File size: 5,346 Bytes
36c50e4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 |
---
license: other
base_model: nvidia/mit-b0
tags:
- vision
- image-segmentation
- generated_from_trainer
model-index:
- name: test_os_counties
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# test_os_counties
This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the rwood-97/os_counties dataset.
It achieves the following results on the evaluation set:
- Loss: 1.8997
- Mean Iou: 0.1075
- Mean Accuracy: 0.4992
- Overall Accuracy: 0.2118
- Accuracy Non-map: 0.9899
- Accuracy Map: 0.0084
- Iou Non-map: 0.2065
- Iou Map: 0.0084
## 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: 6e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Accuracy Non-map | Accuracy Map | Iou Non-map | Iou Map |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:-------------:|:----------------:|:----------------:|:------------:|:-----------:|:-------:|
| 0.4081 | 0.41 | 20 | 0.8267 | 0.2101 | 0.4981 | 0.3478 | 0.7548 | 0.2414 | 0.1934 | 0.2268 |
| 0.3939 | 0.82 | 40 | 0.8190 | 0.2387 | 0.4978 | 0.3898 | 0.6821 | 0.3134 | 0.1881 | 0.2894 |
| 0.3319 | 1.22 | 60 | 0.9487 | 0.1579 | 0.4976 | 0.2759 | 0.8762 | 0.1191 | 0.2005 | 0.1153 |
| 0.3234 | 1.63 | 80 | 0.9275 | 0.1847 | 0.4974 | 0.3120 | 0.8138 | 0.1809 | 0.1969 | 0.1725 |
| 0.3932 | 2.04 | 100 | 0.9053 | 0.1892 | 0.4974 | 0.3183 | 0.8031 | 0.1916 | 0.1962 | 0.1822 |
| 0.3412 | 2.45 | 120 | 0.8265 | 0.2385 | 0.4973 | 0.3895 | 0.6814 | 0.3132 | 0.1878 | 0.2891 |
| 0.2976 | 2.86 | 140 | 1.3713 | 0.1127 | 0.4987 | 0.2182 | 0.9778 | 0.0197 | 0.2058 | 0.0195 |
| 0.2803 | 3.27 | 160 | 1.6436 | 0.1095 | 0.4993 | 0.2143 | 0.9860 | 0.0126 | 0.2064 | 0.0125 |
| 0.3686 | 3.67 | 180 | 1.2379 | 0.1221 | 0.4982 | 0.2298 | 0.9564 | 0.0399 | 0.2047 | 0.0395 |
| 0.3434 | 4.08 | 200 | 1.1857 | 0.1385 | 0.4978 | 0.2507 | 0.9197 | 0.0758 | 0.2028 | 0.0743 |
| 0.2951 | 4.49 | 220 | 1.3947 | 0.1160 | 0.4986 | 0.2223 | 0.9704 | 0.0268 | 0.2054 | 0.0266 |
| 0.2333 | 4.9 | 240 | 1.5480 | 0.1170 | 0.4988 | 0.2236 | 0.9687 | 0.0288 | 0.2054 | 0.0286 |
| 0.2491 | 5.31 | 260 | 1.6563 | 0.1136 | 0.4990 | 0.2194 | 0.9764 | 0.0215 | 0.2058 | 0.0214 |
| 0.2706 | 5.71 | 280 | 1.9766 | 0.1058 | 0.4995 | 0.2098 | 0.9942 | 0.0048 | 0.2068 | 0.0048 |
| 0.2171 | 6.12 | 300 | 1.6191 | 0.1117 | 0.4989 | 0.2170 | 0.9804 | 0.0175 | 0.2060 | 0.0174 |
| 0.2352 | 6.53 | 320 | 1.8075 | 0.1102 | 0.4992 | 0.2151 | 0.9842 | 0.0141 | 0.2062 | 0.0141 |
| 0.2953 | 6.94 | 340 | 1.4709 | 0.1178 | 0.4986 | 0.2245 | 0.9667 | 0.0305 | 0.2053 | 0.0303 |
| 0.2799 | 7.35 | 360 | 1.3843 | 0.1300 | 0.4982 | 0.2399 | 0.9392 | 0.0571 | 0.2038 | 0.0562 |
| 0.2285 | 7.76 | 380 | 1.7799 | 0.1101 | 0.4991 | 0.2150 | 0.9842 | 0.0140 | 0.2062 | 0.0139 |
| 0.3461 | 8.16 | 400 | 1.7282 | 0.1106 | 0.4989 | 0.2157 | 0.9826 | 0.0153 | 0.2061 | 0.0152 |
| 0.1867 | 8.57 | 420 | 2.3356 | 0.1042 | 0.4997 | 0.2079 | 0.9981 | 0.0014 | 0.2070 | 0.0014 |
| 0.1731 | 8.98 | 440 | 2.1465 | 0.1050 | 0.4996 | 0.2089 | 0.9959 | 0.0032 | 0.2069 | 0.0032 |
| 0.224 | 9.39 | 460 | 2.3467 | 0.1047 | 0.4996 | 0.2084 | 0.9969 | 0.0024 | 0.2070 | 0.0024 |
| 0.2199 | 9.8 | 480 | 1.8997 | 0.1075 | 0.4992 | 0.2118 | 0.9899 | 0.0084 | 0.2065 | 0.0084 |
### Framework versions
- Transformers 4.35.0.dev0
- Pytorch 2.1.0
- Datasets 2.14.5
- Tokenizers 0.14.1
|