andrew-healey commited on
Commit
bedbc00
·
verified ·
1 Parent(s): 1933a60

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size10240_seq_len512/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size10240_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", "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": 10240, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0001, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1e-4_total_batch_size10240_seq_len512/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b9fd0bfa27ab267155e5ae2bb7e7ccf3640df9cfec6890472ce93f6d886a52e
3
+ size 964
lr1e-4_total_batch_size10240_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 11.778345 (lr=3.4965e-07) (hash(x)=26375038)
5
+ 1 train 11.759197 (lr=6.9930e-07) (hash(x)=27234506)
6
+ 2 train 11.744032 (lr=1.0490e-06) (hash(x)=23702020)
7
+ 3 train 11.757065 (lr=1.3986e-06) (hash(x)=20782690)
8
+ 4 train 11.786220 (lr=1.7483e-06) (hash(x)=23094976)
9
+ 5 train 11.771325 (lr=2.0979e-06) (hash(x)=26532095)
10
+ 6 train 11.756479 (lr=2.4476e-06) (hash(x)=27151649)
11
+ 7 train 11.764814 (lr=2.7972e-06) (hash(x)=23890908)
12
+ 8 train 11.758749 (lr=3.1469e-06) (hash(x)=25588236)
13
+ 9 train 11.755365 (lr=3.4965e-06) (hash(x)=24011372)
14
+ 10 train 11.743948 (lr=3.8462e-06) (hash(x)=28375581)
15
+ 11 train 11.740856 (lr=4.1958e-06) (hash(x)=24611628)
16
+ 12 train 11.725715 (lr=4.5455e-06) (hash(x)=30298407)
17
+ 13 train 11.714281 (lr=4.8951e-06) (hash(x)=19245352)
18
+ 14 train 11.697369 (lr=5.2448e-06) (hash(x)=28936608)
19
+ 15 train 11.707483 (lr=5.5944e-06) (hash(x)=25767553)
20
+ 16 train 11.693099 (lr=5.9441e-06) (hash(x)=31093473)
21
+ 17 train 11.680680 (lr=6.2937e-06) (hash(x)=24809873)
22
+ 18 train 11.670518 (lr=6.6434e-06) (hash(x)=27852919)
23
+ 19 train 11.644516 (lr=6.9930e-06) (hash(x)=22512166)
24
+ 20 train 11.672844 (lr=7.3427e-06) (hash(x)=27620338)
25
+ 21 train 11.658647 (lr=7.6923e-06) (hash(x)=28092148)
26
+ 22 train 11.619225 (lr=8.0420e-06) (hash(x)=27938767)
27
+ 23 train 11.631838 (lr=8.3916e-06) (hash(x)=24732833)
28
+ 24 train 11.602551 (lr=8.7413e-06) (hash(x)=23200230)
29
+ 25 train 11.601409 (lr=9.0909e-06) (hash(x)=22694718)
30
+ 26 train 11.610548 (lr=9.4406e-06) (hash(x)=23125151)
31
+ 27 train 11.586233 (lr=9.7902e-06) (hash(x)=25129410)
32
+ 28 train 11.572270 (lr=1.0140e-05) (hash(x)=23059154)
33
+ 29 train 11.550849 (lr=1.0490e-05) (hash(x)=27858570)
34
+ 30 train 11.518124 (lr=1.0839e-05) (hash(x)=22402617)
35
+ 31 train 11.506947 (lr=1.1189e-05) (hash(x)=26582391)
36
+ 32 train 11.482800 (lr=1.1538e-05) (hash(x)=26075451)
37
+ 33 train 11.466439 (lr=1.1888e-05) (hash(x)=27279806)
38
+ 34 train 11.473557 (lr=1.2238e-05) (hash(x)=26188136)
39
+ 35 train 11.428185 (lr=1.2587e-05) (hash(x)=26472336)
40
+ 36 train 11.443016 (lr=1.2937e-05) (hash(x)=26376212)
41
+ 37 train 11.427958 (lr=1.3287e-05) (hash(x)=28301589)
42
+ 38 train 11.365438 (lr=1.3636e-05) (hash(x)=28035221)
43
+ 39 train 11.346185 (lr=1.3986e-05) (hash(x)=27654289)
44
+ 40 train 11.380013 (lr=1.4336e-05) (hash(x)=21560904)
45
+ 41 train 11.297602 (lr=1.4685e-05) (hash(x)=24995715)
46
+ 42 train 11.279999 (lr=1.5035e-05) (hash(x)=23792508)
47
+ 43 train 11.270692 (lr=1.5385e-05) (hash(x)=26559876)
48
+ 44 train 11.233850 (lr=1.5734e-05) (hash(x)=25758852)
49
+ 45 train 11.156335 (lr=1.6084e-05) (hash(x)=28364895)
50
+ 46 train 11.154964 (lr=1.6434e-05) (hash(x)=25269299)
51
+ 47 train 11.188103 (lr=1.6783e-05) (hash(x)=21770329)
52
+ 48 train 11.203210 (lr=1.7133e-05) (hash(x)=24785397)
53
+ 49 train 11.098937 (lr=1.7483e-05) (hash(x)=24628606)
54
+ 50 val loss 11.0794
55
+ 50 val perplexity 64819.2617
56
+ 50 train 11.079323 (lr=1.7832e-05) (hash(x)=23181910)
57
+ 51 train 11.032196 (lr=1.8182e-05) (hash(x)=22723459)
58
+ 52 train 11.005144 (lr=1.8531e-05) (hash(x)=25506632)
59
+ 53 train 10.878448 (lr=1.8881e-05) (hash(x)=22897919)
60
+ 54 train 10.932482 (lr=1.9231e-05) (hash(x)=26546719)
61
+ 55 train 11.077948 (lr=1.9580e-05) (hash(x)=30338342)
62
+ 56 train 10.890413 (lr=1.9930e-05) (hash(x)=29175888)
63
+ 57 train 10.917294 (lr=2.0280e-05) (hash(x)=27063280)
64
+ 58 train 10.836851 (lr=2.0629e-05) (hash(x)=32915097)
65
+ 59 train 10.788944 (lr=2.0979e-05) (hash(x)=25678059)
66
+ 60 train 10.729813 (lr=2.1329e-05) (hash(x)=20083773)
67
+ 61 train 10.715242 (lr=2.1678e-05) (hash(x)=21853028)
68
+ 62 train 10.693165 (lr=2.2028e-05) (hash(x)=21808483)
69
+ 63 train 10.728613 (lr=2.2378e-05) (hash(x)=26109335)
70
+ 64 train 10.676506 (lr=2.2727e-05) (hash(x)=24916754)
71
+ 65 train 10.726309 (lr=2.3077e-05) (hash(x)=25334848)
72
+ 66 train 10.694220 (lr=2.3427e-05) (hash(x)=24917705)
73
+ 67 train 10.508959 (lr=2.3776e-05) (hash(x)=25044885)
74
+ 68 train 10.648635 (lr=2.4126e-05) (hash(x)=28747022)
75
+ 69 train 10.562368 (lr=2.4476e-05) (hash(x)=23678349)
76
+ 70 train 10.644717 (lr=2.4825e-05) (hash(x)=28529813)
77
+ 71 train 10.602144 (lr=2.5175e-05) (hash(x)=28870690)
78
+ 72 train 10.548023 (lr=2.5524e-05) (hash(x)=25044834)
79
+ 73 train 10.644623 (lr=2.5874e-05) (hash(x)=32982615)
80
+ 74 train 10.537313 (lr=2.6224e-05) (hash(x)=20970960)
81
+ 75 train 10.640180 (lr=2.6573e-05) (hash(x)=35279941)
82
+ 76 train 10.537445 (lr=2.6923e-05) (hash(x)=25529472)
83
+ 77 train 10.517427 (lr=2.7273e-05) (hash(x)=26400041)
84
+ 78 train 10.425118 (lr=2.7622e-05) (hash(x)=24656138)
85
+ 79 train 10.408859 (lr=2.7972e-05) (hash(x)=25015923)
86
+ 80 train 10.362152 (lr=2.8322e-05) (hash(x)=22924885)
87
+ 81 train 10.421706 (lr=2.8671e-05) (hash(x)=24853586)
88
+ 82 train 10.452950 (lr=2.9021e-05) (hash(x)=27004780)
89
+ 83 train 10.372227 (lr=2.9371e-05) (hash(x)=26740855)
90
+ 84 train 10.261119 (lr=2.9720e-05) (hash(x)=23950114)
91
+ 85 train 10.339935 (lr=3.0070e-05) (hash(x)=24584116)
92
+ 86 train 10.378732 (lr=3.0420e-05) (hash(x)=25718516)
93
+ 87 train 10.335232 (lr=3.0769e-05) (hash(x)=22262151)
94
+ 88 train 10.387228 (lr=3.1119e-05) (hash(x)=23025303)
95
+ 89 train 10.254504 (lr=3.1469e-05) (hash(x)=26627860)
96
+ 90 train 10.259588 (lr=3.1818e-05) (hash(x)=23700532)
97
+ 91 train 10.238726 (lr=3.2168e-05) (hash(x)=26621834)
98
+ 92 train 10.264668 (lr=3.2517e-05) (hash(x)=21821465)
99
+ 93 train 10.334944 (lr=3.2867e-05) (hash(x)=26930630)
100
+ 94 train 10.269863 (lr=3.3217e-05) (hash(x)=19557946)
101
+ 95 train 10.291189 (lr=3.3566e-05) (hash(x)=23234741)
102
+ 96 train 10.318916 (lr=3.3916e-05) (hash(x)=26713563)
103
+ 97 train 10.295780 (lr=3.4266e-05) (hash(x)=28102443)
104
+ 98 train 10.412042 (lr=3.4615e-05) (hash(x)=31689122)
105
+ 99 train 10.392051 (lr=3.4965e-05) (hash(x)=25470563)
106
+ 100 val loss 10.2928
107
+ 100 val perplexity 29517.8477
108
+ 100 train 10.369211 (lr=3.5315e-05) (hash(x)=23757479)
109
+ 101 train 10.217088 (lr=3.5664e-05) (hash(x)=28282861)
110
+ 102 train 10.289894 (lr=3.6014e-05) (hash(x)=26731964)
111
+ 103 train 10.346184 (lr=3.6364e-05) (hash(x)=21799102)
112
+ 104 train 10.238846 (lr=3.6713e-05) (hash(x)=24672077)
113
+ 105 train 10.222236 (lr=3.7063e-05) (hash(x)=23471769)
114
+ 106 train 10.243361 (lr=3.7413e-05) (hash(x)=24058931)
115
+ 107 train 10.222819 (lr=3.7762e-05) (hash(x)=23275479)
116
+ 108 train 10.308283 (lr=3.8112e-05) (hash(x)=24364396)
117
+ 109 train 10.271538 (lr=3.8462e-05) (hash(x)=24473581)
118
+ 110 train 10.267145 (lr=3.8811e-05) (hash(x)=23774644)
119
+ 111 train 10.284931 (lr=3.9161e-05) (hash(x)=22091862)
120
+ 112 train 10.078693 (lr=3.9510e-05) (hash(x)=25482303)
121
+ 113 train 10.243159 (lr=3.9860e-05) (hash(x)=26018202)
122
+ 114 train 10.158385 (lr=4.0210e-05) (hash(x)=24560096)
123
+ 115 train 10.170152 (lr=4.0559e-05) (hash(x)=17430373)
124
+ 116 train 10.030766 (lr=4.0909e-05) (hash(x)=20098681)
125
+ 117 train 10.058788 (lr=4.1259e-05) (hash(x)=24078083)
126
+ 118 train 10.171762 (lr=4.1608e-05) (hash(x)=26595592)
127
+ 119 train 10.109999 (lr=4.1958e-05) (hash(x)=22954861)
128
+ 120 train 10.201167 (lr=4.2308e-05) (hash(x)=29294271)
129
+ 121 train 10.195426 (lr=4.2657e-05) (hash(x)=28695016)
130
+ 122 train 10.131299 (lr=4.3007e-05) (hash(x)=27100115)
131
+ 123 train 10.173187 (lr=4.3357e-05) (hash(x)=26671799)
132
+ 124 train 10.196498 (lr=4.3706e-05) (hash(x)=28137394)
133
+ 125 train 10.164107 (lr=4.4056e-05) (hash(x)=23166092)
134
+ 126 train 10.023652 (lr=4.4406e-05) (hash(x)=25885986)
135
+ 127 train 10.056133 (lr=4.4755e-05) (hash(x)=26924723)
136
+ 128 train 10.120659 (lr=4.5105e-05) (hash(x)=26140590)
137
+ 129 train 10.122700 (lr=4.5455e-05) (hash(x)=25138659)
138
+ 130 train 10.140198 (lr=4.5804e-05) (hash(x)=27224685)
139
+ 131 train 10.107017 (lr=4.6154e-05) (hash(x)=23557495)
140
+ 132 train 10.037539 (lr=4.6503e-05) (hash(x)=23928957)
141
+ 133 train 10.032234 (lr=4.6853e-05) (hash(x)=25288123)
142
+ 134 train 9.966191 (lr=4.7203e-05) (hash(x)=23246294)
143
+ 135 train 9.751441 (lr=4.7552e-05) (hash(x)=27804380)
144
+ 136 train 10.018253 (lr=4.7902e-05) (hash(x)=23089140)
145
+ 137 train 10.144728 (lr=4.8252e-05) (hash(x)=26993263)
146
+ 138 train 10.144464 (lr=4.8601e-05) (hash(x)=27397203)
147
+ 139 train 10.073860 (lr=4.8951e-05) (hash(x)=26654908)
148
+ 140 train 9.989578 (lr=4.9301e-05) (hash(x)=21965022)
149
+ 141 train 9.974422 (lr=4.9650e-05) (hash(x)=27347722)
150
+ 142 train 10.057055 (lr=5.0000e-05) (hash(x)=28314127)
151
+ 143 train 10.056404 (lr=5.0350e-05) (hash(x)=23627518)
152
+ 144 train 9.992710 (lr=5.0699e-05) (hash(x)=25024764)
153
+ 145 train 9.951674 (lr=5.1049e-05) (hash(x)=21768671)
154
+ 146 train 9.991768 (lr=5.1399e-05) (hash(x)=25076667)
155
+ 147 train 9.946714 (lr=5.1748e-05) (hash(x)=21889990)
156
+ 148 train 9.887909 (lr=5.2098e-05) (hash(x)=20932794)
157
+ 149 train 9.894447 (lr=5.2448e-05) (hash(x)=23665838)
158
+ 150 val loss 9.9522
159
+ 150 val perplexity 20997.9062
160
+ 150 train 10.273306 (lr=5.2797e-05) (hash(x)=30223582)
161
+ 151 train 9.919977 (lr=5.3147e-05) (hash(x)=22528001)
162
+ 152 train 9.985833 (lr=5.3497e-05) (hash(x)=25181641)
163
+ 153 train 9.924169 (lr=5.3846e-05) (hash(x)=22487328)
164
+ 154 train 9.982895 (lr=5.4196e-05) (hash(x)=26544630)
165
+ 155 train 10.031384 (lr=5.4545e-05) (hash(x)=29248942)
166
+ 156 train 9.941399 (lr=5.4895e-05) (hash(x)=25052066)
167
+ 157 train 9.846305 (lr=5.5245e-05) (hash(x)=23543273)
168
+ 158 train 9.917454 (lr=5.5594e-05) (hash(x)=27451288)
169
+ 159 train 9.840445 (lr=5.5944e-05) (hash(x)=25533417)
170
+ 160 train 9.880438 (lr=5.6294e-05) (hash(x)=24432899)
171
+ 161 train 9.780600 (lr=5.6643e-05) (hash(x)=25552036)
172
+ 162 train 9.807118 (lr=5.6993e-05) (hash(x)=28149782)
173
+ 163 train 9.775693 (lr=5.7343e-05) (hash(x)=20612533)
174
+ 164 train 9.736376 (lr=5.7692e-05) (hash(x)=19774173)
175
+ 165 train 9.743935 (lr=5.8042e-05) (hash(x)=20216795)
176
+ 166 train 9.897200 (lr=5.8392e-05) (hash(x)=25896435)
177
+ 167 train 9.855928 (lr=5.8741e-05) (hash(x)=24677740)
178
+ 168 train 9.839788 (lr=5.9091e-05) (hash(x)=21835643)
179
+ 169 train 9.751865 (lr=5.9441e-05) (hash(x)=16722715)
180
+ 170 train 9.837046 (lr=5.9790e-05) (hash(x)=26303975)
181
+ 171 train 9.784081 (lr=6.0140e-05) (hash(x)=25347203)
182
+ 172 train 9.766868 (lr=6.0490e-05) (hash(x)=25973417)
183
+ 173 train 9.748303 (lr=6.0839e-05) (hash(x)=24272489)
184
+ 174 train 9.659122 (lr=6.1189e-05) (hash(x)=29260846)
185
+ 175 train 9.739622 (lr=6.1538e-05) (hash(x)=28922363)
186
+ 176 train 9.685648 (lr=6.1888e-05) (hash(x)=26322572)
187
+ 177 train 9.666355 (lr=6.2238e-05) (hash(x)=25510798)
188
+ 178 train 9.733231 (lr=6.2587e-05) (hash(x)=24138948)
189
+ 179 train 9.678607 (lr=6.2937e-05) (hash(x)=25631397)
190
+ 180 train 9.610472 (lr=6.3287e-05) (hash(x)=24001903)
191
+ 181 train 9.601251 (lr=6.3636e-05) (hash(x)=21333676)
192
+ 182 train 9.722335 (lr=6.3986e-05) (hash(x)=30770484)
193
+ 183 train 9.685575 (lr=6.4336e-05) (hash(x)=26237983)
194
+ 184 train 9.619829 (lr=6.4685e-05) (hash(x)=26205744)
195
+ 185 train 9.686939 (lr=6.5035e-05) (hash(x)=25976502)
196
+ 186 train 9.557831 (lr=6.5385e-05) (hash(x)=24080636)
197
+ 187 train 9.755264 (lr=6.5734e-05) (hash(x)=24742645)
198
+ 188 train 9.712208 (lr=6.6084e-05) (hash(x)=28677257)
199
+ 189 train 9.609908 (lr=6.6434e-05) (hash(x)=27086333)
200
+ 190 train 9.590649 (lr=6.6783e-05) (hash(x)=27855233)
201
+ 191 train 9.649851 (lr=6.7133e-05) (hash(x)=26157660)
202
+ 192 train 9.609116 (lr=6.7483e-05) (hash(x)=25046062)
203
+ 193 train 9.603561 (lr=6.7832e-05) (hash(x)=23616370)
204
+ 194 train 9.399410 (lr=6.8182e-05) (hash(x)=24302232)
205
+ 195 train 9.409752 (lr=6.8531e-05) (hash(x)=22582169)
206
+ 196 train 9.487982 (lr=6.8881e-05) (hash(x)=22773833)
207
+ 197 train 9.426292 (lr=6.9231e-05) (hash(x)=26233189)
208
+ 198 train 9.456774 (lr=6.9580e-05) (hash(x)=26431507)
209
+ 199 train 9.581537 (lr=6.9930e-05) (hash(x)=26431960)
210
+ 200 val loss 9.5248
211
+ 200 val perplexity 13695.5342
212
+ 200 train 9.517801 (lr=7.0280e-05) (hash(x)=25112360)
213
+ 201 train 9.644327 (lr=7.0629e-05) (hash(x)=30707498)
214
+ 202 train 9.441256 (lr=7.0979e-05) (hash(x)=22586447)
215
+ 203 train 9.541018 (lr=7.1329e-05) (hash(x)=25716176)
216
+ 204 train 9.554415 (lr=7.1678e-05) (hash(x)=25118971)
217
+ 205 train 9.397882 (lr=7.2028e-05) (hash(x)=26256420)
218
+ 206 train 9.477187 (lr=7.2378e-05) (hash(x)=22633318)
219
+ 207 train 9.487203 (lr=7.2727e-05) (hash(x)=28019788)
220
+ 208 train 9.416048 (lr=7.3077e-05) (hash(x)=27396089)
221
+ 209 train 9.437456 (lr=7.3427e-05) (hash(x)=23510110)
222
+ 210 train 9.490053 (lr=7.3776e-05) (hash(x)=23646926)
223
+ 211 train 9.376534 (lr=7.4126e-05) (hash(x)=20382963)
224
+ 212 train 9.350914 (lr=7.4476e-05) (hash(x)=24304768)
225
+ 213 train 9.273870 (lr=7.4825e-05) (hash(x)=25339466)
226
+ 214 train 9.361347 (lr=7.5175e-05) (hash(x)=22088696)
227
+ 215 train 9.411371 (lr=7.5524e-05) (hash(x)=24489647)
228
+ 216 train 9.305987 (lr=7.5874e-05) (hash(x)=22940357)
229
+ 217 train 9.365932 (lr=7.6224e-05) (hash(x)=22927699)
230
+ 218 train 9.307164 (lr=7.6573e-05) (hash(x)=27631132)
231
+ 219 train 9.324688 (lr=7.6923e-05) (hash(x)=25128502)
232
+ 220 train 9.353636 (lr=7.7273e-05) (hash(x)=23576982)
233
+ 221 train 9.351097 (lr=7.7622e-05) (hash(x)=25808969)
234
+ 222 train 9.301634 (lr=7.7972e-05) (hash(x)=23429962)
235
+ 223 train 9.225149 (lr=7.8322e-05) (hash(x)=25231390)
236
+ 224 train 9.197817 (lr=7.8671e-05) (hash(x)=25395441)
237
+ 225 train 9.177799 (lr=7.9021e-05) (hash(x)=21154388)
238
+ 226 train 9.101061 (lr=7.9371e-05) (hash(x)=24278611)
239
+ 227 train 9.194531 (lr=7.9720e-05) (hash(x)=22708977)
240
+ 228 train 9.436414 (lr=8.0070e-05) (hash(x)=28228490)
241
+ 229 train 9.299029 (lr=8.0420e-05) (hash(x)=27258353)
242
+ 230 train 9.224396 (lr=8.0769e-05) (hash(x)=23252199)
243
+ 231 train 9.151514 (lr=8.1119e-05) (hash(x)=24364920)
244
+ 232 train 9.202085 (lr=8.1469e-05) (hash(x)=23283905)
245
+ 233 train 9.219398 (lr=8.1818e-05) (hash(x)=27243972)
246
+ 234 train 9.339492 (lr=8.2168e-05) (hash(x)=28113043)
247
+ 235 train 9.157563 (lr=8.2517e-05) (hash(x)=24932925)
248
+ 236 train 9.169520 (lr=8.2867e-05) (hash(x)=19001259)
249
+ 237 train 9.235782 (lr=8.3217e-05) (hash(x)=26371091)
250
+ 238 train 9.030504 (lr=8.3566e-05) (hash(x)=24258817)
251
+ 239 train 9.125591 (lr=8.3916e-05) (hash(x)=26913684)
252
+ 240 train 9.166052 (lr=8.4266e-05) (hash(x)=27753043)
253
+ 241 train 9.086603 (lr=8.4615e-05) (hash(x)=29239611)
254
+ 242 train 9.107545 (lr=8.4965e-05) (hash(x)=22984557)
255
+ 243 train 8.805584 (lr=8.5315e-05) (hash(x)=18017792)
256
+ 244 train 9.044850 (lr=8.5664e-05) (hash(x)=23374526)
257
+ 245 train 9.031712 (lr=8.6014e-05) (hash(x)=27574089)
258
+ 246 train 9.082461 (lr=8.6364e-05) (hash(x)=26154906)
259
+ 247 train 9.010483 (lr=8.6713e-05) (hash(x)=28613882)
260
+ 248 train 9.032949 (lr=8.7063e-05) (hash(x)=25272182)
261
+ 249 train 8.988424 (lr=8.7413e-05) (hash(x)=26981914)
262
+ 250 val loss 9.0008
263
+ 250 val perplexity 8109.7710
264
+ 250 train 8.990062 (lr=8.7762e-05) (hash(x)=24232348)
265
+ 251 train 9.054781 (lr=8.8112e-05) (hash(x)=22652243)
266
+ 252 train 9.038454 (lr=8.8462e-05) (hash(x)=25411991)
267
+ 253 train 8.790537 (lr=8.8811e-05) (hash(x)=15446025)
268
+ 254 train 8.872969 (lr=8.9161e-05) (hash(x)=27846176)
269
+ 255 train 8.870403 (lr=8.9510e-05) (hash(x)=27194521)
270
+ 256 train 8.937792 (lr=8.9860e-05) (hash(x)=29430001)
271
+ 257 train 8.670198 (lr=9.0210e-05) (hash(x)=25264518)
272
+ 258 train 8.932689 (lr=9.0559e-05) (hash(x)=25568956)
273
+ 259 train 8.948605 (lr=9.0909e-05) (hash(x)=24936836)
274
+ 260 train 8.929629 (lr=9.1259e-05) (hash(x)=27452099)
275
+ 261 train 8.907574 (lr=9.1608e-05) (hash(x)=28197282)
276
+ 262 train 8.845582 (lr=9.1958e-05) (hash(x)=22931889)
277
+ 263 train 8.879288 (lr=9.2308e-05) (hash(x)=24765578)
278
+ 264 train 8.796483 (lr=9.2657e-05) (hash(x)=25078649)
279
+ 265 train 8.905019 (lr=9.3007e-05) (hash(x)=27448185)
280
+ 266 train 8.918234 (lr=9.3357e-05) (hash(x)=24804856)
281
+ 267 train 8.898142 (lr=9.3706e-05) (hash(x)=23107416)
282
+ 268 train 8.691341 (lr=9.4056e-05) (hash(x)=24960796)
283
+ 269 train 8.845376 (lr=9.4406e-05) (hash(x)=26755138)
284
+ 270 train 8.718391 (lr=9.4755e-05) (hash(x)=24542526)
285
+ 271 train 8.712637 (lr=9.5105e-05) (hash(x)=29314255)
286
+ 272 train 8.639194 (lr=9.5455e-05) (hash(x)=27347755)
287
+ 273 train 8.709740 (lr=9.5804e-05) (hash(x)=22112669)
288
+ 274 train 8.627817 (lr=9.6154e-05) (hash(x)=25161929)
289
+ 275 train 8.733785 (lr=9.6503e-05) (hash(x)=24013079)
290
+ 276 train 8.703789 (lr=9.6853e-05) (hash(x)=26768629)
291
+ 277 train 8.554570 (lr=9.7203e-05) (hash(x)=22537194)
292
+ 278 train 8.568731 (lr=9.7552e-05) (hash(x)=24974360)
293
+ 279 train 8.697974 (lr=9.7902e-05) (hash(x)=25322001)
294
+ 280 train 8.622138 (lr=9.8252e-05) (hash(x)=20562247)
295
+ 281 train 8.796700 (lr=9.8601e-05) (hash(x)=27028126)
296
+ 282 train 8.669106 (lr=9.8951e-05) (hash(x)=25666355)
297
+ 283 train 8.676164 (lr=9.9301e-05) (hash(x)=26690440)
298
+ 284 train 8.580957 (lr=9.9650e-05) (hash(x)=27348418)
299
+ 285 train 8.611712 (lr=1.0000e-04) (hash(x)=26967331)
300
+ 286 train 8.566389 (lr=1.0000e-04) (hash(x)=24765121)
301
+ 287 train 8.557635 (lr=1.0000e-04) (hash(x)=22598512)
302
+ 288 train 8.548866 (lr=9.9998e-05) (hash(x)=23725598)
303
+ 289 train 8.573915 (lr=9.9996e-05) (hash(x)=24697985)
304
+ 290 train 8.520561 (lr=9.9993e-05) (hash(x)=25201962)
305
+ 291 train 8.299683 (lr=9.9989e-05) (hash(x)=18473911)
306
+ 292 train 8.448569 (lr=9.9984e-05) (hash(x)=25982840)
307
+ 293 train 8.496991 (lr=9.9979e-05) (hash(x)=23172124)
308
+ 294 train 8.793334 (lr=9.9972e-05) (hash(x)=34646114)
309
+ 295 train 8.695269 (lr=9.9965e-05) (hash(x)=22322442)
310
+ 296 train 8.521338 (lr=9.9956e-05) (hash(x)=26211794)
311
+ 297 train 8.724833 (lr=9.9947e-05) (hash(x)=29659121)
312
+ 298 train 8.505793 (lr=9.9937e-05) (hash(x)=23538306)
313
+ 299 train 8.459048 (lr=9.9926e-05) (hash(x)=26324153)
314
+ 300 val loss 8.4478
315
+ 300 val perplexity 4665.0020
316
+ 300 train 8.430164 (lr=9.9915e-05) (hash(x)=24910403)
317
+ 301 train 8.463837 (lr=9.9902e-05) (hash(x)=26939970)
318
+ 302 train 8.438341 (lr=9.9889e-05) (hash(x)=27515446)
319
+ 303 train 8.433899 (lr=9.9874e-05) (hash(x)=23446058)
320
+ 304 train 8.540884 (lr=9.9859e-05) (hash(x)=25308253)
321
+ 305 train 8.459402 (lr=9.9843e-05) (hash(x)=29981801)
322
+ 306 train 8.502701 (lr=9.9826e-05) (hash(x)=23470413)
323
+ 307 train 8.508825 (lr=9.9808e-05) (hash(x)=24843741)
324
+ 308 train 8.423241 (lr=9.9789e-05) (hash(x)=27176996)
325
+ 309 train 8.415951 (lr=9.9770e-05) (hash(x)=22855363)
326
+ 310 train 8.254265 (lr=9.9749e-05) (hash(x)=22960957)
327
+ 311 train 8.390139 (lr=9.9728e-05) (hash(x)=23197102)
328
+ 312 train 8.494860 (lr=9.9706e-05) (hash(x)=26132276)
329
+ 313 train 8.336870 (lr=9.9683e-05) (hash(x)=25704919)
330
+ 314 train 8.466761 (lr=9.9659e-05) (hash(x)=22280814)
331
+ 315 train 8.862793 (lr=9.9634e-05) (hash(x)=23054940)
332
+ 316 train 8.786248 (lr=9.9609e-05) (hash(x)=24712416)
333
+ 317 train 8.492140 (lr=9.9582e-05) (hash(x)=24909904)
334
+ 318 train 8.356775 (lr=9.9555e-05) (hash(x)=24321591)
335
+ 319 train 8.169714 (lr=9.9526e-05) (hash(x)=27140757)
336
+ 320 train 8.149944 (lr=9.9497e-05) (hash(x)=26563770)
337
+ 321 train 8.285829 (lr=9.9467e-05) (hash(x)=25993110)
338
+ 322 train 8.277875 (lr=9.9437e-05) (hash(x)=28520222)
339
+ 323 train 8.176364 (lr=9.9405e-05) (hash(x)=25149419)
340
+ 324 train 8.379780 (lr=9.9372e-05) (hash(x)=26112813)
341
+ 325 train 8.328814 (lr=9.9339e-05) (hash(x)=27623643)
342
+ 326 train 8.304392 (lr=9.9305e-05) (hash(x)=29199854)
343
+ 327 train 8.287429 (lr=9.9270e-05) (hash(x)=24727764)
344
+ 328 train 7.916681 (lr=9.9234e-05) (hash(x)=22667179)
345
+ 329 train 7.914082 (lr=9.9197e-05) (hash(x)=23155773)
346
+ 330 train 8.079935 (lr=9.9159e-05) (hash(x)=23453788)
347
+ 331 train 8.178057 (lr=9.9121e-05) (hash(x)=21935931)
348
+ 332 train 8.069702 (lr=9.9081e-05) (hash(x)=24481079)
349
+ 333 train 8.017599 (lr=9.9041e-05) (hash(x)=20054917)
350
+ 334 train 8.200556 (lr=9.9000e-05) (hash(x)=25907741)
351
+ 335 train 8.274841 (lr=9.8958e-05) (hash(x)=27606073)
352
+ 336 train 8.337448 (lr=9.8915e-05) (hash(x)=25202001)
353
+ 337 train 8.271844 (lr=9.8872e-05) (hash(x)=26534487)
354
+ 338 train 8.230125 (lr=9.8827e-05) (hash(x)=24095850)
355
+ 339 train 8.089558 (lr=9.8782e-05) (hash(x)=22450341)
356
+ 340 train 8.399167 (lr=9.8736e-05) (hash(x)=27993763)
357
+ 341 train 8.161674 (lr=9.8689e-05) (hash(x)=25013073)
358
+ 342 train 8.332715 (lr=9.8641e-05) (hash(x)=30934371)
359
+ 343 train 8.449984 (lr=9.8592e-05) (hash(x)=25945859)
360
+ 344 train 8.071671 (lr=9.8543e-05) (hash(x)=25218689)
361
+ 345 train 8.140821 (lr=9.8492e-05) (hash(x)=24418091)
362
+ 346 train 8.079240 (lr=9.8441e-05) (hash(x)=24615466)
363
+ 347 train 8.134432 (lr=9.8389e-05) (hash(x)=21613707)
364
+ 348 train 8.153658 (lr=9.8336e-05) (hash(x)=24348175)
365
+ 349 train 8.162600 (lr=9.8282e-05) (hash(x)=26386157)
366
+ 350 val loss 8.1418
367
+ 350 val perplexity 3435.1836
368
+ 350 train 8.172040 (lr=9.8228e-05) (hash(x)=25771047)
369
+ 351 train 8.109978 (lr=9.8172e-05) (hash(x)=24396519)
370
+ 352 train 8.245684 (lr=9.8116e-05) (hash(x)=25142399)
371
+ 353 train 8.033554 (lr=9.8059e-05) (hash(x)=23265507)
372
+ 354 train 8.065563 (lr=9.8001e-05) (hash(x)=22514858)
373
+ 355 train 8.309155 (lr=9.7942e-05) (hash(x)=29138828)
374
+ 356 train 8.268498 (lr=9.7882e-05) (hash(x)=20531210)
375
+ 357 train 8.077192 (lr=9.7822e-05) (hash(x)=24265353)
376
+ 358 train 8.180635 (lr=9.7761e-05) (hash(x)=24999726)
377
+ 359 train 8.177893 (lr=9.7699e-05) (hash(x)=24685515)
378
+ 360 train 8.023336 (lr=9.7636e-05) (hash(x)=25071701)
379
+ 361 train 8.093445 (lr=9.7572e-05) (hash(x)=24876269)
380
+ 362 train 8.263870 (lr=9.7507e-05) (hash(x)=31733180)
381
+ 363 train 7.867014 (lr=9.7442e-05) (hash(x)=22825749)
382
+ 364 train 8.185184 (lr=9.7376e-05) (hash(x)=26393943)
383
+ 365 train 8.050089 (lr=9.7309e-05) (hash(x)=21239652)
384
+ 366 train 7.929937 (lr=9.7241e-05) (hash(x)=24943881)
385
+ 367 train 8.279860 (lr=9.7172e-05) (hash(x)=28767869)
386
+ 368 train 8.158594 (lr=9.7103e-05) (hash(x)=29341482)
387
+ 369 train 8.096817 (lr=9.7032e-05) (hash(x)=25614301)
388
+ 370 train 8.072987 (lr=9.6961e-05) (hash(x)=26030058)
389
+ 371 train 8.148811 (lr=9.6889e-05) (hash(x)=21159060)
390
+ 372 train 8.019426 (lr=9.6817e-05) (hash(x)=24629937)
391
+ 373 train 8.051239 (lr=9.6743e-05) (hash(x)=26751788)
392
+ 374 train 8.043850 (lr=9.6669e-05) (hash(x)=26012353)
393
+ 375 train 8.003211 (lr=9.6593e-05) (hash(x)=25045397)
394
+ 376 train 7.749918 (lr=9.6518e-05) (hash(x)=23751143)
395
+ 377 train 8.507426 (lr=9.6441e-05) (hash(x)=29989709)
396
+ 378 train 8.149543 (lr=9.6363e-05) (hash(x)=31017246)
397
+ 379 train 8.257023 (lr=9.6285e-05) (hash(x)=25560928)
398
+ 380 train 8.082493 (lr=9.6206e-05) (hash(x)=27352253)
399
+ 381 train 8.196981 (lr=9.6126e-05) (hash(x)=26062687)
400
+ 382 train 8.166460 (lr=9.6045e-05) (hash(x)=27614522)
401
+ 383 train 8.206846 (lr=9.5963e-05) (hash(x)=28959222)
402
+ 384 train 7.981762 (lr=9.5881e-05) (hash(x)=25122598)
403
+ 385 train 7.978647 (lr=9.5798e-05) (hash(x)=25434884)
404
+ 386 train 8.063536 (lr=9.5714e-05) (hash(x)=27824438)
405
+ 387 train 8.058056 (lr=9.5629e-05) (hash(x)=21248405)
406
+ 388 train 7.893837 (lr=9.5544e-05) (hash(x)=26482588)
407
+ 389 train 8.090656 (lr=9.5457e-05) (hash(x)=25044192)
408
+ 390 train 8.102654 (lr=9.5370e-05) (hash(x)=28645524)
409
+ 391 train 7.981835 (lr=9.5282e-05) (hash(x)=25830182)
410
+ 392 train 8.066872 (lr=9.5194e-05) (hash(x)=25585137)
411
+ 393 train 8.017111 (lr=9.5104e-05) (hash(x)=26621419)
412
+ 394 train 7.673712 (lr=9.5014e-05) (hash(x)=22165286)
413
+ 395 train 7.937910 (lr=9.4923e-05) (hash(x)=25173113)
414
+ 396 train 8.015899 (lr=9.4831e-05) (hash(x)=27267091)
415
+ 397 train 7.917826 (lr=9.4739e-05) (hash(x)=24400133)
416
+ 398 train 7.912088 (lr=9.4646e-05) (hash(x)=23103223)
417
+ 399 train 8.003142 (lr=9.4551e-05) (hash(x)=23486277)
418
+ 400 val loss 7.9647
419
+ 400 val perplexity 2877.6982
420
+ 400 train 7.898077 (lr=9.4457e-05) (hash(x)=23421286)
421
+ 401 train 8.014518 (lr=9.4361e-05) (hash(x)=25978130)
422
+ 402 train 7.992720 (lr=9.4265e-05) (hash(x)=25769432)
423
+ 403 train 8.015042 (lr=9.4168e-05) (hash(x)=28916006)
424
+ 404 train 8.014884 (lr=9.4070e-05) (hash(x)=24226056)
425
+ 405 train 8.240057 (lr=9.3971e-05) (hash(x)=33195100)
426
+ 406 train 8.177874 (lr=9.3872e-05) (hash(x)=27730410)
427
+ 407 train 8.036160 (lr=9.3772e-05) (hash(x)=25556929)
428
+ 408 train 7.916506 (lr=9.3671e-05) (hash(x)=25991247)
429
+ 409 train 8.101393 (lr=9.3569e-05) (hash(x)=29536986)
430
+ 410 train 7.962565 (lr=9.3467e-05) (hash(x)=26125216)
431
+ 411 train 8.051405 (lr=9.3364e-05) (hash(x)=29648798)
432
+ 412 train 7.942853 (lr=9.3260e-05) (hash(x)=23195388)
433
+ 413 train 7.938233 (lr=9.3155e-05) (hash(x)=23124767)
434
+ 414 train 7.991045 (lr=9.3050e-05) (hash(x)=25713275)
435
+ 415 train 7.861063 (lr=9.2944e-05) (hash(x)=24976217)
436
+ 416 train 8.049006 (lr=9.2837e-05) (hash(x)=27500801)
437
+ 417 train 7.878024 (lr=9.2729e-05) (hash(x)=23632825)
438
+ 418 train 7.954999 (lr=9.2621e-05) (hash(x)=24456276)
439
+ 419 train 8.110435 (lr=9.2512e-05) (hash(x)=31019606)
440
+ 420 train 8.095550 (lr=9.2402e-05) (hash(x)=25027904)
441
+ 421 train 8.000123 (lr=9.2292e-05) (hash(x)=28236580)
442
+ 422 train 7.923436 (lr=9.2181e-05) (hash(x)=25386473)
443
+ 423 train 7.996744 (lr=9.2069e-05) (hash(x)=28269421)
444
+ 424 train 7.672778 (lr=9.1956e-05) (hash(x)=24308447)
445
+ 425 train 7.883312 (lr=9.1843e-05) (hash(x)=25563279)
446
+ 426 train 8.014158 (lr=9.1729e-05) (hash(x)=26152637)
447
+ 427 train 8.020092 (lr=9.1614e-05) (hash(x)=24865358)
448
+ 428 train 7.809392 (lr=9.1499e-05) (hash(x)=22230964)
449
+ 429 train 7.731553 (lr=9.1382e-05) (hash(x)=22155546)
450
+ 430 train 7.971404 (lr=9.1266e-05) (hash(x)=29142319)
451
+ 431 train 7.781676 (lr=9.1148e-05) (hash(x)=26984272)
452
+ 432 train 7.894748 (lr=9.1030e-05) (hash(x)=27077032)
453
+ 433 train 7.904703 (lr=9.0911e-05) (hash(x)=23193673)
454
+ 434 train 7.957410 (lr=9.0791e-05) (hash(x)=27112558)
455
+ 435 train 7.986318 (lr=9.0671e-05) (hash(x)=24544116)
456
+ 436 train 7.989131 (lr=9.0550e-05) (hash(x)=25890184)
457
+ 437 train 7.839506 (lr=9.0428e-05) (hash(x)=24547533)
458
+ 438 train 7.999065 (lr=9.0306e-05) (hash(x)=27467688)
459
+ 439 train 7.778264 (lr=9.0182e-05) (hash(x)=22059850)
460
+ 440 train 7.992395 (lr=9.0059e-05) (hash(x)=23893130)
461
+ 441 train 7.898538 (lr=8.9934e-05) (hash(x)=25994573)
462
+ 442 train 7.902876 (lr=8.9809e-05) (hash(x)=26577783)
463
+ 443 train 7.895796 (lr=8.9683e-05) (hash(x)=23926632)
464
+ 444 train 7.828486 (lr=8.9557e-05) (hash(x)=23574207)
465
+ 445 train 7.729926 (lr=8.9430e-05) (hash(x)=27111369)
466
+ 446 train 7.941661 (lr=8.9302e-05) (hash(x)=23979820)
467
+ 447 train 7.899097 (lr=8.9173e-05) (hash(x)=27025333)
468
+ 448 train 7.745211 (lr=8.9044e-05) (hash(x)=22846386)
469
+ 449 train 7.694888 (lr=8.8914e-05) (hash(x)=16908068)
470
+ 450 val loss 7.8549
471
+ 450 val perplexity 2578.2766
472
+ 450 train 7.850267 (lr=8.8784e-05) (hash(x)=25664727)
473
+ 451 train 7.899023 (lr=8.8653e-05) (hash(x)=29120407)
474
+ 452 train 7.948782 (lr=8.8521e-05) (hash(x)=25564213)
475
+ 453 train 7.982290 (lr=8.8388e-05) (hash(x)=27092710)
476
+ 454 train 7.846152 (lr=8.8255e-05) (hash(x)=28533197)
477
+ 455 train 7.752515 (lr=8.8122e-05) (hash(x)=23827393)
478
+ 456 train 7.679630 (lr=8.7987e-05) (hash(x)=24154233)
479
+ 457 train 7.931724 (lr=8.7852e-05) (hash(x)=32812727)
480
+ 458 train 7.811368 (lr=8.7717e-05) (hash(x)=26305435)
481
+ 459 train 7.930490 (lr=8.7580e-05) (hash(x)=27230027)
482
+ 460 train 7.787232 (lr=8.7444e-05) (hash(x)=23532437)
483
+ 461 train 7.892492 (lr=8.7306e-05) (hash(x)=25703381)
484
+ 462 train 7.884911 (lr=8.7168e-05) (hash(x)=26961429)
485
+ 463 train 7.707123 (lr=8.7029e-05) (hash(x)=24968260)
486
+ 464 train 7.791509 (lr=8.6890e-05) (hash(x)=24854265)
487
+ 465 train 7.746077 (lr=8.6750e-05) (hash(x)=25676468)
488
+ 466 train 7.868931 (lr=8.6609e-05) (hash(x)=25503865)
489
+ 467 train 7.839814 (lr=8.6468e-05) (hash(x)=27544803)
490
+ 468 train 7.829498 (lr=8.6326e-05) (hash(x)=24658683)
491
+ 469 train 7.842686 (lr=8.6184e-05) (hash(x)=24331407)
492
+ 470 train 7.860165 (lr=8.6041e-05) (hash(x)=21892556)
493
+ 471 train 7.747561 (lr=8.5897e-05) (hash(x)=26540663)
494
+ 472 train 7.771288 (lr=8.5753e-05) (hash(x)=26819462)
495
+ 473 train 7.864779 (lr=8.5608e-05) (hash(x)=25532657)
496
+ 474 train 7.928006 (lr=8.5462e-05) (hash(x)=26515570)
497
+ 475 train 8.046780 (lr=8.5316e-05) (hash(x)=25856625)
498
+ 476 train 7.705584 (lr=8.5170e-05) (hash(x)=25260471)
499
+ 477 train 7.818718 (lr=8.5022e-05) (hash(x)=23437426)
500
+ 478 train 7.920829 (lr=8.4875e-05) (hash(x)=25961833)
501
+ 479 train 7.714629 (lr=8.4726e-05) (hash(x)=23164356)
502
+ 480 train 7.762531 (lr=8.4577e-05) (hash(x)=25052665)
503
+ 481 train 7.852869 (lr=8.4428e-05) (hash(x)=25957896)
504
+ 482 train 7.879270 (lr=8.4278e-05) (hash(x)=24723263)
505
+ 483 train 7.859634 (lr=8.4127e-05) (hash(x)=25646282)
506
+ 484 train 7.842199 (lr=8.3976e-05) (hash(x)=25851993)
507
+ 485 train 7.912834 (lr=8.3824e-05) (hash(x)=21513584)
508
+ 486 train 7.715232 (lr=8.3672e-05) (hash(x)=26017176)
509
+ 487 train 7.871733 (lr=8.3519e-05) (hash(x)=23843233)
510
+ 488 train 7.842238 (lr=8.3366e-05) (hash(x)=21476847)
511
+ 489 train 7.677542 (lr=8.3212e-05) (hash(x)=23226697)
512
+ 490 train 7.451546 (lr=8.3057e-05) (hash(x)=19772969)
513
+ 491 train 7.758326 (lr=8.2902e-05) (hash(x)=24506028)
514
+ 492 train 7.769519 (lr=8.2746e-05) (hash(x)=21077417)
515
+ 493 train 7.749755 (lr=8.2590e-05) (hash(x)=26024321)
516
+ 494 train 7.877829 (lr=8.2434e-05) (hash(x)=27786174)
517
+ 495 train 7.768178 (lr=8.2276e-05) (hash(x)=22781277)
518
+ 496 train 7.831873 (lr=8.2119e-05) (hash(x)=24459895)
519
+ 497 train 7.831679 (lr=8.1960e-05) (hash(x)=24451843)
520
+ 498 train 7.814331 (lr=8.1801e-05) (hash(x)=21116390)
521
+ 499 train 7.728621 (lr=8.1642e-05) (hash(x)=24948650)
522
+ 500 val loss 7.7873
523
+ 500 val perplexity 2409.7715
524
+ 500 train 7.753311 (lr=8.1482e-05) (hash(x)=25617781)
525
+ 501 train 7.697995 (lr=8.1322e-05) (hash(x)=25559534)
526
+ 502 train 7.815330 (lr=8.1161e-05) (hash(x)=25546274)
527
+ 503 train 7.821017 (lr=8.1000e-05) (hash(x)=26373991)
528
+ 504 train 7.759289 (lr=8.0838e-05) (hash(x)=26410662)
529
+ 505 train 7.987728 (lr=8.0676e-05) (hash(x)=27756673)
530
+ 506 train 7.891615 (lr=8.0513e-05) (hash(x)=23661686)
531
+ 507 train 7.762277 (lr=8.0349e-05) (hash(x)=23402396)
532
+ 508 train 7.873063 (lr=8.0186e-05) (hash(x)=26547918)
533
+ 509 train 7.820776 (lr=8.0021e-05) (hash(x)=28250354)
534
+ 510 train 7.845766 (lr=7.9856e-05) (hash(x)=25598861)
535
+ 511 train 7.670006 (lr=7.9691e-05) (hash(x)=21562133)
536
+ 512 train 7.920454 (lr=7.9525e-05) (hash(x)=25312391)
537
+ 513 train 7.931112 (lr=7.9359e-05) (hash(x)=27253861)
538
+ 514 train 7.819691 (lr=7.9192e-05) (hash(x)=26795921)
539
+ 515 train 7.828576 (lr=7.9025e-05) (hash(x)=25009210)
540
+ 516 train 7.749927 (lr=7.8858e-05) (hash(x)=26254538)
541
+ 517 train 7.636095 (lr=7.8689e-05) (hash(x)=23052577)
542
+ 518 train 7.975102 (lr=7.8521e-05) (hash(x)=23667224)
543
+ 519 train 7.779010 (lr=7.8352e-05) (hash(x)=27161811)
544
+ 520 train 7.775244 (lr=7.8182e-05) (hash(x)=26313522)
545
+ 521 train 7.802114 (lr=7.8012e-05) (hash(x)=27302459)
546
+ 522 train 7.938920 (lr=7.7842e-05) (hash(x)=26837963)
547
+ 523 train 7.812513 (lr=7.7671e-05) (hash(x)=29515100)
548
+ 524 train 7.768843 (lr=7.7500e-05) (hash(x)=26674478)
549
+ 525 train 7.716802 (lr=7.7328e-05) (hash(x)=25318001)
550
+ 526 train 7.837920 (lr=7.7156e-05) (hash(x)=25174043)
551
+ 527 train 7.691401 (lr=7.6984e-05) (hash(x)=25305929)
552
+ 528 train 7.879487 (lr=7.6811e-05) (hash(x)=26175477)
553
+ 529 train 7.659453 (lr=7.6637e-05) (hash(x)=24339043)
554
+ 530 train 7.888689 (lr=7.6463e-05) (hash(x)=25412772)
555
+ 531 train 7.739210 (lr=7.6289e-05) (hash(x)=26808543)
556
+ 532 train 7.618253 (lr=7.6115e-05) (hash(x)=24738650)
557
+ 533 train 7.628334 (lr=7.5940e-05) (hash(x)=25763991)
558
+ 534 train 7.905653 (lr=7.5764e-05) (hash(x)=28182190)
559
+ 535 train 7.722682 (lr=7.5588e-05) (hash(x)=25669509)
560
+ 536 train 7.879704 (lr=7.5412e-05) (hash(x)=25616522)
561
+ 537 train 7.836187 (lr=7.5235e-05) (hash(x)=27676869)
562
+ 538 train 7.789387 (lr=7.5058e-05) (hash(x)=26316170)
563
+ 539 train 7.788942 (lr=7.4881e-05) (hash(x)=28435805)
564
+ 540 train 7.758497 (lr=7.4703e-05) (hash(x)=26869022)
565
+ 541 train 7.671610 (lr=7.4525e-05) (hash(x)=23455211)
566
+ 542 train 7.833405 (lr=7.4346e-05) (hash(x)=29321187)
567
+ 543 train 7.896886 (lr=7.4167e-05) (hash(x)=26482758)
568
+ 544 train 7.649776 (lr=7.3988e-05) (hash(x)=26046639)
569
+ 545 train 7.755075 (lr=7.3808e-05) (hash(x)=27068280)
570
+ 546 train 7.683715 (lr=7.3628e-05) (hash(x)=26782091)
571
+ 547 train 7.815842 (lr=7.3448e-05) (hash(x)=24929929)
572
+ 548 train 7.747647 (lr=7.3267e-05) (hash(x)=25950541)
573
+ 549 train 7.677628 (lr=7.3086e-05) (hash(x)=25373676)
574
+ 550 val loss 7.7434
575
+ 550 val perplexity 2306.3184
576
+ 550 train 7.730602 (lr=7.2904e-05) (hash(x)=27283187)
577
+ 551 train 7.709369 (lr=7.2723e-05) (hash(x)=25043037)
578
+ 552 train 7.797846 (lr=7.2540e-05) (hash(x)=27027534)
579
+ 553 train 7.742583 (lr=7.2358e-05) (hash(x)=27906976)
580
+ 554 train 7.743201 (lr=7.2175e-05) (hash(x)=23889709)
581
+ 555 train 8.432260 (lr=7.1992e-05) (hash(x)=29151257)
582
+ 556 train 7.874847 (lr=7.1808e-05) (hash(x)=21307974)
583
+ 557 train 7.779875 (lr=7.1624e-05) (hash(x)=26405613)
584
+ 558 train 7.568102 (lr=7.1440e-05) (hash(x)=24233887)
585
+ 559 train 7.550921 (lr=7.1256e-05) (hash(x)=26400365)
586
+ 560 train 7.816962 (lr=7.1071e-05) (hash(x)=22293114)
587
+ 561 train 7.648985 (lr=7.0886e-05) (hash(x)=23971905)
588
+ 562 train 7.602998 (lr=7.0701e-05) (hash(x)=20076502)
589
+ 563 train 7.731240 (lr=7.0515e-05) (hash(x)=24098812)
590
+ 564 train 7.785452 (lr=7.0329e-05) (hash(x)=27610410)
591
+ 565 train 7.804211 (lr=7.0143e-05) (hash(x)=29402976)
592
+ 566 train 7.471817 (lr=6.9956e-05) (hash(x)=19032564)
593
+ 567 train 7.711669 (lr=6.9769e-05) (hash(x)=20967666)
594
+ 568 train 7.582003 (lr=6.9582e-05) (hash(x)=22666342)
595
+ 569 train 7.758421 (lr=6.9394e-05) (hash(x)=24287610)
596
+ 570 train 7.918939 (lr=6.9207e-05) (hash(x)=28637634)
597
+ 571 train 8.090188 (lr=6.9019e-05) (hash(x)=28667963)
598
+ 572 train 7.871489 (lr=6.8830e-05) (hash(x)=23621685)
599
+ 573 train 7.648467 (lr=6.8642e-05) (hash(x)=25974316)
600
+ 574 train 7.672780 (lr=6.8453e-05) (hash(x)=24487710)
601
+ 575 train 7.964065 (lr=6.8264e-05) (hash(x)=27083417)
602
+ 576 train 7.502601 (lr=6.8075e-05) (hash(x)=22850633)
603
+ 577 train 7.851356 (lr=6.7885e-05) (hash(x)=28605205)
604
+ 578 train 7.648702 (lr=6.7695e-05) (hash(x)=24786468)
605
+ 579 train 7.665360 (lr=6.7505e-05) (hash(x)=23649001)
606
+ 580 train 7.756144 (lr=6.7315e-05) (hash(x)=26473994)
607
+ 581 train 7.546936 (lr=6.7124e-05) (hash(x)=25921700)
608
+ 582 train 7.579129 (lr=6.6933e-05) (hash(x)=24117626)
609
+ 583 train 7.949737 (lr=6.6742e-05) (hash(x)=31338300)
610
+ 584 train 7.620744 (lr=6.6551e-05) (hash(x)=21337692)
611
+ 585 train 7.747930 (lr=6.6360e-05) (hash(x)=23862328)
612
+ 586 train 7.762220 (lr=6.6168e-05) (hash(x)=24380098)
613
+ 587 train 7.634292 (lr=6.5976e-05) (hash(x)=30603174)
614
+ 588 train 7.926414 (lr=6.5784e-05) (hash(x)=30882548)
615
+ 589 train 7.543739 (lr=6.5592e-05) (hash(x)=23603806)
616
+ 590 train 7.716879 (lr=6.5399e-05) (hash(x)=24382442)
617
+ 591 train 7.632149 (lr=6.5206e-05) (hash(x)=24830965)
618
+ 592 train 7.736012 (lr=6.5013e-05) (hash(x)=25547480)
619
+ 593 train 7.808901 (lr=6.4820e-05) (hash(x)=32488729)
620
+ 594 train 8.019091 (lr=6.4627e-05) (hash(x)=31331643)
621
+ 595 train 7.813828 (lr=6.4433e-05) (hash(x)=27269893)
622
+ 596 train 7.630206 (lr=6.4240e-05) (hash(x)=23278725)
623
+ 597 train 7.744902 (lr=6.4046e-05) (hash(x)=25215077)
624
+ 598 train 7.683950 (lr=6.3852e-05) (hash(x)=22442404)
625
+ 599 train 7.632875 (lr=6.3658e-05) (hash(x)=24825041)
626
+ 600 val loss 7.6991
627
+ 600 val perplexity 2206.3130
628
+ 600 train 7.779371 (lr=6.3463e-05) (hash(x)=25590276)
629
+ 601 train 7.699202 (lr=6.3269e-05) (hash(x)=25893804)
630
+ 602 train 7.586247 (lr=6.3074e-05) (hash(x)=22222468)
631
+ 603 train 7.587089 (lr=6.2879e-05) (hash(x)=22724871)
632
+ 604 train 7.615715 (lr=6.2684e-05) (hash(x)=25140744)
633
+ 605 train 7.702605 (lr=6.2489e-05) (hash(x)=27317554)
634
+ 606 train 7.658985 (lr=6.2294e-05) (hash(x)=26692343)
635
+ 607 train 7.472412 (lr=6.2098e-05) (hash(x)=24683717)
636
+ 608 train 7.613105 (lr=6.1903e-05) (hash(x)=26849303)
637
+ 609 train 7.697630 (lr=6.1707e-05) (hash(x)=21584083)
638
+ 610 train 7.819031 (lr=6.1511e-05) (hash(x)=26059939)
639
+ 611 train 7.832165 (lr=6.1315e-05) (hash(x)=26190337)
640
+ 612 train 7.738771 (lr=6.1119e-05) (hash(x)=23663918)
641
+ 613 train 7.650105 (lr=6.0923e-05) (hash(x)=22249019)
642
+ 614 train 7.729262 (lr=6.0726e-05) (hash(x)=25779849)
643
+ 615 train 7.687661 (lr=6.0530e-05) (hash(x)=23086289)
644
+ 616 train 7.813433 (lr=6.0333e-05) (hash(x)=27523941)
645
+ 617 train 7.982520 (lr=6.0137e-05) (hash(x)=28047044)
646
+ 618 train 7.716504 (lr=5.9940e-05) (hash(x)=25125498)
647
+ 619 train 7.569838 (lr=5.9743e-05) (hash(x)=23593875)
648
+ 620 train 7.633795 (lr=5.9546e-05) (hash(x)=25117733)
649
+ 621 train 7.652703 (lr=5.9349e-05) (hash(x)=26059735)
650
+ 622 train 7.631871 (lr=5.9152e-05) (hash(x)=23028679)
651
+ 623 train 7.529600 (lr=5.8955e-05) (hash(x)=22414190)
652
+ 624 train 7.669950 (lr=5.8758e-05) (hash(x)=22223137)
653
+ 625 train 7.823971 (lr=5.8560e-05) (hash(x)=24204591)
654
+ 626 train 7.552416 (lr=5.8363e-05) (hash(x)=22922475)
655
+ 627 train 7.716341 (lr=5.8165e-05) (hash(x)=26758114)
656
+ 628 train 7.607426 (lr=5.7968e-05) (hash(x)=25358064)
657
+ 629 train 7.695556 (lr=5.7770e-05) (hash(x)=20084233)
658
+ 630 train 7.666682 (lr=5.7573e-05) (hash(x)=26395104)
659
+ 631 train 7.493793 (lr=5.7375e-05) (hash(x)=21563184)
660
+ 632 train 7.737299 (lr=5.7177e-05) (hash(x)=25207694)
661
+ 633 train 7.586671 (lr=5.6979e-05) (hash(x)=23861489)
662
+ 634 train 7.829515 (lr=5.6782e-05) (hash(x)=23471007)
663
+ 635 train 7.578623 (lr=5.6584e-05) (hash(x)=19218470)
664
+ 636 train 7.678704 (lr=5.6386e-05) (hash(x)=22944035)
665
+ 637 train 7.729545 (lr=5.6188e-05) (hash(x)=28469562)
666
+ 638 train 7.926088 (lr=5.5990e-05) (hash(x)=26452814)
667
+ 639 train 7.799911 (lr=5.5792e-05) (hash(x)=25588469)
668
+ 640 train 7.666831 (lr=5.5594e-05) (hash(x)=24788298)
669
+ 641 train 7.800921 (lr=5.5396e-05) (hash(x)=25236367)
670
+ 642 train 7.568877 (lr=5.5198e-05) (hash(x)=23069067)
671
+ 643 train 7.938192 (lr=5.5000e-05) (hash(x)=30433767)
672
+ 644 train 8.138870 (lr=5.4802e-05) (hash(x)=29268881)
673
+ 645 train 7.779850 (lr=5.4604e-05) (hash(x)=26302626)
674
+ 646 train 7.897844 (lr=5.4406e-05) (hash(x)=24749682)
675
+ 647 train 7.439370 (lr=5.4208e-05) (hash(x)=17851621)
676
+ 648 train 7.631240 (lr=5.4010e-05) (hash(x)=25620741)
677
+ 649 train 8.140253 (lr=5.3812e-05) (hash(x)=27028191)
678
+ 650 val loss 7.6809
679
+ 650 val perplexity 2166.6021
680
+ 650 train 7.809146 (lr=5.3614e-05) (hash(x)=27590299)
681
+ 651 train 7.908608 (lr=5.3416e-05) (hash(x)=25183690)
682
+ 652 train 7.803892 (lr=5.3218e-05) (hash(x)=31114252)
683
+ 653 train 7.955293 (lr=5.3021e-05) (hash(x)=26182243)
684
+ 654 train 8.035484 (lr=5.2823e-05) (hash(x)=30419908)
685
+ 655 train 7.921749 (lr=5.2625e-05) (hash(x)=27040642)
686
+ 656 train 7.710411 (lr=5.2427e-05) (hash(x)=26240761)
687
+ 657 train 7.732761 (lr=5.2230e-05) (hash(x)=24489607)
688
+ 658 train 7.870761 (lr=5.2032e-05) (hash(x)=25012872)
689
+ 659 train 7.435538 (lr=5.1835e-05) (hash(x)=22183303)
690
+ 660 train 7.731437 (lr=5.1637e-05) (hash(x)=27682633)
691
+ 661 train 7.806808 (lr=5.1440e-05) (hash(x)=28844646)
692
+ 662 train 8.378026 (lr=5.1242e-05) (hash(x)=31082070)
693
+ 663 train 8.013935 (lr=5.1045e-05) (hash(x)=27537063)
694
+ 664 train 7.716995 (lr=5.0848e-05) (hash(x)=26125988)
695
+ 665 train 7.811551 (lr=5.0651e-05) (hash(x)=25446686)
696
+ 666 train 7.818001 (lr=5.0454e-05) (hash(x)=19819857)
697
+ 667 train 7.469242 (lr=5.0257e-05) (hash(x)=19492832)
698
+ 668 train 7.654526 (lr=5.0060e-05) (hash(x)=23741214)
699
+ 669 train 7.656859 (lr=4.9863e-05) (hash(x)=25484958)
700
+ 670 train 7.647113 (lr=4.9667e-05) (hash(x)=25525370)
701
+ 671 train 7.776371 (lr=4.9470e-05) (hash(x)=25951629)
702
+ 672 train 8.036814 (lr=4.9274e-05) (hash(x)=28557663)
703
+ 673 train 7.524099 (lr=4.9077e-05) (hash(x)=23462798)
704
+ 674 train 7.510583 (lr=4.8881e-05) (hash(x)=27320280)
705
+ 675 train 7.778380 (lr=4.8685e-05) (hash(x)=22408682)
706
+ 676 train 7.564774 (lr=4.8489e-05) (hash(x)=22230799)
707
+ 677 train 7.997296 (lr=4.8293e-05) (hash(x)=31317970)
708
+ 678 train 7.575196 (lr=4.8097e-05) (hash(x)=24020983)
709
+ 679 train 7.621708 (lr=4.7902e-05) (hash(x)=27633457)
710
+ 680 train 7.502813 (lr=4.7706e-05) (hash(x)=23766987)
711
+ 681 train 7.669285 (lr=4.7511e-05) (hash(x)=23729283)
712
+ 682 train 7.844398 (lr=4.7316e-05) (hash(x)=25975834)
713
+ 683 train 7.736680 (lr=4.7121e-05) (hash(x)=27635080)
714
+ 684 train 7.714621 (lr=4.6926e-05) (hash(x)=27181156)
715
+ 685 train 7.757821 (lr=4.6731e-05) (hash(x)=25918780)
716
+ 686 train 7.726954 (lr=4.6537e-05) (hash(x)=22370417)
717
+ 687 train 7.568861 (lr=4.6342e-05) (hash(x)=22287596)
718
+ 688 train 7.579237 (lr=4.6148e-05) (hash(x)=21584419)
719
+ 689 train 7.631186 (lr=4.5954e-05) (hash(x)=22694623)
720
+ 690 train 7.762285 (lr=4.5760e-05) (hash(x)=30820846)
721
+ 691 train 8.260849 (lr=4.5567e-05) (hash(x)=35895440)
722
+ 692 train 7.769084 (lr=4.5373e-05) (hash(x)=27058657)
723
+ 693 train 7.669861 (lr=4.5180e-05) (hash(x)=24298886)
724
+ 694 train 7.697334 (lr=4.4987e-05) (hash(x)=27303061)
725
+ 695 train 7.742071 (lr=4.4794e-05) (hash(x)=25888847)
726
+ 696 train 7.602031 (lr=4.4601e-05) (hash(x)=21267641)
727
+ 697 train 8.008384 (lr=4.4408e-05) (hash(x)=30860646)
728
+ 698 train 7.687947 (lr=4.4216e-05) (hash(x)=26881837)
729
+ 699 train 7.782543 (lr=4.4024e-05) (hash(x)=26518800)
730
+ 700 val loss 7.6735
731
+ 700 val perplexity 2150.6499
732
+ 700 train 7.787914 (lr=4.3832e-05) (hash(x)=25863277)
733
+ 701 train 7.571774 (lr=4.3640e-05) (hash(x)=25385523)
734
+ 702 train 7.617644 (lr=4.3449e-05) (hash(x)=23362519)
735
+ 703 train 8.035971 (lr=4.3258e-05) (hash(x)=37519283)
736
+ 704 train 8.073067 (lr=4.3067e-05) (hash(x)=23392584)
737
+ 705 train 8.100092 (lr=4.2876e-05) (hash(x)=20564397)
738
+ 706 train 7.787432 (lr=4.2685e-05) (hash(x)=27065499)
739
+ 707 train 7.688785 (lr=4.2495e-05) (hash(x)=25141945)
740
+ 708 train 7.841002 (lr=4.2305e-05) (hash(x)=25466598)
741
+ 709 train 7.631757 (lr=4.2115e-05) (hash(x)=22717866)
742
+ 710 train 8.228502 (lr=4.1925e-05) (hash(x)=29648992)
743
+ 711 train 7.967737 (lr=4.1736e-05) (hash(x)=28527939)
744
+ 712 train 7.602393 (lr=4.1547e-05) (hash(x)=25134784)
745
+ 713 train 8.101931 (lr=4.1358e-05) (hash(x)=32757059)
746
+ 714 train 7.602934 (lr=4.1170e-05) (hash(x)=24145729)
747
+ 715 train 7.574712 (lr=4.0981e-05) (hash(x)=21942471)
748
+ 716 train 7.719856 (lr=4.0793e-05) (hash(x)=25325444)
749
+ 717 train 8.002458 (lr=4.0606e-05) (hash(x)=24247339)
750
+ 718 train 7.700689 (lr=4.0418e-05) (hash(x)=24142989)
751
+ 719 train 7.662145 (lr=4.0231e-05) (hash(x)=24955630)
752
+ 720 train 7.725366 (lr=4.0044e-05) (hash(x)=25491335)
753
+ 721 train 7.824761 (lr=3.9857e-05) (hash(x)=21368780)
754
+ 722 train 7.746728 (lr=3.9671e-05) (hash(x)=27146278)
755
+ 723 train 7.610024 (lr=3.9485e-05) (hash(x)=25541230)
756
+ 724 train 7.650712 (lr=3.9299e-05) (hash(x)=25375355)
757
+ 725 train 7.530919 (lr=3.9114e-05) (hash(x)=23098806)
758
+ 726 train 7.537552 (lr=3.8929e-05) (hash(x)=22061174)
759
+ 727 train 7.553737 (lr=3.8744e-05) (hash(x)=23501481)
760
+ 728 train 7.749593 (lr=3.8560e-05) (hash(x)=24656430)
761
+ 729 train 7.493769 (lr=3.8376e-05) (hash(x)=23802681)
762
+ 730 train 7.569107 (lr=3.8192e-05) (hash(x)=23496118)
763
+ 731 train 7.717530 (lr=3.8008e-05) (hash(x)=22754988)
764
+ 732 train 7.457503 (lr=3.7825e-05) (hash(x)=21461650)
765
+ 733 train 7.524594 (lr=3.7642e-05) (hash(x)=22183009)
766
+ 734 train 7.631665 (lr=3.7460e-05) (hash(x)=22300616)
767
+ 735 train 7.426915 (lr=3.7277e-05) (hash(x)=23475070)
768
+ 736 train 7.649411 (lr=3.7096e-05) (hash(x)=24508407)
769
+ 737 train 7.707459 (lr=3.6914e-05) (hash(x)=23359061)
770
+ 738 train 7.616834 (lr=3.6733e-05) (hash(x)=22588667)
771
+ 739 train 7.714833 (lr=3.6552e-05) (hash(x)=23609959)
772
+ 740 train 7.682714 (lr=3.6372e-05) (hash(x)=25394852)
773
+ 741 train 7.523766 (lr=3.6192e-05) (hash(x)=22408449)
774
+ 742 train 7.675414 (lr=3.6012e-05) (hash(x)=26947398)
775
+ 743 train 7.414140 (lr=3.5833e-05) (hash(x)=23441847)
776
+ 744 train 7.537478 (lr=3.5654e-05) (hash(x)=23101350)
777
+ 745 train 7.550110 (lr=3.5475e-05) (hash(x)=23676365)
778
+ 746 train 7.719032 (lr=3.5297e-05) (hash(x)=26488811)
779
+ 747 train 7.750334 (lr=3.5119e-05) (hash(x)=23096292)
780
+ 748 train 7.768732 (lr=3.4942e-05) (hash(x)=24512380)
781
+ 749 train 7.696436 (lr=3.4765e-05) (hash(x)=27706294)
782
+ 750 val loss 7.6445
783
+ 750 val perplexity 2089.1770
784
+ 750 train 7.622089 (lr=3.4588e-05) (hash(x)=21892472)
785
+ 751 train 7.614200 (lr=3.4412e-05) (hash(x)=26171093)
786
+ 752 train 7.624597 (lr=3.4236e-05) (hash(x)=23191101)
787
+ 753 train 8.019724 (lr=3.4060e-05) (hash(x)=34946941)
788
+ 754 train 7.825703 (lr=3.3885e-05) (hash(x)=25569942)
789
+ 755 train 7.372855 (lr=3.3711e-05) (hash(x)=20060838)
790
+ 756 train 7.478687 (lr=3.3537e-05) (hash(x)=24581020)
791
+ 757 train 7.482943 (lr=3.3363e-05) (hash(x)=21591702)
792
+ 758 train 7.521377 (lr=3.3189e-05) (hash(x)=21870012)
793
+ 759 train 7.635319 (lr=3.3016e-05) (hash(x)=25451055)
794
+ 760 train 7.383495 (lr=3.2844e-05) (hash(x)=18989695)
795
+ 761 train 7.058500 (lr=3.2672e-05) (hash(x)=18691455)
796
+ 762 train 7.462986 (lr=3.2500e-05) (hash(x)=22406503)
797
+ 763 train 7.654760 (lr=3.2329e-05) (hash(x)=24934792)
798
+ 764 train 7.542640 (lr=3.2158e-05) (hash(x)=24249677)
799
+ 765 train 7.446667 (lr=3.1988e-05) (hash(x)=23864361)
800
+ 766 train 7.445629 (lr=3.1818e-05) (hash(x)=21549554)
801
+ 767 train 7.625589 (lr=3.1648e-05) (hash(x)=25367597)
802
+ 768 train 7.796278 (lr=3.1479e-05) (hash(x)=24409290)
803
+ 769 train 7.565759 (lr=3.1311e-05) (hash(x)=27016702)
804
+ 770 train 7.751408 (lr=3.1142e-05) (hash(x)=25520453)
805
+ 771 train 7.568425 (lr=3.0975e-05) (hash(x)=25022148)
806
+ 772 train 7.745041 (lr=3.0808e-05) (hash(x)=25010319)
807
+ 773 train 7.615832 (lr=3.0641e-05) (hash(x)=24459012)
808
+ 774 train 7.653831 (lr=3.0475e-05) (hash(x)=26528670)
809
+ 775 train 7.645749 (lr=3.0309e-05) (hash(x)=25431365)
810
+ 776 train 7.502379 (lr=3.0144e-05) (hash(x)=24229522)
811
+ 777 train 7.695449 (lr=2.9979e-05) (hash(x)=25958785)
812
+ 778 train 7.592598 (lr=2.9814e-05) (hash(x)=22882640)
813
+ 779 train 7.957471 (lr=2.9651e-05) (hash(x)=29307788)
814
+ 780 train 7.521873 (lr=2.9487e-05) (hash(x)=23167825)
815
+ 781 train 7.594881 (lr=2.9324e-05) (hash(x)=24965435)
816
+ 782 train 7.455843 (lr=2.9162e-05) (hash(x)=23772949)
817
+ 783 train 7.670658 (lr=2.9000e-05) (hash(x)=25651929)
818
+ 784 train 7.601382 (lr=2.8839e-05) (hash(x)=23146724)
819
+ 785 train 7.645479 (lr=2.8678e-05) (hash(x)=28451031)
820
+ 786 train 7.663624 (lr=2.8518e-05) (hash(x)=23796771)
821
+ 787 train 7.685363 (lr=2.8358e-05) (hash(x)=26312387)
822
+ 788 train 7.568400 (lr=2.8199e-05) (hash(x)=21257377)
823
+ 789 train 7.723933 (lr=2.8040e-05) (hash(x)=25326650)
824
+ 790 train 7.604388 (lr=2.7881e-05) (hash(x)=24605603)
825
+ 791 train 7.553473 (lr=2.7724e-05) (hash(x)=23804913)
826
+ 792 train 7.763256 (lr=2.7566e-05) (hash(x)=24322926)
827
+ 793 train 7.762107 (lr=2.7410e-05) (hash(x)=25537529)
828
+ 794 train 7.481017 (lr=2.7254e-05) (hash(x)=25707197)
829
+ 795 train 7.670818 (lr=2.7098e-05) (hash(x)=26115519)
830
+ 796 train 7.592053 (lr=2.6943e-05) (hash(x)=24850654)
831
+ 797 train 7.551762 (lr=2.6788e-05) (hash(x)=25494804)
832
+ 798 train 7.586142 (lr=2.6634e-05) (hash(x)=21972022)
833
+ 799 train 7.866730 (lr=2.6481e-05) (hash(x)=23210747)
834
+ 800 val loss 7.6386
835
+ 800 val perplexity 2076.7825
836
+ 800 train 7.643985 (lr=2.6328e-05) (hash(x)=17064773)
837
+ 801 train 7.603842 (lr=2.6176e-05) (hash(x)=15717069)
838
+ 802 train 7.684797 (lr=2.6024e-05) (hash(x)=16189636)
839
+ 803 train 7.482850 (lr=2.5873e-05) (hash(x)=18463712)
840
+ 804 train 7.635897 (lr=2.5722e-05) (hash(x)=19970072)
841
+ 805 train 7.703615 (lr=2.5572e-05) (hash(x)=20705573)
842
+ 806 train 7.567783 (lr=2.5423e-05) (hash(x)=22913147)
843
+ 807 train 7.639965 (lr=2.5274e-05) (hash(x)=23245699)
844
+ 808 train 7.626671 (lr=2.5125e-05) (hash(x)=27121904)
845
+ 809 train 7.527898 (lr=2.4978e-05) (hash(x)=23444521)
846
+ 810 train 7.475498 (lr=2.4830e-05) (hash(x)=20917937)
847
+ 811 train 7.536948 (lr=2.4684e-05) (hash(x)=21770300)
848
+ 812 train 8.614824 (lr=2.4538e-05) (hash(x)=24642519)
849
+ 813 train 7.791303 (lr=2.4392e-05) (hash(x)=23334569)
850
+ 814 train 7.396803 (lr=2.4247e-05) (hash(x)=21796200)
851
+ 815 train 7.431859 (lr=2.4103e-05) (hash(x)=25387532)
852
+ 816 train 7.342999 (lr=2.3959e-05) (hash(x)=22203733)
853
+ 817 train 7.934273 (lr=2.3816e-05) (hash(x)=29466743)
854
+ 818 train 7.626596 (lr=2.3674e-05) (hash(x)=21756102)
855
+ 819 train 7.667036 (lr=2.3532e-05) (hash(x)=27718703)
856
+ 820 train 7.641782 (lr=2.3391e-05) (hash(x)=22978681)
857
+ 821 train 7.690335 (lr=2.3250e-05) (hash(x)=17631973)
858
+ 822 train 7.754877 (lr=2.3110e-05) (hash(x)=27947891)
859
+ 823 train 7.662114 (lr=2.2971e-05) (hash(x)=27661052)
860
+ 824 train 7.445637 (lr=2.2832e-05) (hash(x)=23061452)
861
+ 825 train 7.539428 (lr=2.2694e-05) (hash(x)=22244096)
862
+ 826 train 7.569288 (lr=2.2556e-05) (hash(x)=25389713)
863
+ 827 train 7.671076 (lr=2.2420e-05) (hash(x)=26538928)
864
+ 828 train 7.745315 (lr=2.2283e-05) (hash(x)=25685569)
865
+ 829 train 7.498966 (lr=2.2148e-05) (hash(x)=26256802)
866
+ 830 train 7.799363 (lr=2.2013e-05) (hash(x)=24412323)
867
+ 831 train 7.659465 (lr=2.1878e-05) (hash(x)=21772654)
868
+ 832 train 7.541817 (lr=2.1745e-05) (hash(x)=25456825)
869
+ 833 train 7.881545 (lr=2.1612e-05) (hash(x)=25026277)
870
+ 834 train 7.800159 (lr=2.1479e-05) (hash(x)=26855439)
871
+ 835 train 7.799935 (lr=2.1347e-05) (hash(x)=26263254)
872
+ 836 train 7.821570 (lr=2.1216e-05) (hash(x)=21917396)
873
+ 837 train 7.614340 (lr=2.1086e-05) (hash(x)=24726314)
874
+ 838 train 7.523699 (lr=2.0956e-05) (hash(x)=24375100)
875
+ 839 train 7.551291 (lr=2.0827e-05) (hash(x)=23552356)
876
+ 840 train 7.658912 (lr=2.0698e-05) (hash(x)=25343112)
877
+ 841 train 7.615695 (lr=2.0570e-05) (hash(x)=26699819)
878
+ 842 train 7.907065 (lr=2.0443e-05) (hash(x)=25218858)
879
+ 843 train 8.049483 (lr=2.0317e-05) (hash(x)=22707141)
880
+ 844 train 7.934984 (lr=2.0191e-05) (hash(x)=29168269)
881
+ 845 train 8.029676 (lr=2.0066e-05) (hash(x)=25676149)
882
+ 846 train 7.573329 (lr=1.9941e-05) (hash(x)=24324401)
883
+ 847 train 7.902899 (lr=1.9818e-05) (hash(x)=34234481)
884
+ 848 train 7.622798 (lr=1.9694e-05) (hash(x)=24511386)
885
+ 849 train 7.629669 (lr=1.9572e-05) (hash(x)=25401624)
886
+ 850 val loss 7.6406
887
+ 850 val perplexity 2080.9448
888
+ 850 train 7.605271 (lr=1.9450e-05) (hash(x)=22340685)
889
+ 851 train 7.933499 (lr=1.9329e-05) (hash(x)=35008670)
890
+ 852 train 7.937935 (lr=1.9209e-05) (hash(x)=29662286)
891
+ 853 train 7.828029 (lr=1.9089e-05) (hash(x)=25682280)
892
+ 854 train 7.813427 (lr=1.8970e-05) (hash(x)=27686498)
893
+ 855 train 7.640196 (lr=1.8852e-05) (hash(x)=25891543)
894
+ 856 train 7.586551 (lr=1.8734e-05) (hash(x)=22538881)
895
+ 857 train 7.651596 (lr=1.8618e-05) (hash(x)=25962407)
896
+ 858 train 7.573223 (lr=1.8501e-05) (hash(x)=24819073)
897
+ 859 train 7.471816 (lr=1.8386e-05) (hash(x)=21971224)
898
+ 860 train 7.443999 (lr=1.8271e-05) (hash(x)=20951317)
899
+ 861 train 7.870314 (lr=1.8157e-05) (hash(x)=30838142)
900
+ 862 train 7.708407 (lr=1.8044e-05) (hash(x)=23102419)
901
+ 863 train 7.700505 (lr=1.7931e-05) (hash(x)=27312186)
902
+ 864 train 8.104451 (lr=1.7819e-05) (hash(x)=30920030)
903
+ 865 train 7.512696 (lr=1.7708e-05) (hash(x)=23657081)
904
+ 866 train 8.013491 (lr=1.7598e-05) (hash(x)=35351214)
905
+ 867 train 7.506133 (lr=1.7488e-05) (hash(x)=22163749)
906
+ 868 train 7.376035 (lr=1.7379e-05) (hash(x)=24629160)
907
+ 869 train 7.493510 (lr=1.7271e-05) (hash(x)=24829818)
908
+ 870 train 7.833752 (lr=1.7163e-05) (hash(x)=25424550)
909
+ 871 train 7.607903 (lr=1.7056e-05) (hash(x)=25942888)
910
+ 872 train 7.725047 (lr=1.6950e-05) (hash(x)=26418501)
911
+ 873 train 7.602732 (lr=1.6845e-05) (hash(x)=23785671)
912
+ 874 train 7.751019 (lr=1.6740e-05) (hash(x)=25097859)
913
+ 875 train 7.584238 (lr=1.6636e-05) (hash(x)=25493916)
914
+ 876 train 7.777052 (lr=1.6533e-05) (hash(x)=19819210)
915
+ 877 train 7.859134 (lr=1.6431e-05) (hash(x)=28618702)
916
+ 878 train 8.025672 (lr=1.6329e-05) (hash(x)=29707907)
917
+ 879 train 7.738074 (lr=1.6228e-05) (hash(x)=24289951)
918
+ 880 train 7.579609 (lr=1.6128e-05) (hash(x)=25147026)
919
+ 881 train 7.699232 (lr=1.6029e-05) (hash(x)=26331486)
920
+ 882 train 7.672817 (lr=1.5930e-05) (hash(x)=25885126)
921
+ 883 train 7.596989 (lr=1.5832e-05) (hash(x)=23648671)
922
+ 884 train 7.445858 (lr=1.5735e-05) (hash(x)=20252079)
923
+ 885 train 7.710497 (lr=1.5639e-05) (hash(x)=27292797)
924
+ 886 train 7.262682 (lr=1.5543e-05) (hash(x)=18317379)
925
+ 887 train 7.594328 (lr=1.5449e-05) (hash(x)=24992055)
926
+ 888 train 7.974067 (lr=1.5354e-05) (hash(x)=24734221)
927
+ 889 train 7.826250 (lr=1.5261e-05) (hash(x)=24398330)
928
+ 890 train 7.597431 (lr=1.5169e-05) (hash(x)=23573984)
929
+ 891 train 7.701459 (lr=1.5077e-05) (hash(x)=26005549)
930
+ 892 train 7.491080 (lr=1.4986e-05) (hash(x)=22583878)
931
+ 893 train 7.778744 (lr=1.4896e-05) (hash(x)=30259737)
932
+ 894 train 7.710117 (lr=1.4806e-05) (hash(x)=25196470)
933
+ 895 train 7.547873 (lr=1.4718e-05) (hash(x)=21046554)
934
+ 896 train 7.741303 (lr=1.4630e-05) (hash(x)=25454371)
935
+ 897 train 7.525836 (lr=1.4543e-05) (hash(x)=18773877)
936
+ 898 train 7.548968 (lr=1.4456e-05) (hash(x)=24645168)
937
+ 899 train 7.727813 (lr=1.4371e-05) (hash(x)=26321539)
938
+ 900 val loss 7.6293
939
+ 900 val perplexity 2057.5811
940
+ 900 train 7.650787 (lr=1.4286e-05) (hash(x)=25722713)
941
+ 901 train 7.556557 (lr=1.4202e-05) (hash(x)=22850804)
942
+ 902 train 7.671810 (lr=1.4119e-05) (hash(x)=22503524)
943
+ 903 train 7.696234 (lr=1.4037e-05) (hash(x)=27807872)
944
+ 904 train 7.704907 (lr=1.3955e-05) (hash(x)=25845590)
945
+ 905 train 7.595258 (lr=1.3874e-05) (hash(x)=25545137)
946
+ 906 train 7.750066 (lr=1.3794e-05) (hash(x)=25471231)
947
+ 907 train 7.544093 (lr=1.3715e-05) (hash(x)=21555364)
948
+ 908 train 7.656128 (lr=1.3637e-05) (hash(x)=24474988)
949
+ 909 train 7.708959 (lr=1.3559e-05) (hash(x)=27243798)
950
+ 910 train 7.579549 (lr=1.3482e-05) (hash(x)=24419085)
951
+ 911 train 7.551946 (lr=1.3407e-05) (hash(x)=26325324)
952
+ 912 train 7.501114 (lr=1.3331e-05) (hash(x)=24822111)
953
+ 913 train 7.563728 (lr=1.3257e-05) (hash(x)=26228728)
954
+ 914 train 7.762024 (lr=1.3183e-05) (hash(x)=24707042)
955
+ 915 train 7.777285 (lr=1.3111e-05) (hash(x)=24956294)
956
+ 916 train 7.440916 (lr=1.3039e-05) (hash(x)=23424737)
957
+ 917 train 7.817878 (lr=1.2968e-05) (hash(x)=25592289)
958
+ 918 train 7.586237 (lr=1.2897e-05) (hash(x)=26826109)
959
+ 919 train 7.616658 (lr=1.2828e-05) (hash(x)=24429389)
960
+ 920 train 7.587954 (lr=1.2759e-05) (hash(x)=24747421)
961
+ 921 train 7.654936 (lr=1.2691e-05) (hash(x)=28733708)
962
+ 922 train 7.738257 (lr=1.2624e-05) (hash(x)=29704803)
963
+ 923 train 7.923301 (lr=1.2558e-05) (hash(x)=27929846)
964
+ 924 train 7.667874 (lr=1.2493e-05) (hash(x)=18961171)
965
+ 925 train 7.620592 (lr=1.2428e-05) (hash(x)=22417476)
966
+ 926 train 7.683595 (lr=1.2364e-05) (hash(x)=26364120)
967
+ 927 train 7.743367 (lr=1.2301e-05) (hash(x)=25282897)
968
+ 928 train 7.543608 (lr=1.2239e-05) (hash(x)=22082503)
969
+ 929 train 7.503415 (lr=1.2178e-05) (hash(x)=22278062)
970
+ 930 train 7.561129 (lr=1.2118e-05) (hash(x)=24271468)
971
+ 931 train 7.501224 (lr=1.2058e-05) (hash(x)=26802508)
972
+ 932 train 8.200258 (lr=1.1999e-05) (hash(x)=33468379)
973
+ 933 train 7.395924 (lr=1.1941e-05) (hash(x)=22094832)
974
+ 934 train 7.582695 (lr=1.1884e-05) (hash(x)=26759850)
975
+ 935 train 7.769766 (lr=1.1828e-05) (hash(x)=28390416)
976
+ 936 train 7.871329 (lr=1.1772e-05) (hash(x)=22727956)
977
+ 937 train 7.807997 (lr=1.1718e-05) (hash(x)=26379331)
978
+ 938 train 7.952809 (lr=1.1664e-05) (hash(x)=25048158)
979
+ 939 train 7.814912 (lr=1.1611e-05) (hash(x)=26292451)
980
+ 940 train 7.667743 (lr=1.1559e-05) (hash(x)=23336031)
981
+ 941 train 7.846426 (lr=1.1508e-05) (hash(x)=26301866)
982
+ 942 train 7.994064 (lr=1.1457e-05) (hash(x)=26687040)
983
+ 943 train 7.546209 (lr=1.1408e-05) (hash(x)=22915344)
984
+ 944 train 7.571156 (lr=1.1359e-05) (hash(x)=19831900)
985
+ 945 train 7.672693 (lr=1.1311e-05) (hash(x)=25448318)
986
+ 946 train 7.633951 (lr=1.1264e-05) (hash(x)=24993116)
987
+ 947 train 7.813816 (lr=1.1218e-05) (hash(x)=24744383)
988
+ 948 train 7.562504 (lr=1.1173e-05) (hash(x)=23731849)
989
+ 949 train 7.509165 (lr=1.1128e-05) (hash(x)=22835201)
990
+ 950 val loss 7.6353
991
+ 950 val perplexity 2070.0051
992
+ 950 train 7.495546 (lr=1.1085e-05) (hash(x)=24652361)
993
+ 951 train 7.731914 (lr=1.1042e-05) (hash(x)=26284678)
994
+ 952 train 7.655455 (lr=1.1000e-05) (hash(x)=27522881)
995
+ 953 train 7.620192 (lr=1.0959e-05) (hash(x)=24745936)
996
+ 954 train 7.479528 (lr=1.0919e-05) (hash(x)=21973969)
997
+ 955 train 7.478342 (lr=1.0879e-05) (hash(x)=21654269)
998
+ 956 train 7.497247 (lr=1.0841e-05) (hash(x)=21708739)
999
+ 957 train 7.853351 (lr=1.0803e-05) (hash(x)=32387999)
1000
+ 958 train 8.322100 (lr=1.0766e-05) (hash(x)=32662682)
1001
+ 959 train 8.309869 (lr=1.0730e-05) (hash(x)=29466268)
1002
+ 960 train 8.248133 (lr=1.0695e-05) (hash(x)=28372973)
1003
+ 961 train 8.277311 (lr=1.0661e-05) (hash(x)=30736727)
1004
+ 962 train 8.227760 (lr=1.0628e-05) (hash(x)=29935562)
1005
+ 963 train 8.268118 (lr=1.0595e-05) (hash(x)=32616762)
1006
+ 964 train 8.214703 (lr=1.0563e-05) (hash(x)=32861403)
1007
+ 965 train 8.246642 (lr=1.0533e-05) (hash(x)=29358578)
1008
+ 966 train 8.272852 (lr=1.0503e-05) (hash(x)=31218966)
1009
+ 967 train 8.345358 (lr=1.0474e-05) (hash(x)=32201138)
1010
+ 968 train 8.241787 (lr=1.0445e-05) (hash(x)=30545133)
1011
+ 969 train 8.452784 (lr=1.0418e-05) (hash(x)=30041621)
1012
+ 970 train 8.246517 (lr=1.0391e-05) (hash(x)=30371023)
1013
+ 971 train 8.412689 (lr=1.0366e-05) (hash(x)=33683468)
1014
+ 972 train 8.312199 (lr=1.0341e-05) (hash(x)=29809830)
1015
+ 973 train 7.724546 (lr=1.0317e-05) (hash(x)=26837755)
1016
+ 974 train 7.711532 (lr=1.0294e-05) (hash(x)=26084089)
1017
+ 975 train 7.841079 (lr=1.0272e-05) (hash(x)=27498125)
1018
+ 976 train 8.094208 (lr=1.0251e-05) (hash(x)=34478395)
1019
+ 977 train 7.736024 (lr=1.0230e-05) (hash(x)=22372507)
1020
+ 978 train 7.669850 (lr=1.0211e-05) (hash(x)=25638879)
1021
+ 979 train 7.523422 (lr=1.0192e-05) (hash(x)=25164318)
1022
+ 980 train 7.698132 (lr=1.0174e-05) (hash(x)=27271900)
1023
+ 981 train 7.613483 (lr=1.0157e-05) (hash(x)=26562947)
1024
+ 982 train 7.641259 (lr=1.0141e-05) (hash(x)=24917192)
1025
+ 983 train 7.553860 (lr=1.0126e-05) (hash(x)=25690221)
1026
+ 984 train 7.796426 (lr=1.0111e-05) (hash(x)=24807841)
1027
+ 985 train 7.533139 (lr=1.0098e-05) (hash(x)=21947305)
1028
+ 986 train 7.543126 (lr=1.0085e-05) (hash(x)=22195089)
1029
+ 987 train 7.663280 (lr=1.0074e-05) (hash(x)=29031550)
1030
+ 988 train 7.773809 (lr=1.0063e-05) (hash(x)=29395823)
1031
+ 989 train 7.704870 (lr=1.0053e-05) (hash(x)=26551589)
1032
+ 990 train 7.666689 (lr=1.0044e-05) (hash(x)=23689487)
1033
+ 991 train 7.758269 (lr=1.0035e-05) (hash(x)=24094093)
1034
+ 992 train 7.499237 (lr=1.0028e-05) (hash(x)=24106453)
1035
+ 993 train 7.556653 (lr=1.0021e-05) (hash(x)=23639076)
1036
+ 994 train 7.665761 (lr=1.0016e-05) (hash(x)=25107314)
1037
+ 995 train 10.815391 (lr=1.0011e-05) (hash(x)=61455960)
1038
+ 996 train 11.865571 (lr=1.0007e-05) (hash(x)=63899424)
1039
+ 997 train 11.879715 (lr=1.0004e-05) (hash(x)=65636572)
1040
+ 998 train 12.410354 (lr=1.0002e-05) (hash(x)=71578682)
1041
+ 999 val loss 7.6360
1042
+ 999 val perplexity 2071.4438
1043
+ 999 train 11.325349 (lr=1.0000e-05) (hash(x)=61104710)
lr1e-4_total_batch_size10240_seq_len512/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a626404b1a9ef230678d63355b4873840a0bb6b7519006b78a2181c07297777
3
+ size 48156226
lr1e-4_total_batch_size10240_seq_len512/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18f9e3e4e543973755d8868c5b69b495310d7f4a689dfa12e0f1d6fa0c28f567
3
+ size 71157574