├── .gitignore ├── LICENSE ├── README.md ├── cifar10 ├── cifar-10.jpg └── log.txt ├── cifar100 ├── cifar-100.jpg └── log.txt ├── models ├── __init__.py └── model.py ├── test.py ├── train.py └── utils └── plot_log.py /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | .Python 11 | env/ 12 | build/ 13 | develop-eggs/ 14 | dist/ 15 | downloads/ 16 | eggs/ 17 | .eggs/ 18 | lib/ 19 | lib64/ 20 | parts/ 21 | sdist/ 22 | var/ 23 | *.egg-info/ 24 | .installed.cfg 25 | *.egg 26 | 27 | # PyInstaller 28 | # Usually these files are written by a python script from a template 29 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 30 | *.manifest 31 | *.spec 32 | 33 | # Installer logs 34 | pip-log.txt 35 | pip-delete-this-directory.txt 36 | 37 | # Unit test / coverage reports 38 | htmlcov/ 39 | .tox/ 40 | .coverage 41 | .coverage.* 42 | .cache 43 | nosetests.xml 44 | coverage.xml 45 | *,cover 46 | .hypothesis/ 47 | 48 | # Translations 49 | *.mo 50 | *.pot 51 | 52 | # Django stuff: 53 | *.log 54 | local_settings.py 55 | 56 | # Flask stuff: 57 | instance/ 58 | .webassets-cache 59 | 60 | # Scrapy stuff: 61 | .scrapy 62 | 63 | # Sphinx documentation 64 | docs/_build/ 65 | 66 | # PyBuilder 67 | target/ 68 | 69 | # IPython Notebook 70 | .ipynb_checkpoints 71 | 72 | # pyenv 73 | .python-version 74 | 75 | # celery beat schedule file 76 | celerybeat-schedule 77 | 78 | # dotenv 79 | .env 80 | 81 | # virtualenv 82 | venv/ 83 | ENV/ 84 | 85 | # Spyder project settings 86 | .spyderproject 87 | 88 | # Rope project settings 89 | .ropeproject 90 | 91 | # Pycharm project 92 | .idea 93 | snapshots 94 | logs 95 | *.pytorch 96 | *.tar.bz 97 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | MIT License 2 | 3 | Copyright (c) 2017 Pau Rodriguez 4 | 5 | Permission is hereby granted, free of charge, to any person obtaining a copy 6 | of this software and associated documentation files (the "Software"), to deal 7 | in the Software without restriction, including without limitation the rights 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 9 | copies of the Software, and to permit persons to whom the Software is 10 | furnished to do so, subject to the following conditions: 11 | 12 | The above copyright notice and this permission notice shall be included in all 13 | copies or substantial portions of the Software. 14 | 15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 21 | SOFTWARE. 22 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # ResNeXt.pytorch 2 | Reproduces ResNet-V3 (Aggregated Residual Transformations for Deep Neural Networks) with pytorch. 3 | 4 | - [x] Tried on pytorch 1.6 5 | - [x] Trains on Cifar10 and Cifar100 6 | - [x] Upload Cifar Training Curves 7 | - [x] Upload Cifar Trained Models 8 | - [x] Pytorch 0.4.0 9 | - [ ] Train Imagenet 10 | 11 | ## Download 12 | ```bash 13 | git clone https://github.com/prlz77/resnext.pytorch 14 | cd resnext.pytorch 15 | # git checkout R4.0 or R3.0 for backwards compatibility (not recommended). 16 | ``` 17 | 18 | ## Usage 19 | To train on Cifar-10 using 2 gpu: 20 | 21 | ```bash 22 | python train.py ~/DATASETS/cifar.python cifar10 -s ./snapshots --log ./logs --ngpu 2 --learning_rate 0.05 -b 128 23 | ``` 24 | It should reach *~3.65%* on Cifar-10, and *~17.77%* on Cifar-100. 25 | 26 | 27 | After train phase, you can check saved model. 28 | 29 | **Thanks to [@AppleHolic](https://github.com/AppleHolic) we have now a test script:** 30 | 31 | To test on Cifar-10 using 2 gpu: 32 | ```bash 33 | python test.py ~/DATASETS/cifar.python cifar10 --ngpu 2 --load ./snapshots/model.pytorch --test_bs 128 34 | ``` 35 | 36 | 37 | ## Configurations 38 | From [the original paper](https://arxiv.org/pdf/1611.05431.pdf): 39 | 40 | | cardinality | base_width | parameters | Error cifar10 | error cifar100 | default | 41 | |:-----------:|:------------:|:----------:|:----------------:|:-----------------:|:-------:| 42 | | 8 | 64 | 34.4M | 3.65 | 17.77 | x | 43 | | 16 | 64 | 68.1M | 3.58 | 17.31 | | 44 | 45 | **Update:** ``widen_factor`` has been disentangled from ``base_width`` because it was confusing. Now widen factor is set to consant 4, and ``base_width`` is the same as in the original paper. 46 | 47 | ## Trained models and curves 48 | [Link](https://mega.nz/#F!wbJXDS6b!YN3hCDi1tT3SdNFrLPm7mA) to trained models corresponding to the following curves: 49 | 50 | **Update:** several commits have been pushed after training the models in Mega, so it is recommended to revert to ``e10c37d8cf7a958048bc0f58cd86c3e8ac4e707d`` 51 | 52 | ![CIFAR-10](https://github.com/prlz77/ResNeXt.pytorch/blob/master/cifar10/cifar-10.jpg) 53 | ![CIFAR-100](https://github.com/prlz77/ResNeXt.pytorch/blob/master/cifar100/cifar-100.jpg) 54 | 55 | ## Other frameworks 56 | * [torch (@facebookresearch)](https://github.com/facebookresearch/ResNeXt). (Original) Cifar and Imagenet 57 | * [caffe (@terrychenism)](https://github.com/terrychenism/ResNeXt). Imagenet 58 | * [MXNet (@dmlc)](https://github.com/dmlc/mxnet/tree/master/example/image-classification#imagenet-1k). Imagenet 59 | 60 | ## Cite 61 | ``` 62 | @article{xie2016aggregated, 63 | title={Aggregated residual transformations for deep neural networks}, 64 | author={Xie, Saining and Girshick, Ross and Doll{\'a}r, Piotr and Tu, Zhuowen and He, Kaiming}, 65 | journal={arXiv preprint arXiv:1611.05431}, 66 | year={2016} 67 | } 68 | ``` 69 | -------------------------------------------------------------------------------- /cifar10/cifar-10.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prlz77/ResNeXt.pytorch/39fb8d03847f26ec02fb9b880ecaaa88db7a7d16/cifar10/cifar-10.jpg -------------------------------------------------------------------------------- /cifar10/log.txt: -------------------------------------------------------------------------------- 1 | {"load": null, "cardinality": 8, "momentum": 0.9, "depth": 29, "batch_size": 128, "decay": 0.0005, "epochs": 300, "dataset": "cifar10", "test_bs": 50, "test": false, "prefetch": 4, "log": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "gamma": 0.1, "save": "./cifar10/", "schedule": [150, 225], "data_path": "/data/cifar.pytorch/"} 2 | {"load": null, "momentum": 0.9, "train_loss": 1.2666670091125556, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.5306, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 0, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 1.2791472765803338, "schedule": [150, 225], "save": "./cifar10/"} 3 | {"load": null, "momentum": 0.9, "train_loss": 0.7998314224231581, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.6704, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 1, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.9537543568015099, "schedule": [150, 225], "save": "./cifar10/"} 4 | {"load": null, "momentum": 0.9, "train_loss": 0.7759610514354602, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.7549, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 2, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.7063407988846302, "schedule": [150, 225], "save": "./cifar10/"} 5 | {"load": null, "momentum": 0.9, "train_loss": 0.5206140929487751, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.7712, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 3, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.6797966969013214, "schedule": [150, 225], "save": "./cifar10/"} 6 | {"load": null, "momentum": 0.9, "train_loss": 0.4793380065969588, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.789, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 4, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.6454205292463303, "schedule": [150, 225], "save": "./cifar10/"} 7 | {"load": null, "momentum": 0.9, "train_loss": 0.45989600575731593, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8032, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 5, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.6044755820930005, "schedule": [150, 225], "save": "./cifar10/"} 8 | {"load": null, "momentum": 0.9, "train_loss": 0.5800966709263453, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8373, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 6, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4976392801105976, "schedule": [150, 225], "save": "./cifar10/"} 9 | {"load": null, "momentum": 0.9, "train_loss": 0.43991572069757756, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8299, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 7, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4972084932029247, "schedule": [150, 225], "save": "./cifar10/"} 10 | {"load": null, "momentum": 0.9, "train_loss": 0.27524289651866535, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.857, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 8, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.42173018455505373, "schedule": [150, 225], "save": "./cifar10/"} 11 | {"load": null, "momentum": 0.9, "train_loss": 0.29595639335035423, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8259, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 9, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5306424726545811, "schedule": [150, 225], "save": "./cifar10/"} 12 | {"load": null, "momentum": 0.9, "train_loss": 0.44155698522194065, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8338, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 10, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.49642379887402055, "schedule": [150, 225], "save": "./cifar10/"} 13 | {"load": null, "momentum": 0.9, "train_loss": 0.3596434793691874, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8593, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 11, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4156754682958126, "schedule": [150, 225], "save": "./cifar10/"} 14 | {"load": null, "momentum": 0.9, "train_loss": 0.2556023064560641, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8652, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 12, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.40670637518167496, "schedule": [150, 225], "save": "./cifar10/"} 15 | {"load": null, "momentum": 0.9, "train_loss": 0.4207796529027028, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8454, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 13, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4752872555702925, "schedule": [150, 225], "save": "./cifar10/"} 16 | {"load": null, "momentum": 0.9, "train_loss": 0.2165364084516546, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8371, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 14, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.474342071339488, "schedule": [150, 225], "save": "./cifar10/"} 17 | {"load": null, "momentum": 0.9, "train_loss": 0.2315336284126644, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8583, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 15, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4287001150846481, "schedule": [150, 225], "save": "./cifar10/"} 18 | {"load": null, "momentum": 0.9, "train_loss": 0.2646744834540136, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8514, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 16, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4465509482473135, "schedule": [150, 225], "save": "./cifar10/"} 19 | {"load": null, "momentum": 0.9, "train_loss": 0.2921852771212345, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8779, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 17, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.36642500828951596, "schedule": [150, 225], "save": "./cifar10/"} 20 | {"load": null, "momentum": 0.9, "train_loss": 0.21879412314593688, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8684, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 18, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.41280961643904446, "schedule": [150, 225], "save": "./cifar10/"} 21 | {"load": null, "momentum": 0.9, "train_loss": 0.19896230402866358, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8368, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 19, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.49646077245473863, "schedule": [150, 225], "save": "./cifar10/"} 22 | {"load": null, "momentum": 0.9, "train_loss": 0.2912619806888304, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8782, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 20, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.36526339072734115, "schedule": [150, 225], "save": "./cifar10/"} 23 | {"load": null, "momentum": 0.9, "train_loss": 0.2728017338612598, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8668, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 21, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4142205213010311, "schedule": [150, 225], "save": "./cifar10/"} 24 | {"load": null, "momentum": 0.9, "train_loss": 0.22708008964558135, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8775, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 22, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3698298180475831, "schedule": [150, 225], "save": "./cifar10/"} 25 | {"load": null, "momentum": 0.9, "train_loss": 0.40216984582135223, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8641, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 23, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4211005153506994, "schedule": [150, 225], "save": "./cifar10/"} 26 | {"load": null, "momentum": 0.9, "train_loss": 0.1972395472641819, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8625, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 24, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4168802173063159, "schedule": [150, 225], "save": "./cifar10/"} 27 | {"load": null, "momentum": 0.9, "train_loss": 0.2991537012041322, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8486, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 25, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5303729448094964, "schedule": [150, 225], "save": "./cifar10/"} 28 | {"load": null, "momentum": 0.9, "train_loss": 0.21409706345804275, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.875, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 26, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3843521739542484, "schedule": [150, 225], "save": "./cifar10/"} 29 | {"load": null, "momentum": 0.9, "train_loss": 0.23494790521916353, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8677, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 27, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39820175781846046, "schedule": [150, 225], "save": "./cifar10/"} 30 | {"load": null, "momentum": 0.9, "train_loss": 0.22066058193076613, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8291, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 28, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.554571478664875, "schedule": [150, 225], "save": "./cifar10/"} 31 | {"load": null, "momentum": 0.9, "train_loss": 0.28190577103158554, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8648, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 29, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.42084941297769546, "schedule": [150, 225], "save": "./cifar10/"} 32 | {"load": null, "momentum": 0.9, "train_loss": 0.2499579104641952, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.874, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 30, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39608553998172286, "schedule": [150, 225], "save": "./cifar10/"} 33 | {"load": null, "momentum": 0.9, "train_loss": 0.2592808899999084, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.868, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 31, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4072450201958418, "schedule": [150, 225], "save": "./cifar10/"} 34 | {"load": null, "momentum": 0.9, "train_loss": 0.3778306236505954, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8401, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 32, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5096460921317338, "schedule": [150, 225], "save": "./cifar10/"} 35 | {"load": null, "momentum": 0.9, "train_loss": 0.19016786456702922, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8752, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 33, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4082094248384237, "schedule": [150, 225], "save": "./cifar10/"} 36 | {"load": null, "momentum": 0.9, "train_loss": 0.2516471352422053, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8742, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 34, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3746392953023314, "schedule": [150, 225], "save": "./cifar10/"} 37 | {"load": null, "momentum": 0.9, "train_loss": 0.2210551840663877, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8873, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 35, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3437679402530193, "schedule": [150, 225], "save": "./cifar10/"} 38 | {"load": null, "momentum": 0.9, "train_loss": 0.22981512471829832, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8459, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 36, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.46623633295297623, "schedule": [150, 225], "save": "./cifar10/"} 39 | {"load": null, "momentum": 0.9, "train_loss": 0.2289894150583223, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8927, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 37, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.33894464664161206, "schedule": [150, 225], "save": "./cifar10/"} 40 | {"load": null, "momentum": 0.9, "train_loss": 0.20401070280418393, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8808, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 38, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.36693722419440744, "schedule": [150, 225], "save": "./cifar10/"} 41 | {"load": null, "momentum": 0.9, "train_loss": 0.17188088542332827, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8777, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 39, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.379287799410522, "schedule": [150, 225], "save": "./cifar10/"} 42 | {"load": null, "momentum": 0.9, "train_loss": 0.11472714774871087, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8849, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 40, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.37114339966326954, "schedule": [150, 225], "save": "./cifar10/"} 43 | {"load": null, "momentum": 0.9, "train_loss": 0.1966571691373571, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8742, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 41, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3835367640852928, "schedule": [150, 225], "save": "./cifar10/"} 44 | {"load": null, "momentum": 0.9, "train_loss": 0.3615269117110839, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8786, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 42, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3886564930528402, "schedule": [150, 225], "save": "./cifar10/"} 45 | {"load": null, "momentum": 0.9, "train_loss": 0.27012504874508103, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.886, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 43, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.35644353114068505, "schedule": [150, 225], "save": "./cifar10/"} 46 | {"load": null, "momentum": 0.9, "train_loss": 0.17787486864107788, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8828, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 44, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3673527769371867, "schedule": [150, 225], "save": "./cifar10/"} 47 | {"load": null, "momentum": 0.9, "train_loss": 0.32979997452645865, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8837, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 45, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.37880981547757986, "schedule": [150, 225], "save": "./cifar10/"} 48 | {"load": null, "momentum": 0.9, "train_loss": 0.19917852024518687, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8264, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 46, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5773442770540714, "schedule": [150, 225], "save": "./cifar10/"} 49 | {"load": null, "momentum": 0.9, "train_loss": 0.20504044231734309, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8297, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 47, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5951879897713661, "schedule": [150, 225], "save": "./cifar10/"} 50 | {"load": null, "momentum": 0.9, "train_loss": 0.22659261316752888, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8884, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 48, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.36136733580380676, "schedule": [150, 225], "save": "./cifar10/"} 51 | {"load": null, "momentum": 0.9, "train_loss": 0.17066084626225797, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8954, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 49, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3260503964871168, "schedule": [150, 225], "save": "./cifar10/"} 52 | {"load": null, "momentum": 0.9, "train_loss": 0.27511456771038856, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8904, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 50, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.35147213041782377, "schedule": [150, 225], "save": "./cifar10/"} 53 | {"load": null, "momentum": 0.9, "train_loss": 0.23779621978697357, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8752, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 51, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.374977396838367, "schedule": [150, 225], "save": "./cifar10/"} 54 | {"load": null, "momentum": 0.9, "train_loss": 0.12766075407255023, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8789, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 52, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.38650466602295636, "schedule": [150, 225], "save": "./cifar10/"} 55 | {"load": null, "momentum": 0.9, "train_loss": 0.22073069155049557, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8769, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 53, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.38949060562998056, "schedule": [150, 225], "save": "./cifar10/"} 56 | {"load": null, "momentum": 0.9, "train_loss": 0.20002711265082007, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8518, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 54, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4738762682676315, "schedule": [150, 225], "save": "./cifar10/"} 57 | {"load": null, "momentum": 0.9, "train_loss": 0.21177464693296574, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8629, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 55, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4314264675974846, "schedule": [150, 225], "save": "./cifar10/"} 58 | {"load": null, "momentum": 0.9, "train_loss": 0.17559030870630596, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8739, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 56, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.41352448862046004, "schedule": [150, 225], "save": "./cifar10/"} 59 | {"load": null, "momentum": 0.9, "train_loss": 0.21205829362439105, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9039, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 57, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.2963742396607995, "schedule": [150, 225], "save": "./cifar10/"} 60 | {"load": null, "momentum": 0.9, "train_loss": 0.19637857317639676, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8538, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 58, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.46538735754787924, "schedule": [150, 225], "save": "./cifar10/"} 61 | {"load": null, "momentum": 0.9, "train_loss": 0.1605801876502218, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8825, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 59, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.37965614508837464, "schedule": [150, 225], "save": "./cifar10/"} 62 | {"load": null, "momentum": 0.9, "train_loss": 0.18361111722665863, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8891, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 60, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.34815994147211315, "schedule": [150, 225], "save": "./cifar10/"} 63 | {"load": null, "momentum": 0.9, "train_loss": 0.15782545230862674, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8819, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 61, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.373977289609611, "schedule": [150, 225], "save": "./cifar10/"} 64 | {"load": null, "momentum": 0.9, "train_loss": 0.19167126447660363, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8953, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 62, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3413768706843257, "schedule": [150, 225], "save": "./cifar10/"} 65 | {"load": null, "momentum": 0.9, "train_loss": 0.17242943391102997, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8946, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 63, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3253706143051386, "schedule": [150, 225], "save": "./cifar10/"} 66 | {"load": null, "momentum": 0.9, "train_loss": 0.2646030789154399, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8915, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 64, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.34078452618792654, "schedule": [150, 225], "save": "./cifar10/"} 67 | {"load": null, "momentum": 0.9, "train_loss": 0.1229577376597983, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9032, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 65, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3051677060499787, "schedule": [150, 225], "save": "./cifar10/"} 68 | {"load": null, "momentum": 0.9, "train_loss": 0.22921766445931308, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8959, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 66, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3327574524283409, "schedule": [150, 225], "save": "./cifar10/"} 69 | {"load": null, "momentum": 0.9, "train_loss": 0.2872623653603659, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8749, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 67, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.41305938344448806, "schedule": [150, 225], "save": "./cifar10/"} 70 | {"load": null, "momentum": 0.9, "train_loss": 0.20919571615894406, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.874, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 68, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4068450893461704, "schedule": [150, 225], "save": "./cifar10/"} 71 | {"load": null, "momentum": 0.9, "train_loss": 0.1213507311717689, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8753, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 69, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.41584196120500566, "schedule": [150, 225], "save": "./cifar10/"} 72 | {"load": null, "momentum": 0.9, "train_loss": 0.1605948107481427, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8816, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 70, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.38609036203473807, "schedule": [150, 225], "save": "./cifar10/"} 73 | {"load": null, "momentum": 0.9, "train_loss": 0.18790415964013923, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8795, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 71, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.38773251257836816, "schedule": [150, 225], "save": "./cifar10/"} 74 | {"load": null, "momentum": 0.9, "train_loss": 0.30140772740513344, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8651, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 72, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.41989620208740236, "schedule": [150, 225], "save": "./cifar10/"} 75 | {"load": null, "momentum": 0.9, "train_loss": 0.16272695991180677, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8968, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 73, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.31620026532560586, "schedule": [150, 225], "save": "./cifar10/"} 76 | {"load": null, "momentum": 0.9, "train_loss": 0.17609269880735548, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8879, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 74, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.34660840656608344, "schedule": [150, 225], "save": "./cifar10/"} 77 | {"load": null, "momentum": 0.9, "train_loss": 0.20077687184124027, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8778, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 75, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.375783121176064, "schedule": [150, 225], "save": "./cifar10/"} 78 | {"load": null, "momentum": 0.9, "train_loss": 0.10699789545633576, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.88, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 76, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3886543399468064, "schedule": [150, 225], "save": "./cifar10/"} 79 | {"load": null, "momentum": 0.9, "train_loss": 0.2879120702888743, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8901, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 77, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3335260050371289, "schedule": [150, 225], "save": "./cifar10/"} 80 | {"load": null, "momentum": 0.9, "train_loss": 0.17149747830899442, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8921, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 78, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.33334875624626875, "schedule": [150, 225], "save": "./cifar10/"} 81 | {"load": null, "momentum": 0.9, "train_loss": 0.0875030072640969, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8933, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 79, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.36230363911017777, "schedule": [150, 225], "save": "./cifar10/"} 82 | {"load": null, "momentum": 0.9, "train_loss": 0.23330159977194, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9029, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 80, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3102672438882291, "schedule": [150, 225], "save": "./cifar10/"} 83 | {"load": null, "momentum": 0.9, "train_loss": 0.16558810308472122, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8842, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 81, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39870804842561486, "schedule": [150, 225], "save": "./cifar10/"} 84 | {"load": null, "momentum": 0.9, "train_loss": 0.16872125519836087, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9007, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 82, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.31337948990985753, "schedule": [150, 225], "save": "./cifar10/"} 85 | {"load": null, "momentum": 0.9, "train_loss": 0.11674132823550214, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.887, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 83, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.35478557251393794, "schedule": [150, 225], "save": "./cifar10/"} 86 | {"load": null, "momentum": 0.9, "train_loss": 0.13097501107315293, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8954, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 84, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.32751773415133356, "schedule": [150, 225], "save": "./cifar10/"} 87 | {"load": null, "momentum": 0.9, "train_loss": 0.1546819382193037, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8954, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 85, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3409549530595541, "schedule": [150, 225], "save": "./cifar10/"} 88 | {"load": null, "momentum": 0.9, "train_loss": 0.1751219203772023, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8751, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 86, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39500596454367043, "schedule": [150, 225], "save": "./cifar10/"} 89 | {"load": null, "momentum": 0.9, "train_loss": 0.16205258246268986, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8735, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 87, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.41318475794047116, "schedule": [150, 225], "save": "./cifar10/"} 90 | {"load": null, "momentum": 0.9, "train_loss": 0.19996387231726503, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8886, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 88, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3652574011310935, "schedule": [150, 225], "save": "./cifar10/"} 91 | {"load": null, "momentum": 0.9, "train_loss": 0.12354783199125331, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8895, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 89, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.36334238983690736, "schedule": [150, 225], "save": "./cifar10/"} 92 | {"load": null, "momentum": 0.9, "train_loss": 0.22534587743405518, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8934, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 90, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3288333482667804, "schedule": [150, 225], "save": "./cifar10/"} 93 | {"load": null, "momentum": 0.9, "train_loss": 0.18516297894897757, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8537, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 91, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.48783326491713525, "schedule": [150, 225], "save": "./cifar10/"} 94 | {"load": null, "momentum": 0.9, "train_loss": 0.11975512378535161, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8893, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 92, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3493700683489442, "schedule": [150, 225], "save": "./cifar10/"} 95 | {"load": null, "momentum": 0.9, "train_loss": 0.21151212265714386, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.874, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 93, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4096935661509633, "schedule": [150, 225], "save": "./cifar10/"} 96 | {"load": null, "momentum": 0.9, "train_loss": 0.1956150033594124, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8785, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 94, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.40291756015270946, "schedule": [150, 225], "save": "./cifar10/"} 97 | {"load": null, "momentum": 0.9, "train_loss": 0.17727864627898382, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8714, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 95, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39307315479964017, "schedule": [150, 225], "save": "./cifar10/"} 98 | {"load": null, "momentum": 0.9, "train_loss": 0.17511357328065574, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8959, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 96, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.324241233561188, "schedule": [150, 225], "save": "./cifar10/"} 99 | {"load": null, "momentum": 0.9, "train_loss": 0.18524342136369043, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8885, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 97, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3558444334194064, "schedule": [150, 225], "save": "./cifar10/"} 100 | {"load": null, "momentum": 0.9, "train_loss": 0.12977151436816675, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8942, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 98, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3282514745742083, "schedule": [150, 225], "save": "./cifar10/"} 101 | {"load": null, "momentum": 0.9, "train_loss": 0.21482057384202427, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8941, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 99, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3290054955333471, "schedule": [150, 225], "save": "./cifar10/"} 102 | {"load": null, "momentum": 0.9, "train_loss": 0.20226657761485006, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8918, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 100, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.34128159334883096, "schedule": [150, 225], "save": "./cifar10/"} 103 | {"load": null, "momentum": 0.9, "train_loss": 0.12182491786239395, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8544, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 101, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4559965577721596, "schedule": [150, 225], "save": "./cifar10/"} 104 | {"load": null, "momentum": 0.9, "train_loss": 0.11325694551649904, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8813, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 102, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3755914867669344, "schedule": [150, 225], "save": "./cifar10/"} 105 | {"load": null, "momentum": 0.9, "train_loss": 0.20330920508273267, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8712, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 103, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.40795027624815705, "schedule": [150, 225], "save": "./cifar10/"} 106 | {"load": null, "momentum": 0.9, "train_loss": 0.1651465525609925, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8995, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 104, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3222890492528677, "schedule": [150, 225], "save": "./cifar10/"} 107 | {"load": null, "momentum": 0.9, "train_loss": 0.15639510424446637, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.891, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 105, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3536194303818047, "schedule": [150, 225], "save": "./cifar10/"} 108 | {"load": null, "momentum": 0.9, "train_loss": 0.11294037399519624, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8879, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 106, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3791441973112524, "schedule": [150, 225], "save": "./cifar10/"} 109 | {"load": null, "momentum": 0.9, "train_loss": 0.08350167126959258, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8902, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 107, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3636630828678608, "schedule": [150, 225], "save": "./cifar10/"} 110 | {"load": null, "momentum": 0.9, "train_loss": 0.1880376473359208, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8945, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 108, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.33088157687336206, "schedule": [150, 225], "save": "./cifar10/"} 111 | {"load": null, "momentum": 0.9, "train_loss": 0.14607817520070865, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8862, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 109, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.34927321318536997, "schedule": [150, 225], "save": "./cifar10/"} 112 | {"load": null, "momentum": 0.9, "train_loss": 0.08664232911678793, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9121, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 110, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.26731032244861125, "schedule": [150, 225], "save": "./cifar10/"} 113 | {"load": null, "momentum": 0.9, "train_loss": 0.09206947153030105, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8978, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 111, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.31119134832173584, "schedule": [150, 225], "save": "./cifar10/"} 114 | {"load": null, "momentum": 0.9, "train_loss": 0.12669824027548152, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8952, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 112, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3335871702991426, "schedule": [150, 225], "save": "./cifar10/"} 115 | {"load": null, "momentum": 0.9, "train_loss": 0.16318103376092846, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8839, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 113, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3741954399831593, "schedule": [150, 225], "save": "./cifar10/"} 116 | {"load": null, "momentum": 0.9, "train_loss": 0.12285949677889885, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8941, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 114, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.33211559694260356, "schedule": [150, 225], "save": "./cifar10/"} 117 | {"load": null, "momentum": 0.9, "train_loss": 0.12395224630625425, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8919, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 115, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3387392172217369, "schedule": [150, 225], "save": "./cifar10/"} 118 | {"load": null, "momentum": 0.9, "train_loss": 0.13896647252614625, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8498, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 116, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5350771675631404, "schedule": [150, 225], "save": "./cifar10/"} 119 | {"load": null, "momentum": 0.9, "train_loss": 0.13315356903465786, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8783, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 117, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3985060391947627, "schedule": [150, 225], "save": "./cifar10/"} 120 | {"load": null, "momentum": 0.9, "train_loss": 0.08764401652462532, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8865, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 118, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3584533652663231, "schedule": [150, 225], "save": "./cifar10/"} 121 | {"load": null, "momentum": 0.9, "train_loss": 0.16211169789547847, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8869, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 119, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3807142147421837, "schedule": [150, 225], "save": "./cifar10/"} 122 | {"load": null, "momentum": 0.9, "train_loss": 0.1464889879307107, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8688, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 120, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.43609306652098895, "schedule": [150, 225], "save": "./cifar10/"} 123 | {"load": null, "momentum": 0.9, "train_loss": 0.09466987165577329, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8888, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 121, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.35028758240863683, "schedule": [150, 225], "save": "./cifar10/"} 124 | {"load": null, "momentum": 0.9, "train_loss": 0.06071041714727968, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9001, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 122, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.31731982685625554, "schedule": [150, 225], "save": "./cifar10/"} 125 | {"load": null, "momentum": 0.9, "train_loss": 0.215032362356139, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8966, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 123, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.33141786299645903, "schedule": [150, 225], "save": "./cifar10/"} 126 | {"load": null, "momentum": 0.9, "train_loss": 0.1461519446886909, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.87, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 124, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.42853931184858085, "schedule": [150, 225], "save": "./cifar10/"} 127 | {"load": null, "momentum": 0.9, "train_loss": 0.1898245200686138, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8988, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 125, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.31869767732918264, "schedule": [150, 225], "save": "./cifar10/"} 128 | {"load": null, "momentum": 0.9, "train_loss": 0.2843209813068225, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8767, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 126, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39254647087305783, "schedule": [150, 225], "save": "./cifar10/"} 129 | {"load": null, "momentum": 0.9, "train_loss": 0.09136279708310463, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8679, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 127, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4361115729063749, "schedule": [150, 225], "save": "./cifar10/"} 130 | {"load": null, "momentum": 0.9, "train_loss": 0.1223677845235346, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8873, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 128, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.36593699622899295, "schedule": [150, 225], "save": "./cifar10/"} 131 | {"load": null, "momentum": 0.9, "train_loss": 0.10681527204030061, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8844, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 129, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.383501938469708, "schedule": [150, 225], "save": "./cifar10/"} 132 | {"load": null, "momentum": 0.9, "train_loss": 0.21693851190628283, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8831, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 130, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39602261926978827, "schedule": [150, 225], "save": "./cifar10/"} 133 | {"load": null, "momentum": 0.9, "train_loss": 0.13469917760529757, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.873, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 131, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4127215235494077, "schedule": [150, 225], "save": "./cifar10/"} 134 | {"load": null, "momentum": 0.9, "train_loss": 0.1844398417695297, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8821, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 132, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3704837433807552, "schedule": [150, 225], "save": "./cifar10/"} 135 | {"load": null, "momentum": 0.9, "train_loss": 0.09085054452604914, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8687, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 133, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.43628038637340066, "schedule": [150, 225], "save": "./cifar10/"} 136 | {"load": null, "momentum": 0.9, "train_loss": 0.14528705694805488, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9039, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 134, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3003537464700639, "schedule": [150, 225], "save": "./cifar10/"} 137 | {"load": null, "momentum": 0.9, "train_loss": 0.10644292748176343, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8397, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 135, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5331770410388708, "schedule": [150, 225], "save": "./cifar10/"} 138 | {"load": null, "momentum": 0.9, "train_loss": 0.3540273318415417, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8222, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 136, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.6082278560847044, "schedule": [150, 225], "save": "./cifar10/"} 139 | {"load": null, "momentum": 0.9, "train_loss": 0.2668474341726477, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8958, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 137, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3309092673100531, "schedule": [150, 225], "save": "./cifar10/"} 140 | {"load": null, "momentum": 0.9, "train_loss": 0.12715305965322896, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8906, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 138, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3532950812764466, "schedule": [150, 225], "save": "./cifar10/"} 141 | {"load": null, "momentum": 0.9, "train_loss": 0.18391348553843478, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8972, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 139, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.32824699811637403, "schedule": [150, 225], "save": "./cifar10/"} 142 | {"load": null, "momentum": 0.9, "train_loss": 0.19643313917293903, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8648, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 140, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4465531235188246, "schedule": [150, 225], "save": "./cifar10/"} 143 | {"load": null, "momentum": 0.9, "train_loss": 0.21314023868929866, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8844, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 141, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3748804623261094, "schedule": [150, 225], "save": "./cifar10/"} 144 | {"load": null, "momentum": 0.9, "train_loss": 0.20913667694968374, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8836, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 142, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3760984083265066, "schedule": [150, 225], "save": "./cifar10/"} 145 | {"load": null, "momentum": 0.9, "train_loss": 0.10423301571818433, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.904, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 143, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.29856251328252253, "schedule": [150, 225], "save": "./cifar10/"} 146 | {"load": null, "momentum": 0.9, "train_loss": 0.19232574138924463, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8921, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 144, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3388327742740512, "schedule": [150, 225], "save": "./cifar10/"} 147 | {"load": null, "momentum": 0.9, "train_loss": 0.1836790052303032, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8791, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 145, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.39107946556061507, "schedule": [150, 225], "save": "./cifar10/"} 148 | {"load": null, "momentum": 0.9, "train_loss": 0.11596160222856008, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8962, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 146, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.3169965928792953, "schedule": [150, 225], "save": "./cifar10/"} 149 | {"load": null, "momentum": 0.9, "train_loss": 0.1945314468338686, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8414, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 147, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.5732587710022926, "schedule": [150, 225], "save": "./cifar10/"} 150 | {"load": null, "momentum": 0.9, "train_loss": 0.1528817942994435, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8729, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 148, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.4132155790552497, "schedule": [150, 225], "save": "./cifar10/"} 151 | {"load": null, "momentum": 0.9, "train_loss": 0.15255173532055044, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.8947, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 149, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.05, "ngpu": 2, "test_loss": 0.33253141488879917, "schedule": [150, 225], "save": "./cifar10/"} 152 | {"load": null, "momentum": 0.9, "train_loss": 0.03615028238232158, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9496, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 150, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.15096881515346466, "schedule": [150, 225], "save": "./cifar10/"} 153 | {"load": null, "momentum": 0.9, "train_loss": 0.04754832921215993, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9536, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 151, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.14550544496625661, "schedule": [150, 225], "save": "./cifar10/"} 154 | {"load": null, "momentum": 0.9, "train_loss": 0.014644483726451066, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9538, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 152, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.14037243416998535, "schedule": [150, 225], "save": "./cifar10/"} 155 | {"load": null, "momentum": 0.9, "train_loss": 0.019153290655001126, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9541, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 153, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1399338852101937, "schedule": [150, 225], "save": "./cifar10/"} 156 | {"load": null, "momentum": 0.9, "train_loss": 0.013610766746836948, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.955, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 154, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13913329060189425, "schedule": [150, 225], "save": "./cifar10/"} 157 | {"load": null, "momentum": 0.9, "train_loss": 0.00523394772492776, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9557, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 155, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13546354456804693, "schedule": [150, 225], "save": "./cifar10/"} 158 | {"load": null, "momentum": 0.9, "train_loss": 0.015223538035952278, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9553, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 156, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13702478352934122, "schedule": [150, 225], "save": "./cifar10/"} 159 | {"load": null, "momentum": 0.9, "train_loss": 0.013036407064227998, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9566, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 157, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13520003878511488, "schedule": [150, 225], "save": "./cifar10/"} 160 | {"load": null, "momentum": 0.9, "train_loss": 0.014989242585551076, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9593, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 158, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1334880225593224, "schedule": [150, 225], "save": "./cifar10/"} 161 | {"load": null, "momentum": 0.9, "train_loss": 0.016111858233063964, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9588, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 159, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13444617738947273, "schedule": [150, 225], "save": "./cifar10/"} 162 | {"load": null, "momentum": 0.9, "train_loss": 0.013112914013205378, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9573, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 160, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13519213111139833, "schedule": [150, 225], "save": "./cifar10/"} 163 | {"load": null, "momentum": 0.9, "train_loss": 0.028680534332089357, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9592, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 161, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13189091073349118, "schedule": [150, 225], "save": "./cifar10/"} 164 | {"load": null, "momentum": 0.9, "train_loss": 0.009748337941692388, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9576, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 162, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1345513264182955, "schedule": [150, 225], "save": "./cifar10/"} 165 | {"load": null, "momentum": 0.9, "train_loss": 0.005939394489199685, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9598, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 163, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13244503145571798, "schedule": [150, 225], "save": "./cifar10/"} 166 | {"load": null, "momentum": 0.9, "train_loss": 0.014099952845669247, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9589, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 164, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13102685233112424, "schedule": [150, 225], "save": "./cifar10/"} 167 | {"load": null, "momentum": 0.9, "train_loss": 0.02226689072145805, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9593, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 165, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13280007203342392, "schedule": [150, 225], "save": "./cifar10/"} 168 | {"load": null, "momentum": 0.9, "train_loss": 0.005650496289805818, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9592, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 166, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13114067380782216, "schedule": [150, 225], "save": "./cifar10/"} 169 | {"load": null, "momentum": 0.9, "train_loss": 0.009140805187832278, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9592, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 167, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1281498529110104, "schedule": [150, 225], "save": "./cifar10/"} 170 | {"load": null, "momentum": 0.9, "train_loss": 0.008319641855014326, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9588, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 168, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13096166118979455, "schedule": [150, 225], "save": "./cifar10/"} 171 | {"load": null, "momentum": 0.9, "train_loss": 0.009590349321707044, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9585, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 169, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13019899057922885, "schedule": [150, 225], "save": "./cifar10/"} 172 | {"load": null, "momentum": 0.9, "train_loss": 0.012690266400570747, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9593, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 170, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1292826535133645, "schedule": [150, 225], "save": "./cifar10/"} 173 | {"load": null, "momentum": 0.9, "train_loss": 0.012490255465796412, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9607, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 171, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12733239047694953, "schedule": [150, 225], "save": "./cifar10/"} 174 | {"load": null, "momentum": 0.9, "train_loss": 0.0061888432040455195, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9589, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 172, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12692215225892142, "schedule": [150, 225], "save": "./cifar10/"} 175 | {"load": null, "momentum": 0.9, "train_loss": 0.002748397854646512, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9599, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 173, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12869582238607108, "schedule": [150, 225], "save": "./cifar10/"} 176 | {"load": null, "momentum": 0.9, "train_loss": 0.00606872151022531, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9602, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 174, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12746965163387358, "schedule": [150, 225], "save": "./cifar10/"} 177 | {"load": null, "momentum": 0.9, "train_loss": 0.05771629247148167, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9594, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 175, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1274120309157297, "schedule": [150, 225], "save": "./cifar10/"} 178 | {"load": null, "momentum": 0.9, "train_loss": 0.0055489084793295665, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9591, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 176, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12859138254541905, "schedule": [150, 225], "save": "./cifar10/"} 179 | {"load": null, "momentum": 0.9, "train_loss": 0.03183253901597907, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9608, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 177, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1291424342733808, "schedule": [150, 225], "save": "./cifar10/"} 180 | {"load": null, "momentum": 0.9, "train_loss": 0.009518718907823246, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9594, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 178, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13206570075592025, "schedule": [150, 225], "save": "./cifar10/"} 181 | {"load": null, "momentum": 0.9, "train_loss": 0.007301705077455144, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9609, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 179, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12677892765961588, "schedule": [150, 225], "save": "./cifar10/"} 182 | {"load": null, "momentum": 0.9, "train_loss": 0.009780723313511309, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9598, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 180, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1262706537102349, "schedule": [150, 225], "save": "./cifar10/"} 183 | {"load": null, "momentum": 0.9, "train_loss": 0.005194970582333101, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9605, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 181, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12534124718979, "schedule": [150, 225], "save": "./cifar10/"} 184 | {"load": null, "momentum": 0.9, "train_loss": 0.01191642656628261, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9607, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 182, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12864306432195008, "schedule": [150, 225], "save": "./cifar10/"} 185 | {"load": null, "momentum": 0.9, "train_loss": 0.007081258101395598, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9606, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 183, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12733098710188642, "schedule": [150, 225], "save": "./cifar10/"} 186 | {"load": null, "momentum": 0.9, "train_loss": 0.004698297708369393, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.962, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 184, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12528964958153666, "schedule": [150, 225], "save": "./cifar10/"} 187 | {"load": null, "momentum": 0.9, "train_loss": 0.0037974580653539246, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9617, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 185, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12434362874832004, "schedule": [150, 225], "save": "./cifar10/"} 188 | {"load": null, "momentum": 0.9, "train_loss": 0.00508814463026649, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9613, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 186, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1280486347572878, "schedule": [150, 225], "save": "./cifar10/"} 189 | {"load": null, "momentum": 0.9, "train_loss": 0.0055566356741545584, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9613, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 187, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12783099338877946, "schedule": [150, 225], "save": "./cifar10/"} 190 | {"load": null, "momentum": 0.9, "train_loss": 0.005829708823591071, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9614, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 188, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12607156791724264, "schedule": [150, 225], "save": "./cifar10/"} 191 | {"load": null, "momentum": 0.9, "train_loss": 0.005424078633391567, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9607, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 189, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12653473261510953, "schedule": [150, 225], "save": "./cifar10/"} 192 | {"load": null, "momentum": 0.9, "train_loss": 0.006423735228571184, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9615, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 190, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12526730289449917, "schedule": [150, 225], "save": "./cifar10/"} 193 | {"load": null, "momentum": 0.9, "train_loss": 0.007047801747978482, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9606, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 191, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12636988142039626, "schedule": [150, 225], "save": "./cifar10/"} 194 | {"load": null, "momentum": 0.9, "train_loss": 0.0034334397490615445, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9602, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 192, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12502824079245328, "schedule": [150, 225], "save": "./cifar10/"} 195 | {"load": null, "momentum": 0.9, "train_loss": 0.004348056688438804, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9618, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 193, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12527566875098273, "schedule": [150, 225], "save": "./cifar10/"} 196 | {"load": null, "momentum": 0.9, "train_loss": 0.005825455701689032, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9616, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 194, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12647594089386985, "schedule": [150, 225], "save": "./cifar10/"} 197 | {"load": null, "momentum": 0.9, "train_loss": 0.0053144325608881346, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9616, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 195, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12610608375631271, "schedule": [150, 225], "save": "./cifar10/"} 198 | {"load": null, "momentum": 0.9, "train_loss": 0.0077305791158875665, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9624, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 196, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12497112267883495, "schedule": [150, 225], "save": "./cifar10/"} 199 | {"load": null, "momentum": 0.9, "train_loss": 0.009941826523715273, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9613, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 197, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12672264263499528, "schedule": [150, 225], "save": "./cifar10/"} 200 | {"load": null, "momentum": 0.9, "train_loss": 0.005525765230336292, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9621, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 198, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12292574040591717, "schedule": [150, 225], "save": "./cifar10/"} 201 | {"load": null, "momentum": 0.9, "train_loss": 0.00891955162918495, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9606, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 199, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12498664696933702, "schedule": [150, 225], "save": "./cifar10/"} 202 | {"load": null, "momentum": 0.9, "train_loss": 0.005222012781741492, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9619, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 200, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12430691993096843, "schedule": [150, 225], "save": "./cifar10/"} 203 | {"load": null, "momentum": 0.9, "train_loss": 0.004983152245434701, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9621, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 201, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12409566399408505, "schedule": [150, 225], "save": "./cifar10/"} 204 | {"load": null, "momentum": 0.9, "train_loss": 0.011417045470639816, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9612, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 202, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12638282892294228, "schedule": [150, 225], "save": "./cifar10/"} 205 | {"load": null, "momentum": 0.9, "train_loss": 0.002970298393662247, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9616, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 203, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12701996973715723, "schedule": [150, 225], "save": "./cifar10/"} 206 | {"load": null, "momentum": 0.9, "train_loss": 0.0051371753310886755, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9606, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 204, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12591514791245573, "schedule": [150, 225], "save": "./cifar10/"} 207 | {"load": null, "momentum": 0.9, "train_loss": 0.00912232278210594, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9628, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 205, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12534397126408295, "schedule": [150, 225], "save": "./cifar10/"} 208 | {"load": null, "momentum": 0.9, "train_loss": 0.0034667741090931735, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9614, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 206, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12168042631819845, "schedule": [150, 225], "save": "./cifar10/"} 209 | {"load": null, "momentum": 0.9, "train_loss": 0.004015523914029347, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9613, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 207, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1274924952117726, "schedule": [150, 225], "save": "./cifar10/"} 210 | {"load": null, "momentum": 0.9, "train_loss": 0.005132932530018102, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9621, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 208, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12663681366946547, "schedule": [150, 225], "save": "./cifar10/"} 211 | {"load": null, "momentum": 0.9, "train_loss": 0.002415187196056505, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9617, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 209, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12432439478230663, "schedule": [150, 225], "save": "./cifar10/"} 212 | {"load": null, "momentum": 0.9, "train_loss": 0.0032579527259303073, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 210, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12666592848603614, "schedule": [150, 225], "save": "./cifar10/"} 213 | {"load": null, "momentum": 0.9, "train_loss": 0.0037481523104012865, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9629, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 211, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12382897237082943, "schedule": [150, 225], "save": "./cifar10/"} 214 | {"load": null, "momentum": 0.9, "train_loss": 0.0045698584608146525, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 212, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12471316132927314, "schedule": [150, 225], "save": "./cifar10/"} 215 | {"load": null, "momentum": 0.9, "train_loss": 0.0024637399790384228, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.962, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 213, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12858842987800018, "schedule": [150, 225], "save": "./cifar10/"} 216 | {"load": null, "momentum": 0.9, "train_loss": 0.0029405974186479983, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9618, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 214, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12541410850011744, "schedule": [150, 225], "save": "./cifar10/"} 217 | {"load": null, "momentum": 0.9, "train_loss": 0.004107762471791486, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 215, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1284248086786829, "schedule": [150, 225], "save": "./cifar10/"} 218 | {"load": null, "momentum": 0.9, "train_loss": 0.011746349386009812, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9616, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 216, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12740027168416418, "schedule": [150, 225], "save": "./cifar10/"} 219 | {"load": null, "momentum": 0.9, "train_loss": 0.006235550944197301, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9616, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 217, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12868588511832058, "schedule": [150, 225], "save": "./cifar10/"} 220 | {"load": null, "momentum": 0.9, "train_loss": 0.007204421223142653, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9631, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 218, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12487805131124333, "schedule": [150, 225], "save": "./cifar10/"} 221 | {"load": null, "momentum": 0.9, "train_loss": 0.0035312326773148436, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9624, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 219, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.12519285651156678, "schedule": [150, 225], "save": "./cifar10/"} 222 | {"load": null, "momentum": 0.9, "train_loss": 0.003340229343831546, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9633, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 220, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1208706572582014, "schedule": [150, 225], "save": "./cifar10/"} 223 | {"load": null, "momentum": 0.9, "train_loss": 0.006139295299580685, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9576, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 221, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.141234974777326, "schedule": [150, 225], "save": "./cifar10/"} 224 | {"load": null, "momentum": 0.9, "train_loss": 0.0057536928468642145, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.962, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 222, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.1315669750282541, "schedule": [150, 225], "save": "./cifar10/"} 225 | {"load": null, "momentum": 0.9, "train_loss": 0.005694523696238231, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9599, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 223, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13175589784979821, "schedule": [150, 225], "save": "./cifar10/"} 226 | {"load": null, "momentum": 0.9, "train_loss": 0.004187146334117733, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9577, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 224, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.005000000000000001, "ngpu": 2, "test_loss": 0.13778194069629535, "schedule": [150, 225], "save": "./cifar10/"} 227 | {"load": null, "momentum": 0.9, "train_loss": 0.0041638850747166166, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9612, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 225, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12769160933326928, "schedule": [150, 225], "save": "./cifar10/"} 228 | {"load": null, "momentum": 0.9, "train_loss": 0.003111808750757101, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9609, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 226, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12653525736532173, "schedule": [150, 225], "save": "./cifar10/"} 229 | {"load": null, "momentum": 0.9, "train_loss": 0.0046697273611018275, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9624, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 227, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.124449753833469, "schedule": [150, 225], "save": "./cifar10/"} 230 | {"load": null, "momentum": 0.9, "train_loss": 0.0038161267743058534, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9621, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 228, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12488175525213592, "schedule": [150, 225], "save": "./cifar10/"} 231 | {"load": null, "momentum": 0.9, "train_loss": 0.006449660471131318, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9622, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 229, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12477041870472022, "schedule": [150, 225], "save": "./cifar10/"} 232 | {"load": null, "momentum": 0.9, "train_loss": 0.00351997003853412, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9623, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 230, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12592551701236515, "schedule": [150, 225], "save": "./cifar10/"} 233 | {"load": null, "momentum": 0.9, "train_loss": 0.003923068963360804, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9635, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 231, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1239228491880931, "schedule": [150, 225], "save": "./cifar10/"} 234 | {"load": null, "momentum": 0.9, "train_loss": 0.002545187494331411, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9616, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 232, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12380864925333299, "schedule": [150, 225], "save": "./cifar10/"} 235 | {"load": null, "momentum": 0.9, "train_loss": 0.005154310385009454, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9623, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 233, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12368534306413494, "schedule": [150, 225], "save": "./cifar10/"} 236 | {"load": null, "momentum": 0.9, "train_loss": 0.0034984258377751465, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9622, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 234, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12362852458958513, "schedule": [150, 225], "save": "./cifar10/"} 237 | {"load": null, "momentum": 0.9, "train_loss": 0.0036831278922873087, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9631, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 235, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12527450532652437, "schedule": [150, 225], "save": "./cifar10/"} 238 | {"load": null, "momentum": 0.9, "train_loss": 0.00353034524431625, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9628, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 236, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12566267148358748, "schedule": [150, 225], "save": "./cifar10/"} 239 | {"load": null, "momentum": 0.9, "train_loss": 0.003889501357524035, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 237, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1228469470073469, "schedule": [150, 225], "save": "./cifar10/"} 240 | {"load": null, "momentum": 0.9, "train_loss": 0.0033881939777356306, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9626, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 238, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12331506436457858, "schedule": [150, 225], "save": "./cifar10/"} 241 | {"load": null, "momentum": 0.9, "train_loss": 0.005029275096238006, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 239, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12214900381979532, "schedule": [150, 225], "save": "./cifar10/"} 242 | {"load": null, "momentum": 0.9, "train_loss": 0.0029654502337527052, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9641, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 240, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1236105169239454, "schedule": [150, 225], "save": "./cifar10/"} 243 | {"load": null, "momentum": 0.9, "train_loss": 0.003927277921781521, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9634, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 241, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12233292169403284, "schedule": [150, 225], "save": "./cifar10/"} 244 | {"load": null, "momentum": 0.9, "train_loss": 0.002870309948664706, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9624, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 242, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1227298795885872, "schedule": [150, 225], "save": "./cifar10/"} 245 | {"load": null, "momentum": 0.9, "train_loss": 0.002705255574614563, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9644, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 243, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1229365301958751, "schedule": [150, 225], "save": "./cifar10/"} 246 | {"load": null, "momentum": 0.9, "train_loss": 0.004482150760149488, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9629, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 244, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12327680965769104, "schedule": [150, 225], "save": "./cifar10/"} 247 | {"load": null, "momentum": 0.9, "train_loss": 0.006738728863013546, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9642, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 245, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12228621399961412, "schedule": [150, 225], "save": "./cifar10/"} 248 | {"load": null, "momentum": 0.9, "train_loss": 0.003348268517565846, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 246, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12298201682744547, "schedule": [150, 225], "save": "./cifar10/"} 249 | {"load": null, "momentum": 0.9, "train_loss": 0.004431468103217517, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.964, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 247, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12213080604793504, "schedule": [150, 225], "save": "./cifar10/"} 250 | {"load": null, "momentum": 0.9, "train_loss": 0.005800364873153012, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9616, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 248, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12389650874771178, "schedule": [150, 225], "save": "./cifar10/"} 251 | {"load": null, "momentum": 0.9, "train_loss": 0.003020123408002102, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9632, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 249, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12277358030667529, "schedule": [150, 225], "save": "./cifar10/"} 252 | {"load": null, "momentum": 0.9, "train_loss": 0.003435328007179868, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9641, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 250, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12259911050554365, "schedule": [150, 225], "save": "./cifar10/"} 253 | {"load": null, "momentum": 0.9, "train_loss": 0.002398559246894967, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9618, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 251, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1245108513138257, "schedule": [150, 225], "save": "./cifar10/"} 254 | {"load": null, "momentum": 0.9, "train_loss": 0.003675810501809577, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9643, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 252, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12200792725547217, "schedule": [150, 225], "save": "./cifar10/"} 255 | {"load": null, "momentum": 0.9, "train_loss": 0.0036602822601106326, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9637, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 253, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12327381319366396, "schedule": [150, 225], "save": "./cifar10/"} 256 | {"load": null, "momentum": 0.9, "train_loss": 0.0029804112581202115, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9636, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 254, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12305690033710562, "schedule": [150, 225], "save": "./cifar10/"} 257 | {"load": null, "momentum": 0.9, "train_loss": 0.003214730049149222, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9637, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 255, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12339711045264266, "schedule": [150, 225], "save": "./cifar10/"} 258 | {"load": null, "momentum": 0.9, "train_loss": 0.003487114125487931, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9628, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 256, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12289890493382699, "schedule": [150, 225], "save": "./cifar10/"} 259 | {"load": null, "momentum": 0.9, "train_loss": 0.0021039995065068985, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9631, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 257, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12333212748635561, "schedule": [150, 225], "save": "./cifar10/"} 260 | {"load": null, "momentum": 0.9, "train_loss": 0.0029836499889249937, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9635, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 258, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1230759689048864, "schedule": [150, 225], "save": "./cifar10/"} 261 | {"load": null, "momentum": 0.9, "train_loss": 0.0024926971708267485, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9634, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 259, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12378971961326897, "schedule": [150, 225], "save": "./cifar10/"} 262 | {"load": null, "momentum": 0.9, "train_loss": 0.003001655611307979, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9631, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 260, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12311361112399027, "schedule": [150, 225], "save": "./cifar10/"} 263 | {"load": null, "momentum": 0.9, "train_loss": 0.0024933579733092777, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9636, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 261, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12235193759901448, "schedule": [150, 225], "save": "./cifar10/"} 264 | {"load": null, "momentum": 0.9, "train_loss": 0.004080867965722056, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9632, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 262, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12385067977244034, "schedule": [150, 225], "save": "./cifar10/"} 265 | {"load": null, "momentum": 0.9, "train_loss": 0.005845987792709447, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 263, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12348166382755153, "schedule": [150, 225], "save": "./cifar10/"} 266 | {"load": null, "momentum": 0.9, "train_loss": 0.003004258034392124, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.963, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 264, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12387946122093126, "schedule": [150, 225], "save": "./cifar10/"} 267 | {"load": null, "momentum": 0.9, "train_loss": 0.0019027536842367377, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9635, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 265, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12221496771206147, "schedule": [150, 225], "save": "./cifar10/"} 268 | {"load": null, "momentum": 0.9, "train_loss": 0.0023761535231858174, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9622, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 266, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12401096355286427, "schedule": [150, 225], "save": "./cifar10/"} 269 | {"load": null, "momentum": 0.9, "train_loss": 0.0032755447491490315, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9629, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 267, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1225827261141967, "schedule": [150, 225], "save": "./cifar10/"} 270 | {"load": null, "momentum": 0.9, "train_loss": 0.008209587475294766, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9639, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 268, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12371349875000305, "schedule": [150, 225], "save": "./cifar10/"} 271 | {"load": null, "momentum": 0.9, "train_loss": 0.0036930687488394114, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9634, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 269, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12257061482174322, "schedule": [150, 225], "save": "./cifar10/"} 272 | {"load": null, "momentum": 0.9, "train_loss": 0.0026023124155518996, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9629, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 270, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1243571346567478, "schedule": [150, 225], "save": "./cifar10/"} 273 | {"load": null, "momentum": 0.9, "train_loss": 0.0024697067443644016, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.963, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 271, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1236650044831913, "schedule": [150, 225], "save": "./cifar10/"} 274 | {"load": null, "momentum": 0.9, "train_loss": 0.002784319862625112, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9641, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 272, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12358801879221573, "schedule": [150, 225], "save": "./cifar10/"} 275 | {"load": null, "momentum": 0.9, "train_loss": 0.004019681803222138, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9628, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 273, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12366870715515688, "schedule": [150, 225], "save": "./cifar10/"} 276 | {"load": null, "momentum": 0.9, "train_loss": 0.007116835699044376, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9636, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 274, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12230663126567379, "schedule": [150, 225], "save": "./cifar10/"} 277 | {"load": null, "momentum": 0.9, "train_loss": 0.004110124265546024, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9643, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 275, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12344820452271961, "schedule": [150, 225], "save": "./cifar10/"} 278 | {"load": null, "momentum": 0.9, "train_loss": 0.0025541486241024525, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9637, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 276, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12303338631056249, "schedule": [150, 225], "save": "./cifar10/"} 279 | {"load": null, "momentum": 0.9, "train_loss": 0.004500292776492055, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9626, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 277, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12348060775082559, "schedule": [150, 225], "save": "./cifar10/"} 280 | {"load": null, "momentum": 0.9, "train_loss": 0.002364718138856009, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9643, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 278, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12286251512821764, "schedule": [150, 225], "save": "./cifar10/"} 281 | {"load": null, "momentum": 0.9, "train_loss": 0.002312139178164735, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9626, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 279, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12326026407186873, "schedule": [150, 225], "save": "./cifar10/"} 282 | {"load": null, "momentum": 0.9, "train_loss": 0.00457396843911867, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9629, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 280, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12456441071000882, "schedule": [150, 225], "save": "./cifar10/"} 283 | {"load": null, "momentum": 0.9, "train_loss": 0.0019944933339618425, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9635, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 281, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12324362804181874, "schedule": [150, 225], "save": "./cifar10/"} 284 | {"load": null, "momentum": 0.9, "train_loss": 0.00343114333083396, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9632, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 282, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12276332213776187, "schedule": [150, 225], "save": "./cifar10/"} 285 | {"load": null, "momentum": 0.9, "train_loss": 0.009143110481513027, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9626, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 283, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12451606411719694, "schedule": [150, 225], "save": "./cifar10/"} 286 | {"load": null, "momentum": 0.9, "train_loss": 0.003093198081709482, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9635, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 284, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12221563402446918, "schedule": [150, 225], "save": "./cifar10/"} 287 | {"load": null, "momentum": 0.9, "train_loss": 0.003213724890594683, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.963, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 285, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12504095760872588, "schedule": [150, 225], "save": "./cifar10/"} 288 | {"load": null, "momentum": 0.9, "train_loss": 0.0028024453870028054, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9623, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 286, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12347501578158698, "schedule": [150, 225], "save": "./cifar10/"} 289 | {"load": null, "momentum": 0.9, "train_loss": 0.0049871711985399235, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9628, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 287, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12486130614764988, "schedule": [150, 225], "save": "./cifar10/"} 290 | {"load": null, "momentum": 0.9, "train_loss": 0.004965986925902667, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9628, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 288, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12362388557055964, "schedule": [150, 225], "save": "./cifar10/"} 291 | {"load": null, "momentum": 0.9, "train_loss": 0.00224579026046494, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9624, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 289, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12548424933454952, "schedule": [150, 225], "save": "./cifar10/"} 292 | {"load": null, "momentum": 0.9, "train_loss": 0.002603057187234684, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.963, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 290, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12430377512937411, "schedule": [150, 225], "save": "./cifar10/"} 293 | {"load": null, "momentum": 0.9, "train_loss": 0.005088397701975614, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9636, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 291, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1234501900197938, "schedule": [150, 225], "save": "./cifar10/"} 294 | {"load": null, "momentum": 0.9, "train_loss": 0.003936855000187874, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9625, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 292, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12488066383637487, "schedule": [150, 225], "save": "./cifar10/"} 295 | {"load": null, "momentum": 0.9, "train_loss": 0.0043818963612073485, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9621, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 293, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12632466240786017, "schedule": [150, 225], "save": "./cifar10/"} 296 | {"load": null, "momentum": 0.9, "train_loss": 0.004207146282157093, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9629, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 294, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12437587867374532, "schedule": [150, 225], "save": "./cifar10/"} 297 | {"load": null, "momentum": 0.9, "train_loss": 0.0037611155573168517, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9627, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 295, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12633475233102218, "schedule": [150, 225], "save": "./cifar10/"} 298 | {"load": null, "momentum": 0.9, "train_loss": 0.002445609683993189, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9614, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 296, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12667138249147683, "schedule": [150, 225], "save": "./cifar10/"} 299 | {"load": null, "momentum": 0.9, "train_loss": 0.0035723799590737697, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9628, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 297, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12635885129100644, "schedule": [150, 225], "save": "./cifar10/"} 300 | {"load": null, "momentum": 0.9, "train_loss": 0.005797809002070756, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9633, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 298, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.12458226824644952, "schedule": [150, 225], "save": "./cifar10/"} 301 | {"load": null, "momentum": 0.9, "train_loss": 0.003527769341146083, "batch_size": 128, "decay": 0.0005, "epochs": 300, "test": false, "log": "cifar10", "test_bs": 50, "gamma": 0.1, "test_accuracy": 0.9623, "data_path": "/data/cifar.pytorch/", "cardinality": 8, "prefetch": 4, "epoch": 299, "depth": 29, "dataset": "cifar10", "widen_factor": 4, "learning_rate": 0.0005000000000000001, "ngpu": 2, "test_loss": 0.1258646929007955, "schedule": [150, 225], "save": "./cifar10/"} 302 | -------------------------------------------------------------------------------- /cifar100/cifar-100.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prlz77/ResNeXt.pytorch/39fb8d03847f26ec02fb9b880ecaaa88db7a7d16/cifar100/cifar-100.jpg -------------------------------------------------------------------------------- /models/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prlz77/ResNeXt.pytorch/39fb8d03847f26ec02fb9b880ecaaa88db7a7d16/models/__init__.py -------------------------------------------------------------------------------- /models/model.py: -------------------------------------------------------------------------------- 1 | # -*- coding: utf-8 -*- 2 | from __future__ import division 3 | 4 | """ 5 | Creates a ResNeXt Model as defined in: 6 | 7 | Xie, S., Girshick, R., Dollár, P., Tu, Z., & He, K. (2016). 8 | Aggregated residual transformations for deep neural networks. 9 | arXiv preprint arXiv:1611.05431. 10 | 11 | """ 12 | 13 | __author__ = "Pau Rodríguez López, ISELAB, CVC-UAB" 14 | __email__ = "pau.rodri1@gmail.com" 15 | 16 | import torch.nn as nn 17 | import torch.nn.functional as F 18 | from torch.nn import init 19 | 20 | 21 | class ResNeXtBottleneck(nn.Module): 22 | """ 23 | RexNeXt bottleneck type C (https://github.com/facebookresearch/ResNeXt/blob/master/models/resnext.lua) 24 | """ 25 | 26 | def __init__(self, in_channels, out_channels, stride, cardinality, base_width, widen_factor): 27 | """ Constructor 28 | 29 | Args: 30 | in_channels: input channel dimensionality 31 | out_channels: output channel dimensionality 32 | stride: conv stride. Replaces pooling layer. 33 | cardinality: num of convolution groups. 34 | base_width: base number of channels in each group. 35 | widen_factor: factor to reduce the input dimensionality before convolution. 36 | """ 37 | super(ResNeXtBottleneck, self).__init__() 38 | width_ratio = out_channels / (widen_factor * 64.) 39 | D = cardinality * int(base_width * width_ratio) 40 | self.conv_reduce = nn.Conv2d(in_channels, D, kernel_size=1, stride=1, padding=0, bias=False) 41 | self.bn_reduce = nn.BatchNorm2d(D) 42 | self.conv_conv = nn.Conv2d(D, D, kernel_size=3, stride=stride, padding=1, groups=cardinality, bias=False) 43 | self.bn = nn.BatchNorm2d(D) 44 | self.conv_expand = nn.Conv2d(D, out_channels, kernel_size=1, stride=1, padding=0, bias=False) 45 | self.bn_expand = nn.BatchNorm2d(out_channels) 46 | 47 | self.shortcut = nn.Sequential() 48 | if in_channels != out_channels: 49 | self.shortcut.add_module('shortcut_conv', 50 | nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=stride, padding=0, 51 | bias=False)) 52 | self.shortcut.add_module('shortcut_bn', nn.BatchNorm2d(out_channels)) 53 | 54 | def forward(self, x): 55 | bottleneck = self.conv_reduce.forward(x) 56 | bottleneck = F.relu(self.bn_reduce.forward(bottleneck), inplace=True) 57 | bottleneck = self.conv_conv.forward(bottleneck) 58 | bottleneck = F.relu(self.bn.forward(bottleneck), inplace=True) 59 | bottleneck = self.conv_expand.forward(bottleneck) 60 | bottleneck = self.bn_expand.forward(bottleneck) 61 | residual = self.shortcut.forward(x) 62 | return F.relu(residual + bottleneck, inplace=True) 63 | 64 | 65 | class CifarResNeXt(nn.Module): 66 | """ 67 | ResNext optimized for the Cifar dataset, as specified in 68 | https://arxiv.org/pdf/1611.05431.pdf 69 | """ 70 | 71 | def __init__(self, cardinality, depth, nlabels, base_width, widen_factor=4): 72 | """ Constructor 73 | 74 | Args: 75 | cardinality: number of convolution groups. 76 | depth: number of layers. 77 | nlabels: number of classes 78 | base_width: base number of channels in each group. 79 | widen_factor: factor to adjust the channel dimensionality 80 | """ 81 | super(CifarResNeXt, self).__init__() 82 | self.cardinality = cardinality 83 | self.depth = depth 84 | self.block_depth = (self.depth - 2) // 9 85 | self.base_width = base_width 86 | self.widen_factor = widen_factor 87 | self.nlabels = nlabels 88 | self.output_size = 64 89 | self.stages = [64, 64 * self.widen_factor, 128 * self.widen_factor, 256 * self.widen_factor] 90 | 91 | self.conv_1_3x3 = nn.Conv2d(3, 64, 3, 1, 1, bias=False) 92 | self.bn_1 = nn.BatchNorm2d(64) 93 | self.stage_1 = self.block('stage_1', self.stages[0], self.stages[1], 1) 94 | self.stage_2 = self.block('stage_2', self.stages[1], self.stages[2], 2) 95 | self.stage_3 = self.block('stage_3', self.stages[2], self.stages[3], 2) 96 | self.classifier = nn.Linear(self.stages[3], nlabels) 97 | init.kaiming_normal(self.classifier.weight) 98 | 99 | for key in self.state_dict(): 100 | if key.split('.')[-1] == 'weight': 101 | if 'conv' in key: 102 | init.kaiming_normal(self.state_dict()[key], mode='fan_out') 103 | if 'bn' in key: 104 | self.state_dict()[key][...] = 1 105 | elif key.split('.')[-1] == 'bias': 106 | self.state_dict()[key][...] = 0 107 | 108 | def block(self, name, in_channels, out_channels, pool_stride=2): 109 | """ Stack n bottleneck modules where n is inferred from the depth of the network. 110 | 111 | Args: 112 | name: string name of the current block. 113 | in_channels: number of input channels 114 | out_channels: number of output channels 115 | pool_stride: factor to reduce the spatial dimensionality in the first bottleneck of the block. 116 | 117 | Returns: a Module consisting of n sequential bottlenecks. 118 | 119 | """ 120 | block = nn.Sequential() 121 | for bottleneck in range(self.block_depth): 122 | name_ = '%s_bottleneck_%d' % (name, bottleneck) 123 | if bottleneck == 0: 124 | block.add_module(name_, ResNeXtBottleneck(in_channels, out_channels, pool_stride, self.cardinality, 125 | self.base_width, self.widen_factor)) 126 | else: 127 | block.add_module(name_, 128 | ResNeXtBottleneck(out_channels, out_channels, 1, self.cardinality, self.base_width, 129 | self.widen_factor)) 130 | return block 131 | 132 | def forward(self, x): 133 | x = self.conv_1_3x3.forward(x) 134 | x = F.relu(self.bn_1.forward(x), inplace=True) 135 | x = self.stage_1.forward(x) 136 | x = self.stage_2.forward(x) 137 | x = self.stage_3.forward(x) 138 | x = F.avg_pool2d(x, 8, 1) 139 | x = x.view(-1, self.stages[3]) 140 | return self.classifier(x) 141 | -------------------------------------------------------------------------------- /test.py: -------------------------------------------------------------------------------- 1 | # -*- coding: utf-8 -*- 2 | from __future__ import division 3 | 4 | """ 5 | Test the ResNeXt Model on Cifar10 and Cifar 100. Implementation as defined in: 6 | 7 | Xie, S., Girshick, R., Dollár, P., Tu, Z., & He, K. (2016). 8 | Aggregated residual transformations for deep neural networks. 9 | arXiv preprint arXiv:1611.05431. 10 | 11 | """ 12 | 13 | __author__ = "Pau Rodríguez López, ISELAB, CVC-UAB" 14 | __email__ = "pau.rodri1@gmail.com" 15 | 16 | __editor__ = "Il-Ji Choi, Vuno. Inc." # test file 17 | __editor_email__ = "choiilji@gmail.com" 18 | 19 | import argparse 20 | import torch 21 | import torch.nn.functional as F 22 | import torchvision.datasets as dset 23 | import torchvision.transforms as transforms 24 | from models.model import CifarResNeXt 25 | 26 | def get_args(): 27 | parser = argparse.ArgumentParser(description='Test ResNeXt on CIFAR', 28 | formatter_class=argparse.ArgumentDefaultsHelpFormatter) 29 | # Positional arguments 30 | parser.add_argument('data_path', type=str, help='Root for the Cifar dataset.') 31 | parser.add_argument('dataset', type=str, choices=['cifar10', 'cifar100'], help='Choose between Cifar10/100.') 32 | # Optimization options 33 | parser.add_argument('--batch_size', '-b', type=int, default=128, help='Batch size.') 34 | parser.add_argument('--test_bs', type=int, default=10) 35 | # Checkpoints 36 | parser.add_argument('--load', '-l', type=str, help='Checkpoint path to resume / test.') 37 | parser.add_argument('--test', '-t', action='store_true', help='Test only flag.') 38 | # Architecture 39 | parser.add_argument('--depth', type=int, default=29, help='Model depth.') 40 | parser.add_argument('--cardinality', type=int, default=8, help='Model cardinality (group).') 41 | parser.add_argument('--base_width', type=int, default=64, help='Number of channels in each group.') 42 | parser.add_argument('--widen_factor', type=int, default=4, help='Widen factor. 4 -> 64, 8 -> 128, ...') 43 | # Acceleration 44 | parser.add_argument('--ngpu', type=int, default=1, help='0 = CPU.') 45 | parser.add_argument('--prefetch', type=int, default=2, help='Pre-fetching threads.') 46 | # i/o 47 | parser.add_argument('--log', type=str, default='./', help='Log folder.') 48 | args = parser.parse_args() 49 | return args 50 | 51 | def test(): 52 | # define default variables 53 | args = get_args()# divide args part and call it as function 54 | mean = [x / 255 for x in [125.3, 123.0, 113.9]] 55 | std = [x / 255 for x in [63.0, 62.1, 66.7]] 56 | state = {k: v for k, v in args._get_kwargs()} 57 | 58 | # prepare test data parts 59 | test_transform = transforms.Compose( 60 | [transforms.ToTensor(), transforms.Normalize(mean, std)]) 61 | test_data = dset.CIFAR100(args.data_path, train=False, transform=test_transform, download=True) 62 | if args.dataset == 'cifar10': 63 | test_data = dset.CIFAR10(args.data_path, train=False, transform=test_transform, download=True) 64 | nlabels = 10 65 | else: 66 | test_data = dset.CIFAR100(args.data_path, train=False, transform=test_transform, download=True) 67 | nlabels = 100 68 | 69 | test_loader = torch.utils.data.DataLoader(test_data, batch_size=args.test_bs, shuffle=False, 70 | num_workers=args.prefetch, pin_memory=True) 71 | 72 | # initialize model and load from checkpoint 73 | net = CifarResNeXt(args.cardinality, args.depth, nlabels, args.base_width, args.widen_factor) 74 | loaded_state_dict = torch.load(args.load) 75 | temp = {} 76 | for key, val in list(loaded_state_dict.items()): 77 | if 'module' in key: 78 | # parsing keys for ignoring 'module.' in keys 79 | temp[key[7:]] = val 80 | else: 81 | temp[key] = val 82 | loaded_state_dict = temp 83 | net.load_state_dict(loaded_state_dict) 84 | 85 | # paralleize model 86 | if args.ngpu > 1: 87 | net = torch.nn.DataParallel(net, device_ids=list(range(args.ngpu))) 88 | if args.ngpu > 0: 89 | net.cuda() 90 | 91 | # use network for evaluation 92 | net.eval() 93 | 94 | # calculation part 95 | loss_avg = 0.0 96 | correct = 0.0 97 | with torch.no_grad(): 98 | for batch_idx, (data, target) in enumerate(test_loader): 99 | data, target = torch.autograd.Variable(data.cuda()), torch.autograd.Variable(target.cuda()) 100 | 101 | # forward 102 | output = net(data) 103 | loss = F.cross_entropy(output, target) 104 | 105 | # accuracy 106 | pred = output.data.max(1)[1] 107 | correct += pred.eq(target.data).sum() 108 | 109 | # test loss average 110 | loss_avg += loss.item() 111 | 112 | state['test_loss'] = loss_avg / len(test_loader) 113 | state['test_accuracy'] = correct / len(test_loader.dataset) 114 | 115 | # finally print state dictionary 116 | print(state) 117 | 118 | if __name__=='__main__': 119 | test() 120 | -------------------------------------------------------------------------------- /train.py: -------------------------------------------------------------------------------- 1 | # -*- coding: utf-8 -*- 2 | from __future__ import division 3 | 4 | """ 5 | Trains a ResNeXt Model on Cifar10 and Cifar 100. Implementation as defined in: 6 | 7 | Xie, S., Girshick, R., Dollár, P., Tu, Z., & He, K. (2016). 8 | Aggregated residual transformations for deep neural networks. 9 | arXiv preprint arXiv:1611.05431. 10 | 11 | """ 12 | 13 | __author__ = "Pau Rodríguez López, ISELAB, CVC-UAB" 14 | __email__ = "pau.rodri1@gmail.com" 15 | 16 | import argparse 17 | import os 18 | import json 19 | import torch 20 | import torch.nn.functional as F 21 | import torchvision.datasets as dset 22 | import torchvision.transforms as transforms 23 | from models.model import CifarResNeXt 24 | 25 | if __name__ == '__main__': 26 | parser = argparse.ArgumentParser(description='Trains ResNeXt on CIFAR', 27 | formatter_class=argparse.ArgumentDefaultsHelpFormatter) 28 | # Positional arguments 29 | parser.add_argument('data_path', type=str, help='Root for the Cifar dataset.') 30 | parser.add_argument('dataset', type=str, choices=['cifar10', 'cifar100'], help='Choose between Cifar10/100.') 31 | # Optimization options 32 | parser.add_argument('--epochs', '-e', type=int, default=300, help='Number of epochs to train.') 33 | parser.add_argument('--batch_size', '-b', type=int, default=128, help='Batch size.') 34 | parser.add_argument('--learning_rate', '-lr', type=float, default=0.1, help='The Learning Rate.') 35 | parser.add_argument('--momentum', '-m', type=float, default=0.9, help='Momentum.') 36 | parser.add_argument('--decay', '-d', type=float, default=0.0005, help='Weight decay (L2 penalty).') 37 | parser.add_argument('--test_bs', type=int, default=10) 38 | parser.add_argument('--schedule', type=int, nargs='+', default=[150, 225], 39 | help='Decrease learning rate at these epochs.') 40 | parser.add_argument('--gamma', type=float, default=0.1, help='LR is multiplied by gamma on schedule.') 41 | # Checkpoints 42 | parser.add_argument('--save', '-s', type=str, default='./', help='Folder to save checkpoints.') 43 | parser.add_argument('--load', '-l', type=str, help='Checkpoint path to resume / test.') 44 | parser.add_argument('--test', '-t', action='store_true', help='Test only flag.') 45 | # Architecture 46 | parser.add_argument('--depth', type=int, default=29, help='Model depth.') 47 | parser.add_argument('--cardinality', type=int, default=8, help='Model cardinality (group).') 48 | parser.add_argument('--base_width', type=int, default=64, help='Number of channels in each group.') 49 | parser.add_argument('--widen_factor', type=int, default=4, help='Widen factor. 4 -> 64, 8 -> 128, ...') 50 | # Acceleration 51 | parser.add_argument('--ngpu', type=int, default=1, help='0 = CPU.') 52 | parser.add_argument('--prefetch', type=int, default=2, help='Pre-fetching threads.') 53 | # i/o 54 | parser.add_argument('--log', type=str, default='./', help='Log folder.') 55 | args = parser.parse_args() 56 | 57 | # Init logger 58 | if not os.path.isdir(args.log): 59 | os.makedirs(args.log) 60 | log = open(os.path.join(args.log, 'log.txt'), 'w') 61 | state = {k: v for k, v in args._get_kwargs()} 62 | log.write(json.dumps(state) + '\n') 63 | 64 | # Calculate number of epochs wrt batch size 65 | args.epochs = args.epochs * 128 // args.batch_size 66 | args.schedule = [x * 128 // args.batch_size for x in args.schedule] 67 | 68 | # Init dataset 69 | if not os.path.isdir(args.data_path): 70 | os.makedirs(args.data_path) 71 | 72 | mean = [x / 255 for x in [125.3, 123.0, 113.9]] 73 | std = [x / 255 for x in [63.0, 62.1, 66.7]] 74 | 75 | train_transform = transforms.Compose( 76 | [transforms.RandomHorizontalFlip(), transforms.RandomCrop(32, padding=4), transforms.ToTensor(), 77 | transforms.Normalize(mean, std)]) 78 | test_transform = transforms.Compose( 79 | [transforms.ToTensor(), transforms.Normalize(mean, std)]) 80 | if args.dataset == 'cifar10': 81 | train_data = dset.CIFAR10(args.data_path, train=True, transform=train_transform, download=True) 82 | test_data = dset.CIFAR10(args.data_path, train=False, transform=test_transform, download=True) 83 | nlabels = 10 84 | else: 85 | train_data = dset.CIFAR100(args.data_path, train=True, transform=train_transform, download=True) 86 | test_data = dset.CIFAR100(args.data_path, train=False, transform=test_transform, download=True) 87 | nlabels = 100 88 | train_loader = torch.utils.data.DataLoader(train_data, batch_size=args.batch_size, shuffle=True, 89 | num_workers=args.prefetch, pin_memory=True) 90 | test_loader = torch.utils.data.DataLoader(test_data, batch_size=args.test_bs, shuffle=False, 91 | num_workers=args.prefetch, pin_memory=True) 92 | 93 | # Init checkpoints 94 | if not os.path.isdir(args.save): 95 | os.makedirs(args.save) 96 | 97 | # Init model, criterion, and optimizer 98 | net = CifarResNeXt(args.cardinality, args.depth, nlabels, args.base_width, args.widen_factor) 99 | print(net) 100 | if args.ngpu > 1: 101 | net = torch.nn.DataParallel(net, device_ids=list(range(args.ngpu))) 102 | 103 | if args.ngpu > 0: 104 | net.cuda() 105 | 106 | optimizer = torch.optim.SGD(net.parameters(), state['learning_rate'], momentum=state['momentum'], 107 | weight_decay=state['decay'], nesterov=True) 108 | 109 | # train function (forward, backward, update) 110 | def train(): 111 | net.train() 112 | loss_avg = 0.0 113 | for batch_idx, (data, target) in enumerate(train_loader): 114 | data, target = torch.autograd.Variable(data.cuda()), torch.autograd.Variable(target.cuda()) 115 | 116 | # forward 117 | output = net(data) 118 | 119 | # backward 120 | optimizer.zero_grad() 121 | loss = F.cross_entropy(output, target) 122 | loss.backward() 123 | optimizer.step() 124 | 125 | # exponential moving average 126 | loss_avg = loss_avg * 0.2 + float(loss) * 0.8 127 | 128 | state['train_loss'] = loss_avg 129 | 130 | 131 | # test function (forward only) 132 | def test(): 133 | net.eval() 134 | loss_avg = 0.0 135 | correct = 0 136 | for batch_idx, (data, target) in enumerate(test_loader): 137 | data, target = torch.autograd.Variable(data.cuda()), torch.autograd.Variable(target.cuda()) 138 | 139 | # forward 140 | output = net(data) 141 | loss = F.cross_entropy(output, target) 142 | 143 | # accuracy 144 | pred = output.data.max(1)[1] 145 | correct += float(pred.eq(target.data).sum()) 146 | 147 | # test loss average 148 | loss_avg += float(loss) 149 | 150 | state['test_loss'] = loss_avg / len(test_loader) 151 | state['test_accuracy'] = correct / len(test_loader.dataset) 152 | 153 | 154 | # Main loop 155 | best_accuracy = 0.0 156 | for epoch in range(args.epochs): 157 | if epoch in args.schedule: 158 | state['learning_rate'] *= args.gamma 159 | for param_group in optimizer.param_groups: 160 | param_group['lr'] = state['learning_rate'] 161 | 162 | state['epoch'] = epoch 163 | train() 164 | test() 165 | if state['test_accuracy'] > best_accuracy: 166 | best_accuracy = state['test_accuracy'] 167 | torch.save(net.state_dict(), os.path.join(args.save, 'model.pytorch')) 168 | log.write('%s\n' % json.dumps(state)) 169 | log.flush() 170 | print(state) 171 | print("Best accuracy: %f" % best_accuracy) 172 | 173 | log.close() 174 | -------------------------------------------------------------------------------- /utils/plot_log.py: -------------------------------------------------------------------------------- 1 | import re 2 | import matplotlib.pyplot as plt 3 | 4 | if __name__=='__main__': 5 | file = open('./logs/log.txt','r') 6 | accuracy = [] 7 | epochs = [] 8 | loss = [] 9 | for line in file: 10 | test_accuracy = re.search('"test_accuracy": ([0]\.[0-9]+)*', line) 11 | if test_accuracy: 12 | accuracy.append(test_accuracy.group(1)) 13 | 14 | epoch = re.search('"epoch": ([0-9]+)*', line) 15 | if epoch: 16 | epochs.append(epoch.group(1)) 17 | 18 | train_loss = re.search('"train_loss": ([0-9]\.[0-9]+)*', line) 19 | if train_loss: 20 | loss.append(train_loss.group(1)) 21 | file.close() 22 | plt.figure('test_accuracy vs epochs') 23 | plt.xlabel('epoch') 24 | plt.ylabel('test_accuracy') 25 | plt.plot(epochs,accuracy,'b*') 26 | plt.plot(epochs,accuracy,'r') 27 | plt.grid(True) 28 | 29 | plt.figure('train_loss vs epochs') 30 | plt.xlabel('epoch') 31 | plt.ylabel('train_loss') 32 | plt.plot(epochs,loss,'b*') 33 | plt.plot(epochs,loss,'y') 34 | plt.grid(True) 35 | 36 | plt.show() 37 | 38 | --------------------------------------------------------------------------------