andrew-healey commited on
Commit
89f9b7f
·
verified ·
1 Parent(s): aac5ba5

Upload folder using huggingface_hub

Browse files
lr1e-4_total_batch_size5120_seq_len64/args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hellaswag": true, "attention_kind": "selective", "log_dir": "proxy_model_sweep/lr1e-4_total_batch_size5120_seq_len64", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": null, "group": "proxy_model_sweep", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1337, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 64, "batch_size": null, "total_batch_size": 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": 0.0001, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "n_embd": 128}
lr1e-4_total_batch_size5120_seq_len64/dataloader_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60582085b67f79a59d68bfd29cd9c94a65421fc83fc6750cd587ed93c2812fa4
3
+ size 964
lr1e-4_total_batch_size5120_seq_len64/log2.txt ADDED
@@ -0,0 +1,1043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ max_steps: 1000
2
+ 0 val loss 11.7175
3
+ 0 val perplexity 122699.7891
4
+ 0 train 11.749433 (lr=3.4965e-07) (hash(x)=3451208)
5
+ 1 train 11.703221 (lr=6.9930e-07) (hash(x)=2985804)
6
+ 2 train 11.790811 (lr=1.0490e-06) (hash(x)=3025796)
7
+ 3 train 11.765208 (lr=1.3986e-06) (hash(x)=3135520)
8
+ 4 train 11.734349 (lr=1.7483e-06) (hash(x)=3142430)
9
+ 5 train 11.739650 (lr=2.0979e-06) (hash(x)=3699147)
10
+ 6 train 11.717319 (lr=2.4476e-06) (hash(x)=3996635)
11
+ 7 train 11.718472 (lr=2.7972e-06) (hash(x)=3276965)
12
+ 8 train 11.684131 (lr=3.1469e-06) (hash(x)=3385917)
13
+ 9 train 11.700359 (lr=3.4965e-06) (hash(x)=2444616)
14
+ 10 train 11.675490 (lr=3.8462e-06) (hash(x)=3699470)
15
+ 11 train 11.671686 (lr=4.1958e-06) (hash(x)=3181634)
16
+ 12 train 11.667338 (lr=4.5455e-06) (hash(x)=3659236)
17
+ 13 train 11.606978 (lr=4.8951e-06) (hash(x)=3550163)
18
+ 14 train 11.640166 (lr=5.2448e-06) (hash(x)=2382853)
19
+ 15 train 11.591650 (lr=5.5944e-06) (hash(x)=2891006)
20
+ 16 train 11.581844 (lr=5.9441e-06) (hash(x)=2756611)
21
+ 17 train 11.566497 (lr=6.2937e-06) (hash(x)=3381491)
22
+ 18 train 11.524179 (lr=6.6434e-06) (hash(x)=3241824)
23
+ 19 train 11.530416 (lr=6.9930e-06) (hash(x)=3846008)
24
+ 20 train 11.493472 (lr=7.3427e-06) (hash(x)=2544106)
25
+ 21 train 11.511292 (lr=7.6923e-06) (hash(x)=3596625)
26
+ 22 train 11.466618 (lr=8.0420e-06) (hash(x)=2587727)
27
+ 23 train 11.424956 (lr=8.3916e-06) (hash(x)=2536675)
28
+ 24 train 11.437775 (lr=8.7413e-06) (hash(x)=3510943)
29
+ 25 train 11.350089 (lr=9.0909e-06) (hash(x)=4321027)
30
+ 26 train 11.319960 (lr=9.4406e-06) (hash(x)=2266940)
31
+ 27 train 11.287392 (lr=9.7902e-06) (hash(x)=2461698)
32
+ 28 train 11.227976 (lr=1.0140e-05) (hash(x)=2931775)
33
+ 29 train 11.271896 (lr=1.0490e-05) (hash(x)=3818493)
34
+ 30 train 11.175318 (lr=1.0839e-05) (hash(x)=2875034)
35
+ 31 train 11.039930 (lr=1.1189e-05) (hash(x)=2954056)
36
+ 32 train 11.141338 (lr=1.1538e-05) (hash(x)=3859846)
37
+ 33 train 11.110346 (lr=1.1888e-05) (hash(x)=3663107)
38
+ 34 train 11.000410 (lr=1.2238e-05) (hash(x)=2312499)
39
+ 35 train 11.050831 (lr=1.2587e-05) (hash(x)=2876503)
40
+ 36 train 11.090109 (lr=1.2937e-05) (hash(x)=2395056)
41
+ 37 train 11.038979 (lr=1.3287e-05) (hash(x)=3158454)
42
+ 38 train 10.920173 (lr=1.3636e-05) (hash(x)=2748606)
43
+ 39 train 10.869217 (lr=1.3986e-05) (hash(x)=2652726)
44
+ 40 train 10.868067 (lr=1.4336e-05) (hash(x)=2758346)
45
+ 41 train 10.762947 (lr=1.4685e-05) (hash(x)=2766194)
46
+ 42 train 10.783828 (lr=1.5035e-05) (hash(x)=4269053)
47
+ 43 train 10.733423 (lr=1.5385e-05) (hash(x)=3865564)
48
+ 44 train 10.644015 (lr=1.5734e-05) (hash(x)=4061515)
49
+ 45 train 10.693297 (lr=1.6084e-05) (hash(x)=3535906)
50
+ 46 train 10.808512 (lr=1.6434e-05) (hash(x)=2220303)
51
+ 47 train 10.635164 (lr=1.6783e-05) (hash(x)=3142323)
52
+ 48 train 10.594684 (lr=1.7133e-05) (hash(x)=2783022)
53
+ 49 train 10.644900 (lr=1.7483e-05) (hash(x)=2747058)
54
+ 50 val loss 10.6209
55
+ 50 val perplexity 40984.0508
56
+ 50 train 10.625107 (lr=1.7832e-05) (hash(x)=2801277)
57
+ 51 train 10.666308 (lr=1.8182e-05) (hash(x)=3057360)
58
+ 52 train 10.526087 (lr=1.8531e-05) (hash(x)=2888775)
59
+ 53 train 10.493000 (lr=1.8881e-05) (hash(x)=2669357)
60
+ 54 train 10.477505 (lr=1.9231e-05) (hash(x)=3679065)
61
+ 55 train 10.344665 (lr=1.9580e-05) (hash(x)=3004989)
62
+ 56 train 10.508859 (lr=1.9930e-05) (hash(x)=2626297)
63
+ 57 train 10.416359 (lr=2.0280e-05) (hash(x)=2440498)
64
+ 58 train 10.450983 (lr=2.0629e-05) (hash(x)=3445074)
65
+ 59 train 10.576163 (lr=2.0979e-05) (hash(x)=2323423)
66
+ 60 train 10.437980 (lr=2.1329e-05) (hash(x)=3274625)
67
+ 61 train 10.424321 (lr=2.1678e-05) (hash(x)=3100868)
68
+ 62 train 10.366983 (lr=2.2028e-05) (hash(x)=3424878)
69
+ 63 train 10.193438 (lr=2.2378e-05) (hash(x)=3207764)
70
+ 64 train 10.174148 (lr=2.2727e-05) (hash(x)=2793064)
71
+ 65 train 10.341216 (lr=2.3077e-05) (hash(x)=3271229)
72
+ 66 train 10.380327 (lr=2.3427e-05) (hash(x)=2868232)
73
+ 67 train 10.276700 (lr=2.3776e-05) (hash(x)=3149634)
74
+ 68 train 10.146085 (lr=2.4126e-05) (hash(x)=2563226)
75
+ 69 train 10.274535 (lr=2.4476e-05) (hash(x)=3774750)
76
+ 70 train 10.281669 (lr=2.4825e-05) (hash(x)=3152496)
77
+ 71 train 10.314386 (lr=2.5175e-05) (hash(x)=4140327)
78
+ 72 train 10.274370 (lr=2.5524e-05) (hash(x)=3543858)
79
+ 73 train 10.173452 (lr=2.5874e-05) (hash(x)=3438929)
80
+ 74 train 10.201087 (lr=2.6224e-05) (hash(x)=3570542)
81
+ 75 train 10.292422 (lr=2.6573e-05) (hash(x)=3390476)
82
+ 76 train 10.201995 (lr=2.6923e-05) (hash(x)=3201153)
83
+ 77 train 10.056100 (lr=2.7273e-05) (hash(x)=3939431)
84
+ 78 train 10.229388 (lr=2.7622e-05) (hash(x)=3471234)
85
+ 79 train 10.147619 (lr=2.7972e-05) (hash(x)=3100847)
86
+ 80 train 10.267204 (lr=2.8322e-05) (hash(x)=3077381)
87
+ 81 train 10.114554 (lr=2.8671e-05) (hash(x)=2942266)
88
+ 82 train 10.014654 (lr=2.9021e-05) (hash(x)=2790512)
89
+ 83 train 10.062372 (lr=2.9371e-05) (hash(x)=3618428)
90
+ 84 train 10.109626 (lr=2.9720e-05) (hash(x)=3342819)
91
+ 85 train 10.111317 (lr=3.0070e-05) (hash(x)=3598964)
92
+ 86 train 10.055500 (lr=3.0420e-05) (hash(x)=2975493)
93
+ 87 train 10.152554 (lr=3.0769e-05) (hash(x)=4449654)
94
+ 88 train 10.117231 (lr=3.1119e-05) (hash(x)=3484308)
95
+ 89 train 10.106617 (lr=3.1469e-05) (hash(x)=2906105)
96
+ 90 train 9.900798 (lr=3.1818e-05) (hash(x)=3548351)
97
+ 91 train 9.858378 (lr=3.2168e-05) (hash(x)=3339267)
98
+ 92 train 9.942442 (lr=3.2517e-05) (hash(x)=4244318)
99
+ 93 train 9.988443 (lr=3.2867e-05) (hash(x)=3382988)
100
+ 94 train 10.244148 (lr=3.3217e-05) (hash(x)=3448763)
101
+ 95 train 9.946208 (lr=3.3566e-05) (hash(x)=2858681)
102
+ 96 train 10.211857 (lr=3.3916e-05) (hash(x)=3844384)
103
+ 97 train 10.193048 (lr=3.4266e-05) (hash(x)=2894041)
104
+ 98 train 9.945816 (lr=3.4615e-05) (hash(x)=2980750)
105
+ 99 train 9.890718 (lr=3.4965e-05) (hash(x)=2218193)
106
+ 100 val loss 9.9941
107
+ 100 val perplexity 21897.5723
108
+ 100 train 10.015549 (lr=3.5315e-05) (hash(x)=3321779)
109
+ 101 train 9.978595 (lr=3.5664e-05) (hash(x)=2522274)
110
+ 102 train 9.876815 (lr=3.6014e-05) (hash(x)=3079028)
111
+ 103 train 9.993232 (lr=3.6364e-05) (hash(x)=3195898)
112
+ 104 train 9.916461 (lr=3.6713e-05) (hash(x)=3711694)
113
+ 105 train 9.946428 (lr=3.7063e-05) (hash(x)=2899839)
114
+ 106 train 9.690989 (lr=3.7413e-05) (hash(x)=2676240)
115
+ 107 train 9.803503 (lr=3.7762e-05) (hash(x)=2500264)
116
+ 108 train 9.816823 (lr=3.8112e-05) (hash(x)=3183657)
117
+ 109 train 9.815461 (lr=3.8462e-05) (hash(x)=3668948)
118
+ 110 train 10.440611 (lr=3.8811e-05) (hash(x)=3537635)
119
+ 111 train 9.794337 (lr=3.9161e-05) (hash(x)=3236150)
120
+ 112 train 9.725899 (lr=3.9510e-05) (hash(x)=2951937)
121
+ 113 train 9.844172 (lr=3.9860e-05) (hash(x)=3586288)
122
+ 114 train 9.881594 (lr=4.0210e-05) (hash(x)=3215993)
123
+ 115 train 9.825887 (lr=4.0559e-05) (hash(x)=3337040)
124
+ 116 train 9.784458 (lr=4.0909e-05) (hash(x)=5246494)
125
+ 117 train 9.877434 (lr=4.1259e-05) (hash(x)=4003369)
126
+ 118 train 9.871281 (lr=4.1608e-05) (hash(x)=2658596)
127
+ 119 train 9.656714 (lr=4.1958e-05) (hash(x)=3408369)
128
+ 120 train 9.766912 (lr=4.2308e-05) (hash(x)=3005779)
129
+ 121 train 9.579286 (lr=4.2657e-05) (hash(x)=3559111)
130
+ 122 train 9.677951 (lr=4.3007e-05) (hash(x)=3501124)
131
+ 123 train 9.585670 (lr=4.3357e-05) (hash(x)=2787317)
132
+ 124 train 9.531790 (lr=4.3706e-05) (hash(x)=3105990)
133
+ 125 train 9.526091 (lr=4.4056e-05) (hash(x)=2957029)
134
+ 126 train 9.688815 (lr=4.4406e-05) (hash(x)=3846628)
135
+ 127 train 9.493155 (lr=4.4755e-05) (hash(x)=3012068)
136
+ 128 train 9.619542 (lr=4.5105e-05) (hash(x)=3399330)
137
+ 129 train 9.602374 (lr=4.5455e-05) (hash(x)=2792231)
138
+ 130 train 9.535743 (lr=4.5804e-05) (hash(x)=3436915)
139
+ 131 train 9.665660 (lr=4.6154e-05) (hash(x)=2240543)
140
+ 132 train 9.817401 (lr=4.6503e-05) (hash(x)=3377630)
141
+ 133 train 9.433772 (lr=4.6853e-05) (hash(x)=2586857)
142
+ 134 train 9.287163 (lr=4.7203e-05) (hash(x)=3807526)
143
+ 135 train 9.369850 (lr=4.7552e-05) (hash(x)=3491889)
144
+ 136 train 9.582163 (lr=4.7902e-05) (hash(x)=4842167)
145
+ 137 train 9.561415 (lr=4.8252e-05) (hash(x)=3727327)
146
+ 138 train 9.387117 (lr=4.8601e-05) (hash(x)=3043815)
147
+ 139 train 9.204242 (lr=4.8951e-05) (hash(x)=2958365)
148
+ 140 train 9.509283 (lr=4.9301e-05) (hash(x)=2987254)
149
+ 141 train 9.547733 (lr=4.9650e-05) (hash(x)=3563076)
150
+ 142 train 9.729171 (lr=5.0000e-05) (hash(x)=3987128)
151
+ 143 train 9.339495 (lr=5.0350e-05) (hash(x)=3437709)
152
+ 144 train 9.396897 (lr=5.0699e-05) (hash(x)=2553092)
153
+ 145 train 9.374376 (lr=5.1049e-05) (hash(x)=2979567)
154
+ 146 train 9.369103 (lr=5.1399e-05) (hash(x)=3005413)
155
+ 147 train 9.509384 (lr=5.1748e-05) (hash(x)=4237046)
156
+ 148 train 9.342746 (lr=5.2098e-05) (hash(x)=3386282)
157
+ 149 train 9.208300 (lr=5.2448e-05) (hash(x)=1995144)
158
+ 150 val loss 9.2533
159
+ 150 val perplexity 10438.8848
160
+ 150 train 9.191036 (lr=5.2797e-05) (hash(x)=3036507)
161
+ 151 train 10.151500 (lr=5.3147e-05) (hash(x)=5321047)
162
+ 152 train 9.692801 (lr=5.3497e-05) (hash(x)=3401355)
163
+ 153 train 9.073463 (lr=5.3846e-05) (hash(x)=3052730)
164
+ 154 train 9.030754 (lr=5.4196e-05) (hash(x)=2582624)
165
+ 155 train 9.417466 (lr=5.4545e-05) (hash(x)=3651404)
166
+ 156 train 9.137393 (lr=5.4895e-05) (hash(x)=3123258)
167
+ 157 train 9.132973 (lr=5.5245e-05) (hash(x)=3313001)
168
+ 158 train 9.186816 (lr=5.5594e-05) (hash(x)=3098619)
169
+ 159 train 9.123348 (lr=5.5944e-05) (hash(x)=3049379)
170
+ 160 train 8.923647 (lr=5.6294e-05) (hash(x)=2847337)
171
+ 161 train 8.956984 (lr=5.6643e-05) (hash(x)=2677265)
172
+ 162 train 8.971138 (lr=5.6993e-05) (hash(x)=3113508)
173
+ 163 train 9.114452 (lr=5.7343e-05) (hash(x)=2128348)
174
+ 164 train 9.193283 (lr=5.7692e-05) (hash(x)=3444033)
175
+ 165 train 8.974720 (lr=5.8042e-05) (hash(x)=4105439)
176
+ 166 train 8.926306 (lr=5.8392e-05) (hash(x)=3038099)
177
+ 167 train 8.941767 (lr=5.8741e-05) (hash(x)=3806489)
178
+ 168 train 8.663692 (lr=5.9091e-05) (hash(x)=2693823)
179
+ 169 train 8.832467 (lr=5.9441e-05) (hash(x)=2852625)
180
+ 170 train 8.769349 (lr=5.9790e-05) (hash(x)=3098982)
181
+ 171 train 8.875973 (lr=6.0140e-05) (hash(x)=3201546)
182
+ 172 train 8.923967 (lr=6.0490e-05) (hash(x)=3116229)
183
+ 173 train 8.725583 (lr=6.0839e-05) (hash(x)=3790597)
184
+ 174 train 8.912278 (lr=6.1189e-05) (hash(x)=2632089)
185
+ 175 train 8.697653 (lr=6.1538e-05) (hash(x)=3147508)
186
+ 176 train 8.907799 (lr=6.1888e-05) (hash(x)=2893416)
187
+ 177 train 8.652203 (lr=6.2238e-05) (hash(x)=3211389)
188
+ 178 train 8.642109 (lr=6.2587e-05) (hash(x)=3945225)
189
+ 179 train 8.648934 (lr=6.2937e-05) (hash(x)=2925153)
190
+ 180 train 8.666121 (lr=6.3287e-05) (hash(x)=2522561)
191
+ 181 train 8.511979 (lr=6.3636e-05) (hash(x)=2975211)
192
+ 182 train 8.641434 (lr=6.3986e-05) (hash(x)=2573158)
193
+ 183 train 8.700521 (lr=6.4336e-05) (hash(x)=4028685)
194
+ 184 train 8.690080 (lr=6.4685e-05) (hash(x)=3005976)
195
+ 185 train 8.456658 (lr=6.5035e-05) (hash(x)=3054572)
196
+ 186 train 8.574030 (lr=6.5385e-05) (hash(x)=2866287)
197
+ 187 train 8.818626 (lr=6.5734e-05) (hash(x)=3855998)
198
+ 188 train 8.488988 (lr=6.6084e-05) (hash(x)=3098080)
199
+ 189 train 8.406698 (lr=6.6434e-05) (hash(x)=2582342)
200
+ 190 train 8.643781 (lr=6.6783e-05) (hash(x)=2446127)
201
+ 191 train 8.369967 (lr=6.7133e-05) (hash(x)=3185893)
202
+ 192 train 8.543923 (lr=6.7483e-05) (hash(x)=2769486)
203
+ 193 train 8.429652 (lr=6.7832e-05) (hash(x)=3561988)
204
+ 194 train 8.545141 (lr=6.8182e-05) (hash(x)=3179595)
205
+ 195 train 8.726631 (lr=6.8531e-05) (hash(x)=3177996)
206
+ 196 train 8.413643 (lr=6.8881e-05) (hash(x)=3136605)
207
+ 197 train 9.022807 (lr=6.9231e-05) (hash(x)=3672699)
208
+ 198 train 8.904700 (lr=6.9580e-05) (hash(x)=3047766)
209
+ 199 train 8.469955 (lr=6.9930e-05) (hash(x)=3683623)
210
+ 200 val loss 8.4466
211
+ 200 val perplexity 4659.3511
212
+ 200 train 8.565431 (lr=7.0280e-05) (hash(x)=3414762)
213
+ 201 train 8.488630 (lr=7.0629e-05) (hash(x)=3133274)
214
+ 202 train 8.278190 (lr=7.0979e-05) (hash(x)=3032641)
215
+ 203 train 8.443985 (lr=7.1329e-05) (hash(x)=4336041)
216
+ 204 train 8.316755 (lr=7.1678e-05) (hash(x)=3649067)
217
+ 205 train 8.468173 (lr=7.2028e-05) (hash(x)=4242233)
218
+ 206 train 8.611753 (lr=7.2378e-05) (hash(x)=2840620)
219
+ 207 train 8.257256 (lr=7.2727e-05) (hash(x)=2323202)
220
+ 208 train 8.375405 (lr=7.3077e-05) (hash(x)=3851549)
221
+ 209 train 8.198908 (lr=7.3427e-05) (hash(x)=3169273)
222
+ 210 train 8.268880 (lr=7.3776e-05) (hash(x)=3689246)
223
+ 211 train 8.149456 (lr=7.4126e-05) (hash(x)=2378162)
224
+ 212 train 7.899050 (lr=7.4476e-05) (hash(x)=2813214)
225
+ 213 train 8.458991 (lr=7.4825e-05) (hash(x)=2580783)
226
+ 214 train 8.192712 (lr=7.5175e-05) (hash(x)=3574963)
227
+ 215 train 8.211306 (lr=7.5524e-05) (hash(x)=2761628)
228
+ 216 train 8.315450 (lr=7.5874e-05) (hash(x)=2160941)
229
+ 217 train 8.173099 (lr=7.6224e-05) (hash(x)=2641942)
230
+ 218 train 8.272324 (lr=7.6573e-05) (hash(x)=2912450)
231
+ 219 train 8.267923 (lr=7.6923e-05) (hash(x)=3761105)
232
+ 220 train 8.147377 (lr=7.7273e-05) (hash(x)=2936363)
233
+ 221 train 8.320019 (lr=7.7622e-05) (hash(x)=2658396)
234
+ 222 train 8.288508 (lr=7.7972e-05) (hash(x)=2438865)
235
+ 223 train 8.024476 (lr=7.8322e-05) (hash(x)=3434176)
236
+ 224 train 7.967316 (lr=7.8671e-05) (hash(x)=3423982)
237
+ 225 train 8.030337 (lr=7.9021e-05) (hash(x)=2646319)
238
+ 226 train 8.197598 (lr=7.9371e-05) (hash(x)=3228229)
239
+ 227 train 8.197890 (lr=7.9720e-05) (hash(x)=2757880)
240
+ 228 train 7.956522 (lr=8.0070e-05) (hash(x)=2715183)
241
+ 229 train 8.016181 (lr=8.0420e-05) (hash(x)=3401657)
242
+ 230 train 8.074086 (lr=8.0769e-05) (hash(x)=2436276)
243
+ 231 train 7.830480 (lr=8.1119e-05) (hash(x)=2554842)
244
+ 232 train 7.875931 (lr=8.1469e-05) (hash(x)=2845436)
245
+ 233 train 7.447624 (lr=8.1818e-05) (hash(x)=2392937)
246
+ 234 train 7.709443 (lr=8.2168e-05) (hash(x)=2667747)
247
+ 235 train 7.850690 (lr=8.2517e-05) (hash(x)=2568457)
248
+ 236 train 7.905157 (lr=8.2867e-05) (hash(x)=3272754)
249
+ 237 train 8.038877 (lr=8.3217e-05) (hash(x)=3599849)
250
+ 238 train 7.935855 (lr=8.3566e-05) (hash(x)=2800155)
251
+ 239 train 7.769154 (lr=8.3916e-05) (hash(x)=3411163)
252
+ 240 train 8.139616 (lr=8.4266e-05) (hash(x)=3573261)
253
+ 241 train 8.038181 (lr=8.4615e-05) (hash(x)=3233934)
254
+ 242 train 7.987463 (lr=8.4965e-05) (hash(x)=3644510)
255
+ 243 train 8.052616 (lr=8.5315e-05) (hash(x)=3663117)
256
+ 244 train 8.165498 (lr=8.5664e-05) (hash(x)=2862470)
257
+ 245 train 7.963323 (lr=8.6014e-05) (hash(x)=3400210)
258
+ 246 train 7.836923 (lr=8.6364e-05) (hash(x)=3538858)
259
+ 247 train 8.195297 (lr=8.6713e-05) (hash(x)=3803878)
260
+ 248 train 7.961594 (lr=8.7063e-05) (hash(x)=2939652)
261
+ 249 train 8.123166 (lr=8.7413e-05) (hash(x)=3784195)
262
+ 250 val loss 7.9152
263
+ 250 val perplexity 2738.6445
264
+ 250 train 8.137356 (lr=8.7762e-05) (hash(x)=2352377)
265
+ 251 train 7.910754 (lr=8.8112e-05) (hash(x)=2647947)
266
+ 252 train 7.875915 (lr=8.8462e-05) (hash(x)=2382162)
267
+ 253 train 7.833811 (lr=8.8811e-05) (hash(x)=2832090)
268
+ 254 train 7.895080 (lr=8.9161e-05) (hash(x)=2760950)
269
+ 255 train 7.885769 (lr=8.9510e-05) (hash(x)=3427494)
270
+ 256 train 8.066689 (lr=8.9860e-05) (hash(x)=3318366)
271
+ 257 train 7.810338 (lr=9.0210e-05) (hash(x)=3833844)
272
+ 258 train 8.025464 (lr=9.0559e-05) (hash(x)=2761594)
273
+ 259 train 7.848320 (lr=9.0909e-05) (hash(x)=2930772)
274
+ 260 train 8.117914 (lr=9.1259e-05) (hash(x)=3191402)
275
+ 261 train 8.013896 (lr=9.1608e-05) (hash(x)=3970232)
276
+ 262 train 8.224127 (lr=9.1958e-05) (hash(x)=2311991)
277
+ 263 train 7.766282 (lr=9.2308e-05) (hash(x)=2811588)
278
+ 264 train 7.870401 (lr=9.2657e-05) (hash(x)=3553207)
279
+ 265 train 7.848941 (lr=9.3007e-05) (hash(x)=2993366)
280
+ 266 train 7.780642 (lr=9.3357e-05) (hash(x)=2874416)
281
+ 267 train 7.753829 (lr=9.3706e-05) (hash(x)=4308511)
282
+ 268 train 7.992799 (lr=9.4056e-05) (hash(x)=3177305)
283
+ 269 train 7.641156 (lr=9.4406e-05) (hash(x)=3864907)
284
+ 270 train 7.697446 (lr=9.4755e-05) (hash(x)=3712643)
285
+ 271 train 7.624561 (lr=9.5105e-05) (hash(x)=3222913)
286
+ 272 train 7.791805 (lr=9.5455e-05) (hash(x)=2514559)
287
+ 273 train 7.856152 (lr=9.5804e-05) (hash(x)=2537083)
288
+ 274 train 8.048437 (lr=9.6154e-05) (hash(x)=3753557)
289
+ 275 train 8.151173 (lr=9.6503e-05) (hash(x)=2888911)
290
+ 276 train 7.890789 (lr=9.6853e-05) (hash(x)=4268306)
291
+ 277 train 8.176405 (lr=9.7203e-05) (hash(x)=4193793)
292
+ 278 train 8.054844 (lr=9.7552e-05) (hash(x)=3157715)
293
+ 279 train 7.878359 (lr=9.7902e-05) (hash(x)=3510963)
294
+ 280 train 7.859636 (lr=9.8252e-05) (hash(x)=3448175)
295
+ 281 train 7.559700 (lr=9.8601e-05) (hash(x)=3664282)
296
+ 282 train 7.760493 (lr=9.8951e-05) (hash(x)=2565399)
297
+ 283 train 7.985130 (lr=9.9301e-05) (hash(x)=3639400)
298
+ 284 train 7.996665 (lr=9.9650e-05) (hash(x)=3977218)
299
+ 285 train 8.130413 (lr=1.0000e-04) (hash(x)=3486873)
300
+ 286 train 7.626995 (lr=1.0000e-04) (hash(x)=2319934)
301
+ 287 train 7.637301 (lr=1.0000e-04) (hash(x)=3226934)
302
+ 288 train 7.608849 (lr=9.9998e-05) (hash(x)=3717947)
303
+ 289 train 7.722213 (lr=9.9996e-05) (hash(x)=3799900)
304
+ 290 train 7.524928 (lr=9.9993e-05) (hash(x)=2168504)
305
+ 291 train 7.829865 (lr=9.9989e-05) (hash(x)=2881290)
306
+ 292 train 7.857509 (lr=9.9984e-05) (hash(x)=3150134)
307
+ 293 train 8.012663 (lr=9.9979e-05) (hash(x)=3382829)
308
+ 294 train 7.645284 (lr=9.9972e-05) (hash(x)=3408834)
309
+ 295 train 7.626407 (lr=9.9965e-05) (hash(x)=2870627)
310
+ 296 train 7.827587 (lr=9.9956e-05) (hash(x)=2715310)
311
+ 297 train 7.579937 (lr=9.9947e-05) (hash(x)=2383672)
312
+ 298 train 7.634856 (lr=9.9937e-05) (hash(x)=2550881)
313
+ 299 train 7.813010 (lr=9.9926e-05) (hash(x)=3237167)
314
+ 300 val loss 7.7709
315
+ 300 val perplexity 2370.4971
316
+ 300 train 8.574001 (lr=9.9915e-05) (hash(x)=4155088)
317
+ 301 train 8.384887 (lr=9.9902e-05) (hash(x)=3865053)
318
+ 302 train 8.073510 (lr=9.9889e-05) (hash(x)=1423062)
319
+ 303 train 7.378293 (lr=9.9874e-05) (hash(x)=3526556)
320
+ 304 train 7.952139 (lr=9.9859e-05) (hash(x)=3714130)
321
+ 305 train 7.779804 (lr=9.9843e-05) (hash(x)=3179225)
322
+ 306 train 7.895935 (lr=9.9826e-05) (hash(x)=2472506)
323
+ 307 train 7.881113 (lr=9.9808e-05) (hash(x)=3497999)
324
+ 308 train 7.991903 (lr=9.9789e-05) (hash(x)=3920000)
325
+ 309 train 8.041852 (lr=9.9770e-05) (hash(x)=3878435)
326
+ 310 train 8.154178 (lr=9.9749e-05) (hash(x)=3632447)
327
+ 311 train 7.974840 (lr=9.9728e-05) (hash(x)=3153837)
328
+ 312 train 7.741848 (lr=9.9706e-05) (hash(x)=2535819)
329
+ 313 train 7.960308 (lr=9.9683e-05) (hash(x)=3931422)
330
+ 314 train 7.761275 (lr=9.9659e-05) (hash(x)=2914933)
331
+ 315 train 7.716047 (lr=9.9634e-05) (hash(x)=2952505)
332
+ 316 train 7.749990 (lr=9.9609e-05) (hash(x)=3618145)
333
+ 317 train 7.887024 (lr=9.9582e-05) (hash(x)=2470884)
334
+ 318 train 7.571192 (lr=9.9555e-05) (hash(x)=2883937)
335
+ 319 train 7.750555 (lr=9.9526e-05) (hash(x)=2653794)
336
+ 320 train 7.754069 (lr=9.9497e-05) (hash(x)=3408195)
337
+ 321 train 7.790888 (lr=9.9467e-05) (hash(x)=3078133)
338
+ 322 train 8.026183 (lr=9.9437e-05) (hash(x)=3566093)
339
+ 323 train 7.957152 (lr=9.9405e-05) (hash(x)=3245055)
340
+ 324 train 7.686990 (lr=9.9372e-05) (hash(x)=2654279)
341
+ 325 train 7.998524 (lr=9.9339e-05) (hash(x)=2503882)
342
+ 326 train 7.777662 (lr=9.9305e-05) (hash(x)=2220025)
343
+ 327 train 7.406109 (lr=9.9270e-05) (hash(x)=2624914)
344
+ 328 train 7.457938 (lr=9.9234e-05) (hash(x)=2841973)
345
+ 329 train 7.395285 (lr=9.9197e-05) (hash(x)=2603463)
346
+ 330 train 7.467191 (lr=9.9159e-05) (hash(x)=2554400)
347
+ 331 train 7.358995 (lr=9.9121e-05) (hash(x)=2647960)
348
+ 332 train 8.072198 (lr=9.9081e-05) (hash(x)=5471918)
349
+ 333 train 7.830614 (lr=9.9041e-05) (hash(x)=2179553)
350
+ 334 train 7.684057 (lr=9.9000e-05) (hash(x)=2729922)
351
+ 335 train 7.752392 (lr=9.8958e-05) (hash(x)=3110930)
352
+ 336 train 7.549194 (lr=9.8915e-05) (hash(x)=2795079)
353
+ 337 train 7.586680 (lr=9.8872e-05) (hash(x)=2783946)
354
+ 338 train 7.482025 (lr=9.8827e-05) (hash(x)=1972781)
355
+ 339 train 7.393521 (lr=9.8782e-05) (hash(x)=2597532)
356
+ 340 train 7.563128 (lr=9.8736e-05) (hash(x)=3501313)
357
+ 341 train 7.999594 (lr=9.8689e-05) (hash(x)=2952253)
358
+ 342 train 7.725410 (lr=9.8641e-05) (hash(x)=3784221)
359
+ 343 train 7.800297 (lr=9.8592e-05) (hash(x)=3297008)
360
+ 344 train 7.776986 (lr=9.8543e-05) (hash(x)=3117650)
361
+ 345 train 7.783531 (lr=9.8492e-05) (hash(x)=3770315)
362
+ 346 train 7.763414 (lr=9.8441e-05) (hash(x)=3812153)
363
+ 347 train 7.706602 (lr=9.8389e-05) (hash(x)=2894320)
364
+ 348 train 7.721105 (lr=9.8336e-05) (hash(x)=3405673)
365
+ 349 train 7.836195 (lr=9.8282e-05) (hash(x)=3630469)
366
+ 350 val loss 7.7386
367
+ 350 val perplexity 2295.2683
368
+ 350 train 7.924701 (lr=9.8228e-05) (hash(x)=2997673)
369
+ 351 train 8.046050 (lr=9.8172e-05) (hash(x)=3316544)
370
+ 352 train 7.921103 (lr=9.8116e-05) (hash(x)=3082557)
371
+ 353 train 7.962475 (lr=9.8059e-05) (hash(x)=3971998)
372
+ 354 train 7.831978 (lr=9.8001e-05) (hash(x)=3318716)
373
+ 355 train 7.706338 (lr=9.7942e-05) (hash(x)=3344675)
374
+ 356 train 7.843235 (lr=9.7882e-05) (hash(x)=3309664)
375
+ 357 train 7.869767 (lr=9.7822e-05) (hash(x)=2829724)
376
+ 358 train 7.783576 (lr=9.7761e-05) (hash(x)=3638490)
377
+ 359 train 7.780287 (lr=9.7699e-05) (hash(x)=4077726)
378
+ 360 train 7.673557 (lr=9.7636e-05) (hash(x)=2648185)
379
+ 361 train 7.683789 (lr=9.7572e-05) (hash(x)=2781854)
380
+ 362 train 7.696439 (lr=9.7507e-05) (hash(x)=2843035)
381
+ 363 train 7.372821 (lr=9.7442e-05) (hash(x)=2932335)
382
+ 364 train 7.286936 (lr=9.7376e-05) (hash(x)=3218475)
383
+ 365 train 8.711348 (lr=9.7309e-05) (hash(x)=3187135)
384
+ 366 train 7.871809 (lr=9.7241e-05) (hash(x)=3914253)
385
+ 367 train 7.822980 (lr=9.7172e-05) (hash(x)=3174796)
386
+ 368 train 7.557837 (lr=9.7103e-05) (hash(x)=2897545)
387
+ 369 train 7.910929 (lr=9.7032e-05) (hash(x)=2950140)
388
+ 370 train 7.548553 (lr=9.6961e-05) (hash(x)=2618408)
389
+ 371 train 7.833662 (lr=9.6889e-05) (hash(x)=3278917)
390
+ 372 train 7.652622 (lr=9.6817e-05) (hash(x)=2607646)
391
+ 373 train 7.635096 (lr=9.6743e-05) (hash(x)=2947042)
392
+ 374 train 7.754431 (lr=9.6669e-05) (hash(x)=3028925)
393
+ 375 train 7.716943 (lr=9.6593e-05) (hash(x)=2993037)
394
+ 376 train 7.765289 (lr=9.6518e-05) (hash(x)=2038984)
395
+ 377 train 8.021362 (lr=9.6441e-05) (hash(x)=3886190)
396
+ 378 train 7.772281 (lr=9.6363e-05) (hash(x)=2901354)
397
+ 379 train 7.715843 (lr=9.6285e-05) (hash(x)=3607426)
398
+ 380 train 7.682667 (lr=9.6206e-05) (hash(x)=3622195)
399
+ 381 train 7.900334 (lr=9.6126e-05) (hash(x)=2721474)
400
+ 382 train 7.448871 (lr=9.6045e-05) (hash(x)=3169777)
401
+ 383 train 7.761519 (lr=9.5963e-05) (hash(x)=2742929)
402
+ 384 train 7.706453 (lr=9.5881e-05) (hash(x)=3161837)
403
+ 385 train 7.680935 (lr=9.5798e-05) (hash(x)=2938987)
404
+ 386 train 7.979664 (lr=9.5714e-05) (hash(x)=2988539)
405
+ 387 train 7.599267 (lr=9.5629e-05) (hash(x)=1784952)
406
+ 388 train 7.277911 (lr=9.5544e-05) (hash(x)=2672094)
407
+ 389 train 7.651691 (lr=9.5457e-05) (hash(x)=2489404)
408
+ 390 train 7.642127 (lr=9.5370e-05) (hash(x)=3072767)
409
+ 391 train 7.471303 (lr=9.5282e-05) (hash(x)=3291574)
410
+ 392 train 7.703460 (lr=9.5194e-05) (hash(x)=3052076)
411
+ 393 train 7.577206 (lr=9.5104e-05) (hash(x)=2924499)
412
+ 394 train 7.445159 (lr=9.5014e-05) (hash(x)=3234250)
413
+ 395 train 7.723222 (lr=9.4923e-05) (hash(x)=3419749)
414
+ 396 train 7.553703 (lr=9.4831e-05) (hash(x)=3159370)
415
+ 397 train 7.832929 (lr=9.4739e-05) (hash(x)=3299358)
416
+ 398 train 7.919606 (lr=9.4646e-05) (hash(x)=3560618)
417
+ 399 train 7.857443 (lr=9.4551e-05) (hash(x)=3755178)
418
+ 400 val loss 7.7339
419
+ 400 val perplexity 2284.4421
420
+ 400 train 7.684278 (lr=9.4457e-05) (hash(x)=2568447)
421
+ 401 train 7.700538 (lr=9.4361e-05) (hash(x)=3934113)
422
+ 402 train 7.792826 (lr=9.4265e-05) (hash(x)=3536573)
423
+ 403 train 8.287138 (lr=9.4168e-05) (hash(x)=3990344)
424
+ 404 train 7.934027 (lr=9.4070e-05) (hash(x)=2910818)
425
+ 405 train 7.548350 (lr=9.3971e-05) (hash(x)=2709366)
426
+ 406 train 7.673820 (lr=9.3872e-05) (hash(x)=2842817)
427
+ 407 train 7.750360 (lr=9.3772e-05) (hash(x)=3632553)
428
+ 408 train 7.888094 (lr=9.3671e-05) (hash(x)=3185053)
429
+ 409 train 7.967241 (lr=9.3569e-05) (hash(x)=3057647)
430
+ 410 train 7.416771 (lr=9.3467e-05) (hash(x)=2314061)
431
+ 411 train 7.620731 (lr=9.3364e-05) (hash(x)=3542024)
432
+ 412 train 7.893754 (lr=9.3260e-05) (hash(x)=3791703)
433
+ 413 train 7.604419 (lr=9.3155e-05) (hash(x)=2707461)
434
+ 414 train 7.648180 (lr=9.3050e-05) (hash(x)=3056759)
435
+ 415 train 8.103285 (lr=9.2944e-05) (hash(x)=3532987)
436
+ 416 train 8.025621 (lr=9.2837e-05) (hash(x)=4056578)
437
+ 417 train 7.655461 (lr=9.2729e-05) (hash(x)=3581161)
438
+ 418 train 7.646413 (lr=9.2621e-05) (hash(x)=1600859)
439
+ 419 train 7.688949 (lr=9.2512e-05) (hash(x)=3455941)
440
+ 420 train 7.906344 (lr=9.2402e-05) (hash(x)=3303318)
441
+ 421 train 7.755214 (lr=9.2292e-05) (hash(x)=3528629)
442
+ 422 train 7.736320 (lr=9.2181e-05) (hash(x)=2948835)
443
+ 423 train 7.490787 (lr=9.2069e-05) (hash(x)=2465412)
444
+ 424 train 7.516855 (lr=9.1956e-05) (hash(x)=3225341)
445
+ 425 train 7.684481 (lr=9.1843e-05) (hash(x)=3705855)
446
+ 426 train 7.610358 (lr=9.1729e-05) (hash(x)=2563141)
447
+ 427 train 7.669765 (lr=9.1614e-05) (hash(x)=3188390)
448
+ 428 train 7.563805 (lr=9.1499e-05) (hash(x)=3499566)
449
+ 429 train 7.662772 (lr=9.1382e-05) (hash(x)=2691682)
450
+ 430 train 7.606159 (lr=9.1266e-05) (hash(x)=2262959)
451
+ 431 train 7.709542 (lr=9.1148e-05) (hash(x)=3870848)
452
+ 432 train 7.740563 (lr=9.1030e-05) (hash(x)=2994145)
453
+ 433 train 7.517117 (lr=9.0911e-05) (hash(x)=2645411)
454
+ 434 train 7.715463 (lr=9.0791e-05) (hash(x)=3593252)
455
+ 435 train 7.567378 (lr=9.0671e-05) (hash(x)=3363163)
456
+ 436 train 7.760517 (lr=9.0550e-05) (hash(x)=3986069)
457
+ 437 train 7.871801 (lr=9.0428e-05) (hash(x)=3091084)
458
+ 438 train 7.742680 (lr=9.0306e-05) (hash(x)=3669915)
459
+ 439 train 7.789462 (lr=9.0182e-05) (hash(x)=3559093)
460
+ 440 train 7.852496 (lr=9.0059e-05) (hash(x)=2808277)
461
+ 441 train 7.603705 (lr=8.9934e-05) (hash(x)=3306636)
462
+ 442 train 7.876592 (lr=8.9809e-05) (hash(x)=2726692)
463
+ 443 train 7.811301 (lr=8.9683e-05) (hash(x)=3118177)
464
+ 444 train 7.738744 (lr=8.9557e-05) (hash(x)=3015629)
465
+ 445 train 7.691771 (lr=8.9430e-05) (hash(x)=3433982)
466
+ 446 train 7.851472 (lr=8.9302e-05) (hash(x)=2753224)
467
+ 447 train 7.708429 (lr=8.9173e-05) (hash(x)=2479541)
468
+ 448 train 7.479002 (lr=8.9044e-05) (hash(x)=2822449)
469
+ 449 train 7.730446 (lr=8.8914e-05) (hash(x)=3024717)
470
+ 450 val loss 7.7223
471
+ 450 val perplexity 2258.1130
472
+ 450 train 7.784658 (lr=8.8784e-05) (hash(x)=2985618)
473
+ 451 train 7.428979 (lr=8.8653e-05) (hash(x)=2680322)
474
+ 452 train 7.436573 (lr=8.8521e-05) (hash(x)=2497480)
475
+ 453 train 7.578758 (lr=8.8388e-05) (hash(x)=3283493)
476
+ 454 train 7.462043 (lr=8.8255e-05) (hash(x)=2602773)
477
+ 455 train 7.637937 (lr=8.8122e-05) (hash(x)=4527404)
478
+ 456 train 8.245945 (lr=8.7987e-05) (hash(x)=2601776)
479
+ 457 train 7.963500 (lr=8.7852e-05) (hash(x)=3612631)
480
+ 458 train 7.989337 (lr=8.7717e-05) (hash(x)=2946281)
481
+ 459 train 7.793707 (lr=8.7580e-05) (hash(x)=4038857)
482
+ 460 train 7.818914 (lr=8.7444e-05) (hash(x)=3059628)
483
+ 461 train 7.692846 (lr=8.7306e-05) (hash(x)=4298752)
484
+ 462 train 7.815265 (lr=8.7168e-05) (hash(x)=3154731)
485
+ 463 train 7.538111 (lr=8.7029e-05) (hash(x)=3236940)
486
+ 464 train 7.710227 (lr=8.6890e-05) (hash(x)=2660300)
487
+ 465 train 7.590045 (lr=8.6750e-05) (hash(x)=3064872)
488
+ 466 train 8.024341 (lr=8.6609e-05) (hash(x)=2939627)
489
+ 467 train 7.811286 (lr=8.6468e-05) (hash(x)=3040813)
490
+ 468 train 8.248059 (lr=8.6326e-05) (hash(x)=3819797)
491
+ 469 train 7.920714 (lr=8.6184e-05) (hash(x)=3451533)
492
+ 470 train 7.666557 (lr=8.6041e-05) (hash(x)=3711520)
493
+ 471 train 7.702642 (lr=8.5897e-05) (hash(x)=2946468)
494
+ 472 train 7.521824 (lr=8.5753e-05) (hash(x)=2689478)
495
+ 473 train 7.413901 (lr=8.5608e-05) (hash(x)=3418499)
496
+ 474 train 8.023088 (lr=8.5462e-05) (hash(x)=3380977)
497
+ 475 train 7.726617 (lr=8.5316e-05) (hash(x)=3500766)
498
+ 476 train 7.579147 (lr=8.5170e-05) (hash(x)=2168497)
499
+ 477 train 7.602959 (lr=8.5022e-05) (hash(x)=2657291)
500
+ 478 train 7.686066 (lr=8.4875e-05) (hash(x)=2965595)
501
+ 479 train 7.755686 (lr=8.4726e-05) (hash(x)=3628445)
502
+ 480 train 7.757569 (lr=8.4577e-05) (hash(x)=3425156)
503
+ 481 train 8.016519 (lr=8.4428e-05) (hash(x)=3673023)
504
+ 482 train 7.738678 (lr=8.4278e-05) (hash(x)=3501380)
505
+ 483 train 7.852448 (lr=8.4127e-05) (hash(x)=4575840)
506
+ 484 train 7.773813 (lr=8.3976e-05) (hash(x)=3166341)
507
+ 485 train 7.699785 (lr=8.3824e-05) (hash(x)=2093331)
508
+ 486 train 7.187489 (lr=8.3672e-05) (hash(x)=2049200)
509
+ 487 train 7.227901 (lr=8.3519e-05) (hash(x)=2372381)
510
+ 488 train 7.371218 (lr=8.3366e-05) (hash(x)=3414285)
511
+ 489 train 7.883346 (lr=8.3212e-05) (hash(x)=2805788)
512
+ 490 train 7.758993 (lr=8.3057e-05) (hash(x)=3849228)
513
+ 491 train 7.725756 (lr=8.2902e-05) (hash(x)=3463844)
514
+ 492 train 7.886817 (lr=8.2746e-05) (hash(x)=3204745)
515
+ 493 train 7.741932 (lr=8.2590e-05) (hash(x)=2618019)
516
+ 494 train 7.701122 (lr=8.2434e-05) (hash(x)=3045488)
517
+ 495 train 7.853068 (lr=8.2276e-05) (hash(x)=4708728)
518
+ 496 train 7.711629 (lr=8.2119e-05) (hash(x)=2453637)
519
+ 497 train 7.700097 (lr=8.1960e-05) (hash(x)=3193265)
520
+ 498 train 7.577980 (lr=8.1801e-05) (hash(x)=3411630)
521
+ 499 train 7.862544 (lr=8.1642e-05) (hash(x)=2766207)
522
+ 500 val loss 7.7082
523
+ 500 val perplexity 2226.5925
524
+ 500 train 7.616567 (lr=8.1482e-05) (hash(x)=3606738)
525
+ 501 train 7.804639 (lr=8.1322e-05) (hash(x)=2330970)
526
+ 502 train 7.792992 (lr=8.1161e-05) (hash(x)=3317886)
527
+ 503 train 7.612552 (lr=8.1000e-05) (hash(x)=2766434)
528
+ 504 train 7.654883 (lr=8.0838e-05) (hash(x)=3524713)
529
+ 505 train 7.970875 (lr=8.0676e-05) (hash(x)=2620477)
530
+ 506 train 7.185073 (lr=8.0513e-05) (hash(x)=2177504)
531
+ 507 train 6.894221 (lr=8.0349e-05) (hash(x)=1459901)
532
+ 508 train 7.281404 (lr=8.0186e-05) (hash(x)=4023924)
533
+ 509 train 7.811387 (lr=8.0021e-05) (hash(x)=3364474)
534
+ 510 train 7.529491 (lr=7.9856e-05) (hash(x)=2282910)
535
+ 511 train 7.786476 (lr=7.9691e-05) (hash(x)=3437140)
536
+ 512 train 7.584352 (lr=7.9525e-05) (hash(x)=3446393)
537
+ 513 train 8.053323 (lr=7.9359e-05) (hash(x)=4050994)
538
+ 514 train 7.369313 (lr=7.9192e-05) (hash(x)=2651397)
539
+ 515 train 7.402227 (lr=7.9025e-05) (hash(x)=2473336)
540
+ 516 train 7.759040 (lr=7.8858e-05) (hash(x)=2909170)
541
+ 517 train 7.728101 (lr=7.8689e-05) (hash(x)=2726907)
542
+ 518 train 8.030631 (lr=7.8521e-05) (hash(x)=4655886)
543
+ 519 train 7.648452 (lr=7.8352e-05) (hash(x)=3191889)
544
+ 520 train 7.851241 (lr=7.8182e-05) (hash(x)=3063304)
545
+ 521 train 7.881397 (lr=7.8012e-05) (hash(x)=2981356)
546
+ 522 train 7.765809 (lr=7.7842e-05) (hash(x)=3680996)
547
+ 523 train 7.955995 (lr=7.7671e-05) (hash(x)=2971592)
548
+ 524 train 7.628803 (lr=7.7500e-05) (hash(x)=2727589)
549
+ 525 train 7.693187 (lr=7.7328e-05) (hash(x)=3369519)
550
+ 526 train 7.838012 (lr=7.7156e-05) (hash(x)=3979663)
551
+ 527 train 7.714499 (lr=7.6984e-05) (hash(x)=3072949)
552
+ 528 train 7.856801 (lr=7.6811e-05) (hash(x)=3331935)
553
+ 529 train 7.602625 (lr=7.6637e-05) (hash(x)=3197904)
554
+ 530 train 7.790174 (lr=7.6463e-05) (hash(x)=3292870)
555
+ 531 train 7.878369 (lr=7.6289e-05) (hash(x)=3895149)
556
+ 532 train 7.975493 (lr=7.6115e-05) (hash(x)=2975885)
557
+ 533 train 7.719143 (lr=7.5940e-05) (hash(x)=3200202)
558
+ 534 train 7.871094 (lr=7.5764e-05) (hash(x)=2709773)
559
+ 535 train 7.745011 (lr=7.5588e-05) (hash(x)=2959066)
560
+ 536 train 7.635227 (lr=7.5412e-05) (hash(x)=3457380)
561
+ 537 train 7.600094 (lr=7.5235e-05) (hash(x)=2978755)
562
+ 538 train 7.730713 (lr=7.5058e-05) (hash(x)=3107300)
563
+ 539 train 7.872008 (lr=7.4881e-05) (hash(x)=3207361)
564
+ 540 train 7.722160 (lr=7.4703e-05) (hash(x)=2533388)
565
+ 541 train 7.640673 (lr=7.4525e-05) (hash(x)=2797327)
566
+ 542 train 7.414634 (lr=7.4346e-05) (hash(x)=2844633)
567
+ 543 train 7.999400 (lr=7.4167e-05) (hash(x)=3047968)
568
+ 544 train 7.735019 (lr=7.3988e-05) (hash(x)=4853965)
569
+ 545 train 7.786438 (lr=7.3808e-05) (hash(x)=2957193)
570
+ 546 train 7.762244 (lr=7.3628e-05) (hash(x)=2899336)
571
+ 547 train 7.532676 (lr=7.3448e-05) (hash(x)=2709892)
572
+ 548 train 7.535606 (lr=7.3267e-05) (hash(x)=2837887)
573
+ 549 train 7.657208 (lr=7.3086e-05) (hash(x)=4039987)
574
+ 550 val loss 7.6951
575
+ 550 val perplexity 2197.6497
576
+ 550 train 7.557309 (lr=7.2904e-05) (hash(x)=2737971)
577
+ 551 train 7.734739 (lr=7.2723e-05) (hash(x)=2674322)
578
+ 552 train 7.609689 (lr=7.2540e-05) (hash(x)=3570711)
579
+ 553 train 7.698533 (lr=7.2358e-05) (hash(x)=3045172)
580
+ 554 train 7.738959 (lr=7.2175e-05) (hash(x)=2833715)
581
+ 555 train 7.569333 (lr=7.1992e-05) (hash(x)=2614055)
582
+ 556 train 7.850594 (lr=7.1808e-05) (hash(x)=3489655)
583
+ 557 train 7.778706 (lr=7.1624e-05) (hash(x)=4519396)
584
+ 558 train 8.011612 (lr=7.1440e-05) (hash(x)=3148019)
585
+ 559 train 7.710378 (lr=7.1256e-05) (hash(x)=2678501)
586
+ 560 train 7.826849 (lr=7.1071e-05) (hash(x)=2846095)
587
+ 561 train 7.445323 (lr=7.0886e-05) (hash(x)=3544564)
588
+ 562 train 7.856921 (lr=7.0701e-05) (hash(x)=4116563)
589
+ 563 train 8.019300 (lr=7.0515e-05) (hash(x)=2047883)
590
+ 564 train 7.814747 (lr=7.0329e-05) (hash(x)=3243076)
591
+ 565 train 7.737323 (lr=7.0143e-05) (hash(x)=2873219)
592
+ 566 train 7.845903 (lr=6.9956e-05) (hash(x)=2881108)
593
+ 567 train 7.755179 (lr=6.9769e-05) (hash(x)=2972085)
594
+ 568 train 7.620904 (lr=6.9582e-05) (hash(x)=2328968)
595
+ 569 train 7.768929 (lr=6.9394e-05) (hash(x)=4015293)
596
+ 570 train 7.806571 (lr=6.9207e-05) (hash(x)=3211990)
597
+ 571 train 7.847847 (lr=6.9019e-05) (hash(x)=3015019)
598
+ 572 train 7.542269 (lr=6.8830e-05) (hash(x)=2413554)
599
+ 573 train 7.690924 (lr=6.8642e-05) (hash(x)=3332166)
600
+ 574 train 7.647289 (lr=6.8453e-05) (hash(x)=2486251)
601
+ 575 train 7.472944 (lr=6.8264e-05) (hash(x)=2127731)
602
+ 576 train 7.684792 (lr=6.8075e-05) (hash(x)=2333731)
603
+ 577 train 7.541497 (lr=6.7885e-05) (hash(x)=3279094)
604
+ 578 train 7.798142 (lr=6.7695e-05) (hash(x)=4180720)
605
+ 579 train 7.808392 (lr=6.7505e-05) (hash(x)=2686206)
606
+ 580 train 7.812500 (lr=6.7315e-05) (hash(x)=3250400)
607
+ 581 train 7.632296 (lr=6.7124e-05) (hash(x)=2691923)
608
+ 582 train 7.242152 (lr=6.6933e-05) (hash(x)=2070780)
609
+ 583 train 7.289696 (lr=6.6742e-05) (hash(x)=2230443)
610
+ 584 train 7.468549 (lr=6.6551e-05) (hash(x)=2108372)
611
+ 585 train 7.775241 (lr=6.6360e-05) (hash(x)=3297222)
612
+ 586 train 7.638796 (lr=6.6168e-05) (hash(x)=3214252)
613
+ 587 train 7.537973 (lr=6.5976e-05) (hash(x)=2258539)
614
+ 588 train 7.881364 (lr=6.5784e-05) (hash(x)=5278979)
615
+ 589 train 8.559334 (lr=6.5592e-05) (hash(x)=4227667)
616
+ 590 train 8.949348 (lr=6.5399e-05) (hash(x)=4465380)
617
+ 591 train 7.618426 (lr=6.5206e-05) (hash(x)=3023900)
618
+ 592 train 7.888482 (lr=6.5013e-05) (hash(x)=4040345)
619
+ 593 train 7.771491 (lr=6.4820e-05) (hash(x)=3212440)
620
+ 594 train 8.012320 (lr=6.4627e-05) (hash(x)=3150459)
621
+ 595 train 8.284838 (lr=6.4433e-05) (hash(x)=3910339)
622
+ 596 train 8.078747 (lr=6.4240e-05) (hash(x)=3272745)
623
+ 597 train 7.660906 (lr=6.4046e-05) (hash(x)=2681147)
624
+ 598 train 7.770867 (lr=6.3852e-05) (hash(x)=2684449)
625
+ 599 train 7.724843 (lr=6.3658e-05) (hash(x)=3077483)
626
+ 600 val loss 7.7065
627
+ 600 val perplexity 2222.7534
628
+ 600 train 7.729784 (lr=6.3463e-05) (hash(x)=3128166)
629
+ 601 train 7.716200 (lr=6.3269e-05) (hash(x)=2435623)
630
+ 602 train 7.666214 (lr=6.3074e-05) (hash(x)=3500606)
631
+ 603 train 7.826246 (lr=6.2879e-05) (hash(x)=3668040)
632
+ 604 train 7.722524 (lr=6.2684e-05) (hash(x)=3530457)
633
+ 605 train 8.061076 (lr=6.2489e-05) (hash(x)=3884772)
634
+ 606 train 7.932255 (lr=6.2294e-05) (hash(x)=3375192)
635
+ 607 train 7.562283 (lr=6.2098e-05) (hash(x)=2480588)
636
+ 608 train 7.920752 (lr=6.1903e-05) (hash(x)=3779663)
637
+ 609 train 7.781637 (lr=6.1707e-05) (hash(x)=2858176)
638
+ 610 train 7.788981 (lr=6.1511e-05) (hash(x)=3372951)
639
+ 611 train 7.827242 (lr=6.1315e-05) (hash(x)=4355904)
640
+ 612 train 7.916214 (lr=6.1119e-05) (hash(x)=3390073)
641
+ 613 train 7.917703 (lr=6.0923e-05) (hash(x)=3226998)
642
+ 614 train 7.824528 (lr=6.0726e-05) (hash(x)=2231691)
643
+ 615 train 7.826435 (lr=6.0530e-05) (hash(x)=2508700)
644
+ 616 train 7.685073 (lr=6.0333e-05) (hash(x)=3156546)
645
+ 617 train 7.798107 (lr=6.0137e-05) (hash(x)=3845366)
646
+ 618 train 7.643012 (lr=5.9940e-05) (hash(x)=2434970)
647
+ 619 train 7.622969 (lr=5.9743e-05) (hash(x)=3142447)
648
+ 620 train 7.606694 (lr=5.9546e-05) (hash(x)=3276912)
649
+ 621 train 7.414536 (lr=5.9349e-05) (hash(x)=2998061)
650
+ 622 train 7.761463 (lr=5.9152e-05) (hash(x)=2596688)
651
+ 623 train 7.649961 (lr=5.8955e-05) (hash(x)=3500594)
652
+ 624 train 8.198074 (lr=5.8758e-05) (hash(x)=3161569)
653
+ 625 train 7.718914 (lr=5.8560e-05) (hash(x)=2084274)
654
+ 626 train 7.550472 (lr=5.8363e-05) (hash(x)=3102855)
655
+ 627 train 7.673632 (lr=5.8165e-05) (hash(x)=3730075)
656
+ 628 train 7.834611 (lr=5.7968e-05) (hash(x)=3761585)
657
+ 629 train 7.969739 (lr=5.7770e-05) (hash(x)=2879419)
658
+ 630 train 8.734007 (lr=5.7573e-05) (hash(x)=2636386)
659
+ 631 train 8.144592 (lr=5.7375e-05) (hash(x)=3585841)
660
+ 632 train 8.287870 (lr=5.7177e-05) (hash(x)=3471713)
661
+ 633 train 8.552389 (lr=5.6979e-05) (hash(x)=2909843)
662
+ 634 train 8.093419 (lr=5.6782e-05) (hash(x)=3781776)
663
+ 635 train 7.744789 (lr=5.6584e-05) (hash(x)=2936395)
664
+ 636 train 7.770822 (lr=5.6386e-05) (hash(x)=3062681)
665
+ 637 train 7.852262 (lr=5.6188e-05) (hash(x)=3899268)
666
+ 638 train 7.654629 (lr=5.5990e-05) (hash(x)=2820006)
667
+ 639 train 7.699104 (lr=5.5792e-05) (hash(x)=2979578)
668
+ 640 train 7.455833 (lr=5.5594e-05) (hash(x)=3668622)
669
+ 641 train 7.513448 (lr=5.5396e-05) (hash(x)=2930025)
670
+ 642 train 7.726802 (lr=5.5198e-05) (hash(x)=3437396)
671
+ 643 train 7.631406 (lr=5.5000e-05) (hash(x)=2976458)
672
+ 644 train 7.602330 (lr=5.4802e-05) (hash(x)=3161473)
673
+ 645 train 7.859408 (lr=5.4604e-05) (hash(x)=3522699)
674
+ 646 train 7.490274 (lr=5.4406e-05) (hash(x)=2594466)
675
+ 647 train 7.561731 (lr=5.4208e-05) (hash(x)=3006372)
676
+ 648 train 7.767431 (lr=5.4010e-05) (hash(x)=2690183)
677
+ 649 train 7.890172 (lr=5.3812e-05) (hash(x)=3527477)
678
+ 650 val loss 7.6975
679
+ 650 val perplexity 2202.8828
680
+ 650 train 7.734461 (lr=5.3614e-05) (hash(x)=3443996)
681
+ 651 train 7.754220 (lr=5.3416e-05) (hash(x)=3252677)
682
+ 652 train 7.714626 (lr=5.3218e-05) (hash(x)=3489611)
683
+ 653 train 7.781690 (lr=5.3021e-05) (hash(x)=4383071)
684
+ 654 train 7.765073 (lr=5.2823e-05) (hash(x)=3901307)
685
+ 655 train 7.626541 (lr=5.2625e-05) (hash(x)=2619193)
686
+ 656 train 7.328915 (lr=5.2427e-05) (hash(x)=3129958)
687
+ 657 train 7.269726 (lr=5.2230e-05) (hash(x)=2670739)
688
+ 658 train 7.071469 (lr=5.2032e-05) (hash(x)=2719534)
689
+ 659 train 7.503823 (lr=5.1835e-05) (hash(x)=2844557)
690
+ 660 train 7.273498 (lr=5.1637e-05) (hash(x)=2459601)
691
+ 661 train 7.576645 (lr=5.1440e-05) (hash(x)=2997987)
692
+ 662 train 7.693564 (lr=5.1242e-05) (hash(x)=3089231)
693
+ 663 train 7.458712 (lr=5.1045e-05) (hash(x)=2264862)
694
+ 664 train 7.309120 (lr=5.0848e-05) (hash(x)=3509149)
695
+ 665 train 7.619244 (lr=5.0651e-05) (hash(x)=3163570)
696
+ 666 train 7.281568 (lr=5.0454e-05) (hash(x)=2195792)
697
+ 667 train 7.321415 (lr=5.0257e-05) (hash(x)=2321001)
698
+ 668 train 7.595953 (lr=5.0060e-05) (hash(x)=2740256)
699
+ 669 train 7.600167 (lr=4.9863e-05) (hash(x)=3544588)
700
+ 670 train 7.770852 (lr=4.9667e-05) (hash(x)=2924570)
701
+ 671 train 7.813039 (lr=4.9470e-05) (hash(x)=3953318)
702
+ 672 train 7.927788 (lr=4.9274e-05) (hash(x)=3012566)
703
+ 673 train 7.746851 (lr=4.9077e-05) (hash(x)=3470444)
704
+ 674 train 7.777851 (lr=4.8881e-05) (hash(x)=3119940)
705
+ 675 train 7.770487 (lr=4.8685e-05) (hash(x)=3813804)
706
+ 676 train 7.742265 (lr=4.8489e-05) (hash(x)=3190117)
707
+ 677 train 7.596648 (lr=4.8293e-05) (hash(x)=3430688)
708
+ 678 train 7.859132 (lr=4.8097e-05) (hash(x)=2814603)
709
+ 679 train 7.284520 (lr=4.7902e-05) (hash(x)=3172114)
710
+ 680 train 8.074198 (lr=4.7706e-05) (hash(x)=5108346)
711
+ 681 train 7.764048 (lr=4.7511e-05) (hash(x)=3707048)
712
+ 682 train 7.586308 (lr=4.7316e-05) (hash(x)=2418366)
713
+ 683 train 7.723191 (lr=4.7121e-05) (hash(x)=3630061)
714
+ 684 train 7.771107 (lr=4.6926e-05) (hash(x)=3906638)
715
+ 685 train 7.883831 (lr=4.6731e-05) (hash(x)=4499523)
716
+ 686 train 8.513484 (lr=4.6537e-05) (hash(x)=2997194)
717
+ 687 train 7.704493 (lr=4.6342e-05) (hash(x)=3624493)
718
+ 688 train 7.527813 (lr=4.6148e-05) (hash(x)=3196325)
719
+ 689 train 7.600121 (lr=4.5954e-05) (hash(x)=3669400)
720
+ 690 train 7.721524 (lr=4.5760e-05) (hash(x)=2952088)
721
+ 691 train 7.614277 (lr=4.5567e-05) (hash(x)=2885835)
722
+ 692 train 7.536109 (lr=4.5373e-05) (hash(x)=2977433)
723
+ 693 train 7.548788 (lr=4.5180e-05) (hash(x)=2906377)
724
+ 694 train 7.698526 (lr=4.4987e-05) (hash(x)=3446862)
725
+ 695 train 7.589169 (lr=4.4794e-05) (hash(x)=3628332)
726
+ 696 train 7.668311 (lr=4.4601e-05) (hash(x)=3518254)
727
+ 697 train 7.613544 (lr=4.4408e-05) (hash(x)=2766583)
728
+ 698 train 7.807809 (lr=4.4216e-05) (hash(x)=3132882)
729
+ 699 train 7.752291 (lr=4.4024e-05) (hash(x)=2844282)
730
+ 700 val loss 7.6692
731
+ 700 val perplexity 2141.2634
732
+ 700 train 7.705235 (lr=4.3832e-05) (hash(x)=3455892)
733
+ 701 train 7.614131 (lr=4.3640e-05) (hash(x)=3834162)
734
+ 702 train 7.689454 (lr=4.3449e-05) (hash(x)=3621421)
735
+ 703 train 7.594209 (lr=4.3258e-05) (hash(x)=2888972)
736
+ 704 train 7.622677 (lr=4.3067e-05) (hash(x)=3936409)
737
+ 705 train 7.959282 (lr=4.2876e-05) (hash(x)=3123026)
738
+ 706 train 7.669536 (lr=4.2685e-05) (hash(x)=2747922)
739
+ 707 train 7.516227 (lr=4.2495e-05) (hash(x)=2904637)
740
+ 708 train 7.599037 (lr=4.2305e-05) (hash(x)=3436413)
741
+ 709 train 7.519498 (lr=4.2115e-05) (hash(x)=2616177)
742
+ 710 train 7.913063 (lr=4.1925e-05) (hash(x)=2950778)
743
+ 711 train 7.947219 (lr=4.1736e-05) (hash(x)=3744709)
744
+ 712 train 7.896531 (lr=4.1547e-05) (hash(x)=3418206)
745
+ 713 train 7.712865 (lr=4.1358e-05) (hash(x)=1995841)
746
+ 714 train 7.610532 (lr=4.1170e-05) (hash(x)=3226340)
747
+ 715 train 7.691141 (lr=4.0981e-05) (hash(x)=3186059)
748
+ 716 train 7.785026 (lr=4.0793e-05) (hash(x)=3973127)
749
+ 717 train 7.616951 (lr=4.0606e-05) (hash(x)=3828964)
750
+ 718 train 7.840805 (lr=4.0418e-05) (hash(x)=3374694)
751
+ 719 train 7.600031 (lr=4.0231e-05) (hash(x)=2564054)
752
+ 720 train 7.681393 (lr=4.0044e-05) (hash(x)=2889182)
753
+ 721 train 7.481635 (lr=3.9857e-05) (hash(x)=2604802)
754
+ 722 train 7.658703 (lr=3.9671e-05) (hash(x)=2631069)
755
+ 723 train 7.600266 (lr=3.9485e-05) (hash(x)=3342173)
756
+ 724 train 7.731849 (lr=3.9299e-05) (hash(x)=3614538)
757
+ 725 train 8.181367 (lr=3.9114e-05) (hash(x)=3616575)
758
+ 726 train 7.319047 (lr=3.8929e-05) (hash(x)=2401284)
759
+ 727 train 7.433576 (lr=3.8744e-05) (hash(x)=3222229)
760
+ 728 train 7.865627 (lr=3.8560e-05) (hash(x)=3053865)
761
+ 729 train 7.701032 (lr=3.8376e-05) (hash(x)=3412947)
762
+ 730 train 7.889108 (lr=3.8192e-05) (hash(x)=2248853)
763
+ 731 train 7.379097 (lr=3.8008e-05) (hash(x)=3292116)
764
+ 732 train 7.504256 (lr=3.7825e-05) (hash(x)=3757215)
765
+ 733 train 7.445259 (lr=3.7642e-05) (hash(x)=3271264)
766
+ 734 train 8.019453 (lr=3.7460e-05) (hash(x)=6026595)
767
+ 735 train 7.942544 (lr=3.7277e-05) (hash(x)=2824555)
768
+ 736 train 7.624786 (lr=3.7096e-05) (hash(x)=3156632)
769
+ 737 train 7.966803 (lr=3.6914e-05) (hash(x)=4644238)
770
+ 738 train 7.729155 (lr=3.6733e-05) (hash(x)=3330900)
771
+ 739 train 7.704562 (lr=3.6552e-05) (hash(x)=3286145)
772
+ 740 train 7.562980 (lr=3.6372e-05) (hash(x)=2981506)
773
+ 741 train 7.783583 (lr=3.6192e-05) (hash(x)=2640920)
774
+ 742 train 8.155554 (lr=3.6012e-05) (hash(x)=3144232)
775
+ 743 train 7.409521 (lr=3.5833e-05) (hash(x)=2694606)
776
+ 744 train 7.525813 (lr=3.5654e-05) (hash(x)=3076354)
777
+ 745 train 7.719265 (lr=3.5475e-05) (hash(x)=4075841)
778
+ 746 train 7.632718 (lr=3.5297e-05) (hash(x)=2890259)
779
+ 747 train 7.692908 (lr=3.5119e-05) (hash(x)=2964484)
780
+ 748 train 7.726641 (lr=3.4942e-05) (hash(x)=3141147)
781
+ 749 train 7.577957 (lr=3.4765e-05) (hash(x)=2502564)
782
+ 750 val loss 7.6492
783
+ 750 val perplexity 2098.9497
784
+ 750 train 7.730612 (lr=3.4588e-05) (hash(x)=2405634)
785
+ 751 train 7.515952 (lr=3.4412e-05) (hash(x)=2797954)
786
+ 752 train 7.206916 (lr=3.4236e-05) (hash(x)=2463151)
787
+ 753 train 7.342766 (lr=3.4060e-05) (hash(x)=3552620)
788
+ 754 train 8.143105 (lr=3.3885e-05) (hash(x)=4445877)
789
+ 755 train 8.206616 (lr=3.3711e-05) (hash(x)=2821863)
790
+ 756 train 7.545918 (lr=3.3537e-05) (hash(x)=2096879)
791
+ 757 train 7.999121 (lr=3.3363e-05) (hash(x)=3709124)
792
+ 758 train 7.917977 (lr=3.3189e-05) (hash(x)=3538344)
793
+ 759 train 7.967587 (lr=3.3016e-05) (hash(x)=2860360)
794
+ 760 train 7.627438 (lr=3.2844e-05) (hash(x)=2233793)
795
+ 761 train 7.725009 (lr=3.2672e-05) (hash(x)=3052447)
796
+ 762 train 7.931804 (lr=3.2500e-05) (hash(x)=3901950)
797
+ 763 train 7.777191 (lr=3.2329e-05) (hash(x)=3794897)
798
+ 764 train 7.827371 (lr=3.2158e-05) (hash(x)=2874773)
799
+ 765 train 7.861567 (lr=3.1988e-05) (hash(x)=3277698)
800
+ 766 train 7.605861 (lr=3.1818e-05) (hash(x)=3706890)
801
+ 767 train 8.217529 (lr=3.1648e-05) (hash(x)=3971772)
802
+ 768 train 7.743663 (lr=3.1479e-05) (hash(x)=3599671)
803
+ 769 train 7.499330 (lr=3.1311e-05) (hash(x)=4210652)
804
+ 770 train 7.602262 (lr=3.1142e-05) (hash(x)=3564381)
805
+ 771 train 7.614793 (lr=3.0975e-05) (hash(x)=3512226)
806
+ 772 train 7.707340 (lr=3.0808e-05) (hash(x)=3766758)
807
+ 773 train 7.714253 (lr=3.0641e-05) (hash(x)=3088734)
808
+ 774 train 7.832429 (lr=3.0475e-05) (hash(x)=2125989)
809
+ 775 train 7.522181 (lr=3.0309e-05) (hash(x)=2815361)
810
+ 776 train 7.338858 (lr=3.0144e-05) (hash(x)=2664781)
811
+ 777 train 7.647563 (lr=2.9979e-05) (hash(x)=2831165)
812
+ 778 train 7.704540 (lr=2.9814e-05) (hash(x)=2982462)
813
+ 779 train 7.824671 (lr=2.9651e-05) (hash(x)=3301986)
814
+ 780 train 7.759628 (lr=2.9487e-05) (hash(x)=4116610)
815
+ 781 train 7.772216 (lr=2.9324e-05) (hash(x)=3405120)
816
+ 782 train 7.565571 (lr=2.9162e-05) (hash(x)=3324806)
817
+ 783 train 7.647415 (lr=2.9000e-05) (hash(x)=2826471)
818
+ 784 train 7.745193 (lr=2.8839e-05) (hash(x)=3689635)
819
+ 785 train 7.667083 (lr=2.8678e-05) (hash(x)=3171450)
820
+ 786 train 7.580828 (lr=2.8518e-05) (hash(x)=3249622)
821
+ 787 train 7.721552 (lr=2.8358e-05) (hash(x)=2547188)
822
+ 788 train 7.207758 (lr=2.8199e-05) (hash(x)=2559165)
823
+ 789 train 7.153617 (lr=2.8040e-05) (hash(x)=2604163)
824
+ 790 train 7.431013 (lr=2.7881e-05) (hash(x)=3277487)
825
+ 791 train 7.745605 (lr=2.7724e-05) (hash(x)=3781597)
826
+ 792 train 7.624376 (lr=2.7566e-05) (hash(x)=3614573)
827
+ 793 train 7.732265 (lr=2.7410e-05) (hash(x)=3359105)
828
+ 794 train 7.546487 (lr=2.7254e-05) (hash(x)=3235782)
829
+ 795 train 7.533357 (lr=2.7098e-05) (hash(x)=2933586)
830
+ 796 train 7.600265 (lr=2.6943e-05) (hash(x)=2702037)
831
+ 797 train 7.447344 (lr=2.6788e-05) (hash(x)=4128427)
832
+ 798 train 8.039268 (lr=2.6634e-05) (hash(x)=3343017)
833
+ 799 train 7.297547 (lr=2.6481e-05) (hash(x)=1400697)
834
+ 800 val loss 7.6288
835
+ 800 val perplexity 2056.4990
836
+ 800 train 7.587326 (lr=2.6328e-05) (hash(x)=1958860)
837
+ 801 train 7.549398 (lr=2.6176e-05) (hash(x)=2541140)
838
+ 802 train 7.702372 (lr=2.6024e-05) (hash(x)=4050777)
839
+ 803 train 7.711732 (lr=2.5873e-05) (hash(x)=2718043)
840
+ 804 train 7.727318 (lr=2.5722e-05) (hash(x)=3444483)
841
+ 805 train 7.652524 (lr=2.5572e-05) (hash(x)=2976357)
842
+ 806 train 7.440941 (lr=2.5423e-05) (hash(x)=2998919)
843
+ 807 train 7.866652 (lr=2.5274e-05) (hash(x)=4188358)
844
+ 808 train 7.629499 (lr=2.5125e-05) (hash(x)=3235819)
845
+ 809 train 7.746847 (lr=2.4978e-05) (hash(x)=2773625)
846
+ 810 train 7.581608 (lr=2.4830e-05) (hash(x)=3716604)
847
+ 811 train 8.428396 (lr=2.4684e-05) (hash(x)=3639169)
848
+ 812 train 7.787879 (lr=2.4538e-05) (hash(x)=3384965)
849
+ 813 train 7.964314 (lr=2.4392e-05) (hash(x)=3486817)
850
+ 814 train 7.826596 (lr=2.4247e-05) (hash(x)=3242208)
851
+ 815 train 7.644888 (lr=2.4103e-05) (hash(x)=3085878)
852
+ 816 train 7.738395 (lr=2.3959e-05) (hash(x)=4775285)
853
+ 817 train 7.536144 (lr=2.3816e-05) (hash(x)=3634909)
854
+ 818 train 7.734795 (lr=2.3674e-05) (hash(x)=2856942)
855
+ 819 train 7.916406 (lr=2.3532e-05) (hash(x)=3858674)
856
+ 820 train 7.622629 (lr=2.3391e-05) (hash(x)=3621957)
857
+ 821 train 7.605330 (lr=2.3250e-05) (hash(x)=3812958)
858
+ 822 train 7.614458 (lr=2.3110e-05) (hash(x)=2617022)
859
+ 823 train 7.899922 (lr=2.2971e-05) (hash(x)=3356513)
860
+ 824 train 7.513614 (lr=2.2832e-05) (hash(x)=2428281)
861
+ 825 train 7.640045 (lr=2.2694e-05) (hash(x)=2684672)
862
+ 826 train 7.606279 (lr=2.2556e-05) (hash(x)=3426018)
863
+ 827 train 7.637681 (lr=2.2420e-05) (hash(x)=3329920)
864
+ 828 train 7.678614 (lr=2.2283e-05) (hash(x)=2641333)
865
+ 829 train 7.672116 (lr=2.2148e-05) (hash(x)=3346748)
866
+ 830 train 7.601799 (lr=2.2013e-05) (hash(x)=3170909)
867
+ 831 train 7.504438 (lr=2.1878e-05) (hash(x)=3350964)
868
+ 832 train 7.824353 (lr=2.1745e-05) (hash(x)=3638825)
869
+ 833 train 7.723385 (lr=2.1612e-05) (hash(x)=3318649)
870
+ 834 train 7.517222 (lr=2.1479e-05) (hash(x)=2247171)
871
+ 835 train 7.562825 (lr=2.1347e-05) (hash(x)=2339818)
872
+ 836 train 7.600541 (lr=2.1216e-05) (hash(x)=3047804)
873
+ 837 train 7.640975 (lr=2.1086e-05) (hash(x)=3070413)
874
+ 838 train 7.756958 (lr=2.0956e-05) (hash(x)=4828968)
875
+ 839 train 7.979217 (lr=2.0827e-05) (hash(x)=3639491)
876
+ 840 train 7.838798 (lr=2.0698e-05) (hash(x)=3367484)
877
+ 841 train 7.787946 (lr=2.0570e-05) (hash(x)=2181128)
878
+ 842 train 7.536163 (lr=2.0443e-05) (hash(x)=2872878)
879
+ 843 train 7.815197 (lr=2.0317e-05) (hash(x)=3420184)
880
+ 844 train 7.610316 (lr=2.0191e-05) (hash(x)=3424340)
881
+ 845 train 7.666057 (lr=2.0066e-05) (hash(x)=3511906)
882
+ 846 train 7.603694 (lr=1.9941e-05) (hash(x)=2684357)
883
+ 847 train 7.812149 (lr=1.9818e-05) (hash(x)=3203986)
884
+ 848 train 7.208911 (lr=1.9694e-05) (hash(x)=2573740)
885
+ 849 train 7.487296 (lr=1.9572e-05) (hash(x)=2810112)
886
+ 850 val loss 7.6176
887
+ 850 val perplexity 2033.7723
888
+ 850 train 7.450761 (lr=1.9450e-05) (hash(x)=2945682)
889
+ 851 train 7.656694 (lr=1.9329e-05) (hash(x)=3738516)
890
+ 852 train 7.669497 (lr=1.9209e-05) (hash(x)=2945498)
891
+ 853 train 7.792475 (lr=1.9089e-05) (hash(x)=2584256)
892
+ 854 train 7.872504 (lr=1.8970e-05) (hash(x)=3962389)
893
+ 855 train 7.639168 (lr=1.8852e-05) (hash(x)=2755250)
894
+ 856 train 7.533469 (lr=1.8734e-05) (hash(x)=2813792)
895
+ 857 train 7.412879 (lr=1.8618e-05) (hash(x)=3085599)
896
+ 858 train 7.323928 (lr=1.8501e-05) (hash(x)=2451315)
897
+ 859 train 7.487380 (lr=1.8386e-05) (hash(x)=3081821)
898
+ 860 train 7.618304 (lr=1.8271e-05) (hash(x)=2281098)
899
+ 861 train 7.783508 (lr=1.8157e-05) (hash(x)=3566151)
900
+ 862 train 7.586865 (lr=1.8044e-05) (hash(x)=3357680)
901
+ 863 train 7.386048 (lr=1.7931e-05) (hash(x)=4133377)
902
+ 864 train 7.616713 (lr=1.7819e-05) (hash(x)=3371626)
903
+ 865 train 7.671497 (lr=1.7708e-05) (hash(x)=3502104)
904
+ 866 train 7.654167 (lr=1.7598e-05) (hash(x)=3179786)
905
+ 867 train 7.647291 (lr=1.7488e-05) (hash(x)=2768070)
906
+ 868 train 7.656028 (lr=1.7379e-05) (hash(x)=3082980)
907
+ 869 train 7.721316 (lr=1.7271e-05) (hash(x)=3640006)
908
+ 870 train 7.865849 (lr=1.7163e-05) (hash(x)=3632476)
909
+ 871 train 7.546025 (lr=1.7056e-05) (hash(x)=2682018)
910
+ 872 train 7.906270 (lr=1.6950e-05) (hash(x)=3857222)
911
+ 873 train 7.560998 (lr=1.6845e-05) (hash(x)=2629721)
912
+ 874 train 7.667650 (lr=1.6740e-05) (hash(x)=3116372)
913
+ 875 train 7.551920 (lr=1.6636e-05) (hash(x)=3149746)
914
+ 876 train 7.672505 (lr=1.6533e-05) (hash(x)=3681364)
915
+ 877 train 7.770788 (lr=1.6431e-05) (hash(x)=3944386)
916
+ 878 train 7.519994 (lr=1.6329e-05) (hash(x)=3493113)
917
+ 879 train 7.405013 (lr=1.6228e-05) (hash(x)=2856601)
918
+ 880 train 7.827049 (lr=1.6128e-05) (hash(x)=3192138)
919
+ 881 train 7.612295 (lr=1.6029e-05) (hash(x)=3191705)
920
+ 882 train 7.694275 (lr=1.5930e-05) (hash(x)=3104887)
921
+ 883 train 7.627579 (lr=1.5832e-05) (hash(x)=3989915)
922
+ 884 train 7.681943 (lr=1.5735e-05) (hash(x)=2442597)
923
+ 885 train 7.611646 (lr=1.5639e-05) (hash(x)=3279917)
924
+ 886 train 7.728721 (lr=1.5543e-05) (hash(x)=2852523)
925
+ 887 train 7.546513 (lr=1.5449e-05) (hash(x)=2981901)
926
+ 888 train 7.369968 (lr=1.5354e-05) (hash(x)=2709365)
927
+ 889 train 7.692051 (lr=1.5261e-05) (hash(x)=3399281)
928
+ 890 train 7.344348 (lr=1.5169e-05) (hash(x)=3527238)
929
+ 891 train 7.546730 (lr=1.5077e-05) (hash(x)=3728099)
930
+ 892 train 7.826212 (lr=1.4986e-05) (hash(x)=3523105)
931
+ 893 train 7.595063 (lr=1.4896e-05) (hash(x)=2521093)
932
+ 894 train 7.704844 (lr=1.4806e-05) (hash(x)=3145786)
933
+ 895 train 7.590614 (lr=1.4718e-05) (hash(x)=3328627)
934
+ 896 train 7.557797 (lr=1.4630e-05) (hash(x)=2756657)
935
+ 897 train 7.357015 (lr=1.4543e-05) (hash(x)=2728827)
936
+ 898 train 7.464124 (lr=1.4456e-05) (hash(x)=2393931)
937
+ 899 train 7.271106 (lr=1.4371e-05) (hash(x)=2562198)
938
+ 900 val loss 7.5990
939
+ 900 val perplexity 1996.2244
940
+ 900 train 7.534462 (lr=1.4286e-05) (hash(x)=3110252)
941
+ 901 train 7.642040 (lr=1.4202e-05) (hash(x)=3588452)
942
+ 902 train 7.634254 (lr=1.4119e-05) (hash(x)=3063440)
943
+ 903 train 7.701039 (lr=1.4037e-05) (hash(x)=3671502)
944
+ 904 train 7.734181 (lr=1.3955e-05) (hash(x)=3212481)
945
+ 905 train 7.769210 (lr=1.3874e-05) (hash(x)=3128347)
946
+ 906 train 7.834904 (lr=1.3794e-05) (hash(x)=3845319)
947
+ 907 train 7.805075 (lr=1.3715e-05) (hash(x)=2890655)
948
+ 908 train 7.580709 (lr=1.3637e-05) (hash(x)=3707729)
949
+ 909 train 7.694618 (lr=1.3559e-05) (hash(x)=3249651)
950
+ 910 train 7.507112 (lr=1.3482e-05) (hash(x)=2852206)
951
+ 911 train 7.586023 (lr=1.3407e-05) (hash(x)=3307897)
952
+ 912 train 7.172713 (lr=1.3331e-05) (hash(x)=2778919)
953
+ 913 train 7.597015 (lr=1.3257e-05) (hash(x)=2618491)
954
+ 914 train 7.605618 (lr=1.3183e-05) (hash(x)=3289311)
955
+ 915 train 7.892837 (lr=1.3111e-05) (hash(x)=4195998)
956
+ 916 train 7.487579 (lr=1.3039e-05) (hash(x)=2824400)
957
+ 917 train 7.549895 (lr=1.2968e-05) (hash(x)=3225031)
958
+ 918 train 7.576835 (lr=1.2897e-05) (hash(x)=3490967)
959
+ 919 train 7.768996 (lr=1.2828e-05) (hash(x)=3326435)
960
+ 920 train 7.532475 (lr=1.2759e-05) (hash(x)=3686113)
961
+ 921 train 7.463759 (lr=1.2691e-05) (hash(x)=2862440)
962
+ 922 train 7.575618 (lr=1.2624e-05) (hash(x)=2626015)
963
+ 923 train 7.726254 (lr=1.2558e-05) (hash(x)=2918467)
964
+ 924 train 7.706365 (lr=1.2493e-05) (hash(x)=2715041)
965
+ 925 train 7.534023 (lr=1.2428e-05) (hash(x)=4141440)
966
+ 926 train 7.341287 (lr=1.2364e-05) (hash(x)=2236345)
967
+ 927 train 7.497160 (lr=1.2301e-05) (hash(x)=2956276)
968
+ 928 train 7.474613 (lr=1.2239e-05) (hash(x)=2701949)
969
+ 929 train 7.579115 (lr=1.2178e-05) (hash(x)=2820808)
970
+ 930 train 7.336380 (lr=1.2118e-05) (hash(x)=4317946)
971
+ 931 train 7.662438 (lr=1.2058e-05) (hash(x)=3053243)
972
+ 932 train 7.559228 (lr=1.1999e-05) (hash(x)=2844025)
973
+ 933 train 7.663240 (lr=1.1941e-05) (hash(x)=3799382)
974
+ 934 train 7.518514 (lr=1.1884e-05) (hash(x)=3505437)
975
+ 935 train 7.712101 (lr=1.1828e-05) (hash(x)=3490211)
976
+ 936 train 7.752571 (lr=1.1772e-05) (hash(x)=3924537)
977
+ 937 train 7.439594 (lr=1.1718e-05) (hash(x)=3421416)
978
+ 938 train 7.647201 (lr=1.1664e-05) (hash(x)=2152062)
979
+ 939 train 7.581032 (lr=1.1611e-05) (hash(x)=3184017)
980
+ 940 train 7.783771 (lr=1.1559e-05) (hash(x)=3792013)
981
+ 941 train 7.471547 (lr=1.1508e-05) (hash(x)=3039663)
982
+ 942 train 7.487977 (lr=1.1457e-05) (hash(x)=3640516)
983
+ 943 train 7.451119 (lr=1.1408e-05) (hash(x)=3166300)
984
+ 944 train 7.413672 (lr=1.1359e-05) (hash(x)=3379396)
985
+ 945 train 7.631753 (lr=1.1311e-05) (hash(x)=2689958)
986
+ 946 train 7.619427 (lr=1.1264e-05) (hash(x)=3644685)
987
+ 947 train 7.617109 (lr=1.1218e-05) (hash(x)=2893184)
988
+ 948 train 7.853418 (lr=1.1173e-05) (hash(x)=4142539)
989
+ 949 train 7.553109 (lr=1.1128e-05) (hash(x)=4102744)
990
+ 950 val loss 7.5876
991
+ 950 val perplexity 1973.5031
992
+ 950 train 7.794882 (lr=1.1085e-05) (hash(x)=3337629)
993
+ 951 train 7.899271 (lr=1.1042e-05) (hash(x)=3617664)
994
+ 952 train 7.460125 (lr=1.1000e-05) (hash(x)=2990263)
995
+ 953 train 7.424429 (lr=1.0959e-05) (hash(x)=2945509)
996
+ 954 train 7.816028 (lr=1.0919e-05) (hash(x)=2829576)
997
+ 955 train 7.342905 (lr=1.0879e-05) (hash(x)=2838328)
998
+ 956 train 7.681808 (lr=1.0841e-05) (hash(x)=2809450)
999
+ 957 train 7.708008 (lr=1.0803e-05) (hash(x)=3459400)
1000
+ 958 train 7.424464 (lr=1.0766e-05) (hash(x)=2872315)
1001
+ 959 train 7.489458 (lr=1.0730e-05) (hash(x)=3992250)
1002
+ 960 train 7.495971 (lr=1.0695e-05) (hash(x)=3131162)
1003
+ 961 train 7.550020 (lr=1.0661e-05) (hash(x)=3066475)
1004
+ 962 train 7.675117 (lr=1.0628e-05) (hash(x)=2730169)
1005
+ 963 train 7.547653 (lr=1.0595e-05) (hash(x)=3164373)
1006
+ 964 train 7.587910 (lr=1.0563e-05) (hash(x)=2453767)
1007
+ 965 train 7.694930 (lr=1.0533e-05) (hash(x)=3075962)
1008
+ 966 train 7.654838 (lr=1.0503e-05) (hash(x)=2890913)
1009
+ 967 train 7.638484 (lr=1.0474e-05) (hash(x)=3796173)
1010
+ 968 train 7.630119 (lr=1.0445e-05) (hash(x)=3445261)
1011
+ 969 train 7.554083 (lr=1.0418e-05) (hash(x)=3677710)
1012
+ 970 train 7.855076 (lr=1.0391e-05) (hash(x)=3195420)
1013
+ 971 train 7.525368 (lr=1.0366e-05) (hash(x)=3076126)
1014
+ 972 train 7.338386 (lr=1.0341e-05) (hash(x)=2856714)
1015
+ 973 train 7.580235 (lr=1.0317e-05) (hash(x)=3808030)
1016
+ 974 train 7.675290 (lr=1.0294e-05) (hash(x)=3500981)
1017
+ 975 train 7.623466 (lr=1.0272e-05) (hash(x)=3195966)
1018
+ 976 train 7.795401 (lr=1.0251e-05) (hash(x)=2954039)
1019
+ 977 train 7.499160 (lr=1.0230e-05) (hash(x)=2790464)
1020
+ 978 train 7.402913 (lr=1.0211e-05) (hash(x)=3189190)
1021
+ 979 train 7.450305 (lr=1.0192e-05) (hash(x)=2837463)
1022
+ 980 train 7.233965 (lr=1.0174e-05) (hash(x)=2340601)
1023
+ 981 train 7.094156 (lr=1.0157e-05) (hash(x)=2418446)
1024
+ 982 train 7.416530 (lr=1.0141e-05) (hash(x)=3899399)
1025
+ 983 train 7.698303 (lr=1.0126e-05) (hash(x)=2910115)
1026
+ 984 train 7.674072 (lr=1.0111e-05) (hash(x)=3373312)
1027
+ 985 train 7.493239 (lr=1.0098e-05) (hash(x)=2548970)
1028
+ 986 train 7.470074 (lr=1.0085e-05) (hash(x)=3355705)
1029
+ 987 train 7.568470 (lr=1.0074e-05) (hash(x)=3834685)
1030
+ 988 train 7.656357 (lr=1.0063e-05) (hash(x)=3783278)
1031
+ 989 train 7.775276 (lr=1.0053e-05) (hash(x)=2952651)
1032
+ 990 train 7.565278 (lr=1.0044e-05) (hash(x)=3207153)
1033
+ 991 train 7.546044 (lr=1.0035e-05) (hash(x)=2734508)
1034
+ 992 train 7.614619 (lr=1.0028e-05) (hash(x)=2633642)
1035
+ 993 train 7.612174 (lr=1.0021e-05) (hash(x)=3436107)
1036
+ 994 train 7.582572 (lr=1.0016e-05) (hash(x)=3340592)
1037
+ 995 train 7.712637 (lr=1.0011e-05) (hash(x)=3584964)
1038
+ 996 train 7.678309 (lr=1.0007e-05) (hash(x)=3281643)
1039
+ 997 train 7.483663 (lr=1.0004e-05) (hash(x)=1899895)
1040
+ 998 train 7.312737 (lr=1.0002e-05) (hash(x)=2864615)
1041
+ 999 val loss 7.5828
1042
+ 999 val perplexity 1964.0270
1043
+ 999 train 7.705114 (lr=1.0000e-05) (hash(x)=3807932)
lr1e-4_total_batch_size5120_seq_len64/model_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:445aaed26a68f265bd0376e9ffe25a56a90662c3b46cd2015ddacf8278ccf9ec
3
+ size 35540418
lr1e-4_total_batch_size5120_seq_len64/optimizer_00999.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4006c693bc77908d2cb5a82092be0c21922bbcef410be9058902a212e51e2ec5
3
+ size 70698822