End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.98,
|
| 4 |
+
"eval_loss": 0.06010868772864342,
|
| 5 |
+
"eval_runtime": 16.9911,
|
| 6 |
+
"eval_samples_per_second": 158.907,
|
| 7 |
+
"eval_steps_per_second": 5.003,
|
| 8 |
+
"total_flos": 1.8124066505760768e+18,
|
| 9 |
+
"train_loss": 0.34462205903571946,
|
| 10 |
+
"train_runtime": 1109.7931,
|
| 11 |
+
"train_samples_per_second": 65.688,
|
| 12 |
+
"train_steps_per_second": 0.514
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.98,
|
| 4 |
+
"eval_loss": 0.06010868772864342,
|
| 5 |
+
"eval_runtime": 16.9911,
|
| 6 |
+
"eval_samples_per_second": 158.907,
|
| 7 |
+
"eval_steps_per_second": 5.003
|
| 8 |
+
}
|
runs/Apr18_08-15-46_d2b6f2f31cc3/events.out.tfevents.1713429293.d2b6f2f31cc3.2876.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a947ab4930cd21222d9849b364f175b600d0d29f1b155f7175c95424e114104f
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.8124066505760768e+18,
|
| 4 |
+
"train_loss": 0.34462205903571946,
|
| 5 |
+
"train_runtime": 1109.7931,
|
| 6 |
+
"train_samples_per_second": 65.688,
|
| 7 |
+
"train_steps_per_second": 0.514
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,456 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.98,
|
| 3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-570",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 570,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05,
|
| 13 |
+
"grad_norm": 5.6030073165893555,
|
| 14 |
+
"learning_rate": 8.771929824561403e-06,
|
| 15 |
+
"loss": 2.3227,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.11,
|
| 20 |
+
"grad_norm": 5.942467212677002,
|
| 21 |
+
"learning_rate": 1.7543859649122806e-05,
|
| 22 |
+
"loss": 2.114,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.16,
|
| 27 |
+
"grad_norm": 9.006362915039062,
|
| 28 |
+
"learning_rate": 2.6315789473684212e-05,
|
| 29 |
+
"loss": 1.7361,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.21,
|
| 34 |
+
"grad_norm": 14.675954818725586,
|
| 35 |
+
"learning_rate": 3.508771929824561e-05,
|
| 36 |
+
"loss": 1.1337,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.26,
|
| 41 |
+
"grad_norm": 13.018393516540527,
|
| 42 |
+
"learning_rate": 4.3859649122807014e-05,
|
| 43 |
+
"loss": 0.6849,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.32,
|
| 48 |
+
"grad_norm": 29.24870491027832,
|
| 49 |
+
"learning_rate": 4.970760233918128e-05,
|
| 50 |
+
"loss": 0.4908,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.37,
|
| 55 |
+
"grad_norm": 17.498170852661133,
|
| 56 |
+
"learning_rate": 4.8732943469785574e-05,
|
| 57 |
+
"loss": 0.4472,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.42,
|
| 62 |
+
"grad_norm": 30.422813415527344,
|
| 63 |
+
"learning_rate": 4.7758284600389865e-05,
|
| 64 |
+
"loss": 0.4151,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.47,
|
| 69 |
+
"grad_norm": 15.24944019317627,
|
| 70 |
+
"learning_rate": 4.678362573099415e-05,
|
| 71 |
+
"loss": 0.3611,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.53,
|
| 76 |
+
"grad_norm": 18.759361267089844,
|
| 77 |
+
"learning_rate": 4.580896686159844e-05,
|
| 78 |
+
"loss": 0.3719,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.58,
|
| 83 |
+
"grad_norm": 14.161396026611328,
|
| 84 |
+
"learning_rate": 4.483430799220273e-05,
|
| 85 |
+
"loss": 0.3395,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.63,
|
| 90 |
+
"grad_norm": 9.216828346252441,
|
| 91 |
+
"learning_rate": 4.3859649122807014e-05,
|
| 92 |
+
"loss": 0.3214,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.68,
|
| 97 |
+
"grad_norm": 10.63085651397705,
|
| 98 |
+
"learning_rate": 4.2884990253411305e-05,
|
| 99 |
+
"loss": 0.327,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.74,
|
| 104 |
+
"grad_norm": 14.545869827270508,
|
| 105 |
+
"learning_rate": 4.1910331384015596e-05,
|
| 106 |
+
"loss": 0.3075,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.79,
|
| 111 |
+
"grad_norm": 14.948440551757812,
|
| 112 |
+
"learning_rate": 4.093567251461988e-05,
|
| 113 |
+
"loss": 0.2626,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.84,
|
| 118 |
+
"grad_norm": 16.263952255249023,
|
| 119 |
+
"learning_rate": 3.996101364522417e-05,
|
| 120 |
+
"loss": 0.2731,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.89,
|
| 125 |
+
"grad_norm": 8.593974113464355,
|
| 126 |
+
"learning_rate": 3.898635477582846e-05,
|
| 127 |
+
"loss": 0.2553,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.95,
|
| 132 |
+
"grad_norm": 11.096720695495605,
|
| 133 |
+
"learning_rate": 3.8011695906432746e-05,
|
| 134 |
+
"loss": 0.2696,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.0,
|
| 139 |
+
"grad_norm": 9.849555015563965,
|
| 140 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 141 |
+
"loss": 0.2994,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.0,
|
| 146 |
+
"eval_accuracy": 0.9603703703703703,
|
| 147 |
+
"eval_loss": 0.12342998385429382,
|
| 148 |
+
"eval_runtime": 17.2042,
|
| 149 |
+
"eval_samples_per_second": 156.938,
|
| 150 |
+
"eval_steps_per_second": 4.941,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.05,
|
| 155 |
+
"grad_norm": 11.277251243591309,
|
| 156 |
+
"learning_rate": 3.606237816764133e-05,
|
| 157 |
+
"loss": 0.2387,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.11,
|
| 162 |
+
"grad_norm": 8.723546028137207,
|
| 163 |
+
"learning_rate": 3.508771929824561e-05,
|
| 164 |
+
"loss": 0.2595,
|
| 165 |
+
"step": 210
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.16,
|
| 169 |
+
"grad_norm": 15.544955253601074,
|
| 170 |
+
"learning_rate": 3.41130604288499e-05,
|
| 171 |
+
"loss": 0.224,
|
| 172 |
+
"step": 220
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.21,
|
| 176 |
+
"grad_norm": 14.086812019348145,
|
| 177 |
+
"learning_rate": 3.313840155945419e-05,
|
| 178 |
+
"loss": 0.235,
|
| 179 |
+
"step": 230
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.26,
|
| 183 |
+
"grad_norm": 9.04621410369873,
|
| 184 |
+
"learning_rate": 3.216374269005848e-05,
|
| 185 |
+
"loss": 0.2281,
|
| 186 |
+
"step": 240
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.32,
|
| 190 |
+
"grad_norm": 11.80008602142334,
|
| 191 |
+
"learning_rate": 3.118908382066277e-05,
|
| 192 |
+
"loss": 0.2342,
|
| 193 |
+
"step": 250
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.37,
|
| 197 |
+
"grad_norm": 3.7134058475494385,
|
| 198 |
+
"learning_rate": 3.0214424951267055e-05,
|
| 199 |
+
"loss": 0.222,
|
| 200 |
+
"step": 260
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.42,
|
| 204 |
+
"grad_norm": 6.554859638214111,
|
| 205 |
+
"learning_rate": 2.9239766081871346e-05,
|
| 206 |
+
"loss": 0.2025,
|
| 207 |
+
"step": 270
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.47,
|
| 211 |
+
"grad_norm": 9.202827453613281,
|
| 212 |
+
"learning_rate": 2.8265107212475634e-05,
|
| 213 |
+
"loss": 0.1814,
|
| 214 |
+
"step": 280
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.53,
|
| 218 |
+
"grad_norm": 8.696596145629883,
|
| 219 |
+
"learning_rate": 2.729044834307992e-05,
|
| 220 |
+
"loss": 0.203,
|
| 221 |
+
"step": 290
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.58,
|
| 225 |
+
"grad_norm": 10.952228546142578,
|
| 226 |
+
"learning_rate": 2.6315789473684212e-05,
|
| 227 |
+
"loss": 0.1701,
|
| 228 |
+
"step": 300
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.63,
|
| 232 |
+
"grad_norm": 10.55245304107666,
|
| 233 |
+
"learning_rate": 2.53411306042885e-05,
|
| 234 |
+
"loss": 0.1732,
|
| 235 |
+
"step": 310
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.68,
|
| 239 |
+
"grad_norm": 12.811739921569824,
|
| 240 |
+
"learning_rate": 2.4366471734892787e-05,
|
| 241 |
+
"loss": 0.1906,
|
| 242 |
+
"step": 320
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.74,
|
| 246 |
+
"grad_norm": 9.685540199279785,
|
| 247 |
+
"learning_rate": 2.3391812865497074e-05,
|
| 248 |
+
"loss": 0.1975,
|
| 249 |
+
"step": 330
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.79,
|
| 253 |
+
"grad_norm": 8.672029495239258,
|
| 254 |
+
"learning_rate": 2.2417153996101365e-05,
|
| 255 |
+
"loss": 0.1704,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.84,
|
| 260 |
+
"grad_norm": 15.387055397033691,
|
| 261 |
+
"learning_rate": 2.1442495126705653e-05,
|
| 262 |
+
"loss": 0.1845,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.89,
|
| 267 |
+
"grad_norm": 9.411192893981934,
|
| 268 |
+
"learning_rate": 2.046783625730994e-05,
|
| 269 |
+
"loss": 0.1906,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.95,
|
| 274 |
+
"grad_norm": 10.436761856079102,
|
| 275 |
+
"learning_rate": 1.949317738791423e-05,
|
| 276 |
+
"loss": 0.1715,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.0,
|
| 281 |
+
"grad_norm": 12.051985740661621,
|
| 282 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 283 |
+
"loss": 0.1853,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.0,
|
| 288 |
+
"eval_accuracy": 0.9740740740740741,
|
| 289 |
+
"eval_loss": 0.07047658413648605,
|
| 290 |
+
"eval_runtime": 16.9051,
|
| 291 |
+
"eval_samples_per_second": 159.715,
|
| 292 |
+
"eval_steps_per_second": 5.028,
|
| 293 |
+
"step": 380
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 2.05,
|
| 297 |
+
"grad_norm": 7.126999855041504,
|
| 298 |
+
"learning_rate": 1.7543859649122806e-05,
|
| 299 |
+
"loss": 0.1961,
|
| 300 |
+
"step": 390
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 2.11,
|
| 304 |
+
"grad_norm": 9.795936584472656,
|
| 305 |
+
"learning_rate": 1.6569200779727097e-05,
|
| 306 |
+
"loss": 0.1833,
|
| 307 |
+
"step": 400
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.16,
|
| 311 |
+
"grad_norm": 8.326896667480469,
|
| 312 |
+
"learning_rate": 1.5594541910331384e-05,
|
| 313 |
+
"loss": 0.1612,
|
| 314 |
+
"step": 410
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.21,
|
| 318 |
+
"grad_norm": 11.762768745422363,
|
| 319 |
+
"learning_rate": 1.4619883040935673e-05,
|
| 320 |
+
"loss": 0.1729,
|
| 321 |
+
"step": 420
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.26,
|
| 325 |
+
"grad_norm": 7.951204299926758,
|
| 326 |
+
"learning_rate": 1.364522417153996e-05,
|
| 327 |
+
"loss": 0.1818,
|
| 328 |
+
"step": 430
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 2.32,
|
| 332 |
+
"grad_norm": 8.159992218017578,
|
| 333 |
+
"learning_rate": 1.267056530214425e-05,
|
| 334 |
+
"loss": 0.1607,
|
| 335 |
+
"step": 440
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 2.37,
|
| 339 |
+
"grad_norm": 11.383317947387695,
|
| 340 |
+
"learning_rate": 1.1695906432748537e-05,
|
| 341 |
+
"loss": 0.1588,
|
| 342 |
+
"step": 450
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 2.42,
|
| 346 |
+
"grad_norm": 4.525303363800049,
|
| 347 |
+
"learning_rate": 1.0721247563352826e-05,
|
| 348 |
+
"loss": 0.1509,
|
| 349 |
+
"step": 460
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 2.47,
|
| 353 |
+
"grad_norm": 5.270012855529785,
|
| 354 |
+
"learning_rate": 9.746588693957115e-06,
|
| 355 |
+
"loss": 0.157,
|
| 356 |
+
"step": 470
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.53,
|
| 360 |
+
"grad_norm": 6.473080158233643,
|
| 361 |
+
"learning_rate": 8.771929824561403e-06,
|
| 362 |
+
"loss": 0.1543,
|
| 363 |
+
"step": 480
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.58,
|
| 367 |
+
"grad_norm": 15.744648933410645,
|
| 368 |
+
"learning_rate": 7.797270955165692e-06,
|
| 369 |
+
"loss": 0.1291,
|
| 370 |
+
"step": 490
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 2.63,
|
| 374 |
+
"grad_norm": 9.5807523727417,
|
| 375 |
+
"learning_rate": 6.82261208576998e-06,
|
| 376 |
+
"loss": 0.1292,
|
| 377 |
+
"step": 500
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.68,
|
| 381 |
+
"grad_norm": 5.534010887145996,
|
| 382 |
+
"learning_rate": 5.8479532163742686e-06,
|
| 383 |
+
"loss": 0.1512,
|
| 384 |
+
"step": 510
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.74,
|
| 388 |
+
"grad_norm": 8.032577514648438,
|
| 389 |
+
"learning_rate": 4.873294346978558e-06,
|
| 390 |
+
"loss": 0.1347,
|
| 391 |
+
"step": 520
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.79,
|
| 395 |
+
"grad_norm": 24.584503173828125,
|
| 396 |
+
"learning_rate": 3.898635477582846e-06,
|
| 397 |
+
"loss": 0.1569,
|
| 398 |
+
"step": 530
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.84,
|
| 402 |
+
"grad_norm": 3.999915599822998,
|
| 403 |
+
"learning_rate": 2.9239766081871343e-06,
|
| 404 |
+
"loss": 0.1664,
|
| 405 |
+
"step": 540
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.89,
|
| 409 |
+
"grad_norm": 9.104424476623535,
|
| 410 |
+
"learning_rate": 1.949317738791423e-06,
|
| 411 |
+
"loss": 0.1938,
|
| 412 |
+
"step": 550
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.95,
|
| 416 |
+
"grad_norm": 8.999449729919434,
|
| 417 |
+
"learning_rate": 9.746588693957115e-07,
|
| 418 |
+
"loss": 0.1519,
|
| 419 |
+
"step": 560
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 3.0,
|
| 423 |
+
"grad_norm": 11.300539016723633,
|
| 424 |
+
"learning_rate": 0.0,
|
| 425 |
+
"loss": 0.158,
|
| 426 |
+
"step": 570
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 3.0,
|
| 430 |
+
"eval_accuracy": 0.98,
|
| 431 |
+
"eval_loss": 0.06010868772864342,
|
| 432 |
+
"eval_runtime": 16.8705,
|
| 433 |
+
"eval_samples_per_second": 160.043,
|
| 434 |
+
"eval_steps_per_second": 5.038,
|
| 435 |
+
"step": 570
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.0,
|
| 439 |
+
"step": 570,
|
| 440 |
+
"total_flos": 1.8124066505760768e+18,
|
| 441 |
+
"train_loss": 0.34462205903571946,
|
| 442 |
+
"train_runtime": 1109.7931,
|
| 443 |
+
"train_samples_per_second": 65.688,
|
| 444 |
+
"train_steps_per_second": 0.514
|
| 445 |
+
}
|
| 446 |
+
],
|
| 447 |
+
"logging_steps": 10,
|
| 448 |
+
"max_steps": 570,
|
| 449 |
+
"num_input_tokens_seen": 0,
|
| 450 |
+
"num_train_epochs": 3,
|
| 451 |
+
"save_steps": 500,
|
| 452 |
+
"total_flos": 1.8124066505760768e+18,
|
| 453 |
+
"train_batch_size": 32,
|
| 454 |
+
"trial_name": null,
|
| 455 |
+
"trial_params": null
|
| 456 |
+
}
|