File size: 6,535 Bytes
5cf3a50 |
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 |
---
license: other
base_model: nvidia/mit-b0
tags:
- generated_from_trainer
model-index:
- name: Deepglobe_segformer_3_400
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. -->
# Deepglobe_segformer_3_400
This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7188
- Mean Iou: 0.4787
- Mean Accuracy: 0.6002
- Overall Accuracy: 0.8270
- Per Category Iou: [0.6299517395035389, 0.8579965667549715, 0.05609795827086041, 0.7364800207464396, 0.535316077547452, 0.5350211742530107, 0.0]
- Per Category Accuracy: [0.9063438304928855, 0.918655830242053, 0.06293362989838948, 0.8511005897416815, 0.6186843655274462, 0.8435526450990776, 0.0]
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Per Category Iou | Per Category Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:-------------:|:----------------:|:---------------------------------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------------------------------------------------------:|
| 1.4953 | 0.25 | 20 | 1.6438 | 0.3034 | 0.4362 | 0.7295 | [0.4372228066060852, 0.7851195516086454, 0.02767277670705226, 0.5318149141289378, 0.15332072216110113, 0.18882192890212937, 0.0] | [0.9054689287550887, 0.881972966540296, 0.030212967162335593, 0.855402152679511, 0.1546016559161411, 0.22575227044586893, 0.0] |
| 1.1619 | 0.49 | 40 | 1.1145 | 0.3522 | 0.4979 | 0.7453 | [0.4963248654368396, 0.7969307673775792, 0.005234675057991406, 0.5334928148352919, 0.40576267255758364, 0.22735397960522108, 0.0] | [0.9263632677194786, 0.8664759596406031, 0.005313323233395412, 0.9408065580353865, 0.4356812098052455, 0.31082581787427777, 0.0] |
| 1.1227 | 0.74 | 60 | 0.8344 | 0.4328 | 0.5514 | 0.8012 | [0.559377748191037, 0.833094745127801, 0.01617350405138789, 0.6810331455551815, 0.5222264810642535, 0.4177578560386505, 0.0] | [0.9190514270956751, 0.9199089912495539, 0.016740285242832744, 0.8991225550844304, 0.5349978690495003, 0.5701733759099568, 0.0] |
| 0.9186 | 0.99 | 80 | 0.8410 | 0.4315 | 0.5671 | 0.7981 | [0.610435962831109, 0.8248261041118935, 0.01287044797979473, 0.590317136673045, 0.5094678255258223, 0.47235101846741195, 0.0] | [0.8831700349552921, 0.8970489608062372, 0.013270642954328338, 0.9399664246329911, 0.5645757812288003, 0.6713455504003415, 0.0] |
| 1.3163 | 1.23 | 100 | 0.7876 | 0.4340 | 0.5585 | 0.8040 | [0.5512733368835496, 0.8404873242788062, 0.013928311212001072, 0.6685287371808809, 0.5155178151904813, 0.44798355714772914, 0.0] | [0.9061603321998727, 0.9229407158322913, 0.014365271974602195, 0.8723671532813903, 0.5859858479347339, 0.6073823083063936, 0.0] |
| 1.0072 | 1.48 | 120 | 0.7143 | 0.4654 | 0.5803 | 0.8279 | [0.6290456933460526, 0.8548414682316915, 0.015319296135995817, 0.6898485050535011, 0.519412217733134, 0.5491789992522889, 0.0] | [0.8906250495558662, 0.9439846826240867, 0.01570476015863979, 0.8294264466203112, 0.6039401035306737, 0.7782847364385814, 0.0] |
| 1.0037 | 1.73 | 140 | 0.7418 | 0.4669 | 0.5881 | 0.8238 | [0.6374852209379004, 0.8516151187065506, 0.04539433499278664, 0.6791015352123421, 0.5125601555696692, 0.5419419009373415, 0.0] | [0.8948436979663406, 0.928989068607313, 0.0491690469068136, 0.7914770913802412, 0.6247819167700217, 0.8274075815898259, 0.0] |
| 0.8331 | 1.98 | 160 | 0.7708 | 0.4602 | 0.6010 | 0.8148 | [0.61646612920421, 0.8431945482344851, 0.029559832548671162, 0.6925368958522767, 0.5197854175542566, 0.519586455316693, 0.0] | [0.9117309779099771, 0.8958172398851775, 0.03212238615858189, 0.914917262843377, 0.6484953254767503, 0.8040762308622557, 0.0] |
| 0.8578 | 2.22 | 180 | 0.6801 | 0.4823 | 0.5954 | 0.8333 | [0.6398254047914631, 0.8574983124445581, 0.04405338937855245, 0.7279725707575311, 0.5436762768533776, 0.5633375351596654, 0.0] | [0.8780062570652506, 0.9352013135786893, 0.04717820922310891, 0.8867957701630768, 0.5987970904145536, 0.8214859465526372, 0.0] |
| 0.7435 | 2.47 | 200 | 0.6822 | 0.4814 | 0.5994 | 0.8322 | [0.6283252084456665, 0.8600981959587153, 0.04592991298814512, 0.7355128664373006, 0.5428628143906776, 0.5573630881468757, 0.0] | [0.9001945988070346, 0.9301202539919514, 0.04953874805801353, 0.8820072095633354, 0.6198719514239059, 0.8137765082953279, 0.0] |
| 0.8242 | 2.72 | 220 | 0.6634 | 0.4859 | 0.5989 | 0.8355 | [0.627805454883537, 0.8613036148884382, 0.05734368620108552, 0.7490310967237723, 0.5390407061468642, 0.5668121927843631, 0.0] | [0.8979554665500735, 0.9377690650548433, 0.06191559475446536, 0.876048855605443, 0.6199629358272637, 0.7987612336682027, 0.0] |
| 1.0752 | 2.96 | 240 | 0.7188 | 0.4787 | 0.6002 | 0.8270 | [0.6299517395035389, 0.8579965667549715, 0.05609795827086041, 0.7364800207464396, 0.535316077547452, 0.5350211742530107, 0.0] | [0.9063438304928855, 0.918655830242053, 0.06293362989838948, 0.8511005897416815, 0.6186843655274462, 0.8435526450990776, 0.0] |
### Framework versions
- Transformers 4.31.0
- Pytorch 2.0.0+cu117
- Datasets 2.10.1
- Tokenizers 0.13.3
|