andrew-healey commited on
Commit
93c7d46
·
verified ·
1 Parent(s): 681d536

Upload folder using huggingface_hub

Browse files
lr3e-3_total_batch_size40960_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr3e-3_total_batch_size40960_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": 40960, "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_size40960_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cec289292cf51a5462e2631af59109e64df3eca2d9e62e1fbd1c53a776a0f874
3
+ size 964
lr3e-3_total_batch_size40960_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 11.697288 (lr=1.0490e-05) (hash(x)=5534142)
5
+ 1 train 11.664479 (lr=2.0979e-05) (hash(x)=6328012)
6
+ 2 train 11.583329 (lr=3.1469e-05) (hash(x)=5556098)
7
+ 3 train 11.452430 (lr=4.1958e-05) (hash(x)=6219138)
8
+ 4 train 11.301478 (lr=5.2448e-05) (hash(x)=5491972)
9
+ 5 train 11.121954 (lr=6.2937e-05) (hash(x)=5917741)
10
+ 6 train 10.950987 (lr=7.3427e-05) (hash(x)=6112458)
11
+ 7 train 10.788260 (lr=8.3916e-05) (hash(x)=6568379)
12
+ 8 train 10.616850 (lr=9.4406e-05) (hash(x)=7386082)
13
+ 9 train 10.523159 (lr=1.0490e-04) (hash(x)=6151674)
14
+ 10 train 10.476165 (lr=1.1538e-04) (hash(x)=7859060)
15
+ 11 train 10.401353 (lr=1.2587e-04) (hash(x)=6391506)
16
+ 12 train 10.381824 (lr=1.3636e-04) (hash(x)=5985675)
17
+ 13 train 10.369776 (lr=1.4685e-04) (hash(x)=6984772)
18
+ 14 train 10.312874 (lr=1.5734e-04) (hash(x)=6603717)
19
+ 15 train 10.217746 (lr=1.6783e-04) (hash(x)=6190579)
20
+ 16 train 10.191489 (lr=1.7832e-04) (hash(x)=7276303)
21
+ 17 train 10.225484 (lr=1.8881e-04) (hash(x)=6630378)
22
+ 18 train 10.206956 (lr=1.9930e-04) (hash(x)=9989744)
23
+ 19 train 10.040196 (lr=2.0979e-04) (hash(x)=6317894)
24
+ 20 train 9.903807 (lr=2.2028e-04) (hash(x)=7947703)
25
+ 21 train 9.804403 (lr=2.3077e-04) (hash(x)=5777498)
26
+ 22 train 9.719789 (lr=2.4126e-04) (hash(x)=7723226)
27
+ 23 train 9.590364 (lr=2.5175e-04) (hash(x)=6491229)
28
+ 24 train 9.598219 (lr=2.6224e-04) (hash(x)=6019389)
29
+ 25 train 9.442539 (lr=2.7273e-04) (hash(x)=4648609)
30
+ 26 train 9.272500 (lr=2.8322e-04) (hash(x)=5558355)
31
+ 27 train 9.208429 (lr=2.9371e-04) (hash(x)=6369326)
32
+ 28 train 9.032048 (lr=3.0420e-04) (hash(x)=4906853)
33
+ 29 train 8.791049 (lr=3.1469e-04) (hash(x)=6437376)
34
+ 30 train 8.860345 (lr=3.2517e-04) (hash(x)=7084093)
35
+ 31 train 8.681697 (lr=3.3566e-04) (hash(x)=7048804)
36
+ 32 train 8.612520 (lr=3.4615e-04) (hash(x)=5752594)
37
+ 33 train 8.299013 (lr=3.5664e-04) (hash(x)=6393520)
38
+ 34 train 8.447812 (lr=3.6713e-04) (hash(x)=6793550)
39
+ 35 train 8.246338 (lr=3.7762e-04) (hash(x)=6581348)
40
+ 36 train 8.132047 (lr=3.8811e-04) (hash(x)=6235837)
41
+ 37 train 8.166009 (lr=3.9860e-04) (hash(x)=7083021)
42
+ 38 train 8.175107 (lr=4.0909e-04) (hash(x)=6155821)
43
+ 39 train 7.954661 (lr=4.1958e-04) (hash(x)=5657123)
44
+ 40 train 7.899768 (lr=4.3007e-04) (hash(x)=5312267)
45
+ 41 train 7.749982 (lr=4.4056e-04) (hash(x)=6830813)
46
+ 42 train 7.766180 (lr=4.5105e-04) (hash(x)=6424933)
47
+ 43 train 7.898534 (lr=4.6154e-04) (hash(x)=6680023)
48
+ 44 train 7.881954 (lr=4.7203e-04) (hash(x)=7523279)
49
+ 45 train 7.812966 (lr=4.8252e-04) (hash(x)=6635925)
50
+ 46 train 7.779627 (lr=4.9301e-04) (hash(x)=5869441)
51
+ 47 train 7.838256 (lr=5.0350e-04) (hash(x)=6349004)
52
+ 48 train 7.683963 (lr=5.1399e-04) (hash(x)=6008454)
53
+ 49 train 7.765456 (lr=5.2448e-04) (hash(x)=7682741)
54
+ 50 val loss 7.7952
55
+ 50 val perplexity 2428.8696
56
+ 50 train 7.868549 (lr=5.3497e-04) (hash(x)=7794028)
57
+ 51 train 7.813786 (lr=5.4545e-04) (hash(x)=6166062)
58
+ 52 train 7.790914 (lr=5.5594e-04) (hash(x)=4811192)
59
+ 53 train 7.684366 (lr=5.6643e-04) (hash(x)=7822680)
60
+ 54 train 7.775131 (lr=5.7692e-04) (hash(x)=6431409)
61
+ 55 train 7.841287 (lr=5.8741e-04) (hash(x)=5323032)
62
+ 56 train 7.632245 (lr=5.9790e-04) (hash(x)=7363286)
63
+ 57 train 7.913535 (lr=6.0839e-04) (hash(x)=5616075)
64
+ 58 train 7.873405 (lr=6.1888e-04) (hash(x)=5672608)
65
+ 59 train 7.701962 (lr=6.2937e-04) (hash(x)=6839867)
66
+ 60 train 7.693372 (lr=6.3986e-04) (hash(x)=4916928)
67
+ 61 train 7.775772 (lr=6.5035e-04) (hash(x)=8353379)
68
+ 62 train 7.728631 (lr=6.6084e-04) (hash(x)=4838871)
69
+ 63 train 7.535529 (lr=6.7133e-04) (hash(x)=6795665)
70
+ 64 train 7.720943 (lr=6.8182e-04) (hash(x)=6678038)
71
+ 65 train 7.777207 (lr=6.9231e-04) (hash(x)=6112215)
72
+ 66 train 7.789667 (lr=7.0280e-04) (hash(x)=5915361)
73
+ 67 train 7.725598 (lr=7.1329e-04) (hash(x)=5771588)
74
+ 68 train 7.674642 (lr=7.2378e-04) (hash(x)=5688829)
75
+ 69 train 7.718995 (lr=7.3427e-04) (hash(x)=5443305)
76
+ 70 train 7.775645 (lr=7.4476e-04) (hash(x)=6901933)
77
+ 71 train 7.672373 (lr=7.5524e-04) (hash(x)=4294425)
78
+ 72 train 7.606356 (lr=7.6573e-04) (hash(x)=4363941)
79
+ 73 train 7.884652 (lr=7.7622e-04) (hash(x)=5486335)
80
+ 74 train 7.879192 (lr=7.8671e-04) (hash(x)=7295165)
81
+ 75 train 7.768381 (lr=7.9720e-04) (hash(x)=6267424)
82
+ 76 train 7.845061 (lr=8.0769e-04) (hash(x)=5270452)
83
+ 77 train 7.673835 (lr=8.1818e-04) (hash(x)=6305297)
84
+ 78 train 7.957179 (lr=8.2867e-04) (hash(x)=6040077)
85
+ 79 train 7.919756 (lr=8.3916e-04) (hash(x)=7130251)
86
+ 80 train 7.608723 (lr=8.4965e-04) (hash(x)=6508350)
87
+ 81 train 7.787894 (lr=8.6014e-04) (hash(x)=5859576)
88
+ 82 train 7.456119 (lr=8.7063e-04) (hash(x)=4892065)
89
+ 83 train 7.596235 (lr=8.8112e-04) (hash(x)=7317289)
90
+ 84 train 7.739205 (lr=8.9161e-04) (hash(x)=5510218)
91
+ 85 train 7.901382 (lr=9.0210e-04) (hash(x)=6752265)
92
+ 86 train 7.637979 (lr=9.1259e-04) (hash(x)=6530867)
93
+ 87 train 7.692719 (lr=9.2308e-04) (hash(x)=5790078)
94
+ 88 train 7.711093 (lr=9.3357e-04) (hash(x)=7614023)
95
+ 89 train 7.736171 (lr=9.4406e-04) (hash(x)=5245146)
96
+ 90 train 7.644411 (lr=9.5455e-04) (hash(x)=5786972)
97
+ 91 train 7.699026 (lr=9.6503e-04) (hash(x)=6346459)
98
+ 92 train 7.723455 (lr=9.7552e-04) (hash(x)=5209234)
99
+ 93 train 7.639918 (lr=9.8601e-04) (hash(x)=5676768)
100
+ 94 train 7.777446 (lr=9.9650e-04) (hash(x)=5863092)
101
+ 95 train 7.799561 (lr=1.0070e-03) (hash(x)=7019489)
102
+ 96 train 7.644855 (lr=1.0175e-03) (hash(x)=5483019)
103
+ 97 train 7.641657 (lr=1.0280e-03) (hash(x)=5968716)
104
+ 98 train 7.519772 (lr=1.0385e-03) (hash(x)=7780194)
105
+ 99 train 7.601771 (lr=1.0490e-03) (hash(x)=3784321)
106
+ 100 val loss 7.6151
107
+ 100 val perplexity 2028.6332
108
+ 100 train 7.635980 (lr=1.0594e-03) (hash(x)=7774109)
109
+ 101 train 7.805457 (lr=1.0699e-03) (hash(x)=6576749)
110
+ 102 train 7.688452 (lr=1.0804e-03) (hash(x)=7072404)
111
+ 103 train 7.597086 (lr=1.0909e-03) (hash(x)=6556025)
112
+ 104 train 7.679251 (lr=1.1014e-03) (hash(x)=7325027)
113
+ 105 train 7.680349 (lr=1.1119e-03) (hash(x)=6373412)
114
+ 106 train 7.577396 (lr=1.1224e-03) (hash(x)=5262868)
115
+ 107 train 7.497384 (lr=1.1329e-03) (hash(x)=8055563)
116
+ 108 train 7.640412 (lr=1.1434e-03) (hash(x)=5932790)
117
+ 109 train 7.599257 (lr=1.1538e-03) (hash(x)=5680154)
118
+ 110 train 7.635151 (lr=1.1643e-03) (hash(x)=5914217)
119
+ 111 train 7.541186 (lr=1.1748e-03) (hash(x)=6424128)
120
+ 112 train 7.487160 (lr=1.1853e-03) (hash(x)=7804089)
121
+ 113 train 7.606859 (lr=1.1958e-03) (hash(x)=6490149)
122
+ 114 train 7.555996 (lr=1.2063e-03) (hash(x)=6283540)
123
+ 115 train 7.532464 (lr=1.2168e-03) (hash(x)=5728337)
124
+ 116 train 7.539703 (lr=1.2273e-03) (hash(x)=7400272)
125
+ 117 train 7.550205 (lr=1.2378e-03) (hash(x)=6473432)
126
+ 118 train 7.672611 (lr=1.2483e-03) (hash(x)=6922646)
127
+ 119 train 7.533684 (lr=1.2587e-03) (hash(x)=7277141)
128
+ 120 train 7.597652 (lr=1.2692e-03) (hash(x)=7443124)
129
+ 121 train 7.591029 (lr=1.2797e-03) (hash(x)=5632735)
130
+ 122 train 7.499315 (lr=1.2902e-03) (hash(x)=5881015)
131
+ 123 train 7.550933 (lr=1.3007e-03) (hash(x)=5809411)
132
+ 124 train 7.556213 (lr=1.3112e-03) (hash(x)=7232964)
133
+ 125 train 7.514816 (lr=1.3217e-03) (hash(x)=6171829)
134
+ 126 train 7.618147 (lr=1.3322e-03) (hash(x)=5884456)
135
+ 127 train 7.569534 (lr=1.3427e-03) (hash(x)=5077031)
136
+ 128 train 7.636033 (lr=1.3531e-03) (hash(x)=6248895)
137
+ 129 train 7.547081 (lr=1.3636e-03) (hash(x)=6918004)
138
+ 130 train 7.592378 (lr=1.3741e-03) (hash(x)=6482466)
139
+ 131 train 7.527458 (lr=1.3846e-03) (hash(x)=6588573)
140
+ 132 train 7.608802 (lr=1.3951e-03) (hash(x)=6136939)
141
+ 133 train 7.488947 (lr=1.4056e-03) (hash(x)=7040157)
142
+ 134 train 7.603480 (lr=1.4161e-03) (hash(x)=6455595)
143
+ 135 train 7.578610 (lr=1.4266e-03) (hash(x)=4122983)
144
+ 136 train 7.509761 (lr=1.4371e-03) (hash(x)=4724052)
145
+ 137 train 7.477651 (lr=1.4476e-03) (hash(x)=6820797)
146
+ 138 train 7.723425 (lr=1.4580e-03) (hash(x)=7477594)
147
+ 139 train 7.456712 (lr=1.4685e-03) (hash(x)=6146305)
148
+ 140 train 7.477085 (lr=1.4790e-03) (hash(x)=6814831)
149
+ 141 train 7.538975 (lr=1.4895e-03) (hash(x)=5548298)
150
+ 142 train 7.728532 (lr=1.5000e-03) (hash(x)=7609416)
151
+ 143 train 7.682565 (lr=1.5105e-03) (hash(x)=7574342)
152
+ 144 train 7.513314 (lr=1.5210e-03) (hash(x)=5906937)
153
+ 145 train 7.551728 (lr=1.5315e-03) (hash(x)=8013471)
154
+ 146 train 7.525507 (lr=1.5420e-03) (hash(x)=6146213)
155
+ 147 train 7.546567 (lr=1.5524e-03) (hash(x)=6709902)
156
+ 148 train 7.667819 (lr=1.5629e-03) (hash(x)=6402571)
157
+ 149 train 7.518766 (lr=1.5734e-03) (hash(x)=7683614)
158
+ 150 val loss 7.5500
159
+ 150 val perplexity 1900.8174
160
+ 150 train 7.517725 (lr=1.5839e-03) (hash(x)=5453026)
161
+ 151 train 7.442407 (lr=1.5944e-03) (hash(x)=6260017)
162
+ 152 train 7.562396 (lr=1.6049e-03) (hash(x)=6591466)
163
+ 153 train 7.576917 (lr=1.6154e-03) (hash(x)=5501472)
164
+ 154 train 7.591479 (lr=1.6259e-03) (hash(x)=5871698)
165
+ 155 train 7.423154 (lr=1.6364e-03) (hash(x)=4282052)
166
+ 156 train 7.518431 (lr=1.6469e-03) (hash(x)=7824660)
167
+ 157 train 7.449726 (lr=1.6573e-03) (hash(x)=5954950)
168
+ 158 train 7.522537 (lr=1.6678e-03) (hash(x)=5484656)
169
+ 159 train 7.659804 (lr=1.6783e-03) (hash(x)=7031328)
170
+ 160 train 7.581904 (lr=1.6888e-03) (hash(x)=9157069)
171
+ 161 train 7.667650 (lr=1.6993e-03) (hash(x)=5996883)
172
+ 162 train 7.749897 (lr=1.7098e-03) (hash(x)=7645447)
173
+ 163 train 7.773850 (lr=1.7203e-03) (hash(x)=6336164)
174
+ 164 train 7.487417 (lr=1.7308e-03) (hash(x)=5224240)
175
+ 165 train 7.816998 (lr=1.7413e-03) (hash(x)=5447801)
176
+ 166 train 7.540844 (lr=1.7517e-03) (hash(x)=4155598)
177
+ 167 train 7.502088 (lr=1.7622e-03) (hash(x)=7482744)
178
+ 168 train 7.553442 (lr=1.7727e-03) (hash(x)=4733414)
179
+ 169 train 7.510495 (lr=1.7832e-03) (hash(x)=7806284)
180
+ 170 train 7.501236 (lr=1.7937e-03) (hash(x)=7236478)
181
+ 171 train 7.516132 (lr=1.8042e-03) (hash(x)=4656255)
182
+ 172 train 7.625044 (lr=1.8147e-03) (hash(x)=9298325)
183
+ 173 train 7.558259 (lr=1.8252e-03) (hash(x)=6287524)
184
+ 174 train 7.599046 (lr=1.8357e-03) (hash(x)=7895970)
185
+ 175 train 7.538168 (lr=1.8462e-03) (hash(x)=8847992)
186
+ 176 train 7.775425 (lr=1.8566e-03) (hash(x)=6338998)
187
+ 177 train 7.715359 (lr=1.8671e-03) (hash(x)=7104011)
188
+ 178 train 7.564546 (lr=1.8776e-03) (hash(x)=5573214)
189
+ 179 train 7.603178 (lr=1.8881e-03) (hash(x)=6882897)
190
+ 180 train 7.560982 (lr=1.8986e-03) (hash(x)=5854242)
191
+ 181 train 7.404141 (lr=1.9091e-03) (hash(x)=6972494)
192
+ 182 train 7.457464 (lr=1.9196e-03) (hash(x)=6796805)
193
+ 183 train 7.359481 (lr=1.9301e-03) (hash(x)=6139120)
194
+ 184 train 7.528379 (lr=1.9406e-03) (hash(x)=6486796)
195
+ 185 train 7.403976 (lr=1.9510e-03) (hash(x)=5578797)
196
+ 186 train 7.478106 (lr=1.9615e-03) (hash(x)=6895326)
197
+ 187 train 7.529750 (lr=1.9720e-03) (hash(x)=7783160)
198
+ 188 train 7.542354 (lr=1.9825e-03) (hash(x)=4867625)
199
+ 189 train 7.369759 (lr=1.9930e-03) (hash(x)=6631123)
200
+ 190 train 7.233572 (lr=2.0035e-03) (hash(x)=6710965)
201
+ 191 train 7.374309 (lr=2.0140e-03) (hash(x)=6294657)
202
+ 192 train 7.523968 (lr=2.0245e-03) (hash(x)=5443606)
203
+ 193 train 7.510797 (lr=2.0350e-03) (hash(x)=7035770)
204
+ 194 train 7.571062 (lr=2.0455e-03) (hash(x)=4899541)
205
+ 195 train 7.418818 (lr=2.0559e-03) (hash(x)=6227120)
206
+ 196 train 7.499205 (lr=2.0664e-03) (hash(x)=5981648)
207
+ 197 train 7.487247 (lr=2.0769e-03) (hash(x)=5558993)
208
+ 198 train 7.546631 (lr=2.0874e-03) (hash(x)=7015671)
209
+ 199 train 7.529815 (lr=2.0979e-03) (hash(x)=5931234)
210
+ 200 val loss 7.5323
211
+ 200 val perplexity 1867.3190
212
+ 200 train 7.493212 (lr=2.1084e-03) (hash(x)=4582159)
213
+ 201 train 7.520218 (lr=2.1189e-03) (hash(x)=5029472)
214
+ 202 train 7.451859 (lr=2.1294e-03) (hash(x)=5434102)
215
+ 203 train 7.771329 (lr=2.1399e-03) (hash(x)=6810638)
216
+ 204 train 7.580019 (lr=2.1503e-03) (hash(x)=6521562)
217
+ 205 train 7.632313 (lr=2.1608e-03) (hash(x)=7226932)
218
+ 206 train 7.510768 (lr=2.1713e-03) (hash(x)=6957849)
219
+ 207 train 7.613171 (lr=2.1818e-03) (hash(x)=5334281)
220
+ 208 train 7.665173 (lr=2.1923e-03) (hash(x)=5887983)
221
+ 209 train 7.550827 (lr=2.2028e-03) (hash(x)=6393576)
222
+ 210 train 7.792660 (lr=2.2133e-03) (hash(x)=5057959)
223
+ 211 train 7.859183 (lr=2.2238e-03) (hash(x)=8595271)
224
+ 212 train 7.661660 (lr=2.2343e-03) (hash(x)=8395167)
225
+ 213 train 7.729256 (lr=2.2448e-03) (hash(x)=7634417)
226
+ 214 train 7.481421 (lr=2.2552e-03) (hash(x)=5257771)
227
+ 215 train 7.641338 (lr=2.2657e-03) (hash(x)=7674238)
228
+ 216 train 7.748575 (lr=2.2762e-03) (hash(x)=5300555)
229
+ 217 train 7.544027 (lr=2.2867e-03) (hash(x)=6066139)
230
+ 218 train 7.618876 (lr=2.2972e-03) (hash(x)=6688212)
231
+ 219 train 7.802570 (lr=2.3077e-03) (hash(x)=6642404)
232
+ 220 train 7.596277 (lr=2.3182e-03) (hash(x)=4961570)
233
+ 221 train 7.471457 (lr=2.3287e-03) (hash(x)=5105190)
234
+ 222 train 7.757119 (lr=2.3392e-03) (hash(x)=7155991)
235
+ 223 train 7.579486 (lr=2.3497e-03) (hash(x)=5132972)
236
+ 224 train 7.583513 (lr=2.3601e-03) (hash(x)=6269946)
237
+ 225 train 7.607416 (lr=2.3706e-03) (hash(x)=5396133)
238
+ 226 train 7.615504 (lr=2.3811e-03) (hash(x)=6394083)
239
+ 227 train 7.553298 (lr=2.3916e-03) (hash(x)=5237804)
240
+ 228 train 7.593100 (lr=2.4021e-03) (hash(x)=5216638)
241
+ 229 train 7.541492 (lr=2.4126e-03) (hash(x)=6735565)
242
+ 230 train 7.650412 (lr=2.4231e-03) (hash(x)=5983575)
243
+ 231 train 7.623351 (lr=2.4336e-03) (hash(x)=5831619)
244
+ 232 train 7.454241 (lr=2.4441e-03) (hash(x)=9109083)
245
+ 233 train 7.695180 (lr=2.4545e-03) (hash(x)=6912967)
246
+ 234 train 7.870037 (lr=2.4650e-03) (hash(x)=5965674)
247
+ 235 train 7.744660 (lr=2.4755e-03) (hash(x)=5822533)
248
+ 236 train 7.637827 (lr=2.4860e-03) (hash(x)=5268216)
249
+ 237 train 7.518311 (lr=2.4965e-03) (hash(x)=6531900)
250
+ 238 train 7.497599 (lr=2.5070e-03) (hash(x)=5426499)
251
+ 239 train 7.942806 (lr=2.5175e-03) (hash(x)=7455372)
252
+ 240 train 8.156732 (lr=2.5280e-03) (hash(x)=8319119)
253
+ 241 train 8.080388 (lr=2.5385e-03) (hash(x)=7592128)
254
+ 242 train 8.075894 (lr=2.5490e-03) (hash(x)=8704170)
255
+ 243 train 7.928718 (lr=2.5594e-03) (hash(x)=7165037)
256
+ 244 train 7.777972 (lr=2.5699e-03) (hash(x)=7678000)
257
+ 245 train 7.721402 (lr=2.5804e-03) (hash(x)=7414788)
258
+ 246 train 7.870728 (lr=2.5909e-03) (hash(x)=7326216)
259
+ 247 train 8.124786 (lr=2.6014e-03) (hash(x)=5857162)
260
+ 248 train 8.669445 (lr=2.6119e-03) (hash(x)=14715669)
261
+ 249 train 11.827425 (lr=2.6224e-03) (hash(x)=16950560)
262
+ 250 val loss 7.8710
263
+ 250 val perplexity 2620.1401
264
+ 250 train 7.989854 (lr=2.6329e-03) (hash(x)=6159125)
265
+ 251 train 8.123094 (lr=2.6434e-03) (hash(x)=5908442)
266
+ 252 train 8.320396 (lr=2.6538e-03) (hash(x)=6668887)
267
+ 253 train 8.633328 (lr=2.6643e-03) (hash(x)=7662975)
268
+ 254 train 8.839959 (lr=2.6748e-03) (hash(x)=6669539)
269
+ 255 train 8.785848 (lr=2.6853e-03) (hash(x)=5015998)
270
+ 256 train 8.638851 (lr=2.6958e-03) (hash(x)=5458386)
271
+ 257 train 8.603579 (lr=2.7063e-03) (hash(x)=4849847)
272
+ 258 train 8.445528 (lr=2.7168e-03) (hash(x)=5125562)
273
+ 259 train 8.360657 (lr=2.7273e-03) (hash(x)=6012047)
274
+ 260 train 8.162292 (lr=2.7378e-03) (hash(x)=6079364)
275
+ 261 train 8.013741 (lr=2.7483e-03) (hash(x)=6030395)
276
+ 262 train 7.955608 (lr=2.7587e-03) (hash(x)=6781762)
277
+ 263 train 7.948049 (lr=2.7692e-03) (hash(x)=5238325)
278
+ 264 train 8.028419 (lr=2.7797e-03) (hash(x)=8611144)
279
+ 265 train 7.911411 (lr=2.7902e-03) (hash(x)=6901605)
280
+ 266 train 7.997202 (lr=2.8007e-03) (hash(x)=5499337)
281
+ 267 train 8.039770 (lr=2.8112e-03) (hash(x)=7134874)
282
+ 268 train 7.981185 (lr=2.8217e-03) (hash(x)=6149038)
283
+ 269 train 7.935495 (lr=2.8322e-03) (hash(x)=6609760)
284
+ 270 train 7.944668 (lr=2.8427e-03) (hash(x)=6652627)
285
+ 271 train 7.942243 (lr=2.8531e-03) (hash(x)=8618579)
286
+ 272 train 7.840546 (lr=2.8636e-03) (hash(x)=5896892)
287
+ 273 train 8.022510 (lr=2.8741e-03) (hash(x)=8945854)
288
+ 274 train 7.825257 (lr=2.8846e-03) (hash(x)=6775103)
289
+ 275 train 7.940297 (lr=2.8951e-03) (hash(x)=6200024)
290
+ 276 train 7.887071 (lr=2.9056e-03) (hash(x)=5238139)
291
+ 277 train 8.038110 (lr=2.9161e-03) (hash(x)=5780948)
292
+ 278 train 8.055530 (lr=2.9266e-03) (hash(x)=6168407)
293
+ 279 train 8.032235 (lr=2.9371e-03) (hash(x)=3775881)
294
+ 280 train 8.044285 (lr=2.9476e-03) (hash(x)=6848709)
295
+ 281 train 8.387005 (lr=2.9580e-03) (hash(x)=6478113)
296
+ 282 train 8.088246 (lr=2.9685e-03) (hash(x)=6483914)
297
+ 283 train 8.029901 (lr=2.9790e-03) (hash(x)=5927795)
298
+ 284 train 8.022542 (lr=2.9895e-03) (hash(x)=8438033)
299
+ 285 train 8.021898 (lr=3.0000e-03) (hash(x)=5684788)
300
+ 286 train 7.883074 (lr=3.0000e-03) (hash(x)=5710182)
301
+ 287 train 7.936989 (lr=3.0000e-03) (hash(x)=6160063)
302
+ 288 train 7.820657 (lr=2.9999e-03) (hash(x)=5955100)
303
+ 289 train 7.738075 (lr=2.9999e-03) (hash(x)=6672309)
304
+ 290 train 8.077601 (lr=2.9998e-03) (hash(x)=9001564)
305
+ 291 train 8.365347 (lr=2.9997e-03) (hash(x)=5454654)
306
+ 292 train 7.838997 (lr=2.9995e-03) (hash(x)=5174062)
307
+ 293 train 7.943873 (lr=2.9994e-03) (hash(x)=6199203)
308
+ 294 train 7.913346 (lr=2.9992e-03) (hash(x)=8352747)
309
+ 295 train 7.974141 (lr=2.9989e-03) (hash(x)=4547106)
310
+ 296 train 7.981630 (lr=2.9987e-03) (hash(x)=6433099)
311
+ 297 train 7.876280 (lr=2.9984e-03) (hash(x)=6424428)
312
+ 298 train 7.949889 (lr=2.9981e-03) (hash(x)=5975712)
313
+ 299 train 7.885099 (lr=2.9978e-03) (hash(x)=5458532)
314
+ 300 val loss 7.9346
315
+ 300 val perplexity 2792.1294
316
+ 300 train 8.091057 (lr=2.9974e-03) (hash(x)=7002774)
317
+ 301 train 8.157508 (lr=2.9971e-03) (hash(x)=8206173)
318
+ 302 train 8.237864 (lr=2.9967e-03) (hash(x)=7784971)
319
+ 303 train 8.187315 (lr=2.9962e-03) (hash(x)=6001843)
320
+ 304 train 8.016881 (lr=2.9958e-03) (hash(x)=5698642)
321
+ 305 train 7.960882 (lr=2.9953e-03) (hash(x)=6600751)
322
+ 306 train 7.928277 (lr=2.9948e-03) (hash(x)=5804610)
323
+ 307 train 7.884192 (lr=2.9942e-03) (hash(x)=5162652)
324
+ 308 train 7.880470 (lr=2.9937e-03) (hash(x)=5297006)
325
+ 309 train 7.876735 (lr=2.9931e-03) (hash(x)=6728323)
326
+ 310 train 7.859600 (lr=2.9925e-03) (hash(x)=3964882)
327
+ 311 train 7.874573 (lr=2.9918e-03) (hash(x)=6735993)
328
+ 312 train 7.962002 (lr=2.9912e-03) (hash(x)=5150316)
329
+ 313 train 7.806353 (lr=2.9905e-03) (hash(x)=5220448)
330
+ 314 train 7.931576 (lr=2.9898e-03) (hash(x)=6971289)
331
+ 315 train 8.017199 (lr=2.9890e-03) (hash(x)=5012069)
332
+ 316 train 7.867236 (lr=2.9883e-03) (hash(x)=6576136)
333
+ 317 train 7.856455 (lr=2.9875e-03) (hash(x)=6077647)
334
+ 318 train 7.942037 (lr=2.9866e-03) (hash(x)=6032381)
335
+ 319 train 7.814801 (lr=2.9858e-03) (hash(x)=8410716)
336
+ 320 train 7.817859 (lr=2.9849e-03) (hash(x)=6212516)
337
+ 321 train 7.805441 (lr=2.9840e-03) (hash(x)=5298545)
338
+ 322 train 7.876772 (lr=2.9831e-03) (hash(x)=5779073)
339
+ 323 train 7.756114 (lr=2.9821e-03) (hash(x)=6340947)
340
+ 324 train 7.554636 (lr=2.9812e-03) (hash(x)=6530441)
341
+ 325 train 7.835091 (lr=2.9802e-03) (hash(x)=7818007)
342
+ 326 train 7.727715 (lr=2.9791e-03) (hash(x)=6277435)
343
+ 327 train 7.735211 (lr=2.9781e-03) (hash(x)=6226320)
344
+ 328 train 7.802394 (lr=2.9770e-03) (hash(x)=5463751)
345
+ 329 train 7.742763 (lr=2.9759e-03) (hash(x)=6579329)
346
+ 330 train 7.820364 (lr=2.9748e-03) (hash(x)=6660540)
347
+ 331 train 7.773661 (lr=2.9736e-03) (hash(x)=6491272)
348
+ 332 train 7.739844 (lr=2.9724e-03) (hash(x)=6766947)
349
+ 333 train 7.954917 (lr=2.9712e-03) (hash(x)=3721218)
350
+ 334 train 7.847244 (lr=2.9700e-03) (hash(x)=5229898)
351
+ 335 train 7.683785 (lr=2.9687e-03) (hash(x)=6054428)
352
+ 336 train 7.815266 (lr=2.9675e-03) (hash(x)=6290464)
353
+ 337 train 7.784539 (lr=2.9662e-03) (hash(x)=6904498)
354
+ 338 train 7.781158 (lr=2.9648e-03) (hash(x)=7765805)
355
+ 339 train 7.754781 (lr=2.9635e-03) (hash(x)=6543764)
356
+ 340 train 7.774087 (lr=2.9621e-03) (hash(x)=7204955)
357
+ 341 train 7.707437 (lr=2.9607e-03) (hash(x)=4423711)
358
+ 342 train 7.654321 (lr=2.9592e-03) (hash(x)=5343477)
359
+ 343 train 7.770920 (lr=2.9578e-03) (hash(x)=6595856)
360
+ 344 train 7.622881 (lr=2.9563e-03) (hash(x)=7229015)
361
+ 345 train 7.709580 (lr=2.9548e-03) (hash(x)=5916034)
362
+ 346 train 7.871807 (lr=2.9532e-03) (hash(x)=6939521)
363
+ 347 train 7.753299 (lr=2.9517e-03) (hash(x)=5492185)
364
+ 348 train 7.673536 (lr=2.9501e-03) (hash(x)=6898006)
365
+ 349 train 7.606372 (lr=2.9485e-03) (hash(x)=5462008)
366
+ 350 val loss 7.7879
367
+ 350 val perplexity 2411.1887
368
+ 350 train 7.680268 (lr=2.9468e-03) (hash(x)=7821902)
369
+ 351 train 7.778730 (lr=2.9452e-03) (hash(x)=6608108)
370
+ 352 train 7.826403 (lr=2.9435e-03) (hash(x)=6295394)
371
+ 353 train 7.734435 (lr=2.9418e-03) (hash(x)=6803449)
372
+ 354 train 7.758473 (lr=2.9400e-03) (hash(x)=7444134)
373
+ 355 train 7.772284 (lr=2.9383e-03) (hash(x)=5361923)
374
+ 356 train 7.768383 (lr=2.9365e-03) (hash(x)=7803039)
375
+ 357 train 7.884455 (lr=2.9347e-03) (hash(x)=7042578)
376
+ 358 train 7.775435 (lr=2.9328e-03) (hash(x)=6647007)
377
+ 359 train 7.812696 (lr=2.9310e-03) (hash(x)=6303658)
378
+ 360 train 7.936314 (lr=2.9291e-03) (hash(x)=6158927)
379
+ 361 train 7.846681 (lr=2.9272e-03) (hash(x)=6419656)
380
+ 362 train 7.743092 (lr=2.9252e-03) (hash(x)=7891271)
381
+ 363 train 7.946332 (lr=2.9233e-03) (hash(x)=7099888)
382
+ 364 train 7.955026 (lr=2.9213e-03) (hash(x)=6352424)
383
+ 365 train 7.870308 (lr=2.9193e-03) (hash(x)=5863145)
384
+ 366 train 7.895033 (lr=2.9172e-03) (hash(x)=6844597)
385
+ 367 train 7.920383 (lr=2.9152e-03) (hash(x)=6009861)
386
+ 368 train 7.836215 (lr=2.9131e-03) (hash(x)=5374957)
387
+ 369 train 7.867153 (lr=2.9110e-03) (hash(x)=3923225)
388
+ 370 train 7.804662 (lr=2.9088e-03) (hash(x)=5848862)
389
+ 371 train 7.823964 (lr=2.9067e-03) (hash(x)=4440634)
390
+ 372 train 7.794880 (lr=2.9045e-03) (hash(x)=4788998)
391
+ 373 train 7.783244 (lr=2.9023e-03) (hash(x)=7378165)
392
+ 374 train 7.833479 (lr=2.9001e-03) (hash(x)=4771243)
393
+ 375 train 7.818946 (lr=2.8978e-03) (hash(x)=5716089)
394
+ 376 train 7.876677 (lr=2.8955e-03) (hash(x)=7991009)
395
+ 377 train 7.890808 (lr=2.8932e-03) (hash(x)=5348607)
396
+ 378 train 7.959656 (lr=2.8909e-03) (hash(x)=6036146)
397
+ 379 train 7.847301 (lr=2.8885e-03) (hash(x)=6658718)
398
+ 380 train 7.788325 (lr=2.8862e-03) (hash(x)=6176253)
399
+ 381 train 7.789000 (lr=2.8838e-03) (hash(x)=4542895)
400
+ 382 train 7.695456 (lr=2.8813e-03) (hash(x)=5549073)
401
+ 383 train 7.968520 (lr=2.8789e-03) (hash(x)=5342783)
402
+ 384 train 7.990794 (lr=2.8764e-03) (hash(x)=6549546)
403
+ 385 train 7.769759 (lr=2.8739e-03) (hash(x)=6170599)
404
+ 386 train 8.068797 (lr=2.8714e-03) (hash(x)=8483134)
405
+ 387 train 8.129075 (lr=2.8689e-03) (hash(x)=4865036)
406
+ 388 train 7.771400 (lr=2.8663e-03) (hash(x)=4433890)
407
+ 389 train 8.219731 (lr=2.8637e-03) (hash(x)=6738597)
408
+ 390 train 7.858744 (lr=2.8611e-03) (hash(x)=5459620)
409
+ 391 train 7.873519 (lr=2.8585e-03) (hash(x)=4604074)
410
+ 392 train 7.863134 (lr=2.8558e-03) (hash(x)=8045501)
411
+ 393 train 7.972947 (lr=2.8531e-03) (hash(x)=6176464)
412
+ 394 train 7.809046 (lr=2.8504e-03) (hash(x)=6127317)
413
+ 395 train 7.832631 (lr=2.8477e-03) (hash(x)=5571462)
414
+ 396 train 7.839383 (lr=2.8449e-03) (hash(x)=6640755)
415
+ 397 train 7.887265 (lr=2.8422e-03) (hash(x)=6147295)
416
+ 398 train 7.834729 (lr=2.8394e-03) (hash(x)=6120461)
417
+ 399 train 7.792903 (lr=2.8365e-03) (hash(x)=5123928)
418
+ 400 val loss 7.8427
419
+ 400 val perplexity 2546.9512
420
+ 400 train 7.808500 (lr=2.8337e-03) (hash(x)=5428813)
421
+ 401 train 7.777775 (lr=2.8308e-03) (hash(x)=6448586)
422
+ 402 train 7.923082 (lr=2.8279e-03) (hash(x)=8116987)
423
+ 403 train 7.910599 (lr=2.8250e-03) (hash(x)=5974342)
424
+ 404 train 7.949418 (lr=2.8221e-03) (hash(x)=5261336)
425
+ 405 train 7.786650 (lr=2.8191e-03) (hash(x)=5307056)
426
+ 406 train 7.714135 (lr=2.8162e-03) (hash(x)=5865119)
427
+ 407 train 7.761190 (lr=2.8131e-03) (hash(x)=6672380)
428
+ 408 train 7.790997 (lr=2.8101e-03) (hash(x)=6886810)
429
+ 409 train 7.903008 (lr=2.8071e-03) (hash(x)=6564071)
430
+ 410 train 7.696156 (lr=2.8040e-03) (hash(x)=7705089)
431
+ 411 train 7.712930 (lr=2.8009e-03) (hash(x)=5773073)
432
+ 412 train 7.691098 (lr=2.7978e-03) (hash(x)=5905976)
433
+ 413 train 8.183324 (lr=2.7947e-03) (hash(x)=6936653)
434
+ 414 train 7.771364 (lr=2.7915e-03) (hash(x)=6592773)
435
+ 415 train 7.711508 (lr=2.7883e-03) (hash(x)=6796551)
436
+ 416 train 7.779583 (lr=2.7851e-03) (hash(x)=6151589)
437
+ 417 train 8.020397 (lr=2.7819e-03) (hash(x)=7045164)
438
+ 418 train 7.825209 (lr=2.7786e-03) (hash(x)=6650506)
439
+ 419 train 7.744246 (lr=2.7754e-03) (hash(x)=7839748)
440
+ 420 train 7.978945 (lr=2.7721e-03) (hash(x)=6028671)
441
+ 421 train 7.835269 (lr=2.7688e-03) (hash(x)=6118408)
442
+ 422 train 7.782845 (lr=2.7654e-03) (hash(x)=5757526)
443
+ 423 train 7.902639 (lr=2.7621e-03) (hash(x)=6030172)
444
+ 424 train 7.931468 (lr=2.7587e-03) (hash(x)=6719882)
445
+ 425 train 7.797464 (lr=2.7553e-03) (hash(x)=5028050)
446
+ 426 train 7.810927 (lr=2.7519e-03) (hash(x)=5682910)
447
+ 427 train 7.888758 (lr=2.7484e-03) (hash(x)=5747891)
448
+ 428 train 7.702429 (lr=2.7450e-03) (hash(x)=5734208)
449
+ 429 train 7.757227 (lr=2.7415e-03) (hash(x)=7024381)
450
+ 430 train 8.019962 (lr=2.7380e-03) (hash(x)=7814361)
451
+ 431 train 7.888383 (lr=2.7344e-03) (hash(x)=5765893)
452
+ 432 train 7.724426 (lr=2.7309e-03) (hash(x)=5572048)
453
+ 433 train 7.897046 (lr=2.7273e-03) (hash(x)=5610058)
454
+ 434 train 7.758879 (lr=2.7237e-03) (hash(x)=4459483)
455
+ 435 train 7.733801 (lr=2.7201e-03) (hash(x)=6758411)
456
+ 436 train 7.803641 (lr=2.7165e-03) (hash(x)=6701381)
457
+ 437 train 7.948761 (lr=2.7128e-03) (hash(x)=5416400)
458
+ 438 train 7.715144 (lr=2.7092e-03) (hash(x)=5919569)
459
+ 439 train 7.862026 (lr=2.7055e-03) (hash(x)=5282870)
460
+ 440 train 7.710243 (lr=2.7018e-03) (hash(x)=8072181)
461
+ 441 train 7.668911 (lr=2.6980e-03) (hash(x)=7996770)
462
+ 442 train 7.716647 (lr=2.6943e-03) (hash(x)=5093519)
463
+ 443 train 7.734970 (lr=2.6905e-03) (hash(x)=6270917)
464
+ 444 train 7.628579 (lr=2.6867e-03) (hash(x)=5852780)
465
+ 445 train 7.481976 (lr=2.6829e-03) (hash(x)=5327728)
466
+ 446 train 7.801276 (lr=2.6791e-03) (hash(x)=6260765)
467
+ 447 train 7.881089 (lr=2.6752e-03) (hash(x)=6212894)
468
+ 448 train 7.761203 (lr=2.6713e-03) (hash(x)=6724010)
469
+ 449 train 7.826178 (lr=2.6674e-03) (hash(x)=6059934)
470
+ 450 val loss 7.8168
471
+ 450 val perplexity 2481.8940
472
+ 450 train 7.595906 (lr=2.6635e-03) (hash(x)=5587382)
473
+ 451 train 7.748790 (lr=2.6596e-03) (hash(x)=7274108)
474
+ 452 train 7.834761 (lr=2.6556e-03) (hash(x)=6085396)
475
+ 453 train 7.796189 (lr=2.6517e-03) (hash(x)=4784666)
476
+ 454 train 7.889256 (lr=2.6477e-03) (hash(x)=6288638)
477
+ 455 train 7.928413 (lr=2.6437e-03) (hash(x)=7714288)
478
+ 456 train 7.970423 (lr=2.6396e-03) (hash(x)=5534748)
479
+ 457 train 7.831335 (lr=2.6356e-03) (hash(x)=6361270)
480
+ 458 train 7.804205 (lr=2.6315e-03) (hash(x)=4682956)
481
+ 459 train 7.753545 (lr=2.6274e-03) (hash(x)=7568935)
482
+ 460 train 7.880300 (lr=2.6233e-03) (hash(x)=7973311)
483
+ 461 train 7.778728 (lr=2.6192e-03) (hash(x)=6987352)
484
+ 462 train 7.682131 (lr=2.6150e-03) (hash(x)=5413904)
485
+ 463 train 7.885342 (lr=2.6109e-03) (hash(x)=6732924)
486
+ 464 train 7.818769 (lr=2.6067e-03) (hash(x)=6219754)
487
+ 465 train 7.810005 (lr=2.6025e-03) (hash(x)=6839070)
488
+ 466 train 7.770814 (lr=2.5983e-03) (hash(x)=6090120)
489
+ 467 train 8.165805 (lr=2.5940e-03) (hash(x)=7115176)
490
+ 468 train 7.910902 (lr=2.5898e-03) (hash(x)=4989299)
491
+ 469 train 7.800844 (lr=2.5855e-03) (hash(x)=7497590)
492
+ 470 train 7.831113 (lr=2.5812e-03) (hash(x)=3977557)
493
+ 471 train 7.770406 (lr=2.5769e-03) (hash(x)=5524165)
494
+ 472 train 7.764602 (lr=2.5726e-03) (hash(x)=5124936)
495
+ 473 train 7.804563 (lr=2.5682e-03) (hash(x)=5522119)
496
+ 474 train 7.729124 (lr=2.5639e-03) (hash(x)=5030865)
497
+ 475 train 7.820582 (lr=2.5595e-03) (hash(x)=6411783)
498
+ 476 train 7.825943 (lr=2.5551e-03) (hash(x)=6581937)
499
+ 477 train 7.886583 (lr=2.5507e-03) (hash(x)=5116910)
500
+ 478 train 7.730342 (lr=2.5462e-03) (hash(x)=7353459)
501
+ 479 train 7.709627 (lr=2.5418e-03) (hash(x)=6344535)
502
+ 480 train 8.042537 (lr=2.5373e-03) (hash(x)=5840169)
503
+ 481 train 7.724423 (lr=2.5328e-03) (hash(x)=6754290)
504
+ 482 train 7.907392 (lr=2.5283e-03) (hash(x)=6484573)
505
+ 483 train 7.680956 (lr=2.5238e-03) (hash(x)=5260125)
506
+ 484 train 7.805238 (lr=2.5193e-03) (hash(x)=4881179)
507
+ 485 train 7.914628 (lr=2.5147e-03) (hash(x)=5001634)
508
+ 486 train 7.891254 (lr=2.5102e-03) (hash(x)=7424810)
509
+ 487 train 7.736274 (lr=2.5056e-03) (hash(x)=5176431)
510
+ 488 train 7.797306 (lr=2.5010e-03) (hash(x)=5421624)
511
+ 489 train 7.866865 (lr=2.4963e-03) (hash(x)=5674439)
512
+ 490 train 7.757630 (lr=2.4917e-03) (hash(x)=5819157)
513
+ 491 train 7.851449 (lr=2.4871e-03) (hash(x)=6506856)
514
+ 492 train 7.725029 (lr=2.4824e-03) (hash(x)=6119124)
515
+ 493 train 7.899455 (lr=2.4777e-03) (hash(x)=6241669)
516
+ 494 train 7.522427 (lr=2.4730e-03) (hash(x)=4992756)
517
+ 495 train 7.920021 (lr=2.4683e-03) (hash(x)=7277927)
518
+ 496 train 7.895043 (lr=2.4636e-03) (hash(x)=6622097)
519
+ 497 train 7.991194 (lr=2.4588e-03) (hash(x)=7026582)
520
+ 498 train 7.919476 (lr=2.4540e-03) (hash(x)=5840987)
521
+ 499 train 7.619853 (lr=2.4493e-03) (hash(x)=4351074)
522
+ 500 val loss 7.8386
523
+ 500 val perplexity 2536.6829
524
+ 500 train 7.766568 (lr=2.4445e-03) (hash(x)=7301477)
525
+ 501 train 7.952703 (lr=2.4397e-03) (hash(x)=6264167)
526
+ 502 train 7.675928 (lr=2.4348e-03) (hash(x)=4673363)
527
+ 503 train 7.913496 (lr=2.4300e-03) (hash(x)=6332240)
528
+ 504 train 7.824941 (lr=2.4251e-03) (hash(x)=5093790)
529
+ 505 train 7.966659 (lr=2.4203e-03) (hash(x)=6890713)
530
+ 506 train 7.759203 (lr=2.4154e-03) (hash(x)=6042691)
531
+ 507 train 7.892343 (lr=2.4105e-03) (hash(x)=7080746)
532
+ 508 train 7.870537 (lr=2.4056e-03) (hash(x)=7609207)
533
+ 509 train 7.641185 (lr=2.4006e-03) (hash(x)=4591273)
534
+ 510 train 7.792077 (lr=2.3957e-03) (hash(x)=8183111)
535
+ 511 train 8.060661 (lr=2.3907e-03) (hash(x)=5893095)
536
+ 512 train 7.898957 (lr=2.3858e-03) (hash(x)=5709077)
537
+ 513 train 7.840493 (lr=2.3808e-03) (hash(x)=6218715)
538
+ 514 train 7.745891 (lr=2.3758e-03) (hash(x)=4722188)
539
+ 515 train 7.626786 (lr=2.3708e-03) (hash(x)=6313422)
540
+ 516 train 7.768368 (lr=2.3657e-03) (hash(x)=6350660)
541
+ 517 train 7.810089 (lr=2.3607e-03) (hash(x)=5371913)
542
+ 518 train 7.943320 (lr=2.3556e-03) (hash(x)=6283595)
543
+ 519 train 7.939716 (lr=2.3506e-03) (hash(x)=6179167)
544
+ 520 train 7.843018 (lr=2.3455e-03) (hash(x)=7865194)
545
+ 521 train 7.945138 (lr=2.3404e-03) (hash(x)=4883760)
546
+ 522 train 7.761275 (lr=2.3353e-03) (hash(x)=5769286)
547
+ 523 train 7.746145 (lr=2.3301e-03) (hash(x)=6309912)
548
+ 524 train 7.746069 (lr=2.3250e-03) (hash(x)=7490849)
549
+ 525 train 7.726550 (lr=2.3198e-03) (hash(x)=5631134)
550
+ 526 train 7.719683 (lr=2.3147e-03) (hash(x)=4754560)
551
+ 527 train 7.664703 (lr=2.3095e-03) (hash(x)=5448702)
552
+ 528 train 7.864285 (lr=2.3043e-03) (hash(x)=7168842)
553
+ 529 train 7.829534 (lr=2.2991e-03) (hash(x)=7469195)
554
+ 530 train 7.682174 (lr=2.2939e-03) (hash(x)=5479500)
555
+ 531 train 7.846877 (lr=2.2887e-03) (hash(x)=5634177)
556
+ 532 train 7.732557 (lr=2.2834e-03) (hash(x)=9779982)
557
+ 533 train 7.652588 (lr=2.2782e-03) (hash(x)=3641890)
558
+ 534 train 7.728912 (lr=2.2729e-03) (hash(x)=7994103)
559
+ 535 train 7.724516 (lr=2.2676e-03) (hash(x)=5358547)
560
+ 536 train 7.701543 (lr=2.2624e-03) (hash(x)=6797424)
561
+ 537 train 7.835644 (lr=2.2571e-03) (hash(x)=4876580)
562
+ 538 train 7.580977 (lr=2.2517e-03) (hash(x)=4867629)
563
+ 539 train 7.703421 (lr=2.2464e-03) (hash(x)=6601967)
564
+ 540 train 7.670922 (lr=2.2411e-03) (hash(x)=5480657)
565
+ 541 train 7.728854 (lr=2.2357e-03) (hash(x)=7127346)
566
+ 542 train 7.835683 (lr=2.2304e-03) (hash(x)=7224794)
567
+ 543 train 7.733553 (lr=2.2250e-03) (hash(x)=5967252)
568
+ 544 train 7.820445 (lr=2.2196e-03) (hash(x)=7415262)
569
+ 545 train 7.709553 (lr=2.2142e-03) (hash(x)=6522544)
570
+ 546 train 7.742133 (lr=2.2088e-03) (hash(x)=5181640)
571
+ 547 train 7.672265 (lr=2.2034e-03) (hash(x)=5916880)
572
+ 548 train 7.642344 (lr=2.1980e-03) (hash(x)=5620397)
573
+ 549 train 7.705572 (lr=2.1926e-03) (hash(x)=5994970)
574
+ 550 val loss 7.7742
575
+ 550 val perplexity 2378.4238
576
+ 550 train 7.761143 (lr=2.1871e-03) (hash(x)=6532717)
577
+ 551 train 7.757055 (lr=2.1817e-03) (hash(x)=6262762)
578
+ 552 train 7.604557 (lr=2.1762e-03) (hash(x)=5495326)
579
+ 553 train 7.883718 (lr=2.1707e-03) (hash(x)=5927868)
580
+ 554 train 7.692981 (lr=2.1653e-03) (hash(x)=5485433)
581
+ 555 train 7.633148 (lr=2.1598e-03) (hash(x)=6462425)
582
+ 556 train 7.671703 (lr=2.1542e-03) (hash(x)=6299518)
583
+ 557 train 7.744657 (lr=2.1487e-03) (hash(x)=5950376)
584
+ 558 train 7.861241 (lr=2.1432e-03) (hash(x)=6296440)
585
+ 559 train 7.957998 (lr=2.1377e-03) (hash(x)=6974102)
586
+ 560 train 7.732212 (lr=2.1321e-03) (hash(x)=6228679)
587
+ 561 train 7.457102 (lr=2.1266e-03) (hash(x)=6120319)
588
+ 562 train 7.882240 (lr=2.1210e-03) (hash(x)=6289494)
589
+ 563 train 7.710309 (lr=2.1154e-03) (hash(x)=6877818)
590
+ 564 train 7.666038 (lr=2.1099e-03) (hash(x)=6490976)
591
+ 565 train 7.780184 (lr=2.1043e-03) (hash(x)=6419450)
592
+ 566 train 7.766951 (lr=2.0987e-03) (hash(x)=6840857)
593
+ 567 train 7.720932 (lr=2.0931e-03) (hash(x)=4837686)
594
+ 568 train 7.756986 (lr=2.0875e-03) (hash(x)=5561493)
595
+ 569 train 7.665317 (lr=2.0818e-03) (hash(x)=5831242)
596
+ 570 train 7.756404 (lr=2.0762e-03) (hash(x)=6851511)
597
+ 571 train 7.806781 (lr=2.0706e-03) (hash(x)=6169616)
598
+ 572 train 7.730547 (lr=2.0649e-03) (hash(x)=5357151)
599
+ 573 train 7.589833 (lr=2.0593e-03) (hash(x)=5471028)
600
+ 574 train 7.640335 (lr=2.0536e-03) (hash(x)=5051894)
601
+ 575 train 7.634428 (lr=2.0479e-03) (hash(x)=4930450)
602
+ 576 train 7.776700 (lr=2.0422e-03) (hash(x)=6537527)
603
+ 577 train 7.687700 (lr=2.0366e-03) (hash(x)=6330572)
604
+ 578 train 7.819798 (lr=2.0309e-03) (hash(x)=5315155)
605
+ 579 train 7.699385 (lr=2.0252e-03) (hash(x)=5416079)
606
+ 580 train 7.635670 (lr=2.0194e-03) (hash(x)=6255865)
607
+ 581 train 7.900740 (lr=2.0137e-03) (hash(x)=6723036)
608
+ 582 train 7.646069 (lr=2.0080e-03) (hash(x)=7107190)
609
+ 583 train 7.738617 (lr=2.0023e-03) (hash(x)=5868958)
610
+ 584 train 7.825630 (lr=1.9965e-03) (hash(x)=7385923)
611
+ 585 train 7.636406 (lr=1.9908e-03) (hash(x)=5373053)
612
+ 586 train 7.672981 (lr=1.9850e-03) (hash(x)=5371748)
613
+ 587 train 7.816061 (lr=1.9793e-03) (hash(x)=5783717)
614
+ 588 train 7.599503 (lr=1.9735e-03) (hash(x)=6388848)
615
+ 589 train 7.592145 (lr=1.9677e-03) (hash(x)=4865068)
616
+ 590 train 7.549097 (lr=1.9620e-03) (hash(x)=5844961)
617
+ 591 train 7.625145 (lr=1.9562e-03) (hash(x)=5989273)
618
+ 592 train 7.569379 (lr=1.9504e-03) (hash(x)=5136202)
619
+ 593 train 7.691973 (lr=1.9446e-03) (hash(x)=5811889)
620
+ 594 train 7.657429 (lr=1.9388e-03) (hash(x)=6554415)
621
+ 595 train 7.692138 (lr=1.9330e-03) (hash(x)=5969272)
622
+ 596 train 7.656627 (lr=1.9272e-03) (hash(x)=7121518)
623
+ 597 train 8.091704 (lr=1.9214e-03) (hash(x)=6363768)
624
+ 598 train 7.703127 (lr=1.9156e-03) (hash(x)=7030284)
625
+ 599 train 7.796325 (lr=1.9097e-03) (hash(x)=7243339)
626
+ 600 val loss 7.7225
627
+ 600 val perplexity 2258.5503
628
+ 600 train 7.859155 (lr=1.9039e-03) (hash(x)=7199493)
629
+ 601 train 7.923041 (lr=1.8981e-03) (hash(x)=5819281)
630
+ 602 train 7.657584 (lr=1.8922e-03) (hash(x)=7285909)
631
+ 603 train 7.682470 (lr=1.8864e-03) (hash(x)=6730878)
632
+ 604 train 7.727514 (lr=1.8805e-03) (hash(x)=5200141)
633
+ 605 train 7.806885 (lr=1.8747e-03) (hash(x)=5208573)
634
+ 606 train 7.887036 (lr=1.8688e-03) (hash(x)=5976779)
635
+ 607 train 7.717017 (lr=1.8629e-03) (hash(x)=5921767)
636
+ 608 train 7.782152 (lr=1.8571e-03) (hash(x)=6596032)
637
+ 609 train 7.807802 (lr=1.8512e-03) (hash(x)=5428058)
638
+ 610 train 7.725619 (lr=1.8453e-03) (hash(x)=5748050)
639
+ 611 train 7.739799 (lr=1.8395e-03) (hash(x)=6397421)
640
+ 612 train 7.807858 (lr=1.8336e-03) (hash(x)=5638644)
641
+ 613 train 7.729673 (lr=1.8277e-03) (hash(x)=7317623)
642
+ 614 train 7.766802 (lr=1.8218e-03) (hash(x)=5254932)
643
+ 615 train 7.848262 (lr=1.8159e-03) (hash(x)=4945141)
644
+ 616 train 7.762687 (lr=1.8100e-03) (hash(x)=7093994)
645
+ 617 train 7.714802 (lr=1.8041e-03) (hash(x)=7511894)
646
+ 618 train 7.706810 (lr=1.7982e-03) (hash(x)=6706832)
647
+ 619 train 7.643578 (lr=1.7923e-03) (hash(x)=6175371)
648
+ 620 train 7.716118 (lr=1.7864e-03) (hash(x)=7089897)
649
+ 621 train 7.715421 (lr=1.7805e-03) (hash(x)=6828202)
650
+ 622 train 7.988283 (lr=1.7746e-03) (hash(x)=7095301)
651
+ 623 train 7.746866 (lr=1.7686e-03) (hash(x)=5652191)
652
+ 624 train 7.758737 (lr=1.7627e-03) (hash(x)=7336431)
653
+ 625 train 7.763969 (lr=1.7568e-03) (hash(x)=6426339)
654
+ 626 train 7.866049 (lr=1.7509e-03) (hash(x)=5874982)
655
+ 627 train 7.799151 (lr=1.7450e-03) (hash(x)=6978359)
656
+ 628 train 7.537475 (lr=1.7390e-03) (hash(x)=4438173)
657
+ 629 train 7.641233 (lr=1.7331e-03) (hash(x)=6475356)
658
+ 630 train 7.853772 (lr=1.7272e-03) (hash(x)=5030069)
659
+ 631 train 7.629866 (lr=1.7212e-03) (hash(x)=6507939)
660
+ 632 train 7.658139 (lr=1.7153e-03) (hash(x)=6373406)
661
+ 633 train 7.788686 (lr=1.7094e-03) (hash(x)=6623130)
662
+ 634 train 7.704153 (lr=1.7034e-03) (hash(x)=6276741)
663
+ 635 train 7.620919 (lr=1.6975e-03) (hash(x)=7169874)
664
+ 636 train 7.670124 (lr=1.6916e-03) (hash(x)=4947348)
665
+ 637 train 7.762075 (lr=1.6856e-03) (hash(x)=7407715)
666
+ 638 train 7.865242 (lr=1.6797e-03) (hash(x)=5453341)
667
+ 639 train 7.669835 (lr=1.6738e-03) (hash(x)=6520161)
668
+ 640 train 7.740260 (lr=1.6678e-03) (hash(x)=6011882)
669
+ 641 train 7.698287 (lr=1.6619e-03) (hash(x)=5573372)
670
+ 642 train 7.858994 (lr=1.6559e-03) (hash(x)=5344827)
671
+ 643 train 7.921936 (lr=1.6500e-03) (hash(x)=6127887)
672
+ 644 train 7.707982 (lr=1.6441e-03) (hash(x)=6052304)
673
+ 645 train 7.702738 (lr=1.6381e-03) (hash(x)=7039733)
674
+ 646 train 7.757112 (lr=1.6322e-03) (hash(x)=7746451)
675
+ 647 train 7.939248 (lr=1.6262e-03) (hash(x)=7694373)
676
+ 648 train 7.845171 (lr=1.6203e-03) (hash(x)=6911004)
677
+ 649 train 7.770084 (lr=1.6144e-03) (hash(x)=8055224)
678
+ 650 val loss 7.7158
679
+ 650 val perplexity 2243.4919
680
+ 650 train 7.784694 (lr=1.6084e-03) (hash(x)=7060476)
681
+ 651 train 7.750010 (lr=1.6025e-03) (hash(x)=6002317)
682
+ 652 train 7.768577 (lr=1.5966e-03) (hash(x)=6524390)
683
+ 653 train 7.769818 (lr=1.5906e-03) (hash(x)=8331083)
684
+ 654 train 7.798645 (lr=1.5847e-03) (hash(x)=7084852)
685
+ 655 train 7.722343 (lr=1.5788e-03) (hash(x)=6190174)
686
+ 656 train 7.688580 (lr=1.5728e-03) (hash(x)=5601780)
687
+ 657 train 7.841041 (lr=1.5669e-03) (hash(x)=4764034)
688
+ 658 train 7.731352 (lr=1.5610e-03) (hash(x)=6361606)
689
+ 659 train 7.708015 (lr=1.5550e-03) (hash(x)=5697877)
690
+ 660 train 7.775334 (lr=1.5491e-03) (hash(x)=6249793)
691
+ 661 train 7.813062 (lr=1.5432e-03) (hash(x)=7424709)
692
+ 662 train 7.823290 (lr=1.5373e-03) (hash(x)=8203076)
693
+ 663 train 7.582362 (lr=1.5314e-03) (hash(x)=5290019)
694
+ 664 train 7.638449 (lr=1.5254e-03) (hash(x)=6278399)
695
+ 665 train 7.727815 (lr=1.5195e-03) (hash(x)=5709192)
696
+ 666 train 7.797440 (lr=1.5136e-03) (hash(x)=5660292)
697
+ 667 train 7.762671 (lr=1.5077e-03) (hash(x)=5563697)
698
+ 668 train 7.749375 (lr=1.5018e-03) (hash(x)=6325245)
699
+ 669 train 7.759284 (lr=1.4959e-03) (hash(x)=6259741)
700
+ 670 train 7.856074 (lr=1.4900e-03) (hash(x)=6238355)
701
+ 671 train 7.818702 (lr=1.4841e-03) (hash(x)=7778802)
702
+ 672 train 7.775535 (lr=1.4782e-03) (hash(x)=6283496)
703
+ 673 train 7.817242 (lr=1.4723e-03) (hash(x)=5325614)
704
+ 674 train 7.708561 (lr=1.4664e-03) (hash(x)=5376057)
705
+ 675 train 7.679363 (lr=1.4605e-03) (hash(x)=5203215)
706
+ 676 train 7.789243 (lr=1.4547e-03) (hash(x)=4893030)
707
+ 677 train 7.701435 (lr=1.4488e-03) (hash(x)=7595573)
708
+ 678 train 7.788373 (lr=1.4429e-03) (hash(x)=7188575)
709
+ 679 train 7.801357 (lr=1.4371e-03) (hash(x)=6858739)
710
+ 680 train 7.703433 (lr=1.4312e-03) (hash(x)=7157032)
711
+ 681 train 7.806026 (lr=1.4253e-03) (hash(x)=5871492)
712
+ 682 train 7.690859 (lr=1.4195e-03) (hash(x)=5636927)
713
+ 683 train 7.658020 (lr=1.4136e-03) (hash(x)=6135391)
714
+ 684 train 7.759703 (lr=1.4078e-03) (hash(x)=3351432)
715
+ 685 train 7.714324 (lr=1.4019e-03) (hash(x)=6781160)
716
+ 686 train 7.750656 (lr=1.3961e-03) (hash(x)=5824767)
717
+ 687 train 7.763284 (lr=1.3903e-03) (hash(x)=4512723)
718
+ 688 train 7.698975 (lr=1.3844e-03) (hash(x)=4600884)
719
+ 689 train 7.637565 (lr=1.3786e-03) (hash(x)=5079210)
720
+ 690 train 7.689184 (lr=1.3728e-03) (hash(x)=4671963)
721
+ 691 train 7.567908 (lr=1.3670e-03) (hash(x)=4169508)
722
+ 692 train 7.821415 (lr=1.3612e-03) (hash(x)=5341676)
723
+ 693 train 7.611031 (lr=1.3554e-03) (hash(x)=7593391)
724
+ 694 train 7.720570 (lr=1.3496e-03) (hash(x)=6723106)
725
+ 695 train 7.821275 (lr=1.3438e-03) (hash(x)=5136994)
726
+ 696 train 7.790675 (lr=1.3380e-03) (hash(x)=6571636)
727
+ 697 train 7.778792 (lr=1.3323e-03) (hash(x)=6975234)
728
+ 698 train 7.605060 (lr=1.3265e-03) (hash(x)=4868479)
729
+ 699 train 7.939438 (lr=1.3207e-03) (hash(x)=7086593)
730
+ 700 val loss 7.7186
731
+ 700 val perplexity 2249.8350
732
+ 700 train 7.775850 (lr=1.3150e-03) (hash(x)=6870808)
733
+ 701 train 8.031322 (lr=1.3092e-03) (hash(x)=7186398)
734
+ 702 train 7.821193 (lr=1.3035e-03) (hash(x)=4907570)
735
+ 703 train 7.706019 (lr=1.2977e-03) (hash(x)=5812649)
736
+ 704 train 7.735859 (lr=1.2920e-03) (hash(x)=6577447)
737
+ 705 train 7.689717 (lr=1.2863e-03) (hash(x)=5923453)
738
+ 706 train 7.933388 (lr=1.2806e-03) (hash(x)=7282962)
739
+ 707 train 7.640251 (lr=1.2748e-03) (hash(x)=4767725)
740
+ 708 train 7.659239 (lr=1.2691e-03) (hash(x)=6894202)
741
+ 709 train 8.095910 (lr=1.2634e-03) (hash(x)=4904377)
742
+ 710 train 7.692556 (lr=1.2578e-03) (hash(x)=5810443)
743
+ 711 train 7.773018 (lr=1.2521e-03) (hash(x)=5193944)
744
+ 712 train 7.695698 (lr=1.2464e-03) (hash(x)=5188803)
745
+ 713 train 7.711354 (lr=1.2407e-03) (hash(x)=7235549)
746
+ 714 train 7.871559 (lr=1.2351e-03) (hash(x)=5692576)
747
+ 715 train 7.793741 (lr=1.2294e-03) (hash(x)=5103770)
748
+ 716 train 7.753967 (lr=1.2238e-03) (hash(x)=6084786)
749
+ 717 train 7.669781 (lr=1.2182e-03) (hash(x)=5949126)
750
+ 718 train 7.742417 (lr=1.2125e-03) (hash(x)=6321938)
751
+ 719 train 7.785174 (lr=1.2069e-03) (hash(x)=6227413)
752
+ 720 train 7.847672 (lr=1.2013e-03) (hash(x)=6795372)
753
+ 721 train 7.700337 (lr=1.1957e-03) (hash(x)=6506400)
754
+ 722 train 7.737967 (lr=1.1901e-03) (hash(x)=4595742)
755
+ 723 train 7.836090 (lr=1.1846e-03) (hash(x)=7342185)
756
+ 724 train 7.867690 (lr=1.1790e-03) (hash(x)=5762071)
757
+ 725 train 7.741219 (lr=1.1734e-03) (hash(x)=6337670)
758
+ 726 train 7.765738 (lr=1.1679e-03) (hash(x)=6831170)
759
+ 727 train 7.777280 (lr=1.1623e-03) (hash(x)=7080870)
760
+ 728 train 7.723921 (lr=1.1568e-03) (hash(x)=6449970)
761
+ 729 train 7.875694 (lr=1.1513e-03) (hash(x)=5588514)
762
+ 730 train 7.696763 (lr=1.1458e-03) (hash(x)=5379746)
763
+ 731 train 7.846350 (lr=1.1402e-03) (hash(x)=8049901)
764
+ 732 train 7.889677 (lr=1.1347e-03) (hash(x)=8398209)
765
+ 733 train 7.784472 (lr=1.1293e-03) (hash(x)=6152707)
766
+ 734 train 7.844168 (lr=1.1238e-03) (hash(x)=7682617)
767
+ 735 train 7.757251 (lr=1.1183e-03) (hash(x)=7053352)
768
+ 736 train 7.801785 (lr=1.1129e-03) (hash(x)=8491505)
769
+ 737 train 7.769586 (lr=1.1074e-03) (hash(x)=6102853)
770
+ 738 train 7.850826 (lr=1.1020e-03) (hash(x)=4166591)
771
+ 739 train 7.738389 (lr=1.0966e-03) (hash(x)=7501152)
772
+ 740 train 7.707316 (lr=1.0912e-03) (hash(x)=8231949)
773
+ 741 train 7.760527 (lr=1.0858e-03) (hash(x)=5741690)
774
+ 742 train 7.892381 (lr=1.0804e-03) (hash(x)=5420426)
775
+ 743 train 7.688899 (lr=1.0750e-03) (hash(x)=5808766)
776
+ 744 train 7.680545 (lr=1.0696e-03) (hash(x)=7533322)
777
+ 745 train 7.714114 (lr=1.0643e-03) (hash(x)=6350197)
778
+ 746 train 7.816427 (lr=1.0589e-03) (hash(x)=6207170)
779
+ 747 train 7.799877 (lr=1.0536e-03) (hash(x)=5811552)
780
+ 748 train 7.739798 (lr=1.0483e-03) (hash(x)=4540607)
781
+ 749 train 7.624123 (lr=1.0429e-03) (hash(x)=5599396)
782
+ 750 val loss 7.7138
783
+ 750 val perplexity 2238.9255
784
+ 750 train 7.616040 (lr=1.0376e-03) (hash(x)=6180984)
785
+ 751 train 7.785501 (lr=1.0324e-03) (hash(x)=6519152)
786
+ 752 train 7.732035 (lr=1.0271e-03) (hash(x)=9018585)
787
+ 753 train 7.774067 (lr=1.0218e-03) (hash(x)=5988202)
788
+ 754 train 7.800897 (lr=1.0166e-03) (hash(x)=7243449)
789
+ 755 train 7.640830 (lr=1.0113e-03) (hash(x)=4377360)
790
+ 756 train 7.712819 (lr=1.0061e-03) (hash(x)=5928295)
791
+ 757 train 7.673607 (lr=1.0009e-03) (hash(x)=4996189)
792
+ 758 train 7.704133 (lr=9.9568e-04) (hash(x)=6829435)
793
+ 759 train 7.787287 (lr=9.9049e-04) (hash(x)=6929283)
794
+ 760 train 7.829751 (lr=9.8531e-04) (hash(x)=7733662)
795
+ 761 train 7.807028 (lr=9.8015e-04) (hash(x)=5975918)
796
+ 762 train 7.674926 (lr=9.7500e-04) (hash(x)=5566111)
797
+ 763 train 7.750078 (lr=9.6986e-04) (hash(x)=7407511)
798
+ 764 train 7.678582 (lr=9.6474e-04) (hash(x)=5896879)
799
+ 765 train 7.654156 (lr=9.5963e-04) (hash(x)=5569196)
800
+ 766 train 7.832763 (lr=9.5453e-04) (hash(x)=6235218)
801
+ 767 train 7.773887 (lr=9.4944e-04) (hash(x)=7467630)
802
+ 768 train 7.744948 (lr=9.4437e-04) (hash(x)=5268563)
803
+ 769 train 7.938095 (lr=9.3932e-04) (hash(x)=5915215)
804
+ 770 train 7.748833 (lr=9.3427e-04) (hash(x)=5439067)
805
+ 771 train 7.719534 (lr=9.2924e-04) (hash(x)=6167203)
806
+ 772 train 7.780579 (lr=9.2423e-04) (hash(x)=4940526)
807
+ 773 train 7.816148 (lr=9.1923e-04) (hash(x)=5897107)
808
+ 774 train 7.680554 (lr=9.1424e-04) (hash(x)=6780557)
809
+ 775 train 7.772544 (lr=9.0927e-04) (hash(x)=7566733)
810
+ 776 train 7.893176 (lr=9.0431e-04) (hash(x)=7939982)
811
+ 777 train 7.671304 (lr=8.9936e-04) (hash(x)=6447254)
812
+ 778 train 7.661762 (lr=8.9443e-04) (hash(x)=5525831)
813
+ 779 train 7.619434 (lr=8.8952e-04) (hash(x)=5059170)
814
+ 780 train 7.700555 (lr=8.8462e-04) (hash(x)=7485229)
815
+ 781 train 7.798194 (lr=8.7973e-04) (hash(x)=5955219)
816
+ 782 train 7.679614 (lr=8.7486e-04) (hash(x)=6625876)
817
+ 783 train 7.943814 (lr=8.7001e-04) (hash(x)=7141085)
818
+ 784 train 8.030111 (lr=8.6516e-04) (hash(x)=5777157)
819
+ 785 train 7.618348 (lr=8.6034e-04) (hash(x)=6236468)
820
+ 786 train 7.477746 (lr=8.5553e-04) (hash(x)=5381032)
821
+ 787 train 7.654172 (lr=8.5073e-04) (hash(x)=6733716)
822
+ 788 train 7.887829 (lr=8.4596e-04) (hash(x)=7073122)
823
+ 789 train 7.852873 (lr=8.4119e-04) (hash(x)=7622163)
824
+ 790 train 7.901239 (lr=8.3644e-04) (hash(x)=6866834)
825
+ 791 train 7.849960 (lr=8.3171e-04) (hash(x)=7008739)
826
+ 792 train 7.771218 (lr=8.2699e-04) (hash(x)=7839835)
827
+ 793 train 7.672283 (lr=8.2229e-04) (hash(x)=5313637)
828
+ 794 train 7.800224 (lr=8.1761e-04) (hash(x)=6410359)
829
+ 795 train 7.666275 (lr=8.1294e-04) (hash(x)=5721008)
830
+ 796 train 7.677934 (lr=8.0829e-04) (hash(x)=6137936)
831
+ 797 train 7.545530 (lr=8.0365e-04) (hash(x)=6881406)
832
+ 798 train 7.760059 (lr=7.9903e-04) (hash(x)=6088931)
833
+ 799 train 7.772440 (lr=7.9443e-04) (hash(x)=5689728)
834
+ 800 val loss 7.6937
835
+ 800 val perplexity 2194.4966
836
+ 800 train 7.812163 (lr=7.8984e-04) (hash(x)=5800778)
837
+ 801 train 7.725514 (lr=7.8527e-04) (hash(x)=4419021)
838
+ 802 train 7.681883 (lr=7.8072e-04) (hash(x)=5544294)
839
+ 803 train 7.742010 (lr=7.7618e-04) (hash(x)=5663946)
840
+ 804 train 7.672286 (lr=7.7167e-04) (hash(x)=4573911)
841
+ 805 train 7.609970 (lr=7.6716e-04) (hash(x)=5879465)
842
+ 806 train 7.707552 (lr=7.6268e-04) (hash(x)=5641517)
843
+ 807 train 7.851873 (lr=7.5821e-04) (hash(x)=6390336)
844
+ 808 train 7.741605 (lr=7.5376e-04) (hash(x)=8412819)
845
+ 809 train 7.853963 (lr=7.4933e-04) (hash(x)=7005536)
846
+ 810 train 7.606152 (lr=7.4491e-04) (hash(x)=6630048)
847
+ 811 train 7.744143 (lr=7.4051e-04) (hash(x)=7267424)
848
+ 812 train 7.653555 (lr=7.3613e-04) (hash(x)=6497645)
849
+ 813 train 7.596004 (lr=7.3177e-04) (hash(x)=5498697)
850
+ 814 train 7.680041 (lr=7.2742e-04) (hash(x)=6323524)
851
+ 815 train 7.634501 (lr=7.2309e-04) (hash(x)=6529460)
852
+ 816 train 7.515810 (lr=7.1878e-04) (hash(x)=5498587)
853
+ 817 train 7.685587 (lr=7.1449e-04) (hash(x)=6828893)
854
+ 818 train 7.777584 (lr=7.1022e-04) (hash(x)=6591055)
855
+ 819 train 7.645197 (lr=7.0596e-04) (hash(x)=6739572)
856
+ 820 train 7.660881 (lr=7.0173e-04) (hash(x)=5614231)
857
+ 821 train 7.718308 (lr=6.9751e-04) (hash(x)=5462474)
858
+ 822 train 7.705780 (lr=6.9331e-04) (hash(x)=5906742)
859
+ 823 train 7.774216 (lr=6.8913e-04) (hash(x)=8249268)
860
+ 824 train 7.774566 (lr=6.8496e-04) (hash(x)=4886590)
861
+ 825 train 7.556327 (lr=6.8082e-04) (hash(x)=6736150)
862
+ 826 train 7.663374 (lr=6.7669e-04) (hash(x)=6601014)
863
+ 827 train 7.729485 (lr=6.7259e-04) (hash(x)=8283020)
864
+ 828 train 7.687828 (lr=6.6850e-04) (hash(x)=6526568)
865
+ 829 train 7.763347 (lr=6.6443e-04) (hash(x)=6609259)
866
+ 830 train 7.653064 (lr=6.6038e-04) (hash(x)=6079648)
867
+ 831 train 7.633299 (lr=6.5635e-04) (hash(x)=6607178)
868
+ 832 train 7.704588 (lr=6.5234e-04) (hash(x)=6815798)
869
+ 833 train 7.931954 (lr=6.4835e-04) (hash(x)=7277017)
870
+ 834 train 7.641302 (lr=6.4437e-04) (hash(x)=5326415)
871
+ 835 train 7.691921 (lr=6.4042e-04) (hash(x)=4215944)
872
+ 836 train 7.759661 (lr=6.3649e-04) (hash(x)=4689010)
873
+ 837 train 7.726243 (lr=6.3257e-04) (hash(x)=5690469)
874
+ 838 train 7.614465 (lr=6.2868e-04) (hash(x)=6699212)
875
+ 839 train 7.620133 (lr=6.2480e-04) (hash(x)=5494876)
876
+ 840 train 7.986465 (lr=6.2095e-04) (hash(x)=10756054)
877
+ 841 train 7.650000 (lr=6.1711e-04) (hash(x)=6553219)
878
+ 842 train 7.527837 (lr=6.1330e-04) (hash(x)=7323893)
879
+ 843 train 7.951657 (lr=6.0950e-04) (hash(x)=5481970)
880
+ 844 train 7.771655 (lr=6.0573e-04) (hash(x)=6102855)
881
+ 845 train 7.729331 (lr=6.0198e-04) (hash(x)=6889406)
882
+ 846 train 7.769063 (lr=5.9824e-04) (hash(x)=4468013)
883
+ 847 train 7.582155 (lr=5.9453e-04) (hash(x)=6621321)
884
+ 848 train 7.725120 (lr=5.9083e-04) (hash(x)=6185717)
885
+ 849 train 7.615608 (lr=5.8716e-04) (hash(x)=5594828)
886
+ 850 val loss 7.6728
887
+ 850 val perplexity 2149.1121
888
+ 850 train 7.730751 (lr=5.8351e-04) (hash(x)=6530853)
889
+ 851 train 7.666372 (lr=5.7988e-04) (hash(x)=5688551)
890
+ 852 train 7.766419 (lr=5.7627e-04) (hash(x)=6979027)
891
+ 853 train 7.660294 (lr=5.7268e-04) (hash(x)=6065040)
892
+ 854 train 7.616200 (lr=5.6911e-04) (hash(x)=6836310)
893
+ 855 train 7.669339 (lr=5.6556e-04) (hash(x)=4871742)
894
+ 856 train 7.593432 (lr=5.6203e-04) (hash(x)=7461300)
895
+ 857 train 7.729896 (lr=5.5853e-04) (hash(x)=5073610)
896
+ 858 train 7.662956 (lr=5.5504e-04) (hash(x)=6725326)
897
+ 859 train 7.780891 (lr=5.5158e-04) (hash(x)=6114900)
898
+ 860 train 7.751688 (lr=5.4813e-04) (hash(x)=5249155)
899
+ 861 train 7.760153 (lr=5.4471e-04) (hash(x)=8179785)
900
+ 862 train 7.668104 (lr=5.4131e-04) (hash(x)=5588627)
901
+ 863 train 7.682823 (lr=5.3793e-04) (hash(x)=7084451)
902
+ 864 train 7.849386 (lr=5.3458e-04) (hash(x)=7478960)
903
+ 865 train 7.740847 (lr=5.3124e-04) (hash(x)=5609947)
904
+ 866 train 7.720728 (lr=5.2793e-04) (hash(x)=5741827)
905
+ 867 train 7.692734 (lr=5.2464e-04) (hash(x)=3359718)
906
+ 868 train 7.591864 (lr=5.2137e-04) (hash(x)=5480813)
907
+ 869 train 7.673711 (lr=5.1812e-04) (hash(x)=4774534)
908
+ 870 train 7.763359 (lr=5.1489e-04) (hash(x)=5895917)
909
+ 871 train 7.635966 (lr=5.1169e-04) (hash(x)=4902539)
910
+ 872 train 7.695893 (lr=5.0851e-04) (hash(x)=5734618)
911
+ 873 train 7.633891 (lr=5.0534e-04) (hash(x)=7086920)
912
+ 874 train 7.868269 (lr=5.0221e-04) (hash(x)=5338668)
913
+ 875 train 8.062423 (lr=4.9909e-04) (hash(x)=5238889)
914
+ 876 train 7.612788 (lr=4.9600e-04) (hash(x)=6244468)
915
+ 877 train 7.659681 (lr=4.9293e-04) (hash(x)=6715860)
916
+ 878 train 7.802299 (lr=4.8988e-04) (hash(x)=6216047)
917
+ 879 train 7.733390 (lr=4.8685e-04) (hash(x)=6565285)
918
+ 880 train 7.687359 (lr=4.8385e-04) (hash(x)=5257335)
919
+ 881 train 7.690683 (lr=4.8087e-04) (hash(x)=6474071)
920
+ 882 train 7.734727 (lr=4.7791e-04) (hash(x)=5705620)
921
+ 883 train 7.639180 (lr=4.7497e-04) (hash(x)=7193697)
922
+ 884 train 7.725647 (lr=4.7206e-04) (hash(x)=6264039)
923
+ 885 train 7.700440 (lr=4.6917e-04) (hash(x)=6303643)
924
+ 886 train 7.649241 (lr=4.6630e-04) (hash(x)=6120402)
925
+ 887 train 7.667858 (lr=4.6346e-04) (hash(x)=4158764)
926
+ 888 train 7.431095 (lr=4.6063e-04) (hash(x)=6173916)
927
+ 889 train 7.686386 (lr=4.5783e-04) (hash(x)=6211683)
928
+ 890 train 7.719081 (lr=4.5506e-04) (hash(x)=4950548)
929
+ 891 train 7.556942 (lr=4.5231e-04) (hash(x)=4885446)
930
+ 892 train 7.683796 (lr=4.4958e-04) (hash(x)=8117411)
931
+ 893 train 7.706108 (lr=4.4687e-04) (hash(x)=7446434)
932
+ 894 train 7.611644 (lr=4.4419e-04) (hash(x)=5371870)
933
+ 895 train 7.578645 (lr=4.4153e-04) (hash(x)=5022800)
934
+ 896 train 7.581103 (lr=4.3889e-04) (hash(x)=5666758)
935
+ 897 train 7.571211 (lr=4.3628e-04) (hash(x)=5872751)
936
+ 898 train 8.002069 (lr=4.3369e-04) (hash(x)=8428026)
937
+ 899 train 7.826220 (lr=4.3113e-04) (hash(x)=5222699)
938
+ 900 val loss 7.6695
939
+ 900 val perplexity 2142.0947
940
+ 900 train 7.701339 (lr=4.2858e-04) (hash(x)=6048200)
941
+ 901 train 7.565121 (lr=4.2607e-04) (hash(x)=5125556)
942
+ 902 train 7.587252 (lr=4.2357e-04) (hash(x)=6198583)
943
+ 903 train 7.550518 (lr=4.2110e-04) (hash(x)=5112805)
944
+ 904 train 7.634262 (lr=4.1865e-04) (hash(x)=6212803)
945
+ 905 train 7.672472 (lr=4.1623e-04) (hash(x)=5642978)
946
+ 906 train 7.661731 (lr=4.1383e-04) (hash(x)=7307168)
947
+ 907 train 7.601252 (lr=4.1146e-04) (hash(x)=6115590)
948
+ 908 train 7.506181 (lr=4.0910e-04) (hash(x)=7014937)
949
+ 909 train 7.870617 (lr=4.0678e-04) (hash(x)=6374852)
950
+ 910 train 7.586648 (lr=4.0447e-04) (hash(x)=5780051)
951
+ 911 train 7.631597 (lr=4.0220e-04) (hash(x)=4762818)
952
+ 912 train 7.488890 (lr=3.9994e-04) (hash(x)=6585420)
953
+ 913 train 7.778216 (lr=3.9771e-04) (hash(x)=6291590)
954
+ 914 train 7.830007 (lr=3.9550e-04) (hash(x)=5941561)
955
+ 915 train 7.640212 (lr=3.9332e-04) (hash(x)=6824617)
956
+ 916 train 7.727508 (lr=3.9116e-04) (hash(x)=5763341)
957
+ 917 train 7.751686 (lr=3.8903e-04) (hash(x)=6030150)
958
+ 918 train 7.679438 (lr=3.8692e-04) (hash(x)=5325490)
959
+ 919 train 7.692722 (lr=3.8484e-04) (hash(x)=6373236)
960
+ 920 train 7.698028 (lr=3.8277e-04) (hash(x)=6189151)
961
+ 921 train 7.653669 (lr=3.8074e-04) (hash(x)=6645157)
962
+ 922 train 7.665819 (lr=3.7873e-04) (hash(x)=6478394)
963
+ 923 train 7.794346 (lr=3.7674e-04) (hash(x)=5615918)
964
+ 924 train 7.687846 (lr=3.7478e-04) (hash(x)=6376988)
965
+ 925 train 7.719791 (lr=3.7284e-04) (hash(x)=5456350)
966
+ 926 train 7.745311 (lr=3.7093e-04) (hash(x)=5491217)
967
+ 927 train 7.570042 (lr=3.6904e-04) (hash(x)=6871163)
968
+ 928 train 7.587512 (lr=3.6718e-04) (hash(x)=7003708)
969
+ 929 train 7.689504 (lr=3.6534e-04) (hash(x)=7271154)
970
+ 930 train 7.658877 (lr=3.6353e-04) (hash(x)=5378966)
971
+ 931 train 7.650168 (lr=3.6174e-04) (hash(x)=5912082)
972
+ 932 train 7.568612 (lr=3.5998e-04) (hash(x)=5199606)
973
+ 933 train 7.601523 (lr=3.5824e-04) (hash(x)=5195464)
974
+ 934 train 7.557374 (lr=3.5653e-04) (hash(x)=7534052)
975
+ 935 train 7.565880 (lr=3.5484e-04) (hash(x)=6594135)
976
+ 936 train 7.635489 (lr=3.5317e-04) (hash(x)=6782051)
977
+ 937 train 7.752236 (lr=3.5154e-04) (hash(x)=6485577)
978
+ 938 train 7.708436 (lr=3.4992e-04) (hash(x)=6356813)
979
+ 939 train 7.595922 (lr=3.4833e-04) (hash(x)=6903068)
980
+ 940 train 7.667468 (lr=3.4677e-04) (hash(x)=6200862)
981
+ 941 train 7.665900 (lr=3.4523e-04) (hash(x)=5873929)
982
+ 942 train 7.675252 (lr=3.4372e-04) (hash(x)=5624440)
983
+ 943 train 7.640039 (lr=3.4224e-04) (hash(x)=7166972)
984
+ 944 train 7.751742 (lr=3.4077e-04) (hash(x)=6973234)
985
+ 945 train 7.591569 (lr=3.3934e-04) (hash(x)=6626956)
986
+ 946 train 7.713542 (lr=3.3793e-04) (hash(x)=5265592)
987
+ 947 train 7.463261 (lr=3.3654e-04) (hash(x)=5569439)
988
+ 948 train 7.588073 (lr=3.3518e-04) (hash(x)=4981059)
989
+ 949 train 7.604094 (lr=3.3385e-04) (hash(x)=7488974)
990
+ 950 val loss 7.6605
991
+ 950 val perplexity 2122.7214
992
+ 950 train 7.684927 (lr=3.3254e-04) (hash(x)=5160526)
993
+ 951 train 7.607478 (lr=3.3125e-04) (hash(x)=5340877)
994
+ 952 train 7.476114 (lr=3.3000e-04) (hash(x)=6333269)
995
+ 953 train 7.668159 (lr=3.2876e-04) (hash(x)=7205151)
996
+ 954 train 7.710987 (lr=3.2756e-04) (hash(x)=7256545)
997
+ 955 train 7.593724 (lr=3.2638e-04) (hash(x)=6374904)
998
+ 956 train 7.634000 (lr=3.2522e-04) (hash(x)=7124927)
999
+ 957 train 7.603136 (lr=3.2409e-04) (hash(x)=5263132)
1000
+ 958 train 7.782009 (lr=3.2299e-04) (hash(x)=8427208)
1001
+ 959 train 7.591995 (lr=3.2191e-04) (hash(x)=6884232)
1002
+ 960 train 7.573591 (lr=3.2085e-04) (hash(x)=7416977)
1003
+ 961 train 7.695164 (lr=3.1983e-04) (hash(x)=6733050)
1004
+ 962 train 7.605251 (lr=3.1883e-04) (hash(x)=5251404)
1005
+ 963 train 7.724405 (lr=3.1785e-04) (hash(x)=8162997)
1006
+ 964 train 7.855420 (lr=3.1690e-04) (hash(x)=7234382)
1007
+ 965 train 7.762786 (lr=3.1598e-04) (hash(x)=6255275)
1008
+ 966 train 7.716793 (lr=3.1508e-04) (hash(x)=4874742)
1009
+ 967 train 7.686601 (lr=3.1421e-04) (hash(x)=6303037)
1010
+ 968 train 7.606568 (lr=3.1336e-04) (hash(x)=5002696)
1011
+ 969 train 7.642520 (lr=3.1254e-04) (hash(x)=7636268)
1012
+ 970 train 7.625826 (lr=3.1174e-04) (hash(x)=5843590)
1013
+ 971 train 7.705189 (lr=3.1098e-04) (hash(x)=5122289)
1014
+ 972 train 7.590402 (lr=3.1023e-04) (hash(x)=5785819)
1015
+ 973 train 7.657833 (lr=3.0952e-04) (hash(x)=7350343)
1016
+ 974 train 7.480105 (lr=3.0882e-04) (hash(x)=7612943)
1017
+ 975 train 7.733970 (lr=3.0816e-04) (hash(x)=7955067)
1018
+ 976 train 7.639070 (lr=3.0752e-04) (hash(x)=5520886)
1019
+ 977 train 7.529278 (lr=3.0691e-04) (hash(x)=5299200)
1020
+ 978 train 7.557958 (lr=3.0632e-04) (hash(x)=7820550)
1021
+ 979 train 7.528782 (lr=3.0576e-04) (hash(x)=3895312)
1022
+ 980 train 7.646386 (lr=3.0522e-04) (hash(x)=4726896)
1023
+ 981 train 7.438365 (lr=3.0471e-04) (hash(x)=5061167)
1024
+ 982 train 7.612452 (lr=3.0423e-04) (hash(x)=7006804)
1025
+ 983 train 7.620883 (lr=3.0377e-04) (hash(x)=7392600)
1026
+ 984 train 7.818458 (lr=3.0334e-04) (hash(x)=7717172)
1027
+ 985 train 7.952965 (lr=3.0294e-04) (hash(x)=5800587)
1028
+ 986 train 7.637414 (lr=3.0256e-04) (hash(x)=5813784)
1029
+ 987 train 7.612877 (lr=3.0221e-04) (hash(x)=5513267)
1030
+ 988 train 7.597995 (lr=3.0188e-04) (hash(x)=7679326)
1031
+ 989 train 7.669272 (lr=3.0158e-04) (hash(x)=4820278)
1032
+ 990 train 7.683223 (lr=3.0131e-04) (hash(x)=5950845)
1033
+ 991 train 7.599290 (lr=3.0106e-04) (hash(x)=6988067)
1034
+ 992 train 7.530273 (lr=3.0084e-04) (hash(x)=4536923)
1035
+ 993 train 7.666174 (lr=3.0064e-04) (hash(x)=6874134)
1036
+ 994 train 7.902870 (lr=3.0047e-04) (hash(x)=7252013)
1037
+ 995 train 7.560585 (lr=3.0033e-04) (hash(x)=6180446)
1038
+ 996 train 7.783281 (lr=3.0021e-04) (hash(x)=5936769)
1039
+ 997 train 7.711161 (lr=3.0012e-04) (hash(x)=5627313)
1040
+ 998 train 7.666163 (lr=3.0005e-04) (hash(x)=7541743)
1041
+ 999 val loss 7.6568
1042
+ 999 val perplexity 2114.9893
1043
+ 999 train 7.999613 (lr=3.0001e-04) (hash(x)=9812956)
lr3e-3_total_batch_size40960_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cdfee1e08815e8812527e8f7dafc84df27bb1c0861081b96ef09deebdde021b
3
+ size 36163010
lr3e-3_total_batch_size40960_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5483fb3a1ca7826d6435974e949856ceab0c3ca31700e2ed7ee0982e8a4f931
3
+ size 70764358