Upload SIGMA_v13 checkpoints and results
Browse files
checkpoints_a3_2_best/best_meta.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 11,
|
| 3 |
+
"step": 3322,
|
| 4 |
+
"mrr": 0.9747689268096245,
|
| 5 |
+
"r1": 0.966,
|
| 6 |
+
"r10": 0.994
|
| 7 |
+
}
|
checkpoints_a3_2_best/best_model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86e65be9ddb62127bee583664ee627ad2cb381917115c2bfb10c369513d6a2c3
|
| 3 |
+
size 58984949
|
results_a3_2/a3_curve.png
ADDED
|
results_a3_2/a3_final.json
ADDED
|
@@ -0,0 +1,273 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"config": {
|
| 3 |
+
"dim": 256,
|
| 4 |
+
"n_steps": 2,
|
| 5 |
+
"top_k": 16,
|
| 6 |
+
"rank": 4,
|
| 7 |
+
"n_heads": 2,
|
| 8 |
+
"warmup_ratio": 0.1,
|
| 9 |
+
"ema_decay": 0.999,
|
| 10 |
+
"batch_size": 128
|
| 11 |
+
},
|
| 12 |
+
"final_metrics": {
|
| 13 |
+
"mrr": 0.9733942612942612,
|
| 14 |
+
"r1": 0.96,
|
| 15 |
+
"r5": 0.992,
|
| 16 |
+
"r10": 0.996
|
| 17 |
+
},
|
| 18 |
+
"best_mrr": 0.9747689268096245,
|
| 19 |
+
"history": {
|
| 20 |
+
"step": [
|
| 21 |
+
151,
|
| 22 |
+
302,
|
| 23 |
+
453,
|
| 24 |
+
604,
|
| 25 |
+
755,
|
| 26 |
+
906,
|
| 27 |
+
1057,
|
| 28 |
+
1208,
|
| 29 |
+
1359,
|
| 30 |
+
1510,
|
| 31 |
+
1661,
|
| 32 |
+
1812,
|
| 33 |
+
1963,
|
| 34 |
+
2114,
|
| 35 |
+
2265,
|
| 36 |
+
2416,
|
| 37 |
+
2567,
|
| 38 |
+
2718,
|
| 39 |
+
2869,
|
| 40 |
+
3020,
|
| 41 |
+
3171,
|
| 42 |
+
3322,
|
| 43 |
+
3473,
|
| 44 |
+
3624,
|
| 45 |
+
3775,
|
| 46 |
+
3926,
|
| 47 |
+
4077,
|
| 48 |
+
4228,
|
| 49 |
+
4379,
|
| 50 |
+
4530,
|
| 51 |
+
4681,
|
| 52 |
+
4832,
|
| 53 |
+
4983,
|
| 54 |
+
5134,
|
| 55 |
+
5285,
|
| 56 |
+
5436,
|
| 57 |
+
5587,
|
| 58 |
+
5738,
|
| 59 |
+
5889,
|
| 60 |
+
6040
|
| 61 |
+
],
|
| 62 |
+
"loss": [
|
| 63 |
+
0.05027373702518198,
|
| 64 |
+
0.04881227528984776,
|
| 65 |
+
0.04610809926807091,
|
| 66 |
+
0.044877655014691764,
|
| 67 |
+
0.043507184142505886,
|
| 68 |
+
0.04242624673030234,
|
| 69 |
+
0.040315509335884195,
|
| 70 |
+
0.0404422130337022,
|
| 71 |
+
0.03869027391520162,
|
| 72 |
+
0.039034821574162964,
|
| 73 |
+
0.035369555644740334,
|
| 74 |
+
0.03615325192199243,
|
| 75 |
+
0.03357540613768117,
|
| 76 |
+
0.034277692702383,
|
| 77 |
+
0.032801381047099634,
|
| 78 |
+
0.032548489850877924,
|
| 79 |
+
0.03131792017116847,
|
| 80 |
+
0.031045349573852212,
|
| 81 |
+
0.029416534149192815,
|
| 82 |
+
0.029242907721697298,
|
| 83 |
+
0.028275871316328742,
|
| 84 |
+
0.028179163338357446,
|
| 85 |
+
0.02576759449570187,
|
| 86 |
+
0.026437007923863384,
|
| 87 |
+
0.025067514110825313,
|
| 88 |
+
0.025000911616159,
|
| 89 |
+
0.02553811805039052,
|
| 90 |
+
0.024959771101369172,
|
| 91 |
+
0.023039054945851397,
|
| 92 |
+
0.02365897029379247,
|
| 93 |
+
0.022970872184890787,
|
| 94 |
+
0.022299184411492767,
|
| 95 |
+
0.0220370011909908,
|
| 96 |
+
0.021843978166826906,
|
| 97 |
+
0.022709016435292384,
|
| 98 |
+
0.022480477449416326,
|
| 99 |
+
0.0229772361587511,
|
| 100 |
+
0.02244766808071772,
|
| 101 |
+
0.02231158111367794,
|
| 102 |
+
0.021746117917709792
|
| 103 |
+
],
|
| 104 |
+
"mrr": [
|
| 105 |
+
0.007296020171075417,
|
| 106 |
+
0.006981434645512146,
|
| 107 |
+
0.007258763825553949,
|
| 108 |
+
0.020727019820485148,
|
| 109 |
+
0.08483017280957106,
|
| 110 |
+
0.2626234882736713,
|
| 111 |
+
0.5265114247854933,
|
| 112 |
+
0.7651447527168945,
|
| 113 |
+
0.8790372316147287,
|
| 114 |
+
0.927393956472425,
|
| 115 |
+
0.943739478573779,
|
| 116 |
+
0.9527592581925244,
|
| 117 |
+
0.963864755964756,
|
| 118 |
+
0.9671448232025848,
|
| 119 |
+
0.9690153613735899,
|
| 120 |
+
0.9693978692201517,
|
| 121 |
+
0.9698194688370101,
|
| 122 |
+
0.9718220349957911,
|
| 123 |
+
0.9745033430763609,
|
| 124 |
+
0.9738482047364401,
|
| 125 |
+
0.9737780177187154,
|
| 126 |
+
0.9747689268096245,
|
| 127 |
+
0.9714856345408978,
|
| 128 |
+
0.9714240665766981,
|
| 129 |
+
0.9714243107769424,
|
| 130 |
+
0.9704240665766981,
|
| 131 |
+
0.9711332276770357,
|
| 132 |
+
0.9713467965454432,
|
| 133 |
+
0.9703259189640768,
|
| 134 |
+
0.9703427234753551,
|
| 135 |
+
0.970055895418923,
|
| 136 |
+
0.9700560934387249,
|
| 137 |
+
0.9703537124863439,
|
| 138 |
+
0.9700203791530108,
|
| 139 |
+
0.9696648235974552,
|
| 140 |
+
0.969672582972583,
|
| 141 |
+
0.9696778461304776,
|
| 142 |
+
0.969672582972583,
|
| 143 |
+
0.9695418367346939,
|
| 144 |
+
0.9695894557823129
|
| 145 |
+
],
|
| 146 |
+
"r1": [
|
| 147 |
+
0.002,
|
| 148 |
+
0.002,
|
| 149 |
+
0.002,
|
| 150 |
+
0.006,
|
| 151 |
+
0.034,
|
| 152 |
+
0.174,
|
| 153 |
+
0.398,
|
| 154 |
+
0.656,
|
| 155 |
+
0.818,
|
| 156 |
+
0.894,
|
| 157 |
+
0.916,
|
| 158 |
+
0.928,
|
| 159 |
+
0.946,
|
| 160 |
+
0.952,
|
| 161 |
+
0.956,
|
| 162 |
+
0.956,
|
| 163 |
+
0.956,
|
| 164 |
+
0.96,
|
| 165 |
+
0.966,
|
| 166 |
+
0.964,
|
| 167 |
+
0.964,
|
| 168 |
+
0.966,
|
| 169 |
+
0.96,
|
| 170 |
+
0.96,
|
| 171 |
+
0.96,
|
| 172 |
+
0.958,
|
| 173 |
+
0.96,
|
| 174 |
+
0.96,
|
| 175 |
+
0.958,
|
| 176 |
+
0.958,
|
| 177 |
+
0.958,
|
| 178 |
+
0.958,
|
| 179 |
+
0.958,
|
| 180 |
+
0.958,
|
| 181 |
+
0.958,
|
| 182 |
+
0.958,
|
| 183 |
+
0.958,
|
| 184 |
+
0.958,
|
| 185 |
+
0.958,
|
| 186 |
+
0.958
|
| 187 |
+
],
|
| 188 |
+
"grad_norm": [
|
| 189 |
+
0.19050836369395255,
|
| 190 |
+
0.19429627671837807,
|
| 191 |
+
0.193626494333148,
|
| 192 |
+
0.2053767490386963,
|
| 193 |
+
0.1933288175612688,
|
| 194 |
+
0.1909618841856718,
|
| 195 |
+
0.19003299728035927,
|
| 196 |
+
0.18759957425296306,
|
| 197 |
+
0.18610450223088265,
|
| 198 |
+
0.20221985757350922,
|
| 199 |
+
0.1663146274536848,
|
| 200 |
+
0.1844347643107176,
|
| 201 |
+
0.1629525911062956,
|
| 202 |
+
0.18658689439296722,
|
| 203 |
+
0.16427210189402103,
|
| 204 |
+
0.16973903842270374,
|
| 205 |
+
0.1690176397562027,
|
| 206 |
+
0.16868569791316987,
|
| 207 |
+
0.15878911174833774,
|
| 208 |
+
0.1576488983631134,
|
| 209 |
+
0.15221125289797782,
|
| 210 |
+
0.157753631696105,
|
| 211 |
+
0.1504822540283203,
|
| 212 |
+
0.14835401758551597,
|
| 213 |
+
0.1520396003127098,
|
| 214 |
+
0.1515271785110235,
|
| 215 |
+
0.14707064650952817,
|
| 216 |
+
0.15523685581982136,
|
| 217 |
+
0.14308214232325553,
|
| 218 |
+
0.14392880015075207,
|
| 219 |
+
0.13569198094308377,
|
| 220 |
+
0.13905953854322434,
|
| 221 |
+
0.13985855013132095,
|
| 222 |
+
0.13739008083939552,
|
| 223 |
+
0.14033469870686532,
|
| 224 |
+
0.14292256250977517,
|
| 225 |
+
0.13980830140411854,
|
| 226 |
+
0.1357831320166588,
|
| 227 |
+
0.13591801874339582,
|
| 228 |
+
0.13607669658958912
|
| 229 |
+
],
|
| 230 |
+
"lr": [
|
| 231 |
+
5e-05,
|
| 232 |
+
0.0001,
|
| 233 |
+
9.98292246503335e-05,
|
| 234 |
+
9.931806517013612e-05,
|
| 235 |
+
9.847001329696653e-05,
|
| 236 |
+
9.729086208503174e-05,
|
| 237 |
+
9.578866633275288e-05,
|
| 238 |
+
9.397368756032445e-05,
|
| 239 |
+
9.185832391312644e-05,
|
| 240 |
+
8.945702546981969e-05,
|
| 241 |
+
8.678619553365659e-05,
|
| 242 |
+
8.386407858128706e-05,
|
| 243 |
+
8.07106356344834e-05,
|
| 244 |
+
7.734740790612136e-05,
|
| 245 |
+
7.379736965185368e-05,
|
| 246 |
+
7.008477123264848e-05,
|
| 247 |
+
6.623497346023418e-05,
|
| 248 |
+
6.227427435703997e-05,
|
| 249 |
+
5.8229729514036705e-05,
|
| 250 |
+
5.4128967273616625e-05,
|
| 251 |
+
5e-05,
|
| 252 |
+
4.5871032726383386e-05,
|
| 253 |
+
4.17702704859633e-05,
|
| 254 |
+
3.772572564296005e-05,
|
| 255 |
+
3.3765026539765834e-05,
|
| 256 |
+
2.991522876735154e-05,
|
| 257 |
+
2.6202630348146324e-05,
|
| 258 |
+
2.2652592093878666e-05,
|
| 259 |
+
1.928936436551661e-05,
|
| 260 |
+
1.6135921418712956e-05,
|
| 261 |
+
1.3213804466343421e-05,
|
| 262 |
+
1.0542974530180327e-05,
|
| 263 |
+
8.141676086873572e-06,
|
| 264 |
+
6.026312439675552e-06,
|
| 265 |
+
4.2113336672471245e-06,
|
| 266 |
+
2.7091379149682685e-06,
|
| 267 |
+
1.5299867030334814e-06,
|
| 268 |
+
6.819348298638839e-07,
|
| 269 |
+
1.7077534966650766e-07,
|
| 270 |
+
0.0
|
| 271 |
+
]
|
| 272 |
+
}
|
| 273 |
+
}
|