Ablation H-MID-H4_svd_fp32-s1
Browse files
H/MID/H4_svd_fp32/seed1/epoch_1_checkpoint.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:40ccc7789910d415da304b5686f80e481c5af7812cc6dfe5310588f9202c80a4
|
| 3 |
+
size 2254319
|
H/MID/H4_svd_fp32/seed1/final_report.json
ADDED
|
@@ -0,0 +1,575 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"config": {
|
| 3 |
+
"group": "H",
|
| 4 |
+
"variant": "H4_svd_fp32",
|
| 5 |
+
"band": "MID",
|
| 6 |
+
"seed": 1,
|
| 7 |
+
"phase": 2,
|
| 8 |
+
"num_epochs": 1,
|
| 9 |
+
"batch_size": 256,
|
| 10 |
+
"overrides": {
|
| 11 |
+
"svd": "fp32"
|
| 12 |
+
},
|
| 13 |
+
"description": "H-MID-H4_svd_fp32-s1"
|
| 14 |
+
},
|
| 15 |
+
"run_config": {
|
| 16 |
+
"matrix_v": 64,
|
| 17 |
+
"D": 8,
|
| 18 |
+
"patch_size": 16,
|
| 19 |
+
"hidden": 64,
|
| 20 |
+
"depth": 1,
|
| 21 |
+
"n_cross_layers": 1,
|
| 22 |
+
"n_heads": 4,
|
| 23 |
+
"max_alpha": 0.2,
|
| 24 |
+
"alpha_init": -2.0,
|
| 25 |
+
"img_size": 64,
|
| 26 |
+
"batch_size": 128,
|
| 27 |
+
"lr": 0.0001,
|
| 28 |
+
"epochs": 1,
|
| 29 |
+
"weight_decay": 0.0,
|
| 30 |
+
"use_cv_ema": true,
|
| 31 |
+
"cv_ema_alpha": 0.01,
|
| 32 |
+
"cv_alignment_epochs": 0,
|
| 33 |
+
"cv_measure_every": 50,
|
| 34 |
+
"cv_sigma_scale": 0.3,
|
| 35 |
+
"boost": 0.5,
|
| 36 |
+
"cross_attn_clip": 0.5,
|
| 37 |
+
"allowed_types": [
|
| 38 |
+
0,
|
| 39 |
+
1,
|
| 40 |
+
2,
|
| 41 |
+
3,
|
| 42 |
+
4,
|
| 43 |
+
5,
|
| 44 |
+
6,
|
| 45 |
+
7,
|
| 46 |
+
8,
|
| 47 |
+
9,
|
| 48 |
+
10,
|
| 49 |
+
11,
|
| 50 |
+
12,
|
| 51 |
+
13,
|
| 52 |
+
14,
|
| 53 |
+
15
|
| 54 |
+
],
|
| 55 |
+
"train_size": 1000000,
|
| 56 |
+
"val_size": 10000,
|
| 57 |
+
"num_workers": 2,
|
| 58 |
+
"report_every": 100,
|
| 59 |
+
"major_report_every": 10,
|
| 60 |
+
"save_every": 5,
|
| 61 |
+
"seed": 1,
|
| 62 |
+
"hf_repo": "AbstractPhil/geolip-svae-batteries",
|
| 63 |
+
"upload": false
|
| 64 |
+
},
|
| 65 |
+
"cv_ema_final": 0.3687735359961555,
|
| 66 |
+
"cv_last": 0.38299983530974047,
|
| 67 |
+
"predicted_band": "MID",
|
| 68 |
+
"expected_band": "MID",
|
| 69 |
+
"band_match": true,
|
| 70 |
+
"test_mse": 0.9426831603050232,
|
| 71 |
+
"recon_ema": 0.9474882988305766,
|
| 72 |
+
"S0": 3.6427149772644043,
|
| 73 |
+
"SD": 2.120894432067871,
|
| 74 |
+
"ratio": 1.7175371414113185,
|
| 75 |
+
"erank": 7.878095626831055,
|
| 76 |
+
"observed_sphere_cv": 0.38667705104474637,
|
| 77 |
+
"uniform_sphere_cv_prediction": 0.35418994230276496,
|
| 78 |
+
"band_deviation": 0.03248710874198141,
|
| 79 |
+
"params_finite": true,
|
| 80 |
+
"num_epochs_run": 1,
|
| 81 |
+
"start_epoch": 0,
|
| 82 |
+
"per_epoch_metrics": [
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1,
|
| 85 |
+
"test_mse": 0.9426831603050232,
|
| 86 |
+
"cv_ema": 0.3687735359961555,
|
| 87 |
+
"observed_sphere_cv": 0.38667705104474637,
|
| 88 |
+
"band_deviation": 0.03248710874198141,
|
| 89 |
+
"erank": 7.878095626831055,
|
| 90 |
+
"params_finite": true,
|
| 91 |
+
"wallclock_seconds": 33.74544167518616,
|
| 92 |
+
"checkpoint_path": "/content/ablations_phase2/H/MID/H4_svd_fp32_s1/epoch_1_checkpoint.pt"
|
| 93 |
+
}
|
| 94 |
+
],
|
| 95 |
+
"params_count": 183339,
|
| 96 |
+
"wallclock_seconds": 33.7468101978302,
|
| 97 |
+
"batches_completed": 3906,
|
| 98 |
+
"batch_limit": 3906,
|
| 99 |
+
"cv_trajectory": [
|
| 100 |
+
{
|
| 101 |
+
"batch": 0,
|
| 102 |
+
"cv": 0.35279552109123474,
|
| 103 |
+
"cv_ema": 0.35279552109123474,
|
| 104 |
+
"recon": 1.7395105361938477
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"batch": 50,
|
| 108 |
+
"cv": 0.3576809269668788,
|
| 109 |
+
"cv_ema": 0.3528443751499912,
|
| 110 |
+
"recon": 1.711482048034668
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"batch": 100,
|
| 114 |
+
"cv": 0.37322904503700627,
|
| 115 |
+
"cv_ema": 0.3530482218488613,
|
| 116 |
+
"recon": 1.3940622806549072
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"batch": 150,
|
| 120 |
+
"cv": 0.3809825488523182,
|
| 121 |
+
"cv_ema": 0.35332756511889585,
|
| 122 |
+
"recon": 1.2634203433990479
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"batch": 200,
|
| 126 |
+
"cv": 0.363367159921018,
|
| 127 |
+
"cv_ema": 0.3534279610669171,
|
| 128 |
+
"recon": 1.2130711078643799
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"batch": 250,
|
| 132 |
+
"cv": 0.36187364844540687,
|
| 133 |
+
"cv_ema": 0.353512417940702,
|
| 134 |
+
"recon": 1.1935745477676392
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"batch": 300,
|
| 138 |
+
"cv": 0.3886440247335108,
|
| 139 |
+
"cv_ema": 0.35386373400863,
|
| 140 |
+
"recon": 1.124927043914795
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"batch": 350,
|
| 144 |
+
"cv": 0.35395367182779147,
|
| 145 |
+
"cv_ema": 0.3538646333868216,
|
| 146 |
+
"recon": 1.1946351528167725
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"batch": 400,
|
| 150 |
+
"cv": 0.3638793832258269,
|
| 151 |
+
"cv_ema": 0.35396478088521166,
|
| 152 |
+
"recon": 0.8293387293815613
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"batch": 450,
|
| 156 |
+
"cv": 0.35708758407648233,
|
| 157 |
+
"cv_ema": 0.35399600891712435,
|
| 158 |
+
"recon": 1.0617161989212036
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"batch": 500,
|
| 162 |
+
"cv": 0.3910815829028044,
|
| 163 |
+
"cv_ema": 0.35436686465698114,
|
| 164 |
+
"recon": 0.9334442615509033
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"batch": 550,
|
| 168 |
+
"cv": 0.41627702368454744,
|
| 169 |
+
"cv_ema": 0.3549859662472568,
|
| 170 |
+
"recon": 0.9824468493461609
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"batch": 600,
|
| 174 |
+
"cv": 0.3663982086871857,
|
| 175 |
+
"cv_ema": 0.3551000886716561,
|
| 176 |
+
"recon": 0.9238033294677734
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"batch": 650,
|
| 180 |
+
"cv": 0.399247404986695,
|
| 181 |
+
"cv_ema": 0.3555415618348065,
|
| 182 |
+
"recon": 0.9988247752189636
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"batch": 700,
|
| 186 |
+
"cv": 0.39021216258577907,
|
| 187 |
+
"cv_ema": 0.35588826784231625,
|
| 188 |
+
"recon": 0.903925895690918
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"batch": 750,
|
| 192 |
+
"cv": 0.40540620068753086,
|
| 193 |
+
"cv_ema": 0.35638344717076836,
|
| 194 |
+
"recon": 0.9849305152893066
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"batch": 800,
|
| 198 |
+
"cv": 0.37046451704866784,
|
| 199 |
+
"cv_ema": 0.3565242578695474,
|
| 200 |
+
"recon": 1.0423660278320312
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"batch": 850,
|
| 204 |
+
"cv": 0.3878175760196294,
|
| 205 |
+
"cv_ema": 0.35683719105104816,
|
| 206 |
+
"recon": 1.0945713520050049
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"batch": 900,
|
| 210 |
+
"cv": 0.38274018948251537,
|
| 211 |
+
"cv_ema": 0.35709622103536287,
|
| 212 |
+
"recon": 0.8816555738449097
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"batch": 950,
|
| 216 |
+
"cv": 0.3919742643192616,
|
| 217 |
+
"cv_ema": 0.35744500146820185,
|
| 218 |
+
"recon": 1.1342581510543823
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"batch": 1000,
|
| 222 |
+
"cv": 0.3777959811394234,
|
| 223 |
+
"cv_ema": 0.35764851126491404,
|
| 224 |
+
"recon": 1.1699141263961792
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"batch": 1050,
|
| 228 |
+
"cv": 0.3758352480620029,
|
| 229 |
+
"cv_ema": 0.3578303786328849,
|
| 230 |
+
"recon": 0.9987677931785583
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"batch": 1100,
|
| 234 |
+
"cv": 0.37015982943065856,
|
| 235 |
+
"cv_ema": 0.3579536731408627,
|
| 236 |
+
"recon": 1.0387179851531982
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"batch": 1150,
|
| 240 |
+
"cv": 0.3353433043945045,
|
| 241 |
+
"cv_ema": 0.35772756945339906,
|
| 242 |
+
"recon": 0.9521224498748779
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"batch": 1200,
|
| 246 |
+
"cv": 0.3805808783400831,
|
| 247 |
+
"cv_ema": 0.3579561025422659,
|
| 248 |
+
"recon": 1.0105128288269043
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"batch": 1250,
|
| 252 |
+
"cv": 0.4072234470663867,
|
| 253 |
+
"cv_ema": 0.3584487759875071,
|
| 254 |
+
"recon": 1.0297743082046509
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"batch": 1300,
|
| 258 |
+
"cv": 0.4203487078894899,
|
| 259 |
+
"cv_ema": 0.3590677753065269,
|
| 260 |
+
"recon": 0.9743930101394653
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"batch": 1350,
|
| 264 |
+
"cv": 0.37797572389455275,
|
| 265 |
+
"cv_ema": 0.35925685479240715,
|
| 266 |
+
"recon": 1.0832836627960205
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"batch": 1400,
|
| 270 |
+
"cv": 0.38594222111459714,
|
| 271 |
+
"cv_ema": 0.3595237084556291,
|
| 272 |
+
"recon": 0.9526357650756836
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"batch": 1450,
|
| 276 |
+
"cv": 0.3505802916382021,
|
| 277 |
+
"cv_ema": 0.35943427428745484,
|
| 278 |
+
"recon": 0.7738136053085327
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"batch": 1500,
|
| 282 |
+
"cv": 0.44269113014676775,
|
| 283 |
+
"cv_ema": 0.36026684284604793,
|
| 284 |
+
"recon": 0.8589984774589539
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"batch": 1550,
|
| 288 |
+
"cv": 0.37018737154919346,
|
| 289 |
+
"cv_ema": 0.3603660481330794,
|
| 290 |
+
"recon": 0.9227110147476196
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"batch": 1600,
|
| 294 |
+
"cv": 0.3525421674258676,
|
| 295 |
+
"cv_ema": 0.36028780932600724,
|
| 296 |
+
"recon": 1.060892105102539
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"batch": 1650,
|
| 300 |
+
"cv": 0.3649825551295781,
|
| 301 |
+
"cv_ema": 0.3603347567840429,
|
| 302 |
+
"recon": 1.0482091903686523
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"batch": 1700,
|
| 306 |
+
"cv": 0.3646609633764666,
|
| 307 |
+
"cv_ema": 0.3603780188499672,
|
| 308 |
+
"recon": 1.0489906072616577
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"batch": 1750,
|
| 312 |
+
"cv": 0.37387205512187766,
|
| 313 |
+
"cv_ema": 0.3605129592126863,
|
| 314 |
+
"recon": 0.7998015284538269
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"batch": 1800,
|
| 318 |
+
"cv": 0.37611169044990334,
|
| 319 |
+
"cv_ema": 0.36066894652505843,
|
| 320 |
+
"recon": 1.087367057800293
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"batch": 1850,
|
| 324 |
+
"cv": 0.3893921655400618,
|
| 325 |
+
"cv_ema": 0.36095617871520846,
|
| 326 |
+
"recon": 0.9795621633529663
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"batch": 1900,
|
| 330 |
+
"cv": 0.3950270117745475,
|
| 331 |
+
"cv_ema": 0.3612968870458018,
|
| 332 |
+
"recon": 1.0220293998718262
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"batch": 1950,
|
| 336 |
+
"cv": 0.376394033411517,
|
| 337 |
+
"cv_ema": 0.36144785850945893,
|
| 338 |
+
"recon": 0.9663494825363159
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"batch": 2000,
|
| 342 |
+
"cv": 0.38089426384321917,
|
| 343 |
+
"cv_ema": 0.3616423225627965,
|
| 344 |
+
"recon": 0.9779435992240906
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"batch": 2050,
|
| 348 |
+
"cv": 0.388021791683153,
|
| 349 |
+
"cv_ema": 0.36190611725400007,
|
| 350 |
+
"recon": 1.0033812522888184
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"batch": 2100,
|
| 354 |
+
"cv": 0.42768796323269614,
|
| 355 |
+
"cv_ema": 0.36256393571378703,
|
| 356 |
+
"recon": 1.0004607439041138
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"batch": 2150,
|
| 360 |
+
"cv": 0.42172557122859716,
|
| 361 |
+
"cv_ema": 0.36315555206893513,
|
| 362 |
+
"recon": 0.9089059233665466
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"batch": 2200,
|
| 366 |
+
"cv": 0.3785602451706307,
|
| 367 |
+
"cv_ema": 0.3633095989999521,
|
| 368 |
+
"recon": 0.976809024810791
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"batch": 2250,
|
| 372 |
+
"cv": 0.4301390431878807,
|
| 373 |
+
"cv_ema": 0.3639778934418314,
|
| 374 |
+
"recon": 0.7599304914474487
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"batch": 2300,
|
| 378 |
+
"cv": 0.383788533873862,
|
| 379 |
+
"cv_ema": 0.36417599984615173,
|
| 380 |
+
"recon": 1.0153391361236572
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"batch": 2350,
|
| 384 |
+
"cv": 0.3546983354508382,
|
| 385 |
+
"cv_ema": 0.3640812232021986,
|
| 386 |
+
"recon": 0.903059720993042
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"batch": 2400,
|
| 390 |
+
"cv": 0.39516122252510144,
|
| 391 |
+
"cv_ema": 0.36439202319542763,
|
| 392 |
+
"recon": 0.9019742012023926
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"batch": 2450,
|
| 396 |
+
"cv": 0.34607938049932213,
|
| 397 |
+
"cv_ema": 0.3642088967684666,
|
| 398 |
+
"recon": 0.8698627352714539
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"batch": 2500,
|
| 402 |
+
"cv": 0.3831538421352598,
|
| 403 |
+
"cv_ema": 0.3643983462221345,
|
| 404 |
+
"recon": 0.8243594169616699
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"batch": 2550,
|
| 408 |
+
"cv": 0.4264567089913737,
|
| 409 |
+
"cv_ema": 0.3650189298498269,
|
| 410 |
+
"recon": 0.9537364840507507
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"batch": 2600,
|
| 414 |
+
"cv": 0.3913787937612435,
|
| 415 |
+
"cv_ema": 0.3652825284889411,
|
| 416 |
+
"recon": 0.9144038558006287
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"batch": 2650,
|
| 420 |
+
"cv": 0.4136208278202612,
|
| 421 |
+
"cv_ema": 0.3657659114822543,
|
| 422 |
+
"recon": 0.9588565230369568
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"batch": 2700,
|
| 426 |
+
"cv": 0.36801115001160944,
|
| 427 |
+
"cv_ema": 0.36578836386754787,
|
| 428 |
+
"recon": 0.851468563079834
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"batch": 2750,
|
| 432 |
+
"cv": 0.3676516343890863,
|
| 433 |
+
"cv_ema": 0.3658069965727633,
|
| 434 |
+
"recon": 0.9601517915725708
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"batch": 2800,
|
| 438 |
+
"cv": 0.4004389144315645,
|
| 439 |
+
"cv_ema": 0.3661533157513513,
|
| 440 |
+
"recon": 0.9909943342208862
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"batch": 2850,
|
| 444 |
+
"cv": 0.37530969319193475,
|
| 445 |
+
"cv_ema": 0.3662448795257571,
|
| 446 |
+
"recon": 0.881865382194519
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"batch": 2900,
|
| 450 |
+
"cv": 0.3695447692846085,
|
| 451 |
+
"cv_ema": 0.36627787842334564,
|
| 452 |
+
"recon": 0.9554804563522339
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"batch": 2950,
|
| 456 |
+
"cv": 0.40221775969136014,
|
| 457 |
+
"cv_ema": 0.3666372772360258,
|
| 458 |
+
"recon": 0.8924350738525391
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"batch": 3000,
|
| 462 |
+
"cv": 0.3710383199979106,
|
| 463 |
+
"cv_ema": 0.36668128766364466,
|
| 464 |
+
"recon": 0.9033645391464233
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"batch": 3050,
|
| 468 |
+
"cv": 0.4029121169432733,
|
| 469 |
+
"cv_ema": 0.3670435959564409,
|
| 470 |
+
"recon": 0.8948702216148376
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"batch": 3100,
|
| 474 |
+
"cv": 0.360623147166502,
|
| 475 |
+
"cv_ema": 0.3669793914685415,
|
| 476 |
+
"recon": 0.9067516326904297
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"batch": 3150,
|
| 480 |
+
"cv": 0.42668781909443554,
|
| 481 |
+
"cv_ema": 0.36757647574480046,
|
| 482 |
+
"recon": 0.8444041013717651
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"batch": 3200,
|
| 486 |
+
"cv": 0.39485158698479617,
|
| 487 |
+
"cv_ema": 0.3678492268572004,
|
| 488 |
+
"recon": 0.9932960271835327
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"batch": 3250,
|
| 492 |
+
"cv": 0.3585830048621906,
|
| 493 |
+
"cv_ema": 0.36775656463725026,
|
| 494 |
+
"recon": 0.9426895380020142
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"batch": 3300,
|
| 498 |
+
"cv": 0.3635846092864066,
|
| 499 |
+
"cv_ema": 0.3677148450837418,
|
| 500 |
+
"recon": 1.0063133239746094
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"batch": 3350,
|
| 504 |
+
"cv": 0.36133573143406994,
|
| 505 |
+
"cv_ema": 0.36765105394724507,
|
| 506 |
+
"recon": 0.8604530096054077
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"batch": 3400,
|
| 510 |
+
"cv": 0.38534020098209126,
|
| 511 |
+
"cv_ema": 0.36782794541759356,
|
| 512 |
+
"recon": 0.8997382521629333
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"batch": 3450,
|
| 516 |
+
"cv": 0.38527143466378355,
|
| 517 |
+
"cv_ema": 0.36800238031005544,
|
| 518 |
+
"recon": 0.9783589243888855
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"batch": 3500,
|
| 522 |
+
"cv": 0.38911879190812004,
|
| 523 |
+
"cv_ema": 0.3682135444260361,
|
| 524 |
+
"recon": 1.0944323539733887
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"batch": 3550,
|
| 528 |
+
"cv": 0.37762123883511656,
|
| 529 |
+
"cv_ema": 0.3683076213701269,
|
| 530 |
+
"recon": 0.8415673971176147
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"batch": 3600,
|
| 534 |
+
"cv": 0.3911455367196811,
|
| 535 |
+
"cv_ema": 0.36853600052362245,
|
| 536 |
+
"recon": 0.935973048210144
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"batch": 3650,
|
| 540 |
+
"cv": 0.3792678290262593,
|
| 541 |
+
"cv_ema": 0.3686433188086488,
|
| 542 |
+
"recon": 0.9496223330497742
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"batch": 3700,
|
| 546 |
+
"cv": 0.3474759005781642,
|
| 547 |
+
"cv_ema": 0.368431644626344,
|
| 548 |
+
"recon": 1.091861605644226
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"batch": 3750,
|
| 552 |
+
"cv": 0.3859051960873772,
|
| 553 |
+
"cv_ema": 0.3686063801409543,
|
| 554 |
+
"recon": 0.9115563631057739
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"batch": 3800,
|
| 558 |
+
"cv": 0.3513181902132327,
|
| 559 |
+
"cv_ema": 0.3684334982416771,
|
| 560 |
+
"recon": 0.9223372340202332
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"batch": 3850,
|
| 564 |
+
"cv": 0.388067274382863,
|
| 565 |
+
"cv_ema": 0.36862983600308896,
|
| 566 |
+
"recon": 1.099926471710205
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"batch": 3900,
|
| 570 |
+
"cv": 0.38299983530974047,
|
| 571 |
+
"cv_ema": 0.3687735359961555,
|
| 572 |
+
"recon": 0.9758589267730713
|
| 573 |
+
}
|
| 574 |
+
]
|
| 575 |
+
}
|
H/MID/H4_svd_fp32/seed1/tensorboard/events.out.tfevents.1776797720.52b79b2c7612.4109.66
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d11987a7cdfa4dc1bf805c80040bee1fba322c90cf6600fc566f74f4d45e2e2
|
| 3 |
+
size 612125
|