andrew-healey commited on
Commit
588dcfc
·
verified ·
1 Parent(s): 43766aa

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size5120_seq_len256/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size5120_seq_len256", "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": 256, "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_len256/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_len256/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7488
3
+ 0 val perplexity 126607.6172
4
+ 0 train 11.745071 (lr=3.4965e-07) (hash(x)=12372582)
5
+ 1 train 11.758301 (lr=6.9930e-07) (hash(x)=12683395)
6
+ 2 train 11.760255 (lr=1.0490e-06) (hash(x)=14107219)
7
+ 3 train 11.729419 (lr=1.3986e-06) (hash(x)=12136211)
8
+ 4 train 11.751030 (lr=1.7483e-06) (hash(x)=12407129)
9
+ 5 train 11.716742 (lr=2.0979e-06) (hash(x)=13068552)
10
+ 6 train 11.766962 (lr=2.4476e-06) (hash(x)=17450022)
11
+ 7 train 11.743189 (lr=2.7972e-06) (hash(x)=9915468)
12
+ 8 train 11.741367 (lr=3.1469e-06) (hash(x)=13548512)
13
+ 9 train 11.741474 (lr=3.4965e-06) (hash(x)=12164183)
14
+ 10 train 11.725018 (lr=3.8462e-06) (hash(x)=13803353)
15
+ 11 train 11.722281 (lr=4.1958e-06) (hash(x)=13629277)
16
+ 12 train 11.714296 (lr=4.5455e-06) (hash(x)=12917812)
17
+ 13 train 11.730721 (lr=4.8951e-06) (hash(x)=13413021)
18
+ 14 train 11.696169 (lr=5.2448e-06) (hash(x)=12655054)
19
+ 15 train 11.683971 (lr=5.5944e-06) (hash(x)=12684830)
20
+ 16 train 11.621240 (lr=5.9441e-06) (hash(x)=13439617)
21
+ 17 train 11.674701 (lr=6.2937e-06) (hash(x)=13262550)
22
+ 18 train 11.661559 (lr=6.6434e-06) (hash(x)=11975280)
23
+ 19 train 11.650582 (lr=6.9930e-06) (hash(x)=12003126)
24
+ 20 train 11.627670 (lr=7.3427e-06) (hash(x)=12313617)
25
+ 21 train 11.580272 (lr=7.6923e-06) (hash(x)=14144015)
26
+ 22 train 11.588894 (lr=8.0420e-06) (hash(x)=11178611)
27
+ 23 train 11.550992 (lr=8.3916e-06) (hash(x)=12014715)
28
+ 24 train 11.558554 (lr=8.7413e-06) (hash(x)=15240676)
29
+ 25 train 11.494004 (lr=9.0909e-06) (hash(x)=14912984)
30
+ 26 train 11.499259 (lr=9.4406e-06) (hash(x)=11487406)
31
+ 27 train 11.473790 (lr=9.7902e-06) (hash(x)=9492167)
32
+ 28 train 11.510754 (lr=1.0140e-05) (hash(x)=10765326)
33
+ 29 train 11.451578 (lr=1.0490e-05) (hash(x)=15241928)
34
+ 30 train 11.439598 (lr=1.0839e-05) (hash(x)=12215703)
35
+ 31 train 11.364937 (lr=1.1189e-05) (hash(x)=12595649)
36
+ 32 train 11.408415 (lr=1.1538e-05) (hash(x)=12726494)
37
+ 33 train 11.346420 (lr=1.1888e-05) (hash(x)=15162520)
38
+ 34 train 11.347469 (lr=1.2238e-05) (hash(x)=10640461)
39
+ 35 train 11.328934 (lr=1.2587e-05) (hash(x)=11980631)
40
+ 36 train 11.333309 (lr=1.2937e-05) (hash(x)=10568919)
41
+ 37 train 11.335015 (lr=1.3287e-05) (hash(x)=13172222)
42
+ 38 train 11.250566 (lr=1.3636e-05) (hash(x)=11575367)
43
+ 39 train 11.226089 (lr=1.3986e-05) (hash(x)=11101031)
44
+ 40 train 11.216280 (lr=1.4336e-05) (hash(x)=10304235)
45
+ 41 train 11.180510 (lr=1.4685e-05) (hash(x)=13275746)
46
+ 42 train 11.105244 (lr=1.5035e-05) (hash(x)=14309786)
47
+ 43 train 11.080940 (lr=1.5385e-05) (hash(x)=14316349)
48
+ 44 train 11.062172 (lr=1.5734e-05) (hash(x)=13999350)
49
+ 45 train 11.090105 (lr=1.6084e-05) (hash(x)=14143687)
50
+ 46 train 11.051409 (lr=1.6434e-05) (hash(x)=11365466)
51
+ 47 train 11.011821 (lr=1.6783e-05) (hash(x)=12036401)
52
+ 48 train 10.966129 (lr=1.7133e-05) (hash(x)=11730055)
53
+ 49 train 10.963387 (lr=1.7483e-05) (hash(x)=11143776)
54
+ 50 val loss 10.8994
55
+ 50 val perplexity 54141.9453
56
+ 50 train 10.913265 (lr=1.7832e-05) (hash(x)=12724270)
57
+ 51 train 10.871304 (lr=1.8182e-05) (hash(x)=12337125)
58
+ 52 train 10.806276 (lr=1.8531e-05) (hash(x)=12381632)
59
+ 53 train 10.815655 (lr=1.8881e-05) (hash(x)=10700109)
60
+ 54 train 10.726335 (lr=1.9231e-05) (hash(x)=13197786)
61
+ 55 train 10.694447 (lr=1.9580e-05) (hash(x)=11823810)
62
+ 56 train 10.774783 (lr=1.9930e-05) (hash(x)=12172918)
63
+ 57 train 10.695652 (lr=2.0280e-05) (hash(x)=12070591)
64
+ 58 train 10.653001 (lr=2.0629e-05) (hash(x)=14199166)
65
+ 59 train 10.706221 (lr=2.0979e-05) (hash(x)=13033959)
66
+ 60 train 10.646333 (lr=2.1329e-05) (hash(x)=12802857)
67
+ 61 train 10.583052 (lr=2.1678e-05) (hash(x)=10786770)
68
+ 62 train 10.596500 (lr=2.2028e-05) (hash(x)=11809534)
69
+ 63 train 10.427666 (lr=2.2378e-05) (hash(x)=13549925)
70
+ 64 train 10.356149 (lr=2.2727e-05) (hash(x)=10914323)
71
+ 65 train 10.341787 (lr=2.3077e-05) (hash(x)=14101838)
72
+ 66 train 10.481901 (lr=2.3427e-05) (hash(x)=11235125)
73
+ 67 train 10.491489 (lr=2.3776e-05) (hash(x)=14220701)
74
+ 68 train 10.312720 (lr=2.4126e-05) (hash(x)=12126282)
75
+ 69 train 10.401936 (lr=2.4476e-05) (hash(x)=14571180)
76
+ 70 train 10.298226 (lr=2.4825e-05) (hash(x)=14173731)
77
+ 71 train 10.436249 (lr=2.5175e-05) (hash(x)=13252262)
78
+ 72 train 10.241750 (lr=2.5524e-05) (hash(x)=14118291)
79
+ 73 train 10.247091 (lr=2.5874e-05) (hash(x)=13676730)
80
+ 74 train 10.323200 (lr=2.6224e-05) (hash(x)=13927111)
81
+ 75 train 10.263255 (lr=2.6573e-05) (hash(x)=14638145)
82
+ 76 train 10.233685 (lr=2.6923e-05) (hash(x)=14977819)
83
+ 77 train 10.228270 (lr=2.7273e-05) (hash(x)=14385711)
84
+ 78 train 10.318845 (lr=2.7622e-05) (hash(x)=12219671)
85
+ 79 train 10.309094 (lr=2.7972e-05) (hash(x)=13749054)
86
+ 80 train 10.321883 (lr=2.8322e-05) (hash(x)=11709001)
87
+ 81 train 10.269677 (lr=2.8671e-05) (hash(x)=12469703)
88
+ 82 train 10.246694 (lr=2.9021e-05) (hash(x)=11330478)
89
+ 83 train 10.217027 (lr=2.9371e-05) (hash(x)=12681760)
90
+ 84 train 10.256632 (lr=2.9720e-05) (hash(x)=15215674)
91
+ 85 train 10.239624 (lr=3.0070e-05) (hash(x)=12423697)
92
+ 86 train 10.149387 (lr=3.0420e-05) (hash(x)=11789526)
93
+ 87 train 10.160826 (lr=3.0769e-05) (hash(x)=14461706)
94
+ 88 train 10.203817 (lr=3.1119e-05) (hash(x)=10984941)
95
+ 89 train 10.160625 (lr=3.1469e-05) (hash(x)=13727016)
96
+ 90 train 10.065689 (lr=3.1818e-05) (hash(x)=12862054)
97
+ 91 train 10.007358 (lr=3.2168e-05) (hash(x)=14235218)
98
+ 92 train 10.132554 (lr=3.2517e-05) (hash(x)=13621196)
99
+ 93 train 10.102695 (lr=3.2867e-05) (hash(x)=12501568)
100
+ 94 train 10.272562 (lr=3.3217e-05) (hash(x)=12274822)
101
+ 95 train 10.121917 (lr=3.3566e-05) (hash(x)=11735995)
102
+ 96 train 10.234634 (lr=3.3916e-05) (hash(x)=13050719)
103
+ 97 train 10.273458 (lr=3.4266e-05) (hash(x)=11569945)
104
+ 98 train 10.068501 (lr=3.4615e-05) (hash(x)=11109133)
105
+ 99 train 10.067688 (lr=3.4965e-05) (hash(x)=11641112)
106
+ 100 val loss 10.0889
107
+ 100 val perplexity 24073.1855
108
+ 100 train 10.115582 (lr=3.5315e-05) (hash(x)=11507240)
109
+ 101 train 10.060678 (lr=3.5664e-05) (hash(x)=11536784)
110
+ 102 train 10.075945 (lr=3.6014e-05) (hash(x)=11514312)
111
+ 103 train 10.202707 (lr=3.6364e-05) (hash(x)=12231701)
112
+ 104 train 10.093767 (lr=3.6713e-05) (hash(x)=10619273)
113
+ 105 train 10.042521 (lr=3.7063e-05) (hash(x)=13521301)
114
+ 106 train 9.912802 (lr=3.7413e-05) (hash(x)=10295729)
115
+ 107 train 9.993573 (lr=3.7762e-05) (hash(x)=10175470)
116
+ 108 train 10.054693 (lr=3.8112e-05) (hash(x)=11948399)
117
+ 109 train 10.043325 (lr=3.8462e-05) (hash(x)=13903066)
118
+ 110 train 10.519420 (lr=3.8811e-05) (hash(x)=16265112)
119
+ 111 train 9.990005 (lr=3.9161e-05) (hash(x)=14117945)
120
+ 112 train 9.940434 (lr=3.9510e-05) (hash(x)=12712901)
121
+ 113 train 10.020107 (lr=3.9860e-05) (hash(x)=14851356)
122
+ 114 train 10.048941 (lr=4.0210e-05) (hash(x)=12337690)
123
+ 115 train 9.977542 (lr=4.0559e-05) (hash(x)=13116544)
124
+ 116 train 9.927397 (lr=4.0909e-05) (hash(x)=16794825)
125
+ 117 train 9.987131 (lr=4.1259e-05) (hash(x)=16366718)
126
+ 118 train 9.963415 (lr=4.1608e-05) (hash(x)=13520931)
127
+ 119 train 9.917052 (lr=4.1958e-05) (hash(x)=12678303)
128
+ 120 train 9.899405 (lr=4.2308e-05) (hash(x)=10020764)
129
+ 121 train 9.781209 (lr=4.2657e-05) (hash(x)=11144768)
130
+ 122 train 10.016682 (lr=4.3007e-05) (hash(x)=10378383)
131
+ 123 train 9.859229 (lr=4.3357e-05) (hash(x)=11927462)
132
+ 124 train 9.830335 (lr=4.3706e-05) (hash(x)=13130532)
133
+ 125 train 9.841994 (lr=4.4056e-05) (hash(x)=10418366)
134
+ 126 train 9.893745 (lr=4.4406e-05) (hash(x)=14920480)
135
+ 127 train 9.891519 (lr=4.4755e-05) (hash(x)=12758461)
136
+ 128 train 9.901266 (lr=4.5105e-05) (hash(x)=14632236)
137
+ 129 train 9.919140 (lr=4.5455e-05) (hash(x)=11680802)
138
+ 130 train 9.867279 (lr=4.5804e-05) (hash(x)=12682477)
139
+ 131 train 9.959133 (lr=4.6154e-05) (hash(x)=11946230)
140
+ 132 train 9.876947 (lr=4.6503e-05) (hash(x)=15160610)
141
+ 133 train 9.754589 (lr=4.6853e-05) (hash(x)=11704519)
142
+ 134 train 9.687277 (lr=4.7203e-05) (hash(x)=12746108)
143
+ 135 train 9.733885 (lr=4.7552e-05) (hash(x)=13701311)
144
+ 136 train 9.762169 (lr=4.7902e-05) (hash(x)=15682136)
145
+ 137 train 9.779931 (lr=4.8252e-05) (hash(x)=12751615)
146
+ 138 train 9.812056 (lr=4.8601e-05) (hash(x)=13420600)
147
+ 139 train 9.701154 (lr=4.8951e-05) (hash(x)=11198382)
148
+ 140 train 9.811452 (lr=4.9301e-05) (hash(x)=11556406)
149
+ 141 train 9.820650 (lr=4.9650e-05) (hash(x)=16075171)
150
+ 142 train 9.924412 (lr=5.0000e-05) (hash(x)=16911864)
151
+ 143 train 9.712167 (lr=5.0350e-05) (hash(x)=14346433)
152
+ 144 train 9.775518 (lr=5.0699e-05) (hash(x)=12105279)
153
+ 145 train 9.734037 (lr=5.1049e-05) (hash(x)=12888806)
154
+ 146 train 9.769306 (lr=5.1399e-05) (hash(x)=11676047)
155
+ 147 train 9.923992 (lr=5.1748e-05) (hash(x)=17855984)
156
+ 148 train 9.805021 (lr=5.2098e-05) (hash(x)=14922202)
157
+ 149 train 9.723958 (lr=5.2448e-05) (hash(x)=9616381)
158
+ 150 val loss 9.6526
159
+ 150 val perplexity 15561.5488
160
+ 150 train 9.714430 (lr=5.2797e-05) (hash(x)=11978542)
161
+ 151 train 10.330320 (lr=5.3147e-05) (hash(x)=18910052)
162
+ 152 train 9.971045 (lr=5.3497e-05) (hash(x)=13545339)
163
+ 153 train 9.589483 (lr=5.3846e-05) (hash(x)=12769596)
164
+ 154 train 9.584196 (lr=5.4196e-05) (hash(x)=12235730)
165
+ 155 train 9.832221 (lr=5.4545e-05) (hash(x)=11885403)
166
+ 156 train 9.669153 (lr=5.4895e-05) (hash(x)=12677961)
167
+ 157 train 9.687387 (lr=5.5245e-05) (hash(x)=12476514)
168
+ 158 train 9.681535 (lr=5.5594e-05) (hash(x)=13656193)
169
+ 159 train 9.645142 (lr=5.5944e-05) (hash(x)=12508350)
170
+ 160 train 9.515568 (lr=5.6294e-05) (hash(x)=11204172)
171
+ 161 train 9.546794 (lr=5.6643e-05) (hash(x)=10745637)
172
+ 162 train 9.533916 (lr=5.6993e-05) (hash(x)=11844793)
173
+ 163 train 9.696778 (lr=5.7343e-05) (hash(x)=10862804)
174
+ 164 train 9.699289 (lr=5.7692e-05) (hash(x)=12061659)
175
+ 165 train 9.579821 (lr=5.8042e-05) (hash(x)=15075084)
176
+ 166 train 9.535169 (lr=5.8392e-05) (hash(x)=12508408)
177
+ 167 train 9.534990 (lr=5.8741e-05) (hash(x)=14043988)
178
+ 168 train 9.327209 (lr=5.9091e-05) (hash(x)=10401493)
179
+ 169 train 9.395159 (lr=5.9441e-05) (hash(x)=12667924)
180
+ 170 train 9.330454 (lr=5.9790e-05) (hash(x)=12778664)
181
+ 171 train 9.450498 (lr=6.0140e-05) (hash(x)=11916406)
182
+ 172 train 9.491482 (lr=6.0490e-05) (hash(x)=11516862)
183
+ 173 train 9.342565 (lr=6.0839e-05) (hash(x)=13406195)
184
+ 174 train 9.501434 (lr=6.1189e-05) (hash(x)=12116183)
185
+ 175 train 9.358486 (lr=6.1538e-05) (hash(x)=11968227)
186
+ 176 train 9.534603 (lr=6.1888e-05) (hash(x)=12134811)
187
+ 177 train 9.391093 (lr=6.2238e-05) (hash(x)=11927490)
188
+ 178 train 9.402212 (lr=6.2587e-05) (hash(x)=12684747)
189
+ 179 train 9.376186 (lr=6.2937e-05) (hash(x)=13290191)
190
+ 180 train 9.463200 (lr=6.3287e-05) (hash(x)=11789931)
191
+ 181 train 9.241015 (lr=6.3636e-05) (hash(x)=11771253)
192
+ 182 train 9.391233 (lr=6.3986e-05) (hash(x)=10693812)
193
+ 183 train 9.392084 (lr=6.4336e-05) (hash(x)=15670776)
194
+ 184 train 9.313274 (lr=6.4685e-05) (hash(x)=13586931)
195
+ 185 train 9.242701 (lr=6.5035e-05) (hash(x)=11306133)
196
+ 186 train 9.292295 (lr=6.5385e-05) (hash(x)=12636637)
197
+ 187 train 9.471985 (lr=6.5734e-05) (hash(x)=14358384)
198
+ 188 train 9.272419 (lr=6.6084e-05) (hash(x)=11557997)
199
+ 189 train 9.227139 (lr=6.6434e-05) (hash(x)=10267428)
200
+ 190 train 9.386654 (lr=6.6783e-05) (hash(x)=11861629)
201
+ 191 train 9.229574 (lr=6.7133e-05) (hash(x)=11289341)
202
+ 192 train 9.251202 (lr=6.7483e-05) (hash(x)=11666650)
203
+ 193 train 9.199606 (lr=6.7832e-05) (hash(x)=14062293)
204
+ 194 train 9.310980 (lr=6.8182e-05) (hash(x)=10785993)
205
+ 195 train 9.356422 (lr=6.8531e-05) (hash(x)=14736057)
206
+ 196 train 9.229067 (lr=6.8881e-05) (hash(x)=12054745)
207
+ 197 train 9.584971 (lr=6.9231e-05) (hash(x)=16036112)
208
+ 198 train 9.507402 (lr=6.9580e-05) (hash(x)=16263589)
209
+ 199 train 9.221516 (lr=6.9930e-05) (hash(x)=11482477)
210
+ 200 val loss 9.2076
211
+ 200 val perplexity 9972.3926
212
+ 200 train 9.293366 (lr=7.0280e-05) (hash(x)=13356719)
213
+ 201 train 9.261099 (lr=7.0629e-05) (hash(x)=12742648)
214
+ 202 train 9.157481 (lr=7.0979e-05) (hash(x)=11963702)
215
+ 203 train 9.203009 (lr=7.1329e-05) (hash(x)=15429973)
216
+ 204 train 9.214949 (lr=7.1678e-05) (hash(x)=13257554)
217
+ 205 train 9.241796 (lr=7.2028e-05) (hash(x)=13972190)
218
+ 206 train 9.260845 (lr=7.2378e-05) (hash(x)=13430094)
219
+ 207 train 9.114950 (lr=7.2727e-05) (hash(x)=10345057)
220
+ 208 train 9.197392 (lr=7.3077e-05) (hash(x)=14689084)
221
+ 209 train 9.064786 (lr=7.3427e-05) (hash(x)=11609727)
222
+ 210 train 9.138233 (lr=7.3776e-05) (hash(x)=13860968)
223
+ 211 train 9.024152 (lr=7.4126e-05) (hash(x)=10816510)
224
+ 212 train 8.839532 (lr=7.4476e-05) (hash(x)=10255921)
225
+ 213 train 9.185515 (lr=7.4825e-05) (hash(x)=11801604)
226
+ 214 train 9.070998 (lr=7.5175e-05) (hash(x)=11442918)
227
+ 215 train 9.087047 (lr=7.5524e-05) (hash(x)=11157124)
228
+ 216 train 9.171734 (lr=7.5874e-05) (hash(x)=9862579)
229
+ 217 train 9.025509 (lr=7.6224e-05) (hash(x)=11666699)
230
+ 218 train 8.960950 (lr=7.6573e-05) (hash(x)=15933079)
231
+ 219 train 9.037056 (lr=7.6923e-05) (hash(x)=12349897)
232
+ 220 train 9.028759 (lr=7.7273e-05) (hash(x)=11150337)
233
+ 221 train 9.021337 (lr=7.7622e-05) (hash(x)=11648826)
234
+ 222 train 9.046867 (lr=7.7972e-05) (hash(x)=11198064)
235
+ 223 train 8.925087 (lr=7.8322e-05) (hash(x)=10750452)
236
+ 224 train 8.902761 (lr=7.8671e-05) (hash(x)=11819859)
237
+ 225 train 8.922842 (lr=7.9021e-05) (hash(x)=12114386)
238
+ 226 train 9.047170 (lr=7.9371e-05) (hash(x)=13023365)
239
+ 227 train 8.950422 (lr=7.9720e-05) (hash(x)=11859950)
240
+ 228 train 8.810795 (lr=8.0070e-05) (hash(x)=12542485)
241
+ 229 train 8.822716 (lr=8.0420e-05) (hash(x)=12837162)
242
+ 230 train 8.896549 (lr=8.0769e-05) (hash(x)=12185415)
243
+ 231 train 8.935103 (lr=8.1119e-05) (hash(x)=8921601)
244
+ 232 train 8.838991 (lr=8.1469e-05) (hash(x)=11572366)
245
+ 233 train 8.426647 (lr=8.1818e-05) (hash(x)=9125058)
246
+ 234 train 8.647223 (lr=8.2168e-05) (hash(x)=12007582)
247
+ 235 train 8.742462 (lr=8.2517e-05) (hash(x)=11333049)
248
+ 236 train 8.714151 (lr=8.2867e-05) (hash(x)=10626272)
249
+ 237 train 8.814505 (lr=8.3217e-05) (hash(x)=13807539)
250
+ 238 train 8.802027 (lr=8.3566e-05) (hash(x)=11429976)
251
+ 239 train 8.647377 (lr=8.3916e-05) (hash(x)=12165500)
252
+ 240 train 8.892940 (lr=8.4266e-05) (hash(x)=15491314)
253
+ 241 train 8.771452 (lr=8.4615e-05) (hash(x)=14343620)
254
+ 242 train 8.725523 (lr=8.4965e-05) (hash(x)=12358970)
255
+ 243 train 8.810232 (lr=8.5315e-05) (hash(x)=13153948)
256
+ 244 train 8.878004 (lr=8.5664e-05) (hash(x)=12935023)
257
+ 245 train 8.674495 (lr=8.6014e-05) (hash(x)=14497853)
258
+ 246 train 8.634483 (lr=8.6364e-05) (hash(x)=12545808)
259
+ 247 train 8.917095 (lr=8.6713e-05) (hash(x)=13569316)
260
+ 248 train 8.775635 (lr=8.7063e-05) (hash(x)=13033944)
261
+ 249 train 8.773870 (lr=8.7413e-05) (hash(x)=14212041)
262
+ 250 val loss 8.6448
263
+ 250 val perplexity 5680.4673
264
+ 250 train 8.854788 (lr=8.7762e-05) (hash(x)=10769873)
265
+ 251 train 8.689565 (lr=8.8112e-05) (hash(x)=11456278)
266
+ 252 train 8.622475 (lr=8.8462e-05) (hash(x)=13141720)
267
+ 253 train 8.521441 (lr=8.8811e-05) (hash(x)=13563812)
268
+ 254 train 8.635357 (lr=8.9161e-05) (hash(x)=12645610)
269
+ 255 train 8.593071 (lr=8.9510e-05) (hash(x)=14043713)
270
+ 256 train 8.727262 (lr=8.9860e-05) (hash(x)=14545052)
271
+ 257 train 8.554062 (lr=9.0210e-05) (hash(x)=13968786)
272
+ 258 train 8.714911 (lr=9.0559e-05) (hash(x)=11573299)
273
+ 259 train 8.590847 (lr=9.0909e-05) (hash(x)=11896948)
274
+ 260 train 8.688370 (lr=9.1259e-05) (hash(x)=11602278)
275
+ 261 train 8.653865 (lr=9.1608e-05) (hash(x)=13298324)
276
+ 262 train 8.765777 (lr=9.1958e-05) (hash(x)=13232135)
277
+ 263 train 8.495005 (lr=9.2308e-05) (hash(x)=11969650)
278
+ 264 train 8.526446 (lr=9.2657e-05) (hash(x)=12202825)
279
+ 265 train 8.442101 (lr=9.3007e-05) (hash(x)=12021926)
280
+ 266 train 8.406853 (lr=9.3357e-05) (hash(x)=11486982)
281
+ 267 train 8.461126 (lr=9.3706e-05) (hash(x)=14387351)
282
+ 268 train 8.540955 (lr=9.4056e-05) (hash(x)=14588570)
283
+ 269 train 8.360983 (lr=9.4406e-05) (hash(x)=13939547)
284
+ 270 train 8.278429 (lr=9.4755e-05) (hash(x)=13336938)
285
+ 271 train 8.178799 (lr=9.5105e-05) (hash(x)=13680255)
286
+ 272 train 8.488401 (lr=9.5455e-05) (hash(x)=11276607)
287
+ 273 train 8.506682 (lr=9.5804e-05) (hash(x)=10982564)
288
+ 274 train 8.671436 (lr=9.6154e-05) (hash(x)=13821541)
289
+ 275 train 8.683128 (lr=9.6503e-05) (hash(x)=11875534)
290
+ 276 train 8.513229 (lr=9.6853e-05) (hash(x)=14296280)
291
+ 277 train 8.687487 (lr=9.7203e-05) (hash(x)=13390260)
292
+ 278 train 8.632055 (lr=9.7552e-05) (hash(x)=13142039)
293
+ 279 train 8.471149 (lr=9.7902e-05) (hash(x)=13871675)
294
+ 280 train 8.447507 (lr=9.8252e-05) (hash(x)=13297982)
295
+ 281 train 8.255009 (lr=9.8601e-05) (hash(x)=10728444)
296
+ 282 train 8.318060 (lr=9.8951e-05) (hash(x)=11539550)
297
+ 283 train 8.469046 (lr=9.9301e-05) (hash(x)=13389922)
298
+ 284 train 8.440298 (lr=9.9650e-05) (hash(x)=13390153)
299
+ 285 train 8.518793 (lr=1.0000e-04) (hash(x)=13458742)
300
+ 286 train 8.281790 (lr=1.0000e-04) (hash(x)=10387541)
301
+ 287 train 8.285935 (lr=1.0000e-04) (hash(x)=12263431)
302
+ 288 train 8.329247 (lr=9.9998e-05) (hash(x)=10405877)
303
+ 289 train 8.351934 (lr=9.9996e-05) (hash(x)=12377887)
304
+ 290 train 8.163462 (lr=9.9993e-05) (hash(x)=10511081)
305
+ 291 train 8.429324 (lr=9.9989e-05) (hash(x)=10141365)
306
+ 292 train 8.357029 (lr=9.9984e-05) (hash(x)=13935127)
307
+ 293 train 8.458170 (lr=9.9979e-05) (hash(x)=13965689)
308
+ 294 train 8.259862 (lr=9.9972e-05) (hash(x)=13301076)
309
+ 295 train 8.216383 (lr=9.9965e-05) (hash(x)=12414778)
310
+ 296 train 8.291527 (lr=9.9956e-05) (hash(x)=12497839)
311
+ 297 train 8.119722 (lr=9.9947e-05) (hash(x)=9727144)
312
+ 298 train 8.144386 (lr=9.9937e-05) (hash(x)=11972438)
313
+ 299 train 8.443722 (lr=9.9926e-05) (hash(x)=11738134)
314
+ 300 val loss 8.2412
315
+ 300 val perplexity 3794.0740
316
+ 300 train 8.849049 (lr=9.9915e-05) (hash(x)=15826242)
317
+ 301 train 8.606176 (lr=9.9902e-05) (hash(x)=14924380)
318
+ 302 train 8.456728 (lr=9.9889e-05) (hash(x)=10605039)
319
+ 303 train 7.965144 (lr=9.9874e-05) (hash(x)=14443665)
320
+ 304 train 8.421545 (lr=9.9859e-05) (hash(x)=14431398)
321
+ 305 train 8.282658 (lr=9.9843e-05) (hash(x)=12873827)
322
+ 306 train 8.280551 (lr=9.9826e-05) (hash(x)=11307508)
323
+ 307 train 8.262413 (lr=9.9808e-05) (hash(x)=13435160)
324
+ 308 train 8.359430 (lr=9.9789e-05) (hash(x)=13345013)
325
+ 309 train 8.434390 (lr=9.9770e-05) (hash(x)=13024409)
326
+ 310 train 8.563750 (lr=9.9749e-05) (hash(x)=14339476)
327
+ 311 train 8.329788 (lr=9.9728e-05) (hash(x)=13100449)
328
+ 312 train 8.182564 (lr=9.9706e-05) (hash(x)=10843771)
329
+ 313 train 8.343727 (lr=9.9683e-05) (hash(x)=13304906)
330
+ 314 train 8.184154 (lr=9.9659e-05) (hash(x)=11167598)
331
+ 315 train 8.155850 (lr=9.9634e-05) (hash(x)=11690244)
332
+ 316 train 8.136557 (lr=9.9609e-05) (hash(x)=11802493)
333
+ 317 train 8.328794 (lr=9.9582e-05) (hash(x)=12027838)
334
+ 318 train 8.019601 (lr=9.9555e-05) (hash(x)=12643274)
335
+ 319 train 8.195408 (lr=9.9526e-05) (hash(x)=13953073)
336
+ 320 train 8.126009 (lr=9.9497e-05) (hash(x)=12882377)
337
+ 321 train 8.111633 (lr=9.9467e-05) (hash(x)=12066345)
338
+ 322 train 8.319790 (lr=9.9437e-05) (hash(x)=14822016)
339
+ 323 train 8.232745 (lr=9.9405e-05) (hash(x)=12657159)
340
+ 324 train 8.065827 (lr=9.9372e-05) (hash(x)=11740673)
341
+ 325 train 8.325679 (lr=9.9339e-05) (hash(x)=11335971)
342
+ 326 train 8.171016 (lr=9.9305e-05) (hash(x)=11550890)
343
+ 327 train 7.937030 (lr=9.9270e-05) (hash(x)=10956041)
344
+ 328 train 7.969145 (lr=9.9234e-05) (hash(x)=9941344)
345
+ 329 train 7.914227 (lr=9.9197e-05) (hash(x)=9766608)
346
+ 330 train 7.977751 (lr=9.9159e-05) (hash(x)=11007490)
347
+ 331 train 7.860253 (lr=9.9121e-05) (hash(x)=11060230)
348
+ 332 train 8.350179 (lr=9.9081e-05) (hash(x)=15459248)
349
+ 333 train 8.177165 (lr=9.9041e-05) (hash(x)=11396268)
350
+ 334 train 8.071403 (lr=9.9000e-05) (hash(x)=11127560)
351
+ 335 train 8.056118 (lr=9.8958e-05) (hash(x)=12840879)
352
+ 336 train 7.971099 (lr=9.8915e-05) (hash(x)=11354437)
353
+ 337 train 7.990168 (lr=9.8872e-05) (hash(x)=10987186)
354
+ 338 train 7.883435 (lr=9.8827e-05) (hash(x)=8972293)
355
+ 339 train 7.930821 (lr=9.8782e-05) (hash(x)=8916599)
356
+ 340 train 7.958641 (lr=9.8736e-05) (hash(x)=10487177)
357
+ 341 train 8.266936 (lr=9.8689e-05) (hash(x)=13235433)
358
+ 342 train 7.982561 (lr=9.8641e-05) (hash(x)=12916423)
359
+ 343 train 8.050522 (lr=9.8592e-05) (hash(x)=14428050)
360
+ 344 train 8.012298 (lr=9.8543e-05) (hash(x)=13204551)
361
+ 345 train 8.023855 (lr=9.8492e-05) (hash(x)=12631329)
362
+ 346 train 8.032810 (lr=9.8441e-05) (hash(x)=13809442)
363
+ 347 train 7.995026 (lr=9.8389e-05) (hash(x)=12787108)
364
+ 348 train 8.026801 (lr=9.8336e-05) (hash(x)=11912209)
365
+ 349 train 8.013836 (lr=9.8282e-05) (hash(x)=14499940)
366
+ 350 val loss 7.9544
367
+ 350 val perplexity 2848.1416
368
+ 350 train 8.115109 (lr=9.8228e-05) (hash(x)=13911589)
369
+ 351 train 8.212200 (lr=9.8172e-05) (hash(x)=14104576)
370
+ 352 train 8.111751 (lr=9.8116e-05) (hash(x)=12414592)
371
+ 353 train 8.098505 (lr=9.8059e-05) (hash(x)=13943168)
372
+ 354 train 8.010720 (lr=9.8001e-05) (hash(x)=12770473)
373
+ 355 train 7.959424 (lr=9.7942e-05) (hash(x)=12623743)
374
+ 356 train 8.087355 (lr=9.7882e-05) (hash(x)=12441115)
375
+ 357 train 8.138383 (lr=9.7822e-05) (hash(x)=12689741)
376
+ 358 train 8.019139 (lr=9.7761e-05) (hash(x)=12703188)
377
+ 359 train 8.004976 (lr=9.7699e-05) (hash(x)=14453314)
378
+ 360 train 7.898977 (lr=9.7636e-05) (hash(x)=12038031)
379
+ 361 train 7.872993 (lr=9.7572e-05) (hash(x)=10824391)
380
+ 362 train 7.893896 (lr=9.7507e-05) (hash(x)=11527801)
381
+ 363 train 7.588005 (lr=9.7442e-05) (hash(x)=10470424)
382
+ 364 train 7.497739 (lr=9.7376e-05) (hash(x)=12820231)
383
+ 365 train 8.764436 (lr=9.7309e-05) (hash(x)=15186277)
384
+ 366 train 7.985453 (lr=9.7241e-05) (hash(x)=13477578)
385
+ 367 train 7.987213 (lr=9.7172e-05) (hash(x)=12918172)
386
+ 368 train 7.732706 (lr=9.7103e-05) (hash(x)=12480955)
387
+ 369 train 8.073925 (lr=9.7032e-05) (hash(x)=11936107)
388
+ 370 train 7.824166 (lr=9.6961e-05) (hash(x)=9742856)
389
+ 371 train 7.952458 (lr=9.6889e-05) (hash(x)=13259459)
390
+ 372 train 7.818108 (lr=9.6817e-05) (hash(x)=11608070)
391
+ 373 train 7.839115 (lr=9.6743e-05) (hash(x)=12918385)
392
+ 374 train 7.953631 (lr=9.6669e-05) (hash(x)=12761761)
393
+ 375 train 7.947084 (lr=9.6593e-05) (hash(x)=12368572)
394
+ 376 train 7.898869 (lr=9.6518e-05) (hash(x)=11587019)
395
+ 377 train 8.152098 (lr=9.6441e-05) (hash(x)=13455578)
396
+ 378 train 7.933698 (lr=9.6363e-05) (hash(x)=12043438)
397
+ 379 train 7.857199 (lr=9.6285e-05) (hash(x)=13505144)
398
+ 380 train 7.873709 (lr=9.6206e-05) (hash(x)=12626641)
399
+ 381 train 8.028501 (lr=9.6126e-05) (hash(x)=13370946)
400
+ 382 train 7.672528 (lr=9.6045e-05) (hash(x)=9854862)
401
+ 383 train 7.934245 (lr=9.5963e-05) (hash(x)=13340443)
402
+ 384 train 7.878504 (lr=9.5881e-05) (hash(x)=12641299)
403
+ 385 train 7.841083 (lr=9.5798e-05) (hash(x)=12806141)
404
+ 386 train 8.097155 (lr=9.5714e-05) (hash(x)=12354021)
405
+ 387 train 7.767145 (lr=9.5629e-05) (hash(x)=11469900)
406
+ 388 train 7.516455 (lr=9.5544e-05) (hash(x)=10099706)
407
+ 389 train 7.784154 (lr=9.5457e-05) (hash(x)=11591533)
408
+ 390 train 7.826637 (lr=9.5370e-05) (hash(x)=11174793)
409
+ 391 train 7.637680 (lr=9.5282e-05) (hash(x)=10742101)
410
+ 392 train 7.823531 (lr=9.5194e-05) (hash(x)=13136821)
411
+ 393 train 7.738689 (lr=9.5104e-05) (hash(x)=11781196)
412
+ 394 train 7.636090 (lr=9.5014e-05) (hash(x)=11847817)
413
+ 395 train 7.848371 (lr=9.4923e-05) (hash(x)=13261517)
414
+ 396 train 7.731215 (lr=9.4831e-05) (hash(x)=11622233)
415
+ 397 train 7.950476 (lr=9.4739e-05) (hash(x)=13592858)
416
+ 398 train 8.043576 (lr=9.4646e-05) (hash(x)=15017875)
417
+ 399 train 7.975801 (lr=9.4551e-05) (hash(x)=13025742)
418
+ 400 val loss 7.8235
419
+ 400 val perplexity 2498.7537
420
+ 400 train 7.823912 (lr=9.4457e-05) (hash(x)=11718604)
421
+ 401 train 7.818813 (lr=9.4361e-05) (hash(x)=12152419)
422
+ 402 train 7.910639 (lr=9.4265e-05) (hash(x)=14069856)
423
+ 403 train 8.363827 (lr=9.4168e-05) (hash(x)=15100390)
424
+ 404 train 7.997580 (lr=9.4070e-05) (hash(x)=12718574)
425
+ 405 train 7.692454 (lr=9.3971e-05) (hash(x)=10708197)
426
+ 406 train 7.824056 (lr=9.3872e-05) (hash(x)=10819941)
427
+ 407 train 7.856200 (lr=9.3772e-05) (hash(x)=14351068)
428
+ 408 train 7.964739 (lr=9.3671e-05) (hash(x)=14377468)
429
+ 409 train 8.074910 (lr=9.3569e-05) (hash(x)=12619936)
430
+ 410 train 7.586822 (lr=9.3467e-05) (hash(x)=7908802)
431
+ 411 train 7.746769 (lr=9.3364e-05) (hash(x)=13445545)
432
+ 412 train 7.991012 (lr=9.3260e-05) (hash(x)=14062361)
433
+ 413 train 7.709524 (lr=9.3155e-05) (hash(x)=11375930)
434
+ 414 train 7.767105 (lr=9.3050e-05) (hash(x)=10519208)
435
+ 415 train 8.178192 (lr=9.2944e-05) (hash(x)=13155663)
436
+ 416 train 8.088257 (lr=9.2837e-05) (hash(x)=15774406)
437
+ 417 train 7.739265 (lr=9.2729e-05) (hash(x)=13404244)
438
+ 418 train 7.795409 (lr=9.2621e-05) (hash(x)=9995421)
439
+ 419 train 7.787715 (lr=9.2512e-05) (hash(x)=13309898)
440
+ 420 train 7.982796 (lr=9.2402e-05) (hash(x)=13892965)
441
+ 421 train 7.875384 (lr=9.2292e-05) (hash(x)=12356526)
442
+ 422 train 7.776760 (lr=9.2181e-05) (hash(x)=12133609)
443
+ 423 train 7.569073 (lr=9.2069e-05) (hash(x)=9157946)
444
+ 424 train 7.604524 (lr=9.1956e-05) (hash(x)=11770021)
445
+ 425 train 7.791361 (lr=9.1843e-05) (hash(x)=11837703)
446
+ 426 train 7.707518 (lr=9.1729e-05) (hash(x)=9337790)
447
+ 427 train 7.726833 (lr=9.1614e-05) (hash(x)=12222186)
448
+ 428 train 7.639417 (lr=9.1499e-05) (hash(x)=11408945)
449
+ 429 train 7.772134 (lr=9.1382e-05) (hash(x)=12588150)
450
+ 430 train 7.689183 (lr=9.1266e-05) (hash(x)=12035009)
451
+ 431 train 7.782581 (lr=9.1148e-05) (hash(x)=13872294)
452
+ 432 train 7.799926 (lr=9.1030e-05) (hash(x)=13115811)
453
+ 433 train 7.596321 (lr=9.0911e-05) (hash(x)=10621792)
454
+ 434 train 7.802772 (lr=9.0791e-05) (hash(x)=11829316)
455
+ 435 train 7.646570 (lr=9.0671e-05) (hash(x)=11824400)
456
+ 436 train 7.790904 (lr=9.0550e-05) (hash(x)=13258944)
457
+ 437 train 7.915294 (lr=9.0428e-05) (hash(x)=12495032)
458
+ 438 train 7.786463 (lr=9.0306e-05) (hash(x)=12623570)
459
+ 439 train 7.854343 (lr=9.0182e-05) (hash(x)=12773612)
460
+ 440 train 7.933664 (lr=9.0059e-05) (hash(x)=11183399)
461
+ 441 train 7.676121 (lr=8.9934e-05) (hash(x)=10808441)
462
+ 442 train 7.908303 (lr=8.9809e-05) (hash(x)=13523831)
463
+ 443 train 7.860227 (lr=8.9683e-05) (hash(x)=12794078)
464
+ 444 train 7.771989 (lr=8.9557e-05) (hash(x)=12098496)
465
+ 445 train 7.738407 (lr=8.9430e-05) (hash(x)=11328897)
466
+ 446 train 7.874512 (lr=8.9302e-05) (hash(x)=12482357)
467
+ 447 train 7.777729 (lr=8.9173e-05) (hash(x)=12235713)
468
+ 448 train 7.556093 (lr=8.9044e-05) (hash(x)=12612778)
469
+ 449 train 7.823787 (lr=8.8914e-05) (hash(x)=11512364)
470
+ 450 val loss 7.7351
471
+ 450 val perplexity 2287.1978
472
+ 450 train 7.842170 (lr=8.8784e-05) (hash(x)=13110980)
473
+ 451 train 7.497395 (lr=8.8653e-05) (hash(x)=10270693)
474
+ 452 train 7.487044 (lr=8.8521e-05) (hash(x)=10333232)
475
+ 453 train 7.635853 (lr=8.8388e-05) (hash(x)=12380104)
476
+ 454 train 7.522508 (lr=8.8255e-05) (hash(x)=12106102)
477
+ 455 train 7.693919 (lr=8.8122e-05) (hash(x)=13686441)
478
+ 456 train 8.260736 (lr=8.7987e-05) (hash(x)=11802878)
479
+ 457 train 7.987062 (lr=8.7852e-05) (hash(x)=14151557)
480
+ 458 train 8.010217 (lr=8.7717e-05) (hash(x)=13785241)
481
+ 459 train 7.868275 (lr=8.7580e-05) (hash(x)=13567687)
482
+ 460 train 7.859231 (lr=8.7444e-05) (hash(x)=12731340)
483
+ 461 train 7.737118 (lr=8.7306e-05) (hash(x)=12718282)
484
+ 462 train 7.852074 (lr=8.7168e-05) (hash(x)=13733829)
485
+ 463 train 7.611156 (lr=8.7029e-05) (hash(x)=13214962)
486
+ 464 train 7.771212 (lr=8.6890e-05) (hash(x)=11298890)
487
+ 465 train 7.646701 (lr=8.6750e-05) (hash(x)=10855282)
488
+ 466 train 8.047655 (lr=8.6609e-05) (hash(x)=14082841)
489
+ 467 train 7.857740 (lr=8.6468e-05) (hash(x)=12404167)
490
+ 468 train 8.250472 (lr=8.6326e-05) (hash(x)=15826094)
491
+ 469 train 7.950135 (lr=8.6184e-05) (hash(x)=13017256)
492
+ 470 train 7.704644 (lr=8.6041e-05) (hash(x)=13446769)
493
+ 471 train 7.749940 (lr=8.5897e-05) (hash(x)=12602681)
494
+ 472 train 7.571863 (lr=8.5753e-05) (hash(x)=10655764)
495
+ 473 train 7.459012 (lr=8.5608e-05) (hash(x)=11916856)
496
+ 474 train 8.039892 (lr=8.5462e-05) (hash(x)=13726124)
497
+ 475 train 7.756803 (lr=8.5316e-05) (hash(x)=13850551)
498
+ 476 train 7.596747 (lr=8.5170e-05) (hash(x)=12614910)
499
+ 477 train 7.636633 (lr=8.5022e-05) (hash(x)=11382479)
500
+ 478 train 7.735344 (lr=8.4875e-05) (hash(x)=13387304)
501
+ 479 train 7.799362 (lr=8.4726e-05) (hash(x)=13829219)
502
+ 480 train 7.802971 (lr=8.4577e-05) (hash(x)=14735902)
503
+ 481 train 8.049239 (lr=8.4428e-05) (hash(x)=13540459)
504
+ 482 train 7.777039 (lr=8.4278e-05) (hash(x)=12591112)
505
+ 483 train 7.864939 (lr=8.4127e-05) (hash(x)=15485762)
506
+ 484 train 7.789549 (lr=8.3976e-05) (hash(x)=12950474)
507
+ 485 train 7.738090 (lr=8.3824e-05) (hash(x)=12268519)
508
+ 486 train 7.264820 (lr=8.3672e-05) (hash(x)=8011105)
509
+ 487 train 7.292631 (lr=8.3519e-05) (hash(x)=9335220)
510
+ 488 train 7.445064 (lr=8.3366e-05) (hash(x)=10948116)
511
+ 489 train 7.913856 (lr=8.3212e-05) (hash(x)=13359270)
512
+ 490 train 7.812678 (lr=8.3057e-05) (hash(x)=12487329)
513
+ 491 train 7.756363 (lr=8.2902e-05) (hash(x)=13162637)
514
+ 492 train 7.908572 (lr=8.2746e-05) (hash(x)=12860332)
515
+ 493 train 7.751006 (lr=8.2590e-05) (hash(x)=13305408)
516
+ 494 train 7.737789 (lr=8.2434e-05) (hash(x)=11796684)
517
+ 495 train 7.884517 (lr=8.2276e-05) (hash(x)=15713183)
518
+ 496 train 7.744094 (lr=8.2119e-05) (hash(x)=11904281)
519
+ 497 train 7.729802 (lr=8.1960e-05) (hash(x)=13536419)
520
+ 498 train 7.612567 (lr=8.1801e-05) (hash(x)=11620971)
521
+ 499 train 7.903674 (lr=8.1642e-05) (hash(x)=13326615)
522
+ 500 val loss 7.7021
523
+ 500 val perplexity 2212.8970
524
+ 500 train 7.669699 (lr=8.1482e-05) (hash(x)=13585589)
525
+ 501 train 7.847817 (lr=8.1322e-05) (hash(x)=12112694)
526
+ 502 train 7.828450 (lr=8.1161e-05) (hash(x)=11524363)
527
+ 503 train 7.666476 (lr=8.1000e-05) (hash(x)=9917297)
528
+ 504 train 7.700769 (lr=8.0838e-05) (hash(x)=13392970)
529
+ 505 train 7.990173 (lr=8.0676e-05) (hash(x)=12656639)
530
+ 506 train 7.220766 (lr=8.0513e-05) (hash(x)=8029433)
531
+ 507 train 6.909282 (lr=8.0349e-05) (hash(x)=6829764)
532
+ 508 train 7.312483 (lr=8.0186e-05) (hash(x)=12312323)
533
+ 509 train 7.850737 (lr=8.0021e-05) (hash(x)=12964320)
534
+ 510 train 7.546320 (lr=7.9856e-05) (hash(x)=12460686)
535
+ 511 train 7.818482 (lr=7.9691e-05) (hash(x)=12688412)
536
+ 512 train 7.594422 (lr=7.9525e-05) (hash(x)=11453733)
537
+ 513 train 8.076509 (lr=7.9359e-05) (hash(x)=14906439)
538
+ 514 train 7.413575 (lr=7.9192e-05) (hash(x)=10903202)
539
+ 515 train 7.448827 (lr=7.9025e-05) (hash(x)=13039934)
540
+ 516 train 7.778856 (lr=7.8858e-05) (hash(x)=11526002)
541
+ 517 train 7.743407 (lr=7.8689e-05) (hash(x)=12020357)
542
+ 518 train 8.064659 (lr=7.8521e-05) (hash(x)=17910848)
543
+ 519 train 7.678874 (lr=7.8352e-05) (hash(x)=12108991)
544
+ 520 train 7.868283 (lr=7.8182e-05) (hash(x)=12549643)
545
+ 521 train 7.914614 (lr=7.8012e-05) (hash(x)=12328329)
546
+ 522 train 7.790588 (lr=7.7842e-05) (hash(x)=12084164)
547
+ 523 train 7.981587 (lr=7.7671e-05) (hash(x)=14740147)
548
+ 524 train 7.648562 (lr=7.7500e-05) (hash(x)=10853071)
549
+ 525 train 7.713499 (lr=7.7328e-05) (hash(x)=11559413)
550
+ 526 train 7.854579 (lr=7.7156e-05) (hash(x)=14655032)
551
+ 527 train 7.750962 (lr=7.6984e-05) (hash(x)=12152216)
552
+ 528 train 7.880701 (lr=7.6811e-05) (hash(x)=14655442)
553
+ 529 train 7.634563 (lr=7.6637e-05) (hash(x)=12708922)
554
+ 530 train 7.819821 (lr=7.6463e-05) (hash(x)=12464208)
555
+ 531 train 7.911556 (lr=7.6289e-05) (hash(x)=14469749)
556
+ 532 train 7.987558 (lr=7.6115e-05) (hash(x)=12899435)
557
+ 533 train 7.760355 (lr=7.5940e-05) (hash(x)=12459368)
558
+ 534 train 7.919544 (lr=7.5764e-05) (hash(x)=11414287)
559
+ 535 train 7.785818 (lr=7.5588e-05) (hash(x)=11421483)
560
+ 536 train 7.645350 (lr=7.5412e-05) (hash(x)=12885164)
561
+ 537 train 7.631777 (lr=7.5235e-05) (hash(x)=12939881)
562
+ 538 train 7.777104 (lr=7.5058e-05) (hash(x)=11989344)
563
+ 539 train 7.895910 (lr=7.4881e-05) (hash(x)=12362234)
564
+ 540 train 7.751802 (lr=7.4703e-05) (hash(x)=11521741)
565
+ 541 train 7.659046 (lr=7.4525e-05) (hash(x)=12322333)
566
+ 542 train 7.442516 (lr=7.4346e-05) (hash(x)=10728195)
567
+ 543 train 8.004432 (lr=7.4167e-05) (hash(x)=13448155)
568
+ 544 train 7.746081 (lr=7.3988e-05) (hash(x)=13495082)
569
+ 545 train 7.813329 (lr=7.3808e-05) (hash(x)=12928941)
570
+ 546 train 7.759360 (lr=7.3628e-05) (hash(x)=11423948)
571
+ 547 train 7.565922 (lr=7.3448e-05) (hash(x)=10971430)
572
+ 548 train 7.565248 (lr=7.3267e-05) (hash(x)=11754720)
573
+ 549 train 7.691195 (lr=7.3086e-05) (hash(x)=13627629)
574
+ 550 val loss 7.6795
575
+ 550 val perplexity 2163.5596
576
+ 550 train 7.601290 (lr=7.2904e-05) (hash(x)=11853600)
577
+ 551 train 7.740928 (lr=7.2723e-05) (hash(x)=11905081)
578
+ 552 train 7.637296 (lr=7.2540e-05) (hash(x)=13718040)
579
+ 553 train 7.719368 (lr=7.2358e-05) (hash(x)=14339184)
580
+ 554 train 7.769798 (lr=7.2175e-05) (hash(x)=14022671)
581
+ 555 train 7.571931 (lr=7.1992e-05) (hash(x)=10917438)
582
+ 556 train 7.860756 (lr=7.1808e-05) (hash(x)=12631504)
583
+ 557 train 7.788311 (lr=7.1624e-05) (hash(x)=13804603)
584
+ 558 train 8.022074 (lr=7.1440e-05) (hash(x)=13445207)
585
+ 559 train 7.732583 (lr=7.1256e-05) (hash(x)=12290857)
586
+ 560 train 7.835542 (lr=7.1071e-05) (hash(x)=12852252)
587
+ 561 train 7.459764 (lr=7.0886e-05) (hash(x)=10211118)
588
+ 562 train 7.876356 (lr=7.0701e-05) (hash(x)=15793523)
589
+ 563 train 8.067602 (lr=7.0515e-05) (hash(x)=13407076)
590
+ 564 train 7.857517 (lr=7.0329e-05) (hash(x)=16676586)
591
+ 565 train 7.769851 (lr=7.0143e-05) (hash(x)=12249945)
592
+ 566 train 7.866961 (lr=6.9956e-05) (hash(x)=11660346)
593
+ 567 train 7.791004 (lr=6.9769e-05) (hash(x)=13234933)
594
+ 568 train 7.638211 (lr=6.9582e-05) (hash(x)=10823532)
595
+ 569 train 7.801475 (lr=6.9394e-05) (hash(x)=15194855)
596
+ 570 train 7.835317 (lr=6.9207e-05) (hash(x)=14843612)
597
+ 571 train 7.875848 (lr=6.9019e-05) (hash(x)=12892366)
598
+ 572 train 7.589688 (lr=6.8830e-05) (hash(x)=11056742)
599
+ 573 train 7.698580 (lr=6.8642e-05) (hash(x)=12609795)
600
+ 574 train 7.662716 (lr=6.8453e-05) (hash(x)=11560140)
601
+ 575 train 7.493523 (lr=6.8264e-05) (hash(x)=9768085)
602
+ 576 train 7.701514 (lr=6.8075e-05) (hash(x)=14009803)
603
+ 577 train 7.557343 (lr=6.7885e-05) (hash(x)=12252104)
604
+ 578 train 7.828304 (lr=6.7695e-05) (hash(x)=14304208)
605
+ 579 train 7.843575 (lr=6.7505e-05) (hash(x)=12692791)
606
+ 580 train 7.839464 (lr=6.7315e-05) (hash(x)=13214828)
607
+ 581 train 7.652492 (lr=6.7124e-05) (hash(x)=11957245)
608
+ 582 train 7.276093 (lr=6.6933e-05) (hash(x)=9605788)
609
+ 583 train 7.339360 (lr=6.6742e-05) (hash(x)=9528974)
610
+ 584 train 7.503553 (lr=6.6551e-05) (hash(x)=11036050)
611
+ 585 train 7.819581 (lr=6.6360e-05) (hash(x)=12231837)
612
+ 586 train 7.669895 (lr=6.6168e-05) (hash(x)=12052445)
613
+ 587 train 7.576273 (lr=6.5976e-05) (hash(x)=10032842)
614
+ 588 train 7.907766 (lr=6.5784e-05) (hash(x)=14951059)
615
+ 589 train 8.591299 (lr=6.5592e-05) (hash(x)=17178097)
616
+ 590 train 8.954432 (lr=6.5399e-05) (hash(x)=18009086)
617
+ 591 train 7.639175 (lr=6.5206e-05) (hash(x)=10720403)
618
+ 592 train 7.906215 (lr=6.5013e-05) (hash(x)=15019211)
619
+ 593 train 7.815034 (lr=6.4820e-05) (hash(x)=13541352)
620
+ 594 train 8.051396 (lr=6.4627e-05) (hash(x)=15227056)
621
+ 595 train 8.336821 (lr=6.4433e-05) (hash(x)=15187742)
622
+ 596 train 8.127712 (lr=6.4240e-05) (hash(x)=12858009)
623
+ 597 train 7.689843 (lr=6.4046e-05) (hash(x)=11971935)
624
+ 598 train 7.788243 (lr=6.3852e-05) (hash(x)=10047324)
625
+ 599 train 7.755938 (lr=6.3658e-05) (hash(x)=12874592)
626
+ 600 val loss 7.6632
627
+ 600 val perplexity 2128.6299
628
+ 600 train 7.741915 (lr=6.3463e-05) (hash(x)=11616298)
629
+ 601 train 7.723926 (lr=6.3269e-05) (hash(x)=12462802)
630
+ 602 train 7.703083 (lr=6.3074e-05) (hash(x)=12864255)
631
+ 603 train 7.853041 (lr=6.2879e-05) (hash(x)=13622545)
632
+ 604 train 7.741603 (lr=6.2684e-05) (hash(x)=14649855)
633
+ 605 train 8.092643 (lr=6.2489e-05) (hash(x)=14255974)
634
+ 606 train 7.953965 (lr=6.2294e-05) (hash(x)=14247081)
635
+ 607 train 7.585118 (lr=6.2098e-05) (hash(x)=11025404)
636
+ 608 train 7.948004 (lr=6.1903e-05) (hash(x)=14204393)
637
+ 609 train 7.790172 (lr=6.1707e-05) (hash(x)=11239636)
638
+ 610 train 7.809032 (lr=6.1511e-05) (hash(x)=12606825)
639
+ 611 train 7.861976 (lr=6.1315e-05) (hash(x)=15190123)
640
+ 612 train 7.960916 (lr=6.1119e-05) (hash(x)=15138106)
641
+ 613 train 7.979341 (lr=6.0923e-05) (hash(x)=11464046)
642
+ 614 train 7.863675 (lr=6.0726e-05) (hash(x)=11150867)
643
+ 615 train 7.850419 (lr=6.0530e-05) (hash(x)=12567779)
644
+ 616 train 7.724325 (lr=6.0333e-05) (hash(x)=11611202)
645
+ 617 train 7.834332 (lr=6.0137e-05) (hash(x)=16046446)
646
+ 618 train 7.670864 (lr=5.9940e-05) (hash(x)=11255030)
647
+ 619 train 7.670742 (lr=5.9743e-05) (hash(x)=12439519)
648
+ 620 train 7.635136 (lr=5.9546e-05) (hash(x)=13534296)
649
+ 621 train 7.451732 (lr=5.9349e-05) (hash(x)=11917658)
650
+ 622 train 7.787736 (lr=5.9152e-05) (hash(x)=11009063)
651
+ 623 train 7.684069 (lr=5.8955e-05) (hash(x)=11826091)
652
+ 624 train 8.209997 (lr=5.8758e-05) (hash(x)=14037695)
653
+ 625 train 7.760157 (lr=5.8560e-05) (hash(x)=10807703)
654
+ 626 train 7.604909 (lr=5.8363e-05) (hash(x)=10870729)
655
+ 627 train 7.717366 (lr=5.8165e-05) (hash(x)=14723866)
656
+ 628 train 7.852451 (lr=5.7968e-05) (hash(x)=13363055)
657
+ 629 train 8.010396 (lr=5.7770e-05) (hash(x)=11315230)
658
+ 630 train 8.763025 (lr=5.7573e-05) (hash(x)=11232721)
659
+ 631 train 8.203745 (lr=5.7375e-05) (hash(x)=11259652)
660
+ 632 train 8.332464 (lr=5.7177e-05) (hash(x)=12186888)
661
+ 633 train 8.606040 (lr=5.6979e-05) (hash(x)=11702582)
662
+ 634 train 8.132478 (lr=5.6782e-05) (hash(x)=12503047)
663
+ 635 train 7.767237 (lr=5.6584e-05) (hash(x)=11135951)
664
+ 636 train 7.794212 (lr=5.6386e-05) (hash(x)=11955498)
665
+ 637 train 7.876265 (lr=5.6188e-05) (hash(x)=11570885)
666
+ 638 train 7.703724 (lr=5.5990e-05) (hash(x)=12155844)
667
+ 639 train 7.734083 (lr=5.5792e-05) (hash(x)=14642293)
668
+ 640 train 7.524054 (lr=5.5594e-05) (hash(x)=13839875)
669
+ 641 train 7.609259 (lr=5.5396e-05) (hash(x)=13037070)
670
+ 642 train 7.765981 (lr=5.5198e-05) (hash(x)=14520991)
671
+ 643 train 7.673241 (lr=5.5000e-05) (hash(x)=12485426)
672
+ 644 train 7.682763 (lr=5.4802e-05) (hash(x)=13014970)
673
+ 645 train 7.900847 (lr=5.4604e-05) (hash(x)=15142358)
674
+ 646 train 7.519234 (lr=5.4406e-05) (hash(x)=10027296)
675
+ 647 train 7.631866 (lr=5.4208e-05) (hash(x)=12923132)
676
+ 648 train 7.822680 (lr=5.4010e-05) (hash(x)=12354083)
677
+ 649 train 7.925977 (lr=5.3812e-05) (hash(x)=13055942)
678
+ 650 val loss 7.6618
679
+ 650 val perplexity 2125.5110
680
+ 650 train 7.754591 (lr=5.3614e-05) (hash(x)=13291036)
681
+ 651 train 7.781103 (lr=5.3416e-05) (hash(x)=13559280)
682
+ 652 train 7.740664 (lr=5.3218e-05) (hash(x)=14086368)
683
+ 653 train 7.807534 (lr=5.3021e-05) (hash(x)=15313622)
684
+ 654 train 7.801032 (lr=5.2823e-05) (hash(x)=13812518)
685
+ 655 train 7.665193 (lr=5.2625e-05) (hash(x)=12169964)
686
+ 656 train 7.355139 (lr=5.2427e-05) (hash(x)=11896906)
687
+ 657 train 7.311297 (lr=5.2230e-05) (hash(x)=10902629)
688
+ 658 train 7.113808 (lr=5.2032e-05) (hash(x)=9987921)
689
+ 659 train 7.530096 (lr=5.1835e-05) (hash(x)=12000182)
690
+ 660 train 7.302106 (lr=5.1637e-05) (hash(x)=8940484)
691
+ 661 train 7.600711 (lr=5.1440e-05) (hash(x)=11345877)
692
+ 662 train 7.713447 (lr=5.1242e-05) (hash(x)=14081920)
693
+ 663 train 7.492238 (lr=5.1045e-05) (hash(x)=10948640)
694
+ 664 train 7.350751 (lr=5.0848e-05) (hash(x)=10980535)
695
+ 665 train 7.637386 (lr=5.0651e-05) (hash(x)=13037309)
696
+ 666 train 7.325884 (lr=5.0454e-05) (hash(x)=9375142)
697
+ 667 train 7.353055 (lr=5.0257e-05) (hash(x)=9666433)
698
+ 668 train 7.614735 (lr=5.0060e-05) (hash(x)=11623444)
699
+ 669 train 7.629223 (lr=4.9863e-05) (hash(x)=14049895)
700
+ 670 train 7.801353 (lr=4.9667e-05) (hash(x)=12689579)
701
+ 671 train 7.830479 (lr=4.9470e-05) (hash(x)=14103013)
702
+ 672 train 7.952965 (lr=4.9274e-05) (hash(x)=11809717)
703
+ 673 train 7.759763 (lr=4.9077e-05) (hash(x)=14058870)
704
+ 674 train 7.819689 (lr=4.8881e-05) (hash(x)=13738049)
705
+ 675 train 7.826292 (lr=4.8685e-05) (hash(x)=13553506)
706
+ 676 train 7.778317 (lr=4.8489e-05) (hash(x)=13526463)
707
+ 677 train 7.624836 (lr=4.8293e-05) (hash(x)=12902075)
708
+ 678 train 7.879182 (lr=4.8097e-05) (hash(x)=12466258)
709
+ 679 train 7.329772 (lr=4.7902e-05) (hash(x)=11135042)
710
+ 680 train 8.119125 (lr=4.7706e-05) (hash(x)=14920026)
711
+ 681 train 7.816418 (lr=4.7511e-05) (hash(x)=13650023)
712
+ 682 train 7.633660 (lr=4.7316e-05) (hash(x)=12991042)
713
+ 683 train 7.763315 (lr=4.7121e-05) (hash(x)=14930403)
714
+ 684 train 7.799829 (lr=4.6926e-05) (hash(x)=15386592)
715
+ 685 train 7.927387 (lr=4.6731e-05) (hash(x)=15485935)
716
+ 686 train 8.549294 (lr=4.6537e-05) (hash(x)=15487294)
717
+ 687 train 7.745435 (lr=4.6342e-05) (hash(x)=12404200)
718
+ 688 train 7.546225 (lr=4.6148e-05) (hash(x)=12070779)
719
+ 689 train 7.621696 (lr=4.5954e-05) (hash(x)=12392953)
720
+ 690 train 7.750398 (lr=4.5760e-05) (hash(x)=12331573)
721
+ 691 train 7.655634 (lr=4.5567e-05) (hash(x)=12280314)
722
+ 692 train 7.569410 (lr=4.5373e-05) (hash(x)=11383559)
723
+ 693 train 7.581852 (lr=4.5180e-05) (hash(x)=10900559)
724
+ 694 train 7.751139 (lr=4.4987e-05) (hash(x)=13353547)
725
+ 695 train 7.626840 (lr=4.4794e-05) (hash(x)=11534055)
726
+ 696 train 7.694957 (lr=4.4601e-05) (hash(x)=14797493)
727
+ 697 train 7.640746 (lr=4.4408e-05) (hash(x)=11997160)
728
+ 698 train 7.843601 (lr=4.4216e-05) (hash(x)=13151054)
729
+ 699 train 7.783576 (lr=4.4024e-05) (hash(x)=11308531)
730
+ 700 val loss 7.6410
731
+ 700 val perplexity 2081.8699
732
+ 700 train 7.722028 (lr=4.3832e-05) (hash(x)=13121368)
733
+ 701 train 7.670710 (lr=4.3640e-05) (hash(x)=12759751)
734
+ 702 train 7.732078 (lr=4.3449e-05) (hash(x)=13535835)
735
+ 703 train 7.623817 (lr=4.3258e-05) (hash(x)=10914085)
736
+ 704 train 7.663767 (lr=4.3067e-05) (hash(x)=10617131)
737
+ 705 train 7.987453 (lr=4.2876e-05) (hash(x)=12776027)
738
+ 706 train 7.705152 (lr=4.2685e-05) (hash(x)=11255028)
739
+ 707 train 7.561122 (lr=4.2495e-05) (hash(x)=12357751)
740
+ 708 train 7.660600 (lr=4.2305e-05) (hash(x)=12772363)
741
+ 709 train 7.570772 (lr=4.2115e-05) (hash(x)=11040798)
742
+ 710 train 7.948753 (lr=4.1925e-05) (hash(x)=13528424)
743
+ 711 train 7.979651 (lr=4.1736e-05) (hash(x)=14570713)
744
+ 712 train 7.938939 (lr=4.1547e-05) (hash(x)=13617082)
745
+ 713 train 7.746921 (lr=4.1358e-05) (hash(x)=11492564)
746
+ 714 train 7.636271 (lr=4.1170e-05) (hash(x)=13806329)
747
+ 715 train 7.731492 (lr=4.0981e-05) (hash(x)=12212797)
748
+ 716 train 7.826818 (lr=4.0793e-05) (hash(x)=12715617)
749
+ 717 train 7.669985 (lr=4.0606e-05) (hash(x)=13496879)
750
+ 718 train 7.887058 (lr=4.0418e-05) (hash(x)=12774919)
751
+ 719 train 7.637987 (lr=4.0231e-05) (hash(x)=11680329)
752
+ 720 train 7.709901 (lr=4.0044e-05) (hash(x)=12001582)
753
+ 721 train 7.522655 (lr=3.9857e-05) (hash(x)=11948575)
754
+ 722 train 7.703846 (lr=3.9671e-05) (hash(x)=11246553)
755
+ 723 train 7.644125 (lr=3.9485e-05) (hash(x)=13116772)
756
+ 724 train 7.780936 (lr=3.9299e-05) (hash(x)=13475222)
757
+ 725 train 8.207045 (lr=3.9114e-05) (hash(x)=15527273)
758
+ 726 train 7.374922 (lr=3.8929e-05) (hash(x)=10143298)
759
+ 727 train 7.494412 (lr=3.8744e-05) (hash(x)=12437096)
760
+ 728 train 7.886576 (lr=3.8560e-05) (hash(x)=13259177)
761
+ 729 train 7.756798 (lr=3.8376e-05) (hash(x)=12518287)
762
+ 730 train 7.931480 (lr=3.8192e-05) (hash(x)=12379439)
763
+ 731 train 7.432437 (lr=3.8008e-05) (hash(x)=11220338)
764
+ 732 train 7.567032 (lr=3.7825e-05) (hash(x)=12710818)
765
+ 733 train 7.492830 (lr=3.7642e-05) (hash(x)=12034192)
766
+ 734 train 8.049908 (lr=3.7460e-05) (hash(x)=17970060)
767
+ 735 train 7.972324 (lr=3.7277e-05) (hash(x)=12654843)
768
+ 736 train 7.666026 (lr=3.7096e-05) (hash(x)=13094012)
769
+ 737 train 8.015420 (lr=3.6914e-05) (hash(x)=16116174)
770
+ 738 train 7.780063 (lr=3.6733e-05) (hash(x)=14740205)
771
+ 739 train 7.769230 (lr=3.6552e-05) (hash(x)=13505076)
772
+ 740 train 7.607024 (lr=3.6372e-05) (hash(x)=13104341)
773
+ 741 train 7.835599 (lr=3.6192e-05) (hash(x)=11820607)
774
+ 742 train 8.214757 (lr=3.6012e-05) (hash(x)=16336439)
775
+ 743 train 7.440488 (lr=3.5833e-05) (hash(x)=10606850)
776
+ 744 train 7.569521 (lr=3.5654e-05) (hash(x)=12134587)
777
+ 745 train 7.750698 (lr=3.5475e-05) (hash(x)=13759635)
778
+ 746 train 7.689027 (lr=3.5297e-05) (hash(x)=12898464)
779
+ 747 train 7.742244 (lr=3.5119e-05) (hash(x)=15357227)
780
+ 748 train 7.783304 (lr=3.4942e-05) (hash(x)=13808568)
781
+ 749 train 7.623234 (lr=3.4765e-05) (hash(x)=11597062)
782
+ 750 val loss 7.6364
783
+ 750 val perplexity 2072.2004
784
+ 750 train 7.791523 (lr=3.4588e-05) (hash(x)=11255261)
785
+ 751 train 7.561778 (lr=3.4412e-05) (hash(x)=11281537)
786
+ 752 train 7.265604 (lr=3.4236e-05) (hash(x)=9979447)
787
+ 753 train 7.402912 (lr=3.4060e-05) (hash(x)=13294828)
788
+ 754 train 8.209104 (lr=3.3885e-05) (hash(x)=15954206)
789
+ 755 train 8.245211 (lr=3.3711e-05) (hash(x)=13253249)
790
+ 756 train 7.588990 (lr=3.3537e-05) (hash(x)=10309283)
791
+ 757 train 8.061670 (lr=3.3363e-05) (hash(x)=16101250)
792
+ 758 train 7.985057 (lr=3.3189e-05) (hash(x)=13933994)
793
+ 759 train 8.019863 (lr=3.3016e-05) (hash(x)=12363856)
794
+ 760 train 7.682524 (lr=3.2844e-05) (hash(x)=11923641)
795
+ 761 train 7.770922 (lr=3.2672e-05) (hash(x)=12206540)
796
+ 762 train 7.983355 (lr=3.2500e-05) (hash(x)=14025490)
797
+ 763 train 7.826716 (lr=3.2329e-05) (hash(x)=13580623)
798
+ 764 train 7.855828 (lr=3.2158e-05) (hash(x)=13885651)
799
+ 765 train 7.912757 (lr=3.1988e-05) (hash(x)=12509602)
800
+ 766 train 7.668511 (lr=3.1818e-05) (hash(x)=12966645)
801
+ 767 train 8.271053 (lr=3.1648e-05) (hash(x)=14001265)
802
+ 768 train 7.820420 (lr=3.1479e-05) (hash(x)=13743669)
803
+ 769 train 7.548445 (lr=3.1311e-05) (hash(x)=14290431)
804
+ 770 train 7.645972 (lr=3.1142e-05) (hash(x)=11211355)
805
+ 771 train 7.668486 (lr=3.0975e-05) (hash(x)=13045278)
806
+ 772 train 7.765580 (lr=3.0808e-05) (hash(x)=15088674)
807
+ 773 train 7.788311 (lr=3.0641e-05) (hash(x)=13371383)
808
+ 774 train 7.876847 (lr=3.0475e-05) (hash(x)=12504111)
809
+ 775 train 7.569955 (lr=3.0309e-05) (hash(x)=10444420)
810
+ 776 train 7.405804 (lr=3.0144e-05) (hash(x)=10607988)
811
+ 777 train 7.713726 (lr=2.9979e-05) (hash(x)=12732892)
812
+ 778 train 7.765675 (lr=2.9814e-05) (hash(x)=12650059)
813
+ 779 train 7.895899 (lr=2.9651e-05) (hash(x)=11832964)
814
+ 780 train 7.811251 (lr=2.9487e-05) (hash(x)=14574038)
815
+ 781 train 7.828371 (lr=2.9324e-05) (hash(x)=11477958)
816
+ 782 train 7.622464 (lr=2.9162e-05) (hash(x)=12979425)
817
+ 783 train 7.698474 (lr=2.9000e-05) (hash(x)=12249732)
818
+ 784 train 7.832039 (lr=2.8839e-05) (hash(x)=14867563)
819
+ 785 train 7.721405 (lr=2.8678e-05) (hash(x)=12132933)
820
+ 786 train 7.645487 (lr=2.8518e-05) (hash(x)=13535569)
821
+ 787 train 7.763117 (lr=2.8358e-05) (hash(x)=12307821)
822
+ 788 train 7.271033 (lr=2.8199e-05) (hash(x)=11278357)
823
+ 789 train 7.229966 (lr=2.8040e-05) (hash(x)=11266264)
824
+ 790 train 7.480228 (lr=2.7881e-05) (hash(x)=12955233)
825
+ 791 train 7.791690 (lr=2.7724e-05) (hash(x)=13803785)
826
+ 792 train 7.667414 (lr=2.7566e-05) (hash(x)=14264824)
827
+ 793 train 7.801067 (lr=2.7410e-05) (hash(x)=13256773)
828
+ 794 train 7.579760 (lr=2.7254e-05) (hash(x)=12650537)
829
+ 795 train 7.585497 (lr=2.7098e-05) (hash(x)=12493316)
830
+ 796 train 7.659342 (lr=2.6943e-05) (hash(x)=11836158)
831
+ 797 train 7.505648 (lr=2.6788e-05) (hash(x)=12991346)
832
+ 798 train 8.081731 (lr=2.6634e-05) (hash(x)=13756621)
833
+ 799 train 7.357153 (lr=2.6481e-05) (hash(x)=11689978)
834
+ 800 val loss 7.6292
835
+ 800 val perplexity 2057.4211
836
+ 800 train 7.631804 (lr=2.6328e-05) (hash(x)=10365069)
837
+ 801 train 7.587431 (lr=2.6176e-05) (hash(x)=12056273)
838
+ 802 train 7.762430 (lr=2.6024e-05) (hash(x)=14348732)
839
+ 803 train 7.778242 (lr=2.5873e-05) (hash(x)=11906763)
840
+ 804 train 7.796702 (lr=2.5722e-05) (hash(x)=12437638)
841
+ 805 train 7.718960 (lr=2.5572e-05) (hash(x)=12621559)
842
+ 806 train 7.492385 (lr=2.5423e-05) (hash(x)=11302699)
843
+ 807 train 7.935616 (lr=2.5274e-05) (hash(x)=13865869)
844
+ 808 train 7.680390 (lr=2.5125e-05) (hash(x)=12380967)
845
+ 809 train 7.794242 (lr=2.4978e-05) (hash(x)=11425874)
846
+ 810 train 7.639746 (lr=2.4830e-05) (hash(x)=12723475)
847
+ 811 train 8.425303 (lr=2.4684e-05) (hash(x)=16513375)
848
+ 812 train 7.843686 (lr=2.4538e-05) (hash(x)=13869844)
849
+ 813 train 8.002068 (lr=2.4392e-05) (hash(x)=14492920)
850
+ 814 train 7.876977 (lr=2.4247e-05) (hash(x)=13793114)
851
+ 815 train 7.704354 (lr=2.4103e-05) (hash(x)=13030618)
852
+ 816 train 7.793812 (lr=2.3959e-05) (hash(x)=14792501)
853
+ 817 train 7.575655 (lr=2.3816e-05) (hash(x)=12816692)
854
+ 818 train 7.789332 (lr=2.3674e-05) (hash(x)=12569165)
855
+ 819 train 8.014382 (lr=2.3532e-05) (hash(x)=13701949)
856
+ 820 train 7.671343 (lr=2.3391e-05) (hash(x)=12255108)
857
+ 821 train 7.669661 (lr=2.3250e-05) (hash(x)=12904618)
858
+ 822 train 7.676630 (lr=2.3110e-05) (hash(x)=12258538)
859
+ 823 train 7.950303 (lr=2.2971e-05) (hash(x)=14299833)
860
+ 824 train 7.550400 (lr=2.2832e-05) (hash(x)=10822163)
861
+ 825 train 7.705056 (lr=2.2694e-05) (hash(x)=12586511)
862
+ 826 train 7.681914 (lr=2.2556e-05) (hash(x)=13939370)
863
+ 827 train 7.682074 (lr=2.2420e-05) (hash(x)=11818854)
864
+ 828 train 7.732458 (lr=2.2283e-05) (hash(x)=13752960)
865
+ 829 train 7.732020 (lr=2.2148e-05) (hash(x)=13063115)
866
+ 830 train 7.676750 (lr=2.2013e-05) (hash(x)=12817199)
867
+ 831 train 7.599615 (lr=2.1878e-05) (hash(x)=12908875)
868
+ 832 train 7.923879 (lr=2.1745e-05) (hash(x)=13871627)
869
+ 833 train 7.770768 (lr=2.1612e-05) (hash(x)=12627550)
870
+ 834 train 7.576340 (lr=2.1479e-05) (hash(x)=11338055)
871
+ 835 train 7.615951 (lr=2.1347e-05) (hash(x)=12634656)
872
+ 836 train 7.663651 (lr=2.1216e-05) (hash(x)=12416256)
873
+ 837 train 7.719417 (lr=2.1086e-05) (hash(x)=13466941)
874
+ 838 train 7.819789 (lr=2.0956e-05) (hash(x)=14648030)
875
+ 839 train 8.041725 (lr=2.0827e-05) (hash(x)=13859304)
876
+ 840 train 7.887707 (lr=2.0698e-05) (hash(x)=13483846)
877
+ 841 train 7.811630 (lr=2.0570e-05) (hash(x)=9637450)
878
+ 842 train 7.579254 (lr=2.0443e-05) (hash(x)=11672159)
879
+ 843 train 7.869338 (lr=2.0317e-05) (hash(x)=14940204)
880
+ 844 train 7.669798 (lr=2.0191e-05) (hash(x)=14028640)
881
+ 845 train 7.726581 (lr=2.0066e-05) (hash(x)=13686717)
882
+ 846 train 7.680332 (lr=1.9941e-05) (hash(x)=11037190)
883
+ 847 train 7.861037 (lr=1.9818e-05) (hash(x)=14396964)
884
+ 848 train 7.285347 (lr=1.9694e-05) (hash(x)=10766097)
885
+ 849 train 7.553684 (lr=1.9572e-05) (hash(x)=12949082)
886
+ 850 val loss 7.6206
887
+ 850 val perplexity 2039.7268
888
+ 850 train 7.520333 (lr=1.9450e-05) (hash(x)=13743434)
889
+ 851 train 7.718529 (lr=1.9329e-05) (hash(x)=13577749)
890
+ 852 train 7.736487 (lr=1.9209e-05) (hash(x)=12752830)
891
+ 853 train 7.867672 (lr=1.9089e-05) (hash(x)=13031728)
892
+ 854 train 7.944017 (lr=1.8970e-05) (hash(x)=13630938)
893
+ 855 train 7.706992 (lr=1.8852e-05) (hash(x)=11269801)
894
+ 856 train 7.588866 (lr=1.8734e-05) (hash(x)=12724302)
895
+ 857 train 7.462291 (lr=1.8618e-05) (hash(x)=11106556)
896
+ 858 train 7.385964 (lr=1.8501e-05) (hash(x)=12499441)
897
+ 859 train 7.535069 (lr=1.8386e-05) (hash(x)=11222906)
898
+ 860 train 7.682183 (lr=1.8271e-05) (hash(x)=10586708)
899
+ 861 train 7.831712 (lr=1.8157e-05) (hash(x)=14052269)
900
+ 862 train 7.673687 (lr=1.8044e-05) (hash(x)=13384690)
901
+ 863 train 7.460972 (lr=1.7931e-05) (hash(x)=14486071)
902
+ 864 train 7.680599 (lr=1.7819e-05) (hash(x)=11976037)
903
+ 865 train 7.717909 (lr=1.7708e-05) (hash(x)=13026662)
904
+ 866 train 7.693939 (lr=1.7598e-05) (hash(x)=13443462)
905
+ 867 train 7.677587 (lr=1.7488e-05) (hash(x)=11065744)
906
+ 868 train 7.715825 (lr=1.7379e-05) (hash(x)=12404324)
907
+ 869 train 7.773833 (lr=1.7271e-05) (hash(x)=13318947)
908
+ 870 train 7.913719 (lr=1.7163e-05) (hash(x)=14278749)
909
+ 871 train 7.606133 (lr=1.7056e-05) (hash(x)=11997361)
910
+ 872 train 7.979546 (lr=1.6950e-05) (hash(x)=16591908)
911
+ 873 train 7.647080 (lr=1.6845e-05) (hash(x)=12437178)
912
+ 874 train 7.743247 (lr=1.6740e-05) (hash(x)=11441868)
913
+ 875 train 7.609517 (lr=1.6636e-05) (hash(x)=12705537)
914
+ 876 train 7.744054 (lr=1.6533e-05) (hash(x)=14200097)
915
+ 877 train 7.825002 (lr=1.6431e-05) (hash(x)=15790088)
916
+ 878 train 7.600862 (lr=1.6329e-05) (hash(x)=11804370)
917
+ 879 train 7.480116 (lr=1.6228e-05) (hash(x)=11010336)
918
+ 880 train 7.895975 (lr=1.6128e-05) (hash(x)=10935364)
919
+ 881 train 7.674376 (lr=1.6029e-05) (hash(x)=11401216)
920
+ 882 train 7.755695 (lr=1.5930e-05) (hash(x)=12837899)
921
+ 883 train 7.683318 (lr=1.5832e-05) (hash(x)=15106564)
922
+ 884 train 7.746137 (lr=1.5735e-05) (hash(x)=12856376)
923
+ 885 train 7.687259 (lr=1.5639e-05) (hash(x)=13019227)
924
+ 886 train 7.798133 (lr=1.5543e-05) (hash(x)=13294195)
925
+ 887 train 7.616749 (lr=1.5449e-05) (hash(x)=10864248)
926
+ 888 train 7.457390 (lr=1.5354e-05) (hash(x)=10779619)
927
+ 889 train 7.757562 (lr=1.5261e-05) (hash(x)=12107995)
928
+ 890 train 7.443310 (lr=1.5169e-05) (hash(x)=11879879)
929
+ 891 train 7.662545 (lr=1.5077e-05) (hash(x)=13284954)
930
+ 892 train 7.903453 (lr=1.4986e-05) (hash(x)=12895131)
931
+ 893 train 7.657013 (lr=1.4896e-05) (hash(x)=12430135)
932
+ 894 train 7.753739 (lr=1.4806e-05) (hash(x)=12661748)
933
+ 895 train 7.642135 (lr=1.4718e-05) (hash(x)=11906151)
934
+ 896 train 7.621687 (lr=1.4630e-05) (hash(x)=11476689)
935
+ 897 train 7.409808 (lr=1.4543e-05) (hash(x)=11697970)
936
+ 898 train 7.538272 (lr=1.4456e-05) (hash(x)=10948364)
937
+ 899 train 7.324598 (lr=1.4371e-05) (hash(x)=8629295)
938
+ 900 val loss 7.6124
939
+ 900 val perplexity 2023.1067
940
+ 900 train 7.604218 (lr=1.4286e-05) (hash(x)=12316587)
941
+ 901 train 7.696288 (lr=1.4202e-05) (hash(x)=13977261)
942
+ 902 train 7.687957 (lr=1.4119e-05) (hash(x)=12479583)
943
+ 903 train 7.765285 (lr=1.4037e-05) (hash(x)=15166897)
944
+ 904 train 7.781142 (lr=1.3955e-05) (hash(x)=12472550)
945
+ 905 train 7.831698 (lr=1.3874e-05) (hash(x)=13444474)
946
+ 906 train 7.866756 (lr=1.3794e-05) (hash(x)=13849646)
947
+ 907 train 7.858720 (lr=1.3715e-05) (hash(x)=13108020)
948
+ 908 train 7.627274 (lr=1.3637e-05) (hash(x)=14061108)
949
+ 909 train 7.737940 (lr=1.3559e-05) (hash(x)=12559086)
950
+ 910 train 7.572672 (lr=1.3482e-05) (hash(x)=11562580)
951
+ 911 train 7.657227 (lr=1.3407e-05) (hash(x)=12338963)
952
+ 912 train 7.256672 (lr=1.3331e-05) (hash(x)=10604668)
953
+ 913 train 7.643756 (lr=1.3257e-05) (hash(x)=10411127)
954
+ 914 train 7.667354 (lr=1.3183e-05) (hash(x)=12560107)
955
+ 915 train 7.943704 (lr=1.3111e-05) (hash(x)=17176736)
956
+ 916 train 7.547956 (lr=1.3039e-05) (hash(x)=13034244)
957
+ 917 train 7.614550 (lr=1.2968e-05) (hash(x)=13586859)
958
+ 918 train 7.629796 (lr=1.2897e-05) (hash(x)=14825083)
959
+ 919 train 7.842947 (lr=1.2828e-05) (hash(x)=13235312)
960
+ 920 train 7.608109 (lr=1.2759e-05) (hash(x)=12555636)
961
+ 921 train 7.532580 (lr=1.2691e-05) (hash(x)=10862031)
962
+ 922 train 7.658683 (lr=1.2624e-05) (hash(x)=12129400)
963
+ 923 train 7.793427 (lr=1.2558e-05) (hash(x)=12498316)
964
+ 924 train 7.783417 (lr=1.2493e-05) (hash(x)=13128811)
965
+ 925 train 7.607649 (lr=1.2428e-05) (hash(x)=15109534)
966
+ 926 train 7.401073 (lr=1.2364e-05) (hash(x)=10731002)
967
+ 927 train 7.581272 (lr=1.2301e-05) (hash(x)=11031343)
968
+ 928 train 7.543287 (lr=1.2239e-05) (hash(x)=11552376)
969
+ 929 train 7.655213 (lr=1.2178e-05) (hash(x)=13030101)
970
+ 930 train 7.398396 (lr=1.2118e-05) (hash(x)=13250524)
971
+ 931 train 7.728891 (lr=1.2058e-05) (hash(x)=12787203)
972
+ 932 train 7.652882 (lr=1.1999e-05) (hash(x)=12056192)
973
+ 933 train 7.736353 (lr=1.1941e-05) (hash(x)=13192715)
974
+ 934 train 7.594948 (lr=1.1884e-05) (hash(x)=13092101)
975
+ 935 train 7.802969 (lr=1.1828e-05) (hash(x)=14597757)
976
+ 936 train 7.806710 (lr=1.1772e-05) (hash(x)=13880794)
977
+ 937 train 7.512589 (lr=1.1718e-05) (hash(x)=11548718)
978
+ 938 train 7.730990 (lr=1.1664e-05) (hash(x)=11354422)
979
+ 939 train 7.643562 (lr=1.1611e-05) (hash(x)=12017158)
980
+ 940 train 7.863441 (lr=1.1559e-05) (hash(x)=15839756)
981
+ 941 train 7.549936 (lr=1.1508e-05) (hash(x)=10164479)
982
+ 942 train 7.560804 (lr=1.1457e-05) (hash(x)=13779450)
983
+ 943 train 7.540797 (lr=1.1408e-05) (hash(x)=13571036)
984
+ 944 train 7.496755 (lr=1.1359e-05) (hash(x)=12178618)
985
+ 945 train 7.715475 (lr=1.1311e-05) (hash(x)=11609529)
986
+ 946 train 7.704891 (lr=1.1264e-05) (hash(x)=14297351)
987
+ 947 train 7.662954 (lr=1.1218e-05) (hash(x)=12289875)
988
+ 948 train 7.908993 (lr=1.1173e-05) (hash(x)=13540167)
989
+ 949 train 7.624613 (lr=1.1128e-05) (hash(x)=13086437)
990
+ 950 val loss 7.6073
991
+ 950 val perplexity 2012.8827
992
+ 950 train 7.851632 (lr=1.1085e-05) (hash(x)=15073523)
993
+ 951 train 7.964910 (lr=1.1042e-05) (hash(x)=13225100)
994
+ 952 train 7.534026 (lr=1.1000e-05) (hash(x)=11311171)
995
+ 953 train 7.476605 (lr=1.0959e-05) (hash(x)=12863127)
996
+ 954 train 7.890977 (lr=1.0919e-05) (hash(x)=15288759)
997
+ 955 train 7.427927 (lr=1.0879e-05) (hash(x)=11162887)
998
+ 956 train 7.737081 (lr=1.0841e-05) (hash(x)=9933157)
999
+ 957 train 7.776907 (lr=1.0803e-05) (hash(x)=12364743)
1000
+ 958 train 7.490522 (lr=1.0766e-05) (hash(x)=10723631)
1001
+ 959 train 7.562864 (lr=1.0730e-05) (hash(x)=12281916)
1002
+ 960 train 7.565658 (lr=1.0695e-05) (hash(x)=12131510)
1003
+ 961 train 7.624058 (lr=1.0661e-05) (hash(x)=12249214)
1004
+ 962 train 7.771938 (lr=1.0628e-05) (hash(x)=12301339)
1005
+ 963 train 7.618944 (lr=1.0595e-05) (hash(x)=12348477)
1006
+ 964 train 7.672257 (lr=1.0563e-05) (hash(x)=13002952)
1007
+ 965 train 7.763545 (lr=1.0533e-05) (hash(x)=12468719)
1008
+ 966 train 7.736159 (lr=1.0503e-05) (hash(x)=11222266)
1009
+ 967 train 7.708023 (lr=1.0474e-05) (hash(x)=13634084)
1010
+ 968 train 7.701601 (lr=1.0445e-05) (hash(x)=14045062)
1011
+ 969 train 7.629251 (lr=1.0418e-05) (hash(x)=11740722)
1012
+ 970 train 7.947063 (lr=1.0391e-05) (hash(x)=14816355)
1013
+ 971 train 7.587160 (lr=1.0366e-05) (hash(x)=11378506)
1014
+ 972 train 7.420127 (lr=1.0341e-05) (hash(x)=10458922)
1015
+ 973 train 7.648345 (lr=1.0317e-05) (hash(x)=13417634)
1016
+ 974 train 7.760894 (lr=1.0294e-05) (hash(x)=13825872)
1017
+ 975 train 7.684702 (lr=1.0272e-05) (hash(x)=11801649)
1018
+ 976 train 7.861512 (lr=1.0251e-05) (hash(x)=11686944)
1019
+ 977 train 7.563744 (lr=1.0230e-05) (hash(x)=10917325)
1020
+ 978 train 7.454514 (lr=1.0211e-05) (hash(x)=11207331)
1021
+ 979 train 7.531972 (lr=1.0192e-05) (hash(x)=10552788)
1022
+ 980 train 7.320363 (lr=1.0174e-05) (hash(x)=8947573)
1023
+ 981 train 7.170998 (lr=1.0157e-05) (hash(x)=9421716)
1024
+ 982 train 7.491410 (lr=1.0141e-05) (hash(x)=13067279)
1025
+ 983 train 7.750579 (lr=1.0126e-05) (hash(x)=11746289)
1026
+ 984 train 7.740094 (lr=1.0111e-05) (hash(x)=12788451)
1027
+ 985 train 7.569104 (lr=1.0098e-05) (hash(x)=9589864)
1028
+ 986 train 7.556940 (lr=1.0085e-05) (hash(x)=12600294)
1029
+ 987 train 7.653209 (lr=1.0074e-05) (hash(x)=13826293)
1030
+ 988 train 7.746078 (lr=1.0063e-05) (hash(x)=14679370)
1031
+ 989 train 7.826417 (lr=1.0053e-05) (hash(x)=12720603)
1032
+ 990 train 7.639474 (lr=1.0044e-05) (hash(x)=12387624)
1033
+ 991 train 7.633626 (lr=1.0035e-05) (hash(x)=10783313)
1034
+ 992 train 7.679803 (lr=1.0028e-05) (hash(x)=11773378)
1035
+ 993 train 7.681911 (lr=1.0021e-05) (hash(x)=13077539)
1036
+ 994 train 7.658948 (lr=1.0016e-05) (hash(x)=12541500)
1037
+ 995 train 7.810250 (lr=1.0011e-05) (hash(x)=12913270)
1038
+ 996 train 7.769026 (lr=1.0007e-05) (hash(x)=11426867)
1039
+ 997 train 7.586505 (lr=1.0004e-05) (hash(x)=11228160)
1040
+ 998 train 7.412054 (lr=1.0002e-05) (hash(x)=11408061)
1041
+ 999 val loss 7.6056
1042
+ 999 val perplexity 2009.4668
1043
+ 999 train 7.766582 (lr=1.0000e-05) (hash(x)=14212467)
lr1e-4_total_batch_size5120_seq_len256/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07477d76dd8cfd9541db73cffc1338541c0cec2c12fbcd6fa4443bc0da9ae104
3
+ size 38587970
lr1e-4_total_batch_size5120_seq_len256/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:945f374f099f48901ad463d74123bee12863b9fbc1bf32b054cdd329eb5ed951
3
+ size 70895430