andrew-healey commited on
Commit
a348a0b
·
verified ·
1 Parent(s): 3073177

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size10240_seq_len64/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size10240_seq_len64", "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": 64, "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_len64/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_len64/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7175
3
+ 0 val perplexity 122699.7891
4
+ 0 train 11.732882 (lr=3.4965e-07) (hash(x)=2985804)
5
+ 1 train 11.791316 (lr=6.9930e-07) (hash(x)=3135520)
6
+ 2 train 11.766814 (lr=1.0490e-06) (hash(x)=3699147)
7
+ 3 train 11.738402 (lr=1.3986e-06) (hash(x)=3276965)
8
+ 4 train 11.721532 (lr=1.7483e-06) (hash(x)=2444616)
9
+ 5 train 11.722276 (lr=2.0979e-06) (hash(x)=3181634)
10
+ 6 train 11.707911 (lr=2.4476e-06) (hash(x)=3550163)
11
+ 7 train 11.694474 (lr=2.7972e-06) (hash(x)=2891006)
12
+ 8 train 11.682332 (lr=3.1469e-06) (hash(x)=3381491)
13
+ 9 train 11.662008 (lr=3.4965e-06) (hash(x)=3846008)
14
+ 10 train 11.682211 (lr=3.8462e-06) (hash(x)=3596625)
15
+ 11 train 11.674217 (lr=4.1958e-06) (hash(x)=2536675)
16
+ 12 train 11.658042 (lr=4.5455e-06) (hash(x)=4321027)
17
+ 13 train 11.612878 (lr=4.8951e-06) (hash(x)=2461698)
18
+ 14 train 11.584406 (lr=5.2448e-06) (hash(x)=3818493)
19
+ 15 train 11.560591 (lr=5.5944e-06) (hash(x)=2954056)
20
+ 16 train 11.565057 (lr=5.9441e-06) (hash(x)=3663107)
21
+ 17 train 11.535749 (lr=6.2937e-06) (hash(x)=2876503)
22
+ 18 train 11.536292 (lr=6.6434e-06) (hash(x)=3158454)
23
+ 19 train 11.480976 (lr=6.9930e-06) (hash(x)=2652726)
24
+ 20 train 11.454184 (lr=7.3427e-06) (hash(x)=2766194)
25
+ 21 train 11.443119 (lr=7.6923e-06) (hash(x)=3865564)
26
+ 22 train 11.401747 (lr=8.0420e-06) (hash(x)=3535906)
27
+ 23 train 11.398273 (lr=8.3916e-06) (hash(x)=3142323)
28
+ 24 train 11.355415 (lr=8.7413e-06) (hash(x)=2747058)
29
+ 25 train 11.362532 (lr=9.0909e-06) (hash(x)=3057360)
30
+ 26 train 11.271920 (lr=9.4406e-06) (hash(x)=2669357)
31
+ 27 train 11.229914 (lr=9.7902e-06) (hash(x)=3004989)
32
+ 28 train 11.224152 (lr=1.0140e-05) (hash(x)=2440498)
33
+ 29 train 11.254957 (lr=1.0490e-05) (hash(x)=2323423)
34
+ 30 train 11.174854 (lr=1.0839e-05) (hash(x)=3100868)
35
+ 31 train 11.089357 (lr=1.1189e-05) (hash(x)=3207764)
36
+ 32 train 11.025432 (lr=1.1538e-05) (hash(x)=3271229)
37
+ 33 train 11.066368 (lr=1.1888e-05) (hash(x)=3149634)
38
+ 34 train 10.952493 (lr=1.2238e-05) (hash(x)=3774750)
39
+ 35 train 10.984518 (lr=1.2587e-05) (hash(x)=4140327)
40
+ 36 train 10.897328 (lr=1.2937e-05) (hash(x)=3438929)
41
+ 37 train 10.912499 (lr=1.3287e-05) (hash(x)=3390476)
42
+ 38 train 10.812688 (lr=1.3636e-05) (hash(x)=3939431)
43
+ 39 train 10.823194 (lr=1.3986e-05) (hash(x)=3100847)
44
+ 40 train 10.795654 (lr=1.4336e-05) (hash(x)=2942266)
45
+ 41 train 10.661864 (lr=1.4685e-05) (hash(x)=3618428)
46
+ 42 train 10.717417 (lr=1.5035e-05) (hash(x)=3598964)
47
+ 43 train 10.681093 (lr=1.5385e-05) (hash(x)=4449654)
48
+ 44 train 10.685473 (lr=1.5734e-05) (hash(x)=2906105)
49
+ 45 train 10.500842 (lr=1.6084e-05) (hash(x)=3339267)
50
+ 46 train 10.543131 (lr=1.6434e-05) (hash(x)=3382988)
51
+ 47 train 10.667154 (lr=1.6783e-05) (hash(x)=2858681)
52
+ 48 train 10.720074 (lr=1.7133e-05) (hash(x)=2894041)
53
+ 49 train 10.502328 (lr=1.7483e-05) (hash(x)=2218193)
54
+ 50 val loss 10.5592
55
+ 50 val perplexity 38528.6133
56
+ 50 train 10.567578 (lr=1.7832e-05) (hash(x)=2522274)
57
+ 51 train 10.515223 (lr=1.8182e-05) (hash(x)=3195898)
58
+ 52 train 10.522652 (lr=1.8531e-05) (hash(x)=2899839)
59
+ 53 train 10.339123 (lr=1.8881e-05) (hash(x)=2500264)
60
+ 54 train 10.422790 (lr=1.9231e-05) (hash(x)=3668948)
61
+ 55 train 10.693527 (lr=1.9580e-05) (hash(x)=3236150)
62
+ 56 train 10.424587 (lr=1.9930e-05) (hash(x)=3586288)
63
+ 57 train 10.476999 (lr=2.0280e-05) (hash(x)=3337040)
64
+ 58 train 10.469962 (lr=2.0629e-05) (hash(x)=4003369)
65
+ 59 train 10.438686 (lr=2.0979e-05) (hash(x)=3408369)
66
+ 60 train 10.366010 (lr=2.1329e-05) (hash(x)=3559111)
67
+ 61 train 10.345132 (lr=2.1678e-05) (hash(x)=2787317)
68
+ 62 train 10.265954 (lr=2.2028e-05) (hash(x)=2957029)
69
+ 63 train 10.336536 (lr=2.2378e-05) (hash(x)=3012068)
70
+ 64 train 10.367845 (lr=2.2727e-05) (hash(x)=2792231)
71
+ 65 train 10.396463 (lr=2.3077e-05) (hash(x)=2240543)
72
+ 66 train 10.440831 (lr=2.3427e-05) (hash(x)=2586857)
73
+ 67 train 10.153658 (lr=2.3776e-05) (hash(x)=3491889)
74
+ 68 train 10.400967 (lr=2.4126e-05) (hash(x)=3727327)
75
+ 69 train 10.157139 (lr=2.4476e-05) (hash(x)=2958365)
76
+ 70 train 10.412663 (lr=2.4825e-05) (hash(x)=3563076)
77
+ 71 train 10.382961 (lr=2.5175e-05) (hash(x)=3437709)
78
+ 72 train 10.324716 (lr=2.5524e-05) (hash(x)=2979567)
79
+ 73 train 10.366711 (lr=2.5874e-05) (hash(x)=4237046)
80
+ 74 train 10.273025 (lr=2.6224e-05) (hash(x)=1995144)
81
+ 75 train 10.487105 (lr=2.6573e-05) (hash(x)=5321047)
82
+ 76 train 10.325771 (lr=2.6923e-05) (hash(x)=3052730)
83
+ 77 train 10.252114 (lr=2.7273e-05) (hash(x)=3651404)
84
+ 78 train 10.226456 (lr=2.7622e-05) (hash(x)=3313001)
85
+ 79 train 10.243535 (lr=2.7972e-05) (hash(x)=3049379)
86
+ 80 train 10.075786 (lr=2.8322e-05) (hash(x)=2677265)
87
+ 81 train 10.177273 (lr=2.8671e-05) (hash(x)=2128348)
88
+ 82 train 10.207405 (lr=2.9021e-05) (hash(x)=4105439)
89
+ 83 train 10.089272 (lr=2.9371e-05) (hash(x)=3806489)
90
+ 84 train 10.028945 (lr=2.9720e-05) (hash(x)=2852625)
91
+ 85 train 10.061119 (lr=3.0070e-05) (hash(x)=3201546)
92
+ 86 train 10.108907 (lr=3.0420e-05) (hash(x)=3790597)
93
+ 87 train 10.095228 (lr=3.0769e-05) (hash(x)=3147508)
94
+ 88 train 10.112831 (lr=3.1119e-05) (hash(x)=3211389)
95
+ 89 train 9.958860 (lr=3.1469e-05) (hash(x)=2925153)
96
+ 90 train 9.963380 (lr=3.1818e-05) (hash(x)=2975211)
97
+ 91 train 10.029872 (lr=3.2168e-05) (hash(x)=4028685)
98
+ 92 train 9.961303 (lr=3.2517e-05) (hash(x)=3054572)
99
+ 93 train 10.027830 (lr=3.2867e-05) (hash(x)=3855998)
100
+ 94 train 9.902808 (lr=3.3217e-05) (hash(x)=2582342)
101
+ 95 train 9.966445 (lr=3.3566e-05) (hash(x)=3185893)
102
+ 96 train 10.002604 (lr=3.3916e-05) (hash(x)=3561988)
103
+ 97 train 10.065540 (lr=3.4266e-05) (hash(x)=3177996)
104
+ 98 train 10.107407 (lr=3.4615e-05) (hash(x)=3672699)
105
+ 99 train 10.113381 (lr=3.4965e-05) (hash(x)=3683623)
106
+ 100 val loss 9.9870
107
+ 100 val perplexity 21741.7773
108
+ 100 train 10.059031 (lr=3.5315e-05) (hash(x)=3133274)
109
+ 101 train 9.801871 (lr=3.5664e-05) (hash(x)=4336041)
110
+ 102 train 9.923193 (lr=3.6014e-05) (hash(x)=4242233)
111
+ 103 train 10.007318 (lr=3.6364e-05) (hash(x)=2323202)
112
+ 104 train 9.880161 (lr=3.6713e-05) (hash(x)=3169273)
113
+ 105 train 9.820302 (lr=3.7063e-05) (hash(x)=2378162)
114
+ 106 train 9.819331 (lr=3.7413e-05) (hash(x)=2580783)
115
+ 107 train 9.865105 (lr=3.7762e-05) (hash(x)=2761628)
116
+ 108 train 9.940001 (lr=3.8112e-05) (hash(x)=2641942)
117
+ 109 train 9.848963 (lr=3.8462e-05) (hash(x)=3761105)
118
+ 110 train 9.873786 (lr=3.8811e-05) (hash(x)=2658396)
119
+ 111 train 9.910437 (lr=3.9161e-05) (hash(x)=3434176)
120
+ 112 train 9.599015 (lr=3.9510e-05) (hash(x)=2646319)
121
+ 113 train 9.849051 (lr=3.9860e-05) (hash(x)=2757880)
122
+ 114 train 9.633743 (lr=4.0210e-05) (hash(x)=3401657)
123
+ 115 train 9.750137 (lr=4.0559e-05) (hash(x)=2554842)
124
+ 116 train 9.524903 (lr=4.0909e-05) (hash(x)=2392937)
125
+ 117 train 9.487617 (lr=4.1259e-05) (hash(x)=2568457)
126
+ 118 train 9.681759 (lr=4.1608e-05) (hash(x)=3599849)
127
+ 119 train 9.573322 (lr=4.1958e-05) (hash(x)=3411163)
128
+ 120 train 9.629271 (lr=4.2308e-05) (hash(x)=3233934)
129
+ 121 train 9.633773 (lr=4.2657e-05) (hash(x)=3663117)
130
+ 122 train 9.607313 (lr=4.3007e-05) (hash(x)=3400210)
131
+ 123 train 9.526174 (lr=4.3357e-05) (hash(x)=3803878)
132
+ 124 train 9.625339 (lr=4.3706e-05) (hash(x)=3784195)
133
+ 125 train 9.515150 (lr=4.4056e-05) (hash(x)=2647947)
134
+ 126 train 9.402566 (lr=4.4406e-05) (hash(x)=2832090)
135
+ 127 train 9.444894 (lr=4.4755e-05) (hash(x)=3427494)
136
+ 128 train 9.553656 (lr=4.5105e-05) (hash(x)=3833844)
137
+ 129 train 9.528149 (lr=4.5455e-05) (hash(x)=2930772)
138
+ 130 train 9.547407 (lr=4.5804e-05) (hash(x)=3970232)
139
+ 131 train 9.480282 (lr=4.6154e-05) (hash(x)=2811588)
140
+ 132 train 9.266448 (lr=4.6503e-05) (hash(x)=2993366)
141
+ 133 train 9.323966 (lr=4.6853e-05) (hash(x)=4308511)
142
+ 134 train 9.293565 (lr=4.7203e-05) (hash(x)=3864907)
143
+ 135 train 8.981662 (lr=4.7552e-05) (hash(x)=3222913)
144
+ 136 train 9.275683 (lr=4.7902e-05) (hash(x)=2537083)
145
+ 137 train 9.459193 (lr=4.8252e-05) (hash(x)=2888911)
146
+ 138 train 9.446100 (lr=4.8601e-05) (hash(x)=4193793)
147
+ 139 train 9.321510 (lr=4.8951e-05) (hash(x)=3510963)
148
+ 140 train 9.244029 (lr=4.9301e-05) (hash(x)=3664282)
149
+ 141 train 9.221552 (lr=4.9650e-05) (hash(x)=3639400)
150
+ 142 train 9.348597 (lr=5.0000e-05) (hash(x)=3486873)
151
+ 143 train 9.188105 (lr=5.0350e-05) (hash(x)=3226934)
152
+ 144 train 9.193623 (lr=5.0699e-05) (hash(x)=3799900)
153
+ 145 train 9.123635 (lr=5.1049e-05) (hash(x)=2881290)
154
+ 146 train 9.154635 (lr=5.1399e-05) (hash(x)=3382829)
155
+ 147 train 9.087515 (lr=5.1748e-05) (hash(x)=2870627)
156
+ 148 train 9.106842 (lr=5.2098e-05) (hash(x)=2383672)
157
+ 149 train 9.045465 (lr=5.2448e-05) (hash(x)=3237167)
158
+ 150 val loss 9.1160
159
+ 150 val perplexity 9099.4277
160
+ 150 train 9.583298 (lr=5.2797e-05) (hash(x)=3865053)
161
+ 151 train 9.114018 (lr=5.3147e-05) (hash(x)=3526556)
162
+ 152 train 9.080541 (lr=5.3497e-05) (hash(x)=3179225)
163
+ 153 train 9.165656 (lr=5.3846e-05) (hash(x)=3497999)
164
+ 154 train 9.098701 (lr=5.4196e-05) (hash(x)=3878435)
165
+ 155 train 9.146305 (lr=5.4545e-05) (hash(x)=3153837)
166
+ 156 train 9.056788 (lr=5.4895e-05) (hash(x)=3931422)
167
+ 157 train 8.840894 (lr=5.5245e-05) (hash(x)=2952505)
168
+ 158 train 8.969137 (lr=5.5594e-05) (hash(x)=2470884)
169
+ 159 train 8.825792 (lr=5.5944e-05) (hash(x)=2653794)
170
+ 160 train 8.773242 (lr=5.6294e-05) (hash(x)=3078133)
171
+ 161 train 8.743739 (lr=5.6643e-05) (hash(x)=3245055)
172
+ 162 train 8.713365 (lr=5.6993e-05) (hash(x)=2503882)
173
+ 163 train 8.793697 (lr=5.7343e-05) (hash(x)=2624914)
174
+ 164 train 8.791755 (lr=5.7692e-05) (hash(x)=2603463)
175
+ 165 train 8.674078 (lr=5.8042e-05) (hash(x)=2647960)
176
+ 166 train 8.881968 (lr=5.8392e-05) (hash(x)=2179553)
177
+ 167 train 8.785327 (lr=5.8741e-05) (hash(x)=3110930)
178
+ 168 train 8.749220 (lr=5.9091e-05) (hash(x)=2783946)
179
+ 169 train 8.610692 (lr=5.9441e-05) (hash(x)=2597532)
180
+ 170 train 8.743508 (lr=5.9790e-05) (hash(x)=2952253)
181
+ 171 train 8.746920 (lr=6.0140e-05) (hash(x)=3297008)
182
+ 172 train 8.696741 (lr=6.0490e-05) (hash(x)=3770315)
183
+ 173 train 8.649079 (lr=6.0839e-05) (hash(x)=2894320)
184
+ 174 train 8.582594 (lr=6.1189e-05) (hash(x)=3630469)
185
+ 175 train 8.660129 (lr=6.1538e-05) (hash(x)=3316544)
186
+ 176 train 8.587421 (lr=6.1888e-05) (hash(x)=3971998)
187
+ 177 train 8.525420 (lr=6.2238e-05) (hash(x)=3344675)
188
+ 178 train 8.609815 (lr=6.2587e-05) (hash(x)=2829724)
189
+ 179 train 8.539787 (lr=6.2937e-05) (hash(x)=4077726)
190
+ 180 train 8.446449 (lr=6.3287e-05) (hash(x)=2781854)
191
+ 181 train 8.351162 (lr=6.3636e-05) (hash(x)=2932335)
192
+ 182 train 8.568048 (lr=6.3986e-05) (hash(x)=3187135)
193
+ 183 train 8.545881 (lr=6.4336e-05) (hash(x)=3174796)
194
+ 184 train 8.450216 (lr=6.4685e-05) (hash(x)=2950140)
195
+ 185 train 8.519510 (lr=6.5035e-05) (hash(x)=3278917)
196
+ 186 train 8.326756 (lr=6.5385e-05) (hash(x)=2947042)
197
+ 187 train 8.511607 (lr=6.5734e-05) (hash(x)=2993037)
198
+ 188 train 8.530926 (lr=6.6084e-05) (hash(x)=3886190)
199
+ 189 train 8.374459 (lr=6.6434e-05) (hash(x)=3607426)
200
+ 190 train 8.402632 (lr=6.6783e-05) (hash(x)=2721474)
201
+ 191 train 8.366653 (lr=6.7133e-05) (hash(x)=2742929)
202
+ 192 train 8.428906 (lr=6.7483e-05) (hash(x)=2938987)
203
+ 193 train 8.352521 (lr=6.7832e-05) (hash(x)=1784952)
204
+ 194 train 8.086535 (lr=6.8182e-05) (hash(x)=2489404)
205
+ 195 train 8.090198 (lr=6.8531e-05) (hash(x)=3291574)
206
+ 196 train 8.193933 (lr=6.8881e-05) (hash(x)=2924499)
207
+ 197 train 8.134123 (lr=6.9231e-05) (hash(x)=3419749)
208
+ 198 train 8.199916 (lr=6.9580e-05) (hash(x)=3299358)
209
+ 199 train 8.352238 (lr=6.9930e-05) (hash(x)=3755178)
210
+ 200 val loss 8.2804
211
+ 200 val perplexity 3945.8938
212
+ 200 train 8.230078 (lr=7.0280e-05) (hash(x)=3934113)
213
+ 201 train 8.468972 (lr=7.0629e-05) (hash(x)=3990344)
214
+ 202 train 8.181870 (lr=7.0979e-05) (hash(x)=2709366)
215
+ 203 train 8.233955 (lr=7.1329e-05) (hash(x)=3632553)
216
+ 204 train 8.381918 (lr=7.1678e-05) (hash(x)=3057647)
217
+ 205 train 8.054997 (lr=7.2028e-05) (hash(x)=3542024)
218
+ 206 train 8.161668 (lr=7.2378e-05) (hash(x)=2707461)
219
+ 207 train 8.232626 (lr=7.2727e-05) (hash(x)=3532987)
220
+ 208 train 8.193607 (lr=7.3077e-05) (hash(x)=3581161)
221
+ 209 train 8.097413 (lr=7.3427e-05) (hash(x)=3455941)
222
+ 210 train 8.195847 (lr=7.3776e-05) (hash(x)=3528629)
223
+ 211 train 8.086587 (lr=7.4126e-05) (hash(x)=2465412)
224
+ 212 train 8.029860 (lr=7.4476e-05) (hash(x)=3705855)
225
+ 213 train 8.006989 (lr=7.4825e-05) (hash(x)=3188390)
226
+ 214 train 8.014701 (lr=7.5175e-05) (hash(x)=2691682)
227
+ 215 train 8.074903 (lr=7.5524e-05) (hash(x)=3870848)
228
+ 216 train 7.973989 (lr=7.5874e-05) (hash(x)=2645411)
229
+ 217 train 8.021168 (lr=7.6224e-05) (hash(x)=3363163)
230
+ 218 train 8.064886 (lr=7.6573e-05) (hash(x)=3091084)
231
+ 219 train 8.068585 (lr=7.6923e-05) (hash(x)=3559093)
232
+ 220 train 8.034380 (lr=7.7273e-05) (hash(x)=3306636)
233
+ 221 train 8.081930 (lr=7.7622e-05) (hash(x)=3118177)
234
+ 222 train 7.946490 (lr=7.7972e-05) (hash(x)=3433982)
235
+ 223 train 7.998009 (lr=7.8322e-05) (hash(x)=2479541)
236
+ 224 train 7.878780 (lr=7.8671e-05) (hash(x)=3024717)
237
+ 225 train 7.901913 (lr=7.9021e-05) (hash(x)=2680322)
238
+ 226 train 7.794431 (lr=7.9371e-05) (hash(x)=3283493)
239
+ 227 train 7.841173 (lr=7.9720e-05) (hash(x)=4527404)
240
+ 228 train 8.243922 (lr=8.0070e-05) (hash(x)=3612631)
241
+ 229 train 8.052505 (lr=8.0420e-05) (hash(x)=4038857)
242
+ 230 train 7.953003 (lr=8.0769e-05) (hash(x)=4298752)
243
+ 231 train 7.922385 (lr=8.1119e-05) (hash(x)=3236940)
244
+ 232 train 7.864476 (lr=8.1469e-05) (hash(x)=3064872)
245
+ 233 train 8.051711 (lr=8.1818e-05) (hash(x)=3040813)
246
+ 234 train 8.139004 (lr=8.2168e-05) (hash(x)=3451533)
247
+ 235 train 7.862817 (lr=8.2517e-05) (hash(x)=2946468)
248
+ 236 train 7.744316 (lr=8.2867e-05) (hash(x)=3418499)
249
+ 237 train 8.016415 (lr=8.3217e-05) (hash(x)=3500766)
250
+ 238 train 7.773504 (lr=8.3566e-05) (hash(x)=2657291)
251
+ 239 train 7.892015 (lr=8.3916e-05) (hash(x)=3628445)
252
+ 240 train 7.997382 (lr=8.4266e-05) (hash(x)=3673023)
253
+ 241 train 7.894599 (lr=8.4615e-05) (hash(x)=4575840)
254
+ 242 train 7.867034 (lr=8.4965e-05) (hash(x)=2093331)
255
+ 243 train 7.410897 (lr=8.5315e-05) (hash(x)=2372381)
256
+ 244 train 7.834157 (lr=8.5664e-05) (hash(x)=2805788)
257
+ 245 train 7.878745 (lr=8.6014e-05) (hash(x)=3463844)
258
+ 246 train 7.949064 (lr=8.6364e-05) (hash(x)=2618019)
259
+ 247 train 7.886242 (lr=8.6713e-05) (hash(x)=4708728)
260
+ 248 train 7.840573 (lr=8.7063e-05) (hash(x)=3193265)
261
+ 249 train 7.849874 (lr=8.7413e-05) (hash(x)=2766207)
262
+ 250 val loss 7.8483
263
+ 250 val perplexity 2561.3799
264
+ 250 train 7.841137 (lr=8.7762e-05) (hash(x)=2330970)
265
+ 251 train 7.808842 (lr=8.8112e-05) (hash(x)=2766434)
266
+ 252 train 7.922590 (lr=8.8462e-05) (hash(x)=2620477)
267
+ 253 train 7.247947 (lr=8.8811e-05) (hash(x)=1459901)
268
+ 254 train 7.685851 (lr=8.9161e-05) (hash(x)=3364474)
269
+ 255 train 7.746861 (lr=8.9510e-05) (hash(x)=3437140)
270
+ 256 train 7.887396 (lr=8.9860e-05) (hash(x)=4050994)
271
+ 257 train 7.504844 (lr=9.0210e-05) (hash(x)=2473336)
272
+ 258 train 7.834835 (lr=9.0559e-05) (hash(x)=2726907)
273
+ 259 train 7.945038 (lr=9.0909e-05) (hash(x)=3191889)
274
+ 260 train 7.913040 (lr=9.1259e-05) (hash(x)=2981356)
275
+ 261 train 7.896052 (lr=9.1608e-05) (hash(x)=2971592)
276
+ 262 train 7.736009 (lr=9.1958e-05) (hash(x)=3369519)
277
+ 263 train 7.863128 (lr=9.2308e-05) (hash(x)=3072949)
278
+ 264 train 7.767752 (lr=9.2657e-05) (hash(x)=3197904)
279
+ 265 train 7.893198 (lr=9.3007e-05) (hash(x)=3895149)
280
+ 266 train 7.881532 (lr=9.3357e-05) (hash(x)=3200202)
281
+ 267 train 7.826669 (lr=9.3706e-05) (hash(x)=2959066)
282
+ 268 train 7.676776 (lr=9.4056e-05) (hash(x)=2978755)
283
+ 269 train 7.838346 (lr=9.4406e-05) (hash(x)=3207361)
284
+ 270 train 7.742742 (lr=9.4755e-05) (hash(x)=2797327)
285
+ 271 train 7.770163 (lr=9.5105e-05) (hash(x)=3047968)
286
+ 272 train 7.767956 (lr=9.5455e-05) (hash(x)=2957193)
287
+ 273 train 7.719382 (lr=9.5804e-05) (hash(x)=2709892)
288
+ 274 train 7.679385 (lr=9.6154e-05) (hash(x)=4039987)
289
+ 275 train 7.733166 (lr=9.6503e-05) (hash(x)=2674322)
290
+ 276 train 7.713228 (lr=9.6853e-05) (hash(x)=3045172)
291
+ 277 train 7.664714 (lr=9.7203e-05) (hash(x)=2614055)
292
+ 278 train 7.786970 (lr=9.7552e-05) (hash(x)=4519396)
293
+ 279 train 7.893984 (lr=9.7902e-05) (hash(x)=2678501)
294
+ 280 train 7.704391 (lr=9.8252e-05) (hash(x)=3544564)
295
+ 281 train 7.978375 (lr=9.8601e-05) (hash(x)=2047883)
296
+ 282 train 7.837559 (lr=9.8951e-05) (hash(x)=2873219)
297
+ 283 train 7.846097 (lr=9.9301e-05) (hash(x)=2972085)
298
+ 284 train 7.718194 (lr=9.9650e-05) (hash(x)=4015293)
299
+ 285 train 7.840848 (lr=1.0000e-04) (hash(x)=3015019)
300
+ 286 train 7.664289 (lr=1.0000e-04) (hash(x)=3332166)
301
+ 287 train 7.614014 (lr=1.0000e-04) (hash(x)=2127731)
302
+ 288 train 7.633722 (lr=9.9998e-05) (hash(x)=3279094)
303
+ 289 train 7.845350 (lr=9.9996e-05) (hash(x)=2686206)
304
+ 290 train 7.758513 (lr=9.9993e-05) (hash(x)=2691923)
305
+ 291 train 7.378424 (lr=9.9989e-05) (hash(x)=2230443)
306
+ 292 train 7.665035 (lr=9.9984e-05) (hash(x)=3297222)
307
+ 293 train 7.632549 (lr=9.9979e-05) (hash(x)=2258539)
308
+ 294 train 8.158305 (lr=9.9972e-05) (hash(x)=4227667)
309
+ 295 train 8.228515 (lr=9.9965e-05) (hash(x)=3023900)
310
+ 296 train 7.855277 (lr=9.9956e-05) (hash(x)=3212440)
311
+ 297 train 8.146846 (lr=9.9947e-05) (hash(x)=3910339)
312
+ 298 train 7.862920 (lr=9.9937e-05) (hash(x)=2681147)
313
+ 299 train 7.772367 (lr=9.9926e-05) (hash(x)=3077483)
314
+ 300 val loss 7.7372
315
+ 300 val perplexity 2292.1555
316
+ 300 train 7.716761 (lr=9.9915e-05) (hash(x)=2435623)
317
+ 301 train 7.798538 (lr=9.9902e-05) (hash(x)=3668040)
318
+ 302 train 7.880817 (lr=9.9889e-05) (hash(x)=3884772)
319
+ 303 train 7.783578 (lr=9.9874e-05) (hash(x)=2480588)
320
+ 304 train 7.852970 (lr=9.9859e-05) (hash(x)=2858176)
321
+ 305 train 7.821873 (lr=9.9843e-05) (hash(x)=4355904)
322
+ 306 train 7.943045 (lr=9.9826e-05) (hash(x)=3226998)
323
+ 307 train 7.850823 (lr=9.9808e-05) (hash(x)=2508700)
324
+ 308 train 7.801064 (lr=9.9789e-05) (hash(x)=3845366)
325
+ 309 train 7.668670 (lr=9.9770e-05) (hash(x)=3142447)
326
+ 310 train 7.545741 (lr=9.9749e-05) (hash(x)=2998061)
327
+ 311 train 7.718345 (lr=9.9728e-05) (hash(x)=3500594)
328
+ 312 train 7.967668 (lr=9.9706e-05) (hash(x)=2084274)
329
+ 313 train 7.661828 (lr=9.9683e-05) (hash(x)=3730075)
330
+ 314 train 7.898852 (lr=9.9659e-05) (hash(x)=2879419)
331
+ 315 train 8.409301 (lr=9.9634e-05) (hash(x)=3585841)
332
+ 316 train 8.387711 (lr=9.9609e-05) (hash(x)=2909843)
333
+ 317 train 7.929680 (lr=9.9582e-05) (hash(x)=2936395)
334
+ 318 train 7.816414 (lr=9.9555e-05) (hash(x)=3899268)
335
+ 319 train 7.684793 (lr=9.9526e-05) (hash(x)=2979578)
336
+ 320 train 7.532546 (lr=9.9497e-05) (hash(x)=2930025)
337
+ 321 train 7.721413 (lr=9.9467e-05) (hash(x)=2976458)
338
+ 322 train 7.763329 (lr=9.9437e-05) (hash(x)=3522699)
339
+ 323 train 7.573723 (lr=9.9405e-05) (hash(x)=3006372)
340
+ 324 train 7.860309 (lr=9.9372e-05) (hash(x)=3527477)
341
+ 325 train 7.781151 (lr=9.9339e-05) (hash(x)=3252677)
342
+ 326 train 7.771098 (lr=9.9305e-05) (hash(x)=4383071)
343
+ 327 train 7.724977 (lr=9.9270e-05) (hash(x)=2619193)
344
+ 328 train 7.323357 (lr=9.9234e-05) (hash(x)=2670739)
345
+ 329 train 7.328302 (lr=9.9197e-05) (hash(x)=2844557)
346
+ 330 train 7.470593 (lr=9.9159e-05) (hash(x)=2997987)
347
+ 331 train 7.622513 (lr=9.9121e-05) (hash(x)=2264862)
348
+ 332 train 7.505060 (lr=9.9081e-05) (hash(x)=3163570)
349
+ 333 train 7.389380 (lr=9.9041e-05) (hash(x)=2321001)
350
+ 334 train 7.657947 (lr=9.9000e-05) (hash(x)=3544588)
351
+ 335 train 7.814527 (lr=9.8958e-05) (hash(x)=3953318)
352
+ 336 train 7.847098 (lr=9.8915e-05) (hash(x)=3470444)
353
+ 337 train 7.805351 (lr=9.8872e-05) (hash(x)=3813804)
354
+ 338 train 7.700163 (lr=9.8827e-05) (hash(x)=3430688)
355
+ 339 train 7.621231 (lr=9.8782e-05) (hash(x)=3172114)
356
+ 340 train 7.934962 (lr=9.8736e-05) (hash(x)=3707048)
357
+ 341 train 7.710791 (lr=9.8689e-05) (hash(x)=3630061)
358
+ 342 train 7.867616 (lr=9.8641e-05) (hash(x)=4499523)
359
+ 343 train 8.100295 (lr=9.8592e-05) (hash(x)=3624493)
360
+ 344 train 7.614330 (lr=9.8543e-05) (hash(x)=3669400)
361
+ 345 train 7.711763 (lr=9.8492e-05) (hash(x)=2885835)
362
+ 346 train 7.580270 (lr=9.8441e-05) (hash(x)=2906377)
363
+ 347 train 7.671575 (lr=9.8389e-05) (hash(x)=3628332)
364
+ 348 train 7.691696 (lr=9.8336e-05) (hash(x)=2766583)
365
+ 349 train 7.786150 (lr=9.8282e-05) (hash(x)=2844282)
366
+ 350 val loss 7.6978
367
+ 350 val perplexity 2203.4082
368
+ 350 train 7.680956 (lr=9.8228e-05) (hash(x)=3834162)
369
+ 351 train 7.678476 (lr=9.8172e-05) (hash(x)=2888972)
370
+ 352 train 7.801340 (lr=9.8116e-05) (hash(x)=3123026)
371
+ 353 train 7.620261 (lr=9.8059e-05) (hash(x)=2904637)
372
+ 354 train 7.607158 (lr=9.8001e-05) (hash(x)=2616177)
373
+ 355 train 7.945455 (lr=9.7942e-05) (hash(x)=3744709)
374
+ 356 train 7.829715 (lr=9.7882e-05) (hash(x)=1995841)
375
+ 357 train 7.679598 (lr=9.7822e-05) (hash(x)=3186059)
376
+ 358 train 7.740538 (lr=9.7761e-05) (hash(x)=3828964)
377
+ 359 train 7.752008 (lr=9.7699e-05) (hash(x)=2564054)
378
+ 360 train 7.628008 (lr=9.7636e-05) (hash(x)=2604802)
379
+ 361 train 7.667844 (lr=9.7572e-05) (hash(x)=3342173)
380
+ 362 train 7.938545 (lr=9.7507e-05) (hash(x)=3616575)
381
+ 363 train 7.457852 (lr=9.7442e-05) (hash(x)=3222229)
382
+ 364 train 7.803905 (lr=9.7376e-05) (hash(x)=3412947)
383
+ 365 train 7.687153 (lr=9.7309e-05) (hash(x)=3292116)
384
+ 366 train 7.523152 (lr=9.7241e-05) (hash(x)=3271264)
385
+ 367 train 7.972862 (lr=9.7172e-05) (hash(x)=2824555)
386
+ 368 train 7.812123 (lr=9.7103e-05) (hash(x)=4644238)
387
+ 369 train 7.768101 (lr=9.7032e-05) (hash(x)=3286145)
388
+ 370 train 7.716519 (lr=9.6961e-05) (hash(x)=2640920)
389
+ 371 train 7.816279 (lr=9.6889e-05) (hash(x)=2694606)
390
+ 372 train 7.660134 (lr=9.6817e-05) (hash(x)=4075841)
391
+ 373 train 7.694595 (lr=9.6743e-05) (hash(x)=2964484)
392
+ 374 train 7.699995 (lr=9.6669e-05) (hash(x)=2502564)
393
+ 375 train 7.656762 (lr=9.6593e-05) (hash(x)=2797954)
394
+ 376 train 7.327978 (lr=9.6518e-05) (hash(x)=3552620)
395
+ 377 train 8.204587 (lr=9.6441e-05) (hash(x)=2821863)
396
+ 378 train 7.802646 (lr=9.6363e-05) (hash(x)=3709124)
397
+ 379 train 7.963010 (lr=9.6285e-05) (hash(x)=2860360)
398
+ 380 train 7.716868 (lr=9.6206e-05) (hash(x)=3052447)
399
+ 381 train 7.897144 (lr=9.6126e-05) (hash(x)=3794897)
400
+ 382 train 7.881827 (lr=9.6045e-05) (hash(x)=3277698)
401
+ 383 train 7.921694 (lr=9.5963e-05) (hash(x)=3971772)
402
+ 384 train 7.655245 (lr=9.5881e-05) (hash(x)=4210652)
403
+ 385 train 7.655211 (lr=9.5798e-05) (hash(x)=3512226)
404
+ 386 train 7.757136 (lr=9.5714e-05) (hash(x)=3088734)
405
+ 387 train 7.710690 (lr=9.5629e-05) (hash(x)=2815361)
406
+ 388 train 7.549426 (lr=9.5544e-05) (hash(x)=2831165)
407
+ 389 train 7.808619 (lr=9.5457e-05) (hash(x)=3301986)
408
+ 390 train 7.804243 (lr=9.5370e-05) (hash(x)=3405120)
409
+ 391 train 7.653032 (lr=9.5282e-05) (hash(x)=2826471)
410
+ 392 train 7.746697 (lr=9.5194e-05) (hash(x)=3171450)
411
+ 393 train 7.701006 (lr=9.5104e-05) (hash(x)=2547188)
412
+ 394 train 7.265084 (lr=9.5014e-05) (hash(x)=2604163)
413
+ 395 train 7.629217 (lr=9.4923e-05) (hash(x)=3781597)
414
+ 396 train 7.712907 (lr=9.4831e-05) (hash(x)=3359105)
415
+ 397 train 7.604039 (lr=9.4739e-05) (hash(x)=2933586)
416
+ 398 train 7.585737 (lr=9.4646e-05) (hash(x)=4128427)
417
+ 399 train 7.714355 (lr=9.4551e-05) (hash(x)=1400697)
418
+ 400 val loss 7.6698
419
+ 400 val perplexity 2142.5583
420
+ 400 train 7.608646 (lr=9.4457e-05) (hash(x)=2541140)
421
+ 401 train 7.740726 (lr=9.4361e-05) (hash(x)=2718043)
422
+ 402 train 7.724235 (lr=9.4265e-05) (hash(x)=2976357)
423
+ 403 train 7.701550 (lr=9.4168e-05) (hash(x)=4188358)
424
+ 404 train 7.743996 (lr=9.4070e-05) (hash(x)=2773625)
425
+ 405 train 8.019323 (lr=9.3971e-05) (hash(x)=3639169)
426
+ 406 train 7.914285 (lr=9.3872e-05) (hash(x)=3486817)
427
+ 407 train 7.785763 (lr=9.3772e-05) (hash(x)=3085878)
428
+ 408 train 7.684727 (lr=9.3671e-05) (hash(x)=3634909)
429
+ 409 train 7.883955 (lr=9.3569e-05) (hash(x)=3858674)
430
+ 410 train 7.658516 (lr=9.3467e-05) (hash(x)=3812958)
431
+ 411 train 7.789698 (lr=9.3364e-05) (hash(x)=3356513)
432
+ 412 train 7.623221 (lr=9.3260e-05) (hash(x)=2684672)
433
+ 413 train 7.671316 (lr=9.3155e-05) (hash(x)=3329920)
434
+ 414 train 7.732847 (lr=9.3050e-05) (hash(x)=3346748)
435
+ 415 train 7.621972 (lr=9.2944e-05) (hash(x)=3350964)
436
+ 416 train 7.822435 (lr=9.2837e-05) (hash(x)=3318649)
437
+ 417 train 7.604810 (lr=9.2729e-05) (hash(x)=2339818)
438
+ 418 train 7.682088 (lr=9.2621e-05) (hash(x)=3070413)
439
+ 419 train 7.916272 (lr=9.2512e-05) (hash(x)=3639491)
440
+ 420 train 7.838907 (lr=9.2402e-05) (hash(x)=2181128)
441
+ 421 train 7.725039 (lr=9.2292e-05) (hash(x)=3420184)
442
+ 422 train 7.677886 (lr=9.2181e-05) (hash(x)=3511906)
443
+ 423 train 7.764006 (lr=9.2069e-05) (hash(x)=3203986)
444
+ 424 train 7.431990 (lr=9.1956e-05) (hash(x)=2810112)
445
+ 425 train 7.612066 (lr=9.1843e-05) (hash(x)=3738516)
446
+ 426 train 7.790484 (lr=9.1729e-05) (hash(x)=2584256)
447
+ 427 train 7.803124 (lr=9.1614e-05) (hash(x)=2755250)
448
+ 428 train 7.542973 (lr=9.1499e-05) (hash(x)=3085599)
449
+ 429 train 7.459343 (lr=9.1382e-05) (hash(x)=3081821)
450
+ 430 train 7.752995 (lr=9.1266e-05) (hash(x)=3566151)
451
+ 431 train 7.572211 (lr=9.1148e-05) (hash(x)=4133377)
452
+ 432 train 7.683636 (lr=9.1030e-05) (hash(x)=3502104)
453
+ 433 train 7.699809 (lr=9.0911e-05) (hash(x)=2768070)
454
+ 434 train 7.738077 (lr=9.0791e-05) (hash(x)=3640006)
455
+ 435 train 7.762598 (lr=9.0671e-05) (hash(x)=2682018)
456
+ 436 train 7.788743 (lr=9.0550e-05) (hash(x)=2629721)
457
+ 437 train 7.626055 (lr=9.0428e-05) (hash(x)=3149746)
458
+ 438 train 7.781774 (lr=9.0306e-05) (hash(x)=3944386)
459
+ 439 train 7.534145 (lr=9.0182e-05) (hash(x)=2856601)
460
+ 440 train 7.778121 (lr=9.0059e-05) (hash(x)=3191705)
461
+ 441 train 7.724093 (lr=8.9934e-05) (hash(x)=3989915)
462
+ 442 train 7.701878 (lr=8.9809e-05) (hash(x)=3279917)
463
+ 443 train 7.693469 (lr=8.9683e-05) (hash(x)=2981901)
464
+ 444 train 7.590870 (lr=8.9557e-05) (hash(x)=3399281)
465
+ 445 train 7.491492 (lr=8.9430e-05) (hash(x)=3728099)
466
+ 446 train 7.752637 (lr=8.9302e-05) (hash(x)=2521093)
467
+ 447 train 7.698043 (lr=8.9173e-05) (hash(x)=3328627)
468
+ 448 train 7.515571 (lr=8.9044e-05) (hash(x)=2728827)
469
+ 449 train 7.428347 (lr=8.8914e-05) (hash(x)=2562198)
470
+ 450 val loss 7.6409
471
+ 450 val perplexity 2081.6406
472
+ 450 train 7.656132 (lr=8.8784e-05) (hash(x)=3588452)
473
+ 451 train 7.711031 (lr=8.8653e-05) (hash(x)=3671502)
474
+ 452 train 7.762094 (lr=8.8521e-05) (hash(x)=3128347)
475
+ 453 train 7.820513 (lr=8.8388e-05) (hash(x)=2890655)
476
+ 454 train 7.639319 (lr=8.8255e-05) (hash(x)=3249651)
477
+ 455 train 7.561127 (lr=8.8122e-05) (hash(x)=3307897)
478
+ 456 train 7.458917 (lr=8.7987e-05) (hash(x)=2618491)
479
+ 457 train 7.800689 (lr=8.7852e-05) (hash(x)=4195998)
480
+ 458 train 7.579572 (lr=8.7717e-05) (hash(x)=3225031)
481
+ 459 train 7.728968 (lr=8.7580e-05) (hash(x)=3326435)
482
+ 460 train 7.566765 (lr=8.7444e-05) (hash(x)=2862440)
483
+ 461 train 7.714997 (lr=8.7306e-05) (hash(x)=2918467)
484
+ 462 train 7.692796 (lr=8.7168e-05) (hash(x)=4141440)
485
+ 463 train 7.482833 (lr=8.7029e-05) (hash(x)=2956276)
486
+ 464 train 7.593665 (lr=8.6890e-05) (hash(x)=2820808)
487
+ 465 train 7.554464 (lr=8.6750e-05) (hash(x)=3053243)
488
+ 466 train 7.698264 (lr=8.6609e-05) (hash(x)=3799382)
489
+ 467 train 7.674680 (lr=8.6468e-05) (hash(x)=3490211)
490
+ 468 train 7.655347 (lr=8.6326e-05) (hash(x)=3421416)
491
+ 469 train 7.664033 (lr=8.6184e-05) (hash(x)=3184017)
492
+ 470 train 7.682900 (lr=8.6041e-05) (hash(x)=3039663)
493
+ 471 train 7.546275 (lr=8.5897e-05) (hash(x)=3166300)
494
+ 472 train 7.580097 (lr=8.5753e-05) (hash(x)=2689958)
495
+ 473 train 7.689719 (lr=8.5608e-05) (hash(x)=2893184)
496
+ 474 train 7.765572 (lr=8.5462e-05) (hash(x)=4102744)
497
+ 475 train 7.905611 (lr=8.5316e-05) (hash(x)=3617664)
498
+ 476 train 7.497960 (lr=8.5170e-05) (hash(x)=2945509)
499
+ 477 train 7.640860 (lr=8.5022e-05) (hash(x)=2838328)
500
+ 478 train 7.749151 (lr=8.4875e-05) (hash(x)=3459400)
501
+ 479 train 7.519720 (lr=8.4726e-05) (hash(x)=3992250)
502
+ 480 train 7.579394 (lr=8.4577e-05) (hash(x)=3066475)
503
+ 481 train 7.689741 (lr=8.4428e-05) (hash(x)=3164373)
504
+ 482 train 7.708325 (lr=8.4278e-05) (hash(x)=3075962)
505
+ 483 train 7.704806 (lr=8.4127e-05) (hash(x)=3796173)
506
+ 484 train 7.647732 (lr=8.3976e-05) (hash(x)=3677710)
507
+ 485 train 7.760486 (lr=8.3824e-05) (hash(x)=3076126)
508
+ 486 train 7.522943 (lr=8.3672e-05) (hash(x)=3808030)
509
+ 487 train 7.708724 (lr=8.3519e-05) (hash(x)=3195966)
510
+ 488 train 7.702411 (lr=8.3366e-05) (hash(x)=2790464)
511
+ 489 train 7.476533 (lr=8.3212e-05) (hash(x)=2837463)
512
+ 490 train 7.258983 (lr=8.3057e-05) (hash(x)=2418446)
513
+ 491 train 7.620979 (lr=8.2902e-05) (hash(x)=2910115)
514
+ 492 train 7.596818 (lr=8.2746e-05) (hash(x)=2548970)
515
+ 493 train 7.580765 (lr=8.2590e-05) (hash(x)=3834685)
516
+ 494 train 7.774438 (lr=8.2434e-05) (hash(x)=2952651)
517
+ 495 train 7.616714 (lr=8.2276e-05) (hash(x)=2734508)
518
+ 496 train 7.665231 (lr=8.2119e-05) (hash(x)=3436107)
519
+ 497 train 7.708118 (lr=8.1960e-05) (hash(x)=3584964)
520
+ 498 train 7.657239 (lr=8.1801e-05) (hash(x)=1899895)
521
+ 499 train 7.575729 (lr=8.1642e-05) (hash(x)=3807932)
522
+ 500 val loss 7.6255
523
+ 500 val perplexity 2049.8896
524
+ 500 train 7.607306 (lr=8.1482e-05) (hash(x)=2966531)
525
+ 501 train 7.518918 (lr=8.1322e-05) (hash(x)=2853929)
526
+ 502 train 7.660302 (lr=8.1161e-05) (hash(x)=3862515)
527
+ 503 train 7.685175 (lr=8.1000e-05) (hash(x)=3085170)
528
+ 504 train 7.619595 (lr=8.0838e-05) (hash(x)=2380249)
529
+ 505 train 7.864886 (lr=8.0676e-05) (hash(x)=3686797)
530
+ 506 train 7.763685 (lr=8.0513e-05) (hash(x)=3048340)
531
+ 507 train 7.615467 (lr=8.0349e-05) (hash(x)=2942042)
532
+ 508 train 7.718357 (lr=8.0186e-05) (hash(x)=3231796)
533
+ 509 train 7.694086 (lr=8.0021e-05) (hash(x)=3284266)
534
+ 510 train 7.698176 (lr=7.9856e-05) (hash(x)=3240109)
535
+ 511 train 7.527674 (lr=7.9691e-05) (hash(x)=2205765)
536
+ 512 train 7.769763 (lr=7.9525e-05) (hash(x)=2544011)
537
+ 513 train 7.834258 (lr=7.9359e-05) (hash(x)=3191330)
538
+ 514 train 7.669385 (lr=7.9192e-05) (hash(x)=3341920)
539
+ 515 train 7.687192 (lr=7.9025e-05) (hash(x)=2809037)
540
+ 516 train 7.591370 (lr=7.8858e-05) (hash(x)=2961572)
541
+ 517 train 7.491327 (lr=7.8689e-05) (hash(x)=4122518)
542
+ 518 train 7.841850 (lr=7.8521e-05) (hash(x)=2956149)
543
+ 519 train 7.653050 (lr=7.8352e-05) (hash(x)=3194389)
544
+ 520 train 7.645664 (lr=7.8182e-05) (hash(x)=3334436)
545
+ 521 train 7.678805 (lr=7.8012e-05) (hash(x)=3250777)
546
+ 522 train 7.842827 (lr=7.7842e-05) (hash(x)=3184988)
547
+ 523 train 7.686967 (lr=7.7671e-05) (hash(x)=2929726)
548
+ 524 train 7.651426 (lr=7.7500e-05) (hash(x)=3993085)
549
+ 525 train 7.577429 (lr=7.7328e-05) (hash(x)=3630741)
550
+ 526 train 7.713064 (lr=7.7156e-05) (hash(x)=2694615)
551
+ 527 train 7.568015 (lr=7.6984e-05) (hash(x)=3381861)
552
+ 528 train 7.776250 (lr=7.6811e-05) (hash(x)=3515431)
553
+ 529 train 7.542511 (lr=7.6637e-05) (hash(x)=2972882)
554
+ 530 train 7.819644 (lr=7.6463e-05) (hash(x)=3642113)
555
+ 531 train 7.601957 (lr=7.6289e-05) (hash(x)=3289777)
556
+ 532 train 7.448237 (lr=7.6115e-05) (hash(x)=3282765)
557
+ 533 train 7.467755 (lr=7.5940e-05) (hash(x)=3131718)
558
+ 534 train 7.806204 (lr=7.5764e-05) (hash(x)=3547470)
559
+ 535 train 7.599351 (lr=7.5588e-05) (hash(x)=3725352)
560
+ 536 train 7.763870 (lr=7.5412e-05) (hash(x)=3303408)
561
+ 537 train 7.696092 (lr=7.5235e-05) (hash(x)=3494555)
562
+ 538 train 7.657309 (lr=7.5058e-05) (hash(x)=2899343)
563
+ 539 train 7.668034 (lr=7.4881e-05) (hash(x)=3380558)
564
+ 540 train 7.636491 (lr=7.4703e-05) (hash(x)=3325171)
565
+ 541 train 7.528662 (lr=7.4525e-05) (hash(x)=3365329)
566
+ 542 train 7.691559 (lr=7.4346e-05) (hash(x)=3656211)
567
+ 543 train 7.795645 (lr=7.4167e-05) (hash(x)=1857655)
568
+ 544 train 7.509477 (lr=7.3988e-05) (hash(x)=3006115)
569
+ 545 train 7.650982 (lr=7.3808e-05) (hash(x)=4154413)
570
+ 546 train 7.551301 (lr=7.3628e-05) (hash(x)=4247423)
571
+ 547 train 7.682871 (lr=7.3448e-05) (hash(x)=2427598)
572
+ 548 train 7.610775 (lr=7.3267e-05) (hash(x)=3985459)
573
+ 549 train 7.529366 (lr=7.3086e-05) (hash(x)=3269139)
574
+ 550 val loss 7.5828
575
+ 550 val perplexity 1964.2050
576
+ 550 train 7.574777 (lr=7.2904e-05) (hash(x)=3684103)
577
+ 551 train 7.556104 (lr=7.2723e-05) (hash(x)=3545128)
578
+ 552 train 7.655097 (lr=7.2540e-05) (hash(x)=2377393)
579
+ 553 train 7.610359 (lr=7.2358e-05) (hash(x)=2997863)
580
+ 554 train 7.629133 (lr=7.2175e-05) (hash(x)=2461726)
581
+ 555 train 8.410295 (lr=7.1992e-05) (hash(x)=3223656)
582
+ 556 train 7.731198 (lr=7.1808e-05) (hash(x)=2898390)
583
+ 557 train 7.653292 (lr=7.1624e-05) (hash(x)=2719317)
584
+ 558 train 7.438307 (lr=7.1440e-05) (hash(x)=3377156)
585
+ 559 train 7.415343 (lr=7.1256e-05) (hash(x)=3001561)
586
+ 560 train 7.696163 (lr=7.1071e-05) (hash(x)=2949932)
587
+ 561 train 7.499958 (lr=7.0886e-05) (hash(x)=2976270)
588
+ 562 train 7.465913 (lr=7.0701e-05) (hash(x)=2329532)
589
+ 563 train 7.619365 (lr=7.0515e-05) (hash(x)=2978010)
590
+ 564 train 7.672235 (lr=7.0329e-05) (hash(x)=1967540)
591
+ 565 train 7.711243 (lr=7.0143e-05) (hash(x)=3706363)
592
+ 566 train 7.310487 (lr=6.9956e-05) (hash(x)=2111182)
593
+ 567 train 7.584514 (lr=6.9769e-05) (hash(x)=2855690)
594
+ 568 train 7.452412 (lr=6.9582e-05) (hash(x)=2813212)
595
+ 569 train 7.657115 (lr=6.9394e-05) (hash(x)=3129189)
596
+ 570 train 7.845830 (lr=6.9207e-05) (hash(x)=2686662)
597
+ 571 train 8.023810 (lr=6.9019e-05) (hash(x)=3850767)
598
+ 572 train 7.745684 (lr=6.8830e-05) (hash(x)=2751052)
599
+ 573 train 7.522029 (lr=6.8642e-05) (hash(x)=3210679)
600
+ 574 train 7.556974 (lr=6.8453e-05) (hash(x)=2053726)
601
+ 575 train 7.874046 (lr=6.8264e-05) (hash(x)=3596049)
602
+ 576 train 7.343398 (lr=6.8075e-05) (hash(x)=1873037)
603
+ 577 train 7.755929 (lr=6.7885e-05) (hash(x)=4644943)
604
+ 578 train 7.520781 (lr=6.7695e-05) (hash(x)=2652674)
605
+ 579 train 7.507738 (lr=6.7505e-05) (hash(x)=3059736)
606
+ 580 train 7.643109 (lr=6.7315e-05) (hash(x)=2766614)
607
+ 581 train 7.410051 (lr=6.7124e-05) (hash(x)=3641800)
608
+ 582 train 7.439238 (lr=6.6933e-05) (hash(x)=2283881)
609
+ 583 train 7.853001 (lr=6.6742e-05) (hash(x)=4338038)
610
+ 584 train 7.489397 (lr=6.6551e-05) (hash(x)=2375255)
611
+ 585 train 7.613111 (lr=6.6360e-05) (hash(x)=2767891)
612
+ 586 train 7.652611 (lr=6.6168e-05) (hash(x)=2443773)
613
+ 587 train 7.508510 (lr=6.5976e-05) (hash(x)=2903160)
614
+ 588 train 7.853428 (lr=6.5784e-05) (hash(x)=2834407)
615
+ 589 train 7.411384 (lr=6.5592e-05) (hash(x)=3368763)
616
+ 590 train 7.608603 (lr=6.5399e-05) (hash(x)=3087542)
617
+ 591 train 7.515251 (lr=6.5206e-05) (hash(x)=3833327)
618
+ 592 train 7.642376 (lr=6.5013e-05) (hash(x)=3915797)
619
+ 593 train 7.698318 (lr=6.4820e-05) (hash(x)=3662874)
620
+ 594 train 7.952372 (lr=6.4627e-05) (hash(x)=2763961)
621
+ 595 train 7.706604 (lr=6.4433e-05) (hash(x)=3813983)
622
+ 596 train 7.510582 (lr=6.4240e-05) (hash(x)=3292460)
623
+ 597 train 7.637842 (lr=6.4046e-05) (hash(x)=2894133)
624
+ 598 train 7.553069 (lr=6.3852e-05) (hash(x)=3563696)
625
+ 599 train 7.506391 (lr=6.3658e-05) (hash(x)=2723409)
626
+ 600 val loss 7.5694
627
+ 600 val perplexity 1938.0405
628
+ 600 train 7.677905 (lr=6.3463e-05) (hash(x)=3003373)
629
+ 601 train 7.581322 (lr=6.3269e-05) (hash(x)=2512396)
630
+ 602 train 7.469974 (lr=6.3074e-05) (hash(x)=3097998)
631
+ 603 train 7.461981 (lr=6.2879e-05) (hash(x)=2915897)
632
+ 604 train 7.488265 (lr=6.2684e-05) (hash(x)=3359730)
633
+ 605 train 7.604180 (lr=6.2489e-05) (hash(x)=2625642)
634
+ 606 train 7.559544 (lr=6.2294e-05) (hash(x)=3809845)
635
+ 607 train 7.311430 (lr=6.2098e-05) (hash(x)=3192701)
636
+ 608 train 7.497617 (lr=6.1903e-05) (hash(x)=3624184)
637
+ 609 train 7.592073 (lr=6.1707e-05) (hash(x)=3098679)
638
+ 610 train 7.696472 (lr=6.1511e-05) (hash(x)=3000283)
639
+ 611 train 7.722456 (lr=6.1315e-05) (hash(x)=3215361)
640
+ 612 train 7.642982 (lr=6.1119e-05) (hash(x)=3304345)
641
+ 613 train 7.550508 (lr=6.0923e-05) (hash(x)=2731677)
642
+ 614 train 7.618631 (lr=6.0726e-05) (hash(x)=3189959)
643
+ 615 train 7.588629 (lr=6.0530e-05) (hash(x)=2308886)
644
+ 616 train 7.705387 (lr=6.0333e-05) (hash(x)=2558019)
645
+ 617 train 7.904314 (lr=6.0137e-05) (hash(x)=2972604)
646
+ 618 train 7.597501 (lr=5.9940e-05) (hash(x)=2857159)
647
+ 619 train 7.454687 (lr=5.9743e-05) (hash(x)=2990250)
648
+ 620 train 7.491796 (lr=5.9546e-05) (hash(x)=3225661)
649
+ 621 train 7.535346 (lr=5.9349e-05) (hash(x)=3957754)
650
+ 622 train 7.505868 (lr=5.9152e-05) (hash(x)=2627604)
651
+ 623 train 7.385798 (lr=5.8955e-05) (hash(x)=1604405)
652
+ 624 train 7.539528 (lr=5.8758e-05) (hash(x)=2120615)
653
+ 625 train 7.742959 (lr=5.8560e-05) (hash(x)=2694534)
654
+ 626 train 7.423064 (lr=5.8363e-05) (hash(x)=3903517)
655
+ 627 train 7.582479 (lr=5.8165e-05) (hash(x)=3128993)
656
+ 628 train 7.475522 (lr=5.7968e-05) (hash(x)=3414013)
657
+ 629 train 7.593332 (lr=5.7770e-05) (hash(x)=2106279)
658
+ 630 train 7.567186 (lr=5.7573e-05) (hash(x)=3301348)
659
+ 631 train 7.369084 (lr=5.7375e-05) (hash(x)=3018995)
660
+ 632 train 7.617870 (lr=5.7177e-05) (hash(x)=2741799)
661
+ 633 train 7.478047 (lr=5.6979e-05) (hash(x)=3174663)
662
+ 634 train 7.711748 (lr=5.6782e-05) (hash(x)=3741591)
663
+ 635 train 7.444135 (lr=5.6584e-05) (hash(x)=2741498)
664
+ 636 train 7.550488 (lr=5.6386e-05) (hash(x)=2425651)
665
+ 637 train 7.593585 (lr=5.6188e-05) (hash(x)=3096799)
666
+ 638 train 7.853400 (lr=5.5990e-05) (hash(x)=4022251)
667
+ 639 train 7.686088 (lr=5.5792e-05) (hash(x)=3219527)
668
+ 640 train 7.535413 (lr=5.5594e-05) (hash(x)=2734164)
669
+ 641 train 7.701591 (lr=5.5396e-05) (hash(x)=3497722)
670
+ 642 train 7.432449 (lr=5.5198e-05) (hash(x)=3479864)
671
+ 643 train 7.841222 (lr=5.5000e-05) (hash(x)=4009470)
672
+ 644 train 8.065748 (lr=5.4802e-05) (hash(x)=2524029)
673
+ 645 train 7.652209 (lr=5.4604e-05) (hash(x)=3682466)
674
+ 646 train 7.819438 (lr=5.4406e-05) (hash(x)=2777735)
675
+ 647 train 7.340162 (lr=5.4208e-05) (hash(x)=2608681)
676
+ 648 train 7.528406 (lr=5.4010e-05) (hash(x)=3042118)
677
+ 649 train 8.093957 (lr=5.3812e-05) (hash(x)=3519924)
678
+ 650 val loss 7.5535
679
+ 650 val perplexity 1907.3500
680
+ 650 train 7.717217 (lr=5.3614e-05) (hash(x)=3895249)
681
+ 651 train 7.813123 (lr=5.3416e-05) (hash(x)=3920891)
682
+ 652 train 7.682590 (lr=5.3218e-05) (hash(x)=4081649)
683
+ 653 train 7.831808 (lr=5.3021e-05) (hash(x)=3529881)
684
+ 654 train 7.975283 (lr=5.2823e-05) (hash(x)=3203465)
685
+ 655 train 7.786102 (lr=5.2625e-05) (hash(x)=2769130)
686
+ 656 train 7.563811 (lr=5.2427e-05) (hash(x)=3174554)
687
+ 657 train 7.604636 (lr=5.2230e-05) (hash(x)=3456391)
688
+ 658 train 7.743663 (lr=5.2032e-05) (hash(x)=3712598)
689
+ 659 train 7.285647 (lr=5.1835e-05) (hash(x)=2718477)
690
+ 660 train 7.618804 (lr=5.1637e-05) (hash(x)=3891056)
691
+ 661 train 7.685174 (lr=5.1440e-05) (hash(x)=3437565)
692
+ 662 train 8.293196 (lr=5.1242e-05) (hash(x)=4285805)
693
+ 663 train 7.937998 (lr=5.1045e-05) (hash(x)=2551440)
694
+ 664 train 7.600989 (lr=5.0848e-05) (hash(x)=2428909)
695
+ 665 train 7.715189 (lr=5.0651e-05) (hash(x)=3000536)
696
+ 666 train 7.709491 (lr=5.0454e-05) (hash(x)=2826118)
697
+ 667 train 7.314011 (lr=5.0257e-05) (hash(x)=2214431)
698
+ 668 train 7.537443 (lr=5.0060e-05) (hash(x)=2560061)
699
+ 669 train 7.521225 (lr=4.9863e-05) (hash(x)=2381553)
700
+ 670 train 7.493166 (lr=4.9667e-05) (hash(x)=4250845)
701
+ 671 train 7.654755 (lr=4.9470e-05) (hash(x)=3831039)
702
+ 672 train 7.936278 (lr=4.9274e-05) (hash(x)=3232763)
703
+ 673 train 7.358849 (lr=4.9077e-05) (hash(x)=3213866)
704
+ 674 train 7.368927 (lr=4.8881e-05) (hash(x)=3341280)
705
+ 675 train 7.654426 (lr=4.8685e-05) (hash(x)=2291514)
706
+ 676 train 7.433568 (lr=4.8489e-05) (hash(x)=2493476)
707
+ 677 train 7.879591 (lr=4.8293e-05) (hash(x)=2742895)
708
+ 678 train 7.393229 (lr=4.8097e-05) (hash(x)=2799504)
709
+ 679 train 7.477520 (lr=4.7902e-05) (hash(x)=4049475)
710
+ 680 train 7.362161 (lr=4.7706e-05) (hash(x)=3473655)
711
+ 681 train 7.540382 (lr=4.7511e-05) (hash(x)=2777910)
712
+ 682 train 7.727781 (lr=4.7316e-05) (hash(x)=3807083)
713
+ 683 train 7.617709 (lr=4.7121e-05) (hash(x)=3764259)
714
+ 684 train 7.586986 (lr=4.6926e-05) (hash(x)=2522605)
715
+ 685 train 7.657949 (lr=4.6731e-05) (hash(x)=3435896)
716
+ 686 train 7.593087 (lr=4.6537e-05) (hash(x)=3061883)
717
+ 687 train 7.420278 (lr=4.6342e-05) (hash(x)=2657923)
718
+ 688 train 7.443089 (lr=4.6148e-05) (hash(x)=2991574)
719
+ 689 train 7.466450 (lr=4.5954e-05) (hash(x)=2376503)
720
+ 690 train 7.640477 (lr=4.5760e-05) (hash(x)=4142078)
721
+ 691 train 8.170329 (lr=4.5567e-05) (hash(x)=4456488)
722
+ 692 train 7.668331 (lr=4.5373e-05) (hash(x)=2380558)
723
+ 693 train 7.534100 (lr=4.5180e-05) (hash(x)=3039607)
724
+ 694 train 7.557611 (lr=4.4987e-05) (hash(x)=3383738)
725
+ 695 train 7.612770 (lr=4.4794e-05) (hash(x)=3809054)
726
+ 696 train 7.472328 (lr=4.4601e-05) (hash(x)=2511697)
727
+ 697 train 7.912368 (lr=4.4408e-05) (hash(x)=2570239)
728
+ 698 train 7.566922 (lr=4.4216e-05) (hash(x)=3387491)
729
+ 699 train 7.650161 (lr=4.4024e-05) (hash(x)=4228032)
730
+ 700 val loss 7.5272
731
+ 700 val perplexity 1857.9701
732
+ 700 train 7.635756 (lr=4.3832e-05) (hash(x)=2575948)
733
+ 701 train 7.421094 (lr=4.3640e-05) (hash(x)=3066253)
734
+ 702 train 7.475317 (lr=4.3449e-05) (hash(x)=2548347)
735
+ 703 train 7.923748 (lr=4.3258e-05) (hash(x)=4680947)
736
+ 704 train 7.956561 (lr=4.3067e-05) (hash(x)=2888890)
737
+ 705 train 8.045838 (lr=4.2876e-05) (hash(x)=2659750)
738
+ 706 train 7.700026 (lr=4.2685e-05) (hash(x)=3054136)
739
+ 707 train 7.564030 (lr=4.2495e-05) (hash(x)=3129429)
740
+ 708 train 7.721849 (lr=4.2305e-05) (hash(x)=2916512)
741
+ 709 train 7.510664 (lr=4.2115e-05) (hash(x)=3107542)
742
+ 710 train 8.103435 (lr=4.1925e-05) (hash(x)=3441957)
743
+ 711 train 7.849812 (lr=4.1736e-05) (hash(x)=3278280)
744
+ 712 train 7.479384 (lr=4.1547e-05) (hash(x)=3417401)
745
+ 713 train 7.987727 (lr=4.1358e-05) (hash(x)=3659027)
746
+ 714 train 7.451620 (lr=4.1170e-05) (hash(x)=2977755)
747
+ 715 train 7.436495 (lr=4.0981e-05) (hash(x)=3042432)
748
+ 716 train 7.568407 (lr=4.0793e-05) (hash(x)=2896648)
749
+ 717 train 7.943799 (lr=4.0606e-05) (hash(x)=3750663)
750
+ 718 train 7.565557 (lr=4.0418e-05) (hash(x)=3063622)
751
+ 719 train 7.505811 (lr=4.0231e-05) (hash(x)=3572379)
752
+ 720 train 7.592162 (lr=4.0044e-05) (hash(x)=3405622)
753
+ 721 train 7.680943 (lr=3.9857e-05) (hash(x)=2309982)
754
+ 722 train 7.600262 (lr=3.9671e-05) (hash(x)=2344571)
755
+ 723 train 7.479119 (lr=3.9485e-05) (hash(x)=3026406)
756
+ 724 train 7.539344 (lr=3.9299e-05) (hash(x)=2973390)
757
+ 725 train 7.372053 (lr=3.9114e-05) (hash(x)=3319051)
758
+ 726 train 7.388613 (lr=3.8929e-05) (hash(x)=2907087)
759
+ 727 train 7.423101 (lr=3.8744e-05) (hash(x)=3395717)
760
+ 728 train 7.632541 (lr=3.8560e-05) (hash(x)=2275521)
761
+ 729 train 7.349487 (lr=3.8376e-05) (hash(x)=2722809)
762
+ 730 train 7.390145 (lr=3.8192e-05) (hash(x)=2838814)
763
+ 731 train 7.580816 (lr=3.8008e-05) (hash(x)=3786453)
764
+ 732 train 7.326694 (lr=3.7825e-05) (hash(x)=2934343)
765
+ 733 train 7.358520 (lr=3.7642e-05) (hash(x)=3464793)
766
+ 734 train 7.505126 (lr=3.7460e-05) (hash(x)=2440316)
767
+ 735 train 7.288782 (lr=3.7277e-05) (hash(x)=2767170)
768
+ 736 train 7.542027 (lr=3.7096e-05) (hash(x)=3006894)
769
+ 737 train 7.588622 (lr=3.6914e-05) (hash(x)=2692955)
770
+ 738 train 7.481993 (lr=3.6733e-05) (hash(x)=3073246)
771
+ 739 train 7.589957 (lr=3.6552e-05) (hash(x)=3016471)
772
+ 740 train 7.563406 (lr=3.6372e-05) (hash(x)=2503422)
773
+ 741 train 7.364474 (lr=3.6192e-05) (hash(x)=2613432)
774
+ 742 train 7.531850 (lr=3.6012e-05) (hash(x)=2384020)
775
+ 743 train 7.242850 (lr=3.5833e-05) (hash(x)=3194616)
776
+ 744 train 7.399672 (lr=3.5654e-05) (hash(x)=1575539)
777
+ 745 train 7.397504 (lr=3.5475e-05) (hash(x)=3706718)
778
+ 746 train 7.620665 (lr=3.5297e-05) (hash(x)=3777137)
779
+ 747 train 7.658068 (lr=3.5119e-05) (hash(x)=3606134)
780
+ 748 train 7.658666 (lr=3.4942e-05) (hash(x)=4277923)
781
+ 749 train 7.569621 (lr=3.4765e-05) (hash(x)=2227921)
782
+ 750 val loss 7.5223
783
+ 750 val perplexity 1848.8098
784
+ 750 train 7.491594 (lr=3.4588e-05) (hash(x)=3567929)
785
+ 751 train 7.471123 (lr=3.4412e-05) (hash(x)=3487452)
786
+ 752 train 7.485277 (lr=3.4236e-05) (hash(x)=2754290)
787
+ 753 train 7.906515 (lr=3.4060e-05) (hash(x)=2626716)
788
+ 754 train 7.685014 (lr=3.3885e-05) (hash(x)=2553251)
789
+ 755 train 7.216325 (lr=3.3711e-05) (hash(x)=2464443)
790
+ 756 train 7.354959 (lr=3.3537e-05) (hash(x)=2409411)
791
+ 757 train 7.337407 (lr=3.3363e-05) (hash(x)=2324629)
792
+ 758 train 7.359771 (lr=3.3189e-05) (hash(x)=2260372)
793
+ 759 train 7.495749 (lr=3.3016e-05) (hash(x)=3170051)
794
+ 760 train 7.247304 (lr=3.2844e-05) (hash(x)=1815587)
795
+ 761 train 6.893094 (lr=3.2672e-05) (hash(x)=2737654)
796
+ 762 train 7.294945 (lr=3.2500e-05) (hash(x)=3277493)
797
+ 763 train 7.523612 (lr=3.2329e-05) (hash(x)=3647630)
798
+ 764 train 7.423681 (lr=3.2158e-05) (hash(x)=2872476)
799
+ 765 train 7.301933 (lr=3.1988e-05) (hash(x)=2874607)
800
+ 766 train 7.305700 (lr=3.1818e-05) (hash(x)=3087201)
801
+ 767 train 7.483623 (lr=3.1648e-05) (hash(x)=2701214)
802
+ 768 train 7.660174 (lr=3.1479e-05) (hash(x)=3448247)
803
+ 769 train 7.451676 (lr=3.1311e-05) (hash(x)=3132192)
804
+ 770 train 7.617448 (lr=3.1142e-05) (hash(x)=2171985)
805
+ 771 train 7.431528 (lr=3.0975e-05) (hash(x)=3293012)
806
+ 772 train 7.581586 (lr=3.0808e-05) (hash(x)=3195715)
807
+ 773 train 7.458855 (lr=3.0641e-05) (hash(x)=2375711)
808
+ 774 train 7.522786 (lr=3.0475e-05) (hash(x)=3752547)
809
+ 775 train 7.494302 (lr=3.0309e-05) (hash(x)=3582457)
810
+ 776 train 7.362676 (lr=3.0144e-05) (hash(x)=2892427)
811
+ 777 train 7.572007 (lr=2.9979e-05) (hash(x)=3465407)
812
+ 778 train 7.467859 (lr=2.9814e-05) (hash(x)=2967905)
813
+ 779 train 7.873896 (lr=2.9651e-05) (hash(x)=2494120)
814
+ 780 train 7.381324 (lr=2.9487e-05) (hash(x)=2364729)
815
+ 781 train 7.446656 (lr=2.9324e-05) (hash(x)=2494320)
816
+ 782 train 7.285203 (lr=2.9162e-05) (hash(x)=3042165)
817
+ 783 train 7.551577 (lr=2.9000e-05) (hash(x)=3144649)
818
+ 784 train 7.450248 (lr=2.8839e-05) (hash(x)=2686204)
819
+ 785 train 7.511783 (lr=2.8678e-05) (hash(x)=3547013)
820
+ 786 train 7.537108 (lr=2.8518e-05) (hash(x)=2415554)
821
+ 787 train 7.541318 (lr=2.8358e-05) (hash(x)=2882311)
822
+ 788 train 7.414227 (lr=2.8199e-05) (hash(x)=2796854)
823
+ 789 train 7.595479 (lr=2.8040e-05) (hash(x)=3479214)
824
+ 790 train 7.466133 (lr=2.7881e-05) (hash(x)=3598923)
825
+ 791 train 7.406702 (lr=2.7724e-05) (hash(x)=2906087)
826
+ 792 train 7.620306 (lr=2.7566e-05) (hash(x)=2551378)
827
+ 793 train 7.622853 (lr=2.7410e-05) (hash(x)=3427643)
828
+ 794 train 7.324991 (lr=2.7254e-05) (hash(x)=3641204)
829
+ 795 train 7.538713 (lr=2.7098e-05) (hash(x)=3218278)
830
+ 796 train 7.455210 (lr=2.6943e-05) (hash(x)=2334992)
831
+ 797 train 7.404364 (lr=2.6788e-05) (hash(x)=3501271)
832
+ 798 train 7.435220 (lr=2.6634e-05) (hash(x)=3698242)
833
+ 799 train 7.762365 (lr=2.6481e-05) (hash(x)=2943967)
834
+ 800 val loss 7.5056
835
+ 800 val perplexity 1818.1195
836
+ 800 train 7.573648 (lr=2.6328e-05) (hash(x)=2286899)
837
+ 801 train 7.504590 (lr=2.6176e-05) (hash(x)=2263236)
838
+ 802 train 7.575791 (lr=2.6024e-05) (hash(x)=2054042)
839
+ 803 train 7.365855 (lr=2.5873e-05) (hash(x)=2540358)
840
+ 804 train 7.540182 (lr=2.5722e-05) (hash(x)=2811996)
841
+ 805 train 7.617575 (lr=2.5572e-05) (hash(x)=2927693)
842
+ 806 train 7.422434 (lr=2.5423e-05) (hash(x)=2390390)
843
+ 807 train 7.499634 (lr=2.5274e-05) (hash(x)=2746040)
844
+ 808 train 7.504352 (lr=2.5125e-05) (hash(x)=3860276)
845
+ 809 train 7.400847 (lr=2.4978e-05) (hash(x)=1625088)
846
+ 810 train 7.321376 (lr=2.4830e-05) (hash(x)=3410298)
847
+ 811 train 7.417576 (lr=2.4684e-05) (hash(x)=2451069)
848
+ 812 train 8.607512 (lr=2.4538e-05) (hash(x)=3908827)
849
+ 813 train 7.666638 (lr=2.4392e-05) (hash(x)=2819525)
850
+ 814 train 7.210447 (lr=2.4247e-05) (hash(x)=3321238)
851
+ 815 train 7.253762 (lr=2.4103e-05) (hash(x)=3058592)
852
+ 816 train 7.172640 (lr=2.3959e-05) (hash(x)=2656240)
853
+ 817 train 7.809765 (lr=2.3816e-05) (hash(x)=2877889)
854
+ 818 train 7.479909 (lr=2.3674e-05) (hash(x)=3343391)
855
+ 819 train 7.530747 (lr=2.3532e-05) (hash(x)=3470216)
856
+ 820 train 7.499367 (lr=2.3391e-05) (hash(x)=2737326)
857
+ 821 train 7.512764 (lr=2.3250e-05) (hash(x)=2307619)
858
+ 822 train 7.613177 (lr=2.3110e-05) (hash(x)=3583913)
859
+ 823 train 7.497216 (lr=2.2971e-05) (hash(x)=3978270)
860
+ 824 train 7.284101 (lr=2.2832e-05) (hash(x)=2875391)
861
+ 825 train 7.379969 (lr=2.2694e-05) (hash(x)=1947808)
862
+ 826 train 7.423354 (lr=2.2556e-05) (hash(x)=2980393)
863
+ 827 train 7.537014 (lr=2.2420e-05) (hash(x)=3646840)
864
+ 828 train 7.649922 (lr=2.2283e-05) (hash(x)=2949294)
865
+ 829 train 7.319931 (lr=2.2148e-05) (hash(x)=3240215)
866
+ 830 train 7.661159 (lr=2.2013e-05) (hash(x)=3536109)
867
+ 831 train 7.461557 (lr=2.1878e-05) (hash(x)=2647166)
868
+ 832 train 7.393826 (lr=2.1745e-05) (hash(x)=3354451)
869
+ 833 train 7.765822 (lr=2.1612e-05) (hash(x)=2780633)
870
+ 834 train 7.653461 (lr=2.1479e-05) (hash(x)=4234876)
871
+ 835 train 7.636785 (lr=2.1347e-05) (hash(x)=2431165)
872
+ 836 train 7.678750 (lr=2.1216e-05) (hash(x)=3190961)
873
+ 837 train 7.461987 (lr=2.1086e-05) (hash(x)=2819514)
874
+ 838 train 7.337444 (lr=2.0956e-05) (hash(x)=2351525)
875
+ 839 train 7.388110 (lr=2.0827e-05) (hash(x)=3583496)
876
+ 840 train 7.511945 (lr=2.0698e-05) (hash(x)=2926687)
877
+ 841 train 7.449077 (lr=2.0570e-05) (hash(x)=3830813)
878
+ 842 train 7.786232 (lr=2.0443e-05) (hash(x)=3381957)
879
+ 843 train 7.963102 (lr=2.0317e-05) (hash(x)=2907503)
880
+ 844 train 7.807460 (lr=2.0191e-05) (hash(x)=4139356)
881
+ 845 train 7.892079 (lr=2.0066e-05) (hash(x)=3178968)
882
+ 846 train 7.412551 (lr=1.9941e-05) (hash(x)=2955919)
883
+ 847 train 7.785867 (lr=1.9818e-05) (hash(x)=4734264)
884
+ 848 train 7.458826 (lr=1.9694e-05) (hash(x)=2414220)
885
+ 849 train 7.455934 (lr=1.9572e-05) (hash(x)=3586561)
886
+ 850 val loss 7.4962
887
+ 850 val perplexity 1801.2391
888
+ 850 train 7.442237 (lr=1.9450e-05) (hash(x)=3015002)
889
+ 851 train 7.805845 (lr=1.9329e-05) (hash(x)=4311226)
890
+ 852 train 7.814644 (lr=1.9209e-05) (hash(x)=4128864)
891
+ 853 train 7.694846 (lr=1.9089e-05) (hash(x)=2626153)
892
+ 854 train 7.670996 (lr=1.8970e-05) (hash(x)=3591968)
893
+ 855 train 7.457673 (lr=1.8852e-05) (hash(x)=3757961)
894
+ 856 train 7.419512 (lr=1.8734e-05) (hash(x)=3235141)
895
+ 857 train 7.482204 (lr=1.8618e-05) (hash(x)=3563903)
896
+ 858 train 7.411292 (lr=1.8501e-05) (hash(x)=3264654)
897
+ 859 train 7.287813 (lr=1.8386e-05) (hash(x)=2892205)
898
+ 860 train 7.236824 (lr=1.8271e-05) (hash(x)=3217394)
899
+ 861 train 7.780243 (lr=1.8157e-05) (hash(x)=3676449)
900
+ 862 train 7.567054 (lr=1.8044e-05) (hash(x)=2509857)
901
+ 863 train 7.546492 (lr=1.7931e-05) (hash(x)=4296014)
902
+ 864 train 8.019617 (lr=1.7819e-05) (hash(x)=3316183)
903
+ 865 train 7.363657 (lr=1.7708e-05) (hash(x)=3097760)
904
+ 866 train 7.897007 (lr=1.7598e-05) (hash(x)=4161428)
905
+ 867 train 7.311205 (lr=1.7488e-05) (hash(x)=2397094)
906
+ 868 train 7.197265 (lr=1.7379e-05) (hash(x)=2904567)
907
+ 869 train 7.375964 (lr=1.7271e-05) (hash(x)=2597681)
908
+ 870 train 7.716062 (lr=1.7163e-05) (hash(x)=3344641)
909
+ 871 train 7.436155 (lr=1.7056e-05) (hash(x)=3713912)
910
+ 872 train 7.580998 (lr=1.6950e-05) (hash(x)=3686614)
911
+ 873 train 7.469752 (lr=1.6845e-05) (hash(x)=3389429)
912
+ 874 train 7.620060 (lr=1.6740e-05) (hash(x)=3786539)
913
+ 875 train 7.415157 (lr=1.6636e-05) (hash(x)=3174315)
914
+ 876 train 7.619607 (lr=1.6533e-05) (hash(x)=3075463)
915
+ 877 train 7.745299 (lr=1.6431e-05) (hash(x)=3065010)
916
+ 878 train 7.947687 (lr=1.6329e-05) (hash(x)=3217514)
917
+ 879 train 7.577914 (lr=1.6228e-05) (hash(x)=2782831)
918
+ 880 train 7.417246 (lr=1.6128e-05) (hash(x)=3275105)
919
+ 881 train 7.539039 (lr=1.6029e-05) (hash(x)=3055236)
920
+ 882 train 7.495480 (lr=1.5930e-05) (hash(x)=2609168)
921
+ 883 train 7.449752 (lr=1.5832e-05) (hash(x)=2552776)
922
+ 884 train 7.245572 (lr=1.5735e-05) (hash(x)=2061814)
923
+ 885 train 7.569818 (lr=1.5639e-05) (hash(x)=3523226)
924
+ 886 train 7.013706 (lr=1.5543e-05) (hash(x)=3355915)
925
+ 887 train 7.404282 (lr=1.5449e-05) (hash(x)=2567876)
926
+ 888 train 7.892108 (lr=1.5354e-05) (hash(x)=2907376)
927
+ 889 train 7.684035 (lr=1.5261e-05) (hash(x)=2263655)
928
+ 890 train 7.427641 (lr=1.5169e-05) (hash(x)=3018007)
929
+ 891 train 7.563244 (lr=1.5077e-05) (hash(x)=3132777)
930
+ 892 train 7.317602 (lr=1.4986e-05) (hash(x)=3152647)
931
+ 893 train 7.630404 (lr=1.4896e-05) (hash(x)=3400701)
932
+ 894 train 7.566764 (lr=1.4806e-05) (hash(x)=3500338)
933
+ 895 train 7.372306 (lr=1.4718e-05) (hash(x)=2431080)
934
+ 896 train 7.573184 (lr=1.4630e-05) (hash(x)=2782909)
935
+ 897 train 7.344350 (lr=1.4543e-05) (hash(x)=3413177)
936
+ 898 train 7.381039 (lr=1.4456e-05) (hash(x)=2629009)
937
+ 899 train 7.613544 (lr=1.4371e-05) (hash(x)=2625679)
938
+ 900 val loss 7.4821
939
+ 900 val perplexity 1775.9376
940
+ 900 train 7.503239 (lr=1.4286e-05) (hash(x)=2642823)
941
+ 901 train 7.405985 (lr=1.4202e-05) (hash(x)=2607650)
942
+ 902 train 7.530386 (lr=1.4119e-05) (hash(x)=3177508)
943
+ 903 train 7.543421 (lr=1.4037e-05) (hash(x)=2932917)
944
+ 904 train 7.568965 (lr=1.3955e-05) (hash(x)=2663836)
945
+ 905 train 7.423760 (lr=1.3874e-05) (hash(x)=2815840)
946
+ 906 train 7.611568 (lr=1.3794e-05) (hash(x)=2623723)
947
+ 907 train 7.352070 (lr=1.3715e-05) (hash(x)=3428759)
948
+ 908 train 7.516800 (lr=1.3637e-05) (hash(x)=3042894)
949
+ 909 train 7.534652 (lr=1.3559e-05) (hash(x)=3091077)
950
+ 910 train 7.400636 (lr=1.3482e-05) (hash(x)=3291516)
951
+ 911 train 7.405321 (lr=1.3407e-05) (hash(x)=2417948)
952
+ 912 train 7.353654 (lr=1.3331e-05) (hash(x)=2513340)
953
+ 913 train 7.391990 (lr=1.3257e-05) (hash(x)=2833643)
954
+ 914 train 7.587106 (lr=1.3183e-05) (hash(x)=3190766)
955
+ 915 train 7.672332 (lr=1.3111e-05) (hash(x)=2474886)
956
+ 916 train 7.235352 (lr=1.3039e-05) (hash(x)=3247777)
957
+ 917 train 7.667085 (lr=1.2968e-05) (hash(x)=2257452)
958
+ 918 train 7.412519 (lr=1.2897e-05) (hash(x)=3290762)
959
+ 919 train 7.462494 (lr=1.2828e-05) (hash(x)=3892170)
960
+ 920 train 7.414679 (lr=1.2759e-05) (hash(x)=2951042)
961
+ 921 train 7.503744 (lr=1.2691e-05) (hash(x)=3592175)
962
+ 922 train 7.594787 (lr=1.2624e-05) (hash(x)=4413156)
963
+ 923 train 7.801903 (lr=1.2558e-05) (hash(x)=2884046)
964
+ 924 train 7.503883 (lr=1.2493e-05) (hash(x)=2392123)
965
+ 925 train 7.448651 (lr=1.2428e-05) (hash(x)=2631284)
966
+ 926 train 7.524350 (lr=1.2364e-05) (hash(x)=2897578)
967
+ 927 train 7.595761 (lr=1.2301e-05) (hash(x)=2912533)
968
+ 928 train 7.350925 (lr=1.2239e-05) (hash(x)=2560638)
969
+ 929 train 7.313988 (lr=1.2178e-05) (hash(x)=2709843)
970
+ 930 train 7.437463 (lr=1.2118e-05) (hash(x)=2811132)
971
+ 931 train 7.322388 (lr=1.2058e-05) (hash(x)=4656065)
972
+ 932 train 8.113042 (lr=1.1999e-05) (hash(x)=3776860)
973
+ 933 train 7.178877 (lr=1.1941e-05) (hash(x)=3083056)
974
+ 934 train 7.436421 (lr=1.1884e-05) (hash(x)=3072511)
975
+ 935 train 7.645499 (lr=1.1828e-05) (hash(x)=3329202)
976
+ 936 train 7.767838 (lr=1.1772e-05) (hash(x)=3889488)
977
+ 937 train 7.684964 (lr=1.1718e-05) (hash(x)=3607079)
978
+ 938 train 7.854150 (lr=1.1664e-05) (hash(x)=2997421)
979
+ 939 train 7.703346 (lr=1.1611e-05) (hash(x)=2922687)
980
+ 940 train 7.541104 (lr=1.1559e-05) (hash(x)=2606887)
981
+ 941 train 7.719440 (lr=1.1508e-05) (hash(x)=3816995)
982
+ 942 train 7.862646 (lr=1.1457e-05) (hash(x)=3290027)
983
+ 943 train 7.355483 (lr=1.1408e-05) (hash(x)=2613839)
984
+ 944 train 7.387838 (lr=1.1359e-05) (hash(x)=2348558)
985
+ 945 train 7.514312 (lr=1.1311e-05) (hash(x)=2973517)
986
+ 946 train 7.432105 (lr=1.1264e-05) (hash(x)=2713424)
987
+ 947 train 7.622721 (lr=1.1218e-05) (hash(x)=3050661)
988
+ 948 train 7.377713 (lr=1.1173e-05) (hash(x)=2974673)
989
+ 949 train 7.316993 (lr=1.1128e-05) (hash(x)=3219654)
990
+ 950 val loss 7.4722
991
+ 950 val perplexity 1758.5087
992
+ 950 train 7.303560 (lr=1.1085e-05) (hash(x)=3110593)
993
+ 951 train 7.586716 (lr=1.1042e-05) (hash(x)=3064396)
994
+ 952 train 7.480153 (lr=1.1000e-05) (hash(x)=3693334)
995
+ 953 train 7.472680 (lr=1.0959e-05) (hash(x)=2691151)
996
+ 954 train 7.288827 (lr=1.0919e-05) (hash(x)=3754942)
997
+ 955 train 7.300562 (lr=1.0879e-05) (hash(x)=3033616)
998
+ 956 train 7.297580 (lr=1.0841e-05) (hash(x)=3287084)
999
+ 957 train 7.727380 (lr=1.0803e-05) (hash(x)=4015530)
1000
+ 958 train 8.217419 (lr=1.0766e-05) (hash(x)=4262983)
1001
+ 959 train 8.227455 (lr=1.0730e-05) (hash(x)=4020245)
1002
+ 960 train 8.154897 (lr=1.0695e-05) (hash(x)=3377032)
1003
+ 961 train 8.179979 (lr=1.0661e-05) (hash(x)=3867253)
1004
+ 962 train 8.123699 (lr=1.0628e-05) (hash(x)=4112510)
1005
+ 963 train 8.183675 (lr=1.0595e-05) (hash(x)=4362570)
1006
+ 964 train 8.108925 (lr=1.0563e-05) (hash(x)=3930313)
1007
+ 965 train 8.125933 (lr=1.0533e-05) (hash(x)=3219449)
1008
+ 966 train 8.175665 (lr=1.0503e-05) (hash(x)=4183096)
1009
+ 967 train 8.216145 (lr=1.0474e-05) (hash(x)=4161781)
1010
+ 968 train 8.117911 (lr=1.0445e-05) (hash(x)=3739940)
1011
+ 969 train 8.326439 (lr=1.0418e-05) (hash(x)=5475477)
1012
+ 970 train 8.126746 (lr=1.0391e-05) (hash(x)=3323796)
1013
+ 971 train 8.305820 (lr=1.0366e-05) (hash(x)=4595815)
1014
+ 972 train 8.201150 (lr=1.0341e-05) (hash(x)=2800363)
1015
+ 973 train 7.550983 (lr=1.0317e-05) (hash(x)=3811908)
1016
+ 974 train 7.536014 (lr=1.0294e-05) (hash(x)=3394824)
1017
+ 975 train 7.659945 (lr=1.0272e-05) (hash(x)=3775118)
1018
+ 976 train 7.946225 (lr=1.0251e-05) (hash(x)=3319189)
1019
+ 977 train 7.550482 (lr=1.0230e-05) (hash(x)=2270334)
1020
+ 978 train 7.467356 (lr=1.0211e-05) (hash(x)=3315180)
1021
+ 979 train 7.314185 (lr=1.0192e-05) (hash(x)=3692980)
1022
+ 980 train 7.507624 (lr=1.0174e-05) (hash(x)=2784313)
1023
+ 981 train 7.427670 (lr=1.0157e-05) (hash(x)=3108034)
1024
+ 982 train 7.437014 (lr=1.0141e-05) (hash(x)=2478481)
1025
+ 983 train 7.352783 (lr=1.0126e-05) (hash(x)=4072794)
1026
+ 984 train 7.659690 (lr=1.0111e-05) (hash(x)=2687805)
1027
+ 985 train 7.352623 (lr=1.0098e-05) (hash(x)=2461586)
1028
+ 986 train 7.369145 (lr=1.0085e-05) (hash(x)=2884749)
1029
+ 987 train 7.515239 (lr=1.0074e-05) (hash(x)=3480755)
1030
+ 988 train 7.630845 (lr=1.0063e-05) (hash(x)=3868079)
1031
+ 989 train 7.544714 (lr=1.0053e-05) (hash(x)=4126260)
1032
+ 990 train 7.507469 (lr=1.0044e-05) (hash(x)=2880733)
1033
+ 991 train 7.584702 (lr=1.0035e-05) (hash(x)=2579288)
1034
+ 992 train 7.316351 (lr=1.0028e-05) (hash(x)=2677892)
1035
+ 993 train 7.384012 (lr=1.0021e-05) (hash(x)=2938560)
1036
+ 994 train 7.519317 (lr=1.0016e-05) (hash(x)=3845646)
1037
+ 995 train 10.906981 (lr=1.0011e-05) (hash(x)=7554381)
1038
+ 996 train 11.985785 (lr=1.0007e-05) (hash(x)=8656217)
1039
+ 997 train 11.994038 (lr=1.0004e-05) (hash(x)=9115187)
1040
+ 998 train 12.539096 (lr=1.0002e-05) (hash(x)=8013822)
1041
+ 999 val loss 7.4692
1042
+ 999 val perplexity 1753.1829
1043
+ 999 train 11.338999 (lr=1.0000e-05) (hash(x)=8849063)
lr1e-4_total_batch_size10240_seq_len64/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ae998309b83430d17db6bb3ba7cd4a4a13f4f306eeb5afab9d4c76b41731940
3
+ size 35540418
lr1e-4_total_batch_size10240_seq_len64/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6684444e5e8b7d2b7f64edacd72fc6baf6c4467cc77d0c348c063fdb1b386b12
3
+ size 70698822