andrew-healey commited on
Commit
32b42a2
·
verified ·
1 Parent(s): 8cac03d

Upload folder using huggingface_hub

Browse files
lr5e-4_total_batch_size20480_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr5e-4_total_batch_size20480_seq_len128", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 128, "batch_size": null, "total_batch_size": 20480, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0005, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr5e-4_total_batch_size20480_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8513131d2bf732f8f6d97ff82d32c505b2a2f3f9f2247bac315bb09ce0e1f97
3
+ size 964
lr5e-4_total_batch_size20480_seq_len128/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.6870
3
+ 0 val perplexity 119016.5000
4
+ 0 train 11.700421 (lr=1.7483e-06) (hash(x)=5113283)
5
+ 1 train 11.672960 (lr=3.4965e-06) (hash(x)=5534142)
6
+ 2 train 11.695552 (lr=5.2448e-06) (hash(x)=6763231)
7
+ 3 train 11.666068 (lr=6.9930e-06) (hash(x)=6328012)
8
+ 4 train 11.654607 (lr=8.7413e-06) (hash(x)=7052161)
9
+ 5 train 11.608361 (lr=1.0490e-05) (hash(x)=5556098)
10
+ 6 train 11.570334 (lr=1.2238e-05) (hash(x)=4696943)
11
+ 7 train 11.518970 (lr=1.3986e-05) (hash(x)=6219138)
12
+ 8 train 11.462629 (lr=1.5734e-05) (hash(x)=6049878)
13
+ 9 train 11.423951 (lr=1.7483e-05) (hash(x)=5491972)
14
+ 10 train 11.366195 (lr=1.9231e-05) (hash(x)=8075458)
15
+ 11 train 11.285904 (lr=2.0979e-05) (hash(x)=5917741)
16
+ 12 train 11.237248 (lr=2.2727e-05) (hash(x)=5750403)
17
+ 13 train 11.162921 (lr=2.4476e-05) (hash(x)=6112458)
18
+ 14 train 11.094389 (lr=2.6224e-05) (hash(x)=5806490)
19
+ 15 train 10.994961 (lr=2.7972e-05) (hash(x)=6568379)
20
+ 16 train 10.910911 (lr=2.9720e-05) (hash(x)=7212403)
21
+ 17 train 10.843513 (lr=3.1469e-05) (hash(x)=7386082)
22
+ 18 train 10.781620 (lr=3.3217e-05) (hash(x)=6359743)
23
+ 19 train 10.729923 (lr=3.4965e-05) (hash(x)=6151674)
24
+ 20 train 10.690209 (lr=3.6713e-05) (hash(x)=7168704)
25
+ 21 train 10.643327 (lr=3.8462e-05) (hash(x)=7859060)
26
+ 22 train 10.557471 (lr=4.0210e-05) (hash(x)=6976239)
27
+ 23 train 10.554563 (lr=4.1958e-05) (hash(x)=6391506)
28
+ 24 train 10.531520 (lr=4.3706e-05) (hash(x)=5245999)
29
+ 25 train 10.479081 (lr=4.5455e-05) (hash(x)=5985675)
30
+ 26 train 10.392241 (lr=4.7203e-05) (hash(x)=4355733)
31
+ 27 train 10.559300 (lr=4.8951e-05) (hash(x)=6984772)
32
+ 28 train 10.405327 (lr=5.0699e-05) (hash(x)=6244096)
33
+ 29 train 10.423216 (lr=5.2448e-05) (hash(x)=6603717)
34
+ 30 train 10.325966 (lr=5.4196e-05) (hash(x)=5669522)
35
+ 31 train 10.309162 (lr=5.5944e-05) (hash(x)=6190579)
36
+ 32 train 10.370431 (lr=5.7692e-05) (hash(x)=4912983)
37
+ 33 train 10.242783 (lr=5.9441e-05) (hash(x)=7276303)
38
+ 34 train 10.306746 (lr=6.1189e-05) (hash(x)=5479947)
39
+ 35 train 10.397912 (lr=6.2937e-05) (hash(x)=6630378)
40
+ 36 train 10.329117 (lr=6.4685e-05) (hash(x)=8097010)
41
+ 37 train 10.381491 (lr=6.6434e-05) (hash(x)=9989744)
42
+ 38 train 10.252186 (lr=6.8182e-05) (hash(x)=6694932)
43
+ 39 train 10.178607 (lr=6.9930e-05) (hash(x)=6317894)
44
+ 40 train 10.088614 (lr=7.1678e-05) (hash(x)=5046327)
45
+ 41 train 10.156977 (lr=7.3427e-05) (hash(x)=7947703)
46
+ 42 train 10.015911 (lr=7.5175e-05) (hash(x)=5916068)
47
+ 43 train 10.109138 (lr=7.6923e-05) (hash(x)=5777498)
48
+ 44 train 10.032684 (lr=7.8671e-05) (hash(x)=5753092)
49
+ 45 train 10.005583 (lr=8.0420e-05) (hash(x)=7723226)
50
+ 46 train 9.986021 (lr=8.2168e-05) (hash(x)=7007016)
51
+ 47 train 9.901617 (lr=8.3916e-05) (hash(x)=6491229)
52
+ 48 train 9.931938 (lr=8.5664e-05) (hash(x)=6390752)
53
+ 49 train 10.034129 (lr=8.7413e-05) (hash(x)=6019389)
54
+ 50 val loss 9.8939
55
+ 50 val perplexity 19809.6738
56
+ 50 train 9.885399 (lr=8.9161e-05) (hash(x)=7662067)
57
+ 51 train 9.873328 (lr=9.0909e-05) (hash(x)=4648609)
58
+ 52 train 9.797134 (lr=9.2657e-05) (hash(x)=4725966)
59
+ 53 train 9.742386 (lr=9.4406e-05) (hash(x)=5558355)
60
+ 54 train 9.767694 (lr=9.6154e-05) (hash(x)=6524933)
61
+ 55 train 9.704561 (lr=9.7902e-05) (hash(x)=6369326)
62
+ 56 train 9.654732 (lr=9.9650e-05) (hash(x)=6022625)
63
+ 57 train 9.575443 (lr=1.0140e-04) (hash(x)=4906853)
64
+ 58 train 9.381954 (lr=1.0315e-04) (hash(x)=5589269)
65
+ 59 train 9.489091 (lr=1.0490e-04) (hash(x)=6437376)
66
+ 60 train 9.515290 (lr=1.0664e-04) (hash(x)=7304892)
67
+ 61 train 9.461693 (lr=1.0839e-04) (hash(x)=7084093)
68
+ 62 train 9.460057 (lr=1.1014e-04) (hash(x)=5908178)
69
+ 63 train 9.228505 (lr=1.1189e-04) (hash(x)=7048804)
70
+ 64 train 9.292789 (lr=1.1364e-04) (hash(x)=6085549)
71
+ 65 train 9.291378 (lr=1.1538e-04) (hash(x)=5752594)
72
+ 66 train 9.076304 (lr=1.1713e-04) (hash(x)=7662769)
73
+ 67 train 8.918578 (lr=1.1888e-04) (hash(x)=6393520)
74
+ 68 train 9.151848 (lr=1.2063e-04) (hash(x)=5942867)
75
+ 69 train 9.127384 (lr=1.2238e-04) (hash(x)=6793550)
76
+ 70 train 8.950273 (lr=1.2413e-04) (hash(x)=7105976)
77
+ 71 train 8.967115 (lr=1.2587e-04) (hash(x)=6581348)
78
+ 72 train 8.880000 (lr=1.2762e-04) (hash(x)=5555598)
79
+ 73 train 8.822878 (lr=1.2937e-04) (hash(x)=6235837)
80
+ 74 train 8.747945 (lr=1.3112e-04) (hash(x)=5591770)
81
+ 75 train 8.946789 (lr=1.3287e-04) (hash(x)=7083021)
82
+ 76 train 8.770114 (lr=1.3462e-04) (hash(x)=7281348)
83
+ 77 train 8.808050 (lr=1.3636e-04) (hash(x)=6155821)
84
+ 78 train 8.625364 (lr=1.3811e-04) (hash(x)=5736241)
85
+ 79 train 8.525559 (lr=1.3986e-04) (hash(x)=5657123)
86
+ 80 train 8.465181 (lr=1.4161e-04) (hash(x)=6452095)
87
+ 81 train 8.402158 (lr=1.4336e-04) (hash(x)=5312267)
88
+ 82 train 8.274101 (lr=1.4510e-04) (hash(x)=5434493)
89
+ 83 train 8.474909 (lr=1.4685e-04) (hash(x)=6830813)
90
+ 84 train 8.312563 (lr=1.4860e-04) (hash(x)=4752771)
91
+ 85 train 8.389462 (lr=1.5035e-04) (hash(x)=6424933)
92
+ 86 train 8.288081 (lr=1.5210e-04) (hash(x)=5786517)
93
+ 87 train 8.339081 (lr=1.5385e-04) (hash(x)=6680023)
94
+ 88 train 8.251625 (lr=1.5559e-04) (hash(x)=6474729)
95
+ 89 train 8.267853 (lr=1.5734e-04) (hash(x)=7523279)
96
+ 90 train 8.057953 (lr=1.5909e-04) (hash(x)=5508768)
97
+ 91 train 8.231356 (lr=1.6084e-04) (hash(x)=6635925)
98
+ 92 train 8.114038 (lr=1.6259e-04) (hash(x)=6663801)
99
+ 93 train 8.085885 (lr=1.6434e-04) (hash(x)=5869441)
100
+ 94 train 8.135342 (lr=1.6608e-04) (hash(x)=6698878)
101
+ 95 train 8.047293 (lr=1.6783e-04) (hash(x)=6349004)
102
+ 96 train 8.049777 (lr=1.6958e-04) (hash(x)=4986137)
103
+ 97 train 7.831696 (lr=1.7133e-04) (hash(x)=6008454)
104
+ 98 train 7.897891 (lr=1.7308e-04) (hash(x)=6550770)
105
+ 99 train 8.033091 (lr=1.7483e-04) (hash(x)=7682741)
106
+ 100 val loss 7.9831
107
+ 100 val perplexity 2930.9805
108
+ 100 train 8.090453 (lr=1.7657e-04) (hash(x)=7892158)
109
+ 101 train 7.958868 (lr=1.7832e-04) (hash(x)=7794028)
110
+ 102 train 7.949870 (lr=1.8007e-04) (hash(x)=7093523)
111
+ 103 train 7.980931 (lr=1.8182e-04) (hash(x)=6166062)
112
+ 104 train 7.919570 (lr=1.8357e-04) (hash(x)=6645781)
113
+ 105 train 7.891222 (lr=1.8531e-04) (hash(x)=4811192)
114
+ 106 train 7.811481 (lr=1.8706e-04) (hash(x)=5970866)
115
+ 107 train 7.818060 (lr=1.8881e-04) (hash(x)=7822680)
116
+ 108 train 7.777383 (lr=1.9056e-04) (hash(x)=6317317)
117
+ 109 train 7.874724 (lr=1.9231e-04) (hash(x)=6431409)
118
+ 110 train 7.892319 (lr=1.9406e-04) (hash(x)=6653337)
119
+ 111 train 7.827804 (lr=1.9580e-04) (hash(x)=5323032)
120
+ 112 train 7.717324 (lr=1.9755e-04) (hash(x)=5125339)
121
+ 113 train 7.660126 (lr=1.9930e-04) (hash(x)=7363286)
122
+ 114 train 8.018383 (lr=2.0105e-04) (hash(x)=7670050)
123
+ 115 train 7.784950 (lr=2.0280e-04) (hash(x)=5616075)
124
+ 116 train 7.829693 (lr=2.0455e-04) (hash(x)=6593764)
125
+ 117 train 7.890211 (lr=2.0629e-04) (hash(x)=5672608)
126
+ 118 train 7.730969 (lr=2.0804e-04) (hash(x)=6828653)
127
+ 119 train 7.721272 (lr=2.0979e-04) (hash(x)=6839867)
128
+ 120 train 7.846440 (lr=2.1154e-04) (hash(x)=8426992)
129
+ 121 train 7.517550 (lr=2.1329e-04) (hash(x)=4916928)
130
+ 122 train 7.730700 (lr=2.1503e-04) (hash(x)=6484050)
131
+ 123 train 7.823503 (lr=2.1678e-04) (hash(x)=8353379)
132
+ 124 train 7.734229 (lr=2.1853e-04) (hash(x)=6552510)
133
+ 125 train 7.706458 (lr=2.2028e-04) (hash(x)=4838871)
134
+ 126 train 7.446773 (lr=2.2203e-04) (hash(x)=3021697)
135
+ 127 train 7.606452 (lr=2.2378e-04) (hash(x)=6795665)
136
+ 128 train 7.620957 (lr=2.2552e-04) (hash(x)=5486935)
137
+ 129 train 7.809128 (lr=2.2727e-04) (hash(x)=6678038)
138
+ 130 train 7.825309 (lr=2.2902e-04) (hash(x)=6657714)
139
+ 131 train 7.716088 (lr=2.3077e-04) (hash(x)=6112215)
140
+ 132 train 7.758881 (lr=2.3252e-04) (hash(x)=8353143)
141
+ 133 train 7.783720 (lr=2.3427e-04) (hash(x)=5915361)
142
+ 134 train 7.680995 (lr=2.3601e-04) (hash(x)=6254885)
143
+ 135 train 7.694855 (lr=2.3776e-04) (hash(x)=5771588)
144
+ 136 train 7.677476 (lr=2.3951e-04) (hash(x)=5561507)
145
+ 137 train 7.624990 (lr=2.4126e-04) (hash(x)=5688829)
146
+ 138 train 7.621283 (lr=2.4301e-04) (hash(x)=5371951)
147
+ 139 train 7.790312 (lr=2.4476e-04) (hash(x)=5443305)
148
+ 140 train 7.751131 (lr=2.4650e-04) (hash(x)=5016757)
149
+ 141 train 7.758002 (lr=2.4825e-04) (hash(x)=6901933)
150
+ 142 train 7.730201 (lr=2.5000e-04) (hash(x)=6019085)
151
+ 143 train 7.565399 (lr=2.5175e-04) (hash(x)=4294425)
152
+ 144 train 7.673890 (lr=2.5350e-04) (hash(x)=6095229)
153
+ 145 train 7.503168 (lr=2.5524e-04) (hash(x)=4363941)
154
+ 146 train 7.580531 (lr=2.5699e-04) (hash(x)=4693798)
155
+ 147 train 8.144917 (lr=2.5874e-04) (hash(x)=5486335)
156
+ 148 train 7.948389 (lr=2.6049e-04) (hash(x)=7907450)
157
+ 149 train 7.779182 (lr=2.6224e-04) (hash(x)=7295165)
158
+ 150 val loss 7.6655
159
+ 150 val perplexity 2133.4790
160
+ 150 train 7.703281 (lr=2.6399e-04) (hash(x)=6814026)
161
+ 151 train 7.792414 (lr=2.6573e-04) (hash(x)=6267424)
162
+ 152 train 7.802181 (lr=2.6748e-04) (hash(x)=8086437)
163
+ 153 train 7.851647 (lr=2.6923e-04) (hash(x)=5270452)
164
+ 154 train 7.695092 (lr=2.7098e-04) (hash(x)=6420820)
165
+ 155 train 7.593840 (lr=2.7273e-04) (hash(x)=6305297)
166
+ 156 train 7.737329 (lr=2.7448e-04) (hash(x)=7889849)
167
+ 157 train 8.117135 (lr=2.7622e-04) (hash(x)=6040077)
168
+ 158 train 8.112368 (lr=2.7797e-04) (hash(x)=5652497)
169
+ 159 train 7.692060 (lr=2.7972e-04) (hash(x)=7130251)
170
+ 160 train 7.550304 (lr=2.8147e-04) (hash(x)=6610177)
171
+ 161 train 7.612594 (lr=2.8322e-04) (hash(x)=6508350)
172
+ 162 train 7.769850 (lr=2.8497e-04) (hash(x)=6994983)
173
+ 163 train 7.727452 (lr=2.8671e-04) (hash(x)=5859576)
174
+ 164 train 7.284773 (lr=2.8846e-04) (hash(x)=5487065)
175
+ 165 train 7.532728 (lr=2.9021e-04) (hash(x)=4892065)
176
+ 166 train 7.427954 (lr=2.9196e-04) (hash(x)=4933674)
177
+ 167 train 7.690159 (lr=2.9371e-04) (hash(x)=7317289)
178
+ 168 train 7.792884 (lr=2.9545e-04) (hash(x)=6811522)
179
+ 169 train 7.619172 (lr=2.9720e-04) (hash(x)=5510218)
180
+ 170 train 7.783511 (lr=2.9895e-04) (hash(x)=7304235)
181
+ 171 train 7.934256 (lr=3.0070e-04) (hash(x)=6752265)
182
+ 172 train 7.629849 (lr=3.0245e-04) (hash(x)=5994476)
183
+ 173 train 7.600955 (lr=3.0420e-04) (hash(x)=6530867)
184
+ 174 train 7.702915 (lr=3.0594e-04) (hash(x)=5746260)
185
+ 175 train 7.658589 (lr=3.0769e-04) (hash(x)=5790078)
186
+ 176 train 7.674985 (lr=3.0944e-04) (hash(x)=6430135)
187
+ 177 train 7.746804 (lr=3.1119e-04) (hash(x)=7614023)
188
+ 178 train 7.730116 (lr=3.1294e-04) (hash(x)=5244861)
189
+ 179 train 7.712088 (lr=3.1469e-04) (hash(x)=5245146)
190
+ 180 train 7.603346 (lr=3.1643e-04) (hash(x)=6415565)
191
+ 181 train 7.655617 (lr=3.1818e-04) (hash(x)=5786972)
192
+ 182 train 7.691521 (lr=3.1993e-04) (hash(x)=6131703)
193
+ 183 train 7.712232 (lr=3.2168e-04) (hash(x)=6346459)
194
+ 184 train 7.754544 (lr=3.2343e-04) (hash(x)=6504131)
195
+ 185 train 7.744644 (lr=3.2517e-04) (hash(x)=5209234)
196
+ 186 train 7.656458 (lr=3.2692e-04) (hash(x)=6163272)
197
+ 187 train 7.648332 (lr=3.2867e-04) (hash(x)=5676768)
198
+ 188 train 7.715640 (lr=3.3042e-04) (hash(x)=6090561)
199
+ 189 train 7.867862 (lr=3.3217e-04) (hash(x)=5863092)
200
+ 190 train 7.778335 (lr=3.3392e-04) (hash(x)=7201312)
201
+ 191 train 7.853775 (lr=3.3566e-04) (hash(x)=7019489)
202
+ 192 train 7.619477 (lr=3.3741e-04) (hash(x)=6225608)
203
+ 193 train 7.712648 (lr=3.3916e-04) (hash(x)=5483019)
204
+ 194 train 7.652710 (lr=3.4091e-04) (hash(x)=5903258)
205
+ 195 train 7.691281 (lr=3.4266e-04) (hash(x)=5968716)
206
+ 196 train 7.687555 (lr=3.4441e-04) (hash(x)=5778017)
207
+ 197 train 7.410178 (lr=3.4615e-04) (hash(x)=7780194)
208
+ 198 train 7.622293 (lr=3.4790e-04) (hash(x)=6140998)
209
+ 199 train 7.609862 (lr=3.4965e-04) (hash(x)=3784321)
210
+ 200 val loss 7.6258
211
+ 200 val perplexity 2050.3394
212
+ 200 train 7.632768 (lr=3.5140e-04) (hash(x)=5860821)
213
+ 201 train 7.672125 (lr=3.5315e-04) (hash(x)=7774109)
214
+ 202 train 7.835864 (lr=3.5490e-04) (hash(x)=7997664)
215
+ 203 train 7.801472 (lr=3.5664e-04) (hash(x)=6576749)
216
+ 204 train 7.720698 (lr=3.5839e-04) (hash(x)=7266655)
217
+ 205 train 7.686074 (lr=3.6014e-04) (hash(x)=7072404)
218
+ 206 train 7.599856 (lr=3.6189e-04) (hash(x)=6626459)
219
+ 207 train 7.601377 (lr=3.6364e-04) (hash(x)=6556025)
220
+ 208 train 7.638923 (lr=3.6538e-04) (hash(x)=5942129)
221
+ 209 train 7.745492 (lr=3.6713e-04) (hash(x)=7325027)
222
+ 210 train 7.741173 (lr=3.6888e-04) (hash(x)=7322467)
223
+ 211 train 7.673203 (lr=3.7063e-04) (hash(x)=6373412)
224
+ 212 train 7.465663 (lr=3.7238e-04) (hash(x)=7332255)
225
+ 213 train 7.744501 (lr=3.7413e-04) (hash(x)=5262868)
226
+ 214 train 7.456234 (lr=3.7587e-04) (hash(x)=5688247)
227
+ 215 train 7.622649 (lr=3.7762e-04) (hash(x)=8055563)
228
+ 216 train 7.651843 (lr=3.7937e-04) (hash(x)=5352405)
229
+ 217 train 7.726893 (lr=3.8112e-04) (hash(x)=5932790)
230
+ 218 train 7.666943 (lr=3.8287e-04) (hash(x)=6249312)
231
+ 219 train 7.623081 (lr=3.8462e-04) (hash(x)=5680154)
232
+ 220 train 7.722910 (lr=3.8636e-04) (hash(x)=8156280)
233
+ 221 train 7.649165 (lr=3.8811e-04) (hash(x)=5914217)
234
+ 222 train 7.484366 (lr=3.8986e-04) (hash(x)=7120215)
235
+ 223 train 7.680924 (lr=3.9161e-04) (hash(x)=6424128)
236
+ 224 train 7.432400 (lr=3.9336e-04) (hash(x)=4390027)
237
+ 225 train 7.641574 (lr=3.9510e-04) (hash(x)=7804089)
238
+ 226 train 7.717457 (lr=3.9685e-04) (hash(x)=7130267)
239
+ 227 train 7.493531 (lr=3.9860e-04) (hash(x)=6490149)
240
+ 228 train 7.587654 (lr=4.0035e-04) (hash(x)=9276378)
241
+ 229 train 7.613081 (lr=4.0210e-04) (hash(x)=6283540)
242
+ 230 train 7.594581 (lr=4.0385e-04) (hash(x)=5781680)
243
+ 231 train 7.525056 (lr=4.0559e-04) (hash(x)=5728337)
244
+ 232 train 7.519223 (lr=4.0734e-04) (hash(x)=6278202)
245
+ 233 train 7.623920 (lr=4.0909e-04) (hash(x)=7400272)
246
+ 234 train 7.590850 (lr=4.1084e-04) (hash(x)=5795128)
247
+ 235 train 7.558988 (lr=4.1259e-04) (hash(x)=6473432)
248
+ 236 train 7.579780 (lr=4.1434e-04) (hash(x)=5201502)
249
+ 237 train 7.788270 (lr=4.1608e-04) (hash(x)=6922646)
250
+ 238 train 7.499532 (lr=4.1783e-04) (hash(x)=5484641)
251
+ 239 train 7.566568 (lr=4.1958e-04) (hash(x)=7277141)
252
+ 240 train 7.573974 (lr=4.2133e-04) (hash(x)=5604614)
253
+ 241 train 7.624212 (lr=4.2308e-04) (hash(x)=7443124)
254
+ 242 train 7.636985 (lr=4.2483e-04) (hash(x)=5601968)
255
+ 243 train 7.525299 (lr=4.2657e-04) (hash(x)=5632735)
256
+ 244 train 7.536751 (lr=4.2832e-04) (hash(x)=5594810)
257
+ 245 train 7.395719 (lr=4.3007e-04) (hash(x)=5881015)
258
+ 246 train 7.534366 (lr=4.3182e-04) (hash(x)=6865361)
259
+ 247 train 7.606976 (lr=4.3357e-04) (hash(x)=5809411)
260
+ 248 train 7.598826 (lr=4.3531e-04) (hash(x)=6293025)
261
+ 249 train 7.542069 (lr=4.3706e-04) (hash(x)=7232964)
262
+ 250 val loss 7.5649
263
+ 250 val perplexity 1929.2784
264
+ 250 train 7.486119 (lr=4.3881e-04) (hash(x)=5646619)
265
+ 251 train 7.601873 (lr=4.4056e-04) (hash(x)=6171829)
266
+ 252 train 7.676800 (lr=4.4231e-04) (hash(x)=7787552)
267
+ 253 train 7.609212 (lr=4.4406e-04) (hash(x)=5884456)
268
+ 254 train 7.627812 (lr=4.4580e-04) (hash(x)=6752162)
269
+ 255 train 7.541821 (lr=4.4755e-04) (hash(x)=5077031)
270
+ 256 train 7.698488 (lr=4.4930e-04) (hash(x)=6883481)
271
+ 257 train 7.584640 (lr=4.5105e-04) (hash(x)=6248895)
272
+ 258 train 7.457194 (lr=4.5280e-04) (hash(x)=6636615)
273
+ 259 train 7.646179 (lr=4.5455e-04) (hash(x)=6918004)
274
+ 260 train 7.561431 (lr=4.5629e-04) (hash(x)=7302874)
275
+ 261 train 7.669982 (lr=4.5804e-04) (hash(x)=6482466)
276
+ 262 train 7.498110 (lr=4.5979e-04) (hash(x)=7298317)
277
+ 263 train 7.533354 (lr=4.6154e-04) (hash(x)=6588573)
278
+ 264 train 7.526618 (lr=4.6329e-04) (hash(x)=6229630)
279
+ 265 train 7.600528 (lr=4.6503e-04) (hash(x)=6136939)
280
+ 266 train 7.357875 (lr=4.6678e-04) (hash(x)=6394645)
281
+ 267 train 7.592115 (lr=4.6853e-04) (hash(x)=7040157)
282
+ 268 train 7.628303 (lr=4.7028e-04) (hash(x)=7041858)
283
+ 269 train 7.529849 (lr=4.7203e-04) (hash(x)=6455595)
284
+ 270 train 7.459441 (lr=4.7378e-04) (hash(x)=6480443)
285
+ 271 train 7.647004 (lr=4.7552e-04) (hash(x)=4122983)
286
+ 272 train 7.458611 (lr=4.7727e-04) (hash(x)=7672190)
287
+ 273 train 7.485595 (lr=4.7902e-04) (hash(x)=4724052)
288
+ 274 train 7.437414 (lr=4.8077e-04) (hash(x)=6654267)
289
+ 275 train 7.420813 (lr=4.8252e-04) (hash(x)=6820797)
290
+ 276 train 7.512196 (lr=4.8427e-04) (hash(x)=6458662)
291
+ 277 train 7.917709 (lr=4.8601e-04) (hash(x)=7477594)
292
+ 278 train 7.554106 (lr=4.8776e-04) (hash(x)=5876574)
293
+ 279 train 7.277621 (lr=4.8951e-04) (hash(x)=6146305)
294
+ 280 train 7.454937 (lr=4.9126e-04) (hash(x)=6162080)
295
+ 281 train 7.411793 (lr=4.9301e-04) (hash(x)=6814831)
296
+ 282 train 7.541778 (lr=4.9476e-04) (hash(x)=7138295)
297
+ 283 train 7.317982 (lr=4.9650e-04) (hash(x)=5548298)
298
+ 284 train 7.428448 (lr=4.9825e-04) (hash(x)=6265078)
299
+ 285 train 7.801420 (lr=5.0000e-04) (hash(x)=7609416)
300
+ 286 train 7.483164 (lr=5.0000e-04) (hash(x)=6909367)
301
+ 287 train 7.597700 (lr=5.0000e-04) (hash(x)=7574342)
302
+ 288 train 7.409695 (lr=4.9999e-04) (hash(x)=8859748)
303
+ 289 train 7.362408 (lr=4.9998e-04) (hash(x)=5906937)
304
+ 290 train 7.386526 (lr=4.9997e-04) (hash(x)=7243024)
305
+ 291 train 7.502868 (lr=4.9995e-04) (hash(x)=8013471)
306
+ 292 train 7.416108 (lr=4.9992e-04) (hash(x)=5659461)
307
+ 293 train 7.455505 (lr=4.9989e-04) (hash(x)=6146213)
308
+ 294 train 7.491917 (lr=4.9986e-04) (hash(x)=6602917)
309
+ 295 train 7.399060 (lr=4.9982e-04) (hash(x)=6709902)
310
+ 296 train 7.526380 (lr=4.9978e-04) (hash(x)=8154341)
311
+ 297 train 7.667249 (lr=4.9974e-04) (hash(x)=6402571)
312
+ 298 train 7.428350 (lr=4.9969e-04) (hash(x)=5605466)
313
+ 299 train 7.383250 (lr=4.9963e-04) (hash(x)=7683614)
314
+ 300 val loss 7.4516
315
+ 300 val perplexity 1722.6920
316
+ 300 train 7.500629 (lr=4.9957e-04) (hash(x)=5421572)
317
+ 301 train 7.343226 (lr=4.9951e-04) (hash(x)=5453026)
318
+ 302 train 7.402356 (lr=4.9944e-04) (hash(x)=6437746)
319
+ 303 train 7.311347 (lr=4.9937e-04) (hash(x)=6260017)
320
+ 304 train 7.410797 (lr=4.9929e-04) (hash(x)=6119053)
321
+ 305 train 7.575393 (lr=4.9921e-04) (hash(x)=6591466)
322
+ 306 train 7.485226 (lr=4.9913e-04) (hash(x)=5652701)
323
+ 307 train 7.442460 (lr=4.9904e-04) (hash(x)=5501472)
324
+ 308 train 7.624926 (lr=4.9895e-04) (hash(x)=6160695)
325
+ 309 train 7.380640 (lr=4.9885e-04) (hash(x)=5871698)
326
+ 310 train 7.343381 (lr=4.9875e-04) (hash(x)=7301184)
327
+ 311 train 7.274087 (lr=4.9864e-04) (hash(x)=4282052)
328
+ 312 train 7.481493 (lr=4.9853e-04) (hash(x)=5679330)
329
+ 313 train 7.329782 (lr=4.9841e-04) (hash(x)=7824660)
330
+ 314 train 7.376566 (lr=4.9829e-04) (hash(x)=4857318)
331
+ 315 train 7.313642 (lr=4.9817e-04) (hash(x)=5954950)
332
+ 316 train 7.372428 (lr=4.9804e-04) (hash(x)=5758178)
333
+ 317 train 7.392262 (lr=4.9791e-04) (hash(x)=5484656)
334
+ 318 train 7.383463 (lr=4.9777e-04) (hash(x)=6818913)
335
+ 319 train 7.589435 (lr=4.9763e-04) (hash(x)=7031328)
336
+ 320 train 7.416356 (lr=4.9749e-04) (hash(x)=6837621)
337
+ 321 train 7.457081 (lr=4.9734e-04) (hash(x)=9157069)
338
+ 322 train 7.698480 (lr=4.9718e-04) (hash(x)=6509536)
339
+ 323 train 7.392669 (lr=4.9702e-04) (hash(x)=5996883)
340
+ 324 train 7.647024 (lr=4.9686e-04) (hash(x)=6850868)
341
+ 325 train 7.569314 (lr=4.9670e-04) (hash(x)=7645447)
342
+ 326 train 7.580283 (lr=4.9652e-04) (hash(x)=6717515)
343
+ 327 train 7.708749 (lr=4.9635e-04) (hash(x)=6336164)
344
+ 328 train 7.413807 (lr=4.9617e-04) (hash(x)=6897445)
345
+ 329 train 7.333851 (lr=4.9598e-04) (hash(x)=5224240)
346
+ 330 train 7.483329 (lr=4.9580e-04) (hash(x)=6612945)
347
+ 331 train 7.923858 (lr=4.9560e-04) (hash(x)=5447801)
348
+ 332 train 7.465648 (lr=4.9541e-04) (hash(x)=7308579)
349
+ 333 train 7.358824 (lr=4.9521e-04) (hash(x)=4155598)
350
+ 334 train 7.355170 (lr=4.9500e-04) (hash(x)=6106773)
351
+ 335 train 7.404686 (lr=4.9479e-04) (hash(x)=7482744)
352
+ 336 train 7.500164 (lr=4.9458e-04) (hash(x)=6310687)
353
+ 337 train 7.362833 (lr=4.9436e-04) (hash(x)=4733414)
354
+ 338 train 7.462912 (lr=4.9414e-04) (hash(x)=6979734)
355
+ 339 train 7.269539 (lr=4.9391e-04) (hash(x)=7806284)
356
+ 340 train 7.276829 (lr=4.9368e-04) (hash(x)=5080536)
357
+ 341 train 7.495099 (lr=4.9344e-04) (hash(x)=7236478)
358
+ 342 train 7.432468 (lr=4.9320e-04) (hash(x)=6619318)
359
+ 343 train 7.337922 (lr=4.9296e-04) (hash(x)=4656255)
360
+ 344 train 7.276386 (lr=4.9271e-04) (hash(x)=5546888)
361
+ 345 train 7.748324 (lr=4.9246e-04) (hash(x)=9298325)
362
+ 346 train 7.459085 (lr=4.9220e-04) (hash(x)=5994232)
363
+ 347 train 7.417453 (lr=4.9194e-04) (hash(x)=6287524)
364
+ 348 train 7.540259 (lr=4.9168e-04) (hash(x)=6200049)
365
+ 349 train 7.444436 (lr=4.9141e-04) (hash(x)=7895970)
366
+ 350 val loss 7.3646
367
+ 350 val perplexity 1579.1465
368
+ 350 train 7.350061 (lr=4.9114e-04) (hash(x)=6051847)
369
+ 351 train 7.506237 (lr=4.9086e-04) (hash(x)=8847992)
370
+ 352 train 7.880729 (lr=4.9058e-04) (hash(x)=5123883)
371
+ 353 train 7.464824 (lr=4.9029e-04) (hash(x)=6338998)
372
+ 354 train 7.467878 (lr=4.9000e-04) (hash(x)=5765866)
373
+ 355 train 7.813435 (lr=4.8971e-04) (hash(x)=7104011)
374
+ 356 train 7.582506 (lr=4.8941e-04) (hash(x)=7438267)
375
+ 357 train 7.317100 (lr=4.8911e-04) (hash(x)=5573214)
376
+ 358 train 7.570335 (lr=4.8880e-04) (hash(x)=6897283)
377
+ 359 train 7.362119 (lr=4.8849e-04) (hash(x)=6882897)
378
+ 360 train 7.479269 (lr=4.8818e-04) (hash(x)=5343615)
379
+ 361 train 7.388541 (lr=4.8786e-04) (hash(x)=5854242)
380
+ 362 train 7.301569 (lr=4.8754e-04) (hash(x)=6317324)
381
+ 363 train 7.239825 (lr=4.8721e-04) (hash(x)=6972494)
382
+ 364 train 7.318973 (lr=4.8688e-04) (hash(x)=5661167)
383
+ 365 train 7.332743 (lr=4.8654e-04) (hash(x)=6796805)
384
+ 366 train 7.158183 (lr=4.8620e-04) (hash(x)=5817209)
385
+ 367 train 7.233025 (lr=4.8586e-04) (hash(x)=6139120)
386
+ 368 train 7.402174 (lr=4.8551e-04) (hash(x)=6121296)
387
+ 369 train 7.376255 (lr=4.8516e-04) (hash(x)=6486796)
388
+ 370 train 7.291081 (lr=4.8481e-04) (hash(x)=5514645)
389
+ 371 train 7.235853 (lr=4.8445e-04) (hash(x)=5578797)
390
+ 372 train 7.232906 (lr=4.8408e-04) (hash(x)=7651738)
391
+ 373 train 7.441844 (lr=4.8372e-04) (hash(x)=6895326)
392
+ 374 train 7.463004 (lr=4.8334e-04) (hash(x)=5490256)
393
+ 375 train 7.318242 (lr=4.8297e-04) (hash(x)=7783160)
394
+ 376 train 7.523637 (lr=4.8259e-04) (hash(x)=5608424)
395
+ 377 train 7.289732 (lr=4.8220e-04) (hash(x)=4867625)
396
+ 378 train 7.199661 (lr=4.8182e-04) (hash(x)=4655472)
397
+ 379 train 7.258174 (lr=4.8142e-04) (hash(x)=6631123)
398
+ 380 train 6.930998 (lr=4.8103e-04) (hash(x)=5751990)
399
+ 381 train 7.244713 (lr=4.8063e-04) (hash(x)=6710965)
400
+ 382 train 7.194671 (lr=4.8022e-04) (hash(x)=5293800)
401
+ 383 train 7.221773 (lr=4.7982e-04) (hash(x)=6294657)
402
+ 384 train 7.393067 (lr=4.7940e-04) (hash(x)=6999538)
403
+ 385 train 7.381361 (lr=4.7899e-04) (hash(x)=5443606)
404
+ 386 train 7.358816 (lr=4.7857e-04) (hash(x)=5650401)
405
+ 387 train 7.354572 (lr=4.7815e-04) (hash(x)=7035770)
406
+ 388 train 7.304999 (lr=4.7772e-04) (hash(x)=6678271)
407
+ 389 train 7.499672 (lr=4.7729e-04) (hash(x)=4899541)
408
+ 390 train 7.237353 (lr=4.7685e-04) (hash(x)=5359260)
409
+ 391 train 7.232718 (lr=4.7641e-04) (hash(x)=6227120)
410
+ 392 train 7.301436 (lr=4.7597e-04) (hash(x)=6643916)
411
+ 393 train 7.356061 (lr=4.7552e-04) (hash(x)=5981648)
412
+ 394 train 7.321610 (lr=4.7507e-04) (hash(x)=6467540)
413
+ 395 train 7.252539 (lr=4.7462e-04) (hash(x)=5558993)
414
+ 396 train 7.430574 (lr=4.7416e-04) (hash(x)=6698959)
415
+ 397 train 7.227236 (lr=4.7369e-04) (hash(x)=7015671)
416
+ 398 train 7.248781 (lr=4.7323e-04) (hash(x)=6800890)
417
+ 399 train 7.412173 (lr=4.7276e-04) (hash(x)=5931234)
418
+ 400 val loss 7.3167
419
+ 400 val perplexity 1505.2485
420
+ 400 train 7.409585 (lr=4.7228e-04) (hash(x)=3995070)
421
+ 401 train 7.332630 (lr=4.7181e-04) (hash(x)=4582159)
422
+ 402 train 7.372652 (lr=4.7132e-04) (hash(x)=7081906)
423
+ 403 train 7.296834 (lr=4.7084e-04) (hash(x)=5029472)
424
+ 404 train 7.319156 (lr=4.7035e-04) (hash(x)=3946232)
425
+ 405 train 7.222072 (lr=4.6986e-04) (hash(x)=5434102)
426
+ 406 train 7.960849 (lr=4.6936e-04) (hash(x)=5340646)
427
+ 407 train 7.133418 (lr=4.6886e-04) (hash(x)=6810638)
428
+ 408 train 7.384606 (lr=4.6835e-04) (hash(x)=6768026)
429
+ 409 train 7.375993 (lr=4.6785e-04) (hash(x)=6521562)
430
+ 410 train 7.327439 (lr=4.6733e-04) (hash(x)=4429434)
431
+ 411 train 7.394525 (lr=4.6682e-04) (hash(x)=7226932)
432
+ 412 train 7.169340 (lr=4.6630e-04) (hash(x)=5065932)
433
+ 413 train 7.331411 (lr=4.6578e-04) (hash(x)=6957849)
434
+ 414 train 7.329439 (lr=4.6525e-04) (hash(x)=6431457)
435
+ 415 train 7.399717 (lr=4.6472e-04) (hash(x)=5334281)
436
+ 416 train 7.406147 (lr=4.6418e-04) (hash(x)=5471747)
437
+ 417 train 7.499147 (lr=4.6365e-04) (hash(x)=5887983)
438
+ 418 train 7.404335 (lr=4.6311e-04) (hash(x)=5075909)
439
+ 419 train 7.202665 (lr=4.6256e-04) (hash(x)=6393576)
440
+ 420 train 7.296270 (lr=4.6201e-04) (hash(x)=6608979)
441
+ 421 train 7.728078 (lr=4.6146e-04) (hash(x)=5057959)
442
+ 422 train 7.663626 (lr=4.6090e-04) (hash(x)=6414858)
443
+ 423 train 7.435847 (lr=4.6034e-04) (hash(x)=8595271)
444
+ 424 train 7.300275 (lr=4.5978e-04) (hash(x)=6500949)
445
+ 425 train 7.502699 (lr=4.5921e-04) (hash(x)=8395167)
446
+ 426 train 7.632969 (lr=4.5864e-04) (hash(x)=6288420)
447
+ 427 train 7.425904 (lr=4.5807e-04) (hash(x)=7634417)
448
+ 428 train 7.306537 (lr=4.5749e-04) (hash(x)=6795561)
449
+ 429 train 7.204391 (lr=4.5691e-04) (hash(x)=5257771)
450
+ 430 train 7.367026 (lr=4.5633e-04) (hash(x)=7506860)
451
+ 431 train 7.397141 (lr=4.5574e-04) (hash(x)=7674238)
452
+ 432 train 7.503457 (lr=4.5515e-04) (hash(x)=5353794)
453
+ 433 train 7.430412 (lr=4.5455e-04) (hash(x)=5300555)
454
+ 434 train 7.117725 (lr=4.5396e-04) (hash(x)=5564419)
455
+ 435 train 7.385005 (lr=4.5335e-04) (hash(x)=6066139)
456
+ 436 train 7.363755 (lr=4.5275e-04) (hash(x)=6863709)
457
+ 437 train 7.352476 (lr=4.5214e-04) (hash(x)=6688212)
458
+ 438 train 7.504986 (lr=4.5153e-04) (hash(x)=7095090)
459
+ 439 train 7.606781 (lr=4.5091e-04) (hash(x)=6642404)
460
+ 440 train 7.314439 (lr=4.5029e-04) (hash(x)=6539227)
461
+ 441 train 7.305208 (lr=4.4967e-04) (hash(x)=4961570)
462
+ 442 train 7.246849 (lr=4.4905e-04) (hash(x)=6606391)
463
+ 443 train 7.048862 (lr=4.4842e-04) (hash(x)=5105190)
464
+ 444 train 7.646637 (lr=4.4778e-04) (hash(x)=5695997)
465
+ 445 train 7.336450 (lr=4.4715e-04) (hash(x)=7155991)
466
+ 446 train 7.309959 (lr=4.4651e-04) (hash(x)=7249727)
467
+ 447 train 7.295984 (lr=4.4587e-04) (hash(x)=5132972)
468
+ 448 train 7.292876 (lr=4.4522e-04) (hash(x)=5381802)
469
+ 449 train 7.317535 (lr=4.4457e-04) (hash(x)=6269946)
470
+ 450 val loss 7.2890
471
+ 450 val perplexity 1464.1219
472
+ 450 train 7.271142 (lr=4.4392e-04) (hash(x)=4976762)
473
+ 451 train 7.361707 (lr=4.4326e-04) (hash(x)=5396133)
474
+ 452 train 7.327783 (lr=4.4260e-04) (hash(x)=6173473)
475
+ 453 train 7.296171 (lr=4.4194e-04) (hash(x)=6394083)
476
+ 454 train 7.372822 (lr=4.4128e-04) (hash(x)=6436030)
477
+ 455 train 7.211841 (lr=4.4061e-04) (hash(x)=5237804)
478
+ 456 train 7.193060 (lr=4.3994e-04) (hash(x)=6540397)
479
+ 457 train 7.444540 (lr=4.3926e-04) (hash(x)=5216638)
480
+ 458 train 7.271979 (lr=4.3858e-04) (hash(x)=5750537)
481
+ 459 train 7.259324 (lr=4.3790e-04) (hash(x)=6735565)
482
+ 460 train 7.287656 (lr=4.3722e-04) (hash(x)=7297488)
483
+ 461 train 7.463627 (lr=4.3653e-04) (hash(x)=5983575)
484
+ 462 train 7.290742 (lr=4.3584e-04) (hash(x)=4808543)
485
+ 463 train 7.405876 (lr=4.3515e-04) (hash(x)=5831619)
486
+ 464 train 7.156505 (lr=4.3445e-04) (hash(x)=6280052)
487
+ 465 train 7.171041 (lr=4.3375e-04) (hash(x)=9109083)
488
+ 466 train 7.431561 (lr=4.3305e-04) (hash(x)=5848604)
489
+ 467 train 7.372941 (lr=4.3234e-04) (hash(x)=6912967)
490
+ 468 train 7.553815 (lr=4.3163e-04) (hash(x)=6615365)
491
+ 469 train 7.607442 (lr=4.3092e-04) (hash(x)=5965674)
492
+ 470 train 7.455511 (lr=4.3020e-04) (hash(x)=6879965)
493
+ 471 train 7.419261 (lr=4.2948e-04) (hash(x)=5822533)
494
+ 472 train 7.278405 (lr=4.2876e-04) (hash(x)=5794462)
495
+ 473 train 7.372121 (lr=4.2804e-04) (hash(x)=5268216)
496
+ 474 train 7.190698 (lr=4.2731e-04) (hash(x)=7058396)
497
+ 475 train 7.303502 (lr=4.2658e-04) (hash(x)=6531900)
498
+ 476 train 7.335746 (lr=4.2585e-04) (hash(x)=5878912)
499
+ 477 train 7.136664 (lr=4.2511e-04) (hash(x)=5426499)
500
+ 478 train 7.347939 (lr=4.2437e-04) (hash(x)=8107004)
501
+ 479 train 8.073548 (lr=4.2363e-04) (hash(x)=7455372)
502
+ 480 train 7.998034 (lr=4.2289e-04) (hash(x)=7727959)
503
+ 481 train 7.919175 (lr=4.2214e-04) (hash(x)=8319119)
504
+ 482 train 7.831254 (lr=4.2139e-04) (hash(x)=7581656)
505
+ 483 train 7.897698 (lr=4.2064e-04) (hash(x)=7592128)
506
+ 484 train 7.883114 (lr=4.1988e-04) (hash(x)=10506633)
507
+ 485 train 7.835151 (lr=4.1912e-04) (hash(x)=8704170)
508
+ 486 train 7.658167 (lr=4.1836e-04) (hash(x)=7461138)
509
+ 487 train 7.588099 (lr=4.1759e-04) (hash(x)=7165037)
510
+ 488 train 7.634685 (lr=4.1683e-04) (hash(x)=4976998)
511
+ 489 train 7.399677 (lr=4.1606e-04) (hash(x)=7678000)
512
+ 490 train 7.460315 (lr=4.1529e-04) (hash(x)=6299409)
513
+ 491 train 7.390481 (lr=4.1451e-04) (hash(x)=7414788)
514
+ 492 train 7.459745 (lr=4.1373e-04) (hash(x)=5063061)
515
+ 493 train 7.374101 (lr=4.1295e-04) (hash(x)=7326216)
516
+ 494 train 7.538434 (lr=4.1217e-04) (hash(x)=7492012)
517
+ 495 train 7.500280 (lr=4.1138e-04) (hash(x)=5857162)
518
+ 496 train 7.296118 (lr=4.1059e-04) (hash(x)=6169001)
519
+ 497 train 9.308700 (lr=4.0980e-04) (hash(x)=14715669)
520
+ 498 train 12.223278 (lr=4.0901e-04) (hash(x)=18256706)
521
+ 499 train 11.954405 (lr=4.0821e-04) (hash(x)=16950560)
522
+ 500 val loss 7.3852
523
+ 500 val perplexity 1611.9194
524
+ 500 train 8.109703 (lr=4.0741e-04) (hash(x)=6113563)
525
+ 501 train 7.068744 (lr=4.0661e-04) (hash(x)=6159125)
526
+ 502 train 7.454783 (lr=4.0581e-04) (hash(x)=5482303)
527
+ 503 train 7.484365 (lr=4.0500e-04) (hash(x)=5908442)
528
+ 504 train 7.354419 (lr=4.0419e-04) (hash(x)=4985045)
529
+ 505 train 7.251081 (lr=4.0338e-04) (hash(x)=6668887)
530
+ 506 train 7.330905 (lr=4.0256e-04) (hash(x)=6123910)
531
+ 507 train 7.420095 (lr=4.0175e-04) (hash(x)=7662975)
532
+ 508 train 7.364473 (lr=4.0093e-04) (hash(x)=7191898)
533
+ 509 train 7.303823 (lr=4.0011e-04) (hash(x)=6669539)
534
+ 510 train 7.355706 (lr=3.9928e-04) (hash(x)=6557481)
535
+ 511 train 7.248978 (lr=3.9846e-04) (hash(x)=5015998)
536
+ 512 train 7.317575 (lr=3.9763e-04) (hash(x)=4964540)
537
+ 513 train 7.176949 (lr=3.9680e-04) (hash(x)=5458386)
538
+ 514 train 7.554805 (lr=3.9596e-04) (hash(x)=6077616)
539
+ 515 train 7.274286 (lr=3.9513e-04) (hash(x)=4849847)
540
+ 516 train 7.352317 (lr=3.9429e-04) (hash(x)=5431422)
541
+ 517 train 7.531367 (lr=3.9345e-04) (hash(x)=5125562)
542
+ 518 train 7.447542 (lr=3.9260e-04) (hash(x)=5587715)
543
+ 519 train 7.502025 (lr=3.9176e-04) (hash(x)=6012047)
544
+ 520 train 7.332932 (lr=3.9091e-04) (hash(x)=5668436)
545
+ 521 train 7.344112 (lr=3.9006e-04) (hash(x)=6079364)
546
+ 522 train 7.462158 (lr=3.8921e-04) (hash(x)=5200504)
547
+ 523 train 7.190444 (lr=3.8836e-04) (hash(x)=6030395)
548
+ 524 train 7.433517 (lr=3.8750e-04) (hash(x)=6855319)
549
+ 525 train 7.204855 (lr=3.8664e-04) (hash(x)=6781762)
550
+ 526 train 7.350496 (lr=3.8578e-04) (hash(x)=6717939)
551
+ 527 train 7.349863 (lr=3.8492e-04) (hash(x)=5238325)
552
+ 528 train 7.255769 (lr=3.8405e-04) (hash(x)=5808069)
553
+ 529 train 7.575129 (lr=3.8319e-04) (hash(x)=8611144)
554
+ 530 train 7.260050 (lr=3.8232e-04) (hash(x)=6028738)
555
+ 531 train 7.298766 (lr=3.8145e-04) (hash(x)=6901605)
556
+ 532 train 7.252521 (lr=3.8057e-04) (hash(x)=6723307)
557
+ 533 train 7.340546 (lr=3.7970e-04) (hash(x)=5499337)
558
+ 534 train 7.275239 (lr=3.7882e-04) (hash(x)=6586372)
559
+ 535 train 7.402943 (lr=3.7794e-04) (hash(x)=7134874)
560
+ 536 train 7.293164 (lr=3.7706e-04) (hash(x)=8415623)
561
+ 537 train 7.318466 (lr=3.7618e-04) (hash(x)=6149038)
562
+ 538 train 7.356337 (lr=3.7529e-04) (hash(x)=6078325)
563
+ 539 train 7.327504 (lr=3.7440e-04) (hash(x)=6609760)
564
+ 540 train 7.226675 (lr=3.7351e-04) (hash(x)=6185253)
565
+ 541 train 7.624069 (lr=3.7262e-04) (hash(x)=6652627)
566
+ 542 train 7.410195 (lr=3.7173e-04) (hash(x)=7758536)
567
+ 543 train 7.383860 (lr=3.7084e-04) (hash(x)=8618579)
568
+ 544 train 7.341766 (lr=3.6994e-04) (hash(x)=6710226)
569
+ 545 train 7.210335 (lr=3.6904e-04) (hash(x)=5896892)
570
+ 546 train 7.228855 (lr=3.6814e-04) (hash(x)=6332995)
571
+ 547 train 7.615961 (lr=3.6724e-04) (hash(x)=8945854)
572
+ 548 train 7.197575 (lr=3.6633e-04) (hash(x)=4329129)
573
+ 549 train 7.258995 (lr=3.6543e-04) (hash(x)=6775103)
574
+ 550 val loss 7.2793
575
+ 550 val perplexity 1449.9744
576
+ 550 train 7.363537 (lr=3.6452e-04) (hash(x)=4465735)
577
+ 551 train 7.254971 (lr=3.6361e-04) (hash(x)=6200024)
578
+ 552 train 7.203400 (lr=3.6270e-04) (hash(x)=5520748)
579
+ 553 train 7.216261 (lr=3.6179e-04) (hash(x)=5238139)
580
+ 554 train 7.461508 (lr=3.6088e-04) (hash(x)=7639817)
581
+ 555 train 7.217396 (lr=3.5996e-04) (hash(x)=5780948)
582
+ 556 train 7.360391 (lr=3.5904e-04) (hash(x)=6733444)
583
+ 557 train 7.305562 (lr=3.5812e-04) (hash(x)=6168407)
584
+ 558 train 7.351892 (lr=3.5720e-04) (hash(x)=6680197)
585
+ 559 train 7.215224 (lr=3.5628e-04) (hash(x)=3775881)
586
+ 560 train 7.074762 (lr=3.5536e-04) (hash(x)=5784096)
587
+ 561 train 7.505771 (lr=3.5443e-04) (hash(x)=6848709)
588
+ 562 train 7.553386 (lr=3.5350e-04) (hash(x)=7449019)
589
+ 563 train 7.653483 (lr=3.5257e-04) (hash(x)=6478113)
590
+ 564 train 7.305277 (lr=3.5164e-04) (hash(x)=5925744)
591
+ 565 train 7.367736 (lr=3.5071e-04) (hash(x)=6483914)
592
+ 566 train 7.258492 (lr=3.4978e-04) (hash(x)=6010847)
593
+ 567 train 7.337683 (lr=3.4885e-04) (hash(x)=5927795)
594
+ 568 train 7.333642 (lr=3.4791e-04) (hash(x)=5896332)
595
+ 569 train 7.307150 (lr=3.4697e-04) (hash(x)=8438033)
596
+ 570 train 7.391054 (lr=3.4603e-04) (hash(x)=6081660)
597
+ 571 train 7.320904 (lr=3.4509e-04) (hash(x)=5684788)
598
+ 572 train 7.210525 (lr=3.4415e-04) (hash(x)=6272055)
599
+ 573 train 7.157257 (lr=3.4321e-04) (hash(x)=5710182)
600
+ 574 train 7.280805 (lr=3.4227e-04) (hash(x)=6219616)
601
+ 575 train 7.237747 (lr=3.4132e-04) (hash(x)=6160063)
602
+ 576 train 7.262324 (lr=3.4037e-04) (hash(x)=5710809)
603
+ 577 train 7.014307 (lr=3.3943e-04) (hash(x)=5955100)
604
+ 578 train 7.025233 (lr=3.3848e-04) (hash(x)=5356088)
605
+ 579 train 7.067662 (lr=3.3753e-04) (hash(x)=6672309)
606
+ 580 train 7.278050 (lr=3.3657e-04) (hash(x)=5841684)
607
+ 581 train 7.612208 (lr=3.3562e-04) (hash(x)=9001564)
608
+ 582 train 8.155598 (lr=3.3467e-04) (hash(x)=6897178)
609
+ 583 train 7.391499 (lr=3.3371e-04) (hash(x)=5454654)
610
+ 584 train 7.158257 (lr=3.3276e-04) (hash(x)=6391454)
611
+ 585 train 7.242211 (lr=3.3180e-04) (hash(x)=5174062)
612
+ 586 train 7.359061 (lr=3.3084e-04) (hash(x)=5197924)
613
+ 587 train 7.319856 (lr=3.2988e-04) (hash(x)=6199203)
614
+ 588 train 7.152327 (lr=3.2892e-04) (hash(x)=5786498)
615
+ 589 train 7.270411 (lr=3.2796e-04) (hash(x)=8352747)
616
+ 590 train 7.553372 (lr=3.2700e-04) (hash(x)=4829700)
617
+ 591 train 7.089417 (lr=3.2603e-04) (hash(x)=4547106)
618
+ 592 train 7.379088 (lr=3.2507e-04) (hash(x)=7686831)
619
+ 593 train 7.255884 (lr=3.2410e-04) (hash(x)=6433099)
620
+ 594 train 7.126493 (lr=3.2313e-04) (hash(x)=4826460)
621
+ 595 train 7.209390 (lr=3.2217e-04) (hash(x)=6424428)
622
+ 596 train 7.087097 (lr=3.2120e-04) (hash(x)=6073605)
623
+ 597 train 7.446998 (lr=3.2023e-04) (hash(x)=5975712)
624
+ 598 train 7.053827 (lr=3.1926e-04) (hash(x)=7030029)
625
+ 599 train 7.345481 (lr=3.1829e-04) (hash(x)=5458532)
626
+ 600 val loss 7.2723
627
+ 600 val perplexity 1439.8539
628
+ 600 train 7.193815 (lr=3.1732e-04) (hash(x)=7823919)
629
+ 601 train 7.623002 (lr=3.1634e-04) (hash(x)=7002774)
630
+ 602 train 7.425569 (lr=3.1537e-04) (hash(x)=7056399)
631
+ 603 train 7.625144 (lr=3.1440e-04) (hash(x)=8206173)
632
+ 604 train 7.478229 (lr=3.1342e-04) (hash(x)=8034846)
633
+ 605 train 7.757761 (lr=3.1244e-04) (hash(x)=7784971)
634
+ 606 train 7.867619 (lr=3.1147e-04) (hash(x)=7118076)
635
+ 607 train 7.218085 (lr=3.1049e-04) (hash(x)=6001843)
636
+ 608 train 7.324508 (lr=3.0951e-04) (hash(x)=5582437)
637
+ 609 train 7.424302 (lr=3.0853e-04) (hash(x)=5698642)
638
+ 610 train 7.278915 (lr=3.0756e-04) (hash(x)=6615301)
639
+ 611 train 7.279538 (lr=3.0658e-04) (hash(x)=6600751)
640
+ 612 train 7.184999 (lr=3.0559e-04) (hash(x)=6880955)
641
+ 613 train 7.259321 (lr=3.0461e-04) (hash(x)=5804610)
642
+ 614 train 7.177356 (lr=3.0363e-04) (hash(x)=6263347)
643
+ 615 train 7.155849 (lr=3.0265e-04) (hash(x)=5162652)
644
+ 616 train 7.209515 (lr=3.0167e-04) (hash(x)=6055710)
645
+ 617 train 7.168361 (lr=3.0068e-04) (hash(x)=5297006)
646
+ 618 train 7.190626 (lr=2.9970e-04) (hash(x)=5923648)
647
+ 619 train 7.204132 (lr=2.9872e-04) (hash(x)=6728323)
648
+ 620 train 7.214056 (lr=2.9773e-04) (hash(x)=6660239)
649
+ 621 train 7.162459 (lr=2.9675e-04) (hash(x)=3964882)
650
+ 622 train 7.224735 (lr=2.9576e-04) (hash(x)=5868942)
651
+ 623 train 7.143094 (lr=2.9477e-04) (hash(x)=6735993)
652
+ 624 train 7.470279 (lr=2.9379e-04) (hash(x)=5926190)
653
+ 625 train 7.164414 (lr=2.9280e-04) (hash(x)=5150316)
654
+ 626 train 7.176792 (lr=2.9181e-04) (hash(x)=7571467)
655
+ 627 train 7.078360 (lr=2.9083e-04) (hash(x)=5220448)
656
+ 628 train 7.119508 (lr=2.8984e-04) (hash(x)=6790397)
657
+ 629 train 7.551075 (lr=2.8885e-04) (hash(x)=6971289)
658
+ 630 train 7.489536 (lr=2.8786e-04) (hash(x)=7123359)
659
+ 631 train 7.285959 (lr=2.8687e-04) (hash(x)=5012069)
660
+ 632 train 7.283583 (lr=2.8589e-04) (hash(x)=6183574)
661
+ 633 train 7.184519 (lr=2.8490e-04) (hash(x)=6576136)
662
+ 634 train 7.193376 (lr=2.8391e-04) (hash(x)=6157346)
663
+ 635 train 7.188956 (lr=2.8292e-04) (hash(x)=6077647)
664
+ 636 train 7.419818 (lr=2.8193e-04) (hash(x)=8369488)
665
+ 637 train 7.209139 (lr=2.8094e-04) (hash(x)=6032381)
666
+ 638 train 7.156520 (lr=2.7995e-04) (hash(x)=5924756)
667
+ 639 train 7.175603 (lr=2.7896e-04) (hash(x)=8410716)
668
+ 640 train 7.185459 (lr=2.7797e-04) (hash(x)=5803440)
669
+ 641 train 7.264145 (lr=2.7698e-04) (hash(x)=6212516)
670
+ 642 train 7.332349 (lr=2.7599e-04) (hash(x)=7165945)
671
+ 643 train 7.047024 (lr=2.7500e-04) (hash(x)=5298545)
672
+ 644 train 7.316221 (lr=2.7401e-04) (hash(x)=6281739)
673
+ 645 train 7.271581 (lr=2.7302e-04) (hash(x)=5779073)
674
+ 646 train 7.132289 (lr=2.7203e-04) (hash(x)=5090107)
675
+ 647 train 7.183856 (lr=2.7104e-04) (hash(x)=6340947)
676
+ 648 train 6.944769 (lr=2.7005e-04) (hash(x)=4741418)
677
+ 649 train 6.879519 (lr=2.6906e-04) (hash(x)=6530441)
678
+ 650 val loss 7.2311
679
+ 650 val perplexity 1381.7054
680
+ 650 train 7.410245 (lr=2.6807e-04) (hash(x)=6396041)
681
+ 651 train 7.086882 (lr=2.6708e-04) (hash(x)=7818007)
682
+ 652 train 7.111167 (lr=2.6609e-04) (hash(x)=6478457)
683
+ 653 train 7.172120 (lr=2.6510e-04) (hash(x)=6277435)
684
+ 654 train 7.139045 (lr=2.6411e-04) (hash(x)=6042369)
685
+ 655 train 7.153538 (lr=2.6313e-04) (hash(x)=6226320)
686
+ 656 train 7.178140 (lr=2.6214e-04) (hash(x)=5826883)
687
+ 657 train 7.282806 (lr=2.6115e-04) (hash(x)=5463751)
688
+ 658 train 7.215002 (lr=2.6016e-04) (hash(x)=6433080)
689
+ 659 train 7.074018 (lr=2.5917e-04) (hash(x)=6579329)
690
+ 660 train 7.071548 (lr=2.5819e-04) (hash(x)=4724703)
691
+ 661 train 7.422923 (lr=2.5720e-04) (hash(x)=6660540)
692
+ 662 train 7.238236 (lr=2.5621e-04) (hash(x)=5918756)
693
+ 663 train 7.157944 (lr=2.5523e-04) (hash(x)=6491272)
694
+ 664 train 7.066785 (lr=2.5424e-04) (hash(x)=5452863)
695
+ 665 train 7.276259 (lr=2.5325e-04) (hash(x)=6766947)
696
+ 666 train 7.639463 (lr=2.5227e-04) (hash(x)=7459504)
697
+ 667 train 7.289299 (lr=2.5128e-04) (hash(x)=3721218)
698
+ 668 train 7.473959 (lr=2.5030e-04) (hash(x)=7929571)
699
+ 669 train 7.159019 (lr=2.4932e-04) (hash(x)=5229898)
700
+ 670 train 7.181027 (lr=2.4833e-04) (hash(x)=6083090)
701
+ 671 train 7.037563 (lr=2.4735e-04) (hash(x)=6054428)
702
+ 672 train 7.206755 (lr=2.4637e-04) (hash(x)=6072563)
703
+ 673 train 7.296102 (lr=2.4539e-04) (hash(x)=6290464)
704
+ 674 train 7.388740 (lr=2.4441e-04) (hash(x)=6314742)
705
+ 675 train 7.059244 (lr=2.4342e-04) (hash(x)=6904498)
706
+ 676 train 7.233764 (lr=2.4244e-04) (hash(x)=5378107)
707
+ 677 train 7.344265 (lr=2.4147e-04) (hash(x)=7765805)
708
+ 678 train 7.155362 (lr=2.4049e-04) (hash(x)=5871280)
709
+ 679 train 7.307432 (lr=2.3951e-04) (hash(x)=6543764)
710
+ 680 train 7.180436 (lr=2.3853e-04) (hash(x)=7369665)
711
+ 681 train 7.267678 (lr=2.3756e-04) (hash(x)=7204955)
712
+ 682 train 7.211934 (lr=2.3658e-04) (hash(x)=5856543)
713
+ 683 train 7.088581 (lr=2.3560e-04) (hash(x)=4423711)
714
+ 684 train 7.074786 (lr=2.3463e-04) (hash(x)=5818153)
715
+ 685 train 7.169012 (lr=2.3366e-04) (hash(x)=5343477)
716
+ 686 train 7.356835 (lr=2.3268e-04) (hash(x)=6581702)
717
+ 687 train 7.190972 (lr=2.3171e-04) (hash(x)=6595856)
718
+ 688 train 7.127554 (lr=2.3074e-04) (hash(x)=5325378)
719
+ 689 train 7.055649 (lr=2.2977e-04) (hash(x)=7229015)
720
+ 690 train 7.222462 (lr=2.2880e-04) (hash(x)=5918186)
721
+ 691 train 7.098609 (lr=2.2783e-04) (hash(x)=5916034)
722
+ 692 train 7.088319 (lr=2.2687e-04) (hash(x)=5280722)
723
+ 693 train 7.493737 (lr=2.2590e-04) (hash(x)=6939521)
724
+ 694 train 7.246545 (lr=2.2493e-04) (hash(x)=5172566)
725
+ 695 train 7.089564 (lr=2.2397e-04) (hash(x)=5492185)
726
+ 696 train 6.947562 (lr=2.2300e-04) (hash(x)=5779011)
727
+ 697 train 7.236990 (lr=2.2204e-04) (hash(x)=6898006)
728
+ 698 train 7.073624 (lr=2.2108e-04) (hash(x)=5779217)
729
+ 699 train 6.940072 (lr=2.2012e-04) (hash(x)=5462008)
730
+ 700 val loss 7.2242
731
+ 700 val perplexity 1372.1897
732
+ 700 train 7.088466 (lr=2.1916e-04) (hash(x)=6243843)
733
+ 701 train 7.150336 (lr=2.1820e-04) (hash(x)=7821902)
734
+ 702 train 7.209005 (lr=2.1724e-04) (hash(x)=5222501)
735
+ 703 train 7.181474 (lr=2.1629e-04) (hash(x)=6608108)
736
+ 704 train 7.262262 (lr=2.1533e-04) (hash(x)=7602879)
737
+ 705 train 7.212751 (lr=2.1438e-04) (hash(x)=6295394)
738
+ 706 train 7.171757 (lr=2.1343e-04) (hash(x)=6579907)
739
+ 707 train 7.052801 (lr=2.1247e-04) (hash(x)=6803449)
740
+ 708 train 7.131286 (lr=2.1152e-04) (hash(x)=5413733)
741
+ 709 train 7.093498 (lr=2.1057e-04) (hash(x)=7444134)
742
+ 710 train 7.252778 (lr=2.0963e-04) (hash(x)=7028652)
743
+ 711 train 7.042346 (lr=2.0868e-04) (hash(x)=5361923)
744
+ 712 train 7.152476 (lr=2.0773e-04) (hash(x)=5478293)
745
+ 713 train 7.147458 (lr=2.0679e-04) (hash(x)=7803039)
746
+ 714 train 7.329758 (lr=2.0585e-04) (hash(x)=6469017)
747
+ 715 train 7.218033 (lr=2.0491e-04) (hash(x)=7042578)
748
+ 716 train 7.283916 (lr=2.0397e-04) (hash(x)=6927957)
749
+ 717 train 7.084883 (lr=2.0303e-04) (hash(x)=6647007)
750
+ 718 train 7.057630 (lr=2.0209e-04) (hash(x)=6230888)
751
+ 719 train 7.336038 (lr=2.0115e-04) (hash(x)=6303658)
752
+ 720 train 7.361563 (lr=2.0022e-04) (hash(x)=6769880)
753
+ 721 train 7.188929 (lr=1.9929e-04) (hash(x)=6158927)
754
+ 722 train 7.204041 (lr=1.9836e-04) (hash(x)=5991631)
755
+ 723 train 7.059233 (lr=1.9743e-04) (hash(x)=6419656)
756
+ 724 train 7.062278 (lr=1.9650e-04) (hash(x)=7739209)
757
+ 725 train 6.998619 (lr=1.9557e-04) (hash(x)=7891271)
758
+ 726 train 7.352912 (lr=1.9464e-04) (hash(x)=7171433)
759
+ 727 train 7.144310 (lr=1.9372e-04) (hash(x)=7099888)
760
+ 728 train 7.159133 (lr=1.9280e-04) (hash(x)=6103730)
761
+ 729 train 7.263193 (lr=1.9188e-04) (hash(x)=6352424)
762
+ 730 train 7.150970 (lr=1.9096e-04) (hash(x)=4893671)
763
+ 731 train 7.118305 (lr=1.9004e-04) (hash(x)=5863145)
764
+ 732 train 7.217144 (lr=1.8912e-04) (hash(x)=5502233)
765
+ 733 train 7.122712 (lr=1.8821e-04) (hash(x)=6844597)
766
+ 734 train 7.143062 (lr=1.8730e-04) (hash(x)=6744526)
767
+ 735 train 7.281025 (lr=1.8639e-04) (hash(x)=6009861)
768
+ 736 train 7.146035 (lr=1.8548e-04) (hash(x)=6152358)
769
+ 737 train 7.067510 (lr=1.8457e-04) (hash(x)=5374957)
770
+ 738 train 7.099597 (lr=1.8367e-04) (hash(x)=6495309)
771
+ 739 train 7.255993 (lr=1.8276e-04) (hash(x)=3923225)
772
+ 740 train 7.023756 (lr=1.8186e-04) (hash(x)=6255243)
773
+ 741 train 7.191382 (lr=1.8096e-04) (hash(x)=5848862)
774
+ 742 train 7.196548 (lr=1.8006e-04) (hash(x)=4535000)
775
+ 743 train 7.074863 (lr=1.7916e-04) (hash(x)=4440634)
776
+ 744 train 7.076263 (lr=1.7827e-04) (hash(x)=5222077)
777
+ 745 train 7.109241 (lr=1.7738e-04) (hash(x)=4788998)
778
+ 746 train 7.111574 (lr=1.7649e-04) (hash(x)=6108350)
779
+ 747 train 7.025684 (lr=1.7560e-04) (hash(x)=7378165)
780
+ 748 train 7.102706 (lr=1.7471e-04) (hash(x)=5475646)
781
+ 749 train 7.099395 (lr=1.7382e-04) (hash(x)=4771243)
782
+ 750 val loss 7.2055
783
+ 750 val perplexity 1346.7758
784
+ 750 train 6.982396 (lr=1.7294e-04) (hash(x)=6668088)
785
+ 751 train 7.221345 (lr=1.7206e-04) (hash(x)=5716089)
786
+ 752 train 7.171526 (lr=1.7118e-04) (hash(x)=6991946)
787
+ 753 train 7.078938 (lr=1.7030e-04) (hash(x)=7991009)
788
+ 754 train 7.177859 (lr=1.6943e-04) (hash(x)=6037018)
789
+ 755 train 7.160283 (lr=1.6855e-04) (hash(x)=5348607)
790
+ 756 train 7.290399 (lr=1.6768e-04) (hash(x)=5190526)
791
+ 757 train 7.169240 (lr=1.6681e-04) (hash(x)=6036146)
792
+ 758 train 7.068059 (lr=1.6595e-04) (hash(x)=5329783)
793
+ 759 train 7.157620 (lr=1.6508e-04) (hash(x)=6658718)
794
+ 760 train 7.038643 (lr=1.6422e-04) (hash(x)=4820939)
795
+ 761 train 7.053461 (lr=1.6336e-04) (hash(x)=6176253)
796
+ 762 train 7.019988 (lr=1.6250e-04) (hash(x)=5316750)
797
+ 763 train 7.120662 (lr=1.6164e-04) (hash(x)=4542895)
798
+ 764 train 6.906596 (lr=1.6079e-04) (hash(x)=5395673)
799
+ 765 train 7.007916 (lr=1.5994e-04) (hash(x)=5549073)
800
+ 766 train 7.245580 (lr=1.5909e-04) (hash(x)=5233723)
801
+ 767 train 7.308990 (lr=1.5824e-04) (hash(x)=5342783)
802
+ 768 train 7.228425 (lr=1.5740e-04) (hash(x)=5552448)
803
+ 769 train 7.369246 (lr=1.5655e-04) (hash(x)=6549546)
804
+ 770 train 7.035216 (lr=1.5571e-04) (hash(x)=3816493)
805
+ 771 train 7.084671 (lr=1.5487e-04) (hash(x)=6170599)
806
+ 772 train 7.568273 (lr=1.5404e-04) (hash(x)=6708759)
807
+ 773 train 7.173415 (lr=1.5320e-04) (hash(x)=8483134)
808
+ 774 train 7.767962 (lr=1.5237e-04) (hash(x)=5583732)
809
+ 775 train 7.164863 (lr=1.5154e-04) (hash(x)=4865036)
810
+ 776 train 7.088977 (lr=1.5072e-04) (hash(x)=7234336)
811
+ 777 train 7.149169 (lr=1.4989e-04) (hash(x)=4433890)
812
+ 778 train 7.451601 (lr=1.4907e-04) (hash(x)=7484636)
813
+ 779 train 7.903167 (lr=1.4825e-04) (hash(x)=6738597)
814
+ 780 train 7.129013 (lr=1.4744e-04) (hash(x)=6752421)
815
+ 781 train 7.321273 (lr=1.4662e-04) (hash(x)=5459620)
816
+ 782 train 7.243683 (lr=1.4581e-04) (hash(x)=6718131)
817
+ 783 train 7.223713 (lr=1.4500e-04) (hash(x)=4604074)
818
+ 784 train 7.085139 (lr=1.4419e-04) (hash(x)=5946124)
819
+ 785 train 7.362234 (lr=1.4339e-04) (hash(x)=8045501)
820
+ 786 train 7.455173 (lr=1.4259e-04) (hash(x)=5712725)
821
+ 787 train 7.190992 (lr=1.4179e-04) (hash(x)=6176464)
822
+ 788 train 7.221046 (lr=1.4099e-04) (hash(x)=6276702)
823
+ 789 train 7.047091 (lr=1.4020e-04) (hash(x)=6127317)
824
+ 790 train 7.199562 (lr=1.3941e-04) (hash(x)=7221374)
825
+ 791 train 7.174991 (lr=1.3862e-04) (hash(x)=5571462)
826
+ 792 train 7.170771 (lr=1.3783e-04) (hash(x)=7518021)
827
+ 793 train 7.174084 (lr=1.3705e-04) (hash(x)=6640755)
828
+ 794 train 7.255888 (lr=1.3627e-04) (hash(x)=6870405)
829
+ 795 train 7.233688 (lr=1.3549e-04) (hash(x)=6147295)
830
+ 796 train 7.105479 (lr=1.3471e-04) (hash(x)=6782583)
831
+ 797 train 7.243057 (lr=1.3394e-04) (hash(x)=6120461)
832
+ 798 train 7.161384 (lr=1.3317e-04) (hash(x)=5893870)
833
+ 799 train 7.109607 (lr=1.3241e-04) (hash(x)=5123928)
834
+ 800 val loss 7.1876
835
+ 800 val perplexity 1322.9208
836
+ 800 train 7.179597 (lr=1.3164e-04) (hash(x)=7037251)
837
+ 801 train 7.171445 (lr=1.3088e-04) (hash(x)=5428813)
838
+ 802 train 7.102864 (lr=1.3012e-04) (hash(x)=5550387)
839
+ 803 train 7.173723 (lr=1.2936e-04) (hash(x)=6448586)
840
+ 804 train 7.321678 (lr=1.2861e-04) (hash(x)=5916934)
841
+ 805 train 7.279310 (lr=1.2786e-04) (hash(x)=8116987)
842
+ 806 train 7.256086 (lr=1.2711e-04) (hash(x)=6210899)
843
+ 807 train 7.352948 (lr=1.2637e-04) (hash(x)=5974342)
844
+ 808 train 7.345386 (lr=1.2563e-04) (hash(x)=7236199)
845
+ 809 train 7.280527 (lr=1.2489e-04) (hash(x)=5261336)
846
+ 810 train 7.103971 (lr=1.2415e-04) (hash(x)=6310720)
847
+ 811 train 7.146641 (lr=1.2342e-04) (hash(x)=5307056)
848
+ 812 train 6.992956 (lr=1.2269e-04) (hash(x)=7356633)
849
+ 813 train 7.108308 (lr=1.2196e-04) (hash(x)=5865119)
850
+ 814 train 7.009826 (lr=1.2124e-04) (hash(x)=5979205)
851
+ 815 train 7.152751 (lr=1.2052e-04) (hash(x)=6672380)
852
+ 816 train 7.140406 (lr=1.1980e-04) (hash(x)=5565030)
853
+ 817 train 7.126254 (lr=1.1908e-04) (hash(x)=6886810)
854
+ 818 train 7.190171 (lr=1.1837e-04) (hash(x)=5926478)
855
+ 819 train 7.388689 (lr=1.1766e-04) (hash(x)=6564071)
856
+ 820 train 7.011282 (lr=1.1695e-04) (hash(x)=5024280)
857
+ 821 train 7.020510 (lr=1.1625e-04) (hash(x)=7705089)
858
+ 822 train 6.995688 (lr=1.1555e-04) (hash(x)=4037371)
859
+ 823 train 7.091742 (lr=1.1485e-04) (hash(x)=5773073)
860
+ 824 train 7.114347 (lr=1.1416e-04) (hash(x)=7158139)
861
+ 825 train 6.951860 (lr=1.1347e-04) (hash(x)=5905976)
862
+ 826 train 7.375481 (lr=1.1278e-04) (hash(x)=13897369)
863
+ 827 train 8.458034 (lr=1.1210e-04) (hash(x)=6936653)
864
+ 828 train 7.153047 (lr=1.1142e-04) (hash(x)=6592246)
865
+ 829 train 7.118033 (lr=1.1074e-04) (hash(x)=6592773)
866
+ 830 train 7.169332 (lr=1.1006e-04) (hash(x)=5461178)
867
+ 831 train 6.943568 (lr=1.0939e-04) (hash(x)=6796551)
868
+ 832 train 7.177258 (lr=1.0872e-04) (hash(x)=4721691)
869
+ 833 train 7.084432 (lr=1.0806e-04) (hash(x)=6151589)
870
+ 834 train 7.238265 (lr=1.0740e-04) (hash(x)=7008995)
871
+ 835 train 7.566961 (lr=1.0674e-04) (hash(x)=7045164)
872
+ 836 train 7.159750 (lr=1.0608e-04) (hash(x)=6614221)
873
+ 837 train 7.197012 (lr=1.0543e-04) (hash(x)=6650506)
874
+ 838 train 7.005482 (lr=1.0478e-04) (hash(x)=6363762)
875
+ 839 train 7.162179 (lr=1.0413e-04) (hash(x)=7839748)
876
+ 840 train 7.087656 (lr=1.0349e-04) (hash(x)=5124077)
877
+ 841 train 7.506143 (lr=1.0285e-04) (hash(x)=6028671)
878
+ 842 train 7.023764 (lr=1.0222e-04) (hash(x)=5427175)
879
+ 843 train 7.339113 (lr=1.0158e-04) (hash(x)=6118408)
880
+ 844 train 7.129816 (lr=1.0095e-04) (hash(x)=6022552)
881
+ 845 train 7.144388 (lr=1.0033e-04) (hash(x)=5757526)
882
+ 846 train 7.568955 (lr=9.9707e-05) (hash(x)=6476549)
883
+ 847 train 6.933665 (lr=9.9088e-05) (hash(x)=6030172)
884
+ 848 train 7.300003 (lr=9.8472e-05) (hash(x)=5917229)
885
+ 849 train 7.381317 (lr=9.7860e-05) (hash(x)=6719882)
886
+ 850 val loss 7.1759
887
+ 850 val perplexity 1307.5314
888
+ 850 train 7.238054 (lr=9.7252e-05) (hash(x)=5441498)
889
+ 851 train 7.101161 (lr=9.6646e-05) (hash(x)=5028050)
890
+ 852 train 7.143072 (lr=9.6044e-05) (hash(x)=8487491)
891
+ 853 train 7.154226 (lr=9.5446e-05) (hash(x)=5682910)
892
+ 854 train 7.322289 (lr=9.4851e-05) (hash(x)=6781362)
893
+ 855 train 7.147408 (lr=9.4260e-05) (hash(x)=5747891)
894
+ 856 train 7.192365 (lr=9.3672e-05) (hash(x)=7627661)
895
+ 857 train 6.864371 (lr=9.3088e-05) (hash(x)=5734208)
896
+ 858 train 7.014743 (lr=9.2507e-05) (hash(x)=6162015)
897
+ 859 train 7.229060 (lr=9.1929e-05) (hash(x)=7024381)
898
+ 860 train 7.429635 (lr=9.1356e-05) (hash(x)=5198498)
899
+ 861 train 7.423097 (lr=9.0785e-05) (hash(x)=7814361)
900
+ 862 train 7.289762 (lr=9.0219e-05) (hash(x)=7394995)
901
+ 863 train 7.173902 (lr=8.9656e-05) (hash(x)=5765893)
902
+ 864 train 6.983388 (lr=8.9096e-05) (hash(x)=4990435)
903
+ 865 train 7.139946 (lr=8.8540e-05) (hash(x)=5572048)
904
+ 866 train 7.274060 (lr=8.7988e-05) (hash(x)=7093293)
905
+ 867 train 7.173663 (lr=8.7439e-05) (hash(x)=5610058)
906
+ 868 train 7.063409 (lr=8.6894e-05) (hash(x)=6771468)
907
+ 869 train 7.102000 (lr=8.6353e-05) (hash(x)=4459483)
908
+ 870 train 6.990169 (lr=8.5815e-05) (hash(x)=4926546)
909
+ 871 train 7.109494 (lr=8.5281e-05) (hash(x)=6758411)
910
+ 872 train 7.072481 (lr=8.4751e-05) (hash(x)=7394277)
911
+ 873 train 7.198714 (lr=8.4224e-05) (hash(x)=6701381)
912
+ 874 train 7.534401 (lr=8.3701e-05) (hash(x)=8120966)
913
+ 875 train 7.049159 (lr=8.3182e-05) (hash(x)=5416400)
914
+ 876 train 7.063768 (lr=8.2666e-05) (hash(x)=6292778)
915
+ 877 train 7.003854 (lr=8.2154e-05) (hash(x)=5919569)
916
+ 878 train 7.318653 (lr=8.1646e-05) (hash(x)=7561740)
917
+ 879 train 7.060314 (lr=8.1142e-05) (hash(x)=5282870)
918
+ 880 train 7.101182 (lr=8.0641e-05) (hash(x)=5473238)
919
+ 881 train 6.974599 (lr=8.0144e-05) (hash(x)=8072181)
920
+ 882 train 7.002427 (lr=7.9651e-05) (hash(x)=6757322)
921
+ 883 train 6.970382 (lr=7.9162e-05) (hash(x)=7996770)
922
+ 884 train 7.051551 (lr=7.8676e-05) (hash(x)=3181021)
923
+ 885 train 7.077705 (lr=7.8195e-05) (hash(x)=5093519)
924
+ 886 train 7.067791 (lr=7.7717e-05) (hash(x)=5391352)
925
+ 887 train 7.110921 (lr=7.7243e-05) (hash(x)=6270917)
926
+ 888 train 7.084269 (lr=7.6772e-05) (hash(x)=5818195)
927
+ 889 train 6.834678 (lr=7.6306e-05) (hash(x)=5852780)
928
+ 890 train 6.753691 (lr=7.5843e-05) (hash(x)=5777928)
929
+ 891 train 6.957564 (lr=7.5384e-05) (hash(x)=5327728)
930
+ 892 train 6.994030 (lr=7.4930e-05) (hash(x)=7651281)
931
+ 893 train 7.197446 (lr=7.4479e-05) (hash(x)=6260765)
932
+ 894 train 7.099965 (lr=7.4031e-05) (hash(x)=6265102)
933
+ 895 train 7.242281 (lr=7.3588e-05) (hash(x)=6212894)
934
+ 896 train 7.025801 (lr=7.3149e-05) (hash(x)=6901967)
935
+ 897 train 7.093394 (lr=7.2714e-05) (hash(x)=6724010)
936
+ 898 train 7.095313 (lr=7.2282e-05) (hash(x)=6299033)
937
+ 899 train 7.249408 (lr=7.1854e-05) (hash(x)=6059934)
938
+ 900 val loss 7.1696
939
+ 900 val perplexity 1299.2709
940
+ 900 train 6.956734 (lr=7.1431e-05) (hash(x)=6029696)
941
+ 901 train 6.891285 (lr=7.1011e-05) (hash(x)=5587382)
942
+ 902 train 7.093894 (lr=7.0595e-05) (hash(x)=5890076)
943
+ 903 train 7.015258 (lr=7.0184e-05) (hash(x)=7274108)
944
+ 904 train 7.089938 (lr=6.9776e-05) (hash(x)=7975528)
945
+ 905 train 7.150680 (lr=6.9372e-05) (hash(x)=6085396)
946
+ 906 train 7.027868 (lr=6.8972e-05) (hash(x)=6724496)
947
+ 907 train 7.018924 (lr=6.8576e-05) (hash(x)=4784666)
948
+ 908 train 7.031313 (lr=6.8184e-05) (hash(x)=7193845)
949
+ 909 train 7.189034 (lr=6.7796e-05) (hash(x)=6288638)
950
+ 910 train 7.163491 (lr=6.7412e-05) (hash(x)=5685285)
951
+ 911 train 7.132307 (lr=6.7033e-05) (hash(x)=7714288)
952
+ 912 train 7.220439 (lr=6.6657e-05) (hash(x)=5983679)
953
+ 913 train 7.180109 (lr=6.6285e-05) (hash(x)=5534748)
954
+ 914 train 7.109372 (lr=6.5917e-05) (hash(x)=6545471)
955
+ 915 train 7.098323 (lr=6.5553e-05) (hash(x)=6361270)
956
+ 916 train 7.098313 (lr=6.5194e-05) (hash(x)=6257744)
957
+ 917 train 7.121905 (lr=6.4838e-05) (hash(x)=4682956)
958
+ 918 train 7.005388 (lr=6.4487e-05) (hash(x)=4521756)
959
+ 919 train 7.138700 (lr=6.4139e-05) (hash(x)=7568935)
960
+ 920 train 7.063842 (lr=6.3796e-05) (hash(x)=7969353)
961
+ 921 train 7.250971 (lr=6.3457e-05) (hash(x)=7973311)
962
+ 922 train 6.977053 (lr=6.3121e-05) (hash(x)=4843937)
963
+ 923 train 7.095287 (lr=6.2790e-05) (hash(x)=6987352)
964
+ 924 train 7.233639 (lr=6.2463e-05) (hash(x)=7179418)
965
+ 925 train 6.882221 (lr=6.2140e-05) (hash(x)=5413904)
966
+ 926 train 7.293038 (lr=6.1822e-05) (hash(x)=6028105)
967
+ 927 train 7.184164 (lr=6.1507e-05) (hash(x)=6732924)
968
+ 928 train 7.079037 (lr=6.1197e-05) (hash(x)=4547651)
969
+ 929 train 7.230199 (lr=6.0890e-05) (hash(x)=6219754)
970
+ 930 train 7.019924 (lr=6.0588e-05) (hash(x)=7465186)
971
+ 931 train 7.161325 (lr=6.0290e-05) (hash(x)=6839070)
972
+ 932 train 7.043342 (lr=5.9996e-05) (hash(x)=5390324)
973
+ 933 train 7.011817 (lr=5.9706e-05) (hash(x)=6090120)
974
+ 934 train 7.653197 (lr=5.9421e-05) (hash(x)=7123181)
975
+ 935 train 7.308783 (lr=5.9139e-05) (hash(x)=7115176)
976
+ 936 train 7.297654 (lr=5.8862e-05) (hash(x)=7016990)
977
+ 937 train 7.075312 (lr=5.8589e-05) (hash(x)=4989299)
978
+ 938 train 7.101528 (lr=5.8320e-05) (hash(x)=6783311)
979
+ 939 train 7.070540 (lr=5.8056e-05) (hash(x)=7497590)
980
+ 940 train 7.288990 (lr=5.7795e-05) (hash(x)=6767514)
981
+ 941 train 7.044348 (lr=5.7539e-05) (hash(x)=3977557)
982
+ 942 train 7.166264 (lr=5.7287e-05) (hash(x)=4806492)
983
+ 943 train 7.091781 (lr=5.7039e-05) (hash(x)=5524165)
984
+ 944 train 7.135848 (lr=5.6796e-05) (hash(x)=4930752)
985
+ 945 train 7.092951 (lr=5.6556e-05) (hash(x)=5124936)
986
+ 946 train 6.994030 (lr=5.6321e-05) (hash(x)=6877351)
987
+ 947 train 7.281543 (lr=5.6090e-05) (hash(x)=5522119)
988
+ 948 train 7.080702 (lr=5.5864e-05) (hash(x)=6183512)
989
+ 949 train 6.984340 (lr=5.5641e-05) (hash(x)=5030865)
990
+ 950 val loss 7.1528
991
+ 950 val perplexity 1277.6420
992
+ 950 train 7.097123 (lr=5.5423e-05) (hash(x)=6619812)
993
+ 951 train 7.098376 (lr=5.5209e-05) (hash(x)=6411783)
994
+ 952 train 7.088872 (lr=5.4999e-05) (hash(x)=6748982)
995
+ 953 train 7.000419 (lr=5.4794e-05) (hash(x)=6581937)
996
+ 954 train 7.102852 (lr=5.4593e-05) (hash(x)=7881022)
997
+ 955 train 7.066958 (lr=5.4396e-05) (hash(x)=5116910)
998
+ 956 train 6.857338 (lr=5.4203e-05) (hash(x)=4880833)
999
+ 957 train 6.900289 (lr=5.4015e-05) (hash(x)=7353459)
1000
+ 958 train 6.972719 (lr=5.3831e-05) (hash(x)=4951302)
1001
+ 959 train 6.879022 (lr=5.3651e-05) (hash(x)=6344535)
1002
+ 960 train 7.476079 (lr=5.3476e-05) (hash(x)=7411763)
1003
+ 961 train 7.066309 (lr=5.3305e-05) (hash(x)=5840169)
1004
+ 962 train 7.168903 (lr=5.3138e-05) (hash(x)=5784788)
1005
+ 963 train 6.974478 (lr=5.2975e-05) (hash(x)=6754290)
1006
+ 964 train 7.261588 (lr=5.2817e-05) (hash(x)=6369963)
1007
+ 965 train 7.335275 (lr=5.2663e-05) (hash(x)=6484573)
1008
+ 966 train 6.908237 (lr=5.2513e-05) (hash(x)=6516338)
1009
+ 967 train 6.942739 (lr=5.2368e-05) (hash(x)=5260125)
1010
+ 968 train 7.083892 (lr=5.2227e-05) (hash(x)=5781624)
1011
+ 969 train 6.876302 (lr=5.2090e-05) (hash(x)=4881179)
1012
+ 970 train 6.999709 (lr=5.1957e-05) (hash(x)=6707821)
1013
+ 971 train 7.147691 (lr=5.1829e-05) (hash(x)=5001634)
1014
+ 972 train 7.025114 (lr=5.1705e-05) (hash(x)=4402949)
1015
+ 973 train 7.027910 (lr=5.1586e-05) (hash(x)=7424810)
1016
+ 974 train 6.900013 (lr=5.1471e-05) (hash(x)=4814803)
1017
+ 975 train 6.823050 (lr=5.1360e-05) (hash(x)=5176431)
1018
+ 976 train 7.021018 (lr=5.1253e-05) (hash(x)=5769393)
1019
+ 977 train 6.965964 (lr=5.1151e-05) (hash(x)=5421624)
1020
+ 978 train 7.193567 (lr=5.1053e-05) (hash(x)=5999773)
1021
+ 979 train 6.874374 (lr=5.0960e-05) (hash(x)=5674439)
1022
+ 980 train 6.981283 (lr=5.0871e-05) (hash(x)=6345351)
1023
+ 981 train 6.917695 (lr=5.0786e-05) (hash(x)=5819157)
1024
+ 982 train 7.177959 (lr=5.0705e-05) (hash(x)=11159355)
1025
+ 983 train 7.118654 (lr=5.0629e-05) (hash(x)=6506856)
1026
+ 984 train 7.013239 (lr=5.0557e-05) (hash(x)=6331023)
1027
+ 985 train 6.918195 (lr=5.0490e-05) (hash(x)=6119124)
1028
+ 986 train 7.119637 (lr=5.0427e-05) (hash(x)=8142528)
1029
+ 987 train 7.305790 (lr=5.0368e-05) (hash(x)=6241669)
1030
+ 988 train 6.927010 (lr=5.0314e-05) (hash(x)=3852991)
1031
+ 989 train 6.660688 (lr=5.0263e-05) (hash(x)=4992756)
1032
+ 990 train 7.372957 (lr=5.0218e-05) (hash(x)=7097145)
1033
+ 991 train 7.146570 (lr=5.0176e-05) (hash(x)=7277927)
1034
+ 992 train 7.228550 (lr=5.0139e-05) (hash(x)=5661400)
1035
+ 993 train 7.135367 (lr=5.0107e-05) (hash(x)=6622097)
1036
+ 994 train 7.267321 (lr=5.0078e-05) (hash(x)=5016804)
1037
+ 995 train 7.375402 (lr=5.0054e-05) (hash(x)=7026582)
1038
+ 996 train 7.121544 (lr=5.0035e-05) (hash(x)=5639136)
1039
+ 997 train 7.285857 (lr=5.0020e-05) (hash(x)=5840987)
1040
+ 998 train 6.901052 (lr=5.0009e-05) (hash(x)=5265410)
1041
+ 999 val loss 7.1461
1042
+ 999 val perplexity 1269.0940
1043
+ 999 train 6.878165 (lr=5.0002e-05) (hash(x)=4351074)
lr5e-4_total_batch_size20480_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b53038dc5fac0f2485226ed66e92424f68462d47185c944b426b4e055762969
3
+ size 36163010
lr5e-4_total_batch_size20480_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f67e7633a253a9907760b0f7561875f5f64e48e8c0922c0fe516c77b5e400fc3
3
+ size 70764358