andrew-healey commited on
Commit
2042dea
·
verified ·
1 Parent(s): 5ed8038

Upload folder using huggingface_hub

Browse files
lr4e-4_total_batch_size5120_seq_len128/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep_2/lr4e-4_total_batch_size5120_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": 5120, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": true, "mup_enable_coord_check_logging": false, "max_lr": 0.0004, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr4e-4_total_batch_size5120_seq_len128/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr4e-4_total_batch_size5120_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 11.683437 (lr=1.3986e-06) (hash(x)=6154740)
5
+ 1 train 11.714834 (lr=2.7972e-06) (hash(x)=6605789)
6
+ 2 train 11.724317 (lr=4.1958e-06) (hash(x)=6696707)
7
+ 3 train 11.702266 (lr=5.5944e-06) (hash(x)=5113283)
8
+ 4 train 11.619727 (lr=6.9930e-06) (hash(x)=5607069)
9
+ 5 train 11.654514 (lr=8.3916e-06) (hash(x)=7007748)
10
+ 6 train 11.585394 (lr=9.7902e-06) (hash(x)=7643299)
11
+ 7 train 11.596355 (lr=1.1189e-05) (hash(x)=5534142)
12
+ 8 train 11.551446 (lr=1.2587e-05) (hash(x)=6539293)
13
+ 9 train 11.563625 (lr=1.3986e-05) (hash(x)=5974274)
14
+ 10 train 11.548813 (lr=1.5385e-05) (hash(x)=7199085)
15
+ 11 train 11.482031 (lr=1.6783e-05) (hash(x)=6763231)
16
+ 12 train 11.380842 (lr=1.8182e-05) (hash(x)=6720398)
17
+ 13 train 11.360128 (lr=1.9580e-05) (hash(x)=6706513)
18
+ 14 train 11.364397 (lr=2.0979e-05) (hash(x)=5727925)
19
+ 15 train 11.290984 (lr=2.2378e-05) (hash(x)=6328012)
20
+ 16 train 11.216109 (lr=2.3776e-05) (hash(x)=5812722)
21
+ 17 train 11.163255 (lr=2.5175e-05) (hash(x)=6183012)
22
+ 18 train 11.194344 (lr=2.6573e-05) (hash(x)=5952539)
23
+ 19 train 11.130493 (lr=2.7972e-05) (hash(x)=7052161)
24
+ 20 train 10.930115 (lr=2.9371e-05) (hash(x)=5810871)
25
+ 21 train 10.979438 (lr=3.0769e-05) (hash(x)=6586466)
26
+ 22 train 10.851637 (lr=3.2168e-05) (hash(x)=5061781)
27
+ 23 train 10.863820 (lr=3.3566e-05) (hash(x)=5556098)
28
+ 24 train 10.915709 (lr=3.4965e-05) (hash(x)=7268535)
29
+ 25 train 10.771225 (lr=3.6364e-05) (hash(x)=8155572)
30
+ 26 train 10.657117 (lr=3.7762e-05) (hash(x)=5019590)
31
+ 27 train 10.572108 (lr=3.9161e-05) (hash(x)=4696943)
32
+ 28 train 10.519890 (lr=4.0559e-05) (hash(x)=5984010)
33
+ 29 train 10.677322 (lr=4.1958e-05) (hash(x)=7649389)
34
+ 30 train 10.422065 (lr=4.3357e-05) (hash(x)=6332817)
35
+ 31 train 10.346922 (lr=4.4755e-05) (hash(x)=6219138)
36
+ 32 train 10.494532 (lr=4.6154e-05) (hash(x)=6895681)
37
+ 33 train 10.375053 (lr=4.7552e-05) (hash(x)=7075794)
38
+ 34 train 10.271528 (lr=4.8951e-05) (hash(x)=5024163)
39
+ 35 train 10.334936 (lr=5.0350e-05) (hash(x)=6049878)
40
+ 36 train 10.403502 (lr=5.1748e-05) (hash(x)=4774943)
41
+ 37 train 10.398000 (lr=5.3147e-05) (hash(x)=6385927)
42
+ 38 train 10.220635 (lr=5.4545e-05) (hash(x)=5070217)
43
+ 39 train 10.221786 (lr=5.5944e-05) (hash(x)=5491972)
44
+ 40 train 10.286844 (lr=5.7343e-05) (hash(x)=5695751)
45
+ 41 train 10.220202 (lr=5.8741e-05) (hash(x)=6061420)
46
+ 42 train 10.138086 (lr=6.0140e-05) (hash(x)=7836710)
47
+ 43 train 10.250729 (lr=6.1538e-05) (hash(x)=8075458)
48
+ 44 train 10.070316 (lr=6.2937e-05) (hash(x)=7190584)
49
+ 45 train 10.125028 (lr=6.4336e-05) (hash(x)=7238769)
50
+ 46 train 10.279922 (lr=6.5734e-05) (hash(x)=4931205)
51
+ 47 train 10.091259 (lr=6.7133e-05) (hash(x)=5917741)
52
+ 48 train 10.034421 (lr=6.8531e-05) (hash(x)=5530785)
53
+ 49 train 10.084161 (lr=6.9930e-05) (hash(x)=5430605)
54
+ 50 val loss 10.0612
55
+ 50 val perplexity 23416.7500
56
+ 50 train 10.069185 (lr=7.1329e-05) (hash(x)=6056774)
57
+ 51 train 9.960612 (lr=7.2727e-05) (hash(x)=5750403)
58
+ 52 train 9.968672 (lr=7.4126e-05) (hash(x)=6187497)
59
+ 53 train 9.958236 (lr=7.5524e-05) (hash(x)=4792640)
60
+ 54 train 10.064965 (lr=7.6923e-05) (hash(x)=6868430)
61
+ 55 train 9.912956 (lr=7.8322e-05) (hash(x)=6112458)
62
+ 56 train 9.852873 (lr=7.9720e-05) (hash(x)=5523601)
63
+ 57 train 9.787319 (lr=8.1119e-05) (hash(x)=5801122)
64
+ 58 train 9.881976 (lr=8.2517e-05) (hash(x)=6826295)
65
+ 59 train 9.995114 (lr=8.3916e-05) (hash(x)=5806490)
66
+ 60 train 9.822445 (lr=8.5315e-05) (hash(x)=6223100)
67
+ 61 train 9.759766 (lr=8.6713e-05) (hash(x)=5754140)
68
+ 62 train 9.739307 (lr=8.8112e-05) (hash(x)=6675367)
69
+ 63 train 9.644345 (lr=8.9510e-05) (hash(x)=6568379)
70
+ 64 train 9.501768 (lr=9.0909e-05) (hash(x)=5648233)
71
+ 65 train 9.672204 (lr=9.2308e-05) (hash(x)=7134932)
72
+ 66 train 9.627056 (lr=9.3706e-05) (hash(x)=5242536)
73
+ 67 train 9.665130 (lr=9.5105e-05) (hash(x)=7212403)
74
+ 68 train 9.505008 (lr=9.6503e-05) (hash(x)=5342967)
75
+ 69 train 9.610952 (lr=9.7902e-05) (hash(x)=7720703)
76
+ 70 train 9.525063 (lr=9.9301e-05) (hash(x)=7351337)
77
+ 71 train 9.558569 (lr=1.0070e-04) (hash(x)=7386082)
78
+ 72 train 9.406364 (lr=1.0210e-04) (hash(x)=6931318)
79
+ 73 train 9.404094 (lr=1.0350e-04) (hash(x)=6848513)
80
+ 74 train 9.399150 (lr=1.0490e-04) (hash(x)=7714274)
81
+ 75 train 9.452235 (lr=1.0629e-04) (hash(x)=6359743)
82
+ 76 train 9.264580 (lr=1.0769e-04) (hash(x)=6801650)
83
+ 77 train 9.226777 (lr=1.0909e-04) (hash(x)=7536538)
84
+ 78 train 9.317209 (lr=1.1049e-04) (hash(x)=6957410)
85
+ 79 train 9.313169 (lr=1.1189e-04) (hash(x)=6151674)
86
+ 80 train 9.268032 (lr=1.1329e-04) (hash(x)=6033844)
87
+ 81 train 9.132256 (lr=1.1469e-04) (hash(x)=6317099)
88
+ 82 train 9.070349 (lr=1.1608e-04) (hash(x)=5357844)
89
+ 83 train 9.122319 (lr=1.1748e-04) (hash(x)=7168704)
90
+ 84 train 9.251307 (lr=1.1888e-04) (hash(x)=7157938)
91
+ 85 train 9.034517 (lr=1.2028e-04) (hash(x)=6736092)
92
+ 86 train 8.936291 (lr=1.2168e-04) (hash(x)=5383012)
93
+ 87 train 8.991364 (lr=1.2308e-04) (hash(x)=7859060)
94
+ 88 train 8.970185 (lr=1.2448e-04) (hash(x)=5869288)
95
+ 89 train 8.944607 (lr=1.2587e-04) (hash(x)=6958223)
96
+ 90 train 8.808808 (lr=1.2727e-04) (hash(x)=6731344)
97
+ 91 train 8.763073 (lr=1.2867e-04) (hash(x)=6976239)
98
+ 92 train 8.726782 (lr=1.3007e-04) (hash(x)=7769441)
99
+ 93 train 8.755550 (lr=1.3147e-04) (hash(x)=6956845)
100
+ 94 train 8.856450 (lr=1.3287e-04) (hash(x)=6638037)
101
+ 95 train 8.550882 (lr=1.3427e-04) (hash(x)=6391506)
102
+ 96 train 8.845491 (lr=1.3566e-04) (hash(x)=7319928)
103
+ 97 train 8.793879 (lr=1.3706e-04) (hash(x)=6113701)
104
+ 98 train 8.449331 (lr=1.3846e-04) (hash(x)=6246453)
105
+ 99 train 8.519605 (lr=1.3986e-04) (hash(x)=5245999)
106
+ 100 val loss 8.5686
107
+ 100 val perplexity 5263.9146
108
+ 100 train 8.578829 (lr=1.4126e-04) (hash(x)=6227256)
109
+ 101 train 8.509744 (lr=1.4266e-04) (hash(x)=5774189)
110
+ 102 train 8.398865 (lr=1.4406e-04) (hash(x)=5588220)
111
+ 103 train 8.480408 (lr=1.4545e-04) (hash(x)=5985675)
112
+ 104 train 8.346600 (lr=1.4685e-04) (hash(x)=6259214)
113
+ 105 train 8.453425 (lr=1.4825e-04) (hash(x)=6451776)
114
+ 106 train 8.139659 (lr=1.4965e-04) (hash(x)=4940406)
115
+ 107 train 8.260358 (lr=1.5105e-04) (hash(x)=4355733)
116
+ 108 train 8.303938 (lr=1.5245e-04) (hash(x)=6186065)
117
+ 109 train 8.337807 (lr=1.5385e-04) (hash(x)=6669188)
118
+ 110 train 9.373795 (lr=1.5524e-04) (hash(x)=7118997)
119
+ 111 train 8.518595 (lr=1.5664e-04) (hash(x)=6984772)
120
+ 112 train 8.155628 (lr=1.5804e-04) (hash(x)=6472877)
121
+ 113 train 8.311499 (lr=1.5944e-04) (hash(x)=7440580)
122
+ 114 train 8.334013 (lr=1.6084e-04) (hash(x)=6378428)
123
+ 115 train 8.339608 (lr=1.6224e-04) (hash(x)=6244096)
124
+ 116 train 8.340956 (lr=1.6364e-04) (hash(x)=9100578)
125
+ 117 train 8.612723 (lr=1.6503e-04) (hash(x)=8381560)
126
+ 118 train 8.327225 (lr=1.6643e-04) (hash(x)=6066072)
127
+ 119 train 8.059075 (lr=1.6783e-04) (hash(x)=6603717)
128
+ 120 train 7.984727 (lr=1.6923e-04) (hash(x)=5138438)
129
+ 121 train 7.869165 (lr=1.7063e-04) (hash(x)=6437516)
130
+ 122 train 8.058741 (lr=1.7203e-04) (hash(x)=5743807)
131
+ 123 train 7.859820 (lr=1.7343e-04) (hash(x)=5669522)
132
+ 124 train 8.065757 (lr=1.7483e-04) (hash(x)=5963623)
133
+ 125 train 7.815839 (lr=1.7622e-04) (hash(x)=5505889)
134
+ 126 train 8.219624 (lr=1.7762e-04) (hash(x)=7607512)
135
+ 127 train 7.993526 (lr=1.7902e-04) (hash(x)=6190579)
136
+ 128 train 8.177776 (lr=1.8042e-04) (hash(x)=7444834)
137
+ 129 train 8.087092 (lr=1.8182e-04) (hash(x)=5153196)
138
+ 130 train 7.981976 (lr=1.8322e-04) (hash(x)=6495326)
139
+ 131 train 8.088462 (lr=1.8462e-04) (hash(x)=4912983)
140
+ 132 train 8.271357 (lr=1.8601e-04) (hash(x)=7094046)
141
+ 133 train 7.855439 (lr=1.8741e-04) (hash(x)=5109125)
142
+ 134 train 7.815718 (lr=1.8881e-04) (hash(x)=7119136)
143
+ 135 train 7.861666 (lr=1.9021e-04) (hash(x)=7276303)
144
+ 136 train 8.110420 (lr=1.9161e-04) (hash(x)=9147123)
145
+ 137 train 8.102427 (lr=1.9301e-04) (hash(x)=6645716)
146
+ 138 train 8.029046 (lr=1.9441e-04) (hash(x)=6347875)
147
+ 139 train 7.806607 (lr=1.9580e-04) (hash(x)=5479947)
148
+ 140 train 7.927036 (lr=1.9720e-04) (hash(x)=5832595)
149
+ 141 train 8.217505 (lr=1.9860e-04) (hash(x)=7850360)
150
+ 142 train 8.590468 (lr=2.0000e-04) (hash(x)=8499683)
151
+ 143 train 8.117359 (lr=2.0140e-04) (hash(x)=6630378)
152
+ 144 train 8.050580 (lr=2.0280e-04) (hash(x)=5340846)
153
+ 145 train 7.878900 (lr=2.0420e-04) (hash(x)=6295358)
154
+ 146 train 7.917219 (lr=2.0559e-04) (hash(x)=5974729)
155
+ 147 train 8.450937 (lr=2.0699e-04) (hash(x)=8097010)
156
+ 148 train 8.198786 (lr=2.0839e-04) (hash(x)=6681250)
157
+ 149 train 7.617529 (lr=2.0979e-04) (hash(x)=3958188)
158
+ 150 val loss 7.8662
159
+ 150 val perplexity 2607.7197
160
+ 150 train 7.818198 (lr=2.1119e-04) (hash(x)=5975539)
161
+ 151 train 9.711157 (lr=2.1259e-04) (hash(x)=9989744)
162
+ 152 train 8.821021 (lr=2.1399e-04) (hash(x)=6748487)
163
+ 153 train 7.778756 (lr=2.1538e-04) (hash(x)=6343832)
164
+ 154 train 7.855833 (lr=2.1678e-04) (hash(x)=5688949)
165
+ 155 train 8.195263 (lr=2.1818e-04) (hash(x)=6694932)
166
+ 156 train 7.828443 (lr=2.1958e-04) (hash(x)=5872835)
167
+ 157 train 7.897624 (lr=2.2098e-04) (hash(x)=6429137)
168
+ 158 train 8.120209 (lr=2.2238e-04) (hash(x)=6999390)
169
+ 159 train 7.968429 (lr=2.2378e-04) (hash(x)=6317894)
170
+ 160 train 7.690858 (lr=2.2517e-04) (hash(x)=5813657)
171
+ 161 train 7.793367 (lr=2.2657e-04) (hash(x)=5165595)
172
+ 162 train 7.734653 (lr=2.2797e-04) (hash(x)=5682633)
173
+ 163 train 7.942009 (lr=2.2937e-04) (hash(x)=5046327)
174
+ 164 train 8.024662 (lr=2.3077e-04) (hash(x)=6195808)
175
+ 165 train 7.881701 (lr=2.3217e-04) (hash(x)=7060363)
176
+ 166 train 7.873632 (lr=2.3357e-04) (hash(x)=5837126)
177
+ 167 train 7.847794 (lr=2.3497e-04) (hash(x)=7947703)
178
+ 168 train 7.484069 (lr=2.3636e-04) (hash(x)=5277865)
179
+ 169 train 7.740170 (lr=2.3776e-04) (hash(x)=5971240)
180
+ 170 train 7.682817 (lr=2.3916e-04) (hash(x)=6590408)
181
+ 171 train 7.842447 (lr=2.4056e-04) (hash(x)=5916068)
182
+ 172 train 7.779253 (lr=2.4196e-04) (hash(x)=6069727)
183
+ 173 train 7.809871 (lr=2.4336e-04) (hash(x)=7426277)
184
+ 174 train 7.972969 (lr=2.4476e-04) (hash(x)=5356513)
185
+ 175 train 7.737049 (lr=2.4615e-04) (hash(x)=5777498)
186
+ 176 train 8.004381 (lr=2.4755e-04) (hash(x)=5989756)
187
+ 177 train 7.739745 (lr=2.4895e-04) (hash(x)=6403331)
188
+ 178 train 7.780223 (lr=2.5035e-04) (hash(x)=7650667)
189
+ 179 train 7.965009 (lr=2.5175e-04) (hash(x)=5753092)
190
+ 180 train 7.714940 (lr=2.5315e-04) (hash(x)=5751457)
191
+ 181 train 7.589414 (lr=2.5455e-04) (hash(x)=5981486)
192
+ 182 train 7.635160 (lr=2.5594e-04) (hash(x)=5918229)
193
+ 183 train 7.883895 (lr=2.5734e-04) (hash(x)=7723226)
194
+ 184 train 7.962802 (lr=2.5874e-04) (hash(x)=6111721)
195
+ 185 train 7.578727 (lr=2.6014e-04) (hash(x)=6069558)
196
+ 186 train 7.837216 (lr=2.6154e-04) (hash(x)=5626336)
197
+ 187 train 8.072463 (lr=2.6294e-04) (hash(x)=7007016)
198
+ 188 train 7.572512 (lr=2.6434e-04) (hash(x)=5883013)
199
+ 189 train 7.457197 (lr=2.6573e-04) (hash(x)=5027958)
200
+ 190 train 7.813180 (lr=2.6713e-04) (hash(x)=5474948)
201
+ 191 train 7.631199 (lr=2.6853e-04) (hash(x)=6491229)
202
+ 192 train 7.726456 (lr=2.6993e-04) (hash(x)=5518341)
203
+ 193 train 7.834788 (lr=2.7133e-04) (hash(x)=7036653)
204
+ 194 train 7.817052 (lr=2.7273e-04) (hash(x)=5527616)
205
+ 195 train 8.100171 (lr=2.7413e-04) (hash(x)=6390752)
206
+ 196 train 7.767852 (lr=2.7552e-04) (hash(x)=6032557)
207
+ 197 train 8.478538 (lr=2.7692e-04) (hash(x)=7758167)
208
+ 198 train 8.345364 (lr=2.7832e-04) (hash(x)=7968070)
209
+ 199 train 7.871463 (lr=2.7972e-04) (hash(x)=6019389)
210
+ 200 val loss 7.7821
211
+ 200 val perplexity 2397.2014
212
+ 200 train 7.883311 (lr=2.8112e-04) (hash(x)=7276744)
213
+ 201 train 7.909128 (lr=2.8252e-04) (hash(x)=6503191)
214
+ 202 train 7.744545 (lr=2.8392e-04) (hash(x)=6006880)
215
+ 203 train 7.957067 (lr=2.8531e-04) (hash(x)=7662067)
216
+ 204 train 7.719955 (lr=2.8671e-04) (hash(x)=6297345)
217
+ 205 train 7.953323 (lr=2.8811e-04) (hash(x)=7901992)
218
+ 206 train 8.076226 (lr=2.8951e-04) (hash(x)=6579950)
219
+ 207 train 7.601086 (lr=2.9091e-04) (hash(x)=4648609)
220
+ 208 train 7.848987 (lr=2.9231e-04) (hash(x)=6903216)
221
+ 209 train 7.687524 (lr=2.9371e-04) (hash(x)=5897288)
222
+ 210 train 7.758507 (lr=2.9510e-04) (hash(x)=7300160)
223
+ 211 train 7.608027 (lr=2.9650e-04) (hash(x)=4725966)
224
+ 212 train 7.430196 (lr=2.9790e-04) (hash(x)=5060060)
225
+ 213 train 7.959236 (lr=2.9930e-04) (hash(x)=6243442)
226
+ 214 train 7.658198 (lr=3.0070e-04) (hash(x)=5893816)
227
+ 215 train 7.678847 (lr=3.0210e-04) (hash(x)=5558355)
228
+ 216 train 7.831263 (lr=3.0350e-04) (hash(x)=4378747)
229
+ 217 train 7.689771 (lr=3.0490e-04) (hash(x)=5400407)
230
+ 218 train 7.959946 (lr=3.0629e-04) (hash(x)=6900554)
231
+ 219 train 7.831897 (lr=3.0769e-04) (hash(x)=6524933)
232
+ 220 train 7.763245 (lr=3.0909e-04) (hash(x)=5409944)
233
+ 221 train 7.847879 (lr=3.1049e-04) (hash(x)=5889724)
234
+ 222 train 7.771236 (lr=3.1189e-04) (hash(x)=4970496)
235
+ 223 train 7.565698 (lr=3.1329e-04) (hash(x)=6369326)
236
+ 224 train 7.664714 (lr=3.1469e-04) (hash(x)=6563975)
237
+ 225 train 7.846850 (lr=3.1608e-04) (hash(x)=5911906)
238
+ 226 train 7.939465 (lr=3.1748e-04) (hash(x)=6111462)
239
+ 227 train 7.945687 (lr=3.1888e-04) (hash(x)=6022625)
240
+ 228 train 7.745234 (lr=3.2028e-04) (hash(x)=5884663)
241
+ 229 train 7.843312 (lr=3.2168e-04) (hash(x)=6584810)
242
+ 230 train 7.801966 (lr=3.2308e-04) (hash(x)=5289998)
243
+ 231 train 7.510679 (lr=3.2448e-04) (hash(x)=4906853)
244
+ 232 train 7.579675 (lr=3.2587e-04) (hash(x)=5862071)
245
+ 233 train 7.219170 (lr=3.2727e-04) (hash(x)=4564564)
246
+ 234 train 7.591276 (lr=3.2867e-04) (hash(x)=5723971)
247
+ 235 train 7.708779 (lr=3.3007e-04) (hash(x)=5589269)
248
+ 236 train 7.665421 (lr=3.3147e-04) (hash(x)=5683368)
249
+ 237 train 7.913085 (lr=3.3287e-04) (hash(x)=6969180)
250
+ 238 train 7.796215 (lr=3.3427e-04) (hash(x)=5341421)
251
+ 239 train 7.567434 (lr=3.3566e-04) (hash(x)=6437376)
252
+ 240 train 8.076303 (lr=3.3706e-04) (hash(x)=7527007)
253
+ 241 train 7.919109 (lr=3.3846e-04) (hash(x)=7116258)
254
+ 242 train 7.877451 (lr=3.3986e-04) (hash(x)=6463930)
255
+ 243 train 7.907961 (lr=3.4126e-04) (hash(x)=7304892)
256
+ 244 train 8.047992 (lr=3.4266e-04) (hash(x)=5881808)
257
+ 245 train 7.837342 (lr=3.4406e-04) (hash(x)=7366721)
258
+ 246 train 7.760063 (lr=3.4545e-04) (hash(x)=6457521)
259
+ 247 train 8.144821 (lr=3.4685e-04) (hash(x)=7084093)
260
+ 248 train 7.808409 (lr=3.4825e-04) (hash(x)=6015300)
261
+ 249 train 8.056373 (lr=3.4965e-04) (hash(x)=7197652)
262
+ 250 val loss 7.7773
263
+ 250 val perplexity 2385.8423
264
+ 250 train 8.080597 (lr=3.5105e-04) (hash(x)=4830828)
265
+ 251 train 7.805977 (lr=3.5245e-04) (hash(x)=5908178)
266
+ 252 train 7.789102 (lr=3.5385e-04) (hash(x)=6299092)
267
+ 253 train 7.798859 (lr=3.5524e-04) (hash(x)=6349974)
268
+ 254 train 7.802348 (lr=3.5664e-04) (hash(x)=5641494)
269
+ 255 train 7.781238 (lr=3.5804e-04) (hash(x)=7048804)
270
+ 256 train 8.006168 (lr=3.5944e-04) (hash(x)=7035244)
271
+ 257 train 7.726725 (lr=3.6084e-04) (hash(x)=7909039)
272
+ 258 train 7.948016 (lr=3.6224e-04) (hash(x)=5470939)
273
+ 259 train 7.773250 (lr=3.6364e-04) (hash(x)=6085549)
274
+ 260 train 8.079906 (lr=3.6503e-04) (hash(x)=5882649)
275
+ 261 train 7.917882 (lr=3.6643e-04) (hash(x)=7463181)
276
+ 262 train 8.210739 (lr=3.6783e-04) (hash(x)=5215574)
277
+ 263 train 7.656866 (lr=3.6923e-04) (hash(x)=5752594)
278
+ 264 train 7.814951 (lr=3.7063e-04) (hash(x)=6416619)
279
+ 265 train 7.814353 (lr=3.7203e-04) (hash(x)=6114601)
280
+ 266 train 7.759092 (lr=3.7343e-04) (hash(x)=5646171)
281
+ 267 train 7.688155 (lr=3.7483e-04) (hash(x)=7662769)
282
+ 268 train 7.963375 (lr=3.7622e-04) (hash(x)=6394376)
283
+ 269 train 7.553571 (lr=3.7762e-04) (hash(x)=7485666)
284
+ 270 train 7.640589 (lr=3.7902e-04) (hash(x)=6636823)
285
+ 271 train 7.575523 (lr=3.8042e-04) (hash(x)=6393520)
286
+ 272 train 7.766035 (lr=3.8182e-04) (hash(x)=5235659)
287
+ 273 train 7.842965 (lr=3.8322e-04) (hash(x)=5586291)
288
+ 274 train 8.059671 (lr=3.8462e-04) (hash(x)=7034674)
289
+ 275 train 8.154101 (lr=3.8601e-04) (hash(x)=5942867)
290
+ 276 train 7.854487 (lr=3.8741e-04) (hash(x)=7812344)
291
+ 277 train 8.128255 (lr=3.8881e-04) (hash(x)=7574027)
292
+ 278 train 8.080032 (lr=3.9021e-04) (hash(x)=5960216)
293
+ 279 train 7.866420 (lr=3.9161e-04) (hash(x)=6793550)
294
+ 280 train 7.837950 (lr=3.9301e-04) (hash(x)=6879965)
295
+ 281 train 7.543940 (lr=3.9441e-04) (hash(x)=6264789)
296
+ 282 train 7.698717 (lr=3.9580e-04) (hash(x)=5175356)
297
+ 283 train 7.926527 (lr=3.9720e-04) (hash(x)=7105976)
298
+ 284 train 7.934340 (lr=3.9860e-04) (hash(x)=7286386)
299
+ 285 train 8.124363 (lr=4.0000e-04) (hash(x)=7244279)
300
+ 286 train 7.606567 (lr=4.0000e-04) (hash(x)=4877247)
301
+ 287 train 7.569118 (lr=4.0000e-04) (hash(x)=6581348)
302
+ 288 train 7.564267 (lr=3.9999e-04) (hash(x)=6397208)
303
+ 289 train 7.678329 (lr=3.9998e-04) (hash(x)=6543825)
304
+ 290 train 7.451445 (lr=3.9997e-04) (hash(x)=6088993)
305
+ 291 train 7.783363 (lr=3.9996e-04) (hash(x)=5555598)
306
+ 292 train 7.851149 (lr=3.9994e-04) (hash(x)=6924240)
307
+ 293 train 8.049478 (lr=3.9991e-04) (hash(x)=7254232)
308
+ 294 train 7.581804 (lr=3.9989e-04) (hash(x)=7689743)
309
+ 295 train 7.550489 (lr=3.9986e-04) (hash(x)=6235837)
310
+ 296 train 7.804015 (lr=3.9983e-04) (hash(x)=5216883)
311
+ 297 train 7.435771 (lr=3.9979e-04) (hash(x)=4738550)
312
+ 298 train 7.616967 (lr=3.9975e-04) (hash(x)=5126918)
313
+ 299 train 7.774863 (lr=3.9971e-04) (hash(x)=5591770)
314
+ 300 val loss 7.7129
315
+ 300 val perplexity 2237.1272
316
+ 300 train 8.668016 (lr=3.9966e-04) (hash(x)=8001573)
317
+ 301 train 8.370278 (lr=3.9961e-04) (hash(x)=8012491)
318
+ 302 train 8.050730 (lr=3.9955e-04) (hash(x)=3997428)
319
+ 303 train 7.358860 (lr=3.9950e-04) (hash(x)=7083021)
320
+ 304 train 7.944624 (lr=3.9944e-04) (hash(x)=7177371)
321
+ 305 train 7.770327 (lr=3.9937e-04) (hash(x)=6456022)
322
+ 306 train 7.872246 (lr=3.9930e-04) (hash(x)=5238861)
323
+ 307 train 7.862182 (lr=3.9923e-04) (hash(x)=7281348)
324
+ 308 train 7.997207 (lr=3.9916e-04) (hash(x)=6848441)
325
+ 309 train 8.047302 (lr=3.9908e-04) (hash(x)=7824794)
326
+ 310 train 8.186137 (lr=3.9900e-04) (hash(x)=7440208)
327
+ 311 train 7.972701 (lr=3.9891e-04) (hash(x)=6155821)
328
+ 312 train 7.693457 (lr=3.9882e-04) (hash(x)=5350461)
329
+ 313 train 7.940413 (lr=3.9873e-04) (hash(x)=7807564)
330
+ 314 train 7.728491 (lr=3.9864e-04) (hash(x)=5992165)
331
+ 315 train 7.688345 (lr=3.9854e-04) (hash(x)=5736241)
332
+ 316 train 7.666516 (lr=3.9843e-04) (hash(x)=6536253)
333
+ 317 train 7.865399 (lr=3.9833e-04) (hash(x)=5688415)
334
+ 318 train 7.497298 (lr=3.9822e-04) (hash(x)=5839705)
335
+ 319 train 7.677273 (lr=3.9811e-04) (hash(x)=5657123)
336
+ 320 train 7.717196 (lr=3.9799e-04) (hash(x)=6354364)
337
+ 321 train 7.809872 (lr=3.9787e-04) (hash(x)=6102985)
338
+ 322 train 8.071723 (lr=3.9775e-04) (hash(x)=7161967)
339
+ 323 train 8.007324 (lr=3.9762e-04) (hash(x)=6452095)
340
+ 324 train 7.631840 (lr=3.9749e-04) (hash(x)=5547396)
341
+ 325 train 7.940474 (lr=3.9736e-04) (hash(x)=5848921)
342
+ 326 train 7.720636 (lr=3.9722e-04) (hash(x)=4890294)
343
+ 327 train 7.432765 (lr=3.9708e-04) (hash(x)=5312267)
344
+ 328 train 7.513749 (lr=3.9694e-04) (hash(x)=5348929)
345
+ 329 train 7.471559 (lr=3.9679e-04) (hash(x)=5025297)
346
+ 330 train 7.505089 (lr=3.9664e-04) (hash(x)=5670052)
347
+ 331 train 7.356798 (lr=3.9648e-04) (hash(x)=5434493)
348
+ 332 train 8.054082 (lr=3.9633e-04) (hash(x)=9122453)
349
+ 333 train 7.766166 (lr=3.9616e-04) (hash(x)=5197474)
350
+ 334 train 7.603662 (lr=3.9600e-04) (hash(x)=5550786)
351
+ 335 train 7.666357 (lr=3.9583e-04) (hash(x)=6830813)
352
+ 336 train 7.490107 (lr=3.9566e-04) (hash(x)=5258435)
353
+ 337 train 7.524698 (lr=3.9549e-04) (hash(x)=5242548)
354
+ 338 train 7.362160 (lr=3.9531e-04) (hash(x)=3754886)
355
+ 339 train 7.215922 (lr=3.9513e-04) (hash(x)=4752771)
356
+ 340 train 7.464192 (lr=3.9494e-04) (hash(x)=5926875)
357
+ 341 train 8.004081 (lr=3.9475e-04) (hash(x)=6691249)
358
+ 342 train 7.723438 (lr=3.9456e-04) (hash(x)=7514623)
359
+ 343 train 7.812165 (lr=3.9437e-04) (hash(x)=6424933)
360
+ 344 train 7.747590 (lr=3.9417e-04) (hash(x)=6347757)
361
+ 345 train 7.708298 (lr=3.9397e-04) (hash(x)=6960120)
362
+ 346 train 7.726888 (lr=3.9376e-04) (hash(x)=7597197)
363
+ 347 train 7.654065 (lr=3.9356e-04) (hash(x)=5786517)
364
+ 348 train 7.732175 (lr=3.9334e-04) (hash(x)=6424363)
365
+ 349 train 7.754531 (lr=3.9313e-04) (hash(x)=7229689)
366
+ 350 val loss 7.6790
367
+ 350 val perplexity 2162.5024
368
+ 350 train 7.852203 (lr=3.9291e-04) (hash(x)=6772284)
369
+ 351 train 7.939065 (lr=3.9269e-04) (hash(x)=6680023)
370
+ 352 train 7.837459 (lr=3.9246e-04) (hash(x)=6218520)
371
+ 353 train 7.896614 (lr=3.9223e-04) (hash(x)=7688393)
372
+ 354 train 7.720498 (lr=3.9200e-04) (hash(x)=6327977)
373
+ 355 train 7.606984 (lr=3.9177e-04) (hash(x)=6474729)
374
+ 356 train 7.730735 (lr=3.9153e-04) (hash(x)=6657266)
375
+ 357 train 7.744442 (lr=3.9129e-04) (hash(x)=5946650)
376
+ 358 train 7.644190 (lr=3.9104e-04) (hash(x)=6734878)
377
+ 359 train 7.726983 (lr=3.9079e-04) (hash(x)=7523279)
378
+ 360 train 7.573905 (lr=3.9054e-04) (hash(x)=6179478)
379
+ 361 train 7.626390 (lr=3.9029e-04) (hash(x)=5531605)
380
+ 362 train 7.613953 (lr=3.9003e-04) (hash(x)=5844980)
381
+ 363 train 7.281720 (lr=3.8977e-04) (hash(x)=5508768)
382
+ 364 train 7.134599 (lr=3.8950e-04) (hash(x)=6337190)
383
+ 365 train 8.774722 (lr=3.8923e-04) (hash(x)=6415982)
384
+ 366 train 7.787969 (lr=3.8896e-04) (hash(x)=6828959)
385
+ 367 train 7.746159 (lr=3.8869e-04) (hash(x)=6635925)
386
+ 368 train 7.471364 (lr=3.8841e-04) (hash(x)=6358540)
387
+ 369 train 7.853199 (lr=3.8813e-04) (hash(x)=5923706)
388
+ 370 train 7.516103 (lr=3.8785e-04) (hash(x)=4843320)
389
+ 371 train 7.742843 (lr=3.8756e-04) (hash(x)=6663801)
390
+ 372 train 7.572280 (lr=3.8727e-04) (hash(x)=5583287)
391
+ 373 train 7.525884 (lr=3.8697e-04) (hash(x)=6113540)
392
+ 374 train 7.653235 (lr=3.8667e-04) (hash(x)=6389936)
393
+ 375 train 7.604330 (lr=3.8637e-04) (hash(x)=5869441)
394
+ 376 train 7.666893 (lr=3.8607e-04) (hash(x)=5148617)
395
+ 377 train 7.929458 (lr=3.8576e-04) (hash(x)=6913454)
396
+ 378 train 7.606462 (lr=3.8545e-04) (hash(x)=5356233)
397
+ 379 train 7.587533 (lr=3.8514e-04) (hash(x)=6698878)
398
+ 380 train 7.561080 (lr=3.8482e-04) (hash(x)=6280665)
399
+ 381 train 7.779469 (lr=3.8450e-04) (hash(x)=6094899)
400
+ 382 train 7.309037 (lr=3.8418e-04) (hash(x)=6249746)
401
+ 383 train 7.636660 (lr=3.8385e-04) (hash(x)=6349004)
402
+ 384 train 7.577253 (lr=3.8352e-04) (hash(x)=7013468)
403
+ 385 train 7.548975 (lr=3.8319e-04) (hash(x)=6045933)
404
+ 386 train 7.871435 (lr=3.8286e-04) (hash(x)=6571735)
405
+ 387 train 7.498989 (lr=3.8252e-04) (hash(x)=4986137)
406
+ 388 train 7.167762 (lr=3.8217e-04) (hash(x)=5244912)
407
+ 389 train 7.498528 (lr=3.8183e-04) (hash(x)=4798229)
408
+ 390 train 7.577496 (lr=3.8148e-04) (hash(x)=5815783)
409
+ 391 train 7.346290 (lr=3.8113e-04) (hash(x)=6008454)
410
+ 392 train 7.551310 (lr=3.8077e-04) (hash(x)=6407333)
411
+ 393 train 7.436758 (lr=3.8042e-04) (hash(x)=5938362)
412
+ 394 train 7.289711 (lr=3.8006e-04) (hash(x)=6077124)
413
+ 395 train 7.600750 (lr=3.7969e-04) (hash(x)=6550770)
414
+ 396 train 7.420817 (lr=3.7933e-04) (hash(x)=6181528)
415
+ 397 train 7.688013 (lr=3.7896e-04) (hash(x)=7055344)
416
+ 398 train 7.801223 (lr=3.7858e-04) (hash(x)=7689348)
417
+ 399 train 7.725225 (lr=3.7821e-04) (hash(x)=7682741)
418
+ 400 val loss 7.6094
419
+ 400 val perplexity 2017.1411
420
+ 400 train 7.511962 (lr=3.7783e-04) (hash(x)=5189401)
421
+ 401 train 7.518741 (lr=3.7744e-04) (hash(x)=7372095)
422
+ 402 train 7.668220 (lr=3.7706e-04) (hash(x)=6838219)
423
+ 403 train 8.134772 (lr=3.7667e-04) (hash(x)=7892158)
424
+ 404 train 7.775495 (lr=3.7628e-04) (hash(x)=5677077)
425
+ 405 train 7.401995 (lr=3.7588e-04) (hash(x)=5929740)
426
+ 406 train 7.558940 (lr=3.7549e-04) (hash(x)=5285055)
427
+ 407 train 7.650312 (lr=3.7509e-04) (hash(x)=7794028)
428
+ 408 train 7.726353 (lr=3.7468e-04) (hash(x)=6640802)
429
+ 409 train 7.822291 (lr=3.7428e-04) (hash(x)=6839923)
430
+ 410 train 7.277285 (lr=3.7387e-04) (hash(x)=4020997)
431
+ 411 train 7.458729 (lr=3.7345e-04) (hash(x)=7093523)
432
+ 412 train 7.784377 (lr=3.7304e-04) (hash(x)=7370045)
433
+ 413 train 7.507353 (lr=3.7262e-04) (hash(x)=5341312)
434
+ 414 train 7.512074 (lr=3.7220e-04) (hash(x)=5589312)
435
+ 415 train 7.986567 (lr=3.7177e-04) (hash(x)=6166062)
436
+ 416 train 7.887094 (lr=3.7135e-04) (hash(x)=8354149)
437
+ 417 train 7.509351 (lr=3.7092e-04) (hash(x)=5992164)
438
+ 418 train 7.547449 (lr=3.7048e-04) (hash(x)=4748657)
439
+ 419 train 7.541067 (lr=3.7005e-04) (hash(x)=6645781)
440
+ 420 train 7.770974 (lr=3.6961e-04) (hash(x)=6689147)
441
+ 421 train 7.642218 (lr=3.6917e-04) (hash(x)=6207724)
442
+ 422 train 7.557092 (lr=3.6872e-04) (hash(x)=6232863)
443
+ 423 train 7.306893 (lr=3.6828e-04) (hash(x)=4811192)
444
+ 424 train 7.341501 (lr=3.6782e-04) (hash(x)=6211761)
445
+ 425 train 7.530269 (lr=3.6737e-04) (hash(x)=6906335)
446
+ 426 train 7.448237 (lr=3.6692e-04) (hash(x)=4922365)
447
+ 427 train 7.491903 (lr=3.6646e-04) (hash(x)=5970866)
448
+ 428 train 7.444016 (lr=3.6599e-04) (hash(x)=6417833)
449
+ 429 train 7.543698 (lr=3.6553e-04) (hash(x)=5238015)
450
+ 430 train 7.432292 (lr=3.6506e-04) (hash(x)=5375451)
451
+ 431 train 7.556702 (lr=3.6459e-04) (hash(x)=7822680)
452
+ 432 train 7.571167 (lr=3.6412e-04) (hash(x)=6173200)
453
+ 433 train 7.307553 (lr=3.6364e-04) (hash(x)=5478019)
454
+ 434 train 7.529016 (lr=3.6316e-04) (hash(x)=5597859)
455
+ 435 train 7.444973 (lr=3.6268e-04) (hash(x)=6317317)
456
+ 436 train 7.626946 (lr=3.6220e-04) (hash(x)=7120467)
457
+ 437 train 7.724288 (lr=3.6171e-04) (hash(x)=5832344)
458
+ 438 train 7.608081 (lr=3.6122e-04) (hash(x)=6813802)
459
+ 439 train 7.667047 (lr=3.6073e-04) (hash(x)=6431409)
460
+ 440 train 7.744073 (lr=3.6023e-04) (hash(x)=5886861)
461
+ 441 train 7.410055 (lr=3.5974e-04) (hash(x)=5900130)
462
+ 442 train 7.698512 (lr=3.5924e-04) (hash(x)=6061563)
463
+ 443 train 7.647983 (lr=3.5873e-04) (hash(x)=6653337)
464
+ 444 train 7.555354 (lr=3.5823e-04) (hash(x)=6575992)
465
+ 445 train 7.568292 (lr=3.5772e-04) (hash(x)=6696784)
466
+ 446 train 7.717435 (lr=3.5721e-04) (hash(x)=6242161)
467
+ 447 train 7.539690 (lr=3.5669e-04) (hash(x)=5323032)
468
+ 448 train 7.336827 (lr=3.5618e-04) (hash(x)=6439367)
469
+ 449 train 7.581370 (lr=3.5566e-04) (hash(x)=5372100)
470
+ 450 val loss 7.5729
471
+ 450 val perplexity 1944.7410
472
+ 450 train 7.643305 (lr=3.5514e-04) (hash(x)=6344849)
473
+ 451 train 7.298000 (lr=3.5461e-04) (hash(x)=5125339)
474
+ 452 train 7.260176 (lr=3.5408e-04) (hash(x)=4876156)
475
+ 453 train 7.400402 (lr=3.5355e-04) (hash(x)=5664463)
476
+ 454 train 7.294392 (lr=3.5302e-04) (hash(x)=6010042)
477
+ 455 train 7.540538 (lr=3.5249e-04) (hash(x)=7363286)
478
+ 456 train 8.173691 (lr=3.5195e-04) (hash(x)=4873268)
479
+ 457 train 7.884413 (lr=3.5141e-04) (hash(x)=7444468)
480
+ 458 train 7.905367 (lr=3.5087e-04) (hash(x)=6698134)
481
+ 459 train 7.703500 (lr=3.5032e-04) (hash(x)=7670050)
482
+ 460 train 7.708179 (lr=3.4977e-04) (hash(x)=6280866)
483
+ 461 train 7.541913 (lr=3.4922e-04) (hash(x)=8189420)
484
+ 462 train 7.666384 (lr=3.4867e-04) (hash(x)=6878292)
485
+ 463 train 7.350158 (lr=3.4812e-04) (hash(x)=5616075)
486
+ 464 train 7.556487 (lr=3.4756e-04) (hash(x)=5839939)
487
+ 465 train 7.437470 (lr=3.4700e-04) (hash(x)=6360924)
488
+ 466 train 7.913586 (lr=3.4644e-04) (hash(x)=6466475)
489
+ 467 train 7.657649 (lr=3.4587e-04) (hash(x)=6593764)
490
+ 468 train 8.092769 (lr=3.4530e-04) (hash(x)=7389698)
491
+ 469 train 7.748988 (lr=3.4473e-04) (hash(x)=6764377)
492
+ 470 train 7.509034 (lr=3.4416e-04) (hash(x)=7293570)
493
+ 471 train 7.544131 (lr=3.4359e-04) (hash(x)=5672608)
494
+ 472 train 7.408926 (lr=3.4301e-04) (hash(x)=5587975)
495
+ 473 train 7.305445 (lr=3.4243e-04) (hash(x)=6985256)
496
+ 474 train 7.932063 (lr=3.4185e-04) (hash(x)=6549231)
497
+ 475 train 7.584744 (lr=3.4127e-04) (hash(x)=6828653)
498
+ 476 train 7.410558 (lr=3.4068e-04) (hash(x)=5478970)
499
+ 477 train 7.449081 (lr=3.4009e-04) (hash(x)=5199845)
500
+ 478 train 7.520986 (lr=3.3950e-04) (hash(x)=6482073)
501
+ 479 train 7.604218 (lr=3.3891e-04) (hash(x)=6839867)
502
+ 480 train 7.603945 (lr=3.3831e-04) (hash(x)=7073454)
503
+ 481 train 7.904164 (lr=3.3771e-04) (hash(x)=6928712)
504
+ 482 train 7.573225 (lr=3.3711e-04) (hash(x)=5962982)
505
+ 483 train 7.646687 (lr=3.3651e-04) (hash(x)=8426992)
506
+ 484 train 7.570432 (lr=3.3590e-04) (hash(x)=6422394)
507
+ 485 train 7.563037 (lr=3.3530e-04) (hash(x)=5330355)
508
+ 486 train 7.035078 (lr=3.3469e-04) (hash(x)=4111491)
509
+ 487 train 7.035528 (lr=3.3408e-04) (hash(x)=4916928)
510
+ 488 train 7.207923 (lr=3.3346e-04) (hash(x)=6610244)
511
+ 489 train 7.738671 (lr=3.3285e-04) (hash(x)=7287928)
512
+ 490 train 7.586032 (lr=3.3223e-04) (hash(x)=6525083)
513
+ 491 train 7.549232 (lr=3.3161e-04) (hash(x)=6484050)
514
+ 492 train 7.689845 (lr=3.3099e-04) (hash(x)=6573424)
515
+ 493 train 7.540273 (lr=3.3036e-04) (hash(x)=5062850)
516
+ 494 train 7.517158 (lr=3.2973e-04) (hash(x)=6215995)
517
+ 495 train 7.628617 (lr=3.2911e-04) (hash(x)=8353379)
518
+ 496 train 7.506550 (lr=3.2847e-04) (hash(x)=6344823)
519
+ 497 train 7.503284 (lr=3.2784e-04) (hash(x)=6805358)
520
+ 498 train 7.402885 (lr=3.2721e-04) (hash(x)=6018070)
521
+ 499 train 7.670665 (lr=3.2657e-04) (hash(x)=6552510)
522
+ 500 val loss 7.5385
523
+ 500 val perplexity 1878.9983
524
+ 500 train 7.472359 (lr=3.2593e-04) (hash(x)=7218108)
525
+ 501 train 7.663666 (lr=3.2529e-04) (hash(x)=5105545)
526
+ 502 train 7.650087 (lr=3.2464e-04) (hash(x)=5997860)
527
+ 503 train 7.435174 (lr=3.2400e-04) (hash(x)=4838871)
528
+ 504 train 7.471886 (lr=3.2335e-04) (hash(x)=7105888)
529
+ 505 train 7.803447 (lr=3.2270e-04) (hash(x)=5945286)
530
+ 506 train 6.972367 (lr=3.2205e-04) (hash(x)=4381720)
531
+ 507 train 6.651798 (lr=3.2140e-04) (hash(x)=3021697)
532
+ 508 train 7.056805 (lr=3.2074e-04) (hash(x)=6826973)
533
+ 509 train 7.653838 (lr=3.2008e-04) (hash(x)=6443070)
534
+ 510 train 7.341478 (lr=3.1943e-04) (hash(x)=5520637)
535
+ 511 train 7.597086 (lr=3.1876e-04) (hash(x)=6795665)
536
+ 512 train 7.391040 (lr=3.1810e-04) (hash(x)=6063891)
537
+ 513 train 7.892898 (lr=3.1744e-04) (hash(x)=7683068)
538
+ 514 train 7.155515 (lr=3.1677e-04) (hash(x)=5526694)
539
+ 515 train 7.179341 (lr=3.1610e-04) (hash(x)=5486935)
540
+ 516 train 7.567066 (lr=3.1543e-04) (hash(x)=6002252)
541
+ 517 train 7.534157 (lr=3.1476e-04) (hash(x)=6115232)
542
+ 518 train 7.810001 (lr=3.1408e-04) (hash(x)=9039136)
543
+ 519 train 7.467688 (lr=3.1341e-04) (hash(x)=6678038)
544
+ 520 train 7.619728 (lr=3.1273e-04) (hash(x)=5383926)
545
+ 521 train 7.672162 (lr=3.1205e-04) (hash(x)=5747629)
546
+ 522 train 7.557688 (lr=3.1137e-04) (hash(x)=7090040)
547
+ 523 train 7.751288 (lr=3.1069e-04) (hash(x)=6657714)
548
+ 524 train 7.519429 (lr=3.1000e-04) (hash(x)=5414034)
549
+ 525 train 7.539899 (lr=3.0931e-04) (hash(x)=6190642)
550
+ 526 train 7.662959 (lr=3.0862e-04) (hash(x)=7431177)
551
+ 527 train 7.527627 (lr=3.0793e-04) (hash(x)=6112215)
552
+ 528 train 7.646883 (lr=3.0724e-04) (hash(x)=6785874)
553
+ 529 train 7.409170 (lr=3.0655e-04) (hash(x)=7314989)
554
+ 530 train 7.627702 (lr=3.0585e-04) (hash(x)=6222614)
555
+ 531 train 7.712043 (lr=3.0516e-04) (hash(x)=8353143)
556
+ 532 train 7.794209 (lr=3.0446e-04) (hash(x)=6752498)
557
+ 533 train 7.523862 (lr=3.0376e-04) (hash(x)=5912570)
558
+ 534 train 7.683999 (lr=3.0306e-04) (hash(x)=5621785)
559
+ 535 train 7.583555 (lr=3.0235e-04) (hash(x)=5915361)
560
+ 536 train 7.408762 (lr=3.0165e-04) (hash(x)=6853672)
561
+ 537 train 7.357994 (lr=3.0094e-04) (hash(x)=6369494)
562
+ 538 train 7.551110 (lr=3.0023e-04) (hash(x)=6039652)
563
+ 539 train 7.680251 (lr=2.9952e-04) (hash(x)=6254885)
564
+ 540 train 7.523490 (lr=2.9881e-04) (hash(x)=5829860)
565
+ 541 train 7.459058 (lr=2.9810e-04) (hash(x)=5425120)
566
+ 542 train 7.214892 (lr=2.9738e-04) (hash(x)=5209746)
567
+ 543 train 7.830770 (lr=2.9667e-04) (hash(x)=5771588)
568
+ 544 train 7.543005 (lr=2.9595e-04) (hash(x)=8298790)
569
+ 545 train 7.579824 (lr=2.9523e-04) (hash(x)=6763967)
570
+ 546 train 7.572336 (lr=2.9451e-04) (hash(x)=4882397)
571
+ 547 train 7.336516 (lr=2.9379e-04) (hash(x)=5561507)
572
+ 548 train 7.339019 (lr=2.9307e-04) (hash(x)=5762755)
573
+ 549 train 7.460457 (lr=2.9234e-04) (hash(x)=7081151)
574
+ 550 val loss 7.4936
575
+ 550 val perplexity 1796.4297
576
+ 550 train 7.356226 (lr=2.9162e-04) (hash(x)=5047520)
577
+ 551 train 7.495140 (lr=2.9089e-04) (hash(x)=5688829)
578
+ 552 train 7.373357 (lr=2.9016e-04) (hash(x)=7008925)
579
+ 553 train 7.523464 (lr=2.8943e-04) (hash(x)=7821554)
580
+ 554 train 7.538155 (lr=2.8870e-04) (hash(x)=6469035)
581
+ 555 train 7.351624 (lr=2.8797e-04) (hash(x)=5371951)
582
+ 556 train 7.655009 (lr=2.8723e-04) (hash(x)=6436050)
583
+ 557 train 7.574912 (lr=2.8650e-04) (hash(x)=8120149)
584
+ 558 train 7.821643 (lr=2.8576e-04) (hash(x)=6189935)
585
+ 559 train 7.515394 (lr=2.8502e-04) (hash(x)=5443305)
586
+ 560 train 7.617784 (lr=2.8428e-04) (hash(x)=6294312)
587
+ 561 train 7.236239 (lr=2.8354e-04) (hash(x)=7066030)
588
+ 562 train 7.650414 (lr=2.8280e-04) (hash(x)=7744989)
589
+ 563 train 7.790557 (lr=2.8206e-04) (hash(x)=5016757)
590
+ 564 train 7.587486 (lr=2.8132e-04) (hash(x)=8001284)
591
+ 565 train 7.524496 (lr=2.8057e-04) (hash(x)=6296163)
592
+ 566 train 7.660974 (lr=2.7982e-04) (hash(x)=6027688)
593
+ 567 train 7.569398 (lr=2.7908e-04) (hash(x)=6901933)
594
+ 568 train 7.437927 (lr=2.7833e-04) (hash(x)=5124305)
595
+ 569 train 7.547812 (lr=2.7758e-04) (hash(x)=8056633)
596
+ 570 train 7.617347 (lr=2.7683e-04) (hash(x)=6677566)
597
+ 571 train 7.645174 (lr=2.7607e-04) (hash(x)=6019085)
598
+ 572 train 7.398640 (lr=2.7532e-04) (hash(x)=5924495)
599
+ 573 train 7.496363 (lr=2.7457e-04) (hash(x)=7003893)
600
+ 574 train 7.461383 (lr=2.7381e-04) (hash(x)=5563075)
601
+ 575 train 7.270102 (lr=2.7306e-04) (hash(x)=4294425)
602
+ 576 train 7.459836 (lr=2.7230e-04) (hash(x)=5677870)
603
+ 577 train 7.345787 (lr=2.7154e-04) (hash(x)=6568540)
604
+ 578 train 7.579265 (lr=2.7078e-04) (hash(x)=7473394)
605
+ 579 train 7.606874 (lr=2.7002e-04) (hash(x)=6095229)
606
+ 580 train 7.610292 (lr=2.6926e-04) (hash(x)=6946168)
607
+ 581 train 7.416889 (lr=2.6850e-04) (hash(x)=5512150)
608
+ 582 train 7.065933 (lr=2.6773e-04) (hash(x)=4380268)
609
+ 583 train 7.130316 (lr=2.6697e-04) (hash(x)=4363941)
610
+ 584 train 7.292055 (lr=2.6620e-04) (hash(x)=4836009)
611
+ 585 train 7.551021 (lr=2.6544e-04) (hash(x)=6457783)
612
+ 586 train 7.422904 (lr=2.6467e-04) (hash(x)=6378784)
613
+ 587 train 7.336568 (lr=2.6390e-04) (hash(x)=4693798)
614
+ 588 train 7.671266 (lr=2.6314e-04) (hash(x)=9193310)
615
+ 589 train 8.412244 (lr=2.6237e-04) (hash(x)=8759806)
616
+ 590 train 8.784805 (lr=2.6160e-04) (hash(x)=9176140)
617
+ 591 train 7.467399 (lr=2.6083e-04) (hash(x)=5486335)
618
+ 592 train 7.676648 (lr=2.6005e-04) (hash(x)=7868436)
619
+ 593 train 7.575103 (lr=2.5928e-04) (hash(x)=6964993)
620
+ 594 train 7.791656 (lr=2.5851e-04) (hash(x)=6741270)
621
+ 595 train 8.048973 (lr=2.5773e-04) (hash(x)=7907450)
622
+ 596 train 7.905000 (lr=2.5696e-04) (hash(x)=6443452)
623
+ 597 train 7.502116 (lr=2.5618e-04) (hash(x)=5236642)
624
+ 598 train 7.662479 (lr=2.5541e-04) (hash(x)=5604795)
625
+ 599 train 7.613657 (lr=2.5463e-04) (hash(x)=7295165)
626
+ 600 val loss 7.5941
627
+ 600 val perplexity 1986.3958
628
+ 600 train 7.598695 (lr=2.5385e-04) (hash(x)=6034549)
629
+ 601 train 7.626319 (lr=2.5307e-04) (hash(x)=4913602)
630
+ 602 train 7.583829 (lr=2.5230e-04) (hash(x)=6271294)
631
+ 603 train 7.656682 (lr=2.5152e-04) (hash(x)=6814026)
632
+ 604 train 7.544633 (lr=2.5074e-04) (hash(x)=6848040)
633
+ 605 train 7.775752 (lr=2.4996e-04) (hash(x)=7042051)
634
+ 606 train 7.718560 (lr=2.4917e-04) (hash(x)=6481002)
635
+ 607 train 7.384568 (lr=2.4839e-04) (hash(x)=6267424)
636
+ 608 train 7.670825 (lr=2.4761e-04) (hash(x)=7665306)
637
+ 609 train 7.614493 (lr=2.4683e-04) (hash(x)=5614727)
638
+ 610 train 7.607079 (lr=2.4604e-04) (hash(x)=7039197)
639
+ 611 train 7.670383 (lr=2.4526e-04) (hash(x)=8086437)
640
+ 612 train 7.734481 (lr=2.4448e-04) (hash(x)=6993846)
641
+ 613 train 7.768439 (lr=2.4369e-04) (hash(x)=5900143)
642
+ 614 train 7.634926 (lr=2.4291e-04) (hash(x)=4841318)
643
+ 615 train 7.609076 (lr=2.4212e-04) (hash(x)=5270452)
644
+ 616 train 7.519147 (lr=2.4133e-04) (hash(x)=5955026)
645
+ 617 train 7.589430 (lr=2.4055e-04) (hash(x)=8617707)
646
+ 618 train 7.421934 (lr=2.3976e-04) (hash(x)=5159401)
647
+ 619 train 7.433770 (lr=2.3897e-04) (hash(x)=6420820)
648
+ 620 train 7.409893 (lr=2.3818e-04) (hash(x)=6628863)
649
+ 621 train 7.219546 (lr=2.3740e-04) (hash(x)=6215930)
650
+ 622 train 7.556403 (lr=2.3661e-04) (hash(x)=5248645)
651
+ 623 train 7.451512 (lr=2.3582e-04) (hash(x)=6305297)
652
+ 624 train 8.019797 (lr=2.3503e-04) (hash(x)=6989704)
653
+ 625 train 7.530154 (lr=2.3424e-04) (hash(x)=4026682)
654
+ 626 train 7.409697 (lr=2.3345e-04) (hash(x)=6367688)
655
+ 627 train 7.472428 (lr=2.3266e-04) (hash(x)=7889849)
656
+ 628 train 7.642259 (lr=2.3187e-04) (hash(x)=7334719)
657
+ 629 train 7.769471 (lr=2.3108e-04) (hash(x)=6141721)
658
+ 630 train 8.547319 (lr=2.3029e-04) (hash(x)=5056572)
659
+ 631 train 7.972347 (lr=2.2950e-04) (hash(x)=6040077)
660
+ 632 train 8.060642 (lr=2.2871e-04) (hash(x)=6866651)
661
+ 633 train 8.293114 (lr=2.2792e-04) (hash(x)=6155449)
662
+ 634 train 7.873876 (lr=2.2713e-04) (hash(x)=6690174)
663
+ 635 train 7.580360 (lr=2.2633e-04) (hash(x)=5652497)
664
+ 636 train 7.625008 (lr=2.2554e-04) (hash(x)=5931050)
665
+ 637 train 7.704155 (lr=2.2475e-04) (hash(x)=6961314)
666
+ 638 train 7.488312 (lr=2.2396e-04) (hash(x)=5973636)
667
+ 639 train 7.489496 (lr=2.2317e-04) (hash(x)=7130251)
668
+ 640 train 7.263834 (lr=2.2238e-04) (hash(x)=7503390)
669
+ 641 train 7.272105 (lr=2.2158e-04) (hash(x)=5927461)
670
+ 642 train 7.510231 (lr=2.2079e-04) (hash(x)=6196741)
671
+ 643 train 7.378845 (lr=2.2000e-04) (hash(x)=6610177)
672
+ 644 train 7.352052 (lr=2.1921e-04) (hash(x)=6635147)
673
+ 645 train 7.656536 (lr=2.1842e-04) (hash(x)=7277580)
674
+ 646 train 7.282983 (lr=2.1762e-04) (hash(x)=5050080)
675
+ 647 train 7.363386 (lr=2.1683e-04) (hash(x)=6508350)
676
+ 648 train 7.595304 (lr=2.1604e-04) (hash(x)=5276338)
677
+ 649 train 7.743566 (lr=2.1525e-04) (hash(x)=6536034)
678
+ 650 val loss 7.4993
679
+ 650 val perplexity 1806.7867
680
+ 650 train 7.540031 (lr=2.1446e-04) (hash(x)=6944772)
681
+ 651 train 7.556524 (lr=2.1367e-04) (hash(x)=6994983)
682
+ 652 train 7.513436 (lr=2.1287e-04) (hash(x)=7172017)
683
+ 653 train 7.572359 (lr=2.1208e-04) (hash(x)=8700721)
684
+ 654 train 7.562931 (lr=2.1129e-04) (hash(x)=6774360)
685
+ 655 train 7.394775 (lr=2.1050e-04) (hash(x)=5859576)
686
+ 656 train 7.122480 (lr=2.0971e-04) (hash(x)=5899275)
687
+ 657 train 7.059240 (lr=2.0892e-04) (hash(x)=5264962)
688
+ 658 train 6.860204 (lr=2.0813e-04) (hash(x)=5679861)
689
+ 659 train 7.304075 (lr=2.0734e-04) (hash(x)=5487065)
690
+ 660 train 7.066036 (lr=2.0655e-04) (hash(x)=4239476)
691
+ 661 train 7.380940 (lr=2.0576e-04) (hash(x)=5731624)
692
+ 662 train 7.522326 (lr=2.0497e-04) (hash(x)=5883465)
693
+ 663 train 7.277260 (lr=2.0418e-04) (hash(x)=4892065)
694
+ 664 train 7.116191 (lr=2.0339e-04) (hash(x)=5858782)
695
+ 665 train 7.418376 (lr=2.0260e-04) (hash(x)=5489496)
696
+ 666 train 7.080954 (lr=2.0182e-04) (hash(x)=4485195)
697
+ 667 train 7.119433 (lr=2.0103e-04) (hash(x)=4933674)
698
+ 668 train 7.377406 (lr=2.0024e-04) (hash(x)=5746292)
699
+ 669 train 7.380608 (lr=1.9945e-04) (hash(x)=7021003)
700
+ 670 train 7.578075 (lr=1.9867e-04) (hash(x)=5876710)
701
+ 671 train 7.622410 (lr=1.9788e-04) (hash(x)=7317289)
702
+ 672 train 7.732048 (lr=1.9709e-04) (hash(x)=5598226)
703
+ 673 train 7.517103 (lr=1.9631e-04) (hash(x)=7869305)
704
+ 674 train 7.592114 (lr=1.9552e-04) (hash(x)=6611408)
705
+ 675 train 7.565008 (lr=1.9474e-04) (hash(x)=6811522)
706
+ 676 train 7.554715 (lr=1.9396e-04) (hash(x)=6704714)
707
+ 677 train 7.417961 (lr=1.9317e-04) (hash(x)=6601423)
708
+ 678 train 7.628778 (lr=1.9239e-04) (hash(x)=6726071)
709
+ 679 train 7.073271 (lr=1.9161e-04) (hash(x)=5510218)
710
+ 680 train 7.881392 (lr=1.9083e-04) (hash(x)=7950952)
711
+ 681 train 7.560656 (lr=1.9004e-04) (hash(x)=7180298)
712
+ 682 train 7.378563 (lr=1.8926e-04) (hash(x)=6068813)
713
+ 683 train 7.513981 (lr=1.8848e-04) (hash(x)=7304235)
714
+ 684 train 7.530277 (lr=1.8770e-04) (hash(x)=7441806)
715
+ 685 train 7.653310 (lr=1.8693e-04) (hash(x)=8111920)
716
+ 686 train 8.254439 (lr=1.8615e-04) (hash(x)=6222783)
717
+ 687 train 7.511406 (lr=1.8537e-04) (hash(x)=6752265)
718
+ 688 train 7.330371 (lr=1.8459e-04) (hash(x)=6147634)
719
+ 689 train 7.367846 (lr=1.8382e-04) (hash(x)=6788720)
720
+ 690 train 7.493330 (lr=1.8304e-04) (hash(x)=6413518)
721
+ 691 train 7.387491 (lr=1.8227e-04) (hash(x)=5994476)
722
+ 692 train 7.327689 (lr=1.8149e-04) (hash(x)=5462082)
723
+ 693 train 7.312479 (lr=1.8072e-04) (hash(x)=5862533)
724
+ 694 train 7.480178 (lr=1.7995e-04) (hash(x)=7132796)
725
+ 695 train 7.392669 (lr=1.7917e-04) (hash(x)=6530867)
726
+ 696 train 7.427022 (lr=1.7840e-04) (hash(x)=7932207)
727
+ 697 train 7.381664 (lr=1.7763e-04) (hash(x)=5914734)
728
+ 698 train 7.570195 (lr=1.7686e-04) (hash(x)=6361594)
729
+ 699 train 7.522934 (lr=1.7610e-04) (hash(x)=5746260)
730
+ 700 val loss 7.4286
731
+ 700 val perplexity 1683.3726
732
+ 700 train 7.453722 (lr=1.7533e-04) (hash(x)=7065743)
733
+ 701 train 7.390742 (lr=1.7456e-04) (hash(x)=6856097)
734
+ 702 train 7.463803 (lr=1.7380e-04) (hash(x)=7588648)
735
+ 703 train 7.390601 (lr=1.7303e-04) (hash(x)=5790078)
736
+ 704 train 7.385214 (lr=1.7227e-04) (hash(x)=6069422)
737
+ 705 train 7.735109 (lr=1.7150e-04) (hash(x)=5805002)
738
+ 706 train 7.402860 (lr=1.7074e-04) (hash(x)=5344711)
739
+ 707 train 7.274633 (lr=1.6998e-04) (hash(x)=6430135)
740
+ 708 train 7.365939 (lr=1.6922e-04) (hash(x)=6317763)
741
+ 709 train 7.301821 (lr=1.6846e-04) (hash(x)=6156715)
742
+ 710 train 7.646048 (lr=1.6770e-04) (hash(x)=6321494)
743
+ 711 train 7.687255 (lr=1.6694e-04) (hash(x)=7614023)
744
+ 712 train 7.692809 (lr=1.6619e-04) (hash(x)=6740380)
745
+ 713 train 7.522471 (lr=1.6543e-04) (hash(x)=4861744)
746
+ 714 train 7.374860 (lr=1.6468e-04) (hash(x)=6542179)
747
+ 715 train 7.430492 (lr=1.6393e-04) (hash(x)=5244861)
748
+ 716 train 7.552281 (lr=1.6317e-04) (hash(x)=7306636)
749
+ 717 train 7.379325 (lr=1.6242e-04) (hash(x)=7163697)
750
+ 718 train 7.581982 (lr=1.6167e-04) (hash(x)=6421642)
751
+ 719 train 7.362127 (lr=1.6092e-04) (hash(x)=5245146)
752
+ 720 train 7.442839 (lr=1.6018e-04) (hash(x)=6046027)
753
+ 721 train 7.225340 (lr=1.5943e-04) (hash(x)=6153866)
754
+ 722 train 7.422318 (lr=1.5868e-04) (hash(x)=5827481)
755
+ 723 train 7.333190 (lr=1.5794e-04) (hash(x)=6415565)
756
+ 724 train 7.498818 (lr=1.5720e-04) (hash(x)=6409570)
757
+ 725 train 7.945497 (lr=1.5646e-04) (hash(x)=7835853)
758
+ 726 train 7.059450 (lr=1.5572e-04) (hash(x)=4827589)
759
+ 727 train 7.184995 (lr=1.5498e-04) (hash(x)=5786972)
760
+ 728 train 7.603070 (lr=1.5424e-04) (hash(x)=6736612)
761
+ 729 train 7.464903 (lr=1.5350e-04) (hash(x)=6733560)
762
+ 730 train 7.637909 (lr=1.5277e-04) (hash(x)=4877208)
763
+ 731 train 7.113995 (lr=1.5203e-04) (hash(x)=6131703)
764
+ 732 train 7.215689 (lr=1.5130e-04) (hash(x)=6533769)
765
+ 733 train 7.212533 (lr=1.5057e-04) (hash(x)=6001331)
766
+ 734 train 7.784758 (lr=1.4984e-04) (hash(x)=10602643)
767
+ 735 train 7.712900 (lr=1.4911e-04) (hash(x)=6346459)
768
+ 736 train 7.368340 (lr=1.4838e-04) (hash(x)=6728215)
769
+ 737 train 7.657687 (lr=1.4766e-04) (hash(x)=8943770)
770
+ 738 train 7.495002 (lr=1.4693e-04) (hash(x)=7141912)
771
+ 739 train 7.455205 (lr=1.4621e-04) (hash(x)=6504131)
772
+ 740 train 7.310729 (lr=1.4549e-04) (hash(x)=6461667)
773
+ 741 train 7.536925 (lr=1.4477e-04) (hash(x)=5869339)
774
+ 742 train 7.941216 (lr=1.4405e-04) (hash(x)=7948065)
775
+ 743 train 7.168941 (lr=1.4333e-04) (hash(x)=5209234)
776
+ 744 train 7.300319 (lr=1.4262e-04) (hash(x)=6372244)
777
+ 745 train 7.478199 (lr=1.4190e-04) (hash(x)=7678937)
778
+ 746 train 7.382864 (lr=1.4119e-04) (hash(x)=6519438)
779
+ 747 train 7.456671 (lr=1.4048e-04) (hash(x)=6163272)
780
+ 748 train 7.490024 (lr=1.3977e-04) (hash(x)=7025209)
781
+ 749 train 7.333843 (lr=1.3906e-04) (hash(x)=5774172)
782
+ 750 val loss 7.3960
783
+ 750 val perplexity 1629.4705
784
+ 750 train 7.503047 (lr=1.3835e-04) (hash(x)=5327301)
785
+ 751 train 7.291592 (lr=1.3765e-04) (hash(x)=5676768)
786
+ 752 train 6.999124 (lr=1.3694e-04) (hash(x)=4823304)
787
+ 753 train 7.102332 (lr=1.3624e-04) (hash(x)=6443895)
788
+ 754 train 7.922614 (lr=1.3554e-04) (hash(x)=8032890)
789
+ 755 train 7.967155 (lr=1.3484e-04) (hash(x)=6090561)
790
+ 756 train 7.320749 (lr=1.3415e-04) (hash(x)=4514262)
791
+ 757 train 7.754841 (lr=1.3345e-04) (hash(x)=7788628)
792
+ 758 train 7.693462 (lr=1.3276e-04) (hash(x)=7239456)
793
+ 759 train 7.744527 (lr=1.3207e-04) (hash(x)=5863092)
794
+ 760 train 7.414846 (lr=1.3138e-04) (hash(x)=5572628)
795
+ 761 train 7.481321 (lr=1.3069e-04) (hash(x)=7141354)
796
+ 762 train 7.679090 (lr=1.3000e-04) (hash(x)=7272638)
797
+ 763 train 7.548131 (lr=1.2931e-04) (hash(x)=7201312)
798
+ 764 train 7.542058 (lr=1.2863e-04) (hash(x)=6015388)
799
+ 765 train 7.633627 (lr=1.2795e-04) (hash(x)=6414278)
800
+ 766 train 7.389520 (lr=1.2727e-04) (hash(x)=7137466)
801
+ 767 train 7.992478 (lr=1.2659e-04) (hash(x)=7019489)
802
+ 768 train 7.542756 (lr=1.2592e-04) (hash(x)=7233453)
803
+ 769 train 7.249399 (lr=1.2524e-04) (hash(x)=7914626)
804
+ 770 train 7.352845 (lr=1.2457e-04) (hash(x)=5764080)
805
+ 771 train 7.396888 (lr=1.2390e-04) (hash(x)=6225608)
806
+ 772 train 7.448322 (lr=1.2323e-04) (hash(x)=8097255)
807
+ 773 train 7.464288 (lr=1.2256e-04) (hash(x)=5998078)
808
+ 774 train 7.598824 (lr=1.2190e-04) (hash(x)=5416254)
809
+ 775 train 7.305884 (lr=1.2124e-04) (hash(x)=5483019)
810
+ 776 train 7.075982 (lr=1.2057e-04) (hash(x)=4702208)
811
+ 777 train 7.426601 (lr=1.1992e-04) (hash(x)=5911642)
812
+ 778 train 7.462656 (lr=1.1926e-04) (hash(x)=6132487)
813
+ 779 train 7.602303 (lr=1.1860e-04) (hash(x)=5903258)
814
+ 780 train 7.524452 (lr=1.1795e-04) (hash(x)=7915382)
815
+ 781 train 7.529359 (lr=1.1730e-04) (hash(x)=5632006)
816
+ 782 train 7.327503 (lr=1.1665e-04) (hash(x)=6518211)
817
+ 783 train 7.411739 (lr=1.1600e-04) (hash(x)=5968716)
818
+ 784 train 7.534944 (lr=1.1536e-04) (hash(x)=7344525)
819
+ 785 train 7.436255 (lr=1.1471e-04) (hash(x)=6401968)
820
+ 786 train 7.360393 (lr=1.1407e-04) (hash(x)=6276127)
821
+ 787 train 7.465710 (lr=1.1343e-04) (hash(x)=5778017)
822
+ 788 train 7.023957 (lr=1.1279e-04) (hash(x)=5387306)
823
+ 789 train 6.992826 (lr=1.1216e-04) (hash(x)=5772567)
824
+ 790 train 7.221425 (lr=1.1153e-04) (hash(x)=6383748)
825
+ 791 train 7.493791 (lr=1.1089e-04) (hash(x)=7780194)
826
+ 792 train 7.362870 (lr=1.1027e-04) (hash(x)=7119030)
827
+ 793 train 7.496387 (lr=1.0964e-04) (hash(x)=6424771)
828
+ 794 train 7.302902 (lr=1.0901e-04) (hash(x)=6540151)
829
+ 795 train 7.295705 (lr=1.0839e-04) (hash(x)=6140998)
830
+ 796 train 7.391833 (lr=1.0777e-04) (hash(x)=6208271)
831
+ 797 train 7.261333 (lr=1.0715e-04) (hash(x)=7859566)
832
+ 798 train 7.814253 (lr=1.0654e-04) (hash(x)=7064477)
833
+ 799 train 7.040195 (lr=1.0592e-04) (hash(x)=3784321)
834
+ 800 val loss 7.3842
835
+ 800 val perplexity 1610.3484
836
+ 800 train 7.389319 (lr=1.0531e-04) (hash(x)=4472758)
837
+ 801 train 7.322342 (lr=1.0470e-04) (hash(x)=5557891)
838
+ 802 train 7.455420 (lr=1.0410e-04) (hash(x)=7969325)
839
+ 803 train 7.474804 (lr=1.0349e-04) (hash(x)=5860821)
840
+ 804 train 7.479715 (lr=1.0289e-04) (hash(x)=6750848)
841
+ 805 train 7.432245 (lr=1.0229e-04) (hash(x)=5674826)
842
+ 806 train 7.225125 (lr=1.0169e-04) (hash(x)=5529163)
843
+ 807 train 7.610019 (lr=1.0109e-04) (hash(x)=7774109)
844
+ 808 train 7.402578 (lr=1.0050e-04) (hash(x)=6762509)
845
+ 809 train 7.511266 (lr=9.9910e-05) (hash(x)=5466449)
846
+ 810 train 7.366781 (lr=9.9321e-05) (hash(x)=7046935)
847
+ 811 train 8.244202 (lr=9.8735e-05) (hash(x)=7997664)
848
+ 812 train 7.527016 (lr=9.8151e-05) (hash(x)=6882594)
849
+ 813 train 7.705925 (lr=9.7569e-05) (hash(x)=7006517)
850
+ 814 train 7.564469 (lr=9.6990e-05) (hash(x)=6832530)
851
+ 815 train 7.388801 (lr=9.6413e-05) (hash(x)=6576749)
852
+ 816 train 7.497406 (lr=9.5838e-05) (hash(x)=8276629)
853
+ 817 train 7.316368 (lr=9.5266e-05) (hash(x)=6896198)
854
+ 818 train 7.508495 (lr=9.4696e-05) (hash(x)=5829252)
855
+ 819 train 7.657996 (lr=9.4129e-05) (hash(x)=7266655)
856
+ 820 train 7.383558 (lr=9.3564e-05) (hash(x)=6015975)
857
+ 821 train 7.364309 (lr=9.3001e-05) (hash(x)=5696696)
858
+ 822 train 7.410369 (lr=9.2441e-05) (hash(x)=5411666)
859
+ 823 train 7.629680 (lr=9.1884e-05) (hash(x)=7072404)
860
+ 824 train 7.315626 (lr=9.1328e-05) (hash(x)=4910095)
861
+ 825 train 7.424473 (lr=9.0776e-05) (hash(x)=6590657)
862
+ 826 train 7.353024 (lr=9.0226e-05) (hash(x)=7665574)
863
+ 827 train 7.405118 (lr=8.9678e-05) (hash(x)=6626459)
864
+ 828 train 7.428163 (lr=8.9133e-05) (hash(x)=6731971)
865
+ 829 train 7.423668 (lr=8.8591e-05) (hash(x)=5644977)
866
+ 830 train 7.382772 (lr=8.8051e-05) (hash(x)=6238654)
867
+ 831 train 7.264947 (lr=8.7513e-05) (hash(x)=6556025)
868
+ 832 train 7.580822 (lr=8.6978e-05) (hash(x)=7175563)
869
+ 833 train 7.459033 (lr=8.6446e-05) (hash(x)=6407492)
870
+ 834 train 7.285389 (lr=8.5916e-05) (hash(x)=4934335)
871
+ 835 train 7.300157 (lr=8.5389e-05) (hash(x)=5942129)
872
+ 836 train 7.337029 (lr=8.4865e-05) (hash(x)=5995643)
873
+ 837 train 7.404269 (lr=8.4343e-05) (hash(x)=6012163)
874
+ 838 train 7.545713 (lr=8.3824e-05) (hash(x)=8464831)
875
+ 839 train 7.774149 (lr=8.3307e-05) (hash(x)=7325027)
876
+ 840 train 7.636600 (lr=8.2793e-05) (hash(x)=6785865)
877
+ 841 train 7.550599 (lr=8.2282e-05) (hash(x)=4425520)
878
+ 842 train 7.334558 (lr=8.1773e-05) (hash(x)=5388267)
879
+ 843 train 7.595325 (lr=8.1267e-05) (hash(x)=7322467)
880
+ 844 train 7.362306 (lr=8.0764e-05) (hash(x)=6681766)
881
+ 845 train 7.458833 (lr=8.0263e-05) (hash(x)=7482800)
882
+ 846 train 7.374401 (lr=7.9765e-05) (hash(x)=5554493)
883
+ 847 train 7.599457 (lr=7.9270e-05) (hash(x)=6373412)
884
+ 848 train 7.021658 (lr=7.8778e-05) (hash(x)=5117517)
885
+ 849 train 7.258224 (lr=7.8288e-05) (hash(x)=6981426)
886
+ 850 val loss 7.3779
887
+ 850 val perplexity 1600.2367
888
+ 850 train 7.218916 (lr=7.7801e-05) (hash(x)=6886188)
889
+ 851 train 7.424624 (lr=7.7317e-05) (hash(x)=7332255)
890
+ 852 train 7.436696 (lr=7.6836e-05) (hash(x)=6172042)
891
+ 853 train 7.589216 (lr=7.6357e-05) (hash(x)=5930894)
892
+ 854 train 7.635510 (lr=7.5881e-05) (hash(x)=7448958)
893
+ 855 train 7.429552 (lr=7.5408e-05) (hash(x)=5262868)
894
+ 856 train 7.302075 (lr=7.4938e-05) (hash(x)=5558427)
895
+ 857 train 7.190957 (lr=7.4470e-05) (hash(x)=5585769)
896
+ 858 train 7.074392 (lr=7.4005e-05) (hash(x)=5838081)
897
+ 859 train 7.253884 (lr=7.3544e-05) (hash(x)=5688247)
898
+ 860 train 7.357229 (lr=7.3085e-05) (hash(x)=5162020)
899
+ 861 train 7.555761 (lr=7.2628e-05) (hash(x)=7462079)
900
+ 862 train 7.404220 (lr=7.2175e-05) (hash(x)=6516108)
901
+ 863 train 7.197952 (lr=7.1725e-05) (hash(x)=8055563)
902
+ 864 train 7.409320 (lr=7.1277e-05) (hash(x)=6271901)
903
+ 865 train 7.435908 (lr=7.0832e-05) (hash(x)=6221701)
904
+ 866 train 7.431785 (lr=7.0390e-05) (hash(x)=5772861)
905
+ 867 train 7.454789 (lr=6.9952e-05) (hash(x)=5352405)
906
+ 868 train 7.438743 (lr=6.9516e-05) (hash(x)=6111630)
907
+ 869 train 7.469532 (lr=6.9082e-05) (hash(x)=6730666)
908
+ 870 train 7.646123 (lr=6.8652e-05) (hash(x)=8698315)
909
+ 871 train 7.321314 (lr=6.8225e-05) (hash(x)=5932790)
910
+ 872 train 7.698706 (lr=6.7801e-05) (hash(x)=8099892)
911
+ 873 train 7.362211 (lr=6.7379e-05) (hash(x)=6223114)
912
+ 874 train 7.465238 (lr=6.6961e-05) (hash(x)=5798363)
913
+ 875 train 7.291901 (lr=6.6545e-05) (hash(x)=6249312)
914
+ 876 train 7.439408 (lr=6.6133e-05) (hash(x)=6929692)
915
+ 877 train 7.550559 (lr=6.5723e-05) (hash(x)=7242827)
916
+ 878 train 7.301716 (lr=6.5317e-05) (hash(x)=6332123)
917
+ 879 train 7.210484 (lr=6.4913e-05) (hash(x)=5680154)
918
+ 880 train 7.595014 (lr=6.4513e-05) (hash(x)=6352331)
919
+ 881 train 7.403541 (lr=6.4115e-05) (hash(x)=6332332)
920
+ 882 train 7.452321 (lr=6.3721e-05) (hash(x)=6180674)
921
+ 883 train 7.404085 (lr=6.3329e-05) (hash(x)=8156280)
922
+ 884 train 7.426672 (lr=6.2941e-05) (hash(x)=5874247)
923
+ 885 train 7.371298 (lr=6.2556e-05) (hash(x)=6659781)
924
+ 886 train 7.493792 (lr=6.2173e-05) (hash(x)=5780147)
925
+ 887 train 7.313876 (lr=6.1794e-05) (hash(x)=5914217)
926
+ 888 train 7.157481 (lr=6.1418e-05) (hash(x)=5411762)
927
+ 889 train 7.472383 (lr=6.1045e-05) (hash(x)=6714110)
928
+ 890 train 7.149372 (lr=6.0675e-05) (hash(x)=5999685)
929
+ 891 train 7.332294 (lr=6.0308e-05) (hash(x)=7120215)
930
+ 892 train 7.582332 (lr=5.9944e-05) (hash(x)=6546587)
931
+ 893 train 7.346202 (lr=5.9583e-05) (hash(x)=6593413)
932
+ 894 train 7.456257 (lr=5.9225e-05) (hash(x)=6369927)
933
+ 895 train 7.350650 (lr=5.8871e-05) (hash(x)=6424128)
934
+ 896 train 7.335809 (lr=5.8519e-05) (hash(x)=5736158)
935
+ 897 train 7.159677 (lr=5.8171e-05) (hash(x)=5903043)
936
+ 898 train 7.245536 (lr=5.7826e-05) (hash(x)=4419128)
937
+ 899 train 7.069800 (lr=5.7484e-05) (hash(x)=4390027)
938
+ 900 val loss 7.3551
939
+ 900 val perplexity 1564.1499
940
+ 900 train 7.308222 (lr=5.7145e-05) (hash(x)=6728135)
941
+ 901 train 7.435823 (lr=5.6809e-05) (hash(x)=6945760)
942
+ 902 train 7.396671 (lr=5.6476e-05) (hash(x)=6081534)
943
+ 903 train 7.469185 (lr=5.6147e-05) (hash(x)=7804089)
944
+ 904 train 7.526792 (lr=5.5821e-05) (hash(x)=6225832)
945
+ 905 train 7.565935 (lr=5.5497e-05) (hash(x)=6273417)
946
+ 906 train 7.607038 (lr=5.5178e-05) (hash(x)=7775633)
947
+ 907 train 7.564097 (lr=5.4861e-05) (hash(x)=7130267)
948
+ 908 train 7.345338 (lr=5.4547e-05) (hash(x)=6554076)
949
+ 909 train 7.456250 (lr=5.4237e-05) (hash(x)=6140697)
950
+ 910 train 7.271626 (lr=5.3930e-05) (hash(x)=6128181)
951
+ 911 train 7.385910 (lr=5.3626e-05) (hash(x)=6490149)
952
+ 912 train 6.928421 (lr=5.3325e-05) (hash(x)=5422426)
953
+ 913 train 7.370100 (lr=5.3028e-05) (hash(x)=4847733)
954
+ 914 train 7.373724 (lr=5.2734e-05) (hash(x)=6349747)
955
+ 915 train 7.668162 (lr=5.2443e-05) (hash(x)=9276378)
956
+ 916 train 7.275504 (lr=5.2155e-05) (hash(x)=6869524)
957
+ 917 train 7.300123 (lr=5.1871e-05) (hash(x)=7082053)
958
+ 918 train 7.355165 (lr=5.1589e-05) (hash(x)=7799351)
959
+ 919 train 7.519971 (lr=5.1311e-05) (hash(x)=6283540)
960
+ 920 train 7.304411 (lr=5.1037e-05) (hash(x)=6660860)
961
+ 921 train 7.242377 (lr=5.0765e-05) (hash(x)=5200956)
962
+ 922 train 7.342465 (lr=5.0497e-05) (hash(x)=5653241)
963
+ 923 train 7.494324 (lr=5.0232e-05) (hash(x)=5781680)
964
+ 924 train 7.461452 (lr=4.9971e-05) (hash(x)=6146666)
965
+ 925 train 7.314250 (lr=4.9712e-05) (hash(x)=8652907)
966
+ 926 train 7.114859 (lr=4.9457e-05) (hash(x)=5311546)
967
+ 927 train 7.249785 (lr=4.9206e-05) (hash(x)=5728337)
968
+ 928 train 7.229902 (lr=4.8957e-05) (hash(x)=5413240)
969
+ 929 train 7.349687 (lr=4.8712e-05) (hash(x)=5832401)
970
+ 930 train 7.109508 (lr=4.8470e-05) (hash(x)=7420230)
971
+ 931 train 7.447092 (lr=4.8232e-05) (hash(x)=6278202)
972
+ 932 train 7.297539 (lr=4.7997e-05) (hash(x)=6189873)
973
+ 933 train 7.422616 (lr=4.7765e-05) (hash(x)=7403868)
974
+ 934 train 7.268861 (lr=4.7537e-05) (hash(x)=6867459)
975
+ 935 train 7.485469 (lr=4.7312e-05) (hash(x)=7400272)
976
+ 936 train 7.508608 (lr=4.7090e-05) (hash(x)=7936275)
977
+ 937 train 7.229595 (lr=4.6871e-05) (hash(x)=5943413)
978
+ 938 train 7.445310 (lr=4.6656e-05) (hash(x)=5442166)
979
+ 939 train 7.338399 (lr=4.6445e-05) (hash(x)=5795128)
980
+ 940 train 7.523449 (lr=4.6236e-05) (hash(x)=6831680)
981
+ 941 train 7.247817 (lr=4.6031e-05) (hash(x)=5519980)
982
+ 942 train 7.243730 (lr=4.5830e-05) (hash(x)=6765461)
983
+ 943 train 7.199632 (lr=4.5631e-05) (hash(x)=6473432)
984
+ 944 train 7.178991 (lr=4.5437e-05) (hash(x)=6490395)
985
+ 945 train 7.411891 (lr=4.5245e-05) (hash(x)=5484669)
986
+ 946 train 7.390506 (lr=4.5057e-05) (hash(x)=7278970)
987
+ 947 train 7.377321 (lr=4.4872e-05) (hash(x)=5201502)
988
+ 948 train 7.628700 (lr=4.4691e-05) (hash(x)=7277915)
989
+ 949 train 7.284154 (lr=4.4513e-05) (hash(x)=7227057)
990
+ 950 val loss 7.3435
991
+ 950 val perplexity 1546.1652
992
+ 950 train 7.572911 (lr=4.4338e-05) (hash(x)=8178996)
993
+ 951 train 7.713424 (lr=4.4167e-05) (hash(x)=6922646)
994
+ 952 train 7.215410 (lr=4.4000e-05) (hash(x)=5941772)
995
+ 953 train 7.174014 (lr=4.3835e-05) (hash(x)=6056673)
996
+ 954 train 7.589973 (lr=4.3674e-05) (hash(x)=6912178)
997
+ 955 train 7.107974 (lr=4.3517e-05) (hash(x)=5484641)
998
+ 956 train 7.452599 (lr=4.3363e-05) (hash(x)=5121117)
999
+ 957 train 7.482172 (lr=4.3212e-05) (hash(x)=7100865)
1000
+ 958 train 7.160964 (lr=4.3065e-05) (hash(x)=5334280)
1001
+ 959 train 7.272405 (lr=4.2921e-05) (hash(x)=7277141)
1002
+ 960 train 7.261897 (lr=4.2781e-05) (hash(x)=6401960)
1003
+ 961 train 7.311950 (lr=4.2644e-05) (hash(x)=6407603)
1004
+ 962 train 7.450472 (lr=4.2510e-05) (hash(x)=5786089)
1005
+ 963 train 7.306540 (lr=4.2380e-05) (hash(x)=5604614)
1006
+ 964 train 7.339895 (lr=4.2253e-05) (hash(x)=5454287)
1007
+ 965 train 7.479698 (lr=4.2130e-05) (hash(x)=6200612)
1008
+ 966 train 7.418107 (lr=4.2010e-05) (hash(x)=5691759)
1009
+ 967 train 7.381812 (lr=4.1894e-05) (hash(x)=7443124)
1010
+ 968 train 7.400855 (lr=4.1781e-05) (hash(x)=6711366)
1011
+ 969 train 7.318673 (lr=4.1672e-05) (hash(x)=6728764)
1012
+ 970 train 7.627553 (lr=4.1566e-05) (hash(x)=6870240)
1013
+ 971 train 7.339466 (lr=4.1463e-05) (hash(x)=5601968)
1014
+ 972 train 7.080988 (lr=4.1364e-05) (hash(x)=5311504)
1015
+ 973 train 7.358827 (lr=4.1269e-05) (hash(x)=6877761)
1016
+ 974 train 7.431441 (lr=4.1177e-05) (hash(x)=6457033)
1017
+ 975 train 7.382578 (lr=4.1088e-05) (hash(x)=5632735)
1018
+ 976 train 7.576962 (lr=4.1003e-05) (hash(x)=6024880)
1019
+ 977 train 7.265551 (lr=4.0921e-05) (hash(x)=5658845)
1020
+ 978 train 7.156719 (lr=4.0843e-05) (hash(x)=6089461)
1021
+ 979 train 7.201392 (lr=4.0768e-05) (hash(x)=5594810)
1022
+ 980 train 7.007952 (lr=4.0697e-05) (hash(x)=4667026)
1023
+ 981 train 6.872203 (lr=4.0629e-05) (hash(x)=4738523)
1024
+ 982 train 7.193507 (lr=4.0564e-05) (hash(x)=7451876)
1025
+ 983 train 7.457832 (lr=4.0503e-05) (hash(x)=5881015)
1026
+ 984 train 7.443588 (lr=4.0446e-05) (hash(x)=6255031)
1027
+ 985 train 7.263505 (lr=4.0392e-05) (hash(x)=5263325)
1028
+ 986 train 7.258874 (lr=4.0341e-05) (hash(x)=6654585)
1029
+ 987 train 7.319571 (lr=4.0294e-05) (hash(x)=6865361)
1030
+ 988 train 7.416778 (lr=4.0251e-05) (hash(x)=6978986)
1031
+ 989 train 7.520934 (lr=4.0211e-05) (hash(x)=5644619)
1032
+ 990 train 7.341804 (lr=4.0174e-05) (hash(x)=6551023)
1033
+ 991 train 7.284855 (lr=4.0141e-05) (hash(x)=5809411)
1034
+ 992 train 7.375855 (lr=4.0112e-05) (hash(x)=5501837)
1035
+ 993 train 7.392744 (lr=4.0085e-05) (hash(x)=7044374)
1036
+ 994 train 7.355122 (lr=4.0063e-05) (hash(x)=6526130)
1037
+ 995 train 7.469051 (lr=4.0044e-05) (hash(x)=6293025)
1038
+ 996 train 7.465250 (lr=4.0028e-05) (hash(x)=6295768)
1039
+ 997 train 7.271845 (lr=4.0016e-05) (hash(x)=5189023)
1040
+ 998 train 7.086488 (lr=4.0007e-05) (hash(x)=6255467)
1041
+ 999 val loss 7.3384
1042
+ 999 val perplexity 1538.2334
1043
+ 999 train 7.425029 (lr=4.0002e-05) (hash(x)=7232964)
lr4e-4_total_batch_size5120_seq_len128/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e4c126dd7306ee2dfcdf417eb536f2a0053309bf30ac57039c4e3f17b6deefc5
3
+ size 36163010
lr4e-4_total_batch_size5120_seq_len128/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c4d2d1dd845e521a51369f102393d366d5e09dfbc2ef525351c681af2a3e272
3
+ size 70764358