andrew-healey commited on
Commit
1f5dc6c
·
verified ·
1 Parent(s): 8cc4804

Upload folder using huggingface_hub

Browse files
lr3e-3_total_batch_size5120_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr3e-3_total_batch_size5120_seq_len512", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 512, "batch_size": null, "total_batch_size": 5120, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.003, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr3e-3_total_batch_size5120_seq_len512/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr3e-3_total_batch_size5120_seq_len512/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7800
3
+ 0 val perplexity 130608.6406
4
+ 0 train 11.762018 (lr=1.0490e-05) (hash(x)=22886834)
5
+ 1 train 11.760261 (lr=2.0979e-05) (hash(x)=26375038)
6
+ 2 train 11.718584 (lr=3.1469e-05) (hash(x)=30598777)
7
+ 3 train 11.717956 (lr=4.1958e-05) (hash(x)=27234506)
8
+ 4 train 11.631213 (lr=5.2448e-05) (hash(x)=27767880)
9
+ 5 train 11.600010 (lr=6.2937e-05) (hash(x)=23702020)
10
+ 6 train 11.535187 (lr=7.3427e-05) (hash(x)=31986844)
11
+ 7 train 11.432528 (lr=8.3916e-05) (hash(x)=20782690)
12
+ 8 train 11.322848 (lr=9.4406e-05) (hash(x)=25201599)
13
+ 9 train 11.262590 (lr=1.0490e-04) (hash(x)=23094976)
14
+ 10 train 11.144441 (lr=1.1538e-04) (hash(x)=23841096)
15
+ 11 train 10.959437 (lr=1.2587e-04) (hash(x)=26532095)
16
+ 12 train 10.889548 (lr=1.3636e-04) (hash(x)=24432298)
17
+ 13 train 10.736397 (lr=1.4685e-04) (hash(x)=27151649)
18
+ 14 train 10.616040 (lr=1.5734e-04) (hash(x)=24596846)
19
+ 15 train 10.477584 (lr=1.6783e-04) (hash(x)=23890908)
20
+ 16 train 10.376515 (lr=1.7832e-04) (hash(x)=28913955)
21
+ 17 train 10.278032 (lr=1.8881e-04) (hash(x)=25588236)
22
+ 18 train 10.349444 (lr=1.9930e-04) (hash(x)=23770034)
23
+ 19 train 10.230525 (lr=2.0979e-04) (hash(x)=24011372)
24
+ 20 train 10.026118 (lr=2.2028e-04) (hash(x)=25441898)
25
+ 21 train 10.201641 (lr=2.3077e-04) (hash(x)=28375581)
26
+ 22 train 10.003260 (lr=2.4126e-04) (hash(x)=24046679)
27
+ 23 train 10.058246 (lr=2.5175e-04) (hash(x)=24611628)
28
+ 24 train 10.012608 (lr=2.6224e-04) (hash(x)=26169030)
29
+ 25 train 9.996007 (lr=2.7273e-04) (hash(x)=30298407)
30
+ 26 train 9.851203 (lr=2.8322e-04) (hash(x)=23711112)
31
+ 27 train 9.637230 (lr=2.9371e-04) (hash(x)=19245352)
32
+ 28 train 9.547346 (lr=3.0420e-04) (hash(x)=21529136)
33
+ 29 train 9.813288 (lr=3.1469e-04) (hash(x)=28936608)
34
+ 30 train 9.446461 (lr=3.2517e-04) (hash(x)=24339013)
35
+ 31 train 9.282298 (lr=3.3566e-04) (hash(x)=25767553)
36
+ 32 train 9.300721 (lr=3.4615e-04) (hash(x)=26439905)
37
+ 33 train 9.572387 (lr=3.5664e-04) (hash(x)=31093473)
38
+ 34 train 9.186543 (lr=3.6713e-04) (hash(x)=25450374)
39
+ 35 train 9.142092 (lr=3.7762e-04) (hash(x)=24809873)
40
+ 36 train 9.229797 (lr=3.8811e-04) (hash(x)=23253252)
41
+ 37 train 9.266380 (lr=3.9860e-04) (hash(x)=27852919)
42
+ 38 train 8.981488 (lr=4.0909e-04) (hash(x)=23327497)
43
+ 39 train 8.870475 (lr=4.1958e-04) (hash(x)=22512166)
44
+ 40 train 9.071976 (lr=4.3007e-04) (hash(x)=22859419)
45
+ 41 train 8.877315 (lr=4.4056e-04) (hash(x)=27620338)
46
+ 42 train 8.834985 (lr=4.5105e-04) (hash(x)=26397837)
47
+ 43 train 8.845454 (lr=4.6154e-04) (hash(x)=28092148)
48
+ 44 train 8.718257 (lr=4.7203e-04) (hash(x)=24662703)
49
+ 45 train 8.935245 (lr=4.8252e-04) (hash(x)=27938767)
50
+ 46 train 9.097150 (lr=4.9301e-04) (hash(x)=26037988)
51
+ 47 train 8.615035 (lr=5.0350e-04) (hash(x)=24732833)
52
+ 48 train 8.663086 (lr=5.1399e-04) (hash(x)=25259526)
53
+ 49 train 8.594772 (lr=5.2448e-04) (hash(x)=23200230)
54
+ 50 val loss 8.5612
55
+ 50 val perplexity 5224.9429
56
+ 50 train 8.558077 (lr=5.3497e-04) (hash(x)=26721357)
57
+ 51 train 8.459423 (lr=5.4545e-04) (hash(x)=22694718)
58
+ 52 train 10.888686 (lr=5.5594e-04) (hash(x)=28066766)
59
+ 53 train 8.425588 (lr=5.6643e-04) (hash(x)=23125151)
60
+ 54 train 8.518469 (lr=5.7692e-04) (hash(x)=27193725)
61
+ 55 train 8.305635 (lr=5.8741e-04) (hash(x)=25129410)
62
+ 56 train 8.362761 (lr=5.9790e-04) (hash(x)=24263988)
63
+ 57 train 8.098498 (lr=6.0839e-04) (hash(x)=23059154)
64
+ 58 train 8.348338 (lr=6.1888e-04) (hash(x)=26063864)
65
+ 59 train 8.520139 (lr=6.2937e-04) (hash(x)=27858570)
66
+ 60 train 8.262812 (lr=6.3986e-04) (hash(x)=23874620)
67
+ 61 train 8.091093 (lr=6.5035e-04) (hash(x)=22402617)
68
+ 62 train 8.146502 (lr=6.6084e-04) (hash(x)=23600822)
69
+ 63 train 7.953904 (lr=6.7133e-04) (hash(x)=26582391)
70
+ 64 train 7.957869 (lr=6.8182e-04) (hash(x)=23225283)
71
+ 65 train 8.362988 (lr=6.9231e-04) (hash(x)=26075451)
72
+ 66 train 7.975824 (lr=7.0280e-04) (hash(x)=24723419)
73
+ 67 train 8.150385 (lr=7.1329e-04) (hash(x)=27279806)
74
+ 68 train 7.992424 (lr=7.2378e-04) (hash(x)=25870391)
75
+ 69 train 8.137382 (lr=7.3427e-04) (hash(x)=26188136)
76
+ 70 train 8.313684 (lr=7.4476e-04) (hash(x)=30373443)
77
+ 71 train 8.235140 (lr=7.5524e-04) (hash(x)=26472336)
78
+ 72 train 8.038657 (lr=7.6573e-04) (hash(x)=26651572)
79
+ 73 train 8.085325 (lr=7.7622e-04) (hash(x)=26376212)
80
+ 74 train 8.086051 (lr=7.8671e-04) (hash(x)=26733350)
81
+ 75 train 8.251344 (lr=7.9720e-04) (hash(x)=28301589)
82
+ 76 train 8.099861 (lr=8.0769e-04) (hash(x)=27599559)
83
+ 77 train 8.119344 (lr=8.1818e-04) (hash(x)=28035221)
84
+ 78 train 8.137954 (lr=8.2867e-04) (hash(x)=25016783)
85
+ 79 train 8.157538 (lr=8.3916e-04) (hash(x)=27654289)
86
+ 80 train 7.964523 (lr=8.4965e-04) (hash(x)=24597558)
87
+ 81 train 7.800640 (lr=8.6014e-04) (hash(x)=21560904)
88
+ 82 train 7.728832 (lr=8.7063e-04) (hash(x)=21983837)
89
+ 83 train 7.945615 (lr=8.8112e-04) (hash(x)=24995715)
90
+ 84 train 8.263102 (lr=8.9161e-04) (hash(x)=29876413)
91
+ 85 train 7.897574 (lr=9.0210e-04) (hash(x)=23792508)
92
+ 86 train 7.825984 (lr=9.1259e-04) (hash(x)=25509120)
93
+ 87 train 8.065989 (lr=9.2308e-04) (hash(x)=26559876)
94
+ 88 train 7.967267 (lr=9.3357e-04) (hash(x)=23569647)
95
+ 89 train 7.990154 (lr=9.4406e-04) (hash(x)=25758852)
96
+ 90 train 7.823780 (lr=9.5455e-04) (hash(x)=25706298)
97
+ 91 train 7.992212 (lr=9.6503e-04) (hash(x)=28364895)
98
+ 92 train 7.853757 (lr=9.7552e-04) (hash(x)=25304663)
99
+ 93 train 7.918347 (lr=9.8601e-04) (hash(x)=25269299)
100
+ 94 train 8.089424 (lr=9.9650e-04) (hash(x)=25870566)
101
+ 95 train 7.689047 (lr=1.0070e-03) (hash(x)=21770329)
102
+ 96 train 8.209849 (lr=1.0175e-03) (hash(x)=27595900)
103
+ 97 train 8.121750 (lr=1.0280e-03) (hash(x)=24785397)
104
+ 98 train 7.855667 (lr=1.0385e-03) (hash(x)=21521480)
105
+ 99 train 7.967109 (lr=1.0490e-03) (hash(x)=24628606)
106
+ 100 val loss 8.0063
107
+ 100 val perplexity 2999.6599
108
+ 100 train 8.093553 (lr=1.0594e-03) (hash(x)=24670150)
109
+ 101 train 7.988746 (lr=1.0699e-03) (hash(x)=23181910)
110
+ 102 train 7.887452 (lr=1.0804e-03) (hash(x)=22714991)
111
+ 103 train 7.823487 (lr=1.0909e-03) (hash(x)=22723459)
112
+ 104 train 7.802074 (lr=1.1014e-03) (hash(x)=21524316)
113
+ 105 train 8.009483 (lr=1.1119e-03) (hash(x)=25506632)
114
+ 106 train 7.677930 (lr=1.1224e-03) (hash(x)=21675672)
115
+ 107 train 7.788574 (lr=1.1329e-03) (hash(x)=22897919)
116
+ 108 train 7.803323 (lr=1.1434e-03) (hash(x)=23321631)
117
+ 109 train 8.006656 (lr=1.1538e-03) (hash(x)=26546719)
118
+ 110 train 9.120775 (lr=1.1643e-03) (hash(x)=31962348)
119
+ 111 train 8.497126 (lr=1.1748e-03) (hash(x)=30338342)
120
+ 112 train 8.788442 (lr=1.1853e-03) (hash(x)=23724471)
121
+ 113 train 8.285411 (lr=1.1958e-03) (hash(x)=29175888)
122
+ 114 train 8.139271 (lr=1.2063e-03) (hash(x)=23256716)
123
+ 115 train 8.268172 (lr=1.2168e-03) (hash(x)=27063280)
124
+ 116 train 8.261287 (lr=1.2273e-03) (hash(x)=31057659)
125
+ 117 train 8.408601 (lr=1.2378e-03) (hash(x)=32915097)
126
+ 118 train 8.235984 (lr=1.2483e-03) (hash(x)=28842717)
127
+ 119 train 7.907484 (lr=1.2587e-03) (hash(x)=25678059)
128
+ 120 train 7.775975 (lr=1.2692e-03) (hash(x)=21593510)
129
+ 121 train 7.690438 (lr=1.2797e-03) (hash(x)=20083773)
130
+ 122 train 7.921832 (lr=1.2902e-03) (hash(x)=23002820)
131
+ 123 train 7.694310 (lr=1.3007e-03) (hash(x)=21853028)
132
+ 124 train 8.000465 (lr=1.3112e-03) (hash(x)=26985625)
133
+ 125 train 7.745120 (lr=1.3217e-03) (hash(x)=21808483)
134
+ 126 train 8.217239 (lr=1.3322e-03) (hash(x)=28873251)
135
+ 127 train 8.079806 (lr=1.3427e-03) (hash(x)=26109335)
136
+ 128 train 8.246306 (lr=1.3531e-03) (hash(x)=26334674)
137
+ 129 train 8.030437 (lr=1.3636e-03) (hash(x)=24916754)
138
+ 130 train 8.003387 (lr=1.3741e-03) (hash(x)=25449624)
139
+ 131 train 8.045049 (lr=1.3846e-03) (hash(x)=25334848)
140
+ 132 train 8.387545 (lr=1.3951e-03) (hash(x)=27484863)
141
+ 133 train 7.850161 (lr=1.4056e-03) (hash(x)=24917705)
142
+ 134 train 7.884520 (lr=1.4161e-03) (hash(x)=25143449)
143
+ 135 train 7.918564 (lr=1.4266e-03) (hash(x)=25044885)
144
+ 136 train 8.276156 (lr=1.4371e-03) (hash(x)=27821028)
145
+ 137 train 8.175544 (lr=1.4476e-03) (hash(x)=28747022)
146
+ 138 train 8.108214 (lr=1.4580e-03) (hash(x)=27182888)
147
+ 139 train 7.898388 (lr=1.4685e-03) (hash(x)=23678349)
148
+ 140 train 8.020082 (lr=1.4790e-03) (hash(x)=23593235)
149
+ 141 train 8.307268 (lr=1.4895e-03) (hash(x)=28529813)
150
+ 142 train 8.655150 (lr=1.5000e-03) (hash(x)=32074661)
151
+ 143 train 8.184884 (lr=1.5105e-03) (hash(x)=28870690)
152
+ 144 train 8.180808 (lr=1.5210e-03) (hash(x)=27307705)
153
+ 145 train 8.060107 (lr=1.5315e-03) (hash(x)=25044834)
154
+ 146 train 8.080556 (lr=1.5420e-03) (hash(x)=23712023)
155
+ 147 train 8.523896 (lr=1.5524e-03) (hash(x)=32982615)
156
+ 148 train 8.319450 (lr=1.5629e-03) (hash(x)=30113660)
157
+ 149 train 7.729886 (lr=1.5734e-03) (hash(x)=20970960)
158
+ 150 val loss 7.9763
159
+ 150 val perplexity 2911.1392
160
+ 150 train 7.968419 (lr=1.5839e-03) (hash(x)=23132684)
161
+ 151 train 9.831520 (lr=1.5944e-03) (hash(x)=35279941)
162
+ 152 train 8.771175 (lr=1.6049e-03) (hash(x)=31227444)
163
+ 153 train 8.117344 (lr=1.6154e-03) (hash(x)=25529472)
164
+ 154 train 8.256622 (lr=1.6259e-03) (hash(x)=24350409)
165
+ 155 train 8.407577 (lr=1.6364e-03) (hash(x)=26400041)
166
+ 156 train 7.959975 (lr=1.6469e-03) (hash(x)=25262621)
167
+ 157 train 7.901357 (lr=1.6573e-03) (hash(x)=24656138)
168
+ 158 train 8.060475 (lr=1.6678e-03) (hash(x)=26803414)
169
+ 159 train 7.781722 (lr=1.6783e-03) (hash(x)=25015923)
170
+ 160 train 7.629598 (lr=1.6888e-03) (hash(x)=23581172)
171
+ 161 train 7.689438 (lr=1.6993e-03) (hash(x)=22924885)
172
+ 162 train 7.843590 (lr=1.7098e-03) (hash(x)=23414296)
173
+ 163 train 8.325418 (lr=1.7203e-03) (hash(x)=24853586)
174
+ 164 train 8.277551 (lr=1.7308e-03) (hash(x)=25000130)
175
+ 165 train 8.070045 (lr=1.7413e-03) (hash(x)=27004780)
176
+ 166 train 8.061103 (lr=1.7517e-03) (hash(x)=26148573)
177
+ 167 train 7.960260 (lr=1.7622e-03) (hash(x)=26740855)
178
+ 168 train 7.533565 (lr=1.7727e-03) (hash(x)=20965419)
179
+ 169 train 7.811303 (lr=1.7832e-03) (hash(x)=23950114)
180
+ 170 train 7.790592 (lr=1.7937e-03) (hash(x)=24951982)
181
+ 171 train 7.968775 (lr=1.8042e-03) (hash(x)=24584116)
182
+ 172 train 7.931394 (lr=1.8147e-03) (hash(x)=24378759)
183
+ 173 train 7.974961 (lr=1.8252e-03) (hash(x)=25718516)
184
+ 174 train 8.108790 (lr=1.8357e-03) (hash(x)=28424396)
185
+ 175 train 7.859569 (lr=1.8462e-03) (hash(x)=22262151)
186
+ 176 train 8.128601 (lr=1.8566e-03) (hash(x)=26438412)
187
+ 177 train 7.802660 (lr=1.8671e-03) (hash(x)=23025303)
188
+ 178 train 7.856927 (lr=1.8776e-03) (hash(x)=24190770)
189
+ 179 train 7.986577 (lr=1.8881e-03) (hash(x)=26627860)
190
+ 180 train 7.811626 (lr=1.8986e-03) (hash(x)=23663439)
191
+ 181 train 7.665682 (lr=1.9091e-03) (hash(x)=23700532)
192
+ 182 train 7.677462 (lr=1.9196e-03) (hash(x)=23075676)
193
+ 183 train 7.867935 (lr=1.9301e-03) (hash(x)=26621834)
194
+ 184 train 7.934873 (lr=1.9406e-03) (hash(x)=29426269)
195
+ 185 train 7.678044 (lr=1.9510e-03) (hash(x)=21821465)
196
+ 186 train 7.962233 (lr=1.9615e-03) (hash(x)=26506130)
197
+ 187 train 8.299834 (lr=1.9720e-03) (hash(x)=26930630)
198
+ 188 train 7.833045 (lr=1.9825e-03) (hash(x)=22993793)
199
+ 189 train 7.705350 (lr=1.9930e-03) (hash(x)=19557946)
200
+ 190 train 7.957364 (lr=2.0035e-03) (hash(x)=23572891)
201
+ 191 train 7.718530 (lr=2.0140e-03) (hash(x)=23234741)
202
+ 192 train 7.859476 (lr=2.0245e-03) (hash(x)=25547951)
203
+ 193 train 7.889356 (lr=2.0350e-03) (hash(x)=26713563)
204
+ 194 train 7.907604 (lr=2.0455e-03) (hash(x)=25913622)
205
+ 195 train 8.211601 (lr=2.0559e-03) (hash(x)=28102443)
206
+ 196 train 7.862226 (lr=2.0664e-03) (hash(x)=23093351)
207
+ 197 train 8.593924 (lr=2.0769e-03) (hash(x)=31689122)
208
+ 198 train 8.329302 (lr=2.0874e-03) (hash(x)=32252517)
209
+ 199 train 8.152904 (lr=2.0979e-03) (hash(x)=25470563)
210
+ 200 val loss 8.1037
211
+ 200 val perplexity 3306.7800
212
+ 200 train 8.183162 (lr=2.1084e-03) (hash(x)=25597614)
213
+ 201 train 8.070544 (lr=2.1189e-03) (hash(x)=23757479)
214
+ 202 train 7.953777 (lr=2.1294e-03) (hash(x)=25140048)
215
+ 203 train 8.158601 (lr=2.1399e-03) (hash(x)=28282861)
216
+ 204 train 7.871141 (lr=2.1503e-03) (hash(x)=24754885)
217
+ 205 train 8.149048 (lr=2.1608e-03) (hash(x)=26731964)
218
+ 206 train 8.189460 (lr=2.1713e-03) (hash(x)=26660561)
219
+ 207 train 7.778045 (lr=2.1818e-03) (hash(x)=21799102)
220
+ 208 train 7.984417 (lr=2.1923e-03) (hash(x)=27025986)
221
+ 209 train 7.833181 (lr=2.2028e-03) (hash(x)=24672077)
222
+ 210 train 7.914753 (lr=2.2133e-03) (hash(x)=25322984)
223
+ 211 train 7.727831 (lr=2.2238e-03) (hash(x)=23471769)
224
+ 212 train 7.516001 (lr=2.2343e-03) (hash(x)=20766491)
225
+ 213 train 8.050658 (lr=2.2448e-03) (hash(x)=24058931)
226
+ 214 train 7.705248 (lr=2.2552e-03) (hash(x)=23872843)
227
+ 215 train 7.821485 (lr=2.2657e-03) (hash(x)=23275479)
228
+ 216 train 7.951135 (lr=2.2762e-03) (hash(x)=24914695)
229
+ 217 train 7.777202 (lr=2.2867e-03) (hash(x)=24364396)
230
+ 218 train 8.099063 (lr=2.2972e-03) (hash(x)=27986474)
231
+ 219 train 7.976544 (lr=2.3077e-03) (hash(x)=24473581)
232
+ 220 train 7.852455 (lr=2.3182e-03) (hash(x)=22974689)
233
+ 221 train 7.961548 (lr=2.3287e-03) (hash(x)=23774644)
234
+ 222 train 7.856726 (lr=2.3392e-03) (hash(x)=23245327)
235
+ 223 train 7.666238 (lr=2.3497e-03) (hash(x)=22091862)
236
+ 224 train 7.762062 (lr=2.3601e-03) (hash(x)=24362839)
237
+ 225 train 7.934511 (lr=2.3706e-03) (hash(x)=25482303)
238
+ 226 train 8.056933 (lr=2.3811e-03) (hash(x)=24911853)
239
+ 227 train 7.983467 (lr=2.3916e-03) (hash(x)=26018202)
240
+ 228 train 7.867131 (lr=2.4021e-03) (hash(x)=26124495)
241
+ 229 train 7.965982 (lr=2.4126e-03) (hash(x)=24560096)
242
+ 230 train 7.890714 (lr=2.4231e-03) (hash(x)=24695331)
243
+ 231 train 7.522696 (lr=2.4336e-03) (hash(x)=17430373)
244
+ 232 train 7.628623 (lr=2.4441e-03) (hash(x)=21813345)
245
+ 233 train 7.250780 (lr=2.4545e-03) (hash(x)=20098681)
246
+ 234 train 7.630932 (lr=2.4650e-03) (hash(x)=25095928)
247
+ 235 train 7.781775 (lr=2.4755e-03) (hash(x)=24078083)
248
+ 236 train 7.817749 (lr=2.4860e-03) (hash(x)=22901505)
249
+ 237 train 8.005433 (lr=2.4965e-03) (hash(x)=26595592)
250
+ 238 train 7.886547 (lr=2.5070e-03) (hash(x)=27663196)
251
+ 239 train 7.650795 (lr=2.5175e-03) (hash(x)=22954861)
252
+ 240 train 8.204656 (lr=2.5280e-03) (hash(x)=30159234)
253
+ 241 train 8.064758 (lr=2.5385e-03) (hash(x)=29294271)
254
+ 242 train 8.074604 (lr=2.5490e-03) (hash(x)=23728322)
255
+ 243 train 8.068228 (lr=2.5594e-03) (hash(x)=28695016)
256
+ 244 train 8.203610 (lr=2.5699e-03) (hash(x)=26702728)
257
+ 245 train 7.969274 (lr=2.5804e-03) (hash(x)=27100115)
258
+ 246 train 7.847144 (lr=2.5909e-03) (hash(x)=25082752)
259
+ 247 train 8.178973 (lr=2.6014e-03) (hash(x)=26671799)
260
+ 248 train 7.892236 (lr=2.6119e-03) (hash(x)=23718946)
261
+ 249 train 8.120234 (lr=2.6224e-03) (hash(x)=28137394)
262
+ 250 val loss 7.8537
263
+ 250 val perplexity 2575.2039
264
+ 250 train 8.162617 (lr=2.6329e-03) (hash(x)=23893495)
265
+ 251 train 7.918488 (lr=2.6434e-03) (hash(x)=23166092)
266
+ 252 train 7.917640 (lr=2.6538e-03) (hash(x)=25907665)
267
+ 253 train 7.904659 (lr=2.6643e-03) (hash(x)=25885986)
268
+ 254 train 7.930789 (lr=2.6748e-03) (hash(x)=25262712)
269
+ 255 train 7.920400 (lr=2.6853e-03) (hash(x)=26924723)
270
+ 256 train 8.144126 (lr=2.6958e-03) (hash(x)=28744736)
271
+ 257 train 7.800294 (lr=2.7063e-03) (hash(x)=26140590)
272
+ 258 train 8.013428 (lr=2.7168e-03) (hash(x)=25780449)
273
+ 259 train 7.819873 (lr=2.7273e-03) (hash(x)=25138659)
274
+ 260 train 8.149627 (lr=2.7378e-03) (hash(x)=27244046)
275
+ 261 train 7.979984 (lr=2.7483e-03) (hash(x)=27224685)
276
+ 262 train 8.293767 (lr=2.7587e-03) (hash(x)=28274477)
277
+ 263 train 7.795988 (lr=2.7692e-03) (hash(x)=23557495)
278
+ 264 train 8.024003 (lr=2.7797e-03) (hash(x)=24680596)
279
+ 265 train 8.132712 (lr=2.7902e-03) (hash(x)=23928957)
280
+ 266 train 8.008063 (lr=2.8007e-03) (hash(x)=23761390)
281
+ 267 train 7.810571 (lr=2.8112e-03) (hash(x)=25288123)
282
+ 268 train 8.154808 (lr=2.8217e-03) (hash(x)=28705502)
283
+ 269 train 7.683310 (lr=2.8322e-03) (hash(x)=23246294)
284
+ 270 train 7.916951 (lr=2.8427e-03) (hash(x)=28639079)
285
+ 271 train 7.809988 (lr=2.8531e-03) (hash(x)=27804380)
286
+ 272 train 7.930638 (lr=2.8636e-03) (hash(x)=24172235)
287
+ 273 train 7.968646 (lr=2.8741e-03) (hash(x)=23089140)
288
+ 274 train 8.236835 (lr=2.8846e-03) (hash(x)=27163701)
289
+ 275 train 8.292235 (lr=2.8951e-03) (hash(x)=26993263)
290
+ 276 train 7.992179 (lr=2.9056e-03) (hash(x)=28224233)
291
+ 277 train 8.266095 (lr=2.9161e-03) (hash(x)=27397203)
292
+ 278 train 8.237755 (lr=2.9266e-03) (hash(x)=27923882)
293
+ 279 train 8.104241 (lr=2.9371e-03) (hash(x)=26654908)
294
+ 280 train 8.136360 (lr=2.9476e-03) (hash(x)=24213147)
295
+ 281 train 7.858224 (lr=2.9580e-03) (hash(x)=21965022)
296
+ 282 train 7.838675 (lr=2.9685e-03) (hash(x)=25465685)
297
+ 283 train 8.078303 (lr=2.9790e-03) (hash(x)=27347722)
298
+ 284 train 8.113561 (lr=2.9895e-03) (hash(x)=26732050)
299
+ 285 train 8.279761 (lr=3.0000e-03) (hash(x)=28314127)
300
+ 286 train 7.710796 (lr=3.0000e-03) (hash(x)=21471186)
301
+ 287 train 7.616193 (lr=3.0000e-03) (hash(x)=23627518)
302
+ 288 train 7.684252 (lr=2.9999e-03) (hash(x)=20870353)
303
+ 289 train 7.822554 (lr=2.9999e-03) (hash(x)=25024764)
304
+ 290 train 7.623837 (lr=2.9998e-03) (hash(x)=20683822)
305
+ 291 train 7.912674 (lr=2.9997e-03) (hash(x)=21768671)
306
+ 292 train 7.972833 (lr=2.9995e-03) (hash(x)=25557309)
307
+ 293 train 8.091243 (lr=2.9994e-03) (hash(x)=25076667)
308
+ 294 train 7.709581 (lr=2.9992e-03) (hash(x)=23765822)
309
+ 295 train 7.720580 (lr=2.9989e-03) (hash(x)=21889990)
310
+ 296 train 7.964566 (lr=2.9987e-03) (hash(x)=26339893)
311
+ 297 train 7.613224 (lr=2.9984e-03) (hash(x)=20932794)
312
+ 298 train 7.802375 (lr=2.9981e-03) (hash(x)=21750070)
313
+ 299 train 7.966420 (lr=2.9978e-03) (hash(x)=23665838)
314
+ 300 val loss 7.9154
315
+ 300 val perplexity 2739.1199
316
+ 300 train 8.801383 (lr=2.9974e-03) (hash(x)=32888061)
317
+ 301 train 8.444455 (lr=2.9971e-03) (hash(x)=30223582)
318
+ 302 train 8.115050 (lr=2.9967e-03) (hash(x)=26908418)
319
+ 303 train 7.596101 (lr=2.9962e-03) (hash(x)=22528001)
320
+ 304 train 8.183784 (lr=2.9958e-03) (hash(x)=27452187)
321
+ 305 train 7.977899 (lr=2.9953e-03) (hash(x)=25181641)
322
+ 306 train 8.108852 (lr=2.9948e-03) (hash(x)=25546593)
323
+ 307 train 8.036695 (lr=2.9942e-03) (hash(x)=22487328)
324
+ 308 train 8.166902 (lr=2.9937e-03) (hash(x)=27804274)
325
+ 309 train 8.195557 (lr=2.9931e-03) (hash(x)=26544630)
326
+ 310 train 8.308525 (lr=2.9925e-03) (hash(x)=27738934)
327
+ 311 train 8.107521 (lr=2.9918e-03) (hash(x)=29248942)
328
+ 312 train 7.888223 (lr=2.9912e-03) (hash(x)=25103452)
329
+ 313 train 8.072104 (lr=2.9905e-03) (hash(x)=25052066)
330
+ 314 train 7.989933 (lr=2.9898e-03) (hash(x)=24481302)
331
+ 315 train 7.927261 (lr=2.9890e-03) (hash(x)=23543273)
332
+ 316 train 7.873052 (lr=2.9883e-03) (hash(x)=25608244)
333
+ 317 train 8.070871 (lr=2.9875e-03) (hash(x)=27451288)
334
+ 318 train 7.691783 (lr=2.9866e-03) (hash(x)=22806491)
335
+ 319 train 7.880995 (lr=2.9858e-03) (hash(x)=25533417)
336
+ 320 train 7.893154 (lr=2.9849e-03) (hash(x)=24557997)
337
+ 321 train 7.883869 (lr=2.9840e-03) (hash(x)=24432899)
338
+ 322 train 8.094897 (lr=2.9831e-03) (hash(x)=27583287)
339
+ 323 train 7.981544 (lr=2.9821e-03) (hash(x)=25552036)
340
+ 324 train 7.780636 (lr=2.9812e-03) (hash(x)=24201868)
341
+ 325 train 8.108803 (lr=2.9802e-03) (hash(x)=28149782)
342
+ 326 train 7.915143 (lr=2.9791e-03) (hash(x)=25529698)
343
+ 327 train 7.690273 (lr=2.9781e-03) (hash(x)=20612533)
344
+ 328 train 7.681419 (lr=2.9770e-03) (hash(x)=20699000)
345
+ 329 train 7.571413 (lr=2.9759e-03) (hash(x)=19774173)
346
+ 330 train 7.536576 (lr=2.9748e-03) (hash(x)=21681646)
347
+ 331 train 7.328868 (lr=2.9736e-03) (hash(x)=20216795)
348
+ 332 train 8.203643 (lr=2.9724e-03) (hash(x)=27697998)
349
+ 333 train 7.940414 (lr=2.9712e-03) (hash(x)=25896435)
350
+ 334 train 7.816444 (lr=2.9700e-03) (hash(x)=21585310)
351
+ 335 train 7.903997 (lr=2.9687e-03) (hash(x)=24677740)
352
+ 336 train 7.704423 (lr=2.9675e-03) (hash(x)=22027900)
353
+ 337 train 7.753099 (lr=2.9662e-03) (hash(x)=21835643)
354
+ 338 train 7.660760 (lr=2.9648e-03) (hash(x)=20558462)
355
+ 339 train 7.559709 (lr=2.9635e-03) (hash(x)=16722715)
356
+ 340 train 7.732024 (lr=2.9621e-03) (hash(x)=20972655)
357
+ 341 train 8.240714 (lr=2.9607e-03) (hash(x)=26303975)
358
+ 342 train 7.994629 (lr=2.9592e-03) (hash(x)=22938170)
359
+ 343 train 7.986473 (lr=2.9578e-03) (hash(x)=25347203)
360
+ 344 train 7.988379 (lr=2.9563e-03) (hash(x)=27398686)
361
+ 345 train 7.912147 (lr=2.9548e-03) (hash(x)=25973417)
362
+ 346 train 7.894852 (lr=2.9532e-03) (hash(x)=26918389)
363
+ 347 train 7.912810 (lr=2.9517e-03) (hash(x)=24272489)
364
+ 348 train 7.920650 (lr=2.9501e-03) (hash(x)=25593714)
365
+ 349 train 8.086181 (lr=2.9485e-03) (hash(x)=29260846)
366
+ 350 val loss 8.0128
367
+ 350 val perplexity 3019.3313
368
+ 350 train 8.225280 (lr=2.9468e-03) (hash(x)=27951602)
369
+ 351 train 8.318820 (lr=2.9452e-03) (hash(x)=28922363)
370
+ 352 train 8.160232 (lr=2.9435e-03) (hash(x)=27210734)
371
+ 353 train 8.148188 (lr=2.9418e-03) (hash(x)=26322572)
372
+ 354 train 8.176604 (lr=2.9400e-03) (hash(x)=27084665)
373
+ 355 train 8.023376 (lr=2.9383e-03) (hash(x)=25510798)
374
+ 356 train 8.113296 (lr=2.9365e-03) (hash(x)=24970921)
375
+ 357 train 8.113414 (lr=2.9347e-03) (hash(x)=24138948)
376
+ 358 train 8.075618 (lr=2.9328e-03) (hash(x)=24790211)
377
+ 359 train 8.073248 (lr=2.9310e-03) (hash(x)=25631397)
378
+ 360 train 7.938953 (lr=2.9291e-03) (hash(x)=23226625)
379
+ 361 train 7.888587 (lr=2.9272e-03) (hash(x)=24001903)
380
+ 362 train 7.967094 (lr=2.9252e-03) (hash(x)=24587948)
381
+ 363 train 7.580091 (lr=2.9233e-03) (hash(x)=21333676)
382
+ 364 train 7.508811 (lr=2.9213e-03) (hash(x)=23673779)
383
+ 365 train 8.827117 (lr=2.9193e-03) (hash(x)=30770484)
384
+ 366 train 8.032710 (lr=2.9172e-03) (hash(x)=26564899)
385
+ 367 train 7.973749 (lr=2.9152e-03) (hash(x)=26237983)
386
+ 368 train 7.745927 (lr=2.9131e-03) (hash(x)=23764356)
387
+ 369 train 8.056606 (lr=2.9110e-03) (hash(x)=26205744)
388
+ 370 train 7.697687 (lr=2.9088e-03) (hash(x)=19208770)
389
+ 371 train 7.964690 (lr=2.9067e-03) (hash(x)=25976502)
390
+ 372 train 7.816530 (lr=2.9045e-03) (hash(x)=23983933)
391
+ 373 train 7.781833 (lr=2.9023e-03) (hash(x)=24080636)
392
+ 374 train 7.873661 (lr=2.9001e-03) (hash(x)=24404047)
393
+ 375 train 7.808597 (lr=2.8978e-03) (hash(x)=24742645)
394
+ 376 train 7.877688 (lr=2.8955e-03) (hash(x)=24159600)
395
+ 377 train 8.084773 (lr=2.8932e-03) (hash(x)=28677257)
396
+ 378 train 7.853652 (lr=2.8909e-03) (hash(x)=25604111)
397
+ 379 train 7.790505 (lr=2.8885e-03) (hash(x)=27086333)
398
+ 380 train 7.834577 (lr=2.8862e-03) (hash(x)=25188207)
399
+ 381 train 8.050297 (lr=2.8838e-03) (hash(x)=27855233)
400
+ 382 train 7.584383 (lr=2.8813e-03) (hash(x)=19470039)
401
+ 383 train 7.939146 (lr=2.8789e-03) (hash(x)=26157660)
402
+ 384 train 7.929790 (lr=2.8764e-03) (hash(x)=25291570)
403
+ 385 train 7.872560 (lr=2.8739e-03) (hash(x)=25046062)
404
+ 386 train 8.130644 (lr=2.8714e-03) (hash(x)=27020337)
405
+ 387 train 7.780260 (lr=2.8689e-03) (hash(x)=23616370)
406
+ 388 train 7.503458 (lr=2.8663e-03) (hash(x)=19113218)
407
+ 389 train 8.143687 (lr=2.8637e-03) (hash(x)=24302232)
408
+ 390 train 8.439845 (lr=2.8611e-03) (hash(x)=22188949)
409
+ 391 train 8.442789 (lr=2.8585e-03) (hash(x)=22582169)
410
+ 392 train 8.902046 (lr=2.8558e-03) (hash(x)=24700570)
411
+ 393 train 8.781085 (lr=2.8531e-03) (hash(x)=22773833)
412
+ 394 train 8.508998 (lr=2.8504e-03) (hash(x)=21875928)
413
+ 395 train 8.833369 (lr=2.8477e-03) (hash(x)=26233189)
414
+ 396 train 8.635261 (lr=2.8449e-03) (hash(x)=24321467)
415
+ 397 train 8.875238 (lr=2.8422e-03) (hash(x)=26431507)
416
+ 398 train 8.937438 (lr=2.8394e-03) (hash(x)=28690877)
417
+ 399 train 8.716694 (lr=2.8365e-03) (hash(x)=26431960)
418
+ 400 val loss 8.4720
419
+ 400 val perplexity 4778.8369
420
+ 400 train 8.350560 (lr=2.8337e-03) (hash(x)=24580300)
421
+ 401 train 8.389447 (lr=2.8308e-03) (hash(x)=25112360)
422
+ 402 train 8.352641 (lr=2.8279e-03) (hash(x)=27597243)
423
+ 403 train 8.932751 (lr=2.8250e-03) (hash(x)=30707498)
424
+ 404 train 8.480392 (lr=2.8221e-03) (hash(x)=28485465)
425
+ 405 train 8.021009 (lr=2.8191e-03) (hash(x)=22586447)
426
+ 406 train 8.085316 (lr=2.8162e-03) (hash(x)=23175270)
427
+ 407 train 8.249582 (lr=2.8131e-03) (hash(x)=25716176)
428
+ 408 train 8.379375 (lr=2.8101e-03) (hash(x)=26861373)
429
+ 409 train 8.404407 (lr=2.8071e-03) (hash(x)=25118971)
430
+ 410 train 7.775939 (lr=2.8040e-03) (hash(x)=19829066)
431
+ 411 train 8.013844 (lr=2.8009e-03) (hash(x)=26256420)
432
+ 412 train 8.196448 (lr=2.7978e-03) (hash(x)=27796153)
433
+ 413 train 7.869557 (lr=2.7947e-03) (hash(x)=22633318)
434
+ 414 train 7.927460 (lr=2.7915e-03) (hash(x)=22589383)
435
+ 415 train 8.396065 (lr=2.7883e-03) (hash(x)=28019788)
436
+ 416 train 8.259700 (lr=2.7851e-03) (hash(x)=28970440)
437
+ 417 train 7.923192 (lr=2.7819e-03) (hash(x)=27396089)
438
+ 418 train 7.903057 (lr=2.7786e-03) (hash(x)=21183513)
439
+ 419 train 7.923952 (lr=2.7754e-03) (hash(x)=23510110)
440
+ 420 train 8.177234 (lr=2.7721e-03) (hash(x)=28833467)
441
+ 421 train 7.970465 (lr=2.7688e-03) (hash(x)=23646926)
442
+ 422 train 7.985233 (lr=2.7654e-03) (hash(x)=24697272)
443
+ 423 train 7.684900 (lr=2.7621e-03) (hash(x)=20382963)
444
+ 424 train 7.741651 (lr=2.7587e-03) (hash(x)=23467595)
445
+ 425 train 7.925645 (lr=2.7553e-03) (hash(x)=24304768)
446
+ 426 train 7.818989 (lr=2.7519e-03) (hash(x)=21392328)
447
+ 427 train 7.892463 (lr=2.7484e-03) (hash(x)=25339466)
448
+ 428 train 7.777961 (lr=2.7450e-03) (hash(x)=22092542)
449
+ 429 train 7.874824 (lr=2.7415e-03) (hash(x)=22088696)
450
+ 430 train 7.818005 (lr=2.7380e-03) (hash(x)=22184471)
451
+ 431 train 7.904382 (lr=2.7344e-03) (hash(x)=24489647)
452
+ 432 train 7.933121 (lr=2.7309e-03) (hash(x)=26794132)
453
+ 433 train 7.742414 (lr=2.7273e-03) (hash(x)=22940357)
454
+ 434 train 7.916858 (lr=2.7237e-03) (hash(x)=23719522)
455
+ 435 train 7.743376 (lr=2.7201e-03) (hash(x)=22927699)
456
+ 436 train 7.945408 (lr=2.7165e-03) (hash(x)=26068576)
457
+ 437 train 8.008735 (lr=2.7128e-03) (hash(x)=27631132)
458
+ 438 train 7.924855 (lr=2.7092e-03) (hash(x)=26739991)
459
+ 439 train 7.962318 (lr=2.7055e-03) (hash(x)=25128502)
460
+ 440 train 8.051358 (lr=2.7018e-03) (hash(x)=25657260)
461
+ 441 train 7.814113 (lr=2.6980e-03) (hash(x)=23576982)
462
+ 442 train 8.024734 (lr=2.6943e-03) (hash(x)=27117886)
463
+ 443 train 7.960138 (lr=2.6905e-03) (hash(x)=25808969)
464
+ 444 train 7.871954 (lr=2.6867e-03) (hash(x)=24738238)
465
+ 445 train 7.814191 (lr=2.6829e-03) (hash(x)=23429962)
466
+ 446 train 7.968754 (lr=2.6791e-03) (hash(x)=25075165)
467
+ 447 train 7.798763 (lr=2.6752e-03) (hash(x)=25231390)
468
+ 448 train 7.635175 (lr=2.6713e-03) (hash(x)=22055054)
469
+ 449 train 7.838171 (lr=2.6674e-03) (hash(x)=25395441)
470
+ 450 val loss 7.8546
471
+ 450 val perplexity 2577.5034
472
+ 450 train 7.871149 (lr=2.6635e-03) (hash(x)=25863209)
473
+ 451 train 7.584259 (lr=2.6596e-03) (hash(x)=21154388)
474
+ 452 train 7.584634 (lr=2.6556e-03) (hash(x)=21600876)
475
+ 453 train 7.731755 (lr=2.6517e-03) (hash(x)=24278611)
476
+ 454 train 7.630913 (lr=2.6477e-03) (hash(x)=23221720)
477
+ 455 train 7.805109 (lr=2.6437e-03) (hash(x)=22708977)
478
+ 456 train 8.433214 (lr=2.6396e-03) (hash(x)=23637758)
479
+ 457 train 8.129076 (lr=2.6356e-03) (hash(x)=28228490)
480
+ 458 train 8.207619 (lr=2.6315e-03) (hash(x)=28638071)
481
+ 459 train 8.037328 (lr=2.6274e-03) (hash(x)=27258353)
482
+ 460 train 7.980030 (lr=2.6233e-03) (hash(x)=26604728)
483
+ 461 train 7.931153 (lr=2.6192e-03) (hash(x)=23252199)
484
+ 462 train 8.050430 (lr=2.6150e-03) (hash(x)=26441427)
485
+ 463 train 7.783237 (lr=2.6109e-03) (hash(x)=24364920)
486
+ 464 train 7.901041 (lr=2.6067e-03) (hash(x)=25623792)
487
+ 465 train 7.769861 (lr=2.6025e-03) (hash(x)=23283905)
488
+ 466 train 8.197944 (lr=2.5983e-03) (hash(x)=26025267)
489
+ 467 train 8.006180 (lr=2.5940e-03) (hash(x)=27243972)
490
+ 468 train 8.341283 (lr=2.5898e-03) (hash(x)=30449945)
491
+ 469 train 8.037642 (lr=2.5855e-03) (hash(x)=28113043)
492
+ 470 train 7.862119 (lr=2.5812e-03) (hash(x)=25182521)
493
+ 471 train 7.919229 (lr=2.5769e-03) (hash(x)=24932925)
494
+ 472 train 7.590020 (lr=2.5726e-03) (hash(x)=20353098)
495
+ 473 train 7.570726 (lr=2.5682e-03) (hash(x)=19001259)
496
+ 474 train 8.176160 (lr=2.5639e-03) (hash(x)=27585685)
497
+ 475 train 7.873738 (lr=2.5595e-03) (hash(x)=26371091)
498
+ 476 train 7.753615 (lr=2.5551e-03) (hash(x)=24891798)
499
+ 477 train 7.763587 (lr=2.5507e-03) (hash(x)=24258817)
500
+ 478 train 7.855278 (lr=2.5462e-03) (hash(x)=24330263)
501
+ 479 train 7.935857 (lr=2.5418e-03) (hash(x)=26913684)
502
+ 480 train 7.897950 (lr=2.5373e-03) (hash(x)=26338455)
503
+ 481 train 8.133029 (lr=2.5328e-03) (hash(x)=27753043)
504
+ 482 train 7.914570 (lr=2.5283e-03) (hash(x)=26123289)
505
+ 483 train 8.044521 (lr=2.5238e-03) (hash(x)=29239611)
506
+ 484 train 7.953783 (lr=2.5193e-03) (hash(x)=26553003)
507
+ 485 train 7.817093 (lr=2.5147e-03) (hash(x)=22984557)
508
+ 486 train 7.299364 (lr=2.5102e-03) (hash(x)=16947491)
509
+ 487 train 7.316417 (lr=2.5056e-03) (hash(x)=18017792)
510
+ 488 train 7.516350 (lr=2.5010e-03) (hash(x)=19918608)
511
+ 489 train 8.038016 (lr=2.4963e-03) (hash(x)=23374526)
512
+ 490 train 7.852394 (lr=2.4917e-03) (hash(x)=25009505)
513
+ 491 train 7.893287 (lr=2.4871e-03) (hash(x)=27574089)
514
+ 492 train 8.019051 (lr=2.4824e-03) (hash(x)=24122664)
515
+ 493 train 7.867798 (lr=2.4777e-03) (hash(x)=26154906)
516
+ 494 train 7.803337 (lr=2.4730e-03) (hash(x)=25192767)
517
+ 495 train 7.975006 (lr=2.4683e-03) (hash(x)=28613882)
518
+ 496 train 7.868186 (lr=2.4636e-03) (hash(x)=23547219)
519
+ 497 train 7.858337 (lr=2.4588e-03) (hash(x)=25272182)
520
+ 498 train 7.746517 (lr=2.4540e-03) (hash(x)=24992761)
521
+ 499 train 8.013553 (lr=2.4493e-03) (hash(x)=26981914)
522
+ 500 val loss 7.8500
523
+ 500 val perplexity 2565.7830
524
+ 500 train 7.775487 (lr=2.4445e-03) (hash(x)=22051933)
525
+ 501 train 7.929214 (lr=2.4397e-03) (hash(x)=24232348)
526
+ 502 train 7.886884 (lr=2.4348e-03) (hash(x)=23158331)
527
+ 503 train 7.715297 (lr=2.4300e-03) (hash(x)=22652243)
528
+ 504 train 7.783985 (lr=2.4251e-03) (hash(x)=23805602)
529
+ 505 train 8.090744 (lr=2.4203e-03) (hash(x)=25411991)
530
+ 506 train 7.302019 (lr=2.4154e-03) (hash(x)=18827215)
531
+ 507 train 7.032907 (lr=2.4105e-03) (hash(x)=15446025)
532
+ 508 train 7.426906 (lr=2.4056e-03) (hash(x)=20516263)
533
+ 509 train 7.939737 (lr=2.4006e-03) (hash(x)=27846176)
534
+ 510 train 7.686200 (lr=2.3957e-03) (hash(x)=23342449)
535
+ 511 train 7.925257 (lr=2.3907e-03) (hash(x)=27194521)
536
+ 512 train 7.723377 (lr=2.3858e-03) (hash(x)=23008284)
537
+ 513 train 8.207184 (lr=2.3808e-03) (hash(x)=29430001)
538
+ 514 train 7.538835 (lr=2.3758e-03) (hash(x)=22579319)
539
+ 515 train 7.585490 (lr=2.3708e-03) (hash(x)=25264518)
540
+ 516 train 7.911718 (lr=2.3657e-03) (hash(x)=25359075)
541
+ 517 train 7.855951 (lr=2.3607e-03) (hash(x)=25568956)
542
+ 518 train 8.159354 (lr=2.3556e-03) (hash(x)=32004108)
543
+ 519 train 7.846079 (lr=2.3506e-03) (hash(x)=24936836)
544
+ 520 train 7.952471 (lr=2.3455e-03) (hash(x)=27263338)
545
+ 521 train 8.000583 (lr=2.3404e-03) (hash(x)=27452099)
546
+ 522 train 7.863687 (lr=2.3353e-03) (hash(x)=25965406)
547
+ 523 train 8.062163 (lr=2.3301e-03) (hash(x)=28197282)
548
+ 524 train 7.757804 (lr=2.3250e-03) (hash(x)=22466209)
549
+ 525 train 7.833768 (lr=2.3198e-03) (hash(x)=22931889)
550
+ 526 train 7.950346 (lr=2.3147e-03) (hash(x)=26903920)
551
+ 527 train 7.855747 (lr=2.3095e-03) (hash(x)=24765578)
552
+ 528 train 7.989187 (lr=2.3043e-03) (hash(x)=27811359)
553
+ 529 train 7.713878 (lr=2.2991e-03) (hash(x)=25078649)
554
+ 530 train 7.929274 (lr=2.2939e-03) (hash(x)=25572416)
555
+ 531 train 7.965531 (lr=2.2887e-03) (hash(x)=27448185)
556
+ 532 train 8.080774 (lr=2.2834e-03) (hash(x)=25923719)
557
+ 533 train 7.851192 (lr=2.2782e-03) (hash(x)=24804856)
558
+ 534 train 8.000152 (lr=2.2729e-03) (hash(x)=23207829)
559
+ 535 train 7.905900 (lr=2.2676e-03) (hash(x)=23107416)
560
+ 536 train 7.779315 (lr=2.2624e-03) (hash(x)=26739531)
561
+ 537 train 7.746040 (lr=2.2571e-03) (hash(x)=24960796)
562
+ 538 train 7.870990 (lr=2.2517e-03) (hash(x)=24667802)
563
+ 539 train 8.014147 (lr=2.2464e-03) (hash(x)=26755138)
564
+ 540 train 7.878678 (lr=2.2411e-03) (hash(x)=25537132)
565
+ 541 train 7.781481 (lr=2.2357e-03) (hash(x)=24542526)
566
+ 542 train 7.583717 (lr=2.2304e-03) (hash(x)=21296355)
567
+ 543 train 8.103040 (lr=2.2250e-03) (hash(x)=29314255)
568
+ 544 train 7.863697 (lr=2.2196e-03) (hash(x)=26001799)
569
+ 545 train 7.916538 (lr=2.2142e-03) (hash(x)=27347755)
570
+ 546 train 7.864781 (lr=2.2088e-03) (hash(x)=25107798)
571
+ 547 train 7.687782 (lr=2.2034e-03) (hash(x)=22112669)
572
+ 548 train 7.646765 (lr=2.1980e-03) (hash(x)=21897967)
573
+ 549 train 7.759351 (lr=2.1926e-03) (hash(x)=25161929)
574
+ 550 val loss 7.8140
575
+ 550 val perplexity 2475.1101
576
+ 550 train 7.708199 (lr=2.1871e-03) (hash(x)=27465106)
577
+ 551 train 7.798253 (lr=2.1817e-03) (hash(x)=24013079)
578
+ 552 train 7.717198 (lr=2.1762e-03) (hash(x)=23142015)
579
+ 553 train 7.807951 (lr=2.1707e-03) (hash(x)=26768629)
580
+ 554 train 7.882773 (lr=2.1653e-03) (hash(x)=26393383)
581
+ 555 train 7.698513 (lr=2.1598e-03) (hash(x)=22537194)
582
+ 556 train 7.960217 (lr=2.1542e-03) (hash(x)=24046036)
583
+ 557 train 7.934395 (lr=2.1487e-03) (hash(x)=24974360)
584
+ 558 train 8.190494 (lr=2.1432e-03) (hash(x)=28379928)
585
+ 559 train 7.930035 (lr=2.1377e-03) (hash(x)=25322001)
586
+ 560 train 8.024381 (lr=2.1321e-03) (hash(x)=26622031)
587
+ 561 train 7.629407 (lr=2.1266e-03) (hash(x)=20562247)
588
+ 562 train 8.061297 (lr=2.1210e-03) (hash(x)=27381885)
589
+ 563 train 8.183684 (lr=2.1154e-03) (hash(x)=27028126)
590
+ 564 train 8.019798 (lr=2.1099e-03) (hash(x)=28882928)
591
+ 565 train 7.976343 (lr=2.1043e-03) (hash(x)=25666355)
592
+ 566 train 8.029963 (lr=2.0987e-03) (hash(x)=24330810)
593
+ 567 train 7.948012 (lr=2.0931e-03) (hash(x)=26690440)
594
+ 568 train 7.773222 (lr=2.0875e-03) (hash(x)=22923592)
595
+ 569 train 7.939546 (lr=2.0818e-03) (hash(x)=27348418)
596
+ 570 train 8.026034 (lr=2.0762e-03) (hash(x)=28849848)
597
+ 571 train 8.082133 (lr=2.0706e-03) (hash(x)=26967331)
598
+ 572 train 7.794582 (lr=2.0649e-03) (hash(x)=22831467)
599
+ 573 train 7.878971 (lr=2.0593e-03) (hash(x)=24765121)
600
+ 574 train 7.854972 (lr=2.0536e-03) (hash(x)=24331857)
601
+ 575 train 7.691541 (lr=2.0479e-03) (hash(x)=22598512)
602
+ 576 train 7.875478 (lr=2.0422e-03) (hash(x)=25149353)
603
+ 577 train 7.690314 (lr=2.0366e-03) (hash(x)=23725598)
604
+ 578 train 7.979048 (lr=2.0309e-03) (hash(x)=26449557)
605
+ 579 train 8.041533 (lr=2.0252e-03) (hash(x)=24697985)
606
+ 580 train 8.048884 (lr=2.0194e-03) (hash(x)=26923059)
607
+ 581 train 7.859270 (lr=2.0137e-03) (hash(x)=25201962)
608
+ 582 train 7.487677 (lr=2.0080e-03) (hash(x)=20931520)
609
+ 583 train 7.482011 (lr=2.0023e-03) (hash(x)=18473911)
610
+ 584 train 7.710074 (lr=1.9965e-03) (hash(x)=21306267)
611
+ 585 train 7.987695 (lr=1.9908e-03) (hash(x)=25982840)
612
+ 586 train 7.830758 (lr=1.9850e-03) (hash(x)=25364874)
613
+ 587 train 7.708626 (lr=1.9793e-03) (hash(x)=23172124)
614
+ 588 train 8.013745 (lr=1.9735e-03) (hash(x)=27876897)
615
+ 589 train 8.584651 (lr=1.9677e-03) (hash(x)=34646114)
616
+ 590 train 8.921221 (lr=1.9620e-03) (hash(x)=35153576)
617
+ 591 train 7.814104 (lr=1.9562e-03) (hash(x)=22322442)
618
+ 592 train 8.074634 (lr=1.9504e-03) (hash(x)=27907331)
619
+ 593 train 7.938614 (lr=1.9446e-03) (hash(x)=26211794)
620
+ 594 train 8.154360 (lr=1.9388e-03) (hash(x)=29291512)
621
+ 595 train 8.391970 (lr=1.9330e-03) (hash(x)=29659121)
622
+ 596 train 8.173673 (lr=1.9272e-03) (hash(x)=29674399)
623
+ 597 train 7.842803 (lr=1.9214e-03) (hash(x)=23538306)
624
+ 598 train 7.956140 (lr=1.9156e-03) (hash(x)=21991524)
625
+ 599 train 7.936260 (lr=1.9097e-03) (hash(x)=26324153)
626
+ 600 val loss 7.8651
627
+ 600 val perplexity 2604.6477
628
+ 600 train 7.874713 (lr=1.9039e-03) (hash(x)=23712082)
629
+ 601 train 7.835712 (lr=1.8981e-03) (hash(x)=24910403)
630
+ 602 train 7.869873 (lr=1.8922e-03) (hash(x)=26737205)
631
+ 603 train 8.050651 (lr=1.8864e-03) (hash(x)=26939970)
632
+ 604 train 7.928214 (lr=1.8805e-03) (hash(x)=27651943)
633
+ 605 train 8.231305 (lr=1.8747e-03) (hash(x)=27515446)
634
+ 606 train 8.162839 (lr=1.8688e-03) (hash(x)=26753129)
635
+ 607 train 7.785797 (lr=1.8629e-03) (hash(x)=23446058)
636
+ 608 train 8.166451 (lr=1.8571e-03) (hash(x)=27587849)
637
+ 609 train 7.990160 (lr=1.8512e-03) (hash(x)=25308253)
638
+ 610 train 8.034392 (lr=1.8453e-03) (hash(x)=26615098)
639
+ 611 train 8.037909 (lr=1.8395e-03) (hash(x)=29981801)
640
+ 612 train 8.120684 (lr=1.8336e-03) (hash(x)=29592345)
641
+ 613 train 8.168851 (lr=1.8277e-03) (hash(x)=23470413)
642
+ 614 train 8.072867 (lr=1.8218e-03) (hash(x)=24742370)
643
+ 615 train 8.045070 (lr=1.8159e-03) (hash(x)=24843741)
644
+ 616 train 7.910275 (lr=1.8100e-03) (hash(x)=25192548)
645
+ 617 train 8.028848 (lr=1.8041e-03) (hash(x)=27176996)
646
+ 618 train 7.861001 (lr=1.7982e-03) (hash(x)=23964552)
647
+ 619 train 7.852095 (lr=1.7923e-03) (hash(x)=22855363)
648
+ 620 train 7.825708 (lr=1.7864e-03) (hash(x)=26332996)
649
+ 621 train 7.609999 (lr=1.7805e-03) (hash(x)=22960957)
650
+ 622 train 7.976560 (lr=1.7746e-03) (hash(x)=22752597)
651
+ 623 train 7.836250 (lr=1.7686e-03) (hash(x)=23197102)
652
+ 624 train 8.365653 (lr=1.7627e-03) (hash(x)=27383319)
653
+ 625 train 7.894339 (lr=1.7568e-03) (hash(x)=26132276)
654
+ 626 train 7.753107 (lr=1.7509e-03) (hash(x)=19810497)
655
+ 627 train 7.921194 (lr=1.7450e-03) (hash(x)=25704919)
656
+ 628 train 8.084976 (lr=1.7390e-03) (hash(x)=27174264)
657
+ 629 train 8.113889 (lr=1.7331e-03) (hash(x)=22280814)
658
+ 630 train 8.837309 (lr=1.7272e-03) (hash(x)=24279448)
659
+ 631 train 8.242345 (lr=1.7212e-03) (hash(x)=23054940)
660
+ 632 train 8.336913 (lr=1.7153e-03) (hash(x)=23942400)
661
+ 633 train 8.558918 (lr=1.7094e-03) (hash(x)=24712416)
662
+ 634 train 8.169353 (lr=1.7034e-03) (hash(x)=24564658)
663
+ 635 train 7.918312 (lr=1.6975e-03) (hash(x)=24909904)
664
+ 636 train 7.926790 (lr=1.6916e-03) (hash(x)=23049534)
665
+ 637 train 8.034298 (lr=1.6856e-03) (hash(x)=24321591)
666
+ 638 train 7.894272 (lr=1.6797e-03) (hash(x)=26153298)
667
+ 639 train 7.925186 (lr=1.6738e-03) (hash(x)=27140757)
668
+ 640 train 7.719663 (lr=1.6678e-03) (hash(x)=25115907)
669
+ 641 train 7.725295 (lr=1.6619e-03) (hash(x)=26563770)
670
+ 642 train 7.967435 (lr=1.6559e-03) (hash(x)=28089252)
671
+ 643 train 7.939944 (lr=1.6500e-03) (hash(x)=25993110)
672
+ 644 train 7.836481 (lr=1.6441e-03) (hash(x)=25150008)
673
+ 645 train 8.139997 (lr=1.6381e-03) (hash(x)=28520222)
674
+ 646 train 7.760880 (lr=1.6322e-03) (hash(x)=21349943)
675
+ 647 train 7.862237 (lr=1.6262e-03) (hash(x)=25149419)
676
+ 648 train 7.997051 (lr=1.6203e-03) (hash(x)=25730641)
677
+ 649 train 8.188603 (lr=1.6144e-03) (hash(x)=26112813)
678
+ 650 val loss 7.9197
679
+ 650 val perplexity 2751.0627
680
+ 650 train 7.963179 (lr=1.6084e-03) (hash(x)=25907805)
681
+ 651 train 8.083537 (lr=1.6025e-03) (hash(x)=27623643)
682
+ 652 train 7.984403 (lr=1.5966e-03) (hash(x)=26484959)
683
+ 653 train 8.117291 (lr=1.5906e-03) (hash(x)=29199854)
684
+ 654 train 8.065578 (lr=1.5847e-03) (hash(x)=28369628)
685
+ 655 train 7.912399 (lr=1.5788e-03) (hash(x)=24727764)
686
+ 656 train 7.541097 (lr=1.5728e-03) (hash(x)=22610673)
687
+ 657 train 7.459715 (lr=1.5669e-03) (hash(x)=22667179)
688
+ 658 train 7.255647 (lr=1.5610e-03) (hash(x)=18477300)
689
+ 659 train 7.781798 (lr=1.5550e-03) (hash(x)=23155773)
690
+ 660 train 7.518725 (lr=1.5491e-03) (hash(x)=19461032)
691
+ 661 train 7.834856 (lr=1.5432e-03) (hash(x)=23453788)
692
+ 662 train 7.885877 (lr=1.5373e-03) (hash(x)=24543466)
693
+ 663 train 7.627668 (lr=1.5314e-03) (hash(x)=21935931)
694
+ 664 train 7.464038 (lr=1.5254e-03) (hash(x)=19910292)
695
+ 665 train 7.757491 (lr=1.5195e-03) (hash(x)=24481079)
696
+ 666 train 7.405905 (lr=1.5136e-03) (hash(x)=18922411)
697
+ 667 train 7.471527 (lr=1.5077e-03) (hash(x)=20054917)
698
+ 668 train 7.772372 (lr=1.5018e-03) (hash(x)=24850470)
699
+ 669 train 7.938658 (lr=1.4959e-03) (hash(x)=25907741)
700
+ 670 train 8.110090 (lr=1.4900e-03) (hash(x)=26873522)
701
+ 671 train 8.165524 (lr=1.4841e-03) (hash(x)=27606073)
702
+ 672 train 8.280410 (lr=1.4782e-03) (hash(x)=26209645)
703
+ 673 train 8.036986 (lr=1.4723e-03) (hash(x)=25202001)
704
+ 674 train 8.105918 (lr=1.4664e-03) (hash(x)=25569462)
705
+ 675 train 8.089462 (lr=1.4605e-03) (hash(x)=26534487)
706
+ 676 train 8.015327 (lr=1.4547e-03) (hash(x)=26455057)
707
+ 677 train 7.860612 (lr=1.4488e-03) (hash(x)=24095850)
708
+ 678 train 8.097091 (lr=1.4429e-03) (hash(x)=25287752)
709
+ 679 train 7.595596 (lr=1.4371e-03) (hash(x)=22450341)
710
+ 680 train 8.299162 (lr=1.4312e-03) (hash(x)=29004853)
711
+ 681 train 8.082239 (lr=1.4253e-03) (hash(x)=27993763)
712
+ 682 train 7.843437 (lr=1.4195e-03) (hash(x)=26382658)
713
+ 683 train 7.966227 (lr=1.4136e-03) (hash(x)=25013073)
714
+ 684 train 8.110447 (lr=1.4078e-03) (hash(x)=30595809)
715
+ 685 train 8.133888 (lr=1.4019e-03) (hash(x)=30934371)
716
+ 686 train 8.646014 (lr=1.3961e-03) (hash(x)=33060834)
717
+ 687 train 7.970989 (lr=1.3903e-03) (hash(x)=25945859)
718
+ 688 train 7.781629 (lr=1.3844e-03) (hash(x)=23375678)
719
+ 689 train 7.825696 (lr=1.3786e-03) (hash(x)=25218689)
720
+ 690 train 7.995567 (lr=1.3728e-03) (hash(x)=28127397)
721
+ 691 train 7.863518 (lr=1.3670e-03) (hash(x)=24418091)
722
+ 692 train 7.740605 (lr=1.3612e-03) (hash(x)=22761099)
723
+ 693 train 7.779025 (lr=1.3554e-03) (hash(x)=24615466)
724
+ 694 train 7.932081 (lr=1.3496e-03) (hash(x)=24699240)
725
+ 695 train 7.783148 (lr=1.3438e-03) (hash(x)=21613707)
726
+ 696 train 7.956924 (lr=1.3380e-03) (hash(x)=24977554)
727
+ 697 train 7.854001 (lr=1.3323e-03) (hash(x)=24348175)
728
+ 698 train 7.969176 (lr=1.3265e-03) (hash(x)=25102767)
729
+ 699 train 7.900780 (lr=1.3207e-03) (hash(x)=26386157)
730
+ 700 val loss 7.8504
731
+ 700 val perplexity 2566.7717
732
+ 700 train 7.880644 (lr=1.3150e-03) (hash(x)=26423460)
733
+ 701 train 7.858600 (lr=1.3092e-03) (hash(x)=25771047)
734
+ 702 train 7.893390 (lr=1.3035e-03) (hash(x)=26999875)
735
+ 703 train 7.791124 (lr=1.2977e-03) (hash(x)=24396519)
736
+ 704 train 7.802820 (lr=1.2920e-03) (hash(x)=22588122)
737
+ 705 train 8.125775 (lr=1.2863e-03) (hash(x)=25142399)
738
+ 706 train 7.850517 (lr=1.2806e-03) (hash(x)=20440214)
739
+ 707 train 7.711430 (lr=1.2748e-03) (hash(x)=23265507)
740
+ 708 train 7.824207 (lr=1.2691e-03) (hash(x)=24563470)
741
+ 709 train 7.724874 (lr=1.2634e-03) (hash(x)=22514858)
742
+ 710 train 8.113721 (lr=1.2578e-03) (hash(x)=26691212)
743
+ 711 train 8.098650 (lr=1.2521e-03) (hash(x)=29138828)
744
+ 712 train 8.048411 (lr=1.2464e-03) (hash(x)=28028528)
745
+ 713 train 7.844974 (lr=1.2407e-03) (hash(x)=20531210)
746
+ 714 train 7.777299 (lr=1.2351e-03) (hash(x)=25075352)
747
+ 715 train 7.890312 (lr=1.2294e-03) (hash(x)=24265353)
748
+ 716 train 7.968892 (lr=1.2238e-03) (hash(x)=24635726)
749
+ 717 train 7.810697 (lr=1.2182e-03) (hash(x)=24999726)
750
+ 718 train 8.036819 (lr=1.2125e-03) (hash(x)=27412910)
751
+ 719 train 7.812912 (lr=1.2069e-03) (hash(x)=24685515)
752
+ 720 train 7.881141 (lr=1.2013e-03) (hash(x)=23780329)
753
+ 721 train 7.752870 (lr=1.1957e-03) (hash(x)=25071701)
754
+ 722 train 7.856833 (lr=1.1901e-03) (hash(x)=23767130)
755
+ 723 train 7.874541 (lr=1.1846e-03) (hash(x)=24876269)
756
+ 724 train 7.975599 (lr=1.1790e-03) (hash(x)=26405773)
757
+ 725 train 8.332828 (lr=1.1734e-03) (hash(x)=31733180)
758
+ 726 train 7.626472 (lr=1.1679e-03) (hash(x)=21337509)
759
+ 727 train 7.773012 (lr=1.1623e-03) (hash(x)=22825749)
760
+ 728 train 8.051696 (lr=1.1568e-03) (hash(x)=28638695)
761
+ 729 train 7.927997 (lr=1.1513e-03) (hash(x)=26393943)
762
+ 730 train 8.084762 (lr=1.1458e-03) (hash(x)=27563583)
763
+ 731 train 7.650697 (lr=1.1402e-03) (hash(x)=21239652)
764
+ 732 train 7.743244 (lr=1.1347e-03) (hash(x)=23986428)
765
+ 733 train 7.677908 (lr=1.1293e-03) (hash(x)=24943881)
766
+ 734 train 8.175119 (lr=1.1238e-03) (hash(x)=29691448)
767
+ 735 train 8.128870 (lr=1.1183e-03) (hash(x)=28767869)
768
+ 736 train 7.860191 (lr=1.1129e-03) (hash(x)=23628188)
769
+ 737 train 8.156771 (lr=1.1074e-03) (hash(x)=29341482)
770
+ 738 train 7.959488 (lr=1.1020e-03) (hash(x)=30336570)
771
+ 739 train 7.911068 (lr=1.0966e-03) (hash(x)=25614301)
772
+ 740 train 7.755516 (lr=1.0912e-03) (hash(x)=24160500)
773
+ 741 train 7.955553 (lr=1.0858e-03) (hash(x)=26030058)
774
+ 742 train 8.270194 (lr=1.0804e-03) (hash(x)=29243936)
775
+ 743 train 7.639997 (lr=1.0750e-03) (hash(x)=21159060)
776
+ 744 train 7.710566 (lr=1.0696e-03) (hash(x)=23701853)
777
+ 745 train 7.879870 (lr=1.0643e-03) (hash(x)=24629937)
778
+ 746 train 7.785533 (lr=1.0589e-03) (hash(x)=25110108)
779
+ 747 train 7.881416 (lr=1.0536e-03) (hash(x)=26751788)
780
+ 748 train 7.931753 (lr=1.0483e-03) (hash(x)=26430427)
781
+ 749 train 7.778419 (lr=1.0429e-03) (hash(x)=26012353)
782
+ 750 val loss 7.8195
783
+ 750 val perplexity 2488.6594
784
+ 750 train 7.948356 (lr=1.0376e-03) (hash(x)=22735910)
785
+ 751 train 7.714054 (lr=1.0324e-03) (hash(x)=25045397)
786
+ 752 train 7.445165 (lr=1.0271e-03) (hash(x)=21554427)
787
+ 753 train 7.576070 (lr=1.0218e-03) (hash(x)=23751143)
788
+ 754 train 8.282954 (lr=1.0166e-03) (hash(x)=28602273)
789
+ 755 train 8.311439 (lr=1.0113e-03) (hash(x)=29989709)
790
+ 756 train 7.726705 (lr=1.0061e-03) (hash(x)=22331648)
791
+ 757 train 8.146661 (lr=1.0009e-03) (hash(x)=31017246)
792
+ 758 train 8.091401 (lr=9.9568e-04) (hash(x)=29709045)
793
+ 759 train 8.095680 (lr=9.9049e-04) (hash(x)=25560928)
794
+ 760 train 7.811249 (lr=9.8531e-04) (hash(x)=25075464)
795
+ 761 train 7.871785 (lr=9.8015e-04) (hash(x)=27352253)
796
+ 762 train 8.048056 (lr=9.7500e-04) (hash(x)=28187891)
797
+ 763 train 7.955202 (lr=9.6986e-04) (hash(x)=26062687)
798
+ 764 train 7.971401 (lr=9.6474e-04) (hash(x)=27427811)
799
+ 765 train 8.055823 (lr=9.5963e-04) (hash(x)=27614522)
800
+ 766 train 7.789059 (lr=9.5453e-04) (hash(x)=26129544)
801
+ 767 train 8.343370 (lr=9.4944e-04) (hash(x)=28959222)
802
+ 768 train 7.923635 (lr=9.4437e-04) (hash(x)=26860067)
803
+ 769 train 7.690973 (lr=9.3932e-04) (hash(x)=25122598)
804
+ 770 train 7.795001 (lr=9.3427e-04) (hash(x)=25245030)
805
+ 771 train 7.811031 (lr=9.2924e-04) (hash(x)=25434884)
806
+ 772 train 7.891242 (lr=9.2423e-04) (hash(x)=27732790)
807
+ 773 train 7.900373 (lr=9.1923e-04) (hash(x)=27824438)
808
+ 774 train 7.986076 (lr=9.1424e-04) (hash(x)=27201953)
809
+ 775 train 7.643562 (lr=9.0927e-04) (hash(x)=21248405)
810
+ 776 train 7.524879 (lr=9.0431e-04) (hash(x)=22805934)
811
+ 777 train 7.834993 (lr=8.9936e-04) (hash(x)=26482588)
812
+ 778 train 7.868607 (lr=8.9443e-04) (hash(x)=24153691)
813
+ 779 train 8.036806 (lr=8.8952e-04) (hash(x)=25044192)
814
+ 780 train 7.913841 (lr=8.8462e-04) (hash(x)=25910078)
815
+ 781 train 7.933911 (lr=8.7973e-04) (hash(x)=28645524)
816
+ 782 train 7.733356 (lr=8.7486e-04) (hash(x)=24368498)
817
+ 783 train 7.814007 (lr=8.7001e-04) (hash(x)=25830182)
818
+ 784 train 7.939144 (lr=8.6516e-04) (hash(x)=29181807)
819
+ 785 train 7.807549 (lr=8.6034e-04) (hash(x)=25585137)
820
+ 786 train 7.785738 (lr=8.5553e-04) (hash(x)=24798246)
821
+ 787 train 7.894016 (lr=8.5073e-04) (hash(x)=26621419)
822
+ 788 train 7.432233 (lr=8.4596e-04) (hash(x)=21446891)
823
+ 789 train 7.403937 (lr=8.4119e-04) (hash(x)=22165286)
824
+ 790 train 7.647320 (lr=8.3644e-04) (hash(x)=23477219)
825
+ 791 train 7.891818 (lr=8.3171e-04) (hash(x)=25173113)
826
+ 792 train 7.751995 (lr=8.2699e-04) (hash(x)=25853788)
827
+ 793 train 7.906290 (lr=8.2229e-04) (hash(x)=27267091)
828
+ 794 train 7.716648 (lr=8.1761e-04) (hash(x)=23743694)
829
+ 795 train 7.717668 (lr=8.1294e-04) (hash(x)=24400133)
830
+ 796 train 7.773590 (lr=8.0829e-04) (hash(x)=23663639)
831
+ 797 train 7.620860 (lr=8.0365e-04) (hash(x)=23103223)
832
+ 798 train 8.234634 (lr=7.9903e-04) (hash(x)=28748411)
833
+ 799 train 7.494544 (lr=7.9443e-04) (hash(x)=23486277)
834
+ 800 val loss 7.8118
835
+ 800 val perplexity 2469.6035
836
+ 800 train 7.774972 (lr=7.8984e-04) (hash(x)=25678518)
837
+ 801 train 7.702906 (lr=7.8527e-04) (hash(x)=23421286)
838
+ 802 train 7.900369 (lr=7.8072e-04) (hash(x)=26054104)
839
+ 803 train 7.866107 (lr=7.7618e-04) (hash(x)=25978130)
840
+ 804 train 7.920819 (lr=7.7167e-04) (hash(x)=26006525)
841
+ 805 train 7.864540 (lr=7.6716e-04) (hash(x)=25769432)
842
+ 806 train 7.682384 (lr=7.6268e-04) (hash(x)=22430795)
843
+ 807 train 8.058696 (lr=7.5821e-04) (hash(x)=28916006)
844
+ 808 train 7.866414 (lr=7.5376e-04) (hash(x)=25166800)
845
+ 809 train 7.953792 (lr=7.4933e-04) (hash(x)=24226056)
846
+ 810 train 7.827980 (lr=7.4491e-04) (hash(x)=25057934)
847
+ 811 train 8.580385 (lr=7.4051e-04) (hash(x)=33195100)
848
+ 812 train 8.025210 (lr=7.3613e-04) (hash(x)=26312888)
849
+ 813 train 8.206060 (lr=7.3177e-04) (hash(x)=27730410)
850
+ 814 train 8.093561 (lr=7.2742e-04) (hash(x)=27372474)
851
+ 815 train 7.839372 (lr=7.2309e-04) (hash(x)=25556929)
852
+ 816 train 7.952575 (lr=7.1878e-04) (hash(x)=26909985)
853
+ 817 train 7.758191 (lr=7.1449e-04) (hash(x)=25991247)
854
+ 818 train 7.981721 (lr=7.1022e-04) (hash(x)=27438141)
855
+ 819 train 8.133853 (lr=7.0596e-04) (hash(x)=29536986)
856
+ 820 train 7.821153 (lr=7.0173e-04) (hash(x)=24478391)
857
+ 821 train 7.793408 (lr=6.9751e-04) (hash(x)=26125216)
858
+ 822 train 7.819049 (lr=6.9331e-04) (hash(x)=26422130)
859
+ 823 train 8.135567 (lr=6.8913e-04) (hash(x)=29648798)
860
+ 824 train 7.634712 (lr=6.8496e-04) (hash(x)=21247770)
861
+ 825 train 7.804277 (lr=6.8082e-04) (hash(x)=23195388)
862
+ 826 train 7.796803 (lr=6.7669e-04) (hash(x)=25796725)
863
+ 827 train 7.795616 (lr=6.7259e-04) (hash(x)=23124767)
864
+ 828 train 7.888444 (lr=6.6850e-04) (hash(x)=25233464)
865
+ 829 train 7.842243 (lr=6.6443e-04) (hash(x)=25713275)
866
+ 830 train 7.800237 (lr=6.6038e-04) (hash(x)=25550167)
867
+ 831 train 7.713193 (lr=6.5635e-04) (hash(x)=24976217)
868
+ 832 train 8.014360 (lr=6.5234e-04) (hash(x)=28536827)
869
+ 833 train 7.931560 (lr=6.4835e-04) (hash(x)=27500801)
870
+ 834 train 7.750980 (lr=6.4437e-04) (hash(x)=25545765)
871
+ 835 train 7.757008 (lr=6.4042e-04) (hash(x)=23632825)
872
+ 836 train 7.815829 (lr=6.3649e-04) (hash(x)=25708009)
873
+ 837 train 7.846014 (lr=6.3257e-04) (hash(x)=24456276)
874
+ 838 train 7.974057 (lr=6.2868e-04) (hash(x)=29189855)
875
+ 839 train 8.159251 (lr=6.2480e-04) (hash(x)=31019606)
876
+ 840 train 7.987680 (lr=6.2095e-04) (hash(x)=26328013)
877
+ 841 train 7.919805 (lr=6.1711e-04) (hash(x)=25027904)
878
+ 842 train 7.713434 (lr=6.1330e-04) (hash(x)=23734189)
879
+ 843 train 8.024440 (lr=6.0950e-04) (hash(x)=28236580)
880
+ 844 train 7.784794 (lr=6.0573e-04) (hash(x)=26509780)
881
+ 845 train 7.840784 (lr=6.0198e-04) (hash(x)=25386473)
882
+ 846 train 7.801578 (lr=5.9824e-04) (hash(x)=24052671)
883
+ 847 train 7.962530 (lr=5.9453e-04) (hash(x)=28269421)
884
+ 848 train 7.459465 (lr=5.9083e-04) (hash(x)=22251724)
885
+ 849 train 7.702300 (lr=5.8716e-04) (hash(x)=24308447)
886
+ 850 val loss 7.8000
887
+ 850 val perplexity 2440.5828
888
+ 850 train 7.666711 (lr=5.8351e-04) (hash(x)=24242830)
889
+ 851 train 7.809421 (lr=5.7988e-04) (hash(x)=25563279)
890
+ 852 train 7.860954 (lr=5.7627e-04) (hash(x)=26354481)
891
+ 853 train 7.968563 (lr=5.7268e-04) (hash(x)=26152637)
892
+ 854 train 8.037789 (lr=5.6911e-04) (hash(x)=28051025)
893
+ 855 train 7.816593 (lr=5.6556e-04) (hash(x)=24865358)
894
+ 856 train 7.742359 (lr=5.6203e-04) (hash(x)=24288911)
895
+ 857 train 7.643682 (lr=5.5853e-04) (hash(x)=22230964)
896
+ 858 train 7.512618 (lr=5.5504e-04) (hash(x)=21303832)
897
+ 859 train 7.677808 (lr=5.5158e-04) (hash(x)=22155546)
898
+ 860 train 7.826478 (lr=5.4813e-04) (hash(x)=25296428)
899
+ 861 train 7.975361 (lr=5.4471e-04) (hash(x)=29142319)
900
+ 862 train 7.808402 (lr=5.4131e-04) (hash(x)=25545430)
901
+ 863 train 7.695236 (lr=5.3793e-04) (hash(x)=26984272)
902
+ 864 train 7.811595 (lr=5.3458e-04) (hash(x)=25429005)
903
+ 865 train 7.885201 (lr=5.3124e-04) (hash(x)=27077032)
904
+ 866 train 7.857076 (lr=5.2793e-04) (hash(x)=26494424)
905
+ 867 train 7.810881 (lr=5.2464e-04) (hash(x)=23193673)
906
+ 868 train 7.832891 (lr=5.2137e-04) (hash(x)=25075134)
907
+ 869 train 7.947797 (lr=5.1812e-04) (hash(x)=27112558)
908
+ 870 train 8.042686 (lr=5.1489e-04) (hash(x)=27436608)
909
+ 871 train 7.738876 (lr=5.1169e-04) (hash(x)=24544116)
910
+ 872 train 8.128941 (lr=5.0851e-04) (hash(x)=31632686)
911
+ 873 train 7.719167 (lr=5.0534e-04) (hash(x)=25890184)
912
+ 874 train 7.745496 (lr=5.0221e-04) (hash(x)=22887555)
913
+ 875 train 7.703789 (lr=4.9909e-04) (hash(x)=24547533)
914
+ 876 train 7.868585 (lr=4.9600e-04) (hash(x)=26553496)
915
+ 877 train 7.978447 (lr=4.9293e-04) (hash(x)=27467688)
916
+ 878 train 7.734955 (lr=4.8988e-04) (hash(x)=24766934)
917
+ 879 train 7.608982 (lr=4.8685e-04) (hash(x)=22059850)
918
+ 880 train 7.986363 (lr=4.8385e-04) (hash(x)=22871702)
919
+ 881 train 7.811349 (lr=4.8087e-04) (hash(x)=23893130)
920
+ 882 train 7.904501 (lr=4.7791e-04) (hash(x)=25125691)
921
+ 883 train 7.809424 (lr=4.7497e-04) (hash(x)=25994573)
922
+ 884 train 7.864077 (lr=4.7206e-04) (hash(x)=26076345)
923
+ 885 train 7.814703 (lr=4.6917e-04) (hash(x)=26577783)
924
+ 886 train 7.912715 (lr=4.6630e-04) (hash(x)=27395225)
925
+ 887 train 7.726818 (lr=4.6346e-04) (hash(x)=23926632)
926
+ 888 train 7.606515 (lr=4.6063e-04) (hash(x)=21737239)
927
+ 889 train 7.847373 (lr=4.5783e-04) (hash(x)=23574207)
928
+ 890 train 7.511499 (lr=4.5506e-04) (hash(x)=24365231)
929
+ 891 train 7.717731 (lr=4.5231e-04) (hash(x)=27111369)
930
+ 892 train 7.981834 (lr=4.4958e-04) (hash(x)=27290015)
931
+ 893 train 7.758650 (lr=4.4687e-04) (hash(x)=23979820)
932
+ 894 train 7.852948 (lr=4.4419e-04) (hash(x)=26450121)
933
+ 895 train 7.798074 (lr=4.4153e-04) (hash(x)=27025333)
934
+ 896 train 7.763801 (lr=4.3889e-04) (hash(x)=23624605)
935
+ 897 train 7.590566 (lr=4.3628e-04) (hash(x)=22846386)
936
+ 898 train 7.674617 (lr=4.3369e-04) (hash(x)=22970561)
937
+ 899 train 7.428049 (lr=4.3113e-04) (hash(x)=16908068)
938
+ 900 val loss 7.7754
939
+ 900 val perplexity 2381.3789
940
+ 900 train 7.744481 (lr=4.2858e-04) (hash(x)=24661446)
941
+ 901 train 7.874582 (lr=4.2607e-04) (hash(x)=25664727)
942
+ 902 train 7.822357 (lr=4.2357e-04) (hash(x)=25667011)
943
+ 903 train 7.892105 (lr=4.2110e-04) (hash(x)=29120407)
944
+ 904 train 7.839740 (lr=4.1865e-04) (hash(x)=23385735)
945
+ 905 train 7.876457 (lr=4.1623e-04) (hash(x)=25564213)
946
+ 906 train 7.961173 (lr=4.1383e-04) (hash(x)=25413898)
947
+ 907 train 7.919232 (lr=4.1146e-04) (hash(x)=27092710)
948
+ 908 train 7.697218 (lr=4.0910e-04) (hash(x)=25789923)
949
+ 909 train 7.838205 (lr=4.0678e-04) (hash(x)=28533197)
950
+ 910 train 7.650478 (lr=4.0447e-04) (hash(x)=22982996)
951
+ 911 train 7.712682 (lr=4.0220e-04) (hash(x)=23827393)
952
+ 912 train 7.379932 (lr=3.9994e-04) (hash(x)=21242640)
953
+ 913 train 7.776184 (lr=3.9771e-04) (hash(x)=24154233)
954
+ 914 train 7.777840 (lr=3.9550e-04) (hash(x)=24331967)
955
+ 915 train 8.050603 (lr=3.9332e-04) (hash(x)=32812727)
956
+ 916 train 7.682878 (lr=3.9116e-04) (hash(x)=23572994)
957
+ 917 train 7.744878 (lr=3.8903e-04) (hash(x)=26305435)
958
+ 918 train 7.799212 (lr=3.8692e-04) (hash(x)=26268355)
959
+ 919 train 7.932331 (lr=3.8484e-04) (hash(x)=27230027)
960
+ 920 train 7.719583 (lr=3.8277e-04) (hash(x)=23885377)
961
+ 921 train 7.643239 (lr=3.8074e-04) (hash(x)=23532437)
962
+ 922 train 7.790372 (lr=3.7873e-04) (hash(x)=25577034)
963
+ 923 train 7.895665 (lr=3.7674e-04) (hash(x)=25703381)
964
+ 924 train 7.881172 (lr=3.7478e-04) (hash(x)=27113866)
965
+ 925 train 7.739983 (lr=3.7284e-04) (hash(x)=26961429)
966
+ 926 train 7.488126 (lr=3.7093e-04) (hash(x)=21355372)
967
+ 927 train 7.687079 (lr=3.6904e-04) (hash(x)=24968260)
968
+ 928 train 7.685682 (lr=3.6718e-04) (hash(x)=25357517)
969
+ 929 train 7.752792 (lr=3.6534e-04) (hash(x)=24854265)
970
+ 930 train 7.471004 (lr=3.6353e-04) (hash(x)=21102770)
971
+ 931 train 7.818035 (lr=3.6174e-04) (hash(x)=25676468)
972
+ 932 train 7.750749 (lr=3.5998e-04) (hash(x)=22809869)
973
+ 933 train 7.879001 (lr=3.5824e-04) (hash(x)=25503865)
974
+ 934 train 7.724644 (lr=3.5653e-04) (hash(x)=24853995)
975
+ 935 train 7.922441 (lr=3.5484e-04) (hash(x)=27544803)
976
+ 936 train 7.908697 (lr=3.5317e-04) (hash(x)=25981933)
977
+ 937 train 7.651652 (lr=3.5154e-04) (hash(x)=24658683)
978
+ 938 train 7.838176 (lr=3.4992e-04) (hash(x)=23855201)
979
+ 939 train 7.729529 (lr=3.4833e-04) (hash(x)=24331407)
980
+ 940 train 8.007385 (lr=3.4677e-04) (hash(x)=29265551)
981
+ 941 train 7.627910 (lr=3.4523e-04) (hash(x)=21892556)
982
+ 942 train 7.705571 (lr=3.4372e-04) (hash(x)=27183405)
983
+ 943 train 7.678262 (lr=3.4224e-04) (hash(x)=26540663)
984
+ 944 train 7.614240 (lr=3.4077e-04) (hash(x)=25718393)
985
+ 945 train 7.800020 (lr=3.3934e-04) (hash(x)=26819462)
986
+ 946 train 7.837686 (lr=3.3793e-04) (hash(x)=27427540)
987
+ 947 train 7.774130 (lr=3.3654e-04) (hash(x)=25532657)
988
+ 948 train 8.007024 (lr=3.3518e-04) (hash(x)=27641372)
989
+ 949 train 7.775647 (lr=3.3385e-04) (hash(x)=26515570)
990
+ 950 val loss 7.7595
991
+ 950 val perplexity 2343.7549
992
+ 950 train 7.944380 (lr=3.3254e-04) (hash(x)=26911957)
993
+ 951 train 8.076479 (lr=3.3125e-04) (hash(x)=25856625)
994
+ 952 train 7.632799 (lr=3.3000e-04) (hash(x)=25219129)
995
+ 953 train 7.621408 (lr=3.2876e-04) (hash(x)=25260471)
996
+ 954 train 8.034060 (lr=3.2756e-04) (hash(x)=29373370)
997
+ 955 train 7.537371 (lr=3.2638e-04) (hash(x)=23437426)
998
+ 956 train 7.842114 (lr=3.2522e-04) (hash(x)=23769521)
999
+ 957 train 7.866672 (lr=3.2409e-04) (hash(x)=25961833)
1000
+ 958 train 7.591864 (lr=3.2299e-04) (hash(x)=23582666)
1001
+ 959 train 7.687799 (lr=3.2191e-04) (hash(x)=23164356)
1002
+ 960 train 7.689415 (lr=3.2085e-04) (hash(x)=24443114)
1003
+ 961 train 7.729993 (lr=3.1983e-04) (hash(x)=25052665)
1004
+ 962 train 7.887467 (lr=3.1883e-04) (hash(x)=27802272)
1005
+ 963 train 7.758301 (lr=3.1785e-04) (hash(x)=25957896)
1006
+ 964 train 7.811704 (lr=3.1690e-04) (hash(x)=26737251)
1007
+ 965 train 7.877586 (lr=3.1598e-04) (hash(x)=24723263)
1008
+ 966 train 7.815750 (lr=3.1508e-04) (hash(x)=24707011)
1009
+ 967 train 7.808415 (lr=3.1421e-04) (hash(x)=25646282)
1010
+ 968 train 7.820584 (lr=3.1336e-04) (hash(x)=27544665)
1011
+ 969 train 7.750632 (lr=3.1254e-04) (hash(x)=25851993)
1012
+ 970 train 8.032521 (lr=3.1174e-04) (hash(x)=29059700)
1013
+ 971 train 7.687979 (lr=3.1098e-04) (hash(x)=21513584)
1014
+ 972 train 7.531812 (lr=3.1023e-04) (hash(x)=23151267)
1015
+ 973 train 7.772121 (lr=3.0952e-04) (hash(x)=26017176)
1016
+ 974 train 7.860256 (lr=3.0882e-04) (hash(x)=26979518)
1017
+ 975 train 7.801322 (lr=3.0816e-04) (hash(x)=23843233)
1018
+ 976 train 7.908103 (lr=3.0752e-04) (hash(x)=24193010)
1019
+ 977 train 7.666911 (lr=3.0691e-04) (hash(x)=21476847)
1020
+ 978 train 7.539356 (lr=3.0632e-04) (hash(x)=21366504)
1021
+ 979 train 7.648708 (lr=3.0576e-04) (hash(x)=23226697)
1022
+ 980 train 7.394418 (lr=3.0522e-04) (hash(x)=19961773)
1023
+ 981 train 7.219347 (lr=3.0471e-04) (hash(x)=19772969)
1024
+ 982 train 7.546047 (lr=3.0423e-04) (hash(x)=23110142)
1025
+ 983 train 7.830426 (lr=3.0377e-04) (hash(x)=24506028)
1026
+ 984 train 7.761864 (lr=3.0334e-04) (hash(x)=25480731)
1027
+ 985 train 7.531164 (lr=3.0294e-04) (hash(x)=21077417)
1028
+ 986 train 7.638887 (lr=3.0256e-04) (hash(x)=23686713)
1029
+ 987 train 7.771228 (lr=3.0221e-04) (hash(x)=26024321)
1030
+ 988 train 7.841058 (lr=3.0188e-04) (hash(x)=27424109)
1031
+ 989 train 7.908731 (lr=3.0158e-04) (hash(x)=27786174)
1032
+ 990 train 7.732447 (lr=3.0131e-04) (hash(x)=25232502)
1033
+ 991 train 7.730201 (lr=3.0106e-04) (hash(x)=22781277)
1034
+ 992 train 7.777014 (lr=3.0084e-04) (hash(x)=26184527)
1035
+ 993 train 7.815203 (lr=3.0064e-04) (hash(x)=24459895)
1036
+ 994 train 7.742420 (lr=3.0047e-04) (hash(x)=25244624)
1037
+ 995 train 7.884809 (lr=3.0033e-04) (hash(x)=24451843)
1038
+ 996 train 7.857201 (lr=3.0021e-04) (hash(x)=22129897)
1039
+ 997 train 7.672322 (lr=3.0012e-04) (hash(x)=21116390)
1040
+ 998 train 7.484179 (lr=3.0005e-04) (hash(x)=20650070)
1041
+ 999 val loss 7.7514
1042
+ 999 val perplexity 2324.8494
1043
+ 999 train 7.898430 (lr=3.0001e-04) (hash(x)=24948650)
lr3e-3_total_batch_size5120_seq_len512/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06b304f048efadc53c4a9c453406b226a1e851e1f5fc9e167d265effa634e1b5
3
+ size 48156226
lr3e-3_total_batch_size5120_seq_len512/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d509ac9dd3f90ecc4cb815d02e48986ade3e64c28d25079056b455c2309e4de
3
+ size 71157574