andrew-healey commited on
Commit
595416c
·
verified ·
1 Parent(s): 1dd5f6c

Upload folder using huggingface_hub

Browse files
lr1.5e-3_total_batch_size20480_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr1.5e-3_total_batch_size20480_seq_len128", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 128, "batch_size": null, "total_batch_size": 20480, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0015, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1.5e-3_total_batch_size20480_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8513131d2bf732f8f6d97ff82d32c505b2a2f3f9f2247bac315bb09ce0e1f97
3
+ size 964
lr1.5e-3_total_batch_size20480_seq_len128/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.6870
3
+ 0 val perplexity 119016.5000
4
+ 0 train 11.700421 (lr=5.2448e-06) (hash(x)=5113283)
5
+ 1 train 11.656199 (lr=1.0490e-05) (hash(x)=5534142)
6
+ 2 train 11.654834 (lr=1.5734e-05) (hash(x)=6763231)
7
+ 3 train 11.592807 (lr=2.0979e-05) (hash(x)=6328012)
8
+ 4 train 11.545564 (lr=2.6224e-05) (hash(x)=7052161)
9
+ 5 train 11.430749 (lr=3.1469e-05) (hash(x)=5556098)
10
+ 6 train 11.355437 (lr=3.6713e-05) (hash(x)=4696943)
11
+ 7 train 11.196596 (lr=4.1958e-05) (hash(x)=6219138)
12
+ 8 train 11.065966 (lr=4.7203e-05) (hash(x)=6049878)
13
+ 9 train 10.958699 (lr=5.2448e-05) (hash(x)=5491972)
14
+ 10 train 10.830196 (lr=5.7692e-05) (hash(x)=8075458)
15
+ 11 train 10.709135 (lr=6.2937e-05) (hash(x)=5917741)
16
+ 12 train 10.648876 (lr=6.8182e-05) (hash(x)=5750403)
17
+ 13 train 10.577779 (lr=7.3427e-05) (hash(x)=6112458)
18
+ 14 train 10.522706 (lr=7.8671e-05) (hash(x)=5806490)
19
+ 15 train 10.427026 (lr=8.3916e-05) (hash(x)=6568379)
20
+ 16 train 10.360064 (lr=8.9161e-05) (hash(x)=7212403)
21
+ 17 train 10.332555 (lr=9.4406e-05) (hash(x)=7386082)
22
+ 18 train 10.297807 (lr=9.9650e-05) (hash(x)=6359743)
23
+ 19 train 10.256972 (lr=1.0490e-04) (hash(x)=6151674)
24
+ 20 train 10.270123 (lr=1.1014e-04) (hash(x)=7168704)
25
+ 21 train 10.248352 (lr=1.1538e-04) (hash(x)=7859060)
26
+ 22 train 10.173741 (lr=1.2063e-04) (hash(x)=6976239)
27
+ 23 train 10.182279 (lr=1.2587e-04) (hash(x)=6391506)
28
+ 24 train 10.174531 (lr=1.3112e-04) (hash(x)=5245999)
29
+ 25 train 10.114564 (lr=1.3636e-04) (hash(x)=5985675)
30
+ 26 train 10.014374 (lr=1.4161e-04) (hash(x)=4355733)
31
+ 27 train 10.180069 (lr=1.4685e-04) (hash(x)=6984772)
32
+ 28 train 9.990271 (lr=1.5210e-04) (hash(x)=6244096)
33
+ 29 train 9.989829 (lr=1.5734e-04) (hash(x)=6603717)
34
+ 30 train 9.847124 (lr=1.6259e-04) (hash(x)=5669522)
35
+ 31 train 9.777426 (lr=1.6783e-04) (hash(x)=6190579)
36
+ 32 train 9.803539 (lr=1.7308e-04) (hash(x)=4912983)
37
+ 33 train 9.629395 (lr=1.7832e-04) (hash(x)=7276303)
38
+ 34 train 9.661088 (lr=1.8357e-04) (hash(x)=5479947)
39
+ 35 train 9.722701 (lr=1.8881e-04) (hash(x)=6630378)
40
+ 36 train 9.570948 (lr=1.9406e-04) (hash(x)=8097010)
41
+ 37 train 9.622107 (lr=1.9930e-04) (hash(x)=9989744)
42
+ 38 train 9.425660 (lr=2.0455e-04) (hash(x)=6694932)
43
+ 39 train 9.280789 (lr=2.0979e-04) (hash(x)=6317894)
44
+ 40 train 9.084509 (lr=2.1503e-04) (hash(x)=5046327)
45
+ 41 train 9.103763 (lr=2.2028e-04) (hash(x)=7947703)
46
+ 42 train 8.868105 (lr=2.2552e-04) (hash(x)=5916068)
47
+ 43 train 8.919394 (lr=2.3077e-04) (hash(x)=5777498)
48
+ 44 train 8.794020 (lr=2.3601e-04) (hash(x)=5753092)
49
+ 45 train 8.692651 (lr=2.4126e-04) (hash(x)=7723226)
50
+ 46 train 8.642941 (lr=2.4650e-04) (hash(x)=7007016)
51
+ 47 train 8.470806 (lr=2.5175e-04) (hash(x)=6491229)
52
+ 48 train 8.488767 (lr=2.5699e-04) (hash(x)=6390752)
53
+ 49 train 8.663141 (lr=2.6224e-04) (hash(x)=6019389)
54
+ 50 val loss 8.4014
55
+ 50 val perplexity 4453.0786
56
+ 50 train 8.432234 (lr=2.6748e-04) (hash(x)=7662067)
57
+ 51 train 8.372618 (lr=2.7273e-04) (hash(x)=4648609)
58
+ 52 train 8.232944 (lr=2.7797e-04) (hash(x)=4725966)
59
+ 53 train 8.143088 (lr=2.8322e-04) (hash(x)=5558355)
60
+ 54 train 8.215803 (lr=2.8846e-04) (hash(x)=6524933)
61
+ 55 train 8.115613 (lr=2.9371e-04) (hash(x)=6369326)
62
+ 56 train 8.116792 (lr=2.9895e-04) (hash(x)=6022625)
63
+ 57 train 7.987829 (lr=3.0420e-04) (hash(x)=4906853)
64
+ 58 train 7.746727 (lr=3.0944e-04) (hash(x)=5589269)
65
+ 59 train 7.945351 (lr=3.1469e-04) (hash(x)=6437376)
66
+ 60 train 8.118293 (lr=3.1993e-04) (hash(x)=7304892)
67
+ 61 train 8.057110 (lr=3.2517e-04) (hash(x)=7084093)
68
+ 62 train 8.070441 (lr=3.3042e-04) (hash(x)=5908178)
69
+ 63 train 7.866910 (lr=3.3566e-04) (hash(x)=7048804)
70
+ 64 train 7.907369 (lr=3.4091e-04) (hash(x)=6085549)
71
+ 65 train 8.007548 (lr=3.4615e-04) (hash(x)=5752594)
72
+ 66 train 7.814554 (lr=3.5140e-04) (hash(x)=7662769)
73
+ 67 train 7.743587 (lr=3.5664e-04) (hash(x)=6393520)
74
+ 68 train 7.959010 (lr=3.6189e-04) (hash(x)=5942867)
75
+ 69 train 7.988454 (lr=3.6713e-04) (hash(x)=6793550)
76
+ 70 train 7.777963 (lr=3.7238e-04) (hash(x)=7105976)
77
+ 71 train 7.838023 (lr=3.7762e-04) (hash(x)=6581348)
78
+ 72 train 7.664802 (lr=3.8287e-04) (hash(x)=5555598)
79
+ 73 train 7.773661 (lr=3.8811e-04) (hash(x)=6235837)
80
+ 74 train 7.690621 (lr=3.9336e-04) (hash(x)=5591770)
81
+ 75 train 8.078658 (lr=3.9860e-04) (hash(x)=7083021)
82
+ 76 train 7.860896 (lr=4.0385e-04) (hash(x)=7281348)
83
+ 77 train 8.053528 (lr=4.0909e-04) (hash(x)=6155821)
84
+ 78 train 7.783010 (lr=4.1434e-04) (hash(x)=5736241)
85
+ 79 train 7.748882 (lr=4.1958e-04) (hash(x)=5657123)
86
+ 80 train 7.916852 (lr=4.2483e-04) (hash(x)=6452095)
87
+ 81 train 7.714300 (lr=4.3007e-04) (hash(x)=5312267)
88
+ 82 train 7.433357 (lr=4.3531e-04) (hash(x)=5434493)
89
+ 83 train 7.815344 (lr=4.4056e-04) (hash(x)=6830813)
90
+ 84 train 7.480091 (lr=4.4580e-04) (hash(x)=4752771)
91
+ 85 train 7.766188 (lr=4.5105e-04) (hash(x)=6424933)
92
+ 86 train 7.757049 (lr=4.5629e-04) (hash(x)=5786517)
93
+ 87 train 7.899373 (lr=4.6154e-04) (hash(x)=6680023)
94
+ 88 train 7.822234 (lr=4.6678e-04) (hash(x)=6474729)
95
+ 89 train 7.791986 (lr=4.7203e-04) (hash(x)=7523279)
96
+ 90 train 7.600240 (lr=4.7727e-04) (hash(x)=5508768)
97
+ 91 train 7.909948 (lr=4.8252e-04) (hash(x)=6635925)
98
+ 92 train 7.707299 (lr=4.8776e-04) (hash(x)=6663801)
99
+ 93 train 7.684121 (lr=4.9301e-04) (hash(x)=5869441)
100
+ 94 train 7.811765 (lr=4.9825e-04) (hash(x)=6698878)
101
+ 95 train 7.675604 (lr=5.0350e-04) (hash(x)=6349004)
102
+ 96 train 7.722186 (lr=5.0874e-04) (hash(x)=4986137)
103
+ 97 train 7.504371 (lr=5.1399e-04) (hash(x)=6008454)
104
+ 98 train 7.580455 (lr=5.1923e-04) (hash(x)=6550770)
105
+ 99 train 7.760457 (lr=5.2448e-04) (hash(x)=7682741)
106
+ 100 val loss 7.7111
107
+ 100 val perplexity 2233.0059
108
+ 100 train 7.834822 (lr=5.2972e-04) (hash(x)=7892158)
109
+ 101 train 7.708977 (lr=5.3497e-04) (hash(x)=7794028)
110
+ 102 train 7.690384 (lr=5.4021e-04) (hash(x)=7093523)
111
+ 103 train 7.784535 (lr=5.4545e-04) (hash(x)=6166062)
112
+ 104 train 7.719335 (lr=5.5070e-04) (hash(x)=6645781)
113
+ 105 train 7.682852 (lr=5.5594e-04) (hash(x)=4811192)
114
+ 106 train 7.612810 (lr=5.6119e-04) (hash(x)=5970866)
115
+ 107 train 7.621126 (lr=5.6643e-04) (hash(x)=7822680)
116
+ 108 train 7.610049 (lr=5.7168e-04) (hash(x)=6317317)
117
+ 109 train 7.736577 (lr=5.7692e-04) (hash(x)=6431409)
118
+ 110 train 7.722856 (lr=5.8217e-04) (hash(x)=6653337)
119
+ 111 train 7.686261 (lr=5.8741e-04) (hash(x)=5323032)
120
+ 112 train 7.562115 (lr=5.9266e-04) (hash(x)=5125339)
121
+ 113 train 7.479772 (lr=5.9790e-04) (hash(x)=7363286)
122
+ 114 train 7.958846 (lr=6.0315e-04) (hash(x)=7670050)
123
+ 115 train 7.678079 (lr=6.0839e-04) (hash(x)=5616075)
124
+ 116 train 7.742112 (lr=6.1364e-04) (hash(x)=6593764)
125
+ 117 train 7.814540 (lr=6.1888e-04) (hash(x)=5672608)
126
+ 118 train 7.644459 (lr=6.2413e-04) (hash(x)=6828653)
127
+ 119 train 7.599339 (lr=6.2937e-04) (hash(x)=6839867)
128
+ 120 train 7.796554 (lr=6.3462e-04) (hash(x)=8426992)
129
+ 121 train 7.418078 (lr=6.3986e-04) (hash(x)=4916928)
130
+ 122 train 7.648375 (lr=6.4510e-04) (hash(x)=6484050)
131
+ 123 train 7.737336 (lr=6.5035e-04) (hash(x)=8353379)
132
+ 124 train 7.678030 (lr=6.5559e-04) (hash(x)=6552510)
133
+ 125 train 7.643674 (lr=6.6084e-04) (hash(x)=4838871)
134
+ 126 train 7.406717 (lr=6.6608e-04) (hash(x)=3021697)
135
+ 127 train 7.555677 (lr=6.7133e-04) (hash(x)=6795665)
136
+ 128 train 7.560263 (lr=6.7657e-04) (hash(x)=5486935)
137
+ 129 train 7.788219 (lr=6.8182e-04) (hash(x)=6678038)
138
+ 130 train 7.813459 (lr=6.8706e-04) (hash(x)=6657714)
139
+ 131 train 7.664540 (lr=6.9231e-04) (hash(x)=6112215)
140
+ 132 train 7.728020 (lr=6.9755e-04) (hash(x)=8353143)
141
+ 133 train 7.754289 (lr=7.0280e-04) (hash(x)=5915361)
142
+ 134 train 7.653434 (lr=7.0804e-04) (hash(x)=6254885)
143
+ 135 train 7.659563 (lr=7.1329e-04) (hash(x)=5771588)
144
+ 136 train 7.625689 (lr=7.1853e-04) (hash(x)=5561507)
145
+ 137 train 7.567125 (lr=7.2378e-04) (hash(x)=5688829)
146
+ 138 train 7.594048 (lr=7.2902e-04) (hash(x)=5371951)
147
+ 139 train 7.779230 (lr=7.3427e-04) (hash(x)=5443305)
148
+ 140 train 7.720368 (lr=7.3951e-04) (hash(x)=5016757)
149
+ 141 train 7.723347 (lr=7.4476e-04) (hash(x)=6901933)
150
+ 142 train 7.686146 (lr=7.5000e-04) (hash(x)=6019085)
151
+ 143 train 7.532293 (lr=7.5524e-04) (hash(x)=4294425)
152
+ 144 train 7.627684 (lr=7.6049e-04) (hash(x)=6095229)
153
+ 145 train 7.433883 (lr=7.6573e-04) (hash(x)=4363941)
154
+ 146 train 7.540045 (lr=7.7098e-04) (hash(x)=4693798)
155
+ 147 train 8.160391 (lr=7.7622e-04) (hash(x)=5486335)
156
+ 148 train 7.901616 (lr=7.8147e-04) (hash(x)=7907450)
157
+ 149 train 7.728026 (lr=7.8671e-04) (hash(x)=7295165)
158
+ 150 val loss 7.6103
159
+ 150 val perplexity 2018.9358
160
+ 150 train 7.663303 (lr=7.9196e-04) (hash(x)=6814026)
161
+ 151 train 7.754788 (lr=7.9720e-04) (hash(x)=6267424)
162
+ 152 train 7.794881 (lr=8.0245e-04) (hash(x)=8086437)
163
+ 153 train 7.809658 (lr=8.0769e-04) (hash(x)=5270452)
164
+ 154 train 7.640813 (lr=8.1294e-04) (hash(x)=6420820)
165
+ 155 train 7.531672 (lr=8.1818e-04) (hash(x)=6305297)
166
+ 156 train 7.708363 (lr=8.2343e-04) (hash(x)=7889849)
167
+ 157 train 8.138688 (lr=8.2867e-04) (hash(x)=6040077)
168
+ 158 train 8.114298 (lr=8.3392e-04) (hash(x)=5652497)
169
+ 159 train 7.683683 (lr=8.3916e-04) (hash(x)=7130251)
170
+ 160 train 7.499966 (lr=8.4441e-04) (hash(x)=6610177)
171
+ 161 train 7.557239 (lr=8.4965e-04) (hash(x)=6508350)
172
+ 162 train 7.751796 (lr=8.5490e-04) (hash(x)=6994983)
173
+ 163 train 7.698402 (lr=8.6014e-04) (hash(x)=5859576)
174
+ 164 train 7.255237 (lr=8.6538e-04) (hash(x)=5487065)
175
+ 165 train 7.536467 (lr=8.7063e-04) (hash(x)=4892065)
176
+ 166 train 7.456326 (lr=8.7587e-04) (hash(x)=4933674)
177
+ 167 train 7.657593 (lr=8.8112e-04) (hash(x)=7317289)
178
+ 168 train 7.766937 (lr=8.8636e-04) (hash(x)=6811522)
179
+ 169 train 7.570502 (lr=8.9161e-04) (hash(x)=5510218)
180
+ 170 train 7.740545 (lr=8.9685e-04) (hash(x)=7304235)
181
+ 171 train 7.877504 (lr=9.0210e-04) (hash(x)=6752265)
182
+ 172 train 7.596398 (lr=9.0734e-04) (hash(x)=5994476)
183
+ 173 train 7.552845 (lr=9.1259e-04) (hash(x)=6530867)
184
+ 174 train 7.636429 (lr=9.1783e-04) (hash(x)=5746260)
185
+ 175 train 7.597188 (lr=9.2308e-04) (hash(x)=5790078)
186
+ 176 train 7.614393 (lr=9.2832e-04) (hash(x)=6430135)
187
+ 177 train 7.673423 (lr=9.3357e-04) (hash(x)=7614023)
188
+ 178 train 7.662975 (lr=9.3881e-04) (hash(x)=5244861)
189
+ 179 train 7.644286 (lr=9.4406e-04) (hash(x)=5245146)
190
+ 180 train 7.526990 (lr=9.4930e-04) (hash(x)=6415565)
191
+ 181 train 7.585247 (lr=9.5455e-04) (hash(x)=5786972)
192
+ 182 train 7.622163 (lr=9.5979e-04) (hash(x)=6131703)
193
+ 183 train 7.627586 (lr=9.6503e-04) (hash(x)=6346459)
194
+ 184 train 7.657121 (lr=9.7028e-04) (hash(x)=6504131)
195
+ 185 train 7.646118 (lr=9.7552e-04) (hash(x)=5209234)
196
+ 186 train 7.556009 (lr=9.8077e-04) (hash(x)=6163272)
197
+ 187 train 7.564787 (lr=9.8601e-04) (hash(x)=5676768)
198
+ 188 train 7.651370 (lr=9.9126e-04) (hash(x)=6090561)
199
+ 189 train 7.767120 (lr=9.9650e-04) (hash(x)=5863092)
200
+ 190 train 7.686440 (lr=1.0017e-03) (hash(x)=7201312)
201
+ 191 train 7.756906 (lr=1.0070e-03) (hash(x)=7019489)
202
+ 192 train 7.553111 (lr=1.0122e-03) (hash(x)=6225608)
203
+ 193 train 7.646795 (lr=1.0175e-03) (hash(x)=5483019)
204
+ 194 train 7.591859 (lr=1.0227e-03) (hash(x)=5903258)
205
+ 195 train 7.634333 (lr=1.0280e-03) (hash(x)=5968716)
206
+ 196 train 7.602778 (lr=1.0332e-03) (hash(x)=5778017)
207
+ 197 train 7.340197 (lr=1.0385e-03) (hash(x)=7780194)
208
+ 198 train 7.544653 (lr=1.0437e-03) (hash(x)=6140998)
209
+ 199 train 7.517815 (lr=1.0490e-03) (hash(x)=3784321)
210
+ 200 val loss 7.5342
211
+ 200 val perplexity 1870.9634
212
+ 200 train 7.539381 (lr=1.0542e-03) (hash(x)=5860821)
213
+ 201 train 7.595202 (lr=1.0594e-03) (hash(x)=7774109)
214
+ 202 train 7.801271 (lr=1.0647e-03) (hash(x)=7997664)
215
+ 203 train 7.755901 (lr=1.0699e-03) (hash(x)=6576749)
216
+ 204 train 7.654162 (lr=1.0752e-03) (hash(x)=7266655)
217
+ 205 train 7.606019 (lr=1.0804e-03) (hash(x)=7072404)
218
+ 206 train 7.535525 (lr=1.0857e-03) (hash(x)=6626459)
219
+ 207 train 7.557119 (lr=1.0909e-03) (hash(x)=6556025)
220
+ 208 train 7.597475 (lr=1.0962e-03) (hash(x)=5942129)
221
+ 209 train 7.705553 (lr=1.1014e-03) (hash(x)=7325027)
222
+ 210 train 7.717119 (lr=1.1066e-03) (hash(x)=7322467)
223
+ 211 train 7.659253 (lr=1.1119e-03) (hash(x)=6373412)
224
+ 212 train 7.434551 (lr=1.1171e-03) (hash(x)=7332255)
225
+ 213 train 7.697508 (lr=1.1224e-03) (hash(x)=5262868)
226
+ 214 train 7.400546 (lr=1.1276e-03) (hash(x)=5688247)
227
+ 215 train 7.610068 (lr=1.1329e-03) (hash(x)=8055563)
228
+ 216 train 7.609561 (lr=1.1381e-03) (hash(x)=5352405)
229
+ 217 train 7.646708 (lr=1.1434e-03) (hash(x)=5932790)
230
+ 218 train 7.614678 (lr=1.1486e-03) (hash(x)=6249312)
231
+ 219 train 7.539387 (lr=1.1538e-03) (hash(x)=5680154)
232
+ 220 train 7.642365 (lr=1.1591e-03) (hash(x)=8156280)
233
+ 221 train 7.574601 (lr=1.1643e-03) (hash(x)=5914217)
234
+ 222 train 7.424339 (lr=1.1696e-03) (hash(x)=7120215)
235
+ 223 train 7.601686 (lr=1.1748e-03) (hash(x)=6424128)
236
+ 224 train 7.378928 (lr=1.1801e-03) (hash(x)=4390027)
237
+ 225 train 7.577817 (lr=1.1853e-03) (hash(x)=7804089)
238
+ 226 train 7.708376 (lr=1.1906e-03) (hash(x)=7130267)
239
+ 227 train 7.496459 (lr=1.1958e-03) (hash(x)=6490149)
240
+ 228 train 7.495287 (lr=1.2010e-03) (hash(x)=9276378)
241
+ 229 train 7.525814 (lr=1.2063e-03) (hash(x)=6283540)
242
+ 230 train 7.524784 (lr=1.2115e-03) (hash(x)=5781680)
243
+ 231 train 7.466030 (lr=1.2168e-03) (hash(x)=5728337)
244
+ 232 train 7.455902 (lr=1.2220e-03) (hash(x)=6278202)
245
+ 233 train 7.527832 (lr=1.2273e-03) (hash(x)=7400272)
246
+ 234 train 7.511873 (lr=1.2325e-03) (hash(x)=5795128)
247
+ 235 train 7.507438 (lr=1.2378e-03) (hash(x)=6473432)
248
+ 236 train 7.519451 (lr=1.2430e-03) (hash(x)=5201502)
249
+ 237 train 7.738815 (lr=1.2483e-03) (hash(x)=6922646)
250
+ 238 train 7.456667 (lr=1.2535e-03) (hash(x)=5484641)
251
+ 239 train 7.523755 (lr=1.2587e-03) (hash(x)=7277141)
252
+ 240 train 7.539600 (lr=1.2640e-03) (hash(x)=5604614)
253
+ 241 train 7.608048 (lr=1.2692e-03) (hash(x)=7443124)
254
+ 242 train 7.623682 (lr=1.2745e-03) (hash(x)=5601968)
255
+ 243 train 7.498574 (lr=1.2797e-03) (hash(x)=5632735)
256
+ 244 train 7.517906 (lr=1.2850e-03) (hash(x)=5594810)
257
+ 245 train 7.340271 (lr=1.2902e-03) (hash(x)=5881015)
258
+ 246 train 7.507755 (lr=1.2955e-03) (hash(x)=6865361)
259
+ 247 train 7.609521 (lr=1.3007e-03) (hash(x)=5809411)
260
+ 248 train 7.586152 (lr=1.3059e-03) (hash(x)=6293025)
261
+ 249 train 7.516971 (lr=1.3112e-03) (hash(x)=7232964)
262
+ 250 val loss 7.5351
263
+ 250 val perplexity 1872.6637
264
+ 250 train 7.461758 (lr=1.3164e-03) (hash(x)=5646619)
265
+ 251 train 7.580719 (lr=1.3217e-03) (hash(x)=6171829)
266
+ 252 train 7.650293 (lr=1.3269e-03) (hash(x)=7787552)
267
+ 253 train 7.611538 (lr=1.3322e-03) (hash(x)=5884456)
268
+ 254 train 7.632044 (lr=1.3374e-03) (hash(x)=6752162)
269
+ 255 train 7.538652 (lr=1.3427e-03) (hash(x)=5077031)
270
+ 256 train 7.706270 (lr=1.3479e-03) (hash(x)=6883481)
271
+ 257 train 7.599019 (lr=1.3531e-03) (hash(x)=6248895)
272
+ 258 train 7.483520 (lr=1.3584e-03) (hash(x)=6636615)
273
+ 259 train 7.666373 (lr=1.3636e-03) (hash(x)=6918004)
274
+ 260 train 7.586901 (lr=1.3689e-03) (hash(x)=7302874)
275
+ 261 train 7.689678 (lr=1.3741e-03) (hash(x)=6482466)
276
+ 262 train 7.543999 (lr=1.3794e-03) (hash(x)=7298317)
277
+ 263 train 7.587693 (lr=1.3846e-03) (hash(x)=6588573)
278
+ 264 train 7.578894 (lr=1.3899e-03) (hash(x)=6229630)
279
+ 265 train 7.627550 (lr=1.3951e-03) (hash(x)=6136939)
280
+ 266 train 7.423788 (lr=1.4003e-03) (hash(x)=6394645)
281
+ 267 train 7.642031 (lr=1.4056e-03) (hash(x)=7040157)
282
+ 268 train 7.693124 (lr=1.4108e-03) (hash(x)=7041858)
283
+ 269 train 7.618255 (lr=1.4161e-03) (hash(x)=6455595)
284
+ 270 train 7.571319 (lr=1.4213e-03) (hash(x)=6480443)
285
+ 271 train 7.720523 (lr=1.4266e-03) (hash(x)=4122983)
286
+ 272 train 7.541777 (lr=1.4318e-03) (hash(x)=7672190)
287
+ 273 train 7.578089 (lr=1.4371e-03) (hash(x)=4724052)
288
+ 274 train 7.515188 (lr=1.4423e-03) (hash(x)=6654267)
289
+ 275 train 7.519406 (lr=1.4476e-03) (hash(x)=6820797)
290
+ 276 train 7.594303 (lr=1.4528e-03) (hash(x)=6458662)
291
+ 277 train 7.952461 (lr=1.4580e-03) (hash(x)=7477594)
292
+ 278 train 7.632566 (lr=1.4633e-03) (hash(x)=5876574)
293
+ 279 train 7.364169 (lr=1.4685e-03) (hash(x)=6146305)
294
+ 280 train 7.530088 (lr=1.4738e-03) (hash(x)=6162080)
295
+ 281 train 7.478974 (lr=1.4790e-03) (hash(x)=6814831)
296
+ 282 train 7.633631 (lr=1.4843e-03) (hash(x)=7138295)
297
+ 283 train 7.375920 (lr=1.4895e-03) (hash(x)=5548298)
298
+ 284 train 7.499508 (lr=1.4948e-03) (hash(x)=6265078)
299
+ 285 train 7.904729 (lr=1.5000e-03) (hash(x)=7609416)
300
+ 286 train 7.580350 (lr=1.5000e-03) (hash(x)=6909367)
301
+ 287 train 7.669079 (lr=1.5000e-03) (hash(x)=7574342)
302
+ 288 train 7.480936 (lr=1.5000e-03) (hash(x)=8859748)
303
+ 289 train 7.454789 (lr=1.4999e-03) (hash(x)=5906937)
304
+ 290 train 7.484822 (lr=1.4999e-03) (hash(x)=7243024)
305
+ 291 train 7.597754 (lr=1.4998e-03) (hash(x)=8013471)
306
+ 292 train 7.495422 (lr=1.4998e-03) (hash(x)=5659461)
307
+ 293 train 7.549082 (lr=1.4997e-03) (hash(x)=6146213)
308
+ 294 train 7.569899 (lr=1.4996e-03) (hash(x)=6602917)
309
+ 295 train 7.517811 (lr=1.4995e-03) (hash(x)=6709902)
310
+ 296 train 7.602410 (lr=1.4993e-03) (hash(x)=8154341)
311
+ 297 train 7.744252 (lr=1.4992e-03) (hash(x)=6402571)
312
+ 298 train 7.552544 (lr=1.4991e-03) (hash(x)=5605466)
313
+ 299 train 7.501730 (lr=1.4989e-03) (hash(x)=7683614)
314
+ 300 val loss 7.5344
315
+ 300 val perplexity 1871.2882
316
+ 300 train 7.573627 (lr=1.4987e-03) (hash(x)=5421572)
317
+ 301 train 7.421243 (lr=1.4985e-03) (hash(x)=5453026)
318
+ 302 train 7.496454 (lr=1.4983e-03) (hash(x)=6437746)
319
+ 303 train 7.405605 (lr=1.4981e-03) (hash(x)=6260017)
320
+ 304 train 7.507078 (lr=1.4979e-03) (hash(x)=6119053)
321
+ 305 train 7.680285 (lr=1.4976e-03) (hash(x)=6591466)
322
+ 306 train 7.603331 (lr=1.4974e-03) (hash(x)=5652701)
323
+ 307 train 7.579115 (lr=1.4971e-03) (hash(x)=5501472)
324
+ 308 train 7.741616 (lr=1.4968e-03) (hash(x)=6160695)
325
+ 309 train 7.495920 (lr=1.4965e-03) (hash(x)=5871698)
326
+ 310 train 7.495649 (lr=1.4962e-03) (hash(x)=7301184)
327
+ 311 train 7.419578 (lr=1.4959e-03) (hash(x)=4282052)
328
+ 312 train 7.662704 (lr=1.4956e-03) (hash(x)=5679330)
329
+ 313 train 7.534294 (lr=1.4952e-03) (hash(x)=7824660)
330
+ 314 train 7.588016 (lr=1.4949e-03) (hash(x)=4857318)
331
+ 315 train 7.533437 (lr=1.4945e-03) (hash(x)=5954950)
332
+ 316 train 7.565056 (lr=1.4941e-03) (hash(x)=5758178)
333
+ 317 train 7.590521 (lr=1.4937e-03) (hash(x)=5484656)
334
+ 318 train 7.592374 (lr=1.4933e-03) (hash(x)=6818913)
335
+ 319 train 7.796222 (lr=1.4929e-03) (hash(x)=7031328)
336
+ 320 train 7.687158 (lr=1.4925e-03) (hash(x)=6837621)
337
+ 321 train 7.711910 (lr=1.4920e-03) (hash(x)=9157069)
338
+ 322 train 7.939399 (lr=1.4915e-03) (hash(x)=6509536)
339
+ 323 train 7.652686 (lr=1.4911e-03) (hash(x)=5996883)
340
+ 324 train 7.913796 (lr=1.4906e-03) (hash(x)=6850868)
341
+ 325 train 7.859757 (lr=1.4901e-03) (hash(x)=7645447)
342
+ 326 train 7.863148 (lr=1.4896e-03) (hash(x)=6717515)
343
+ 327 train 7.953239 (lr=1.4890e-03) (hash(x)=6336164)
344
+ 328 train 7.683919 (lr=1.4885e-03) (hash(x)=6897445)
345
+ 329 train 7.608979 (lr=1.4880e-03) (hash(x)=5224240)
346
+ 330 train 7.757701 (lr=1.4874e-03) (hash(x)=6612945)
347
+ 331 train 8.205700 (lr=1.4868e-03) (hash(x)=5447801)
348
+ 332 train 7.743669 (lr=1.4862e-03) (hash(x)=7308579)
349
+ 333 train 7.592071 (lr=1.4856e-03) (hash(x)=4155598)
350
+ 334 train 7.612006 (lr=1.4850e-03) (hash(x)=6106773)
351
+ 335 train 7.675923 (lr=1.4844e-03) (hash(x)=7482744)
352
+ 336 train 7.728288 (lr=1.4837e-03) (hash(x)=6310687)
353
+ 337 train 7.641386 (lr=1.4831e-03) (hash(x)=4733414)
354
+ 338 train 7.767277 (lr=1.4824e-03) (hash(x)=6979734)
355
+ 339 train 7.571215 (lr=1.4817e-03) (hash(x)=7806284)
356
+ 340 train 7.541497 (lr=1.4810e-03) (hash(x)=5080536)
357
+ 341 train 7.748325 (lr=1.4803e-03) (hash(x)=7236478)
358
+ 342 train 7.685267 (lr=1.4796e-03) (hash(x)=6619318)
359
+ 343 train 7.587360 (lr=1.4789e-03) (hash(x)=4656255)
360
+ 344 train 7.549957 (lr=1.4781e-03) (hash(x)=5546888)
361
+ 345 train 7.971974 (lr=1.4774e-03) (hash(x)=9298325)
362
+ 346 train 7.682766 (lr=1.4766e-03) (hash(x)=5994232)
363
+ 347 train 7.666240 (lr=1.4758e-03) (hash(x)=6287524)
364
+ 348 train 7.767812 (lr=1.4750e-03) (hash(x)=6200049)
365
+ 349 train 7.692766 (lr=1.4742e-03) (hash(x)=7895970)
366
+ 350 val loss 7.6207
367
+ 350 val perplexity 2039.9320
368
+ 350 train 7.620523 (lr=1.4734e-03) (hash(x)=6051847)
369
+ 351 train 7.810394 (lr=1.4726e-03) (hash(x)=8847992)
370
+ 352 train 8.035326 (lr=1.4717e-03) (hash(x)=5123883)
371
+ 353 train 7.709318 (lr=1.4709e-03) (hash(x)=6338998)
372
+ 354 train 7.707136 (lr=1.4700e-03) (hash(x)=5765866)
373
+ 355 train 8.052864 (lr=1.4691e-03) (hash(x)=7104011)
374
+ 356 train 7.806067 (lr=1.4682e-03) (hash(x)=7438267)
375
+ 357 train 7.570807 (lr=1.4673e-03) (hash(x)=5573214)
376
+ 358 train 7.830725 (lr=1.4664e-03) (hash(x)=6897283)
377
+ 359 train 7.659793 (lr=1.4655e-03) (hash(x)=6882897)
378
+ 360 train 7.741349 (lr=1.4645e-03) (hash(x)=5343615)
379
+ 361 train 7.646974 (lr=1.4636e-03) (hash(x)=5854242)
380
+ 362 train 7.592349 (lr=1.4626e-03) (hash(x)=6317324)
381
+ 363 train 7.546738 (lr=1.4616e-03) (hash(x)=6972494)
382
+ 364 train 7.593963 (lr=1.4606e-03) (hash(x)=5661167)
383
+ 365 train 7.611187 (lr=1.4596e-03) (hash(x)=6796805)
384
+ 366 train 7.467405 (lr=1.4586e-03) (hash(x)=5817209)
385
+ 367 train 7.523379 (lr=1.4576e-03) (hash(x)=6139120)
386
+ 368 train 7.682480 (lr=1.4565e-03) (hash(x)=6121296)
387
+ 369 train 7.663208 (lr=1.4555e-03) (hash(x)=6486796)
388
+ 370 train 7.613709 (lr=1.4544e-03) (hash(x)=5514645)
389
+ 371 train 7.532767 (lr=1.4533e-03) (hash(x)=5578797)
390
+ 372 train 7.545802 (lr=1.4522e-03) (hash(x)=7651738)
391
+ 373 train 7.740781 (lr=1.4511e-03) (hash(x)=6895326)
392
+ 374 train 7.747188 (lr=1.4500e-03) (hash(x)=5490256)
393
+ 375 train 7.610308 (lr=1.4489e-03) (hash(x)=7783160)
394
+ 376 train 7.812620 (lr=1.4478e-03) (hash(x)=5608424)
395
+ 377 train 7.590793 (lr=1.4466e-03) (hash(x)=4867625)
396
+ 378 train 7.494847 (lr=1.4454e-03) (hash(x)=4655472)
397
+ 379 train 7.566562 (lr=1.4443e-03) (hash(x)=6631123)
398
+ 380 train 7.244875 (lr=1.4431e-03) (hash(x)=5751990)
399
+ 381 train 7.537903 (lr=1.4419e-03) (hash(x)=6710965)
400
+ 382 train 7.519583 (lr=1.4407e-03) (hash(x)=5293800)
401
+ 383 train 7.532013 (lr=1.4394e-03) (hash(x)=6294657)
402
+ 384 train 7.683675 (lr=1.4382e-03) (hash(x)=6999538)
403
+ 385 train 7.662749 (lr=1.4370e-03) (hash(x)=5443606)
404
+ 386 train 7.659564 (lr=1.4357e-03) (hash(x)=5650401)
405
+ 387 train 7.629635 (lr=1.4344e-03) (hash(x)=7035770)
406
+ 388 train 7.591470 (lr=1.4332e-03) (hash(x)=6678271)
407
+ 389 train 7.790829 (lr=1.4319e-03) (hash(x)=4899541)
408
+ 390 train 7.563005 (lr=1.4306e-03) (hash(x)=5359260)
409
+ 391 train 7.594091 (lr=1.4292e-03) (hash(x)=6227120)
410
+ 392 train 7.625456 (lr=1.4279e-03) (hash(x)=6643916)
411
+ 393 train 7.680484 (lr=1.4266e-03) (hash(x)=5981648)
412
+ 394 train 7.655550 (lr=1.4252e-03) (hash(x)=6467540)
413
+ 395 train 7.630481 (lr=1.4238e-03) (hash(x)=5558993)
414
+ 396 train 7.778150 (lr=1.4225e-03) (hash(x)=6698959)
415
+ 397 train 7.598825 (lr=1.4211e-03) (hash(x)=7015671)
416
+ 398 train 7.606699 (lr=1.4197e-03) (hash(x)=6800890)
417
+ 399 train 7.727904 (lr=1.4183e-03) (hash(x)=5931234)
418
+ 400 val loss 7.6630
419
+ 400 val perplexity 2128.1924
420
+ 400 train 7.625745 (lr=1.4169e-03) (hash(x)=3995070)
421
+ 401 train 7.537912 (lr=1.4154e-03) (hash(x)=4582159)
422
+ 402 train 7.553117 (lr=1.4140e-03) (hash(x)=7081906)
423
+ 403 train 7.660930 (lr=1.4125e-03) (hash(x)=5029472)
424
+ 404 train 7.654960 (lr=1.4110e-03) (hash(x)=3946232)
425
+ 405 train 7.567132 (lr=1.4096e-03) (hash(x)=5434102)
426
+ 406 train 8.256627 (lr=1.4081e-03) (hash(x)=5340646)
427
+ 407 train 7.579591 (lr=1.4066e-03) (hash(x)=6810638)
428
+ 408 train 7.768649 (lr=1.4051e-03) (hash(x)=6768026)
429
+ 409 train 7.794454 (lr=1.4035e-03) (hash(x)=6521562)
430
+ 410 train 7.788802 (lr=1.4020e-03) (hash(x)=4429434)
431
+ 411 train 7.810375 (lr=1.4005e-03) (hash(x)=7226932)
432
+ 412 train 7.611440 (lr=1.3989e-03) (hash(x)=5065932)
433
+ 413 train 7.758048 (lr=1.3973e-03) (hash(x)=6957849)
434
+ 414 train 7.746157 (lr=1.3957e-03) (hash(x)=6431457)
435
+ 415 train 7.825485 (lr=1.3942e-03) (hash(x)=5334281)
436
+ 416 train 7.808317 (lr=1.3926e-03) (hash(x)=5471747)
437
+ 417 train 7.954847 (lr=1.3909e-03) (hash(x)=5887983)
438
+ 418 train 7.831559 (lr=1.3893e-03) (hash(x)=5075909)
439
+ 419 train 7.592844 (lr=1.3877e-03) (hash(x)=6393576)
440
+ 420 train 7.740710 (lr=1.3860e-03) (hash(x)=6608979)
441
+ 421 train 8.037004 (lr=1.3844e-03) (hash(x)=5057959)
442
+ 422 train 8.071031 (lr=1.3827e-03) (hash(x)=6414858)
443
+ 423 train 7.830724 (lr=1.3810e-03) (hash(x)=8595271)
444
+ 424 train 7.708473 (lr=1.3793e-03) (hash(x)=6500949)
445
+ 425 train 7.849252 (lr=1.3776e-03) (hash(x)=8395167)
446
+ 426 train 7.958614 (lr=1.3759e-03) (hash(x)=6288420)
447
+ 427 train 7.766601 (lr=1.3742e-03) (hash(x)=7634417)
448
+ 428 train 7.640208 (lr=1.3725e-03) (hash(x)=6795561)
449
+ 429 train 7.557518 (lr=1.3707e-03) (hash(x)=5257771)
450
+ 430 train 7.709934 (lr=1.3690e-03) (hash(x)=7506860)
451
+ 431 train 7.759339 (lr=1.3672e-03) (hash(x)=7674238)
452
+ 432 train 7.840971 (lr=1.3654e-03) (hash(x)=5353794)
453
+ 433 train 7.745647 (lr=1.3637e-03) (hash(x)=5300555)
454
+ 434 train 7.509052 (lr=1.3619e-03) (hash(x)=5564419)
455
+ 435 train 7.767920 (lr=1.3601e-03) (hash(x)=6066139)
456
+ 436 train 7.703519 (lr=1.3582e-03) (hash(x)=6863709)
457
+ 437 train 7.690256 (lr=1.3564e-03) (hash(x)=6688212)
458
+ 438 train 7.815314 (lr=1.3546e-03) (hash(x)=7095090)
459
+ 439 train 7.883111 (lr=1.3527e-03) (hash(x)=6642404)
460
+ 440 train 7.652912 (lr=1.3509e-03) (hash(x)=6539227)
461
+ 441 train 7.634340 (lr=1.3490e-03) (hash(x)=4961570)
462
+ 442 train 7.596195 (lr=1.3471e-03) (hash(x)=6606391)
463
+ 443 train 7.407789 (lr=1.3452e-03) (hash(x)=5105190)
464
+ 444 train 7.951695 (lr=1.3434e-03) (hash(x)=5695997)
465
+ 445 train 7.648779 (lr=1.3414e-03) (hash(x)=7155991)
466
+ 446 train 7.633060 (lr=1.3395e-03) (hash(x)=7249727)
467
+ 447 train 7.635561 (lr=1.3376e-03) (hash(x)=5132972)
468
+ 448 train 7.637180 (lr=1.3357e-03) (hash(x)=5381802)
469
+ 449 train 7.658418 (lr=1.3337e-03) (hash(x)=6269946)
470
+ 450 val loss 7.6594
471
+ 450 val perplexity 2120.4897
472
+ 450 train 7.621373 (lr=1.3318e-03) (hash(x)=4976762)
473
+ 451 train 7.693218 (lr=1.3298e-03) (hash(x)=5396133)
474
+ 452 train 7.660570 (lr=1.3278e-03) (hash(x)=6173473)
475
+ 453 train 7.664090 (lr=1.3258e-03) (hash(x)=6394083)
476
+ 454 train 7.690687 (lr=1.3238e-03) (hash(x)=6436030)
477
+ 455 train 7.569850 (lr=1.3218e-03) (hash(x)=5237804)
478
+ 456 train 7.543784 (lr=1.3198e-03) (hash(x)=6540397)
479
+ 457 train 7.784432 (lr=1.3178e-03) (hash(x)=5216638)
480
+ 458 train 7.628320 (lr=1.3158e-03) (hash(x)=5750537)
481
+ 459 train 7.610534 (lr=1.3137e-03) (hash(x)=6735565)
482
+ 460 train 7.647701 (lr=1.3117e-03) (hash(x)=7297488)
483
+ 461 train 7.819501 (lr=1.3096e-03) (hash(x)=5983575)
484
+ 462 train 7.655029 (lr=1.3075e-03) (hash(x)=4808543)
485
+ 463 train 7.769736 (lr=1.3054e-03) (hash(x)=5831619)
486
+ 464 train 7.541070 (lr=1.3033e-03) (hash(x)=6280052)
487
+ 465 train 7.527079 (lr=1.3012e-03) (hash(x)=9109083)
488
+ 466 train 7.768382 (lr=1.2991e-03) (hash(x)=5848604)
489
+ 467 train 7.687286 (lr=1.2970e-03) (hash(x)=6912967)
490
+ 468 train 7.849314 (lr=1.2949e-03) (hash(x)=6615365)
491
+ 469 train 7.854143 (lr=1.2928e-03) (hash(x)=5965674)
492
+ 470 train 7.746822 (lr=1.2906e-03) (hash(x)=6879965)
493
+ 471 train 7.740748 (lr=1.2885e-03) (hash(x)=5822533)
494
+ 472 train 7.637462 (lr=1.2863e-03) (hash(x)=5794462)
495
+ 473 train 7.709984 (lr=1.2841e-03) (hash(x)=5268216)
496
+ 474 train 7.568446 (lr=1.2819e-03) (hash(x)=7058396)
497
+ 475 train 7.618642 (lr=1.2797e-03) (hash(x)=6531900)
498
+ 476 train 7.648572 (lr=1.2775e-03) (hash(x)=5878912)
499
+ 477 train 7.468830 (lr=1.2753e-03) (hash(x)=5426499)
500
+ 478 train 7.648506 (lr=1.2731e-03) (hash(x)=8107004)
501
+ 479 train 8.346062 (lr=1.2709e-03) (hash(x)=7455372)
502
+ 480 train 8.293447 (lr=1.2687e-03) (hash(x)=7727959)
503
+ 481 train 8.222755 (lr=1.2664e-03) (hash(x)=8319119)
504
+ 482 train 8.147004 (lr=1.2642e-03) (hash(x)=7581656)
505
+ 483 train 8.157889 (lr=1.2619e-03) (hash(x)=7592128)
506
+ 484 train 8.139569 (lr=1.2596e-03) (hash(x)=10506633)
507
+ 485 train 8.048663 (lr=1.2574e-03) (hash(x)=8704170)
508
+ 486 train 7.847312 (lr=1.2551e-03) (hash(x)=7461138)
509
+ 487 train 7.791181 (lr=1.2528e-03) (hash(x)=7165037)
510
+ 488 train 7.863379 (lr=1.2505e-03) (hash(x)=4976998)
511
+ 489 train 7.581710 (lr=1.2482e-03) (hash(x)=7678000)
512
+ 490 train 7.680907 (lr=1.2459e-03) (hash(x)=6299409)
513
+ 491 train 7.626888 (lr=1.2435e-03) (hash(x)=7414788)
514
+ 492 train 7.677799 (lr=1.2412e-03) (hash(x)=5063061)
515
+ 493 train 7.633603 (lr=1.2389e-03) (hash(x)=7326216)
516
+ 494 train 7.766290 (lr=1.2365e-03) (hash(x)=7492012)
517
+ 495 train 7.744939 (lr=1.2341e-03) (hash(x)=5857162)
518
+ 496 train 7.562132 (lr=1.2318e-03) (hash(x)=6169001)
519
+ 497 train 9.465097 (lr=1.2294e-03) (hash(x)=14715669)
520
+ 498 train 12.258412 (lr=1.2270e-03) (hash(x)=18256706)
521
+ 499 train 11.962095 (lr=1.2246e-03) (hash(x)=16950560)
522
+ 500 val loss 7.7678
523
+ 500 val perplexity 2363.3708
524
+ 500 train 8.443453 (lr=1.2222e-03) (hash(x)=6113563)
525
+ 501 train 7.495465 (lr=1.2198e-03) (hash(x)=6159125)
526
+ 502 train 7.933568 (lr=1.2174e-03) (hash(x)=5482303)
527
+ 503 train 8.031115 (lr=1.2150e-03) (hash(x)=5908442)
528
+ 504 train 7.984268 (lr=1.2126e-03) (hash(x)=4985045)
529
+ 505 train 7.890134 (lr=1.2101e-03) (hash(x)=6668887)
530
+ 506 train 7.894129 (lr=1.2077e-03) (hash(x)=6123910)
531
+ 507 train 7.973477 (lr=1.2052e-03) (hash(x)=7662975)
532
+ 508 train 7.917876 (lr=1.2028e-03) (hash(x)=7191898)
533
+ 509 train 7.841074 (lr=1.2003e-03) (hash(x)=6669539)
534
+ 510 train 7.856494 (lr=1.1978e-03) (hash(x)=6557481)
535
+ 511 train 7.769319 (lr=1.1954e-03) (hash(x)=5015998)
536
+ 512 train 7.805245 (lr=1.1929e-03) (hash(x)=4964540)
537
+ 513 train 7.681995 (lr=1.1904e-03) (hash(x)=5458386)
538
+ 514 train 7.955750 (lr=1.1879e-03) (hash(x)=6077616)
539
+ 515 train 7.745504 (lr=1.1854e-03) (hash(x)=4849847)
540
+ 516 train 7.783379 (lr=1.1829e-03) (hash(x)=5431422)
541
+ 517 train 7.926404 (lr=1.1803e-03) (hash(x)=5125562)
542
+ 518 train 7.883621 (lr=1.1778e-03) (hash(x)=5587715)
543
+ 519 train 7.926979 (lr=1.1753e-03) (hash(x)=6012047)
544
+ 520 train 7.788620 (lr=1.1727e-03) (hash(x)=5668436)
545
+ 521 train 7.801256 (lr=1.1702e-03) (hash(x)=6079364)
546
+ 522 train 7.951874 (lr=1.1676e-03) (hash(x)=5200504)
547
+ 523 train 7.681855 (lr=1.1651e-03) (hash(x)=6030395)
548
+ 524 train 7.931819 (lr=1.1625e-03) (hash(x)=6855319)
549
+ 525 train 7.703115 (lr=1.1599e-03) (hash(x)=6781762)
550
+ 526 train 7.832451 (lr=1.1573e-03) (hash(x)=6717939)
551
+ 527 train 7.825798 (lr=1.1548e-03) (hash(x)=5238325)
552
+ 528 train 7.699074 (lr=1.1522e-03) (hash(x)=5808069)
553
+ 529 train 8.007508 (lr=1.1496e-03) (hash(x)=8611144)
554
+ 530 train 7.710258 (lr=1.1470e-03) (hash(x)=6028738)
555
+ 531 train 7.715397 (lr=1.1443e-03) (hash(x)=6901605)
556
+ 532 train 7.684081 (lr=1.1417e-03) (hash(x)=6723307)
557
+ 533 train 7.768530 (lr=1.1391e-03) (hash(x)=5499337)
558
+ 534 train 7.744495 (lr=1.1365e-03) (hash(x)=6586372)
559
+ 535 train 7.821444 (lr=1.1338e-03) (hash(x)=7134874)
560
+ 536 train 7.739589 (lr=1.1312e-03) (hash(x)=8415623)
561
+ 537 train 7.769959 (lr=1.1285e-03) (hash(x)=6149038)
562
+ 538 train 7.798083 (lr=1.1259e-03) (hash(x)=6078325)
563
+ 539 train 7.758498 (lr=1.1232e-03) (hash(x)=6609760)
564
+ 540 train 7.654715 (lr=1.1205e-03) (hash(x)=6185253)
565
+ 541 train 7.976846 (lr=1.1179e-03) (hash(x)=6652627)
566
+ 542 train 7.799532 (lr=1.1152e-03) (hash(x)=7758536)
567
+ 543 train 7.784450 (lr=1.1125e-03) (hash(x)=8618579)
568
+ 544 train 7.733902 (lr=1.1098e-03) (hash(x)=6710226)
569
+ 545 train 7.610837 (lr=1.1071e-03) (hash(x)=5896892)
570
+ 546 train 7.616485 (lr=1.1044e-03) (hash(x)=6332995)
571
+ 547 train 8.013546 (lr=1.1017e-03) (hash(x)=8945854)
572
+ 548 train 7.571414 (lr=1.0990e-03) (hash(x)=4329129)
573
+ 549 train 7.642760 (lr=1.0963e-03) (hash(x)=6775103)
574
+ 550 val loss 7.6454
575
+ 550 val perplexity 2091.0547
576
+ 550 train 7.737250 (lr=1.0936e-03) (hash(x)=4465735)
577
+ 551 train 7.604080 (lr=1.0908e-03) (hash(x)=6200024)
578
+ 552 train 7.557221 (lr=1.0881e-03) (hash(x)=5520748)
579
+ 553 train 7.583283 (lr=1.0854e-03) (hash(x)=5238139)
580
+ 554 train 7.848519 (lr=1.0826e-03) (hash(x)=7639817)
581
+ 555 train 7.583924 (lr=1.0799e-03) (hash(x)=5780948)
582
+ 556 train 7.726992 (lr=1.0771e-03) (hash(x)=6733444)
583
+ 557 train 7.658147 (lr=1.0744e-03) (hash(x)=6168407)
584
+ 558 train 7.694695 (lr=1.0716e-03) (hash(x)=6680197)
585
+ 559 train 7.616935 (lr=1.0688e-03) (hash(x)=3775881)
586
+ 560 train 7.441613 (lr=1.0661e-03) (hash(x)=5784096)
587
+ 561 train 7.891470 (lr=1.0633e-03) (hash(x)=6848709)
588
+ 562 train 8.013200 (lr=1.0605e-03) (hash(x)=7449019)
589
+ 563 train 8.056321 (lr=1.0577e-03) (hash(x)=6478113)
590
+ 564 train 7.650788 (lr=1.0549e-03) (hash(x)=5925744)
591
+ 565 train 7.696724 (lr=1.0521e-03) (hash(x)=6483914)
592
+ 566 train 7.615714 (lr=1.0493e-03) (hash(x)=6010847)
593
+ 567 train 7.691716 (lr=1.0465e-03) (hash(x)=5927795)
594
+ 568 train 7.700327 (lr=1.0437e-03) (hash(x)=5896332)
595
+ 569 train 7.672825 (lr=1.0409e-03) (hash(x)=8438033)
596
+ 570 train 7.757077 (lr=1.0381e-03) (hash(x)=6081660)
597
+ 571 train 7.687593 (lr=1.0353e-03) (hash(x)=5684788)
598
+ 572 train 7.614841 (lr=1.0325e-03) (hash(x)=6272055)
599
+ 573 train 7.570560 (lr=1.0296e-03) (hash(x)=5710182)
600
+ 574 train 7.667852 (lr=1.0268e-03) (hash(x)=6219616)
601
+ 575 train 7.647057 (lr=1.0240e-03) (hash(x)=6160063)
602
+ 576 train 7.623658 (lr=1.0211e-03) (hash(x)=5710809)
603
+ 577 train 7.426347 (lr=1.0183e-03) (hash(x)=5955100)
604
+ 578 train 7.431763 (lr=1.0154e-03) (hash(x)=5356088)
605
+ 579 train 7.490788 (lr=1.0126e-03) (hash(x)=6672309)
606
+ 580 train 7.654601 (lr=1.0097e-03) (hash(x)=5841684)
607
+ 581 train 7.947323 (lr=1.0069e-03) (hash(x)=9001564)
608
+ 582 train 8.523070 (lr=1.0040e-03) (hash(x)=6897178)
609
+ 583 train 7.749501 (lr=1.0011e-03) (hash(x)=5454654)
610
+ 584 train 7.533551 (lr=9.9827e-04) (hash(x)=6391454)
611
+ 585 train 7.543225 (lr=9.9540e-04) (hash(x)=5174062)
612
+ 586 train 7.634393 (lr=9.9252e-04) (hash(x)=5197924)
613
+ 587 train 7.701547 (lr=9.8964e-04) (hash(x)=6199203)
614
+ 588 train 7.542275 (lr=9.8676e-04) (hash(x)=5786498)
615
+ 589 train 7.652937 (lr=9.8387e-04) (hash(x)=8352747)
616
+ 590 train 7.913832 (lr=9.8099e-04) (hash(x)=4829700)
617
+ 591 train 7.491575 (lr=9.7810e-04) (hash(x)=4547106)
618
+ 592 train 7.751403 (lr=9.7520e-04) (hash(x)=7686831)
619
+ 593 train 7.636143 (lr=9.7230e-04) (hash(x)=6433099)
620
+ 594 train 7.512798 (lr=9.6940e-04) (hash(x)=4826460)
621
+ 595 train 7.599208 (lr=9.6650e-04) (hash(x)=6424428)
622
+ 596 train 7.482791 (lr=9.6360e-04) (hash(x)=6073605)
623
+ 597 train 7.774817 (lr=9.6069e-04) (hash(x)=5975712)
624
+ 598 train 7.454946 (lr=9.5778e-04) (hash(x)=7030029)
625
+ 599 train 7.691435 (lr=9.5486e-04) (hash(x)=5458532)
626
+ 600 val loss 7.6450
627
+ 600 val perplexity 2090.1704
628
+ 600 train 7.576479 (lr=9.5195e-04) (hash(x)=7823919)
629
+ 601 train 7.953879 (lr=9.4903e-04) (hash(x)=7002774)
630
+ 602 train 7.784065 (lr=9.4611e-04) (hash(x)=7056399)
631
+ 603 train 7.957798 (lr=9.4319e-04) (hash(x)=8206173)
632
+ 604 train 7.798918 (lr=9.4026e-04) (hash(x)=8034846)
633
+ 605 train 8.072700 (lr=9.3733e-04) (hash(x)=7784971)
634
+ 606 train 8.340412 (lr=9.3440e-04) (hash(x)=7118076)
635
+ 607 train 7.576227 (lr=9.3147e-04) (hash(x)=6001843)
636
+ 608 train 7.658208 (lr=9.2854e-04) (hash(x)=5582437)
637
+ 609 train 7.757119 (lr=9.2560e-04) (hash(x)=5698642)
638
+ 610 train 7.627432 (lr=9.2267e-04) (hash(x)=6615301)
639
+ 611 train 7.645289 (lr=9.1973e-04) (hash(x)=6600751)
640
+ 612 train 7.555856 (lr=9.1678e-04) (hash(x)=6880955)
641
+ 613 train 7.619359 (lr=9.1384e-04) (hash(x)=5804610)
642
+ 614 train 7.544451 (lr=9.1090e-04) (hash(x)=6263347)
643
+ 615 train 7.540831 (lr=9.0795e-04) (hash(x)=5162652)
644
+ 616 train 7.557220 (lr=9.0500e-04) (hash(x)=6055710)
645
+ 617 train 7.524939 (lr=9.0205e-04) (hash(x)=5297006)
646
+ 618 train 7.544474 (lr=8.9910e-04) (hash(x)=5923648)
647
+ 619 train 7.567671 (lr=8.9615e-04) (hash(x)=6728323)
648
+ 620 train 7.576616 (lr=8.9319e-04) (hash(x)=6660239)
649
+ 621 train 7.537615 (lr=8.9024e-04) (hash(x)=3964882)
650
+ 622 train 7.597541 (lr=8.8728e-04) (hash(x)=5868942)
651
+ 623 train 7.536258 (lr=8.8432e-04) (hash(x)=6735993)
652
+ 624 train 7.813656 (lr=8.8136e-04) (hash(x)=5926190)
653
+ 625 train 7.529661 (lr=8.7840e-04) (hash(x)=5150316)
654
+ 626 train 7.554354 (lr=8.7544e-04) (hash(x)=7571467)
655
+ 627 train 7.458978 (lr=8.7248e-04) (hash(x)=5220448)
656
+ 628 train 7.477483 (lr=8.6952e-04) (hash(x)=6790397)
657
+ 629 train 7.892003 (lr=8.6655e-04) (hash(x)=6971289)
658
+ 630 train 7.865046 (lr=8.6359e-04) (hash(x)=7123359)
659
+ 631 train 7.676325 (lr=8.6062e-04) (hash(x)=5012069)
660
+ 632 train 7.645431 (lr=8.5766e-04) (hash(x)=6183574)
661
+ 633 train 7.556254 (lr=8.5469e-04) (hash(x)=6576136)
662
+ 634 train 7.586408 (lr=8.5172e-04) (hash(x)=6157346)
663
+ 635 train 7.575245 (lr=8.4876e-04) (hash(x)=6077647)
664
+ 636 train 7.789124 (lr=8.4579e-04) (hash(x)=8369488)
665
+ 637 train 7.617842 (lr=8.4282e-04) (hash(x)=6032381)
666
+ 638 train 7.538023 (lr=8.3985e-04) (hash(x)=5924756)
667
+ 639 train 7.577548 (lr=8.3688e-04) (hash(x)=8410716)
668
+ 640 train 7.551620 (lr=8.3391e-04) (hash(x)=5803440)
669
+ 641 train 7.632987 (lr=8.3094e-04) (hash(x)=6212516)
670
+ 642 train 7.699341 (lr=8.2797e-04) (hash(x)=7165945)
671
+ 643 train 7.438562 (lr=8.2500e-04) (hash(x)=5298545)
672
+ 644 train 7.668376 (lr=8.2203e-04) (hash(x)=6281739)
673
+ 645 train 7.654253 (lr=8.1906e-04) (hash(x)=5779073)
674
+ 646 train 7.527207 (lr=8.1609e-04) (hash(x)=5090107)
675
+ 647 train 7.587166 (lr=8.1312e-04) (hash(x)=6340947)
676
+ 648 train 7.364984 (lr=8.1015e-04) (hash(x)=4741418)
677
+ 649 train 7.325338 (lr=8.0718e-04) (hash(x)=6530441)
678
+ 650 val loss 7.6361
679
+ 650 val perplexity 2071.5544
680
+ 650 train 7.840420 (lr=8.0421e-04) (hash(x)=6396041)
681
+ 651 train 7.508986 (lr=8.0124e-04) (hash(x)=7818007)
682
+ 652 train 7.488750 (lr=7.9828e-04) (hash(x)=6478457)
683
+ 653 train 7.541171 (lr=7.9531e-04) (hash(x)=6277435)
684
+ 654 train 7.540680 (lr=7.9234e-04) (hash(x)=6042369)
685
+ 655 train 7.538291 (lr=7.8938e-04) (hash(x)=6226320)
686
+ 656 train 7.554942 (lr=7.8641e-04) (hash(x)=5826883)
687
+ 657 train 7.671491 (lr=7.8345e-04) (hash(x)=5463751)
688
+ 658 train 7.602089 (lr=7.8048e-04) (hash(x)=6433080)
689
+ 659 train 7.463463 (lr=7.7752e-04) (hash(x)=6579329)
690
+ 660 train 7.473896 (lr=7.7456e-04) (hash(x)=4724703)
691
+ 661 train 7.781272 (lr=7.7160e-04) (hash(x)=6660540)
692
+ 662 train 7.605873 (lr=7.6864e-04) (hash(x)=5918756)
693
+ 663 train 7.552368 (lr=7.6568e-04) (hash(x)=6491272)
694
+ 664 train 7.484970 (lr=7.6272e-04) (hash(x)=5452863)
695
+ 665 train 7.659924 (lr=7.5976e-04) (hash(x)=6766947)
696
+ 666 train 7.968570 (lr=7.5681e-04) (hash(x)=7459504)
697
+ 667 train 7.677453 (lr=7.5385e-04) (hash(x)=3721218)
698
+ 668 train 7.841445 (lr=7.5090e-04) (hash(x)=7929571)
699
+ 669 train 7.558710 (lr=7.4795e-04) (hash(x)=5229898)
700
+ 670 train 7.594363 (lr=7.4500e-04) (hash(x)=6083090)
701
+ 671 train 7.448383 (lr=7.4205e-04) (hash(x)=6054428)
702
+ 672 train 7.587810 (lr=7.3910e-04) (hash(x)=6072563)
703
+ 673 train 7.721066 (lr=7.3616e-04) (hash(x)=6290464)
704
+ 674 train 7.788572 (lr=7.3322e-04) (hash(x)=6314742)
705
+ 675 train 7.465023 (lr=7.3027e-04) (hash(x)=6904498)
706
+ 676 train 7.598791 (lr=7.2733e-04) (hash(x)=5378107)
707
+ 677 train 7.698100 (lr=7.2440e-04) (hash(x)=7765805)
708
+ 678 train 7.533567 (lr=7.2146e-04) (hash(x)=5871280)
709
+ 679 train 7.679964 (lr=7.1853e-04) (hash(x)=6543764)
710
+ 680 train 7.574137 (lr=7.1560e-04) (hash(x)=7369665)
711
+ 681 train 7.653553 (lr=7.1267e-04) (hash(x)=7204955)
712
+ 682 train 7.609015 (lr=7.0974e-04) (hash(x)=5856543)
713
+ 683 train 7.487786 (lr=7.0681e-04) (hash(x)=4423711)
714
+ 684 train 7.443988 (lr=7.0389e-04) (hash(x)=5818153)
715
+ 685 train 7.553670 (lr=7.0097e-04) (hash(x)=5343477)
716
+ 686 train 7.713219 (lr=6.9805e-04) (hash(x)=6581702)
717
+ 687 train 7.566154 (lr=6.9514e-04) (hash(x)=6595856)
718
+ 688 train 7.521027 (lr=6.9222e-04) (hash(x)=5325378)
719
+ 689 train 7.420962 (lr=6.8931e-04) (hash(x)=7229015)
720
+ 690 train 7.605985 (lr=6.8640e-04) (hash(x)=5918186)
721
+ 691 train 7.498294 (lr=6.8350e-04) (hash(x)=5916034)
722
+ 692 train 7.496569 (lr=6.8060e-04) (hash(x)=5280722)
723
+ 693 train 7.917385 (lr=6.7770e-04) (hash(x)=6939521)
724
+ 694 train 7.619590 (lr=6.7480e-04) (hash(x)=5172566)
725
+ 695 train 7.495289 (lr=6.7190e-04) (hash(x)=5492185)
726
+ 696 train 7.365283 (lr=6.6901e-04) (hash(x)=5779011)
727
+ 697 train 7.616727 (lr=6.6613e-04) (hash(x)=6898006)
728
+ 698 train 7.470603 (lr=6.6324e-04) (hash(x)=5779217)
729
+ 699 train 7.336102 (lr=6.6036e-04) (hash(x)=5462008)
730
+ 700 val loss 7.6225
731
+ 700 val perplexity 2043.7544
732
+ 700 train 7.462539 (lr=6.5748e-04) (hash(x)=6243843)
733
+ 701 train 7.540549 (lr=6.5460e-04) (hash(x)=7821902)
734
+ 702 train 7.612834 (lr=6.5173e-04) (hash(x)=5222501)
735
+ 703 train 7.602217 (lr=6.4886e-04) (hash(x)=6608108)
736
+ 704 train 7.705032 (lr=6.4600e-04) (hash(x)=7602879)
737
+ 705 train 7.618752 (lr=6.4314e-04) (hash(x)=6295394)
738
+ 706 train 7.600069 (lr=6.4028e-04) (hash(x)=6579907)
739
+ 707 train 7.489980 (lr=6.3742e-04) (hash(x)=6803449)
740
+ 708 train 7.550358 (lr=6.3457e-04) (hash(x)=5413733)
741
+ 709 train 7.527575 (lr=6.3172e-04) (hash(x)=7444134)
742
+ 710 train 7.647399 (lr=6.2888e-04) (hash(x)=7028652)
743
+ 711 train 7.462183 (lr=6.2604e-04) (hash(x)=5361923)
744
+ 712 train 7.569002 (lr=6.2320e-04) (hash(x)=5478293)
745
+ 713 train 7.536158 (lr=6.2037e-04) (hash(x)=7803039)
746
+ 714 train 7.724354 (lr=6.1754e-04) (hash(x)=6469017)
747
+ 715 train 7.611896 (lr=6.1472e-04) (hash(x)=7042578)
748
+ 716 train 7.663189 (lr=6.1190e-04) (hash(x)=6927957)
749
+ 717 train 7.504277 (lr=6.0908e-04) (hash(x)=6647007)
750
+ 718 train 7.477106 (lr=6.0627e-04) (hash(x)=6230888)
751
+ 719 train 7.702402 (lr=6.0346e-04) (hash(x)=6303658)
752
+ 720 train 7.739843 (lr=6.0066e-04) (hash(x)=6769880)
753
+ 721 train 7.608555 (lr=5.9786e-04) (hash(x)=6158927)
754
+ 722 train 7.596186 (lr=5.9507e-04) (hash(x)=5991631)
755
+ 723 train 7.476029 (lr=5.9228e-04) (hash(x)=6419656)
756
+ 724 train 7.471935 (lr=5.8949e-04) (hash(x)=7739209)
757
+ 725 train 7.424521 (lr=5.8671e-04) (hash(x)=7891271)
758
+ 726 train 7.759793 (lr=5.8393e-04) (hash(x)=7171433)
759
+ 727 train 7.544987 (lr=5.8116e-04) (hash(x)=7099888)
760
+ 728 train 7.577643 (lr=5.7839e-04) (hash(x)=6103730)
761
+ 729 train 7.703739 (lr=5.7563e-04) (hash(x)=6352424)
762
+ 730 train 7.565410 (lr=5.7288e-04) (hash(x)=4893671)
763
+ 731 train 7.498900 (lr=5.7012e-04) (hash(x)=5863145)
764
+ 732 train 7.607841 (lr=5.6737e-04) (hash(x)=5502233)
765
+ 733 train 7.504966 (lr=5.6463e-04) (hash(x)=6844597)
766
+ 734 train 7.534903 (lr=5.6189e-04) (hash(x)=6744526)
767
+ 735 train 7.686005 (lr=5.5916e-04) (hash(x)=6009861)
768
+ 736 train 7.573351 (lr=5.5643e-04) (hash(x)=6152358)
769
+ 737 train 7.461759 (lr=5.5371e-04) (hash(x)=5374957)
770
+ 738 train 7.501070 (lr=5.5100e-04) (hash(x)=6495309)
771
+ 739 train 7.637182 (lr=5.4828e-04) (hash(x)=3923225)
772
+ 740 train 7.448307 (lr=5.4558e-04) (hash(x)=6255243)
773
+ 741 train 7.557774 (lr=5.4288e-04) (hash(x)=5848862)
774
+ 742 train 7.599139 (lr=5.4018e-04) (hash(x)=4535000)
775
+ 743 train 7.494648 (lr=5.3749e-04) (hash(x)=4440634)
776
+ 744 train 7.465955 (lr=5.3481e-04) (hash(x)=5222077)
777
+ 745 train 7.495412 (lr=5.3213e-04) (hash(x)=4788998)
778
+ 746 train 7.500865 (lr=5.2946e-04) (hash(x)=6108350)
779
+ 747 train 7.423533 (lr=5.2679e-04) (hash(x)=7378165)
780
+ 748 train 7.499710 (lr=5.2413e-04) (hash(x)=5475646)
781
+ 749 train 7.529293 (lr=5.2147e-04) (hash(x)=4771243)
782
+ 750 val loss 7.5893
783
+ 750 val perplexity 1977.0096
784
+ 750 train 7.397238 (lr=5.1882e-04) (hash(x)=6668088)
785
+ 751 train 7.590489 (lr=5.1618e-04) (hash(x)=5716089)
786
+ 752 train 7.562916 (lr=5.1354e-04) (hash(x)=6991946)
787
+ 753 train 7.448546 (lr=5.1091e-04) (hash(x)=7991009)
788
+ 754 train 7.569211 (lr=5.0828e-04) (hash(x)=6037018)
789
+ 755 train 7.537712 (lr=5.0566e-04) (hash(x)=5348607)
790
+ 756 train 7.666280 (lr=5.0305e-04) (hash(x)=5190526)
791
+ 757 train 7.569938 (lr=5.0044e-04) (hash(x)=6036146)
792
+ 758 train 7.443770 (lr=4.9784e-04) (hash(x)=5329783)
793
+ 759 train 7.552356 (lr=4.9525e-04) (hash(x)=6658718)
794
+ 760 train 7.467426 (lr=4.9266e-04) (hash(x)=4820939)
795
+ 761 train 7.446311 (lr=4.9008e-04) (hash(x)=6176253)
796
+ 762 train 7.435369 (lr=4.8750e-04) (hash(x)=5316750)
797
+ 763 train 7.516755 (lr=4.8493e-04) (hash(x)=4542895)
798
+ 764 train 7.365388 (lr=4.8237e-04) (hash(x)=5395673)
799
+ 765 train 7.404041 (lr=4.7981e-04) (hash(x)=5549073)
800
+ 766 train 7.654234 (lr=4.7726e-04) (hash(x)=5233723)
801
+ 767 train 7.677663 (lr=4.7472e-04) (hash(x)=5342783)
802
+ 768 train 7.625177 (lr=4.7219e-04) (hash(x)=5552448)
803
+ 769 train 7.748279 (lr=4.6966e-04) (hash(x)=6549546)
804
+ 770 train 7.440083 (lr=4.6714e-04) (hash(x)=3816493)
805
+ 771 train 7.488873 (lr=4.6462e-04) (hash(x)=6170599)
806
+ 772 train 7.953141 (lr=4.6211e-04) (hash(x)=6708759)
807
+ 773 train 7.587832 (lr=4.5961e-04) (hash(x)=8483134)
808
+ 774 train 8.170832 (lr=4.5712e-04) (hash(x)=5583732)
809
+ 775 train 7.560757 (lr=4.5463e-04) (hash(x)=4865036)
810
+ 776 train 7.487570 (lr=4.5215e-04) (hash(x)=7234336)
811
+ 777 train 7.449942 (lr=4.4968e-04) (hash(x)=4433890)
812
+ 778 train 7.758094 (lr=4.4722e-04) (hash(x)=7484636)
813
+ 779 train 8.282282 (lr=4.4476e-04) (hash(x)=6738597)
814
+ 780 train 7.519242 (lr=4.4231e-04) (hash(x)=6752421)
815
+ 781 train 7.703094 (lr=4.3987e-04) (hash(x)=5459620)
816
+ 782 train 7.611434 (lr=4.3743e-04) (hash(x)=6718131)
817
+ 783 train 7.609176 (lr=4.3500e-04) (hash(x)=4604074)
818
+ 784 train 7.482567 (lr=4.3258e-04) (hash(x)=5946124)
819
+ 785 train 7.721173 (lr=4.3017e-04) (hash(x)=8045501)
820
+ 786 train 7.842819 (lr=4.2776e-04) (hash(x)=5712725)
821
+ 787 train 7.580128 (lr=4.2537e-04) (hash(x)=6176464)
822
+ 788 train 7.630579 (lr=4.2298e-04) (hash(x)=6276702)
823
+ 789 train 7.444261 (lr=4.2060e-04) (hash(x)=6127317)
824
+ 790 train 7.592384 (lr=4.1822e-04) (hash(x)=7221374)
825
+ 791 train 7.568966 (lr=4.1586e-04) (hash(x)=5571462)
826
+ 792 train 7.571847 (lr=4.1350e-04) (hash(x)=7518021)
827
+ 793 train 7.574230 (lr=4.1115e-04) (hash(x)=6640755)
828
+ 794 train 7.660143 (lr=4.0880e-04) (hash(x)=6870405)
829
+ 795 train 7.625866 (lr=4.0647e-04) (hash(x)=6147295)
830
+ 796 train 7.511385 (lr=4.0414e-04) (hash(x)=6782583)
831
+ 797 train 7.619304 (lr=4.0183e-04) (hash(x)=6120461)
832
+ 798 train 7.563759 (lr=3.9952e-04) (hash(x)=5893870)
833
+ 799 train 7.497763 (lr=3.9722e-04) (hash(x)=5123928)
834
+ 800 val loss 7.5837
835
+ 800 val perplexity 1965.9197
836
+ 800 train 7.555704 (lr=3.9492e-04) (hash(x)=7037251)
837
+ 801 train 7.566846 (lr=3.9264e-04) (hash(x)=5428813)
838
+ 802 train 7.485403 (lr=3.9036e-04) (hash(x)=5550387)
839
+ 803 train 7.583901 (lr=3.8809e-04) (hash(x)=6448586)
840
+ 804 train 7.697530 (lr=3.8583e-04) (hash(x)=5916934)
841
+ 805 train 7.666185 (lr=3.8358e-04) (hash(x)=8116987)
842
+ 806 train 7.624870 (lr=3.8134e-04) (hash(x)=6210899)
843
+ 807 train 7.726729 (lr=3.7910e-04) (hash(x)=5974342)
844
+ 808 train 7.744931 (lr=3.7688e-04) (hash(x)=7236199)
845
+ 809 train 7.696422 (lr=3.7466e-04) (hash(x)=5261336)
846
+ 810 train 7.512304 (lr=3.7245e-04) (hash(x)=6310720)
847
+ 811 train 7.552894 (lr=3.7026e-04) (hash(x)=5307056)
848
+ 812 train 7.396852 (lr=3.6806e-04) (hash(x)=7356633)
849
+ 813 train 7.492779 (lr=3.6588e-04) (hash(x)=5865119)
850
+ 814 train 7.403041 (lr=3.6371e-04) (hash(x)=5979205)
851
+ 815 train 7.554278 (lr=3.6155e-04) (hash(x)=6672380)
852
+ 816 train 7.561201 (lr=3.5939e-04) (hash(x)=5565030)
853
+ 817 train 7.517117 (lr=3.5725e-04) (hash(x)=6886810)
854
+ 818 train 7.567135 (lr=3.5511e-04) (hash(x)=5926478)
855
+ 819 train 7.754086 (lr=3.5298e-04) (hash(x)=6564071)
856
+ 820 train 7.419878 (lr=3.5086e-04) (hash(x)=5024280)
857
+ 821 train 7.438396 (lr=3.4875e-04) (hash(x)=7705089)
858
+ 822 train 7.412940 (lr=3.4665e-04) (hash(x)=4037371)
859
+ 823 train 7.511008 (lr=3.4456e-04) (hash(x)=5773073)
860
+ 824 train 7.507897 (lr=3.4248e-04) (hash(x)=7158139)
861
+ 825 train 7.372197 (lr=3.4041e-04) (hash(x)=5905976)
862
+ 826 train 7.747830 (lr=3.3835e-04) (hash(x)=13897369)
863
+ 827 train 8.723680 (lr=3.3629e-04) (hash(x)=6936653)
864
+ 828 train 7.549129 (lr=3.3425e-04) (hash(x)=6592246)
865
+ 829 train 7.505181 (lr=3.3221e-04) (hash(x)=6592773)
866
+ 830 train 7.583286 (lr=3.3019e-04) (hash(x)=5461178)
867
+ 831 train 7.372470 (lr=3.2817e-04) (hash(x)=6796551)
868
+ 832 train 7.574071 (lr=3.2617e-04) (hash(x)=4721691)
869
+ 833 train 7.497634 (lr=3.2417e-04) (hash(x)=6151589)
870
+ 834 train 7.637172 (lr=3.2219e-04) (hash(x)=7008995)
871
+ 835 train 7.927523 (lr=3.2021e-04) (hash(x)=7045164)
872
+ 836 train 7.545493 (lr=3.1824e-04) (hash(x)=6614221)
873
+ 837 train 7.592062 (lr=3.1629e-04) (hash(x)=6650506)
874
+ 838 train 7.424112 (lr=3.1434e-04) (hash(x)=6363762)
875
+ 839 train 7.557069 (lr=3.1240e-04) (hash(x)=7839748)
876
+ 840 train 7.492625 (lr=3.1047e-04) (hash(x)=5124077)
877
+ 841 train 7.925664 (lr=3.0856e-04) (hash(x)=6028671)
878
+ 842 train 7.423744 (lr=3.0665e-04) (hash(x)=5427175)
879
+ 843 train 7.709420 (lr=3.0475e-04) (hash(x)=6118408)
880
+ 844 train 7.540825 (lr=3.0286e-04) (hash(x)=6022552)
881
+ 845 train 7.536815 (lr=3.0099e-04) (hash(x)=5757526)
882
+ 846 train 7.959790 (lr=2.9912e-04) (hash(x)=6476549)
883
+ 847 train 7.370636 (lr=2.9726e-04) (hash(x)=6030172)
884
+ 848 train 7.656048 (lr=2.9542e-04) (hash(x)=5917229)
885
+ 849 train 7.745681 (lr=2.9358e-04) (hash(x)=6719882)
886
+ 850 val loss 7.5796
887
+ 850 val perplexity 1957.7819
888
+ 850 train 7.622768 (lr=2.9175e-04) (hash(x)=5441498)
889
+ 851 train 7.512782 (lr=2.8994e-04) (hash(x)=5028050)
890
+ 852 train 7.558966 (lr=2.8813e-04) (hash(x)=8487491)
891
+ 853 train 7.544215 (lr=2.8634e-04) (hash(x)=5682910)
892
+ 854 train 7.736517 (lr=2.8455e-04) (hash(x)=6781362)
893
+ 855 train 7.539141 (lr=2.8278e-04) (hash(x)=5747891)
894
+ 856 train 7.601834 (lr=2.8102e-04) (hash(x)=7627661)
895
+ 857 train 7.283474 (lr=2.7926e-04) (hash(x)=5734208)
896
+ 858 train 7.397617 (lr=2.7752e-04) (hash(x)=6162015)
897
+ 859 train 7.627619 (lr=2.7579e-04) (hash(x)=7024381)
898
+ 860 train 7.796052 (lr=2.7407e-04) (hash(x)=5198498)
899
+ 861 train 7.749965 (lr=2.7236e-04) (hash(x)=7814361)
900
+ 862 train 7.675684 (lr=2.7066e-04) (hash(x)=7394995)
901
+ 863 train 7.560795 (lr=2.6897e-04) (hash(x)=5765893)
902
+ 864 train 7.387345 (lr=2.6729e-04) (hash(x)=4990435)
903
+ 865 train 7.511179 (lr=2.6562e-04) (hash(x)=5572048)
904
+ 866 train 7.664811 (lr=2.6396e-04) (hash(x)=7093293)
905
+ 867 train 7.571619 (lr=2.6232e-04) (hash(x)=5610058)
906
+ 868 train 7.465153 (lr=2.6068e-04) (hash(x)=6771468)
907
+ 869 train 7.491187 (lr=2.5906e-04) (hash(x)=4459483)
908
+ 870 train 7.413995 (lr=2.5745e-04) (hash(x)=4926546)
909
+ 871 train 7.521644 (lr=2.5584e-04) (hash(x)=6758411)
910
+ 872 train 7.455358 (lr=2.5425e-04) (hash(x)=7394277)
911
+ 873 train 7.590964 (lr=2.5267e-04) (hash(x)=6701381)
912
+ 874 train 7.916701 (lr=2.5110e-04) (hash(x)=8120966)
913
+ 875 train 7.460240 (lr=2.4955e-04) (hash(x)=5416400)
914
+ 876 train 7.477219 (lr=2.4800e-04) (hash(x)=6292778)
915
+ 877 train 7.408143 (lr=2.4646e-04) (hash(x)=5919569)
916
+ 878 train 7.704479 (lr=2.4494e-04) (hash(x)=7561740)
917
+ 879 train 7.487830 (lr=2.4343e-04) (hash(x)=5282870)
918
+ 880 train 7.499385 (lr=2.4192e-04) (hash(x)=5473238)
919
+ 881 train 7.409494 (lr=2.4043e-04) (hash(x)=8072181)
920
+ 882 train 7.410217 (lr=2.3895e-04) (hash(x)=6757322)
921
+ 883 train 7.394196 (lr=2.3749e-04) (hash(x)=7996770)
922
+ 884 train 7.466973 (lr=2.3603e-04) (hash(x)=3181021)
923
+ 885 train 7.502339 (lr=2.3458e-04) (hash(x)=5093519)
924
+ 886 train 7.471228 (lr=2.3315e-04) (hash(x)=5391352)
925
+ 887 train 7.501481 (lr=2.3173e-04) (hash(x)=6270917)
926
+ 888 train 7.508416 (lr=2.3032e-04) (hash(x)=5818195)
927
+ 889 train 7.230391 (lr=2.2892e-04) (hash(x)=5852780)
928
+ 890 train 7.160896 (lr=2.2753e-04) (hash(x)=5777928)
929
+ 891 train 7.374829 (lr=2.2615e-04) (hash(x)=5327728)
930
+ 892 train 7.411976 (lr=2.2479e-04) (hash(x)=7651281)
931
+ 893 train 7.607723 (lr=2.2344e-04) (hash(x)=6260765)
932
+ 894 train 7.500413 (lr=2.2209e-04) (hash(x)=6265102)
933
+ 895 train 7.650255 (lr=2.2076e-04) (hash(x)=6212894)
934
+ 896 train 7.437178 (lr=2.1945e-04) (hash(x)=6901967)
935
+ 897 train 7.501818 (lr=2.1814e-04) (hash(x)=6724010)
936
+ 898 train 7.492483 (lr=2.1685e-04) (hash(x)=6299033)
937
+ 899 train 7.647099 (lr=2.1556e-04) (hash(x)=6059934)
938
+ 900 val loss 7.5669
939
+ 900 val perplexity 1933.2069
940
+ 900 train 7.378261 (lr=2.1429e-04) (hash(x)=6029696)
941
+ 901 train 7.303301 (lr=2.1303e-04) (hash(x)=5587382)
942
+ 902 train 7.497371 (lr=2.1179e-04) (hash(x)=5890076)
943
+ 903 train 7.449915 (lr=2.1055e-04) (hash(x)=7274108)
944
+ 904 train 7.500693 (lr=2.0933e-04) (hash(x)=7975528)
945
+ 905 train 7.539095 (lr=2.0812e-04) (hash(x)=6085396)
946
+ 906 train 7.434281 (lr=2.0692e-04) (hash(x)=6724496)
947
+ 907 train 7.435217 (lr=2.0573e-04) (hash(x)=4784666)
948
+ 908 train 7.440199 (lr=2.0455e-04) (hash(x)=7193845)
949
+ 909 train 7.579875 (lr=2.0339e-04) (hash(x)=6288638)
950
+ 910 train 7.569043 (lr=2.0224e-04) (hash(x)=5685285)
951
+ 911 train 7.536479 (lr=2.0110e-04) (hash(x)=7714288)
952
+ 912 train 7.612863 (lr=1.9997e-04) (hash(x)=5983679)
953
+ 913 train 7.579433 (lr=1.9885e-04) (hash(x)=5534748)
954
+ 914 train 7.517544 (lr=1.9775e-04) (hash(x)=6545471)
955
+ 915 train 7.518990 (lr=1.9666e-04) (hash(x)=6361270)
956
+ 916 train 7.514293 (lr=1.9558e-04) (hash(x)=6257744)
957
+ 917 train 7.525247 (lr=1.9451e-04) (hash(x)=4682956)
958
+ 918 train 7.413147 (lr=1.9346e-04) (hash(x)=4521756)
959
+ 919 train 7.564897 (lr=1.9242e-04) (hash(x)=7568935)
960
+ 920 train 7.471497 (lr=1.9139e-04) (hash(x)=7969353)
961
+ 921 train 7.646673 (lr=1.9037e-04) (hash(x)=7973311)
962
+ 922 train 7.386485 (lr=1.8936e-04) (hash(x)=4843937)
963
+ 923 train 7.502776 (lr=1.8837e-04) (hash(x)=6987352)
964
+ 924 train 7.595049 (lr=1.8739e-04) (hash(x)=7179418)
965
+ 925 train 7.296122 (lr=1.8642e-04) (hash(x)=5413904)
966
+ 926 train 7.653337 (lr=1.8547e-04) (hash(x)=6028105)
967
+ 927 train 7.589232 (lr=1.8452e-04) (hash(x)=6732924)
968
+ 928 train 7.501529 (lr=1.8359e-04) (hash(x)=4547651)
969
+ 929 train 7.626449 (lr=1.8267e-04) (hash(x)=6219754)
970
+ 930 train 7.452164 (lr=1.8176e-04) (hash(x)=7465186)
971
+ 931 train 7.560885 (lr=1.8087e-04) (hash(x)=6839070)
972
+ 932 train 7.454355 (lr=1.7999e-04) (hash(x)=5390324)
973
+ 933 train 7.414926 (lr=1.7912e-04) (hash(x)=6090120)
974
+ 934 train 7.994825 (lr=1.7826e-04) (hash(x)=7123181)
975
+ 935 train 7.701993 (lr=1.7742e-04) (hash(x)=7115176)
976
+ 936 train 7.703364 (lr=1.7659e-04) (hash(x)=7016990)
977
+ 937 train 7.472511 (lr=1.7577e-04) (hash(x)=4989299)
978
+ 938 train 7.510314 (lr=1.7496e-04) (hash(x)=6783311)
979
+ 939 train 7.470667 (lr=1.7417e-04) (hash(x)=7497590)
980
+ 940 train 7.688316 (lr=1.7339e-04) (hash(x)=6767514)
981
+ 941 train 7.464921 (lr=1.7262e-04) (hash(x)=3977557)
982
+ 942 train 7.567982 (lr=1.7186e-04) (hash(x)=4806492)
983
+ 943 train 7.476320 (lr=1.7112e-04) (hash(x)=5524165)
984
+ 944 train 7.543875 (lr=1.7039e-04) (hash(x)=4930752)
985
+ 945 train 7.524450 (lr=1.6967e-04) (hash(x)=5124936)
986
+ 946 train 7.411867 (lr=1.6896e-04) (hash(x)=6877351)
987
+ 947 train 7.672591 (lr=1.6827e-04) (hash(x)=5522119)
988
+ 948 train 7.490720 (lr=1.6759e-04) (hash(x)=6183512)
989
+ 949 train 7.408257 (lr=1.6692e-04) (hash(x)=5030865)
990
+ 950 val loss 7.5601
991
+ 950 val perplexity 1919.9873
992
+ 950 train 7.500966 (lr=1.6627e-04) (hash(x)=6619812)
993
+ 951 train 7.511671 (lr=1.6563e-04) (hash(x)=6411783)
994
+ 952 train 7.516887 (lr=1.6500e-04) (hash(x)=6748982)
995
+ 953 train 7.419617 (lr=1.6438e-04) (hash(x)=6581937)
996
+ 954 train 7.526221 (lr=1.6378e-04) (hash(x)=7881022)
997
+ 955 train 7.487035 (lr=1.6319e-04) (hash(x)=5116910)
998
+ 956 train 7.278900 (lr=1.6261e-04) (hash(x)=4880833)
999
+ 957 train 7.331339 (lr=1.6205e-04) (hash(x)=7353459)
1000
+ 958 train 7.407221 (lr=1.6149e-04) (hash(x)=4951302)
1001
+ 959 train 7.323879 (lr=1.6095e-04) (hash(x)=6344535)
1002
+ 960 train 7.895007 (lr=1.6043e-04) (hash(x)=7411763)
1003
+ 961 train 7.489126 (lr=1.5991e-04) (hash(x)=5840169)
1004
+ 962 train 7.568331 (lr=1.5941e-04) (hash(x)=5784788)
1005
+ 963 train 7.397121 (lr=1.5893e-04) (hash(x)=6754290)
1006
+ 964 train 7.637458 (lr=1.5845e-04) (hash(x)=6369963)
1007
+ 965 train 7.711027 (lr=1.5799e-04) (hash(x)=6484573)
1008
+ 966 train 7.337222 (lr=1.5754e-04) (hash(x)=6516338)
1009
+ 967 train 7.382394 (lr=1.5710e-04) (hash(x)=5260125)
1010
+ 968 train 7.494167 (lr=1.5668e-04) (hash(x)=5781624)
1011
+ 969 train 7.310863 (lr=1.5627e-04) (hash(x)=4881179)
1012
+ 970 train 7.418378 (lr=1.5587e-04) (hash(x)=6707821)
1013
+ 971 train 7.554861 (lr=1.5549e-04) (hash(x)=5001634)
1014
+ 972 train 7.415101 (lr=1.5512e-04) (hash(x)=4402949)
1015
+ 973 train 7.441772 (lr=1.5476e-04) (hash(x)=7424810)
1016
+ 974 train 7.286432 (lr=1.5441e-04) (hash(x)=4814803)
1017
+ 975 train 7.219942 (lr=1.5408e-04) (hash(x)=5176431)
1018
+ 976 train 7.409331 (lr=1.5376e-04) (hash(x)=5769393)
1019
+ 977 train 7.354766 (lr=1.5345e-04) (hash(x)=5421624)
1020
+ 978 train 7.571158 (lr=1.5316e-04) (hash(x)=5999773)
1021
+ 979 train 7.332287 (lr=1.5288e-04) (hash(x)=5674439)
1022
+ 980 train 7.408945 (lr=1.5261e-04) (hash(x)=6345351)
1023
+ 981 train 7.344377 (lr=1.5236e-04) (hash(x)=5819157)
1024
+ 982 train 7.541343 (lr=1.5212e-04) (hash(x)=11159355)
1025
+ 983 train 7.520033 (lr=1.5189e-04) (hash(x)=6506856)
1026
+ 984 train 7.441432 (lr=1.5167e-04) (hash(x)=6331023)
1027
+ 985 train 7.332491 (lr=1.5147e-04) (hash(x)=6119124)
1028
+ 986 train 7.531219 (lr=1.5128e-04) (hash(x)=8142528)
1029
+ 987 train 7.681267 (lr=1.5110e-04) (hash(x)=6241669)
1030
+ 988 train 7.350418 (lr=1.5094e-04) (hash(x)=3852991)
1031
+ 989 train 7.121955 (lr=1.5079e-04) (hash(x)=4992756)
1032
+ 990 train 7.765850 (lr=1.5065e-04) (hash(x)=7097145)
1033
+ 991 train 7.562995 (lr=1.5053e-04) (hash(x)=7277927)
1034
+ 992 train 7.618833 (lr=1.5042e-04) (hash(x)=5661400)
1035
+ 993 train 7.550998 (lr=1.5032e-04) (hash(x)=6622097)
1036
+ 994 train 7.666943 (lr=1.5024e-04) (hash(x)=5016804)
1037
+ 995 train 7.728390 (lr=1.5016e-04) (hash(x)=7026582)
1038
+ 996 train 7.524142 (lr=1.5010e-04) (hash(x)=5639136)
1039
+ 997 train 7.696088 (lr=1.5006e-04) (hash(x)=5840987)
1040
+ 998 train 7.327505 (lr=1.5003e-04) (hash(x)=5265410)
1041
+ 999 val loss 7.5521
1042
+ 999 val perplexity 1904.7271
1043
+ 999 train 7.315741 (lr=1.5001e-04) (hash(x)=4351074)
lr1.5e-3_total_batch_size20480_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6108a3fb09a96ee441bd0473836e4be971c040257d52cc9ca2692f1b00896c29
3
+ size 36163010
lr1.5e-3_total_batch_size20480_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d44338d40cc252c9db750d7cfad3a73b4863b7147221ed8621912cb92c794800
3
+ size 70764358