-05) (hash(x)=27398686)
361
+ 345 train 8.202487 (lr=9.8492e-05) (hash(x)=25973417)
362
+ 346 train 8.213170 (lr=9.8441e-05) (hash(x)=26918389)
363
+ 347 train 8.165987 (lr=9.8389e-05) (hash(x)=24272489)
364
+ 348 train 8.128397 (lr=9.8336e-05) (hash(x)=25593714)
365
+ 349 train 8.190018 (lr=9.8282e-05) (hash(x)=29260846)
366
+ 350 val loss 8.1857
367
+ 350 val perplexity 3589.1199
368
+ 350 train 8.295154 (lr=9.8228e-05) (hash(x)=27951602)
369
+ 351 train 8.371558 (lr=9.8172e-05) (hash(x)=28922363)
370
+ 352 train 8.259330 (lr=9.8116e-05) (hash(x)=27210734)
371
+ 353 train 8.243551 (lr=9.8059e-05) (hash(x)=26322572)
372
+ 354 train 8.207410 (lr=9.8001e-05) (hash(x)=27084665)
373
+ 355 train 8.143613 (lr=9.7942e-05) (hash(x)=25510798)
374
+ 356 train 8.216857 (lr=9.7882e-05) (hash(x)=24970921)
375
+ 357 train 8.301144 (lr=9.7822e-05) (hash(x)=24138948)
376
+ 358 train 8.189494 (lr=9.7761e-05) (hash(x)=24790211)
377
+ 359 train 8.157221 (lr=9.7699e-05) (hash(x)=25631397)
378
+ 360 train 8.086594 (lr=9.7636e-05) (hash(x)=23226625)
379
+ 361 train 8.072731 (lr=9.7572e-05) (hash(x)=24001903)
380
+ 362 train 8.117575 (lr=9.7507e-05) (hash(x)=24587948)
381
+ 363 train 7.867568 (lr=9.7442e-05) (hash(x)=21333676)
382
+ 364 train 7.776052 (lr=9.7376e-05) (hash(x)=23673779)
383
+ 365 train 8.871776 (lr=9.7309e-05) (hash(x)=30770484)
384
+ 366 train 8.208301 (lr=9.7241e-05) (hash(x)=26564899)
385
+ 367 train 8.182689 (lr=9.7172e-05) (hash(x)=26237983)
386
+ 368 train 7.952806 (lr=9.7103e-05) (hash(x)=23764356)
387
+ 369 train 8.220566 (lr=9.7032e-05) (hash(x)=26205744)
388
+ 370 train 8.058969 (lr=9.6961e-05) (hash(x)=19208770)
389
+ 371 train 8.147968 (lr=9.6889e-05) (hash(x)=25976502)
390
+ 372 train 7.978828 (lr=9.6817e-05) (hash(x)=23983933)
391
+ 373 train 8.032316 (lr=9.6743e-05) (hash(x)=24080636)
392
+ 374 train 8.158792 (lr=9.6669e-05) (hash(x)=24404047)
393
+ 375 train 8.126396 (lr=9.6593e-05) (hash(x)=24742645)
394
+ 376 train 8.102300 (lr=9.6518e-05) (hash(x)=24159600)
395
+ 377 train 8.285780 (lr=9.6441e-05) (hash(x)=28677257)
396
+ 378 train 8.120142 (lr=9.6363e-05) (hash(x)=25604111)
397
+ 379 train 8.015782 (lr=9.6285e-05) (hash(x)=27086333)
398
+ 380 train 8.021477 (lr=9.6206e-05) (hash(x)=25188207)
399
+ 381 train 8.152806 (lr=9.6126e-05) (hash(x)=27855233)
400
+ 382 train 7.846625 (lr=9.6045e-05) (hash(x)=19470039)
401
+ 383 train 8.083662 (lr=9.5963e-05) (hash(x)=26157660)
402
+ 384 train 8.032929 (lr=9.5881e-05) (hash(x)=25291570)
403
+ 385 train 8.013102 (lr=9.5798e-05) (hash(x)=25046062)
404
+ 386 train 8.249227 (lr=9.5714e-05) (hash(x)=27020337)
405
+ 387 train 7.891252 (lr=9.5629e-05) (hash(x)=23616370)
406
+ 388 train 7.638515 (lr=9.5544e-05) (hash(x)=19113218)
407
+ 389 train 7.959029 (lr=9.5457e-05) (hash(x)=24302232)
408
+ 390 train 7.958316 (lr=9.5370e-05) (hash(x)=22188949)
409
+ 391 train 7.765428 (lr=9.5282e-05) (hash(x)=22582169)
410
+ 392 train 8.031525 (lr=9.5194e-05) (hash(x)=24700570)
411
+ 393 train 7.859241 (lr=9.5104e-05) (hash(x)=22773833)
412
+ 394 train 7.769252 (lr=9.5014e-05) (hash(x)=21875928)
413
+ 395 train 7.980147 (lr=9.4923e-05) (hash(x)=26233189)
414
+ 396 train 7.845827 (lr=9.4831e-05) (hash(x)=24321467)
415
+ 397 train 8.034213 (lr=9.4739e-05) (hash(x)=26431507)
416
+ 398 train 8.146808 (lr=9.4646e-05) (hash(x)=28690877)
417
+ 399 train 8.107545 (lr=9.4551e-05) (hash(x)=26431960)
418
+ 400 val loss 8.0014
419
+ 400 val perplexity 2985.1116
420
+ 400 train 7.976574 (lr=9.4457e-05) (hash(x)=24580300)
421
+ 401 train 7.956069 (lr=9.4361e-05) (hash(x)=25112360)
422
+ 402 train 8.010159 (lr=9.4265e-05) (hash(x)=27597243)
423
+ 403 train 8.426721 (lr=9.4168e-05) (hash(x)=30707498)
424
+ 404 train 8.048132 (lr=9.4070e-05) (hash(x)=28485465)
425
+ 405 train 7.812109 (lr=9.3971e-05) (hash(x)=22586447)
426
+ 406 train 7.980748 (lr=9.3872e-05) (hash(x)=23175270)
427
+ 407 train 8.004735 (lr=9.3772e-05) (hash(x)=25716176)
428
+ 408 train 8.085403 (lr=9.3671e-05) (hash(x)=26861373)
429
+ 409 train 8.213449 (lr=9.3569e-05) (hash(x)=25118971)
430
+ 410 train 7.745518 (lr=9.3467e-05) (hash(x)=19829066)
431
+ 411 train 7.852029 (lr=9.3364e-05) (hash(x)=26256420)
432
+ 412 train 8.073081 (lr=9.3260e-05) (hash(x)=27796153)
433
+ 413 train 7.850477 (lr=9.3155e-05) (hash(x)=22633318)
434
+ 414 train 7.894708 (lr=9.3050e-05) (hash(x)=22589383)
435
+ 415 train 8.255014 (lr=9.2944e-05) (hash(x)=28019788)
436
+ 416 train 8.198844 (lr=9.2837e-05) (hash(x)=28970440)
437
+ 417 train 7.862769 (lr=9.2729e-05) (hash(x)=27396089)
438
+ 418 train 7.904654 (lr=9.2621e-05) (hash(x)=21183513)
439
+ 419 train 7.919446 (lr=9.2512e-05) (hash(x)=23510110)
440
+ 420 train 8.108882 (lr=9.2402e-05) (hash(x)=28833467)
441
+ 421 train 7.986089 (lr=9.2292e-05) (hash(x)=23646926)
442
+ 422 train 7.939774 (lr=9.2181e-05) (hash(x)=24697272)
443
+ 423 train 7.742516 (lr=9.2069e-05) (hash(x)=20382963)
444
+ 424 train 7.781633 (lr=9.1956e-05) (hash(x)=23467595)
445
+ 425 train 7.903539 (lr=9.1843e-05) (hash(x)=24304768)
446
+ 426 train 7.814140 (lr=9.1729e-05) (hash(x)=21392328)
447
+ 427 train 7.887437 (lr=9.1614e-05) (hash(x)=25339466)
448
+ 428 train 7.776502 (lr=9.1499e-05) (hash(x)=22092542)
449
+ 429 train 7.904911 (lr=9.1382e-05) (hash(x)=22088696)
450
+ 430 train 7.834120 (lr=9.1266e-05) (hash(x)=22184471)
451
+ 431 train 7.928495 (lr=9.1148e-05) (hash(x)=24489647)
452
+ 432 train 7.930354 (lr=9.1030e-05) (hash(x)=26794132)
453
+ 433 train 7.746942 (lr=9.0911e-05) (hash(x)=22940357)
454
+ 434 train 7.941865 (lr=9.0791e-05) (hash(x)=23719522)
455
+ 435 train 7.764498 (lr=9.0671e-05) (hash(x)=22927699)
456
+ 436 train 7.907698 (lr=9.0550e-05) (hash(x)=26068576)
457
+ 437 train 8.046297 (lr=9.0428e-05) (hash(x)=27631132)
458
+ 438 train 7.906267 (lr=9.0306e-05) (hash(x)=26739991)
459
+ 439 train 7.977521 (lr=9.0182e-05) (hash(x)=25128502)
460
+ 440 train 8.069044 (lr=9.0059e-05) (hash(x)=25657260)
461
+ 441 train 7.836212 (lr=8.9934e-05) (hash(x)=23576982)
462
+ 442 train 7.990134 (lr=8.9809e-05) (hash(x)=27117886)
463
+ 443 train 8.010843 (lr=8.9683e-05) (hash(x)=25808969)
464
+ 444 train 7.935843 (lr=8.9557e-05) (hash(x)=24738238)
465
+ 445 train 7.883171 (lr=8.9430e-05) (hash(x)=23429962)
466
+ 446 train 7.979735 (lr=8.9302e-05) (hash(x)=25075165)
467
+ 447 train 7.834359 (lr=8.9173e-05) (hash(x)=25231390)
468
+ 448 train 7.664243 (lr=8.9044e-05) (hash(x)=22055054)
469
+ 449 train 7.916121 (lr=8.8914e-05) (hash(x)=25395441)
470
+ 450 val loss 7.8922
471
+ 450 val perplexity 2676.2678
472
+ 450 train 7.911904 (lr=8.8784e-05) (hash(x)=25863209)
473
+ 451 train 7.621748 (lr=8.8653e-05) (hash(x)=21154388)
474
+ 452 train 7.628849 (lr=8.8521e-05) (hash(x)=21600876)
475
+ 453 train 7.762184 (lr=8.8388e-05) (hash(x)=24278611)
476
+ 454 train 7.666582 (lr=8.8255e-05) (hash(x)=23221720)
477
+ 455 train 7.840531 (lr=8.8122e-05) (hash(x)=22708977)
478
+ 456 train 8.421048 (lr=8.7987e-05) (hash(x)=23637758)
479
+ 457 train 8.081376 (lr=8.7852e-05) (hash(x)=28228490)
480
+ 458 train 8.107905 (lr=8.7717e-05) (hash(x)=28638071)
481
+ 459 train 7.977960 (lr=8.7580e-05) (hash(x)=27258353)
482
+ 460 train 7.930032 (lr=8.7444e-05) (hash(x)=26604728)
483
+ 461 train 7.842766 (lr=8.7306e-05) (hash(x)=23252199)
484
+ 462 train 7.946175 (lr=8.7168e-05) (hash(x)=26441427)
485
+ 463 train 7.723299 (lr=8.7029e-05) (hash(x)=24364920)
486
+ 464 train 7.897938 (lr=8.6890e-05) (hash(x)=25623792)
487
+ 465 train 7.799504 (lr=8.6750e-05) (hash(x)=23283905)
488
+ 466 train 8.152890 (lr=8.6609e-05) (hash(x)=26025267)
489
+ 467 train 7.910948 (lr=8.6468e-05) (hash(x)=27243972)
490
+ 468 train 8.359487 (lr=8.6326e-05) (hash(x)=30449945)
491
+ 469 train 8.064601 (lr=8.6184e-05) (hash(x)=28113043)
492
+ 470 train 7.829891 (lr=8.6041e-05) (hash(x)=25182521)
493
+ 471 train 7.873804 (lr=8.5897e-05) (hash(x)=24932925)
494
+ 472 train 7.707780 (lr=8.5753e-05) (hash(x)=20353098)
495
+ 473 train 7.639574 (lr=8.5608e-05) (hash(x)=19001259)
496
+ 474 train 8.139350 (lr=8.5462e-05) (hash(x)=27585685)
497
+ 475 train 7.892091 (lr=8.5316e-05) (hash(x)=26371091)
498
+ 476 train 7.719181 (lr=8.5170e-05) (hash(x)=24891798)
499
+ 477 train 7.770285 (lr=8.5022e-05) (hash(x)=24258817)
500
+ 478 train 7.862567 (lr=8.4875e-05) (hash(x)=24330263)
501
+ 479 train 7.919698 (lr=8.4726e-05) (hash(x)=26913684)
502
+ 480 train 7.891245 (lr=8.4577e-05) (hash(x)=26338455)
503
+ 481 train 8.102972 (lr=8.4428e-05) (hash(x)=27753043)
504
+ 482 train 7.851596 (lr=8.4278e-05) (hash(x)=26123289)
505
+ 483 train 7.966156 (lr=8.4127e-05) (hash(x)=29239611)
506
+ 484 train 7.872824 (lr=8.3976e-05) (hash(x)=26553003)
507
+ 485 train 7.848386 (lr=8.3824e-05) (hash(x)=22984557)
508
+ 486 train 7.399796 (lr=8.3672e-05) (hash(x)=16947491)
509
+ 487 train 7.404505 (lr=8.3519e-05) (hash(x)=18017792)
510
+ 488 train 7.578960 (lr=8.3366e-05) (hash(x)=19918608)
511
+ 489 train 7.990076 (lr=8.3212e-05) (hash(x)=23374526)
512
+ 490 train 7.874756 (lr=8.3057e-05) (hash(x)=25009505)
513
+ 491 train 7.833429 (lr=8.2902e-05) (hash(x)=27574089)
514
+ 492 train 7.983060 (lr=8.2746e-05) (hash(x)=24122664)
515
+ 493 train 7.856584 (lr=8.2590e-05) (hash(x)=26154906)
516
+ 494 train 7.816597 (lr=8.2434e-05) (hash(x)=25192767)
517
+ 495 train 7.950326 (lr=8.2276e-05) (hash(x)=28613882)
518
+ 496 train 7.841900 (lr=8.2119e-05) (hash(x)=23547219)
519
+ 497 train 7.826818 (lr=8.1960e-05) (hash(x)=25272182)
520
+ 498 train 7.712712 (lr=8.1801e-05) (hash(x)=24992761)
521
+ 499 train 7.973860 (lr=8.1642e-05) (hash(x)=26981914)
522
+ 500 val loss 7.8347
523
+ 500 val perplexity 2526.7705
524
+ 500 train 7.746886 (lr=8.1482e-05) (hash(x)=22051933)
525
+ 501 train 7.952049 (lr=8.1322e-05) (hash(x)=24232348)
526
+ 502 train 7.950436 (lr=8.1161e-05) (hash(x)=23158331)
527
+ 503 train 7.760433 (lr=8.1000e-05) (hash(x)=22652243)
528
+ 504 train 7.795718 (lr=8.0838e-05) (hash(x)=23805602)
529
+ 505 train 8.063145 (lr=8.0676e-05) (hash(x)=25411991)
530
+ 506 train 7.392728 (lr=8.0513e-05) (hash(x)=18827215)
531
+ 507 train 7.116994 (lr=8.0349e-05) (hash(x)=15446025)
532
+ 508 train 7.447006 (lr=8.0186e-05) (hash(x)=20516263)
533
+ 509 train 7.945335 (lr=8.0021e-05) (hash(x)=27846176)
534
+ 510 train 7.676935 (lr=7.9856e-05) (hash(x)=23342449)
535
+ 511 train 7.896316 (lr=7.9691e-05) (hash(x)=27194521)
536
+ 512 train 7.720159 (lr=7.9525e-05) (hash(x)=23008284)
537
+ 513 train 8.125158 (lr=7.9359e-05) (hash(x)=29430001)
538
+ 514 train 7.509356 (lr=7.9192e-05) (hash(x)=22579319)
539
+ 515 train 7.565389 (lr=7.9025e-05) (hash(x)=25264518)
540
+ 516 train 7.874396 (lr=7.8858e-05) (hash(x)=25359075)
541
+ 517 train 7.846001 (lr=7.8689e-05) (hash(x)=25568956)
542
+ 518 train 8.119682 (lr=7.8521e-05) (hash(x)=32004108)
543
+ 519 train 7.787384 (lr=7.8352e-05) (hash(x)=24936836)
544
+ 520 train 7.921185 (lr=7.8182e-05) (hash(x)=27263338)
545
+ 521 train 7.964056 (lr=7.8012e-05) (hash(x)=27452099)
546
+ 522 train 7.851326 (lr=7.7842e-05) (hash(x)=25965406)
547
+ 523 train 8.052362 (lr=7.7671e-05) (hash(x)=28197282)
548
+ 524 train 7.747787 (lr=7.7500e-05) (hash(x)=22466209)
549
+ 525 train 7.817917 (lr=7.7328e-05) (hash(x)=22931889)
550
+ 526 train 7.955327 (lr=7.7156e-05) (hash(x)=26903920)
551
+ 527 train 7.838830 (lr=7.6984e-05) (hash(x)=24765578)
552
+ 528 train 7.932610 (lr=7.6811e-05) (hash(x)=27811359)
553
+ 529 train 7.712749 (lr=7.6637e-05) (hash(x)=25078649)
554
+ 530 train 7.894597 (lr=7.6463e-05) (hash(x)=25572416)
555
+ 531 train 7.973712 (lr=7.6289e-05) (hash(x)=27448185)
556
+ 532 train 8.037421 (lr=7.6115e-05) (hash(x)=25923719)
557
+ 533 train 7.832639 (lr=7.5940e-05) (hash(x)=24804856)
558
+ 534 train 7.968217 (lr=7.5764e-05) (hash(x)=23207829)
559
+ 535 train 7.869740 (lr=7.5588e-05) (hash(x)=23107416)
560
+ 536 train 7.734635 (lr=7.5412e-05) (hash(x)=26739531)
561
+ 537 train 7.724672 (lr=7.5235e-05) (hash(x)=24960796)
562
+ 538 train 7.868000 (lr=7.5058e-05) (hash(x)=24667802)
563
+ 539 train 7.966385 (lr=7.4881e-05) (hash(x)=26755138)
564
+ 540 train 7.826601 (lr=7.4703e-05) (hash(x)=25537132)
565
+ 541 train 7.748081 (lr=7.4525e-05) (hash(x)=24542526)
566
+ 542 train 7.565048 (lr=7.4346e-05) (hash(x)=21296355)
567
+ 543 train 7.997856 (lr=7.4167e-05) (hash(x)=29314255)
568
+ 544 train 7.807579 (lr=7.3988e-05) (hash(x)=26001799)
569
+ 545 train 7.868913 (lr=7.3808e-05) (hash(x)=27347755)
570
+ 546 train 7.838994 (lr=7.3628e-05) (hash(x)=25107798)
571
+ 547 train 7.660203 (lr=7.3448e-05) (hash(x)=22112669)
572
+ 548 train 7.637968 (lr=7.3267e-05) (hash(x)=21897967)
573
+ 549 train 7.745965 (lr=7.3086e-05) (hash(x)=25161929)
574
+ 550 val loss 7.7996
575
+ 550 val perplexity 2439.6309
576
+ 550 train 7.703557 (lr=7.2904e-05) (hash(x)=27465106)
577
+ 551 train 7.813867 (lr=7.2723e-05) (hash(x)=24013079)
578
+ 552 train 7.740500 (lr=7.2540e-05) (hash(x)=23142015)
579
+ 553 train 7.788625 (lr=7.2358e-05) (hash(x)=26768629)
580
+ 554 train 7.826801 (lr=7.2175e-05) (hash(x)=26393383)
581
+ 555 train 7.594063 (lr=7.1992e-05) (hash(x)=22537194)
582
+ 556 train 7.861438 (lr=7.1808e-05) (hash(x)=24046036)
583
+ 557 train 7.814086 (lr=7.1624e-05) (hash(x)=24974360)
584
+ 558 train 8.030440 (lr=7.1440e-05) (hash(x)=28379928)
585
+ 559 train 7.820210 (lr=7.1256e-05) (hash(x)=25322001)
586
+ 560 train 7.908100 (lr=7.1071e-05) (hash(x)=26622031)
587
+ 561 train 7.558341 (lr=7.0886e-05) (hash(x)=20562247)
588
+ 562 train 7.947896 (lr=7.0701e-05) (hash(x)=27381885)
589
+ 563 train 8.091944 (lr=7.0515e-05) (hash(x)=27028126)
590
+ 564 train 7.897391 (lr=7.0329e-05) (hash(x)=28882928)
591
+ 565 train 7.861022 (lr=7.0143e-05) (hash(x)=25666355)
592
+ 566 train 7.939337 (lr=6.9956e-05) (hash(x)=24330810)
593
+ 567 train 7.870960 (lr=6.9769e-05) (hash(x)=26690440)
594
+ 568 train 7.736110 (lr=6.9582e-05) (hash(x)=22923592)
595
+ 569 train 7.859322 (lr=6.9394e-05) (hash(x)=27348418)
596
+ 570 train 7.885109 (lr=6.9207e-05) (hash(x)=28849848)
597
+ 571 train 7.940651 (lr=6.9019e-05) (hash(x)=26967331)
598
+ 572 train 7.683392 (lr=6.8830e-05) (hash(x)=22831467)
599
+ 573 train 7.779733 (lr=6.8642e-05) (hash(x)=24765121)
600
+ 574 train 7.767337 (lr=6.8453e-05) (hash(x)=24331857)
601
+ 575 train 7.598742 (lr=6.8264e-05) (hash(x)=22598512)
602
+ 576 train 7.794003 (lr=6.8075e-05) (hash(x)=25149353)
603
+ 577 train 7.668240 (lr=6.7885e-05) (hash(x)=23725598)
604
+ 578 train 7.897913 (lr=6.7695e-05) (hash(x)=26449557)
605
+ 579 train 7.889452 (lr=6.7505e-05) (hash(x)=24697985)
606
+ 580 train 7.886296 (lr=6.7315e-05) (hash(x)=26923059)
607
+ 581 train 7.715932 (lr=6.7124e-05) (hash(x)=25201962)
608
+ 582 train 7.381770 (lr=6.6933e-05) (hash(x)=20931520)
609
+ 583 train 7.455095 (lr=6.6742e-05) (hash(x)=18473911)
610
+ 584 train 7.599690 (lr=6.6551e-05) (hash(x)=21306267)
611
+ 585 train 7.876476 (lr=6.6360e-05) (hash(x)=25982840)
612
+ 586 train 7.758805 (lr=6.6168e-05) (hash(x)=25364874)
613
+ 587 train 7.669289 (lr=6.5976e-05) (hash(x)=23172124)
614
+ 588 train 7.936721 (lr=6.5784e-05) (hash(x)=27876897)
615
+ 589 train 8.529624 (lr=6.5592e-05) (hash(x)=34646114)
616
+ 590 train 8.864356 (lr=6.5399e-05) (hash(x)=35153576)
617
+ 591 train 7.673059 (lr=6.5206e-05) (hash(x)=22322442)
618
+ 592 train 7.960739 (lr=6.5013e-05) (hash(x)=27907331)
619
+ 593 train 7.855909 (lr=6.4820e-05) (hash(x)=26211794)
620
+ 594 train 8.065513 (lr=6.4627e-05) (hash(x)=29291512)
621
+ 595 train 8.316224 (lr=6.4433e-05) (hash(x)=29659121)
622
+ 596 train 8.120142 (lr=6.4240e-05) (hash(x)=29674399)
623
+ 597 train 7.760258 (lr=6.4046e-05) (hash(x)=23538306)
624
+ 598 train 7.880569 (lr=6.3852e-05) (hash(x)=21991524)
625
+ 599 train 7.834962 (lr=6.3658e-05) (hash(x)=26324153)
626
+ 600 val loss 7.7977
627
+ 600 val perplexity 2435.0308
628
+ 600 train 7.797328 (lr=6.3463e-05) (hash(x)=23712082)
629
+ 601 train 7.822736 (lr=6.3269e-05) (hash(x)=24910403)
630
+ 602 train 7.792645 (lr=6.3074e-05) (hash(x)=26737205)
631
+ 603 train 7.931361 (lr=6.2879e-05) (hash(x)=26939970)
632
+ 604 train 7.797231 (lr=6.2684e-05) (hash(x)=27651943)
633
+ 605 train 8.092806 (lr=6.2489e-05) (hash(x)=27515446)
634
+ 606 train 7.978427 (lr=6.2294e-05) (hash(x)=26753129)
635
+ 607 train 7.692255 (lr=6.2098e-05) (hash(x)=23446058)
636
+ 608 train 7.942758 (lr=6.1903e-05) (hash(x)=27587849)
637
+ 609 train 7.870189 (lr=6.1707e-05) (hash(x)=25308253)
638
+ 610 train 7.896657 (lr=6.1511e-05) (hash(x)=26615098)
639
+ 611 train 7.917583 (lr=6.1315e-05) (hash(x)=29981801)
640
+ 612 train 8.006940 (lr=6.1119e-05) (hash(x)=29592345)
641
+ 613 train 8.021389 (lr=6.0923e-05) (hash(x)=23470413)
642
+ 614 train 7.915616 (lr=6.0726e-05) (hash(x)=24742370)
643
+ 615 train 7.905232 (lr=6.0530e-05) (hash(x)=24843741)
644
+ 616 train 7.815538 (lr=6.0333e-05) (hash(x)=25192548)
645
+ 617 train 7.904895 (lr=6.0137e-05) (hash(x)=27176996)
646
+ 618 train 7.761873 (lr=5.9940e-05) (hash(x)=23964552)
647
+ 619 train 7.738393 (lr=5.9743e-05) (hash(x)=22855363)
648
+ 620 train 7.699829 (lr=5.9546e-05) (hash(x)=26332996)
649
+ 621 train 7.549547 (lr=5.9349e-05) (hash(x)=22960957)
650
+ 622 train 7.865096 (lr=5.9152e-05) (hash(x)=22752597)
651
+ 623 train 7.763538 (lr=5.8955e-05) (hash(x)=23197102)
652
+ 624 train 8.231557 (lr=5.8758e-05) (hash(x)=27383319)
653
+ 625 train 7.831302 (lr=5.8560e-05) (hash(x)=26132276)
654
+ 626 train 7.677574 (lr=5.8363e-05) (hash(x)=19810497)
655
+ 627 train 7.784789 (lr=5.8165e-05) (hash(x)=25704919)
656
+ 628 train 7.923711 (lr=5.7968e-05) (hash(x)=27174264)
657
+ 629 train 8.030566 (lr=5.7770e-05) (hash(x)=22280814)
658
+ 630 train 8.742656 (lr=5.7573e-05) (hash(x)=24279448)
659
+ 631 train 8.203459 (lr=5.7375e-05) (hash(x)=23054940)
660
+ 632 train 8.310336 (lr=5.7177e-05) (hash(x)=23942400)
661
+ 633 train 8.586176 (lr=5.6979e-05) (hash(x)=24712416)
662
+ 634 train 8.168160 (lr=5.6782e-05) (hash(x)=24564658)
663
+ 635 train 7.836517 (lr=5.6584e-05) (hash(x)=24909904)
664
+ 636 train 7.908140 (lr=5.6386e-05) (hash(x)=23049534)
665
+ 637 train 7.928827 (lr=5.6188e-05) (hash(x)=24321591)
666
+ 638 train 7.706109 (lr=5.5990e-05) (hash(x)=26153298)
667
+ 639 train 7.784331 (lr=5.5792e-05) (hash(x)=27140757)
668
+ 640 train 7.570251 (lr=5.5594e-05) (hash(x)=25115907)
669
+ 641 train 7.655694 (lr=5.5396e-05) (hash(x)=26563770)
670
+ 642 train 7.829694 (lr=5.5198e-05) (hash(x)=28089252)
671
+ 643 train 7.732410 (lr=5.5000e-05) (hash(x)=25993110)
672
+ 644 train 7.719652 (lr=5.4802e-05) (hash(x)=25150008)
673
+ 645 train 7.939255 (lr=5.4604e-05) (hash(x)=28520222)
674
+ 646 train 7.591432 (lr=5.4406e-05) (hash(x)=21349943)
675
+ 647 train 7.670441 (lr=5.4208e-05) (hash(x)=25149419)
676
+ 648 train 7.877708 (lr=5.4010e-05) (hash(x)=25730641)
677
+ 649 train 7.989602 (lr=5.3812e-05) (hash(x)=26112813)
678
+ 650 val loss 7.7694
679
+ 650 val perplexity 2366.9810
680
+ 650 train 7.796289 (lr=5.3614e-05) (hash(x)=25907805)
681
+ 651 train 7.851594 (lr=5.3416e-05) (hash(x)=27623643)
682
+ 652 train 7.793159 (lr=5.3218e-05) (hash(x)=26484959)
683
+ 653 train 7.865926 (lr=5.3021e-05) (hash(x)=29199854)
684
+ 654 train 7.838643 (lr=5.2823e-05) (hash(x)=28369628)
685
+ 655 train 7.713974 (lr=5.2625e-05) (hash(x)=24727764)
686
+ 656 train 7.439061 (lr=5.2427e-05) (hash(x)=22610673)
687
+ 657 train 7.361540 (lr=5.2230e-05) (hash(x)=22667179)
688
+ 658 train 7.179814 (lr=5.2032e-05) (hash(x)=18477300)
689
+ 659 train 7.595999 (lr=5.1835e-05) (hash(x)=23155773)
690
+ 660 train 7.403195 (lr=5.1637e-05) (hash(x)=19461032)
691
+ 661 train 7.671582 (lr=5.1440e-05) (hash(x)=23453788)
692
+ 662 train 7.763507 (lr=5.1242e-05) (hash(x)=24543466)
693
+ 663 train 7.561336 (lr=5.1045e-05) (hash(x)=21935931)
694
+ 664 train 7.440926 (lr=5.0848e-05) (hash(x)=19910292)
695
+ 665 train 7.686597 (lr=5.0651e-05) (hash(x)=24481079)
696
+ 666 train 7.427635 (lr=5.0454e-05) (hash(x)=18922411)
697
+ 667 train 7.444908 (lr=5.0257e-05) (hash(x)=20054917)
698
+ 668 train 7.676207 (lr=5.0060e-05) (hash(x)=24850470)
699
+ 669 train 7.705040 (lr=4.9863e-05) (hash(x)=25907741)
700
+ 670 train 7.840762 (lr=4.9667e-05) (hash(x)=26873522)
701
+ 671 train 7.869147 (lr=4.9470e-05) (hash(x)=27606073)
702
+ 672 train 7.961496 (lr=4.9274e-05) (hash(x)=26209645)
703
+ 673 train 7.816560 (lr=4.9077e-05) (hash(x)=25202001)
704
+ 674 train 7.856000 (lr=4.8881e-05) (hash(x)=25569462)
705
+ 675 train 7.838399 (lr=4.8685e-05) (hash(x)=26534487)
706
+ 676 train 7.814643 (lr=4.8489e-05) (hash(x)=26455057)
707
+ 677 train 7.691425 (lr=4.8293e-05) (hash(x)=24095850)
708
+ 678 train 7.922995 (lr=4.8097e-05) (hash(x)=25287752)
709
+ 679 train 7.418669 (lr=4.7902e-05) (hash(x)=22450341)
710
+ 680 train 8.130902 (lr=4.7706e-05) (hash(x)=29004853)
711
+ 681 train 7.865321 (lr=4.7511e-05) (hash(x)=27993763)
712
+ 682 train 7.686388 (lr=4.7316e-05) (hash(x)=26382658)
713
+ 683 train 7.847646 (lr=4.7121e-05) (hash(x)=25013073)
714
+ 684 train 7.898720 (lr=4.6926e-05) (hash(x)=30595809)
715
+ 685 train 8.007148 (lr=4.6731e-05) (hash(x)=30934371)
716
+ 686 train 8.530044 (lr=4.6537e-05) (hash(x)=33060834)
717
+ 687 train 7.802766 (lr=4.6342e-05) (hash(x)=25945859)
718
+ 688 train 7.646629 (lr=4.6148e-05) (hash(x)=23375678)
719
+ 689 train 7.695192 (lr=4.5954e-05) (hash(x)=25218689)
720
+ 690 train 7.802976 (lr=4.5760e-05) (hash(x)=28127397)
721
+ 691 train 7.747813 (lr=4.5567e-05) (hash(x)=24418091)
722
+ 692 train 7.660182 (lr=4.5373e-05) (hash(x)=22761099)
723
+ 693 train 7.654142 (lr=4.5180e-05) (hash(x)=24615466)
724
+ 694 train 7.817244 (lr=4.4987e-05) (hash(x)=24699240)
725
+ 695 train 7.693007 (lr=4.4794e-05) (hash(x)=21613707)
726
+ 696 train 7.755492 (lr=4.4601e-05) (hash(x)=24977554)
727
+ 697 train 7.725454 (lr=4.4408e-05) (hash(x)=24348175)
728
+ 698 train 7.857873 (lr=4.4216e-05) (hash(x)=25102767)
729
+ 699 train 7.821219 (lr=4.4024e-05) (hash(x)=26386157)
730
+ 700 val loss 7.7386
731
+ 700 val perplexity 2295.3679
732
+ 700 train 7.773688 (lr=4.3832e-05) (hash(x)=26423460)
733
+ 701 train 7.719083 (lr=4.3640e-05) (hash(x)=25771047)
734
+ 702 train 7.762843 (lr=4.3449e-05) (hash(x)=26999875)
735
+ 703 train 7.680254 (lr=4.3258e-05) (hash(x)=24396519)
736
+ 704 train 7.723744 (lr=4.3067e-05) (hash(x)=22588122)
737
+ 705 train 8.010463 (lr=4.2876e-05) (hash(x)=25142399)
738
+ 706 train 7.767162 (lr=4.2685e-05) (hash(x)=20440214)
739
+ 707 train 7.619082 (lr=4.2495e-05) (hash(x)=23265507)
740
+ 708 train 7.690048 (lr=4.2305e-05) (hash(x)=24563470)
741
+ 709 train 7.619761 (lr=4.2115e-05) (hash(x)=22514858)
742
+ 710 train 7.967416 (lr=4.1925e-05) (hash(x)=26691212)
743
+ 711 train 8.012252 (lr=4.1736e-05) (hash(x)=29138828)
744
+ 712 train 7.986040 (lr=4.1547e-05) (hash(x)=28028528)
745
+ 713 train 7.768848 (lr=4.1358e-05) (hash(x)=20531210)
746
+ 714 train 7.694386 (lr=4.1170e-05) (hash(x)=25075352)
747
+ 715 train 7.780620 (lr=4.0981e-05) (hash(x)=24265353)
748
+ 716 train 7.875197 (lr=4.0793e-05) (hash(x)=24635726)
749
+ 717 train 7.729631 (lr=4.0606e-05) (hash(x)=24999726)
750
+ 718 train 7.910581 (lr=4.0418e-05) (hash(x)=27412910)
751
+ 719 train 7.713174 (lr=4.0231e-05) (hash(x)=24685515)
752
+ 720 train 7.776260 (lr=4.0044e-05) (hash(x)=23780329)
753
+ 721 train 7.590933 (lr=3.9857e-05) (hash(x)=25071701)
754
+ 722 train 7.747319 (lr=3.9671e-05) (hash(x)=23767130)
755
+ 723 train 7.699047 (lr=3.9485e-05) (hash(x)=24876269)
756
+ 724 train 7.840742 (lr=3.9299e-05) (hash(x)=26405773)
757
+ 725 train 8.203056 (lr=3.9114e-05) (hash(x)=31733180)
758
+ 726 train 7.470693 (lr=3.8929e-05) (hash(x)=21337509)
759
+ 727 train 7.560100 (lr=3.8744e-05) (hash(x)=22825749)
760
+ 728 train 7.909895 (lr=3.8560e-05) (hash(x)=28638695)
761
+ 729 train 7.796455 (lr=3.8376e-05) (hash(x)=26393943)
762
+ 730 train 7.966617 (lr=3.8192e-05) (hash(x)=27563583)
763
+ 731 train 7.510880 (lr=3.8008e-05) (hash(x)=21239652)
764
+ 732 train 7.648170 (lr=3.7825e-05) (hash(x)=23986428)
765
+ 733 train 7.532730 (lr=3.7642e-05) (hash(x)=24943881)
766
+ 734 train 8.057973 (lr=3.7460e-05) (hash(x)=29691448)
767
+ 735 train 7.991315 (lr=3.7277e-05) (hash(x)=28767869)
768
+ 736 train 7.733481 (lr=3.7096e-05) (hash(x)=23628188)
769
+ 737 train 8.038494 (lr=3.6914e-05) (hash(x)=29341482)
770
+ 738 train 7.843400 (lr=3.6733e-05) (hash(x)=30336570)
771
+ 739 train 7.780446 (lr=3.6552e-05) (hash(x)=25614301)
772
+ 740 train 7.692434 (lr=3.6372e-05) (hash(x)=24160500)
773
+ 741 train 7.880836 (lr=3.6192e-05) (hash(x)=26030058)
774
+ 742 train 8.228282 (lr=3.6012e-05) (hash(x)=29243936)
775
+ 743 train 7.530138 (lr=3.5833e-05) (hash(x)=21159060)
776
+ 744 train 7.632191 (lr=3.5654e-05) (hash(x)=23701853)
777
+ 745 train 7.807756 (lr=3.5475e-05) (hash(x)=24629937)
778
+ 746 train 7.709841 (lr=3.5297e-05) (hash(x)=25110108)
779
+ 747 train 7.778179 (lr=3.5119e-05) (hash(x)=26751788)
780
+ 748 train 7.820279 (lr=3.4942e-05) (hash(x)=26430427)
781
+ 749 train 7.655323 (lr=3.4765e-05) (hash(x)=26012353)
782
+ 750 val loss 7.7309
783
+ 750 val perplexity 2277.6953
784
+ 750 train 7.835443 (lr=3.4588e-05) (hash(x)=22735910)
785
+ 751 train 7.604541 (lr=3.4412e-05) (hash(x)=25045397)
786
+ 752 train 7.344478 (lr=3.4236e-05) (hash(x)=21554427)
787
+ 753 train 7.446912 (lr=3.4060e-05) (hash(x)=23751143)
788
+ 754 train 8.229322 (lr=3.3885e-05) (hash(x)=28602273)
789
+ 755 train 8.266418 (lr=3.3711e-05) (hash(x)=29989709)
790
+ 756 train 7.653561 (lr=3.3537e-05) (hash(x)=22331648)
791
+ 757 train 8.089641 (lr=3.3363e-05) (hash(x)=31017246)
792
+ 758 train 8.011683 (lr=3.3189e-05) (hash(x)=29709045)
793
+ 759 train 8.050717 (lr=3.3016e-05) (hash(x)=25560928)
794
+ 760 train 7.735901 (lr=3.2844e-05) (hash(x)=25075464)
795
+ 761 train 7.824506 (lr=3.2672e-05) (hash(x)=27352253)
796
+ 762 train 8.027542 (lr=3.2500e-05) (hash(x)=28187891)
797
+ 763 train 7.872363 (lr=3.2329e-05) (hash(x)=26062687)
798
+ 764 train 7.905521 (lr=3.2158e-05) (hash(x)=27427811)
799
+ 765 train 7.968201 (lr=3.1988e-05) (hash(x)=27614522)
800
+ 766 train 7.747192 (lr=3.1818e-05) (hash(x)=26129544)
801
+ 767 train 8.278567 (lr=3.1648e-05) (hash(x)=28959222)
802
+ 768 train 7.847680 (lr=3.1479e-05) (hash(x)=26860067)
803
+ 769 train 7.634090 (lr=3.1311e-05) (hash(x)=25122598)
804
+ 770 train 7.719129 (lr=3.1142e-05) (hash(x)=25245030)
805
+ 771 train 7.724206 (lr=3.0975e-05) (hash(x)=25434884)
806
+ 772 train 7.828176 (lr=3.0808e-05) (hash(x)=27732790)
807
+ 773 train 7.858775 (lr=3.0641e-05) (hash(x)=27824438)
808
+ 774 train 7.928463 (lr=3.0475e-05) (hash(x)=27201953)
809
+ 775 train 7.636010 (lr=3.0309e-05) (hash(x)=21248405)
810
+ 776 train 7.501904 (lr=3.0144e-05) (hash(x)=22805934)
811
+ 777 train 7.743581 (lr=2.9979e-05) (hash(x)=26482588)
812
+ 778 train 7.816232 (lr=2.9814e-05) (hash(x)=24153691)
813
+ 779 train 7.940343 (lr=2.9651e-05) (hash(x)=25044192)
814
+ 780 train 7.879991 (lr=2.9487e-05) (hash(x)=25910078)
815
+ 781 train 7.878275 (lr=2.9324e-05) (hash(x)=28645524)
816
+ 782 train 7.691223 (lr=2.9162e-05) (hash(x)=24368498)
817
+ 783 train 7.759683 (lr=2.9000e-05) (hash(x)=25830182)
818
+ 784 train 7.841648 (lr=2.8839e-05) (hash(x)=29181807)
819
+ 785 train 7.756452 (lr=2.8678e-05) (hash(x)=25585137)
820
+ 786 train 7.711185 (lr=2.8518e-05) (hash(x)=24798246)
821
+ 787 train 7.806797 (lr=2.8358e-05) (hash(x)=26621419)
822
+ 788 train 7.386952 (lr=2.8199e-05) (hash(x)=21446891)
823
+ 789 train 7.342375 (lr=2.8040e-05) (hash(x)=22165286)
824
+ 790 train 7.553163 (lr=2.7881e-05) (hash(x)=23477219)
825
+ 791 train 7.834488 (lr=2.7724e-05) (hash(x)=25173113)
826
+ 792 train 7.735875 (lr=2.7566e-05) (hash(x)=25853788)
827
+ 793 train 7.829154 (lr=2.7410e-05) (hash(x)=27267091)
828
+ 794 train 7.650575 (lr=2.7254e-05) (hash(x)=23743694)
829
+ 795 train 7.643759 (lr=2.7098e-05) (hash(x)=24400133)
830
+ 796 train 7.697875 (lr=2.6943e-05) (hash(x)=23663639)
831
+ 797 train 7.573288 (lr=2.6788e-05) (hash(x)=23103223)
832
+ 798 train 8.120043 (lr=2.6634e-05) (hash(x)=28748411)
833
+ 799 train 7.443572 (lr=2.6481e-05) (hash(x)=23486277)
834
+ 800 val loss 7.7178
835
+ 800 val perplexity 2247.8994
836
+ 800 train 7.692224 (lr=2.6328e-05) (hash(x)=25678518)
837
+ 801 train 7.636841 (lr=2.6176e-05) (hash(x)=23421286)
838
+ 802 train 7.807379 (lr=2.6024e-05) (hash(x)=26054104)
839
+ 803 train 7.818646 (lr=2.5873e-05) (hash(x)=25978130)
840
+ 804 train 7.831809 (lr=2.5722e-05) (hash(x)=26006525)
841
+ 805 train 7.767652 (lr=2.5572e-05) (hash(x)=25769432)
842
+ 806 train 7.568003 (lr=2.5423e-05) (hash(x)=22430795)
843
+ 807 train 7.994542 (lr=2.5274e-05) (hash(x)=28916006)
844
+ 808 train 7.729517 (lr=2.5125e-05) (hash(x)=25166800)
845
+ 809 train 7.843672 (lr=2.4978e-05) (hash(x)=24226056)
846
+ 810 train 7.699248 (lr=2.4830e-05) (hash(x)=25057934)
847
+ 811 train 8.461952 (lr=2.4684e-05) (hash(x)=33195100)
848
+ 812 train 7.884610 (lr=2.4538e-05) (hash(x)=26312888)
849
+ 813 train 8.051663 (lr=2.4392e-05) (hash(x)=27730410)
850
+ 814 train 7.910318 (lr=2.4247e-05) (hash(x)=27372474)
851
+ 815 train 7.786591 (lr=2.4103e-05) (hash(x)=25556929)
852
+ 816 train 7.821799 (lr=2.3959e-05) (hash(x)=26909985)
853
+ 817 train 7.644506 (lr=2.3816e-05) (hash(x)=25991247)
854
+ 818 train 7.850621 (lr=2.3674e-05) (hash(x)=27438141)
855
+ 819 train 8.030757 (lr=2.3532e-05) (hash(x)=29536986)
856
+ 820 train 7.737701 (lr=2.3391e-05) (hash(x)=24478391)
857
+ 821 train 7.733479 (lr=2.3250e-05) (hash(x)=26125216)
858
+ 822 train 7.774321 (lr=2.3110e-05) (hash(x)=26422130)
859
+ 823 train 7.981024 (lr=2.2971e-05) (hash(x)=29648798)
860
+ 824 train 7.627000 (lr=2.2832e-05) (hash(x)=21247770)
861
+ 825 train 7.761687 (lr=2.2694e-05) (hash(x)=23195388)
862
+ 826 train 7.716430 (lr=2.2556e-05) (hash(x)=25796725)
863
+ 827 train 7.757127 (lr=2.2420e-05) (hash(x)=23124767)
864
+ 828 train 7.802899 (lr=2.2283e-05) (hash(x)=25233464)
865
+ 829 train 7.779785 (lr=2.2148e-05) (hash(x)=25713275)
866
+ 830 train 7.724813 (lr=2.2013e-05) (hash(x)=25550167)
867
+ 831 train 7.627042 (lr=2.1878e-05) (hash(x)=24976217)
868
+ 832 train 7.934575 (lr=2.1745e-05) (hash(x)=28536827)
869
+ 833 train 7.820785 (lr=2.1612e-05) (hash(x)=27500801)
870
+ 834 train 7.631102 (lr=2.1479e-05) (hash(x)=25545765)
871
+ 835 train 7.673883 (lr=2.1347e-05) (hash(x)=23632825)
872
+ 836 train 7.720844 (lr=2.1216e-05) (hash(x)=25708009)
873
+ 837 train 7.765067 (lr=2.1086e-05) (hash(x)=24456276)
874
+ 838 train 7.867883 (lr=2.0956e-05) (hash(x)=29189855)
875
+ 839 train 8.104510 (lr=2.0827e-05) (hash(x)=31019606)
876
+ 840 train 7.951012 (lr=2.0698e-05) (hash(x)=26328013)
877
+ 841 train 7.875017 (lr=2.0570e-05) (hash(x)=25027904)
878
+ 842 train 7.664558 (lr=2.0443e-05) (hash(x)=23734189)
879
+ 843 train 7.955652 (lr=2.0317e-05) (hash(x)=28236580)
880
+ 844 train 7.725315 (lr=2.0191e-05) (hash(x)=26509780)
881
+ 845 train 7.784470 (lr=2.0066e-05) (hash(x)=25386473)
882
+ 846 train 7.757417 (lr=1.9941e-05) (hash(x)=24052671)
883
+ 847 train 7.945009 (lr=1.9818e-05) (hash(x)=28269421)
884
+ 848 train 7.363342 (lr=1.9694e-05) (hash(x)=22251724)
885
+ 849 train 7.619383 (lr=1.9572e-05) (hash(x)=24308447)
886
+ 850 val loss 7.7190
887
+ 850 val perplexity 2250.7041
888
+ 850 train 7.594337 (lr=1.9450e-05) (hash(x)=24242830)
889
+ 851 train 7.782412 (lr=1.9329e-05) (hash(x)=25563279)
890
+ 852 train 7.782153 (lr=1.9209e-05) (hash(x)=26354481)
891
+ 853 train 7.865166 (lr=1.9089e-05) (hash(x)=26152637)
892
+ 854 train 7.970394 (lr=1.8970e-05) (hash(x)=28051025)
893
+ 855 train 7.752617 (lr=1.8852e-05) (hash(x)=24865358)
894
+ 856 train 7.636621 (lr=1.8734e-05) (hash(x)=24288911)
895
+ 857 train 7.543935 (lr=1.8618e-05) (hash(x)=22230964)
896
+ 858 train 7.454135 (lr=1.8501e-05) (hash(x)=21303832)
897
+ 859 train 7.614290 (lr=1.8386e-05) (hash(x)=22155546)
898
+ 860 train 7.724576 (lr=1.8271e-05) (hash(x)=25296428)
899
+ 861 train 7.870574 (lr=1.8157e-05) (hash(x)=29142319)
900
+ 862 train 7.713500 (lr=1.8044e-05) (hash(x)=25545430)
901
+ 863 train 7.531435 (lr=1.7931e-05) (hash(x)=26984272)
902
+ 864 train 7.726352 (lr=1.7819e-05) (hash(x)=25429005)
903
+ 865 train 7.790149 (lr=1.7708e-05) (hash(x)=27077032)
904
+ 866 train 7.728251 (lr=1.7598e-05) (hash(x)=26494424)
905
+ 867 train 7.723320 (lr=1.7488e-05) (hash(x)=23193673)
906
+ 868 train 7.763917 (lr=1.7379e-05) (hash(x)=25075134)
907
+ 869 train 7.826239 (lr=1.7271e-05) (hash(x)=27112558)
908
+ 870 train 7.967346 (lr=1.7163e-05) (hash(x)=27436608)
909
+ 871 train 7.673750 (lr=1.7056e-05) (hash(x)=24544116)
910
+ 872 train 7.990724 (lr=1.6950e-05) (hash(x)=31632686)
911
+ 873 train 7.694386 (lr=1.6845e-05) (hash(x)=25890184)
912
+ 874 train 7.774828 (lr=1.6740e-05) (hash(x)=22887555)
913
+ 875 train 7.669539 (lr=1.6636e-05) (hash(x)=24547533)
914
+ 876 train 7.793178 (lr=1.6533e-05) (hash(x)=26553496)
915
+ 877 train 7.888308 (lr=1.6431e-05) (hash(x)=27467688)
916
+ 878 train 7.639868 (lr=1.6329e-05) (hash(x)=24766934)
917
+ 879 train 7.550095 (lr=1.6228e-05) (hash(x)=22059850)
918
+ 880 train 7.906858 (lr=1.6128e-05) (hash(x)=22871702)
919
+ 881 train 7.733748 (lr=1.6029e-05) (hash(x)=23893130)
920
+ 882 train 7.818442 (lr=1.5930e-05) (hash(x)=25125691)
921
+ 883 train 7.750296 (lr=1.5832e-05) (hash(x)=25994573)
922
+ 884 train 7.794983 (lr=1.5735e-05) (hash(x)=26076345)
923
+ 885 train 7.733073 (lr=1.5639e-05) (hash(x)=26577783)
924
+ 886 train 7.851583 (lr=1.5543e-05) (hash(x)=27395225)
925
+ 887 train 7.677545 (lr=1.5449e-05) (hash(x)=23926632)
926
+ 888 train 7.542912 (lr=1.5354e-05) (hash(x)=21737239)
927
+ 889 train 7.800176 (lr=1.5261e-05) (hash(x)=23574207)
928
+ 890 train 7.526219 (lr=1.5169e-05) (hash(x)=24365231)
929
+ 891 train 7.699145 (lr=1.5077e-05) (hash(x)=27111369)
930
+ 892 train 7.935370 (lr=1.4986e-05) (hash(x)=27290015)
931
+ 893 train 7.701787 (lr=1.4896e-05) (hash(x)=23979820)
932
+ 894 train 7.800383 (lr=1.4806e-05) (hash(x)=26450121)
933
+ 895 train 7.705535 (lr=1.4718e-05) (hash(x)=27025333)
934
+ 896 train 7.683148 (lr=1.4630e-05) (hash(x)=23624605)
935
+ 897 train 7.491131 (lr=1.4543e-05) (hash(x)=22846386)
936
+ 898 train 7.611001 (lr=1.4456e-05) (hash(x)=22970561)
937
+ 899 train 7.414521 (lr=1.4371e-05) (hash(x)=16908068)
938
+ 900 val loss 7.7042
939
+ 900 val perplexity 2217.7024
940
+ 900 train 7.669951 (lr=1.4286e-05) (hash(x)=24661446)
941
+ 901 train 7.739336 (lr=1.4202e-05) (hash(x)=25664727)
942
+ 902 train 7.737306 (lr=1.4119e-05) (hash(x)=25667011)
943
+ 903 train 7.813867 (lr=1.4037e-05) (hash(x)=29120407)
944
+ 904 train 7.802650 (lr=1.3955e-05) (hash(x)=23385735)
945
+ 905 train 7.853625 (lr=1.3874e-05) (hash(x)=25564213)
946
+ 906 train 7.925617 (lr=1.3794e-05) (hash(x)=25413898)
947
+ 907 train 7.897359 (lr=1.3715e-05) (hash(x)=27092710)
948
+ 908 train 7.694216 (lr=1.3637e-05) (hash(x)=25789923)
949
+ 909 train 7.799297 (lr=1.3559e-05) (hash(x)=28533197)
950
+ 910 train 7.624286 (lr=1.3482e-05) (hash(x)=22982996)
951
+ 911 train 7.711966 (lr=1.3407e-05) (hash(x)=23827393)
952
+ 912 train 7.339430 (lr=1.3331e-05) (hash(x)=21242640)
953
+ 913 train 7.708120 (lr=1.3257e-05) (hash(x)=24154233)
954
+ 914 train 7.717951 (lr=1.3183e-05) (hash(x)=24331967)
955
+ 915 train 7.946733 (lr=1.3111e-05) (hash(x)=32812727)
956
+ 916 train 7.605516 (lr=1.3039e-05) (hash(x)=23572994)
957
+ 917 train 7.650234 (lr=1.2968e-05) (hash(x)=26305435)
958
+ 918 train 7.697694 (lr=1.2897e-05) (hash(x)=26268355)
959
+ 919 train 7.867050 (lr=1.2828e-05) (hash(x)=27230027)
960
+ 920 train 7.676013 (lr=1.2759e-05) (hash(x)=23885377)
961
+ 921 train 7.581641 (lr=1.2691e-05) (hash(x)=23532437)
962
+ 922 train 7.691435 (lr=1.2624e-05) (hash(x)=25577034)
963
+ 923 train 7.822213 (lr=1.2558e-05) (hash(x)=25703381)
964
+ 924 train 7.817732 (lr=1.2493e-05) (hash(x)=27113866)
965
+ 925 train 7.662338 (lr=1.2428e-05) (hash(x)=26961429)
966
+ 926 train 7.465372 (lr=1.2364e-05) (hash(x)=21355372)
967
+ 927 train 7.635878 (lr=1.2301e-05) (hash(x)=24968260)
968
+ 928 train 7.619939 (lr=1.2239e-05) (hash(x)=25357517)
969
+ 929 train 7.676708 (lr=1.2178e-05) (hash(x)=24854265)
970
+ 930 train 7.436241 (lr=1.2118e-05) (hash(x)=21102770)
971
+ 931 train 7.770073 (lr=1.2058e-05) (hash(x)=25676468)
972
+ 932 train 7.685634 (lr=1.1999e-05) (hash(x)=22809869)
973
+ 933 train 7.775471 (lr=1.1941e-05) (hash(x)=25503865)
974
+ 934 train 7.650397 (lr=1.1884e-05) (hash(x)=24853995)
975
+ 935 train 7.795400 (lr=1.1828e-05) (hash(x)=27544803)
976
+ 936 train 7.841106 (lr=1.1772e-05) (hash(x)=25981933)
977
+ 937 train 7.555672 (lr=1.1718e-05) (hash(x)=24658683)
978
+ 938 train 7.756471 (lr=1.1664e-05) (hash(x)=23855201)
979
+ 939 train 7.695769 (lr=1.1611e-05) (hash(x)=24331407)
980
+ 940 train 7.892645 (lr=1.1559e-05) (hash(x)=29265551)
981
+ 941 train 7.596489 (lr=1.1508e-05) (hash(x)=21892556)
982
+ 942 train 7.610655 (lr=1.1457e-05) (hash(x)=27183405)
983
+ 943 train 7.610963 (lr=1.1408e-05) (hash(x)=26540663)
984
+ 944 train 7.572990 (lr=1.1359e-05) (hash(x)=25718393)
985
+ 945 train 7.752506 (lr=1.1311e-05) (hash(x)=26819462)
986
+ 946 train 7.745122 (lr=1.1264e-05) (hash(x)=27427540)
987
+ 947 train 7.722134 (lr=1.1218e-05) (hash(x)=25532657)
988
+ 948 train 7.964739 (lr=1.1173e-05) (hash(x)=27641372)
989
+ 949 train 7.688563 (lr=1.1128e-05) (hash(x)=26515570)
990
+ 950 val loss 7.7007
991
+ 950 val perplexity 2209.8940
992
+ 950 train 7.902711 (lr=1.1085e-05) (hash(x)=26911957)
993
+ 951 train 8.005329 (lr=1.1042e-05) (hash(x)=25856625)
994
+ 952 train 7.598493 (lr=1.1000e-05) (hash(x)=25219129)
995
+ 953 train 7.549696 (lr=1.0959e-05) (hash(x)=25260471)
996
+ 954 train 7.919159 (lr=1.0919e-05) (hash(x)=29373370)
997
+ 955 train 7.492577 (lr=1.0879e-05) (hash(x)=23437426)
998
+ 956 train 7.790935 (lr=1.0841e-05) (hash(x)=23769521)
999
+ 957 train 7.818006 (lr=1.0803e-05) (hash(x)=25961833)
1000
+ 958 train 7.551431 (lr=1.0766e-05) (hash(x)=23582666)
1001
+ 959 train 7.612159 (lr=1.0730e-05) (hash(x)=23164356)
1002
+ 960 train 7.600656 (lr=1.0695e-05) (hash(x)=24443114)
1003
+ 961 train 7.676736 (lr=1.0661e-05) (hash(x)=25052665)
1004
+ 962 train 7.806877 (lr=1.0628e-05) (hash(x)=27802272)
1005
+ 963 train 7.676126 (lr=1.0595e-05) (hash(x)=25957896)
1006
+ 964 train 7.725984 (lr=1.0563e-05) (hash(x)=26737251)
1007
+ 965 train 7.826626 (lr=1.0533e-05) (hash(x)=24723263)
1008
+ 966 train 7.767490 (lr=1.0503e-05) (hash(x)=24707011)
1009
+ 967 train 7.730196 (lr=1.0474e-05) (hash(x)=25646282)
1010
+ 968 train 7.746767 (lr=1.0445e-05) (hash(x)=27544665)
1011
+ 969 train 7.704871 (lr=1.0418e-05) (hash(x)=25851993)
1012
+ 970 train 7.958484 (lr=1.0391e-05) (hash(x)=29059700)
1013
+ 971 train 7.649293 (lr=1.0366e-05) (hash(x)=21513584)
1014
+ 972 train 7.492728 (lr=1.0341e-05) (hash(x)=23151267)
1015
+ 973 train 7.687877 (lr=1.0317e-05) (hash(x)=26017176)
1016
+ 974 train 7.794328 (lr=1.0294e-05) (hash(x)=26979518)
1017
+ 975 train 7.747479 (lr=1.0272e-05) (hash(x)=23843233)
1018
+ 976 train 7.881741 (lr=1.0251e-05) (hash(x)=24193010)
1019
+ 977 train 7.616655 (lr=1.0230e-05) (hash(x)=21476847)
1020
+ 978 train 7.511085 (lr=1.0211e-05) (hash(x)=21366504)
1021
+ 979 train 7.605165 (lr=1.0192e-05) (hash(x)=23226697)
1022
+ 980 train 7.391058 (lr=1.0174e-05) (hash(x)=19961773)
1023
+ 981 train 7.243707 (lr=1.0157e-05) (hash(x)=19772969)
1024
+ 982 train 7.547717 (lr=1.0141e-05) (hash(x)=23110142)
1025
+ 983 train 7.788452 (lr=1.0126e-05) (hash(x)=24506028)
1026
+ 984 train 7.760216 (lr=1.0111e-05) (hash(x)=25480731)
1027
+ 985 train 7.616599 (lr=1.0098e-05) (hash(x)=21077417)
1028
+ 986 train 7.592271 (lr=1.0085e-05) (hash(x)=23686713)
1029
+ 987 train 7.703046 (lr=1.0074e-05) (hash(x)=26024321)
1030
+ 988 train 7.771310 (lr=1.0063e-05) (hash(x)=27424109)
1031
+ 989 train 7.845740 (lr=1.0053e-05) (hash(x)=27786174)
1032
+ 990 train 7.678023 (lr=1.0044e-05) (hash(x)=25232502)
1033
+ 991 train 7.682714 (lr=1.0035e-05) (hash(x)=22781277)
1034
+ 992 train 7.722869 (lr=1.0028e-05) (hash(x)=26184527)
1035
+ 993 train 7.739462 (lr=1.0021e-05) (hash(x)=24459895)
1036
+ 994 train 7.692230 (lr=1.0016e-05) (hash(x)=25244624)
1037
+ 995 train 7.827173 (lr=1.0011e-05) (hash(x)=24451843)
1038
+ 996 train 7.789174 (lr=1.0007e-05) (hash(x)=22129897)
1039
+ 997 train 7.632939 (lr=1.0004e-05) (hash(x)=21116390)
1040
+ 998 train 7.501942 (lr=1.0002e-05) (hash(x)=20650070)
1041
+ 999 val loss 7.6894
1042
+ 999 val perplexity 2185.1177
1043
+ 999 train 7.815773 (lr=1.0000e-05) (hash(x)=24948650)
lr1e-4_total_batch_size5120_seq_len512/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1b2a4901b469a388a053e0ce97cd5e70376310bffc0dcfc0025eb3d373d6b21
3
+ size 48156226
lr1e-4_total_batch_size5120_seq_len512/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19cacf7a7cbedcce0b8dce22a42ec55015ebff3f28b678e590088e902a724c83
3
+ size 71157574