--- license: apache-2.0 base_model: facebook/dinov2-large tags: - generated_from_trainer model-index: - name: drone-DinoVdeau-from-probs-large-2024_11_15-batch-size32_freeze_probs results: [] --- # drone-DinoVdeau-from-probs-large-2024_11_15-batch-size32_freeze_probs This model is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.4668 - Rmse: 0.1546 - Mae: 0.1143 - Kl Divergence: 0.3931 - Explained Variance: 0.4690 - Learning Rate: 0.0000 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.001 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 150 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Rmse | Mae | Kl Divergence | Explained Variance | Rate | |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:-------------:|:------------------:|:------:| | No log | 1.0 | 219 | 0.4855 | 0.1771 | 0.1364 | 0.3101 | 0.3433 | 0.001 | | No log | 2.0 | 438 | 0.4760 | 0.1688 | 0.1247 | 0.5077 | 0.3891 | 0.001 | | 0.5195 | 3.0 | 657 | 0.4777 | 0.1707 | 0.1230 | 0.7896 | 0.3848 | 0.001 | | 0.5195 | 4.0 | 876 | 0.4743 | 0.1672 | 0.1238 | 0.4932 | 0.4037 | 0.001 | | 0.4742 | 5.0 | 1095 | 0.4746 | 0.1669 | 0.1277 | 0.2901 | 0.4132 | 0.001 | | 0.4742 | 6.0 | 1314 | 0.4750 | 0.1674 | 0.1253 | 0.4399 | 0.4022 | 0.001 | | 0.4706 | 7.0 | 1533 | 0.4745 | 0.1671 | 0.1259 | 0.4868 | 0.4020 | 0.001 | | 0.4706 | 8.0 | 1752 | 0.4742 | 0.1672 | 0.1257 | 0.3241 | 0.4111 | 0.001 | | 0.4706 | 9.0 | 1971 | 0.4730 | 0.1658 | 0.1236 | 0.4560 | 0.4107 | 0.001 | | 0.4678 | 10.0 | 2190 | 0.4751 | 0.1679 | 0.1269 | 0.2141 | 0.4190 | 0.001 | | 0.4678 | 11.0 | 2409 | 0.4733 | 0.1663 | 0.1265 | 0.2530 | 0.4189 | 0.001 | | 0.4674 | 12.0 | 2628 | 0.4758 | 0.1684 | 0.1264 | 0.3966 | 0.4074 | 0.001 | | 0.4674 | 13.0 | 2847 | 0.4722 | 0.1650 | 0.1223 | 0.6055 | 0.4142 | 0.001 | | 0.4676 | 14.0 | 3066 | 0.4747 | 0.1666 | 0.1250 | 0.4203 | 0.4071 | 0.001 | | 0.4676 | 15.0 | 3285 | 0.4733 | 0.1662 | 0.1227 | 0.6553 | 0.4153 | 0.001 | | 0.4663 | 16.0 | 3504 | 0.4735 | 0.1656 | 0.1241 | 0.3576 | 0.4176 | 0.001 | | 0.4663 | 17.0 | 3723 | 0.4722 | 0.1643 | 0.1221 | 0.4545 | 0.4231 | 0.001 | | 0.4663 | 18.0 | 3942 | 0.4724 | 0.1647 | 0.1225 | 0.4902 | 0.4209 | 0.001 | | 0.4655 | 19.0 | 4161 | 0.4729 | 0.1650 | 0.1261 | 0.3158 | 0.4224 | 0.001 | | 0.4655 | 20.0 | 4380 | 0.4697 | 0.1623 | 0.1203 | 0.4574 | 0.4342 | 0.0001 | | 0.4635 | 21.0 | 4599 | 0.4689 | 0.1613 | 0.1197 | 0.4569 | 0.4383 | 0.0001 | | 0.4635 | 22.0 | 4818 | 0.4691 | 0.1617 | 0.1202 | 0.4535 | 0.4374 | 0.0001 | | 0.4615 | 23.0 | 5037 | 0.4691 | 0.1614 | 0.1210 | 0.2971 | 0.4442 | 0.0001 | | 0.4615 | 24.0 | 5256 | 0.4692 | 0.1616 | 0.1196 | 0.3916 | 0.4406 | 0.0001 | | 0.4615 | 25.0 | 5475 | 0.4677 | 0.1601 | 0.1181 | 0.4516 | 0.4465 | 0.0001 | | 0.4601 | 26.0 | 5694 | 0.4680 | 0.1605 | 0.1171 | 0.6089 | 0.4434 | 0.0001 | | 0.4601 | 27.0 | 5913 | 0.4675 | 0.1600 | 0.1182 | 0.4741 | 0.4461 | 0.0001 | | 0.4585 | 28.0 | 6132 | 0.4681 | 0.1606 | 0.1200 | 0.3356 | 0.4489 | 0.0001 | | 0.4585 | 29.0 | 6351 | 0.4678 | 0.1603 | 0.1181 | 0.4330 | 0.4460 | 0.0001 | | 0.4578 | 30.0 | 6570 | 0.4680 | 0.1602 | 0.1194 | 0.3160 | 0.4504 | 0.0001 | | 0.4578 | 31.0 | 6789 | 0.4677 | 0.1600 | 0.1179 | 0.4190 | 0.4468 | 0.0001 | | 0.4579 | 32.0 | 7008 | 0.4675 | 0.1598 | 0.1188 | 0.3706 | 0.4504 | 0.0001 | | 0.4579 | 33.0 | 7227 | 0.4671 | 0.1593 | 0.1181 | 0.3504 | 0.4546 | 0.0001 | | 0.4579 | 34.0 | 7446 | 0.4670 | 0.1594 | 0.1180 | 0.3881 | 0.4533 | 0.0001 | | 0.4569 | 35.0 | 7665 | 0.4663 | 0.1587 | 0.1166 | 0.4398 | 0.4556 | 0.0001 | | 0.4569 | 36.0 | 7884 | 0.4666 | 0.1587 | 0.1170 | 0.4382 | 0.4544 | 0.0001 | | 0.4572 | 37.0 | 8103 | 0.4658 | 0.1581 | 0.1163 | 0.4330 | 0.4594 | 0.0001 | | 0.4572 | 38.0 | 8322 | 0.4659 | 0.1583 | 0.1162 | 0.4878 | 0.4567 | 0.0001 | | 0.4572 | 39.0 | 8541 | 0.4670 | 0.1595 | 0.1178 | 0.3791 | 0.4552 | 0.0001 | | 0.4572 | 40.0 | 8760 | 0.4665 | 0.1588 | 0.1178 | 0.3889 | 0.4568 | 0.0001 | | 0.4572 | 41.0 | 8979 | 0.4666 | 0.1589 | 0.1184 | 0.3222 | 0.4591 | 0.0001 | | 0.4559 | 42.0 | 9198 | 0.4655 | 0.1579 | 0.1164 | 0.4262 | 0.4607 | 0.0001 | | 0.4559 | 43.0 | 9417 | 0.4656 | 0.1579 | 0.1162 | 0.4611 | 0.4603 | 0.0001 | | 0.4554 | 44.0 | 9636 | 0.4656 | 0.1580 | 0.1164 | 0.4586 | 0.4616 | 0.0001 | | 0.4554 | 45.0 | 9855 | 0.4660 | 0.1583 | 0.1158 | 0.4368 | 0.4597 | 0.0001 | | 0.4557 | 46.0 | 10074 | 0.4660 | 0.1582 | 0.1164 | 0.4118 | 0.4604 | 0.0001 | | 0.4557 | 47.0 | 10293 | 0.4652 | 0.1577 | 0.1154 | 0.5424 | 0.4614 | 0.0001 | | 0.4551 | 48.0 | 10512 | 0.4660 | 0.1586 | 0.1160 | 0.5251 | 0.4596 | 0.0001 | | 0.4551 | 49.0 | 10731 | 0.4660 | 0.1585 | 0.1161 | 0.5007 | 0.4572 | 0.0001 | | 0.4551 | 50.0 | 10950 | 0.4666 | 0.1586 | 0.1185 | 0.2424 | 0.4659 | 0.0001 | | 0.4545 | 51.0 | 11169 | 0.4661 | 0.1584 | 0.1162 | 0.4171 | 0.4589 | 0.0001 | | 0.4545 | 52.0 | 11388 | 0.4650 | 0.1575 | 0.1155 | 0.4912 | 0.4630 | 0.0001 | | 0.4548 | 53.0 | 11607 | 0.4654 | 0.1578 | 0.1169 | 0.4030 | 0.4644 | 0.0001 | | 0.4548 | 54.0 | 11826 | 0.4661 | 0.1585 | 0.1153 | 0.4811 | 0.4595 | 0.0001 | | 0.455 | 55.0 | 12045 | 0.4653 | 0.1576 | 0.1167 | 0.3774 | 0.4638 | 0.0001 | | 0.455 | 56.0 | 12264 | 0.4654 | 0.1575 | 0.1176 | 0.3254 | 0.4670 | 0.0001 | | 0.455 | 57.0 | 12483 | 0.4654 | 0.1575 | 0.1162 | 0.3649 | 0.4662 | 0.0001 | | 0.4531 | 58.0 | 12702 | 0.4665 | 0.1584 | 0.1166 | 0.4075 | 0.4607 | 0.0001 | | 0.4531 | 59.0 | 12921 | 0.4652 | 0.1575 | 0.1157 | 0.4202 | 0.4654 | 1e-05 | | 0.4538 | 60.0 | 13140 | 0.4653 | 0.1571 | 0.1157 | 0.4084 | 0.4669 | 1e-05 | | 0.4538 | 61.0 | 13359 | 0.4654 | 0.1573 | 0.1153 | 0.4497 | 0.4661 | 1e-05 | | 0.4529 | 62.0 | 13578 | 0.4648 | 0.1568 | 0.1153 | 0.4112 | 0.4682 | 1e-05 | | 0.4529 | 63.0 | 13797 | 0.4648 | 0.1567 | 0.1152 | 0.3748 | 0.4702 | 1e-05 | | 0.4527 | 64.0 | 14016 | 0.4652 | 0.1571 | 0.1162 | 0.3044 | 0.4721 | 1e-05 | | 0.4527 | 65.0 | 14235 | 0.4648 | 0.1569 | 0.1153 | 0.4685 | 0.4670 | 1e-05 | | 0.4527 | 66.0 | 14454 | 0.4650 | 0.1573 | 0.1148 | 0.5087 | 0.4671 | 1e-05 | | 0.4531 | 67.0 | 14673 | 0.4646 | 0.1568 | 0.1155 | 0.4274 | 0.4690 | 1e-05 | | 0.4531 | 68.0 | 14892 | 0.4646 | 0.1566 | 0.1144 | 0.4969 | 0.4680 | 1e-05 | | 0.452 | 69.0 | 15111 | 0.4644 | 0.1564 | 0.1145 | 0.4480 | 0.4696 | 1e-05 | | 0.452 | 70.0 | 15330 | 0.4648 | 0.1567 | 0.1150 | 0.4291 | 0.4692 | 1e-05 | | 0.4524 | 71.0 | 15549 | 0.4645 | 0.1565 | 0.1156 | 0.3797 | 0.4711 | 1e-05 | | 0.4524 | 72.0 | 15768 | 0.4647 | 0.1569 | 0.1150 | 0.4280 | 0.4690 | 1e-05 | | 0.4524 | 73.0 | 15987 | 0.4641 | 0.1563 | 0.1142 | 0.4592 | 0.4707 | 1e-05 | | 0.4515 | 74.0 | 16206 | 0.4642 | 0.1564 | 0.1151 | 0.4321 | 0.4706 | 1e-05 | | 0.4515 | 75.0 | 16425 | 0.4645 | 0.1565 | 0.1152 | 0.3843 | 0.4708 | 1e-05 | | 0.4521 | 76.0 | 16644 | 0.4646 | 0.1569 | 0.1147 | 0.5216 | 0.4675 | 1e-05 | | 0.4521 | 77.0 | 16863 | 0.4648 | 0.1569 | 0.1152 | 0.4094 | 0.4691 | 1e-05 | | 0.4519 | 78.0 | 17082 | 0.4643 | 0.1564 | 0.1149 | 0.4399 | 0.4709 | 1e-05 | | 0.4519 | 79.0 | 17301 | 0.4646 | 0.1567 | 0.1147 | 0.4178 | 0.4697 | 1e-05 | | 0.4517 | 80.0 | 17520 | 0.4644 | 0.1564 | 0.1150 | 0.4373 | 0.4700 | 0.0000 | | 0.4517 | 81.0 | 17739 | 0.4645 | 0.1567 | 0.1151 | 0.4701 | 0.4688 | 0.0000 | | 0.4517 | 82.0 | 17958 | 0.4644 | 0.1565 | 0.1146 | 0.4601 | 0.4703 | 0.0000 | | 0.4514 | 83.0 | 18177 | 0.4646 | 0.1567 | 0.1147 | 0.4511 | 0.4684 | 0.0000 | ### Framework versions - Transformers 4.41.0 - Pytorch 2.5.0+cu124 - Datasets 3.0.2 - Tokenizers 0.19.1