andrew-healey commited on
Commit
0199a3f
·
verified ·
1 Parent(s): 2042dea

Upload folder using huggingface_hub

Browse files
lr4e-4_total_batch_size5120_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr4e-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_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": 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": 0.0004, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr4e-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
lr4e-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=1.3986e-06) (hash(x)=22886834)
5
+ 1 train 11.761712 (lr=2.7972e-06) (hash(x)=26375038)
6
+ 2 train 11.727182 (lr=4.1958e-06) (hash(x)=30598777)
7
+ 3 train 11.770659 (lr=5.5944e-06) (hash(x)=27234506)
8
+ 4 train 11.720979 (lr=6.9930e-06) (hash(x)=27767880)
9
+ 5 train 11.710020 (lr=8.3916e-06) (hash(x)=23702020)
10
+ 6 train 11.721946 (lr=9.7902e-06) (hash(x)=31986844)
11
+ 7 train 11.732994 (lr=1.1189e-05) (hash(x)=20782690)
12
+ 8 train 11.743386 (lr=1.2587e-05) (hash(x)=25201599)
13
+ 9 train 11.751269 (lr=1.3986e-05) (hash(x)=23094976)
14
+ 10 train 11.695582 (lr=1.5385e-05) (hash(x)=23841096)
15
+ 11 train 11.662376 (lr=1.6783e-05) (hash(x)=26532095)
16
+ 12 train 11.674286 (lr=1.8182e-05) (hash(x)=24432298)
17
+ 13 train 11.622149 (lr=1.9580e-05) (hash(x)=27151649)
18
+ 14 train 11.604691 (lr=2.0979e-05) (hash(x)=24596846)
19
+ 15 train 11.598269 (lr=2.2378e-05) (hash(x)=23890908)
20
+ 16 train 11.581548 (lr=2.3776e-05) (hash(x)=28913955)
21
+ 17 train 11.529721 (lr=2.5175e-05) (hash(x)=25588236)
22
+ 18 train 11.477553 (lr=2.6573e-05) (hash(x)=23770034)
23
+ 19 train 11.472409 (lr=2.7972e-05) (hash(x)=24011372)
24
+ 20 train 11.386425 (lr=2.9371e-05) (hash(x)=25441898)
25
+ 21 train 11.389735 (lr=3.0769e-05) (hash(x)=28375581)
26
+ 22 train 11.334499 (lr=3.2168e-05) (hash(x)=24046679)
27
+ 23 train 11.338481 (lr=3.3566e-05) (hash(x)=24611628)
28
+ 24 train 11.331514 (lr=3.4965e-05) (hash(x)=26169030)
29
+ 25 train 11.245864 (lr=3.6364e-05) (hash(x)=30298407)
30
+ 26 train 11.196999 (lr=3.7762e-05) (hash(x)=23711112)
31
+ 27 train 11.038733 (lr=3.9161e-05) (hash(x)=19245352)
32
+ 28 train 11.025330 (lr=4.0559e-05) (hash(x)=21529136)
33
+ 29 train 11.088989 (lr=4.1958e-05) (hash(x)=28936608)
34
+ 30 train 10.900310 (lr=4.3357e-05) (hash(x)=24339013)
35
+ 31 train 10.748546 (lr=4.4755e-05) (hash(x)=25767553)
36
+ 32 train 10.761757 (lr=4.6154e-05) (hash(x)=26439905)
37
+ 33 train 10.887057 (lr=4.7552e-05) (hash(x)=31093473)
38
+ 34 train 10.661058 (lr=4.8951e-05) (hash(x)=25450374)
39
+ 35 train 10.662196 (lr=5.0350e-05) (hash(x)=24809873)
40
+ 36 train 10.709335 (lr=5.1748e-05) (hash(x)=23253252)
41
+ 37 train 10.737584 (lr=5.3147e-05) (hash(x)=27852919)
42
+ 38 train 10.527649 (lr=5.4545e-05) (hash(x)=23327497)
43
+ 39 train 10.543945 (lr=5.5944e-05) (hash(x)=22512166)
44
+ 40 train 10.617921 (lr=5.7343e-05) (hash(x)=22859419)
45
+ 41 train 10.413960 (lr=5.8741e-05) (hash(x)=27620338)
46
+ 42 train 10.410434 (lr=6.0140e-05) (hash(x)=26397837)
47
+ 43 train 10.450401 (lr=6.1538e-05) (hash(x)=28092148)
48
+ 44 train 10.365738 (lr=6.2937e-05) (hash(x)=24662703)
49
+ 45 train 10.439790 (lr=6.4336e-05) (hash(x)=27938767)
50
+ 46 train 10.517088 (lr=6.5734e-05) (hash(x)=26037988)
51
+ 47 train 10.282906 (lr=6.7133e-05) (hash(x)=24732833)
52
+ 48 train 10.324231 (lr=6.8531e-05) (hash(x)=25259526)
53
+ 49 train 10.334154 (lr=6.9930e-05) (hash(x)=23200230)
54
+ 50 val loss 10.2945
55
+ 50 val perplexity 29568.9004
56
+ 50 train 10.303683 (lr=7.1329e-05) (hash(x)=26721357)
57
+ 51 train 10.305330 (lr=7.2727e-05) (hash(x)=22694718)
58
+ 52 train 10.243316 (lr=7.4126e-05) (hash(x)=28066766)
59
+ 53 train 10.203763 (lr=7.5524e-05) (hash(x)=23125151)
60
+ 54 train 10.273292 (lr=7.6923e-05) (hash(x)=27193725)
61
+ 55 train 10.159582 (lr=7.8322e-05) (hash(x)=25129410)
62
+ 56 train 10.192543 (lr=7.9720e-05) (hash(x)=24263988)
63
+ 57 train 10.051816 (lr=8.1119e-05) (hash(x)=23059154)
64
+ 58 train 10.176322 (lr=8.2517e-05) (hash(x)=26063864)
65
+ 59 train 10.316957 (lr=8.3916e-05) (hash(x)=27858570)
66
+ 60 train 10.088168 (lr=8.5315e-05) (hash(x)=23874620)
67
+ 61 train 10.084573 (lr=8.6713e-05) (hash(x)=22402617)
68
+ 62 train 10.073235 (lr=8.8112e-05) (hash(x)=23600822)
69
+ 63 train 9.873080 (lr=8.9510e-05) (hash(x)=26582391)
70
+ 64 train 9.895378 (lr=9.0909e-05) (hash(x)=23225283)
71
+ 65 train 10.087008 (lr=9.2308e-05) (hash(x)=26075451)
72
+ 66 train 10.011303 (lr=9.3706e-05) (hash(x)=24723419)
73
+ 67 train 9.929804 (lr=9.5105e-05) (hash(x)=27279806)
74
+ 68 train 9.862325 (lr=9.6503e-05) (hash(x)=25870391)
75
+ 69 train 9.944585 (lr=9.7902e-05) (hash(x)=26188136)
76
+ 70 train 9.857930 (lr=9.9301e-05) (hash(x)=30373443)
77
+ 71 train 9.947771 (lr=1.0070e-04) (hash(x)=26472336)
78
+ 72 train 9.856348 (lr=1.0210e-04) (hash(x)=26651572)
79
+ 73 train 9.795497 (lr=1.0350e-04) (hash(x)=26376212)
80
+ 74 train 9.796396 (lr=1.0490e-04) (hash(x)=26733350)
81
+ 75 train 9.920378 (lr=1.0629e-04) (hash(x)=28301589)
82
+ 76 train 9.783683 (lr=1.0769e-04) (hash(x)=27599559)
83
+ 77 train 9.708537 (lr=1.0909e-04) (hash(x)=28035221)
84
+ 78 train 9.779418 (lr=1.1049e-04) (hash(x)=25016783)
85
+ 79 train 9.794131 (lr=1.1189e-04) (hash(x)=27654289)
86
+ 80 train 9.793641 (lr=1.1329e-04) (hash(x)=24597558)
87
+ 81 train 9.727638 (lr=1.1469e-04) (hash(x)=21560904)
88
+ 82 train 9.533305 (lr=1.1608e-04) (hash(x)=21983837)
89
+ 83 train 9.613105 (lr=1.1748e-04) (hash(x)=24995715)
90
+ 84 train 9.711059 (lr=1.1888e-04) (hash(x)=29876413)
91
+ 85 train 9.570414 (lr=1.2028e-04) (hash(x)=23792508)
92
+ 86 train 9.483134 (lr=1.2168e-04) (hash(x)=25509120)
93
+ 87 train 9.688270 (lr=1.2308e-04) (hash(x)=26559876)
94
+ 88 train 9.592445 (lr=1.2448e-04) (hash(x)=23569647)
95
+ 89 train 9.585902 (lr=1.2587e-04) (hash(x)=25758852)
96
+ 90 train 9.456041 (lr=1.2727e-04) (hash(x)=25706298)
97
+ 91 train 9.404999 (lr=1.2867e-04) (hash(x)=28364895)
98
+ 92 train 9.399004 (lr=1.3007e-04) (hash(x)=25304663)
99
+ 93 train 9.500826 (lr=1.3147e-04) (hash(x)=25269299)
100
+ 94 train 9.620717 (lr=1.3287e-04) (hash(x)=25870566)
101
+ 95 train 9.366263 (lr=1.3427e-04) (hash(x)=21770329)
102
+ 96 train 9.600039 (lr=1.3566e-04) (hash(x)=27595900)
103
+ 97 train 9.579823 (lr=1.3706e-04) (hash(x)=24785397)
104
+ 98 train 9.328650 (lr=1.3846e-04) (hash(x)=21521480)
105
+ 99 train 9.378031 (lr=1.3986e-04) (hash(x)=24628606)
106
+ 100 val loss 9.3878
107
+ 100 val perplexity 11942.1201
108
+ 100 train 9.429362 (lr=1.4126e-04) (hash(x)=24670150)
109
+ 101 train 9.399013 (lr=1.4266e-04) (hash(x)=23181910)
110
+ 102 train 9.258065 (lr=1.4406e-04) (hash(x)=22714991)
111
+ 103 train 9.323733 (lr=1.4545e-04) (hash(x)=22723459)
112
+ 104 train 9.236087 (lr=1.4685e-04) (hash(x)=21524316)
113
+ 105 train 9.277870 (lr=1.4825e-04) (hash(x)=25506632)
114
+ 106 train 9.022864 (lr=1.4965e-04) (hash(x)=21675672)
115
+ 107 train 9.132483 (lr=1.5105e-04) (hash(x)=22897919)
116
+ 108 train 9.160258 (lr=1.5245e-04) (hash(x)=23321631)
117
+ 109 train 9.191370 (lr=1.5385e-04) (hash(x)=26546719)
118
+ 110 train 9.890983 (lr=1.5524e-04) (hash(x)=31962348)
119
+ 111 train 9.228190 (lr=1.5664e-04) (hash(x)=30338342)
120
+ 112 train 9.137326 (lr=1.5804e-04) (hash(x)=23724471)
121
+ 113 train 9.116721 (lr=1.5944e-04) (hash(x)=29175888)
122
+ 114 train 9.259405 (lr=1.6084e-04) (hash(x)=23256716)
123
+ 115 train 9.176568 (lr=1.6224e-04) (hash(x)=27063280)
124
+ 116 train 9.134725 (lr=1.6364e-04) (hash(x)=31057659)
125
+ 117 train 9.288605 (lr=1.6503e-04) (hash(x)=32915097)
126
+ 118 train 9.063211 (lr=1.6643e-04) (hash(x)=28842717)
127
+ 119 train 8.955272 (lr=1.6783e-04) (hash(x)=25678059)
128
+ 120 train 8.896905 (lr=1.6923e-04) (hash(x)=21593510)
129
+ 121 train 8.738500 (lr=1.7063e-04) (hash(x)=20083773)
130
+ 122 train 8.879544 (lr=1.7203e-04) (hash(x)=23002820)
131
+ 123 train 8.769031 (lr=1.7343e-04) (hash(x)=21853028)
132
+ 124 train 8.858190 (lr=1.7483e-04) (hash(x)=26985625)
133
+ 125 train 8.700994 (lr=1.7622e-04) (hash(x)=21808483)
134
+ 126 train 8.982949 (lr=1.7762e-04) (hash(x)=28873251)
135
+ 127 train 8.795117 (lr=1.7902e-04) (hash(x)=26109335)
136
+ 128 train 8.859117 (lr=1.8042e-04) (hash(x)=26334674)
137
+ 129 train 8.799540 (lr=1.8182e-04) (hash(x)=24916754)
138
+ 130 train 8.779594 (lr=1.8322e-04) (hash(x)=25449624)
139
+ 131 train 8.848719 (lr=1.8462e-04) (hash(x)=25334848)
140
+ 132 train 9.000388 (lr=1.8601e-04) (hash(x)=27484863)
141
+ 133 train 8.669944 (lr=1.8741e-04) (hash(x)=24917705)
142
+ 134 train 8.596847 (lr=1.8881e-04) (hash(x)=25143449)
143
+ 135 train 8.635745 (lr=1.9021e-04) (hash(x)=25044885)
144
+ 136 train 8.857414 (lr=1.9161e-04) (hash(x)=27821028)
145
+ 137 train 8.767868 (lr=1.9301e-04) (hash(x)=28747022)
146
+ 138 train 8.713805 (lr=1.9441e-04) (hash(x)=27182888)
147
+ 139 train 8.535649 (lr=1.9580e-04) (hash(x)=23678349)
148
+ 140 train 8.629613 (lr=1.9720e-04) (hash(x)=23593235)
149
+ 141 train 8.814885 (lr=1.9860e-04) (hash(x)=28529813)
150
+ 142 train 9.049083 (lr=2.0000e-04) (hash(x)=32074661)
151
+ 143 train 8.634256 (lr=2.0140e-04) (hash(x)=28870690)
152
+ 144 train 8.632668 (lr=2.0280e-04) (hash(x)=27307705)
153
+ 145 train 8.547359 (lr=2.0420e-04) (hash(x)=25044834)
154
+ 146 train 8.605483 (lr=2.0559e-04) (hash(x)=23712023)
155
+ 147 train 8.919071 (lr=2.0699e-04) (hash(x)=32982615)
156
+ 148 train 8.720126 (lr=2.0839e-04) (hash(x)=30113660)
157
+ 149 train 8.300377 (lr=2.0979e-04) (hash(x)=20970960)
158
+ 150 val loss 8.4402
159
+ 150 val perplexity 4629.6831
160
+ 150 train 8.405448 (lr=2.1119e-04) (hash(x)=23132684)
161
+ 151 train 9.974734 (lr=2.1259e-04) (hash(x)=35279941)
162
+ 152 train 9.209045 (lr=2.1399e-04) (hash(x)=31227444)
163
+ 153 train 8.347156 (lr=2.1538e-04) (hash(x)=25529472)
164
+ 154 train 8.381366 (lr=2.1678e-04) (hash(x)=24350409)
165
+ 155 train 8.701562 (lr=2.1818e-04) (hash(x)=26400041)
166
+ 156 train 8.346365 (lr=2.1958e-04) (hash(x)=25262621)
167
+ 157 train 8.370531 (lr=2.2098e-04) (hash(x)=24656138)
168
+ 158 train 8.490586 (lr=2.2238e-04) (hash(x)=26803414)
169
+ 159 train 8.388156 (lr=2.2378e-04) (hash(x)=25015923)
170
+ 160 train 8.195048 (lr=2.2517e-04) (hash(x)=23581172)
171
+ 161 train 8.282908 (lr=2.2657e-04) (hash(x)=22924885)
172
+ 162 train 8.206004 (lr=2.2797e-04) (hash(x)=23414296)
173
+ 163 train 8.466092 (lr=2.2937e-04) (hash(x)=24853586)
174
+ 164 train 8.483652 (lr=2.3077e-04) (hash(x)=25000130)
175
+ 165 train 8.285308 (lr=2.3217e-04) (hash(x)=27004780)
176
+ 166 train 8.306430 (lr=2.3357e-04) (hash(x)=26148573)
177
+ 167 train 8.231824 (lr=2.3497e-04) (hash(x)=26740855)
178
+ 168 train 7.896873 (lr=2.3636e-04) (hash(x)=20965419)
179
+ 169 train 8.137344 (lr=2.3776e-04) (hash(x)=23950114)
180
+ 170 train 8.115554 (lr=2.3916e-04) (hash(x)=24951982)
181
+ 171 train 8.220584 (lr=2.4056e-04) (hash(x)=24584116)
182
+ 172 train 8.152628 (lr=2.4196e-04) (hash(x)=24378759)
183
+ 173 train 8.123774 (lr=2.4336e-04) (hash(x)=25718516)
184
+ 174 train 8.236064 (lr=2.4476e-04) (hash(x)=28424396)
185
+ 175 train 8.052516 (lr=2.4615e-04) (hash(x)=22262151)
186
+ 176 train 8.257121 (lr=2.4755e-04) (hash(x)=26438412)
187
+ 177 train 8.046446 (lr=2.4895e-04) (hash(x)=23025303)
188
+ 178 train 8.028555 (lr=2.5035e-04) (hash(x)=24190770)
189
+ 179 train 8.180405 (lr=2.5175e-04) (hash(x)=26627860)
190
+ 180 train 8.040138 (lr=2.5315e-04) (hash(x)=23663439)
191
+ 181 train 7.871807 (lr=2.5455e-04) (hash(x)=23700532)
192
+ 182 train 7.905663 (lr=2.5594e-04) (hash(x)=23075676)
193
+ 183 train 8.090796 (lr=2.5734e-04) (hash(x)=26621834)
194
+ 184 train 8.149928 (lr=2.5874e-04) (hash(x)=29426269)
195
+ 185 train 7.901958 (lr=2.6014e-04) (hash(x)=21821465)
196
+ 186 train 8.072777 (lr=2.6154e-04) (hash(x)=26506130)
197
+ 187 train 8.282500 (lr=2.6294e-04) (hash(x)=26930630)
198
+ 188 train 7.875179 (lr=2.6434e-04) (hash(x)=22993793)
199
+ 189 train 7.775772 (lr=2.6573e-04) (hash(x)=19557946)
200
+ 190 train 8.020312 (lr=2.6713e-04) (hash(x)=23572891)
201
+ 191 train 7.862176 (lr=2.6853e-04) (hash(x)=23234741)
202
+ 192 train 7.907533 (lr=2.6993e-04) (hash(x)=25547951)
203
+ 193 train 7.968283 (lr=2.7133e-04) (hash(x)=26713563)
204
+ 194 train 7.990827 (lr=2.7273e-04) (hash(x)=25913622)
205
+ 195 train 8.252502 (lr=2.7413e-04) (hash(x)=28102443)
206
+ 196 train 7.980217 (lr=2.7552e-04) (hash(x)=23093351)
207
+ 197 train 8.627557 (lr=2.7692e-04) (hash(x)=31689122)
208
+ 198 train 8.477883 (lr=2.7832e-04) (hash(x)=32252517)
209
+ 199 train 8.018550 (lr=2.7972e-04) (hash(x)=25470563)
210
+ 200 val loss 7.9532
211
+ 200 val perplexity 2844.7117
212
+ 200 train 8.059333 (lr=2.8112e-04) (hash(x)=25597614)
213
+ 201 train 8.087310 (lr=2.8252e-04) (hash(x)=23757479)
214
+ 202 train 7.913934 (lr=2.8392e-04) (hash(x)=25140048)
215
+ 203 train 8.094152 (lr=2.8531e-04) (hash(x)=28282861)
216
+ 204 train 7.887373 (lr=2.8671e-04) (hash(x)=24754885)
217
+ 205 train 8.118670 (lr=2.8811e-04) (hash(x)=26731964)
218
+ 206 train 8.191173 (lr=2.8951e-04) (hash(x)=26660561)
219
+ 207 train 7.769304 (lr=2.9091e-04) (hash(x)=21799102)
220
+ 208 train 7.982753 (lr=2.9231e-04) (hash(x)=27025986)
221
+ 209 train 7.824104 (lr=2.9371e-04) (hash(x)=24672077)
222
+ 210 train 7.908099 (lr=2.9510e-04) (hash(x)=25322984)
223
+ 211 train 7.741774 (lr=2.9650e-04) (hash(x)=23471769)
224
+ 212 train 7.586157 (lr=2.9790e-04) (hash(x)=20766491)
225
+ 213 train 8.071383 (lr=2.9930e-04) (hash(x)=24058931)
226
+ 214 train 7.764444 (lr=3.0070e-04) (hash(x)=23872843)
227
+ 215 train 7.806528 (lr=3.0210e-04) (hash(x)=23275479)
228
+ 216 train 7.934415 (lr=3.0350e-04) (hash(x)=24914695)
229
+ 217 train 7.788907 (lr=3.0490e-04) (hash(x)=24364396)
230
+ 218 train 8.022510 (lr=3.0629e-04) (hash(x)=27986474)
231
+ 219 train 7.948164 (lr=3.0769e-04) (hash(x)=24473581)
232
+ 220 train 7.870291 (lr=3.0909e-04) (hash(x)=22974689)
233
+ 221 train 7.935555 (lr=3.1049e-04) (hash(x)=23774644)
234
+ 222 train 7.893469 (lr=3.1189e-04) (hash(x)=23245327)
235
+ 223 train 7.716710 (lr=3.1329e-04) (hash(x)=22091862)
236
+ 224 train 7.780473 (lr=3.1469e-04) (hash(x)=24362839)
237
+ 225 train 7.886597 (lr=3.1608e-04) (hash(x)=25482303)
238
+ 226 train 7.978626 (lr=3.1748e-04) (hash(x)=24911853)
239
+ 227 train 7.964651 (lr=3.1888e-04) (hash(x)=26018202)
240
+ 228 train 7.794415 (lr=3.2028e-04) (hash(x)=26124495)
241
+ 229 train 7.887305 (lr=3.2168e-04) (hash(x)=24560096)
242
+ 230 train 7.879899 (lr=3.2308e-04) (hash(x)=24695331)
243
+ 231 train 7.587526 (lr=3.2448e-04) (hash(x)=17430373)
244
+ 232 train 7.669260 (lr=3.2587e-04) (hash(x)=21813345)
245
+ 233 train 7.264420 (lr=3.2727e-04) (hash(x)=20098681)
246
+ 234 train 7.608260 (lr=3.2867e-04) (hash(x)=25095928)
247
+ 235 train 7.735494 (lr=3.3007e-04) (hash(x)=24078083)
248
+ 236 train 7.754076 (lr=3.3147e-04) (hash(x)=22901505)
249
+ 237 train 7.927379 (lr=3.3287e-04) (hash(x)=26595592)
250
+ 238 train 7.880264 (lr=3.3427e-04) (hash(x)=27663196)
251
+ 239 train 7.649796 (lr=3.3566e-04) (hash(x)=22954861)
252
+ 240 train 8.070422 (lr=3.3706e-04) (hash(x)=30159234)
253
+ 241 train 7.971080 (lr=3.3846e-04) (hash(x)=29294271)
254
+ 242 train 7.925672 (lr=3.3986e-04) (hash(x)=23728322)
255
+ 243 train 7.981122 (lr=3.4126e-04) (hash(x)=28695016)
256
+ 244 train 8.086043 (lr=3.4266e-04) (hash(x)=26702728)
257
+ 245 train 7.916951 (lr=3.4406e-04) (hash(x)=27100115)
258
+ 246 train 7.818809 (lr=3.4545e-04) (hash(x)=25082752)
259
+ 247 train 8.172284 (lr=3.4685e-04) (hash(x)=26671799)
260
+ 248 train 7.850059 (lr=3.4825e-04) (hash(x)=23718946)
261
+ 249 train 8.070689 (lr=3.4965e-04) (hash(x)=28137394)
262
+ 250 val loss 7.8559
263
+ 250 val perplexity 2580.8499
264
+ 250 train 8.103876 (lr=3.5105e-04) (hash(x)=23893495)
265
+ 251 train 7.889508 (lr=3.5245e-04) (hash(x)=23166092)
266
+ 252 train 7.866229 (lr=3.5385e-04) (hash(x)=25907665)
267
+ 253 train 7.842268 (lr=3.5524e-04) (hash(x)=25885986)
268
+ 254 train 7.877132 (lr=3.5664e-04) (hash(x)=25262712)
269
+ 255 train 7.850058 (lr=3.5804e-04) (hash(x)=26924723)
270
+ 256 train 8.035642 (lr=3.5944e-04) (hash(x)=28744736)
271
+ 257 train 7.765707 (lr=3.6084e-04) (hash(x)=26140590)
272
+ 258 train 7.961320 (lr=3.6224e-04) (hash(x)=25780449)
273
+ 259 train 7.810725 (lr=3.6364e-04) (hash(x)=25138659)
274
+ 260 train 8.095762 (lr=3.6503e-04) (hash(x)=27244046)
275
+ 261 train 7.906032 (lr=3.6643e-04) (hash(x)=27224685)
276
+ 262 train 8.183617 (lr=3.6783e-04) (hash(x)=28274477)
277
+ 263 train 7.708787 (lr=3.6923e-04) (hash(x)=23557495)
278
+ 264 train 7.865103 (lr=3.7063e-04) (hash(x)=24680596)
279
+ 265 train 7.888007 (lr=3.7203e-04) (hash(x)=23928957)
280
+ 266 train 7.810831 (lr=3.7343e-04) (hash(x)=23761390)
281
+ 267 train 7.712389 (lr=3.7483e-04) (hash(x)=25288123)
282
+ 268 train 8.012771 (lr=3.7622e-04) (hash(x)=28705502)
283
+ 269 train 7.592998 (lr=3.7762e-04) (hash(x)=23246294)
284
+ 270 train 7.680918 (lr=3.7902e-04) (hash(x)=28639079)
285
+ 271 train 7.571797 (lr=3.8042e-04) (hash(x)=27804380)
286
+ 272 train 7.834014 (lr=3.8182e-04) (hash(x)=24172235)
287
+ 273 train 7.866663 (lr=3.8322e-04) (hash(x)=23089140)
288
+ 274 train 8.077326 (lr=3.8462e-04) (hash(x)=27163701)
289
+ 275 train 8.187195 (lr=3.8601e-04) (hash(x)=26993263)
290
+ 276 train 7.932938 (lr=3.8741e-04) (hash(x)=28224233)
291
+ 277 train 8.204275 (lr=3.8881e-04) (hash(x)=27397203)
292
+ 278 train 8.113960 (lr=3.9021e-04) (hash(x)=27923882)
293
+ 279 train 7.928970 (lr=3.9161e-04) (hash(x)=26654908)
294
+ 280 train 7.893444 (lr=3.9301e-04) (hash(x)=24213147)
295
+ 281 train 7.574605 (lr=3.9441e-04) (hash(x)=21965022)
296
+ 282 train 7.770429 (lr=3.9580e-04) (hash(x)=25465685)
297
+ 283 train 7.998641 (lr=3.9720e-04) (hash(x)=27347722)
298
+ 284 train 7.986525 (lr=3.9860e-04) (hash(x)=26732050)
299
+ 285 train 8.135093 (lr=4.0000e-04) (hash(x)=28314127)
300
+ 286 train 7.715266 (lr=4.0000e-04) (hash(x)=21471186)
301
+ 287 train 7.738185 (lr=4.0000e-04) (hash(x)=23627518)
302
+ 288 train 7.751136 (lr=3.9999e-04) (hash(x)=20870353)
303
+ 289 train 7.763999 (lr=3.9998e-04) (hash(x)=25024764)
304
+ 290 train 7.529289 (lr=3.9997e-04) (hash(x)=20683822)
305
+ 291 train 7.866104 (lr=3.9996e-04) (hash(x)=21768671)
306
+ 292 train 7.891206 (lr=3.9994e-04) (hash(x)=25557309)
307
+ 293 train 8.020423 (lr=3.9991e-04) (hash(x)=25076667)
308
+ 294 train 7.684668 (lr=3.9989e-04) (hash(x)=23765822)
309
+ 295 train 7.668614 (lr=3.9986e-04) (hash(x)=21889990)
310
+ 296 train 7.862082 (lr=3.9983e-04) (hash(x)=26339893)
311
+ 297 train 7.557837 (lr=3.9979e-04) (hash(x)=20932794)
312
+ 298 train 7.740137 (lr=3.9975e-04) (hash(x)=21750070)
313
+ 299 train 7.850974 (lr=3.9971e-04) (hash(x)=23665838)
314
+ 300 val loss 7.8239
315
+ 300 val perplexity 2499.6367
316
+ 300 train 8.578812 (lr=3.9966e-04) (hash(x)=32888061)
317
+ 301 train 8.357419 (lr=3.9961e-04) (hash(x)=30223582)
318
+ 302 train 8.082840 (lr=3.9955e-04) (hash(x)=26908418)
319
+ 303 train 7.517303 (lr=3.9950e-04) (hash(x)=22528001)
320
+ 304 train 8.064641 (lr=3.9944e-04) (hash(x)=27452187)
321
+ 305 train 7.883252 (lr=3.9937e-04) (hash(x)=25181641)
322
+ 306 train 7.971297 (lr=3.9930e-04) (hash(x)=25546593)
323
+ 307 train 7.918054 (lr=3.9923e-04) (hash(x)=22487328)
324
+ 308 train 8.057729 (lr=3.9916e-04) (hash(x)=27804274)
325
+ 309 train 8.124686 (lr=3.9908e-04) (hash(x)=26544630)
326
+ 310 train 8.252448 (lr=3.9900e-04) (hash(x)=27738934)
327
+ 311 train 8.035403 (lr=3.9891e-04) (hash(x)=29248942)
328
+ 312 train 7.800590 (lr=3.9882e-04) (hash(x)=25103452)
329
+ 313 train 8.029451 (lr=3.9873e-04) (hash(x)=25052066)
330
+ 314 train 7.901668 (lr=3.9864e-04) (hash(x)=24481302)
331
+ 315 train 7.856209 (lr=3.9854e-04) (hash(x)=23543273)
332
+ 316 train 7.838129 (lr=3.9843e-04) (hash(x)=25608244)
333
+ 317 train 8.050049 (lr=3.9833e-04) (hash(x)=27451288)
334
+ 318 train 7.674034 (lr=3.9822e-04) (hash(x)=22806491)
335
+ 319 train 7.828311 (lr=3.9811e-04) (hash(x)=25533417)
336
+ 320 train 7.811433 (lr=3.9799e-04) (hash(x)=24557997)
337
+ 321 train 7.852129 (lr=3.9787e-04) (hash(x)=24432899)
338
+ 322 train 8.060210 (lr=3.9775e-04) (hash(x)=27583287)
339
+ 323 train 8.004189 (lr=3.9762e-04) (hash(x)=25552036)
340
+ 324 train 7.741361 (lr=3.9749e-04) (hash(x)=24201868)
341
+ 325 train 8.097655 (lr=3.9736e-04) (hash(x)=28149782)
342
+ 326 train 7.856027 (lr=3.9722e-04) (hash(x)=25529698)
343
+ 327 train 7.492310 (lr=3.9708e-04) (hash(x)=20612533)
344
+ 328 train 7.544360 (lr=3.9694e-04) (hash(x)=20699000)
345
+ 329 train 7.472562 (lr=3.9679e-04) (hash(x)=19774173)
346
+ 330 train 7.491451 (lr=3.9664e-04) (hash(x)=21681646)
347
+ 331 train 7.343315 (lr=3.9648e-04) (hash(x)=20216795)
348
+ 332 train 8.146749 (lr=3.9633e-04) (hash(x)=27697998)
349
+ 333 train 7.882796 (lr=3.9616e-04) (hash(x)=25896435)
350
+ 334 train 7.771783 (lr=3.9600e-04) (hash(x)=21585310)
351
+ 335 train 7.838141 (lr=3.9583e-04) (hash(x)=24677740)
352
+ 336 train 7.707747 (lr=3.9566e-04) (hash(x)=22027900)
353
+ 337 train 7.713793 (lr=3.9549e-04) (hash(x)=21835643)
354
+ 338 train 7.598275 (lr=3.9531e-04) (hash(x)=20558462)
355
+ 339 train 7.485750 (lr=3.9513e-04) (hash(x)=16722715)
356
+ 340 train 7.645551 (lr=3.9494e-04) (hash(x)=20972655)
357
+ 341 train 8.130673 (lr=3.9475e-04) (hash(x)=26303975)
358
+ 342 train 7.835675 (lr=3.9456e-04) (hash(x)=22938170)
359
+ 343 train 7.885873 (lr=3.9437e-04) (hash(x)=25347203)
360
+ 344 train 7.852840 (lr=3.9417e-04) (hash(x)=27398686)
361
+ 345 train 7.877520 (lr=3.9397e-04) (hash(x)=25973417)
362
+ 346 train 7.889467 (lr=3.9376e-04) (hash(x)=26918389)
363
+ 347 train 7.854825 (lr=3.9356e-04) (hash(x)=24272489)
364
+ 348 train 7.849720 (lr=3.9334e-04) (hash(x)=25593714)
365
+ 349 train 7.899414 (lr=3.9313e-04) (hash(x)=29260846)
366
+ 350 val loss 7.8182
367
+ 350 val perplexity 2485.4043
368
+ 350 train 7.962194 (lr=3.9291e-04) (hash(x)=27951602)
369
+ 351 train 8.058135 (lr=3.9269e-04) (hash(x)=28922363)
370
+ 352 train 7.918921 (lr=3.9246e-04) (hash(x)=27210734)
371
+ 353 train 7.986090 (lr=3.9223e-04) (hash(x)=26322572)
372
+ 354 train 7.854704 (lr=3.9200e-04) (hash(x)=27084665)
373
+ 355 train 7.752638 (lr=3.9177e-04) (hash(x)=25510798)
374
+ 356 train 7.862178 (lr=3.9153e-04) (hash(x)=24970921)
375
+ 357 train 7.898376 (lr=3.9129e-04) (hash(x)=24138948)
376
+ 358 train 7.770271 (lr=3.9104e-04) (hash(x)=24790211)
377
+ 359 train 7.915450 (lr=3.9079e-04) (hash(x)=25631397)
378
+ 360 train 7.693350 (lr=3.9054e-04) (hash(x)=23226625)
379
+ 361 train 7.789601 (lr=3.9029e-04) (hash(x)=24001903)
380
+ 362 train 7.766843 (lr=3.9003e-04) (hash(x)=24587948)
381
+ 363 train 7.431453 (lr=3.8977e-04) (hash(x)=21333676)
382
+ 364 train 7.274203 (lr=3.8950e-04) (hash(x)=23673779)
383
+ 365 train 8.841041 (lr=3.8923e-04) (hash(x)=30770484)
384
+ 366 train 7.888419 (lr=3.8896e-04) (hash(x)=26564899)
385
+ 367 train 7.858365 (lr=3.8869e-04) (hash(x)=26237983)
386
+ 368 train 7.593766 (lr=3.8841e-04) (hash(x)=23764356)
387
+ 369 train 7.937840 (lr=3.8813e-04) (hash(x)=26205744)
388
+ 370 train 7.656406 (lr=3.8785e-04) (hash(x)=19208770)
389
+ 371 train 7.848132 (lr=3.8756e-04) (hash(x)=25976502)
390
+ 372 train 7.707138 (lr=3.8727e-04) (hash(x)=23983933)
391
+ 373 train 7.702058 (lr=3.8697e-04) (hash(x)=24080636)
392
+ 374 train 7.790575 (lr=3.8667e-04) (hash(x)=24404047)
393
+ 375 train 7.741900 (lr=3.8637e-04) (hash(x)=24742645)
394
+ 376 train 7.781909 (lr=3.8607e-04) (hash(x)=24159600)
395
+ 377 train 8.023674 (lr=3.8576e-04) (hash(x)=28677257)
396
+ 378 train 7.734386 (lr=3.8545e-04) (hash(x)=25604111)
397
+ 379 train 7.711198 (lr=3.8514e-04) (hash(x)=27086333)
398
+ 380 train 7.704967 (lr=3.8482e-04) (hash(x)=25188207)
399
+ 381 train 7.913597 (lr=3.8450e-04) (hash(x)=27855233)
400
+ 382 train 7.494791 (lr=3.8418e-04) (hash(x)=19470039)
401
+ 383 train 7.747125 (lr=3.8385e-04) (hash(x)=26157660)
402
+ 384 train 7.691605 (lr=3.8352e-04) (hash(x)=25291570)
403
+ 385 train 7.674989 (lr=3.8319e-04) (hash(x)=25046062)
404
+ 386 train 7.945024 (lr=3.8286e-04) (hash(x)=27020337)
405
+ 387 train 7.615159 (lr=3.8252e-04) (hash(x)=23616370)
406
+ 388 train 7.293591 (lr=3.8217e-04) (hash(x)=19113218)
407
+ 389 train 7.641786 (lr=3.8183e-04) (hash(x)=24302232)
408
+ 390 train 7.667743 (lr=3.8148e-04) (hash(x)=22188949)
409
+ 391 train 7.445746 (lr=3.8113e-04) (hash(x)=22582169)
410
+ 392 train 7.613426 (lr=3.8077e-04) (hash(x)=24700570)
411
+ 393 train 7.549601 (lr=3.8042e-04) (hash(x)=22773833)
412
+ 394 train 7.426831 (lr=3.8006e-04) (hash(x)=21875928)
413
+ 395 train 7.724616 (lr=3.7969e-04) (hash(x)=26233189)
414
+ 396 train 7.547325 (lr=3.7933e-04) (hash(x)=24321467)
415
+ 397 train 7.824846 (lr=3.7896e-04) (hash(x)=26431507)
416
+ 398 train 7.925646 (lr=3.7858e-04) (hash(x)=28690877)
417
+ 399 train 7.846927 (lr=3.7821e-04) (hash(x)=26431960)
418
+ 400 val loss 7.7337
419
+ 400 val perplexity 2283.9282
420
+ 400 train 7.715025 (lr=3.7783e-04) (hash(x)=24580300)
421
+ 401 train 7.624078 (lr=3.7744e-04) (hash(x)=25112360)
422
+ 402 train 7.716455 (lr=3.7706e-04) (hash(x)=27597243)
423
+ 403 train 8.216340 (lr=3.7667e-04) (hash(x)=30707498)
424
+ 404 train 7.837725 (lr=3.7628e-04) (hash(x)=28485465)
425
+ 405 train 7.484231 (lr=3.7588e-04) (hash(x)=22586447)
426
+ 406 train 7.655731 (lr=3.7549e-04) (hash(x)=23175270)
427
+ 407 train 7.745759 (lr=3.7509e-04) (hash(x)=25716176)
428
+ 408 train 7.811182 (lr=3.7468e-04) (hash(x)=26861373)
429
+ 409 train 7.855147 (lr=3.7428e-04) (hash(x)=25118971)
430
+ 410 train 7.421111 (lr=3.7387e-04) (hash(x)=19829066)
431
+ 411 train 7.569578 (lr=3.7345e-04) (hash(x)=26256420)
432
+ 412 train 7.855487 (lr=3.7304e-04) (hash(x)=27796153)
433
+ 413 train 7.610954 (lr=3.7262e-04) (hash(x)=22633318)
434
+ 414 train 7.629806 (lr=3.7220e-04) (hash(x)=22589383)
435
+ 415 train 8.077329 (lr=3.7177e-04) (hash(x)=28019788)
436
+ 416 train 7.952090 (lr=3.7135e-04) (hash(x)=28970440)
437
+ 417 train 7.646776 (lr=3.7092e-04) (hash(x)=27396089)
438
+ 418 train 7.689483 (lr=3.7048e-04) (hash(x)=21183513)
439
+ 419 train 7.668569 (lr=3.7005e-04) (hash(x)=23510110)
440
+ 420 train 7.877487 (lr=3.6961e-04) (hash(x)=28833467)
441
+ 421 train 7.720227 (lr=3.6917e-04) (hash(x)=23646926)
442
+ 422 train 7.660770 (lr=3.6872e-04) (hash(x)=24697272)
443
+ 423 train 7.455293 (lr=3.6828e-04) (hash(x)=20382963)
444
+ 424 train 7.469532 (lr=3.6782e-04) (hash(x)=23467595)
445
+ 425 train 7.679106 (lr=3.6737e-04) (hash(x)=24304768)
446
+ 426 train 7.589017 (lr=3.6692e-04) (hash(x)=21392328)
447
+ 427 train 7.641454 (lr=3.6646e-04) (hash(x)=25339466)
448
+ 428 train 7.565831 (lr=3.6599e-04) (hash(x)=22092542)
449
+ 429 train 7.667256 (lr=3.6553e-04) (hash(x)=22088696)
450
+ 430 train 7.573459 (lr=3.6506e-04) (hash(x)=22184471)
451
+ 431 train 7.659507 (lr=3.6459e-04) (hash(x)=24489647)
452
+ 432 train 7.680092 (lr=3.6412e-04) (hash(x)=26794132)
453
+ 433 train 7.529958 (lr=3.6364e-04) (hash(x)=22940357)
454
+ 434 train 7.678494 (lr=3.6316e-04) (hash(x)=23719522)
455
+ 435 train 7.538326 (lr=3.6268e-04) (hash(x)=22927699)
456
+ 436 train 7.708934 (lr=3.6220e-04) (hash(x)=26068576)
457
+ 437 train 7.827714 (lr=3.6171e-04) (hash(x)=27631132)
458
+ 438 train 7.717595 (lr=3.6122e-04) (hash(x)=26739991)
459
+ 439 train 7.799654 (lr=3.6073e-04) (hash(x)=25128502)
460
+ 440 train 7.828781 (lr=3.6023e-04) (hash(x)=25657260)
461
+ 441 train 7.586043 (lr=3.5974e-04) (hash(x)=23576982)
462
+ 442 train 7.782579 (lr=3.5924e-04) (hash(x)=27117886)
463
+ 443 train 7.725891 (lr=3.5873e-04) (hash(x)=25808969)
464
+ 444 train 7.613228 (lr=3.5823e-04) (hash(x)=24738238)
465
+ 445 train 7.655702 (lr=3.5772e-04) (hash(x)=23429962)
466
+ 446 train 7.873582 (lr=3.5721e-04) (hash(x)=25075165)
467
+ 447 train 7.622217 (lr=3.5669e-04) (hash(x)=25231390)
468
+ 448 train 7.461713 (lr=3.5618e-04) (hash(x)=22055054)
469
+ 449 train 7.683003 (lr=3.5566e-04) (hash(x)=25395441)
470
+ 450 val loss 7.6865
471
+ 450 val perplexity 2178.7461
472
+ 450 train 7.733472 (lr=3.5514e-04) (hash(x)=25863209)
473
+ 451 train 7.400863 (lr=3.5461e-04) (hash(x)=21154388)
474
+ 452 train 7.370954 (lr=3.5408e-04) (hash(x)=21600876)
475
+ 453 train 7.508171 (lr=3.5355e-04) (hash(x)=24278611)
476
+ 454 train 7.380705 (lr=3.5302e-04) (hash(x)=23221720)
477
+ 455 train 7.594847 (lr=3.5249e-04) (hash(x)=22708977)
478
+ 456 train 8.140361 (lr=3.5195e-04) (hash(x)=23637758)
479
+ 457 train 7.927143 (lr=3.5141e-04) (hash(x)=28228490)
480
+ 458 train 7.943452 (lr=3.5087e-04) (hash(x)=28638071)
481
+ 459 train 7.833799 (lr=3.5032e-04) (hash(x)=27258353)
482
+ 460 train 7.820802 (lr=3.4977e-04) (hash(x)=26604728)
483
+ 461 train 7.667146 (lr=3.4922e-04) (hash(x)=23252199)
484
+ 462 train 7.766150 (lr=3.4867e-04) (hash(x)=26441427)
485
+ 463 train 7.480884 (lr=3.4812e-04) (hash(x)=24364920)
486
+ 464 train 7.630258 (lr=3.4756e-04) (hash(x)=25623792)
487
+ 465 train 7.541378 (lr=3.4700e-04) (hash(x)=23283905)
488
+ 466 train 8.007805 (lr=3.4644e-04) (hash(x)=26025267)
489
+ 467 train 7.767820 (lr=3.4587e-04) (hash(x)=27243972)
490
+ 468 train 8.122204 (lr=3.4530e-04) (hash(x)=30449945)
491
+ 469 train 7.768497 (lr=3.4473e-04) (hash(x)=28113043)
492
+ 470 train 7.641873 (lr=3.4416e-04) (hash(x)=25182521)
493
+ 471 train 7.713246 (lr=3.4359e-04) (hash(x)=24932925)
494
+ 472 train 7.536747 (lr=3.4301e-04) (hash(x)=20353098)
495
+ 473 train 7.433877 (lr=3.4243e-04) (hash(x)=19001259)
496
+ 474 train 8.001169 (lr=3.4185e-04) (hash(x)=27585685)
497
+ 475 train 7.670650 (lr=3.4127e-04) (hash(x)=26371091)
498
+ 476 train 7.475910 (lr=3.4068e-04) (hash(x)=24891798)
499
+ 477 train 7.507851 (lr=3.4009e-04) (hash(x)=24258817)
500
+ 478 train 7.571007 (lr=3.3950e-04) (hash(x)=24330263)
501
+ 479 train 7.699295 (lr=3.3891e-04) (hash(x)=26913684)
502
+ 480 train 7.652080 (lr=3.3831e-04) (hash(x)=26338455)
503
+ 481 train 7.958058 (lr=3.3771e-04) (hash(x)=27753043)
504
+ 482 train 7.678844 (lr=3.3711e-04) (hash(x)=26123289)
505
+ 483 train 7.772736 (lr=3.3651e-04) (hash(x)=29239611)
506
+ 484 train 7.680997 (lr=3.3590e-04) (hash(x)=26553003)
507
+ 485 train 7.650609 (lr=3.3530e-04) (hash(x)=22984557)
508
+ 486 train 7.160617 (lr=3.3469e-04) (hash(x)=16947491)
509
+ 487 train 7.191340 (lr=3.3408e-04) (hash(x)=18017792)
510
+ 488 train 7.321163 (lr=3.3346e-04) (hash(x)=19918608)
511
+ 489 train 7.771875 (lr=3.3285e-04) (hash(x)=23374526)
512
+ 490 train 7.628438 (lr=3.3223e-04) (hash(x)=25009505)
513
+ 491 train 7.644535 (lr=3.3161e-04) (hash(x)=27574089)
514
+ 492 train 7.747769 (lr=3.3099e-04) (hash(x)=24122664)
515
+ 493 train 7.648665 (lr=3.3036e-04) (hash(x)=26154906)
516
+ 494 train 7.569263 (lr=3.2973e-04) (hash(x)=25192767)
517
+ 495 train 7.713017 (lr=3.2911e-04) (hash(x)=28613882)
518
+ 496 train 7.629711 (lr=3.2847e-04) (hash(x)=23547219)
519
+ 497 train 7.627475 (lr=3.2784e-04) (hash(x)=25272182)
520
+ 498 train 7.468216 (lr=3.2721e-04) (hash(x)=24992761)
521
+ 499 train 7.762365 (lr=3.2657e-04) (hash(x)=26981914)
522
+ 500 val loss 7.6332
523
+ 500 val perplexity 2065.6125
524
+ 500 train 7.552813 (lr=3.2593e-04) (hash(x)=22051933)
525
+ 501 train 7.720116 (lr=3.2529e-04) (hash(x)=24232348)
526
+ 502 train 7.708518 (lr=3.2464e-04) (hash(x)=23158331)
527
+ 503 train 7.524069 (lr=3.2400e-04) (hash(x)=22652243)
528
+ 504 train 7.596242 (lr=3.2335e-04) (hash(x)=23805602)
529
+ 505 train 7.887978 (lr=3.2270e-04) (hash(x)=25411991)
530
+ 506 train 7.169883 (lr=3.2205e-04) (hash(x)=18827215)
531
+ 507 train 6.905998 (lr=3.2140e-04) (hash(x)=15446025)
532
+ 508 train 7.230747 (lr=3.2074e-04) (hash(x)=20516263)
533
+ 509 train 7.731800 (lr=3.2008e-04) (hash(x)=27846176)
534
+ 510 train 7.463365 (lr=3.1943e-04) (hash(x)=23342449)
535
+ 511 train 7.707950 (lr=3.1876e-04) (hash(x)=27194521)
536
+ 512 train 7.502069 (lr=3.1810e-04) (hash(x)=23008284)
537
+ 513 train 7.951359 (lr=3.1744e-04) (hash(x)=29430001)
538
+ 514 train 7.269432 (lr=3.1677e-04) (hash(x)=22579319)
539
+ 515 train 7.343285 (lr=3.1610e-04) (hash(x)=25264518)
540
+ 516 train 7.692428 (lr=3.1543e-04) (hash(x)=25359075)
541
+ 517 train 7.661928 (lr=3.1476e-04) (hash(x)=25568956)
542
+ 518 train 7.899338 (lr=3.1408e-04) (hash(x)=32004108)
543
+ 519 train 7.622131 (lr=3.1341e-04) (hash(x)=24936836)
544
+ 520 train 7.746236 (lr=3.1273e-04) (hash(x)=27263338)
545
+ 521 train 7.776847 (lr=3.1205e-04) (hash(x)=27452099)
546
+ 522 train 7.688045 (lr=3.1137e-04) (hash(x)=25965406)
547
+ 523 train 7.863934 (lr=3.1069e-04) (hash(x)=28197282)
548
+ 524 train 7.582237 (lr=3.1000e-04) (hash(x)=22466209)
549
+ 525 train 7.638658 (lr=3.0931e-04) (hash(x)=22931889)
550
+ 526 train 7.762384 (lr=3.0862e-04) (hash(x)=26903920)
551
+ 527 train 7.614944 (lr=3.0793e-04) (hash(x)=24765578)
552
+ 528 train 7.742456 (lr=3.0724e-04) (hash(x)=27811359)
553
+ 529 train 7.539230 (lr=3.0655e-04) (hash(x)=25078649)
554
+ 530 train 7.756212 (lr=3.0585e-04) (hash(x)=25572416)
555
+ 531 train 7.852325 (lr=3.0516e-04) (hash(x)=27448185)
556
+ 532 train 7.905980 (lr=3.0446e-04) (hash(x)=25923719)
557
+ 533 train 7.666679 (lr=3.0376e-04) (hash(x)=24804856)
558
+ 534 train 7.778291 (lr=3.0306e-04) (hash(x)=23207829)
559
+ 535 train 7.676983 (lr=3.0235e-04) (hash(x)=23107416)
560
+ 536 train 7.570381 (lr=3.0165e-04) (hash(x)=26739531)
561
+ 537 train 7.563313 (lr=3.0094e-04) (hash(x)=24960796)
562
+ 538 train 7.657387 (lr=3.0023e-04) (hash(x)=24667802)
563
+ 539 train 7.797616 (lr=2.9952e-04) (hash(x)=26755138)
564
+ 540 train 7.655032 (lr=2.9881e-04) (hash(x)=25537132)
565
+ 541 train 7.583833 (lr=2.9810e-04) (hash(x)=24542526)
566
+ 542 train 7.389100 (lr=2.9738e-04) (hash(x)=21296355)
567
+ 543 train 7.889041 (lr=2.9667e-04) (hash(x)=29314255)
568
+ 544 train 7.618531 (lr=2.9595e-04) (hash(x)=26001799)
569
+ 545 train 7.677786 (lr=2.9523e-04) (hash(x)=27347755)
570
+ 546 train 7.670935 (lr=2.9451e-04) (hash(x)=25107798)
571
+ 547 train 7.477783 (lr=2.9379e-04) (hash(x)=22112669)
572
+ 548 train 7.449146 (lr=2.9307e-04) (hash(x)=21897967)
573
+ 549 train 7.563628 (lr=2.9234e-04) (hash(x)=25161929)
574
+ 550 val loss 7.6074
575
+ 550 val perplexity 2012.9604
576
+ 550 train 7.482878 (lr=2.9162e-04) (hash(x)=27465106)
577
+ 551 train 7.617856 (lr=2.9089e-04) (hash(x)=24013079)
578
+ 552 train 7.533206 (lr=2.9016e-04) (hash(x)=23142015)
579
+ 553 train 7.614569 (lr=2.8943e-04) (hash(x)=26768629)
580
+ 554 train 7.650775 (lr=2.8870e-04) (hash(x)=26393383)
581
+ 555 train 7.420618 (lr=2.8797e-04) (hash(x)=22537194)
582
+ 556 train 7.702990 (lr=2.8723e-04) (hash(x)=24046036)
583
+ 557 train 7.661326 (lr=2.8650e-04) (hash(x)=24974360)
584
+ 558 train 7.928740 (lr=2.8576e-04) (hash(x)=28379928)
585
+ 559 train 7.642494 (lr=2.8502e-04) (hash(x)=25322001)
586
+ 560 train 7.759382 (lr=2.8428e-04) (hash(x)=26622031)
587
+ 561 train 7.457041 (lr=2.8354e-04) (hash(x)=20562247)
588
+ 562 train 7.800043 (lr=2.8280e-04) (hash(x)=27381885)
589
+ 563 train 7.926505 (lr=2.8206e-04) (hash(x)=27028126)
590
+ 564 train 7.738486 (lr=2.8132e-04) (hash(x)=28882928)
591
+ 565 train 7.641026 (lr=2.8057e-04) (hash(x)=25666355)
592
+ 566 train 7.742987 (lr=2.7982e-04) (hash(x)=24330810)
593
+ 567 train 7.672560 (lr=2.7908e-04) (hash(x)=26690440)
594
+ 568 train 7.522315 (lr=2.7833e-04) (hash(x)=22923592)
595
+ 569 train 7.640737 (lr=2.7758e-04) (hash(x)=27348418)
596
+ 570 train 7.679882 (lr=2.7683e-04) (hash(x)=28849848)
597
+ 571 train 7.733823 (lr=2.7607e-04) (hash(x)=26967331)
598
+ 572 train 7.468964 (lr=2.7532e-04) (hash(x)=22831467)
599
+ 573 train 7.602956 (lr=2.7457e-04) (hash(x)=24765121)
600
+ 574 train 7.601668 (lr=2.7381e-04) (hash(x)=24331857)
601
+ 575 train 7.459981 (lr=2.7306e-04) (hash(x)=22598512)
602
+ 576 train 7.616667 (lr=2.7230e-04) (hash(x)=25149353)
603
+ 577 train 7.463432 (lr=2.7154e-04) (hash(x)=23725598)
604
+ 578 train 7.664311 (lr=2.7078e-04) (hash(x)=26449557)
605
+ 579 train 7.699889 (lr=2.7002e-04) (hash(x)=24697985)
606
+ 580 train 7.714749 (lr=2.6926e-04) (hash(x)=26923059)
607
+ 581 train 7.490506 (lr=2.6850e-04) (hash(x)=25201962)
608
+ 582 train 7.147884 (lr=2.6773e-04) (hash(x)=20931520)
609
+ 583 train 7.239325 (lr=2.6697e-04) (hash(x)=18473911)
610
+ 584 train 7.415911 (lr=2.6620e-04) (hash(x)=21306267)
611
+ 585 train 7.647130 (lr=2.6544e-04) (hash(x)=25982840)
612
+ 586 train 7.526642 (lr=2.6467e-04) (hash(x)=25364874)
613
+ 587 train 7.426514 (lr=2.6390e-04) (hash(x)=23172124)
614
+ 588 train 7.753421 (lr=2.6314e-04) (hash(x)=27876897)
615
+ 589 train 8.374115 (lr=2.6237e-04) (hash(x)=34646114)
616
+ 590 train 8.658839 (lr=2.6160e-04) (hash(x)=35153576)
617
+ 591 train 7.548964 (lr=2.6083e-04) (hash(x)=22322442)
618
+ 592 train 7.779138 (lr=2.6005e-04) (hash(x)=27907331)
619
+ 593 train 7.685335 (lr=2.5928e-04) (hash(x)=26211794)
620
+ 594 train 7.863433 (lr=2.5851e-04) (hash(x)=29291512)
621
+ 595 train 8.115520 (lr=2.5773e-04) (hash(x)=29659121)
622
+ 596 train 7.971896 (lr=2.5696e-04) (hash(x)=29674399)
623
+ 597 train 7.640299 (lr=2.5618e-04) (hash(x)=23538306)
624
+ 598 train 7.740874 (lr=2.5541e-04) (hash(x)=21991524)
625
+ 599 train 7.670537 (lr=2.5463e-04) (hash(x)=26324153)
626
+ 600 val loss 7.6318
627
+ 600 val perplexity 2062.6853
628
+ 600 train 7.635567 (lr=2.5385e-04) (hash(x)=23712082)
629
+ 601 train 7.690489 (lr=2.5307e-04) (hash(x)=24910403)
630
+ 602 train 7.612943 (lr=2.5230e-04) (hash(x)=26737205)
631
+ 603 train 7.738811 (lr=2.5152e-04) (hash(x)=26939970)
632
+ 604 train 7.610899 (lr=2.5074e-04) (hash(x)=27651943)
633
+ 605 train 7.906340 (lr=2.4996e-04) (hash(x)=27515446)
634
+ 606 train 7.791934 (lr=2.4917e-04) (hash(x)=26753129)
635
+ 607 train 7.473592 (lr=2.4839e-04) (hash(x)=23446058)
636
+ 608 train 7.793958 (lr=2.4761e-04) (hash(x)=27587849)
637
+ 609 train 7.673238 (lr=2.4683e-04) (hash(x)=25308253)
638
+ 610 train 7.665001 (lr=2.4604e-04) (hash(x)=26615098)
639
+ 611 train 7.721785 (lr=2.4526e-04) (hash(x)=29981801)
640
+ 612 train 7.790768 (lr=2.4448e-04) (hash(x)=29592345)
641
+ 613 train 7.813720 (lr=2.4369e-04) (hash(x)=23470413)
642
+ 614 train 7.725273 (lr=2.4291e-04) (hash(x)=24742370)
643
+ 615 train 7.705728 (lr=2.4212e-04) (hash(x)=24843741)
644
+ 616 train 7.591117 (lr=2.4133e-04) (hash(x)=25192548)
645
+ 617 train 7.704877 (lr=2.4055e-04) (hash(x)=27176996)
646
+ 618 train 7.530110 (lr=2.3976e-04) (hash(x)=23964552)
647
+ 619 train 7.521988 (lr=2.3897e-04) (hash(x)=22855363)
648
+ 620 train 7.482961 (lr=2.3818e-04) (hash(x)=26332996)
649
+ 621 train 7.292532 (lr=2.3740e-04) (hash(x)=22960957)
650
+ 622 train 7.652774 (lr=2.3661e-04) (hash(x)=22752597)
651
+ 623 train 7.524491 (lr=2.3582e-04) (hash(x)=23197102)
652
+ 624 train 8.018651 (lr=2.3503e-04) (hash(x)=27383319)
653
+ 625 train 7.603845 (lr=2.3424e-04) (hash(x)=26132276)
654
+ 626 train 7.502547 (lr=2.3345e-04) (hash(x)=19810497)
655
+ 627 train 7.574701 (lr=2.3266e-04) (hash(x)=25704919)
656
+ 628 train 7.746508 (lr=2.3187e-04) (hash(x)=27174264)
657
+ 629 train 7.846406 (lr=2.3108e-04) (hash(x)=22280814)
658
+ 630 train 8.588954 (lr=2.3029e-04) (hash(x)=24279448)
659
+ 631 train 8.050984 (lr=2.2950e-04) (hash(x)=23054940)
660
+ 632 train 8.167277 (lr=2.2871e-04) (hash(x)=23942400)
661
+ 633 train 8.409505 (lr=2.2792e-04) (hash(x)=24712416)
662
+ 634 train 8.008598 (lr=2.2713e-04) (hash(x)=24564658)
663
+ 635 train 7.684293 (lr=2.2633e-04) (hash(x)=24909904)
664
+ 636 train 7.805206 (lr=2.2554e-04) (hash(x)=23049534)
665
+ 637 train 7.833169 (lr=2.2475e-04) (hash(x)=24321591)
666
+ 638 train 7.628246 (lr=2.2396e-04) (hash(x)=26153298)
667
+ 639 train 7.606736 (lr=2.2317e-04) (hash(x)=27140757)
668
+ 640 train 7.378213 (lr=2.2238e-04) (hash(x)=25115907)
669
+ 641 train 7.361232 (lr=2.2158e-04) (hash(x)=26563770)
670
+ 642 train 7.611556 (lr=2.2079e-04) (hash(x)=28089252)
671
+ 643 train 7.471763 (lr=2.2000e-04) (hash(x)=25993110)
672
+ 644 train 7.484096 (lr=2.1921e-04) (hash(x)=25150008)
673
+ 645 train 7.747834 (lr=2.1842e-04) (hash(x)=28520222)
674
+ 646 train 7.399921 (lr=2.1762e-04) (hash(x)=21349943)
675
+ 647 train 7.451051 (lr=2.1683e-04) (hash(x)=25149419)
676
+ 648 train 7.687094 (lr=2.1604e-04) (hash(x)=25730641)
677
+ 649 train 7.807935 (lr=2.1525e-04) (hash(x)=26112813)
678
+ 650 val loss 7.5658
679
+ 650 val perplexity 1930.9489
680
+ 650 train 7.606884 (lr=2.1446e-04) (hash(x)=25907805)
681
+ 651 train 7.638460 (lr=2.1367e-04) (hash(x)=27623643)
682
+ 652 train 7.600853 (lr=2.1287e-04) (hash(x)=26484959)
683
+ 653 train 7.646362 (lr=2.1208e-04) (hash(x)=29199854)
684
+ 654 train 7.660801 (lr=2.1129e-04) (hash(x)=28369628)
685
+ 655 train 7.521089 (lr=2.1050e-04) (hash(x)=24727764)
686
+ 656 train 7.274818 (lr=2.0971e-04) (hash(x)=22610673)
687
+ 657 train 7.199239 (lr=2.0892e-04) (hash(x)=22667179)
688
+ 658 train 7.021843 (lr=2.0813e-04) (hash(x)=18477300)
689
+ 659 train 7.381120 (lr=2.0734e-04) (hash(x)=23155773)
690
+ 660 train 7.180364 (lr=2.0655e-04) (hash(x)=19461032)
691
+ 661 train 7.447155 (lr=2.0576e-04) (hash(x)=23453788)
692
+ 662 train 7.589633 (lr=2.0497e-04) (hash(x)=24543466)
693
+ 663 train 7.370703 (lr=2.0418e-04) (hash(x)=21935931)
694
+ 664 train 7.216922 (lr=2.0339e-04) (hash(x)=19910292)
695
+ 665 train 7.509130 (lr=2.0260e-04) (hash(x)=24481079)
696
+ 666 train 7.194754 (lr=2.0182e-04) (hash(x)=18922411)
697
+ 667 train 7.217264 (lr=2.0103e-04) (hash(x)=20054917)
698
+ 668 train 7.460618 (lr=2.0024e-04) (hash(x)=24850470)
699
+ 669 train 7.479710 (lr=1.9945e-04) (hash(x)=25907741)
700
+ 670 train 7.629954 (lr=1.9867e-04) (hash(x)=26873522)
701
+ 671 train 7.661283 (lr=1.9788e-04) (hash(x)=27606073)
702
+ 672 train 7.786950 (lr=1.9709e-04) (hash(x)=26209645)
703
+ 673 train 7.601289 (lr=1.9631e-04) (hash(x)=25202001)
704
+ 674 train 7.660721 (lr=1.9552e-04) (hash(x)=25569462)
705
+ 675 train 7.631531 (lr=1.9474e-04) (hash(x)=26534487)
706
+ 676 train 7.635434 (lr=1.9396e-04) (hash(x)=26455057)
707
+ 677 train 7.517467 (lr=1.9317e-04) (hash(x)=24095850)
708
+ 678 train 7.732952 (lr=1.9239e-04) (hash(x)=25287752)
709
+ 679 train 7.239507 (lr=1.9161e-04) (hash(x)=22450341)
710
+ 680 train 7.960056 (lr=1.9083e-04) (hash(x)=29004853)
711
+ 681 train 7.638196 (lr=1.9004e-04) (hash(x)=27993763)
712
+ 682 train 7.478987 (lr=1.8926e-04) (hash(x)=26382658)
713
+ 683 train 7.607319 (lr=1.8848e-04) (hash(x)=25013073)
714
+ 684 train 7.624572 (lr=1.8770e-04) (hash(x)=30595809)
715
+ 685 train 7.715837 (lr=1.8693e-04) (hash(x)=30934371)
716
+ 686 train 8.308474 (lr=1.8615e-04) (hash(x)=33060834)
717
+ 687 train 7.586534 (lr=1.8537e-04) (hash(x)=25945859)
718
+ 688 train 7.430475 (lr=1.8459e-04) (hash(x)=23375678)
719
+ 689 train 7.474863 (lr=1.8382e-04) (hash(x)=25218689)
720
+ 690 train 7.599796 (lr=1.8304e-04) (hash(x)=28127397)
721
+ 691 train 7.497808 (lr=1.8227e-04) (hash(x)=24418091)
722
+ 692 train 7.386284 (lr=1.8149e-04) (hash(x)=22761099)
723
+ 693 train 7.410190 (lr=1.8072e-04) (hash(x)=24615466)
724
+ 694 train 7.580059 (lr=1.7995e-04) (hash(x)=24699240)
725
+ 695 train 7.470056 (lr=1.7917e-04) (hash(x)=21613707)
726
+ 696 train 7.544530 (lr=1.7840e-04) (hash(x)=24977554)
727
+ 697 train 7.487504 (lr=1.7763e-04) (hash(x)=24348175)
728
+ 698 train 7.692122 (lr=1.7686e-04) (hash(x)=25102767)
729
+ 699 train 7.620491 (lr=1.7610e-04) (hash(x)=26386157)
730
+ 700 val loss 7.5246
731
+ 700 val perplexity 1853.1364
732
+ 700 train 7.543725 (lr=1.7533e-04) (hash(x)=26423460)
733
+ 701 train 7.480064 (lr=1.7456e-04) (hash(x)=25771047)
734
+ 702 train 7.555481 (lr=1.7380e-04) (hash(x)=26999875)
735
+ 703 train 7.490953 (lr=1.7303e-04) (hash(x)=24396519)
736
+ 704 train 7.495464 (lr=1.7227e-04) (hash(x)=22588122)
737
+ 705 train 7.827551 (lr=1.7150e-04) (hash(x)=25142399)
738
+ 706 train 7.535623 (lr=1.7074e-04) (hash(x)=20440214)
739
+ 707 train 7.407912 (lr=1.6998e-04) (hash(x)=23265507)
740
+ 708 train 7.518503 (lr=1.6922e-04) (hash(x)=24563470)
741
+ 709 train 7.430501 (lr=1.6846e-04) (hash(x)=22514858)
742
+ 710 train 7.748348 (lr=1.6770e-04) (hash(x)=26691212)
743
+ 711 train 7.779922 (lr=1.6694e-04) (hash(x)=29138828)
744
+ 712 train 7.761325 (lr=1.6619e-04) (hash(x)=28028528)
745
+ 713 train 7.682018 (lr=1.6543e-04) (hash(x)=20531210)
746
+ 714 train 7.491971 (lr=1.6468e-04) (hash(x)=25075352)
747
+ 715 train 7.532562 (lr=1.6393e-04) (hash(x)=24265353)
748
+ 716 train 7.642853 (lr=1.6317e-04) (hash(x)=24635726)
749
+ 717 train 7.501414 (lr=1.6242e-04) (hash(x)=24999726)
750
+ 718 train 7.680169 (lr=1.6167e-04) (hash(x)=27412910)
751
+ 719 train 7.471292 (lr=1.6092e-04) (hash(x)=24685515)
752
+ 720 train 7.557440 (lr=1.6018e-04) (hash(x)=23780329)
753
+ 721 train 7.384611 (lr=1.5943e-04) (hash(x)=25071701)
754
+ 722 train 7.523477 (lr=1.5868e-04) (hash(x)=23767130)
755
+ 723 train 7.475683 (lr=1.5794e-04) (hash(x)=24876269)
756
+ 724 train 7.613349 (lr=1.5720e-04) (hash(x)=26405773)
757
+ 725 train 7.997049 (lr=1.5646e-04) (hash(x)=31733180)
758
+ 726 train 7.183338 (lr=1.5572e-04) (hash(x)=21337509)
759
+ 727 train 7.313069 (lr=1.5498e-04) (hash(x)=22825749)
760
+ 728 train 7.698401 (lr=1.5424e-04) (hash(x)=28638695)
761
+ 729 train 7.562607 (lr=1.5350e-04) (hash(x)=26393943)
762
+ 730 train 7.731818 (lr=1.5277e-04) (hash(x)=27563583)
763
+ 731 train 7.258891 (lr=1.5203e-04) (hash(x)=21239652)
764
+ 732 train 7.382724 (lr=1.5130e-04) (hash(x)=23986428)
765
+ 733 train 7.364917 (lr=1.5057e-04) (hash(x)=24943881)
766
+ 734 train 7.895998 (lr=1.4984e-04) (hash(x)=29691448)
767
+ 735 train 7.816323 (lr=1.4911e-04) (hash(x)=28767869)
768
+ 736 train 7.495233 (lr=1.4838e-04) (hash(x)=23628188)
769
+ 737 train 7.791477 (lr=1.4766e-04) (hash(x)=29341482)
770
+ 738 train 7.649465 (lr=1.4693e-04) (hash(x)=30336570)
771
+ 739 train 7.582391 (lr=1.4621e-04) (hash(x)=25614301)
772
+ 740 train 7.445729 (lr=1.4549e-04) (hash(x)=24160500)
773
+ 741 train 7.624176 (lr=1.4477e-04) (hash(x)=26030058)
774
+ 742 train 7.984622 (lr=1.4405e-04) (hash(x)=29243936)
775
+ 743 train 7.305840 (lr=1.4333e-04) (hash(x)=21159060)
776
+ 744 train 7.427970 (lr=1.4262e-04) (hash(x)=23701853)
777
+ 745 train 7.605230 (lr=1.4190e-04) (hash(x)=24629937)
778
+ 746 train 7.512506 (lr=1.4119e-04) (hash(x)=25110108)
779
+ 747 train 7.578247 (lr=1.4048e-04) (hash(x)=26751788)
780
+ 748 train 7.615915 (lr=1.3977e-04) (hash(x)=26430427)
781
+ 749 train 7.473232 (lr=1.3906e-04) (hash(x)=26012353)
782
+ 750 val loss 7.5066
783
+ 750 val perplexity 1820.0834
784
+ 750 train 7.580135 (lr=1.3835e-04) (hash(x)=22735910)
785
+ 751 train 7.406583 (lr=1.3765e-04) (hash(x)=25045397)
786
+ 752 train 7.106474 (lr=1.3694e-04) (hash(x)=21554427)
787
+ 753 train 7.224483 (lr=1.3624e-04) (hash(x)=23751143)
788
+ 754 train 7.987680 (lr=1.3554e-04) (hash(x)=28602273)
789
+ 755 train 8.035553 (lr=1.3484e-04) (hash(x)=29989709)
790
+ 756 train 7.432967 (lr=1.3415e-04) (hash(x)=22331648)
791
+ 757 train 7.845795 (lr=1.3345e-04) (hash(x)=31017246)
792
+ 758 train 7.792151 (lr=1.3276e-04) (hash(x)=29709045)
793
+ 759 train 7.839722 (lr=1.3207e-04) (hash(x)=25560928)
794
+ 760 train 7.504821 (lr=1.3138e-04) (hash(x)=25075464)
795
+ 761 train 7.616666 (lr=1.3069e-04) (hash(x)=27352253)
796
+ 762 train 7.819825 (lr=1.3000e-04) (hash(x)=28187891)
797
+ 763 train 7.661468 (lr=1.2931e-04) (hash(x)=26062687)
798
+ 764 train 7.653458 (lr=1.2863e-04) (hash(x)=27427811)
799
+ 765 train 7.739742 (lr=1.2795e-04) (hash(x)=27614522)
800
+ 766 train 7.502387 (lr=1.2727e-04) (hash(x)=26129544)
801
+ 767 train 8.022285 (lr=1.2659e-04) (hash(x)=28959222)
802
+ 768 train 7.644228 (lr=1.2592e-04) (hash(x)=26860067)
803
+ 769 train 7.393959 (lr=1.2524e-04) (hash(x)=25122598)
804
+ 770 train 7.465302 (lr=1.2457e-04) (hash(x)=25245030)
805
+ 771 train 7.480988 (lr=1.2390e-04) (hash(x)=25434884)
806
+ 772 train 7.569217 (lr=1.2323e-04) (hash(x)=27732790)
807
+ 773 train 7.580595 (lr=1.2256e-04) (hash(x)=27824438)
808
+ 774 train 7.708383 (lr=1.2190e-04) (hash(x)=27201953)
809
+ 775 train 7.435576 (lr=1.2124e-04) (hash(x)=21248405)
810
+ 776 train 7.218540 (lr=1.2057e-04) (hash(x)=22805934)
811
+ 777 train 7.516550 (lr=1.1992e-04) (hash(x)=26482588)
812
+ 778 train 7.531166 (lr=1.1926e-04) (hash(x)=24153691)
813
+ 779 train 7.662698 (lr=1.1860e-04) (hash(x)=25044192)
814
+ 780 train 7.615449 (lr=1.1795e-04) (hash(x)=25910078)
815
+ 781 train 7.603136 (lr=1.1730e-04) (hash(x)=28645524)
816
+ 782 train 7.444355 (lr=1.1665e-04) (hash(x)=24368498)
817
+ 783 train 7.523260 (lr=1.1600e-04) (hash(x)=25830182)
818
+ 784 train 7.632308 (lr=1.1536e-04) (hash(x)=29181807)
819
+ 785 train 7.542404 (lr=1.1471e-04) (hash(x)=25585137)
820
+ 786 train 7.475906 (lr=1.1407e-04) (hash(x)=24798246)
821
+ 787 train 7.553417 (lr=1.1343e-04) (hash(x)=26621419)
822
+ 788 train 7.155941 (lr=1.1279e-04) (hash(x)=21446891)
823
+ 789 train 7.117670 (lr=1.1216e-04) (hash(x)=22165286)
824
+ 790 train 7.304485 (lr=1.1153e-04) (hash(x)=23477219)
825
+ 791 train 7.597849 (lr=1.1089e-04) (hash(x)=25173113)
826
+ 792 train 7.474068 (lr=1.1027e-04) (hash(x)=25853788)
827
+ 793 train 7.565894 (lr=1.0964e-04) (hash(x)=27267091)
828
+ 794 train 7.419462 (lr=1.0901e-04) (hash(x)=23743694)
829
+ 795 train 7.403450 (lr=1.0839e-04) (hash(x)=24400133)
830
+ 796 train 7.482179 (lr=1.0777e-04) (hash(x)=23663639)
831
+ 797 train 7.364186 (lr=1.0715e-04) (hash(x)=23103223)
832
+ 798 train 7.858798 (lr=1.0654e-04) (hash(x)=28748411)
833
+ 799 train 7.178535 (lr=1.0592e-04) (hash(x)=23486277)
834
+ 800 val loss 7.4713
835
+ 800 val perplexity 1756.9087
836
+ 800 train 7.427179 (lr=1.0531e-04) (hash(x)=25678518)
837
+ 801 train 7.394596 (lr=1.0470e-04) (hash(x)=23421286)
838
+ 802 train 7.532321 (lr=1.0410e-04) (hash(x)=26054104)
839
+ 803 train 7.542554 (lr=1.0349e-04) (hash(x)=25978130)
840
+ 804 train 7.575222 (lr=1.0289e-04) (hash(x)=26006525)
841
+ 805 train 7.505678 (lr=1.0229e-04) (hash(x)=25769432)
842
+ 806 train 7.323430 (lr=1.0169e-04) (hash(x)=22430795)
843
+ 807 train 7.711505 (lr=1.0109e-04) (hash(x)=28916006)
844
+ 808 train 7.501318 (lr=1.0050e-04) (hash(x)=25166800)
845
+ 809 train 7.617902 (lr=9.9910e-05) (hash(x)=24226056)
846
+ 810 train 7.459242 (lr=9.9321e-05) (hash(x)=25057934)
847
+ 811 train 8.260686 (lr=9.8735e-05) (hash(x)=33195100)
848
+ 812 train 7.630991 (lr=9.8151e-05) (hash(x)=26312888)
849
+ 813 train 7.792807 (lr=9.7569e-05) (hash(x)=27730410)
850
+ 814 train 7.666480 (lr=9.6990e-05) (hash(x)=27372474)
851
+ 815 train 7.497070 (lr=9.6413e-05) (hash(x)=25556929)
852
+ 816 train 7.600433 (lr=9.5838e-05) (hash(x)=26909985)
853
+ 817 train 7.441005 (lr=9.5266e-05) (hash(x)=25991247)
854
+ 818 train 7.606810 (lr=9.4696e-05) (hash(x)=27438141)
855
+ 819 train 7.754126 (lr=9.4129e-05) (hash(x)=29536986)
856
+ 820 train 7.493300 (lr=9.3564e-05) (hash(x)=24478391)
857
+ 821 train 7.464437 (lr=9.3001e-05) (hash(x)=26125216)
858
+ 822 train 7.499551 (lr=9.2441e-05) (hash(x)=26422130)
859
+ 823 train 7.689269 (lr=9.1884e-05) (hash(x)=29648798)
860
+ 824 train 7.373250 (lr=9.1328e-05) (hash(x)=21247770)
861
+ 825 train 7.509620 (lr=9.0776e-05) (hash(x)=23195388)
862
+ 826 train 7.449591 (lr=9.0226e-05) (hash(x)=25796725)
863
+ 827 train 7.513509 (lr=8.9678e-05) (hash(x)=23124767)
864
+ 828 train 7.521177 (lr=8.9133e-05) (hash(x)=25233464)
865
+ 829 train 7.514431 (lr=8.8591e-05) (hash(x)=25713275)
866
+ 830 train 7.462730 (lr=8.8051e-05) (hash(x)=25550167)
867
+ 831 train 7.367322 (lr=8.7513e-05) (hash(x)=24976217)
868
+ 832 train 7.675774 (lr=8.6978e-05) (hash(x)=28536827)
869
+ 833 train 7.570114 (lr=8.6446e-05) (hash(x)=27500801)
870
+ 834 train 7.392635 (lr=8.5916e-05) (hash(x)=25545765)
871
+ 835 train 7.394947 (lr=8.5389e-05) (hash(x)=23632825)
872
+ 836 train 7.436456 (lr=8.4865e-05) (hash(x)=25708009)
873
+ 837 train 7.495131 (lr=8.4343e-05) (hash(x)=24456276)
874
+ 838 train 7.611032 (lr=8.3824e-05) (hash(x)=29189855)
875
+ 839 train 7.839775 (lr=8.3307e-05) (hash(x)=31019606)
876
+ 840 train 7.711384 (lr=8.2793e-05) (hash(x)=26328013)
877
+ 841 train 7.619121 (lr=8.2282e-05) (hash(x)=25027904)
878
+ 842 train 7.417558 (lr=8.1773e-05) (hash(x)=23734189)
879
+ 843 train 7.645135 (lr=8.1267e-05) (hash(x)=28236580)
880
+ 844 train 7.446213 (lr=8.0764e-05) (hash(x)=26509780)
881
+ 845 train 7.523695 (lr=8.0263e-05) (hash(x)=25386473)
882
+ 846 train 7.477362 (lr=7.9765e-05) (hash(x)=24052671)
883
+ 847 train 7.678062 (lr=7.9270e-05) (hash(x)=28269421)
884
+ 848 train 7.121700 (lr=7.8778e-05) (hash(x)=22251724)
885
+ 849 train 7.359891 (lr=7.8288e-05) (hash(x)=24308447)
886
+ 850 val loss 7.4623
887
+ 850 val perplexity 1741.1320
888
+ 850 train 7.324823 (lr=7.7801e-05) (hash(x)=24242830)
889
+ 851 train 7.524683 (lr=7.7317e-05) (hash(x)=25563279)
890
+ 852 train 7.550127 (lr=7.6836e-05) (hash(x)=26354481)
891
+ 853 train 7.668492 (lr=7.6357e-05) (hash(x)=26152637)
892
+ 854 train 7.713246 (lr=7.5881e-05) (hash(x)=28051025)
893
+ 855 train 7.527201 (lr=7.5408e-05) (hash(x)=24865358)
894
+ 856 train 7.390286 (lr=7.4938e-05) (hash(x)=24288911)
895
+ 857 train 7.257581 (lr=7.4470e-05) (hash(x)=22230964)
896
+ 858 train 7.191998 (lr=7.4005e-05) (hash(x)=21303832)
897
+ 859 train 7.339108 (lr=7.3544e-05) (hash(x)=22155546)
898
+ 860 train 7.463250 (lr=7.3085e-05) (hash(x)=25296428)
899
+ 861 train 7.611850 (lr=7.2628e-05) (hash(x)=29142319)
900
+ 862 train 7.468327 (lr=7.2175e-05) (hash(x)=25545430)
901
+ 863 train 7.264093 (lr=7.1725e-05) (hash(x)=26984272)
902
+ 864 train 7.451824 (lr=7.1277e-05) (hash(x)=25429005)
903
+ 865 train 7.520978 (lr=7.0832e-05) (hash(x)=27077032)
904
+ 866 train 7.526020 (lr=7.0390e-05) (hash(x)=26494424)
905
+ 867 train 7.552418 (lr=6.9952e-05) (hash(x)=23193673)
906
+ 868 train 7.521654 (lr=6.9516e-05) (hash(x)=25075134)
907
+ 869 train 7.552069 (lr=6.9082e-05) (hash(x)=27112558)
908
+ 870 train 7.709271 (lr=6.8652e-05) (hash(x)=27436608)
909
+ 871 train 7.417000 (lr=6.8225e-05) (hash(x)=24544116)
910
+ 872 train 7.785616 (lr=6.7801e-05) (hash(x)=31632686)
911
+ 873 train 7.438150 (lr=6.7379e-05) (hash(x)=25890184)
912
+ 874 train 7.572937 (lr=6.6961e-05) (hash(x)=22887555)
913
+ 875 train 7.402617 (lr=6.6545e-05) (hash(x)=24547533)
914
+ 876 train 7.542262 (lr=6.6133e-05) (hash(x)=26553496)
915
+ 877 train 7.633232 (lr=6.5723e-05) (hash(x)=27467688)
916
+ 878 train 7.399890 (lr=6.5317e-05) (hash(x)=24766934)
917
+ 879 train 7.308873 (lr=6.4913e-05) (hash(x)=22059850)
918
+ 880 train 7.652802 (lr=6.4513e-05) (hash(x)=22871702)
919
+ 881 train 7.507041 (lr=6.4115e-05) (hash(x)=23893130)
920
+ 882 train 7.527823 (lr=6.3721e-05) (hash(x)=25125691)
921
+ 883 train 7.492677 (lr=6.3329e-05) (hash(x)=25994573)
922
+ 884 train 7.534267 (lr=6.2941e-05) (hash(x)=26076345)
923
+ 885 train 7.459936 (lr=6.2556e-05) (hash(x)=26577783)
924
+ 886 train 7.560666 (lr=6.2173e-05) (hash(x)=27395225)
925
+ 887 train 7.392650 (lr=6.1794e-05) (hash(x)=23926632)
926
+ 888 train 7.257697 (lr=6.1418e-05) (hash(x)=21737239)
927
+ 889 train 7.556491 (lr=6.1045e-05) (hash(x)=23574207)
928
+ 890 train 7.227996 (lr=6.0675e-05) (hash(x)=24365231)
929
+ 891 train 7.412708 (lr=6.0308e-05) (hash(x)=27111369)
930
+ 892 train 7.669170 (lr=5.9944e-05) (hash(x)=27290015)
931
+ 893 train 7.448936 (lr=5.9583e-05) (hash(x)=23979820)
932
+ 894 train 7.539602 (lr=5.9225e-05) (hash(x)=26450121)
933
+ 895 train 7.453219 (lr=5.8871e-05) (hash(x)=27025333)
934
+ 896 train 7.424850 (lr=5.8519e-05) (hash(x)=23624605)
935
+ 897 train 7.276289 (lr=5.8171e-05) (hash(x)=22846386)
936
+ 898 train 7.316283 (lr=5.7826e-05) (hash(x)=22970561)
937
+ 899 train 7.169788 (lr=5.7484e-05) (hash(x)=16908068)
938
+ 900 val loss 7.4430
939
+ 900 val perplexity 1707.7836
940
+ 900 train 7.380563 (lr=5.7145e-05) (hash(x)=24661446)
941
+ 901 train 7.535089 (lr=5.6809e-05) (hash(x)=25664727)
942
+ 902 train 7.496626 (lr=5.6476e-05) (hash(x)=25667011)
943
+ 903 train 7.558334 (lr=5.6147e-05) (hash(x)=29120407)
944
+ 904 train 7.619312 (lr=5.5821e-05) (hash(x)=23385735)
945
+ 905 train 7.658131 (lr=5.5497e-05) (hash(x)=25564213)
946
+ 906 train 7.669969 (lr=5.5178e-05) (hash(x)=25413898)
947
+ 907 train 7.626058 (lr=5.4861e-05) (hash(x)=27092710)
948
+ 908 train 7.429039 (lr=5.4547e-05) (hash(x)=25789923)
949
+ 909 train 7.532323 (lr=5.4237e-05) (hash(x)=28533197)
950
+ 910 train 7.360776 (lr=5.3930e-05) (hash(x)=22982996)
951
+ 911 train 7.461704 (lr=5.3626e-05) (hash(x)=23827393)
952
+ 912 train 7.048573 (lr=5.3325e-05) (hash(x)=21242640)
953
+ 913 train 7.465303 (lr=5.3028e-05) (hash(x)=24154233)
954
+ 914 train 7.468999 (lr=5.2734e-05) (hash(x)=24331967)
955
+ 915 train 7.712711 (lr=5.2443e-05) (hash(x)=32812727)
956
+ 916 train 7.366163 (lr=5.2155e-05) (hash(x)=23572994)
957
+ 917 train 7.410667 (lr=5.1871e-05) (hash(x)=26305435)
958
+ 918 train 7.441031 (lr=5.1589e-05) (hash(x)=26268355)
959
+ 919 train 7.627819 (lr=5.1311e-05) (hash(x)=27230027)
960
+ 920 train 7.399662 (lr=5.1037e-05) (hash(x)=23885377)
961
+ 921 train 7.359638 (lr=5.0765e-05) (hash(x)=23532437)
962
+ 922 train 7.437222 (lr=5.0497e-05) (hash(x)=25577034)
963
+ 923 train 7.585698 (lr=5.0232e-05) (hash(x)=25703381)
964
+ 924 train 7.568883 (lr=4.9971e-05) (hash(x)=27113866)
965
+ 925 train 7.409882 (lr=4.9712e-05) (hash(x)=26961429)
966
+ 926 train 7.213144 (lr=4.9457e-05) (hash(x)=21355372)
967
+ 927 train 7.361459 (lr=4.9206e-05) (hash(x)=24968260)
968
+ 928 train 7.340534 (lr=4.8957e-05) (hash(x)=25357517)
969
+ 929 train 7.432979 (lr=4.8712e-05) (hash(x)=24854265)
970
+ 930 train 7.228682 (lr=4.8470e-05) (hash(x)=21102770)
971
+ 931 train 7.519708 (lr=4.8232e-05) (hash(x)=25676468)
972
+ 932 train 7.398635 (lr=4.7997e-05) (hash(x)=22809869)
973
+ 933 train 7.513713 (lr=4.7765e-05) (hash(x)=25503865)
974
+ 934 train 7.381870 (lr=4.7537e-05) (hash(x)=24853995)
975
+ 935 train 7.580278 (lr=4.7312e-05) (hash(x)=27544803)
976
+ 936 train 7.614295 (lr=4.7090e-05) (hash(x)=25981933)
977
+ 937 train 7.307379 (lr=4.6871e-05) (hash(x)=24658683)
978
+ 938 train 7.527739 (lr=4.6656e-05) (hash(x)=23855201)
979
+ 939 train 7.433908 (lr=4.6445e-05) (hash(x)=24331407)
980
+ 940 train 7.641080 (lr=4.6236e-05) (hash(x)=29265551)
981
+ 941 train 7.324793 (lr=4.6031e-05) (hash(x)=21892556)
982
+ 942 train 7.339611 (lr=4.5830e-05) (hash(x)=27183405)
983
+ 943 train 7.310318 (lr=4.5631e-05) (hash(x)=26540663)
984
+ 944 train 7.273851 (lr=4.5437e-05) (hash(x)=25718393)
985
+ 945 train 7.509436 (lr=4.5245e-05) (hash(x)=26819462)
986
+ 946 train 7.484136 (lr=4.5057e-05) (hash(x)=27427540)
987
+ 947 train 7.458839 (lr=4.4872e-05) (hash(x)=25532657)
988
+ 948 train 7.666618 (lr=4.4691e-05) (hash(x)=27641372)
989
+ 949 train 7.439123 (lr=4.4513e-05) (hash(x)=26515570)
990
+ 950 val loss 7.4335
991
+ 950 val perplexity 1691.6606
992
+ 950 train 7.620336 (lr=4.4338e-05) (hash(x)=26911957)
993
+ 951 train 7.764505 (lr=4.4167e-05) (hash(x)=25856625)
994
+ 952 train 7.314861 (lr=4.4000e-05) (hash(x)=25219129)
995
+ 953 train 7.281584 (lr=4.3835e-05) (hash(x)=25260471)
996
+ 954 train 7.673806 (lr=4.3674e-05) (hash(x)=29373370)
997
+ 955 train 7.227283 (lr=4.3517e-05) (hash(x)=23437426)
998
+ 956 train 7.533296 (lr=4.3363e-05) (hash(x)=23769521)
999
+ 957 train 7.549512 (lr=4.3212e-05) (hash(x)=25961833)
1000
+ 958 train 7.280580 (lr=4.3065e-05) (hash(x)=23582666)
1001
+ 959 train 7.372135 (lr=4.2921e-05) (hash(x)=23164356)
1002
+ 960 train 7.335988 (lr=4.2781e-05) (hash(x)=24443114)
1003
+ 961 train 7.428780 (lr=4.2644e-05) (hash(x)=25052665)
1004
+ 962 train 7.527150 (lr=4.2510e-05) (hash(x)=27802272)
1005
+ 963 train 7.393819 (lr=4.2380e-05) (hash(x)=25957896)
1006
+ 964 train 7.411240 (lr=4.2253e-05) (hash(x)=26737251)
1007
+ 965 train 7.556139 (lr=4.2130e-05) (hash(x)=24723263)
1008
+ 966 train 7.500630 (lr=4.2010e-05) (hash(x)=24707011)
1009
+ 967 train 7.463414 (lr=4.1894e-05) (hash(x)=25646282)
1010
+ 968 train 7.518849 (lr=4.1781e-05) (hash(x)=27544665)
1011
+ 969 train 7.421739 (lr=4.1672e-05) (hash(x)=25851993)
1012
+ 970 train 7.724885 (lr=4.1566e-05) (hash(x)=29059700)
1013
+ 971 train 7.420932 (lr=4.1463e-05) (hash(x)=21513584)
1014
+ 972 train 7.192557 (lr=4.1364e-05) (hash(x)=23151267)
1015
+ 973 train 7.450660 (lr=4.1269e-05) (hash(x)=26017176)
1016
+ 974 train 7.537882 (lr=4.1177e-05) (hash(x)=26979518)
1017
+ 975 train 7.468009 (lr=4.1088e-05) (hash(x)=23843233)
1018
+ 976 train 7.650197 (lr=4.1003e-05) (hash(x)=24193010)
1019
+ 977 train 7.354966 (lr=4.0921e-05) (hash(x)=21476847)
1020
+ 978 train 7.254924 (lr=4.0843e-05) (hash(x)=21366504)
1021
+ 979 train 7.323423 (lr=4.0768e-05) (hash(x)=23226697)
1022
+ 980 train 7.145752 (lr=4.0697e-05) (hash(x)=19961773)
1023
+ 981 train 7.038109 (lr=4.0629e-05) (hash(x)=19772969)
1024
+ 982 train 7.315346 (lr=4.0564e-05) (hash(x)=23110142)
1025
+ 983 train 7.546760 (lr=4.0503e-05) (hash(x)=24506028)
1026
+ 984 train 7.554185 (lr=4.0446e-05) (hash(x)=25480731)
1027
+ 985 train 7.365884 (lr=4.0392e-05) (hash(x)=21077417)
1028
+ 986 train 7.351929 (lr=4.0341e-05) (hash(x)=23686713)
1029
+ 987 train 7.424734 (lr=4.0294e-05) (hash(x)=26024321)
1030
+ 988 train 7.535515 (lr=4.0251e-05) (hash(x)=27424109)
1031
+ 989 train 7.626081 (lr=4.0211e-05) (hash(x)=27786174)
1032
+ 990 train 7.419867 (lr=4.0174e-05) (hash(x)=25232502)
1033
+ 991 train 7.392072 (lr=4.0141e-05) (hash(x)=22781277)
1034
+ 992 train 7.467430 (lr=4.0112e-05) (hash(x)=26184527)
1035
+ 993 train 7.494178 (lr=4.0085e-05) (hash(x)=24459895)
1036
+ 994 train 7.457307 (lr=4.0063e-05) (hash(x)=25244624)
1037
+ 995 train 7.551075 (lr=4.0044e-05) (hash(x)=24451843)
1038
+ 996 train 7.546313 (lr=4.0028e-05) (hash(x)=22129897)
1039
+ 997 train 7.352875 (lr=4.0016e-05) (hash(x)=21116390)
1040
+ 998 train 7.159772 (lr=4.0007e-05) (hash(x)=20650070)
1041
+ 999 val loss 7.4237
1042
+ 999 val perplexity 1675.1754
1043
+ 999 train 7.530208 (lr=4.0002e-05) (hash(x)=24948650)
lr4e-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:aca1b6204bddf1a639e2316e05ab3ff82ebb9a6050a7ff11dc050cc48f784493
3
+ size 48156226
lr4e-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:a7954d66c7be52d5c1f802ea6c8c904f76102feb5547b2261ee7ac813b366bde
3
+ size 71157574