diff --git a/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..598d65c66010457af40a76c5b641acc0ce1f1461 --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..58f3a54a00003c93e5d75c35de3fcbfdd082b8c9 --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..94a096195c06d2bad763c41f54dbf7227599ef8b --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/README.md b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..94a096195c06d2bad763c41f54dbf7227599ef8b --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/config.json b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9ade3be99540a4e3d4643bc3442e7a46b461d20e --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ca6ed2f4a1229d21291d93f1934e7160c23f89f2 --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/README.md b/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8b8f455b794be39589206305ca1a740e99d752c2 --- /dev/null +++ b/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/config.json b/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f78fd7a9eb99c1adf9bf4c761a98bcba9cb9f99 --- /dev/null +++ b/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8d9636b5b7ef9318dab1601103bff7e40937b2ea --- /dev/null +++ b/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fae9a9cc3053b5936d45ecae91b1a2edabf051ba --- /dev/null +++ b/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar10-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8dfa4deeb3c448307889a584850270270706b10c --- /dev/null +++ b/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..acbad7473b8e84235773c84b96dc29cdb56af499 --- /dev/null +++ b/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar100-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7d60b59b5be7c6a903da11e724a8dc4a55b45c3d --- /dev/null +++ b/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..acbad7473b8e84235773c84b96dc29cdb56af499 --- /dev/null +++ b/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar100-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..95bb009d63293662d48aae533d3937db4bd93a48 --- /dev/null +++ b/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..acbad7473b8e84235773c84b96dc29cdb56af499 --- /dev/null +++ b/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d36a31c9d49067d719654c2e53df892e5937ecfd --- /dev/null +++ b/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..acbad7473b8e84235773c84b96dc29cdb56af499 --- /dev/null +++ b/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c99d01cadbf73a4293ccc907b2900b2b0d09118c --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4d2d193ef01ae3cc88be42b32876e4b826d9772b --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..178724417f57fd8ea337fa747e1f0130c25902e1 --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c59b60bf98aec6a560a35d085f03cf9158b429f3 --- /dev/null +++ b/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..aad95549a3effef08aa283c9f8aaaa2f8d702967 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/README.md b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b41949934c863d312f89201da43423022acd76cd --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/config.json b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f4b4ad2690cdc5591354d254789342f7fea2d6d5 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d5e595d40df5bd6dd145ccb3d4e7c1c2a4109333 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e6851102db373d3568235fa17e3ce199c20b2dc3 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e6851102db373d3568235fa17e3ce199c20b2dc3 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cee8e06ce13fa5525fb8517c300e637bc05dba7 --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-dtd-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md b/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f9934e63240213af15fa4eae27b328ce630f7574 --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json b/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1de185c4a239f93470f739721235a30bac76536e --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f7b3e542924e83e476ed485975b858ec99c001e2 --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1de185c4a239f93470f739721235a30bac76536e --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f7b3e542924e83e476ed485975b858ec99c001e2 --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1de185c4a239f93470f739721235a30bac76536e --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-emnist_letters-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b38d97a56d26e7e6ea2bb481f949be6d6e33c7f5 --- /dev/null +++ b/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1de185c4a239f93470f739721235a30bac76536e --- /dev/null +++ b/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..13f28b0f8718c9ab28f3fa61b535712344834b2c --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3fbee75f8255e01f77ec81c2c941301ad8f54ddc --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7624c38cee713e0366a905e321e1be83705adf67 --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3fbee75f8255e01f77ec81c2c941301ad8f54ddc --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7624c38cee713e0366a905e321e1be83705adf67 --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3fbee75f8255e01f77ec81c2c941301ad8f54ddc --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c48c2df17318b268fbd672974ddb7504fc21c233 --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3fbee75f8255e01f77ec81c2c941301ad8f54ddc --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..423d126c174e116dabd223993a23290d869ffa90 --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..99626e1d853a8525a84a8d4ca7c121c77c398eed --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..76b1a62dbf904c8722976b329a12f832e5820de2 --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..99626e1d853a8525a84a8d4ca7c121c77c398eed --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..09d833f40b1f597b117956e1c13686bbf5348a39 --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..99626e1d853a8525a84a8d4ca7c121c77c398eed --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c19b60c3bb5344e26c4307d9984370de587502da --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..99626e1d853a8525a84a8d4ca7c121c77c398eed --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4820924d0c9c35cc7f5988d8547433969888893e --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..99626e1d853a8525a84a8d4ca7c121c77c398eed --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/README.md b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1bc2c0584e839298cf7635b5b550392a2a75a9a2 --- /dev/null +++ b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/config.json b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1bc2c0584e839298cf7635b5b550392a2a75a9a2 --- /dev/null +++ b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..47eac08a8c078440e59da7e5cf3d481a4bc189fd --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..47eac08a8c078440e59da7e5cf3d481a4bc189fd --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..750285cf1dbef322b764e7f233e02e3ec567b2ae --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..81ae378b0786d377e1a38245aae7328fa9a08c19 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..68173be56ecc7f7da88dacef9d10a6c6d8151603 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1d9c70628bd4028bc1dafba699afc9504fe52ce0 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1d9c70628bd4028bc1dafba699afc9504fe52ce0 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..59dc67fd6cf0c4451dcc617b2de75eda18ebad1f --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..874518c652ec15eb6e18b82397e47e4f485a6af7 --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d48302a436f74a903d847ba02c3a3ef8b2c62632 --- /dev/null +++ b/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..002738700083d19b6f2089a9c2d0959b8ee2b0bb --- /dev/null +++ b/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-food101-batch_size_128_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md b/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..11f18db8c69c41cc45889ab82fdc7587194ddfa5 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json b/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a882d09145c7a66ff6c167ae2959b9472a2d7eb5 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2743dbb7b634d5308a52f8e188bb9c344fce9a19 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2743dbb7b634d5308a52f8e188bb9c344fce9a19 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c21b6b8daa1ee0a58c2e77507a7a2cbdbd3d0a58 --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/README.md b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a39e299d867f4b259bf762dcd31b3224f849c735 --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/config.json b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..006cb20af967f7e9ea75d5b405b1da30cda8e2b7 --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f177623c985b9494508bbf057acdfda413a114df --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..53161b832eb7e51bd11bce461528d4642c58be00 --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-gtsrb-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..15fe64fba741cffe2ba9a1834c4ee5da05f642dc --- /dev/null +++ b/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55c0a9dcaa76aba329cf2d4fb9ec5d866be9da8a --- /dev/null +++ b/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-kmnist-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b32b8e91eaceb491e2ce1b36a4f90a2c2384b424 --- /dev/null +++ b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55c0a9dcaa76aba329cf2d4fb9ec5d866be9da8a --- /dev/null +++ b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b32b8e91eaceb491e2ce1b36a4f90a2c2384b424 --- /dev/null +++ b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55c0a9dcaa76aba329cf2d4fb9ec5d866be9da8a --- /dev/null +++ b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c635e86462d7c6e67930b5e677702b52db4439bd --- /dev/null +++ b/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: mnist + pretrained: true +test_datasets: null +train_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: train +type: transformers +val_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: test +``` + diff --git a/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-mnist-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..73b609082c616c57a669b2be2045ff6bd1cf2c69 --- /dev/null +++ b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6e9072520d9ab1e64883a84e11a3a69f3f2a85ee --- /dev/null +++ b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b773896b1151bbf3717647bb1b318e7d3155568e --- /dev/null +++ b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6e9072520d9ab1e64883a84e11a3a69f3f2a85ee --- /dev/null +++ b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/README.md b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ed3b7c187b782b980d942aa976b02856d469c51f --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/config.json b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ed3b7c187b782b980d942aa976b02856d469c51f --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..227c796b4a5388c3a5e39990b92db6b163cf1298 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..227c796b4a5388c3a5e39990b92db6b163cf1298 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..01a4a55d8a7cf9476312e26eeebb57c416bdc02d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.27 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.27 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md b/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7d8b6519cb2e133640ceccc687a9c3eafcc2ee7b --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..69fa3871a161cfb61d653c6ce9465f4bb094dc0b --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..71f62818d3d5bd11eee682972dcecef20f710166 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..522ffcbe8021f310895943a05f373458ecfe31ce --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73d4c5807d169cb2e901ab29de6ec9f57efd9347 --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-oxford_flowers102-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e823ad92f586b79984de749645112a75fbd1b3ab --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3cb201dc7dfa380d2b50c3d20326ba2521864600 --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/README.md b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e823ad92f586b79984de749645112a75fbd1b3ab --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/config.json b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3cb201dc7dfa380d2b50c3d20326ba2521864600 --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0c09f71619efb156540f80b0c3e7947cc1d3205e --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3cb201dc7dfa380d2b50c3d20326ba2521864600 --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0c09f71619efb156540f80b0c3e7947cc1d3205e --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3cb201dc7dfa380d2b50c3d20326ba2521864600 --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-pcam-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..34f6be4fe3d8186518eae783f43972e72bec72a0 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..054ed535def906c22cb0ab0c6b25c9c88b195f06 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6fc916b70d52a60686c0be1303afd291a61fab4b --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..054ed535def906c22cb0ab0c6b25c9c88b195f06 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md b/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2970f6f529145b8a673bd63972e652b92610c0b7 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json b/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..054ed535def906c22cb0ab0c6b25c9c88b195f06 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md b/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d9fce3e0cb095a78cc061fa6b2dd3142e8ab3ee6 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json b/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..054ed535def906c22cb0ab0c6b25c9c88b195f06 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a801f12dcc9db1d642334ab075000a05d38b006c --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..054ed535def906c22cb0ab0c6b25c9c88b195f06 --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c72c8f027bf9d6ff5204a3179bc1e8e24f37927b --- /dev/null +++ b/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1ccd6b3d4942b65d4b1b5d4a4f27c0403ba41898 --- /dev/null +++ b/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6bfb47a67f2a8a7dbe5e28291e610d93e224c360 --- /dev/null +++ b/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1ccd6b3d4942b65d4b1b5d4a4f27c0403ba41898 --- /dev/null +++ b/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-resisc45-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6bfb47a67f2a8a7dbe5e28291e610d93e224c360 --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1ccd6b3d4942b65d4b1b5d4a4f27c0403ba41898 --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ac8a68aeb05f18cb1598fbef8450a24d9afd8d40 --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1ccd6b3d4942b65d4b1b5d4a4f27c0403ba41898 --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/README.md b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6edbcf53fc180fac628dd12a9ae686e884fc9557 --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/config.json b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1ccd6b3d4942b65d4b1b5d4a4f27c0403ba41898 --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c599186d25b1db74da0bc8d3bc05c4b05ad4b21a --- /dev/null +++ b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1ccd6b3d4942b65d4b1b5d4a4f27c0403ba41898 --- /dev/null +++ b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..22ff63b66cd197a1abddf646b522d19195419d74 --- /dev/null +++ b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1ccd6b3d4942b65d4b1b5d4a4f27c0403ba41898 --- /dev/null +++ b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-resisc45-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e867725306478435294224413c7fe86284d594ba --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..eb624463359f1dbca1baefcf1dda0a8d33f41113 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..09d352310265c0dc7627b32f3592bce22db40063 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e21805c7ff136f487586145e67608d3d23bcd1d4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..aabeeb949b3b90c40d86375d99f41262e266f3a5 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..aabeeb949b3b90c40d86375d99f41262e266f3a5 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e867725306478435294224413c7fe86284d594ba --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/README.md b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..eb624463359f1dbca1baefcf1dda0a8d33f41113 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/config.json b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..09d352310265c0dc7627b32f3592bce22db40063 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/README.md b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..09d352310265c0dc7627b32f3592bce22db40063 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/config.json b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..56db396f6294473e3dd45e739086cc3210400cb3 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8737b0d7806002d04443d62a80afecec31b1a09e --- /dev/null +++ b/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1e0f59991da08604f363224d408031555ecd60f4 --- /dev/null +++ b/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..32ffd24993f9112aee430366f23e2f1b32b34543 --- /dev/null +++ b/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4e96dfc41ebe7919b6d046db75114c5d465f222d --- /dev/null +++ b/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stl10-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/README.md b/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7579e51377aba6cd2fe08654745e695fca0fc45a --- /dev/null +++ b/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/config.json b/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4e96dfc41ebe7919b6d046db75114c5d465f222d --- /dev/null +++ b/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stl10-batch_size_128_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/README.md b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0e564e59b34e0b7c9a4d17af5e9c13dee258877b --- /dev/null +++ b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/config.json b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4e96dfc41ebe7919b6d046db75114c5d465f222d --- /dev/null +++ b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3e72c3eec5903b8cad63744200bd030c58397445 --- /dev/null +++ b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4e96dfc41ebe7919b6d046db75114c5d465f222d --- /dev/null +++ b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-stl10-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md b/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3df05d72b72894c1e105f87b5bf20e52e1af01cd --- /dev/null +++ b/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json b/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..235472e739b23ead2b88b4266434b8f42502d7ea --- /dev/null +++ b/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-sun397-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f4d99a0521dab12e7875073ec100c6a350c31930 --- /dev/null +++ b/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..235472e739b23ead2b88b4266434b8f42502d7ea --- /dev/null +++ b/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-sun397-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a0e2c1326d83f0f17d0ed57feaf648e8901cfddf --- /dev/null +++ b/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..235472e739b23ead2b88b4266434b8f42502d7ea --- /dev/null +++ b/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..92034c59308eb77a6735b3297602a366f7483956 --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..235472e739b23ead2b88b4266434b8f42502d7ea --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..491880bd5c13c29628035e496524389b957cd0cc --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..235472e739b23ead2b88b4266434b8f42502d7ea --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..491880bd5c13c29628035e496524389b957cd0cc --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..235472e739b23ead2b88b4266434b8f42502d7ea --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-sun397-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..94895ab1d277e42cfaf326f5ef220ffb30b33d10 --- /dev/null +++ b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..94895ab1d277e42cfaf326f5ef220ffb30b33d10 --- /dev/null +++ b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c7c683b1bcf2e467344fba85036b8d887decf37b --- /dev/null +++ b/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-svhn-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4e523f49113272a7f6b974a0a562b7b2205c735f --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..310dd91ccba15136ca1f387fd727a1fd0acec01b --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..14256c813b1a51e517f3b349cc52572959efcd8d --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..14256c813b1a51e517f3b349cc52572959efcd8d --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-152 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-152 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499ddc3e4e5dc0456a3110b1e6f22201afc1936e --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 8, + 36, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet152-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md b/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e4fa76f3e849ea903cb1fb7ba8f44bd8f6779672 --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json b/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a6fd7768d9fade7e7d7df7ed5cd1e93ae4baec32 --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..74a87c9c96fbb52c02ab1e55c3dcedf4f008672f --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a6fd7768d9fade7e7d7df7ed5cd1e93ae4baec32 --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..74a87c9c96fbb52c02ab1e55c3dcedf4f008672f --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a6fd7768d9fade7e7d7df7ed5cd1e93ae4baec32 --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..df952a11a6cda6b745ba69010d2ba6adc253d003 --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a6fd7768d9fade7e7d7df7ed5cd1e93ae4baec32 --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..96a63f55452f4a962f0c0d97ea337eb12c354891 --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a6fd7768d9fade7e7d7df7ed5cd1e93ae4baec32 --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..27610e1442306b2c86104d4448cb21d1e8fa3e47 --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a6fd7768d9fade7e7d7df7ed5cd1e93ae4baec32 --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c4140f611cca8a74105397a584a990b471395862 --- /dev/null +++ b/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a6fd7768d9fade7e7d7df7ed5cd1e93ae4baec32 --- /dev/null +++ b/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a31d091b885b5f0b1f2bc5c6728c19c5ef7810ba --- /dev/null +++ b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..5c3f41b9ed6c00df33ffaef448e5b6c8ac8a8d65 --- /dev/null +++ b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..300973acb0b6f068c2fedfebf31ee1515822619a --- /dev/null +++ b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..5c3f41b9ed6c00df33ffaef448e5b6c8ac8a8d65 --- /dev/null +++ b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar100-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4bdcae872307708c4b4edf0dd3ed7af81c2c8176 --- /dev/null +++ b/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..5c3f41b9ed6c00df33ffaef448e5b6c8ac8a8d65 --- /dev/null +++ b/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-cifar100-batch_size_256_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4bfb9ba0bb05ac60860bf3755abb5f0a7b97439b --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0033693c4d4d993b2e27bf771066639c2d5a6d35 --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0033693c4d4d993b2e27bf771066639c2d5a6d35 --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0033693c4d4d993b2e27bf771066639c2d5a6d35 --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2f5ec3f8a8314ecf22baa185412448d62ef05b62 --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_128_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..159e4c887b3a318c51428571f2609be8875899cd --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3d008d6aebff8c76ef8ad93ca5e1e48c6deb932b --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4bbd86b95bb7795442a5546ebab38cd680b845f7 --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ce6cd6d95d0a7c1905aabe792d0a5c4e2f515a --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-dtd-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b58e05cfeca75429e5274ec3a6364d3b74cf376f --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..13f9e0905aba09ddfb07306023093f9a4322f7a7 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6070ed69b48468be40ece2206375c7c5f7b4b9ff --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..13f9e0905aba09ddfb07306023093f9a4322f7a7 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..eefd66ed5587f6970966be69be2a373f0cb6f354 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..13f9e0905aba09ddfb07306023093f9a4322f7a7 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8a0bc9f32b9846f9fe270a070ae6466b5a81a442 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..13f9e0905aba09ddfb07306023093f9a4322f7a7 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-emnist_letters-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md b/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..dd353d6c9a959a371277eb805ac27ac0ccd009f4 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json b/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..13f9e0905aba09ddfb07306023093f9a4322f7a7 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-emnist_letters-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9d1a7b69575aefac7a71cda1ed083bf2857ada48 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..13f9e0905aba09ddfb07306023093f9a4322f7a7 --- /dev/null +++ b/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-emnist_letters-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f4b4c8334c2fce028ed33145e4b1d7e761b23c76 --- /dev/null +++ b/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..cf61d0357a8d4fef0a1593df0f7b561cb9c545a1 --- /dev/null +++ b/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-eurosat-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/README.md b/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..43be32a09a314609c1853e7948db7d017efc187d --- /dev/null +++ b/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/config.json b/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..cf61d0357a8d4fef0a1593df0f7b561cb9c545a1 --- /dev/null +++ b/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..cc5bb681ca738270c858dd2c66d5cbd93cd32906 --- /dev/null +++ b/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..cf61d0357a8d4fef0a1593df0f7b561cb9c545a1 --- /dev/null +++ b/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-eurosat-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/README.md b/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..44eda2d430cb746737c92b4b80d3bba01ef263f1 --- /dev/null +++ b/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/config.json b/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..cf61d0357a8d4fef0a1593df0f7b561cb9c545a1 --- /dev/null +++ b/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-eurosat-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..cb304519918c7456e43dc54bd5f959e371d9f0cb --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..cb304519918c7456e43dc54bd5f959e371d9f0cb --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..40c17a79fd8ea745e01c005d3d3cd5158f8a8e9e --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..40c17a79fd8ea745e01c005d3d3cd5158f8a8e9e --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b4f3fbdd67e11339992c84492ff7bd2fe3b0b0d1 --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d8ec8f942a38313b3fd420fb8ed8557e1c0f8305 --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2a07b0b46f4614b333df1abe3ff2def4f41a3124 --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2a07b0b46f4614b333df1abe3ff2def4f41a3124 --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..95053393425e99bf8747d9695556f752509f4b0b --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fashion_mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2062580452c7b5265a6c0b2196854379d150d098 --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md b/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a200119f1c27aa173487977da3bd43d44a8e043f --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json b/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ee92d52ec9d4ec324182c69a7f2a5bb1cd25c49b --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_128_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md b/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..388758c91f8f7f7513224e09df1f0f42dcba012e --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json b/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..99aa4c77f48a2dbd9cf0e9be65a01057e05d2eda --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3bdfa3cc92d4ffbf0637f7d54e237dd9a6d72ea0 --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e2e2421f6623efcac518b0a894c759827bf41381 --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..85fbd7f3a9c3294db12c949fd218582c9bb398ae --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..22e59066ad230eec264387b1be102e8570be9eb3 --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce988daea92558cbffcf8c943ef8039e1e9a5ec2 --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-fer2013-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md b/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b4d674c2de5d2f82246866380d728493ccb04310 --- /dev/null +++ b/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json b/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_128_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..787dc7b6a074f43f57cd8ecf24b4e5ae6fc1ffc0 --- /dev/null +++ b/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..02d19bf3d9c40ecb83be4e7aae5e1154189a0bf8 --- /dev/null +++ b/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ed6f08a0f0be6a97374f36f364089108155a5a9f --- /dev/null +++ b/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..fafef6980400d53d766b9fc8238f37fd8dd902aa --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d59ec223d7c9aaf32cad6630ea2f5209610af02b --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..28a2df107a9aa1acc3fabbafbcfb3f15a7ec1db2 --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e857d2a3c5d3032483c647844c95ab36e288c8f5 --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2bba7874d0acf5076869a6fd62eb0fbe2d2041dc --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-food101-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b5a8f730e4c1761c1101bb34890b7c718e7419b6 --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/README.md b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..74e48924bea9084dacd084e3f1b56decf2539df1 --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/config.json b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d92071e1c5de28cbd6990e26bda79549de85c7f0 --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-gtsrb-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..bf8cb4fb462bf2784239a95a905d4581d1c23972 --- /dev/null +++ b/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-gtsrb-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..db3405a5a270f6ec12603369bb439f0f3fffe77f --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..bd0fa88a93f315700a0973edb4e7184f89df10c4 --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d3ed5e229e33c090e719bd3ae7bef40e50057581 --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..62569c7f31f507ee6f3882e2441b38c75ccc59f3 --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9de0aad1bc05bc7b91e5994c0e58446bbdc72606 --- /dev/null +++ b/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..54cae3827bd20f790075f990d3712f8342d20c79 --- /dev/null +++ b/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-kmnist-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..57f9349edfbba27dc1cfb5fda7ff3fadd416e9ce --- /dev/null +++ b/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..54cae3827bd20f790075f990d3712f8342d20c79 --- /dev/null +++ b/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..fb58d33ad4f26beef04724bacb9ba1243e42dd22 --- /dev/null +++ b/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..54cae3827bd20f790075f990d3712f8342d20c79 --- /dev/null +++ b/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-kmnist-batch_size_64_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md b/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..22fba3adf4826ec10d102618dfed50979214c417 --- /dev/null +++ b/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json b/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..54cae3827bd20f790075f990d3712f8342d20c79 --- /dev/null +++ b/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-kmnist-batch_size_64_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/README.md b/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4e6002194ec078fe1da0cedde9c16e8e171281c8 --- /dev/null +++ b/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: mnist + pretrained: true +test_datasets: null +train_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: train +type: transformers +val_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: test +``` + diff --git a/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/config.json b/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..018e7d4d1ef4a7995d309411e18d7bc5d8d5f4a9 --- /dev/null +++ b/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ee299e3cd0d28371cf52885a12e798d7d5f73841 --- /dev/null +++ b/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: mnist + pretrained: true +test_datasets: null +train_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: train +type: transformers +val_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: test +``` + diff --git a/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..018e7d4d1ef4a7995d309411e18d7bc5d8d5f4a9 --- /dev/null +++ b/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-mnist-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md b/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4ab3cc5674f7f1fac3de716639ebbbf7013a3500 --- /dev/null +++ b/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: mnist + pretrained: true +test_datasets: null +train_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: train +type: transformers +val_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: test +``` + diff --git a/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json b/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..018e7d4d1ef4a7995d309411e18d7bc5d8d5f4a9 --- /dev/null +++ b/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-mnist-batch_size_64_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/README.md b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..5e9200639dd20d5ec1e019e2aa8a6d7f4c5f4c72 --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/config.json b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fd54d9408eb0bfa96cb8fb062f40297ff8e043a3 --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..100afb94797a0a0c05ebec13e027faaa707d5ca4 --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fd54d9408eb0bfa96cb8fb062f40297ff8e043a3 --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e85a54f655341f8db5cedecf58cc6c3b16deff51 --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fd54d9408eb0bfa96cb8fb062f40297ff8e043a3 --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..11b76c284b1bcb0ff3314b7cadac0d6016f4eebf --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fd54d9408eb0bfa96cb8fb062f40297ff8e043a3 --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ee58cd7d8fecdf824eb5a302708ed4c92b59f710 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f938618d830ad3512a63af4ea0f8ca5e6b26a242 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..77bd689e0f2ac41eac2c231f7aeef5f5b507f7b5 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f938618d830ad3512a63af4ea0f8ca5e6b26a242 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7deba0858c042a83e7f1af5619e6ca0d2a7db30f --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f938618d830ad3512a63af4ea0f8ca5e6b26a242 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..56dc9d58df84f6f2b7b9fbac1214b02d63903d9e --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f938618d830ad3512a63af4ea0f8ca5e6b26a242 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7c0396af621941375188a6cede1cb6f42f734294 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f938618d830ad3512a63af4ea0f8ca5e6b26a242 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..54b5a688fc3aec6c2ef318c9b84f1f7c7ffbd931 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f938618d830ad3512a63af4ea0f8ca5e6b26a242 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ac40b54a9ba0e0f3590a082921fe7cc882054861 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f938618d830ad3512a63af4ea0f8ca5e6b26a242 --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f3bba378b7afae1e6367c27dd6aa9a1dedde2b01 --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..af7ef4427b07ea4c1e108606639855234dfa1a8d --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..103221d9a35baf9aeb95fd9e555f8e73234301d4 --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..af7ef4427b07ea4c1e108606639855234dfa1a8d --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..487f68f32bd03553ce8763e919457db846cf89b3 --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..487f68f32bd03553ce8763e919457db846cf89b3 --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..af7ef4427b07ea4c1e108606639855234dfa1a8d --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-pcam-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/README.md b/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b1f6e2569481bd78f9f619ea8b93e4117006baea --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/config.json b/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7de77b1b23edcbdb1f59b2b21726aa56f03fbfdb --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8a9ee3044a145ad8827344f22ad0593e54eeae45 --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7de77b1b23edcbdb1f59b2b21726aa56f03fbfdb --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..31f73042c5c30b82c7a58af5e0e93d38a03ddb5b --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7de77b1b23edcbdb1f59b2b21726aa56f03fbfdb --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a19d753957022b3eb575c71e7255e217c161e6ae --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7de77b1b23edcbdb1f59b2b21726aa56f03fbfdb --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a409370e37b70b2c99b9efffb15caca2efe067e6 --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7de77b1b23edcbdb1f59b2b21726aa56f03fbfdb --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md b/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..67e6a3f25aa679dfbead614215a0670757238607 --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json b/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7de77b1b23edcbdb1f59b2b21726aa56f03fbfdb --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..90c395d128519c14d3de1b6fda0f5891b0261518 --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7de77b1b23edcbdb1f59b2b21726aa56f03fbfdb --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e8d4360844b428fb79361c83e3e88b8215e8b606 --- /dev/null +++ b/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..92840a13704fe0d58f3cc8774361fed5e980dc70 --- /dev/null +++ b/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-resisc45-batch_size_128_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c99d827e5f38fbb645e24408d4ee4da604d6d1d8 --- /dev/null +++ b/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..92840a13704fe0d58f3cc8774361fed5e980dc70 --- /dev/null +++ b/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-resisc45-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/README.md b/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..42f86ccd359bff5102ae53c3a64acafa8f998698 --- /dev/null +++ b/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/config.json b/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ea904dca37c3e1bd676cdab8c844dff369a4e286 --- /dev/null +++ b/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stanford-cars-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md b/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..386c7584cdd37e1f1b6ab5518708db3326b49925 --- /dev/null +++ b/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json b/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ea904dca37c3e1bd676cdab8c844dff369a4e286 --- /dev/null +++ b/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stanford-cars-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/README.md b/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..49ccebcffeea8d03df7419e1c1de629573b7a0b0 --- /dev/null +++ b/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/config.json b/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ea904dca37c3e1bd676cdab8c844dff369a4e286 --- /dev/null +++ b/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c16f86ca41b0f65351d030d672412dace5cbb168 --- /dev/null +++ b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..29c93b2c12685a5b5caf379e93026d934c412298 --- /dev/null +++ b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c16f86ca41b0f65351d030d672412dace5cbb168 --- /dev/null +++ b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..29c93b2c12685a5b5caf379e93026d934c412298 --- /dev/null +++ b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stl10-batch_size_128_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..54ca01c4ce66340debe2a855e57d0b8608fd8b86 --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..29c93b2c12685a5b5caf379e93026d934c412298 --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c9113682c7b2afca8ddb52dbdd0cf01584b1ae35 --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..29c93b2c12685a5b5caf379e93026d934c412298 --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..5db23dec110254bbadd12966499d4d9f56a5f598 --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..29c93b2c12685a5b5caf379e93026d934c412298 --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stl10-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..5ec8010f3e2669c0054d724a781ab82e7103513a --- /dev/null +++ b/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..29c93b2c12685a5b5caf379e93026d934c412298 --- /dev/null +++ b/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-stl10-batch_size_64_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-stl10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet18-stl10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..29c93b2c12685a5b5caf379e93026d934c412298 --- /dev/null +++ b/resnet18-stl10-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..45116283da7a8a58a4d726f0c2e22891c463d504 --- /dev/null +++ b/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..91d95b05ddb0695139a119b3e9795bc4ffaad39d --- /dev/null +++ b/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0d64b076319f6c1ee4eec26033bef27fde7c6671 --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/README.md b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c269195caea5618716d278ed25479619f26306bc --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/config.json b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c269195caea5618716d278ed25479619f26306bc --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..137b8f7423dd5589da7f354b6de1a3b7e624964d --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..567e988a7b27e133e691210ba4b962def7d3cc1c --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md b/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f3391716cb2e82edc04c7ba7c04e716e18b15e23 --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json b/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d6ab5c4dc220546478ce5918919b5ca9db5bd2c4 --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..cb3943181dbf0aaf4206e5e4301d053f3ba34d41 --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..886da3497a7114ff31c2408ec559a773285de738 --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-sun397-batch_size_64_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f9503cdc1fff7a952c9731049690b86b5f256087 --- /dev/null +++ b/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..018e7d4d1ef4a7995d309411e18d7bc5d8d5f4a9 --- /dev/null +++ b/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-svhn-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/README.md b/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9948241d54e71345c78e8584706d6acf3a3cdb2c --- /dev/null +++ b/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/config.json b/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..018e7d4d1ef4a7995d309411e18d7bc5d8d5f4a9 --- /dev/null +++ b/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-svhn-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..53596b41d0b2f9de877c978c228f90a07176da9c --- /dev/null +++ b/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-18 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-18 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..018e7d4d1ef4a7995d309411e18d7bc5d8d5f4a9 --- /dev/null +++ b/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 2, + 2, + 2, + 2 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 64, + 128, + 256, + 512 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "basic", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet18-svhn-batch_size_64_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a61fd76e22dd732711a16b4551668f74593936c1 --- /dev/null +++ b/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1ff9b42153446695881aa454c47f61b2f1841aed --- /dev/null +++ b/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_128_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/README.md b/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a84f571ce87f5acf266851f85ee5f81c29b00c47 --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/config.json b/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2b196b5bc9f5d377699c75b0de1c9f5fbe0bb3cb --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3accabd97c2bea652e16d70949c86917e999b9d4 --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9515ca40c69e99770e058c97d6497fc426bcfa46 --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..305e6dadd068ce46d65b0a4cc929c96d6105d8f1 --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..accd7c370f31ebee6024929454d0efe4b5f054f6 --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a4fe5d1c3301943a00598ea79401ddb261e699d7 --- /dev/null +++ b/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_64_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ff6e4b8dc941bf4b842d64b75189f15253336b61 --- /dev/null +++ b/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar10 + pretrained: true +test_datasets: null +train_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: train +type: transformers +val_datasets: + cifar10: + _target_: datasets.load_dataset + path: tanganke/cifar10 + split: test +``` + diff --git a/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3c885e0d0eac58ee224a40f544b024413974088f --- /dev/null +++ b/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "automobile", + "2": "bird", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "frog", + "7": "horse", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "automobile": 1, + "bird": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "frog": 6, + "horse": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar10-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b8186a90ad6eb362ef0de9546029463a5ff86605 --- /dev/null +++ b/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4eba7ae9631d2b0e864ded80157f76e37582afd4 --- /dev/null +++ b/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar100-batch_size_128_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/README.md b/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6035a866e12aac16ca4ff4a1e8f248bc936546cb --- /dev/null +++ b/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/config.json b/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4eba7ae9631d2b0e864ded80157f76e37582afd4 --- /dev/null +++ b/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar100-batch_size_256_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md b/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..11a92642d13eeb7f663240e108ad998769411124 --- /dev/null +++ b/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json b/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4eba7ae9631d2b0e864ded80157f76e37582afd4 --- /dev/null +++ b/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar100-batch_size_256_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a2f74a5245a96b95632fb4943cb4a84e4e6db540 --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4eba7ae9631d2b0e864ded80157f76e37582afd4 --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c258d0277de4c336312c77cb93d7f8785388bc0f --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4eba7ae9631d2b0e864ded80157f76e37582afd4 --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..71020878478cfadcb01c3402acc9a0e6edcbd99c --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset cifar100. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: cifar100 + pretrained: true +test_datasets: null +train_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: train +type: transformers +val_datasets: + cifar100: + _target_: datasets.load_dataset + path: tanganke/cifar100 + split: test +``` + diff --git a/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4eba7ae9631d2b0e864ded80157f76e37582afd4 --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,243 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple", + "1": "aquarium_fish", + "2": "baby", + "3": "bear", + "4": "beaver", + "5": "bed", + "6": "bee", + "7": "beetle", + "8": "bicycle", + "9": "bottle", + "10": "bowl", + "11": "boy", + "12": "bridge", + "13": "bus", + "14": "butterfly", + "15": "camel", + "16": "can", + "17": "castle", + "18": "caterpillar", + "19": "cattle", + "20": "chair", + "21": "chimpanzee", + "22": "clock", + "23": "cloud", + "24": "cockroach", + "25": "couch", + "26": "cra", + "27": "crocodile", + "28": "cup", + "29": "dinosaur", + "30": "dolphin", + "31": "elephant", + "32": "flatfish", + "33": "forest", + "34": "fox", + "35": "girl", + "36": "hamster", + "37": "house", + "38": "kangaroo", + "39": "keyboard", + "40": "lamp", + "41": "lawn_mower", + "42": "leopard", + "43": "lion", + "44": "lizard", + "45": "lobster", + "46": "man", + "47": "maple_tree", + "48": "motorcycle", + "49": "mountain", + "50": "mouse", + "51": "mushroom", + "52": "oak_tree", + "53": "orange", + "54": "orchid", + "55": "otter", + "56": "palm_tree", + "57": "pear", + "58": "pickup_truck", + "59": "pine_tree", + "60": "plain", + "61": "plate", + "62": "poppy", + "63": "porcupine", + "64": "possum", + "65": "rabbit", + "66": "raccoon", + "67": "ray", + "68": "road", + "69": "rocket", + "70": "rose", + "71": "sea", + "72": "seal", + "73": "shark", + "74": "shrew", + "75": "skunk", + "76": "skyscraper", + "77": "snail", + "78": "snake", + "79": "spider", + "80": "squirrel", + "81": "streetcar", + "82": "sunflower", + "83": "sweet_pepper", + "84": "table", + "85": "tank", + "86": "telephone", + "87": "television", + "88": "tiger", + "89": "tractor", + "90": "train", + "91": "trout", + "92": "tulip", + "93": "turtle", + "94": "wardrobe", + "95": "whale", + "96": "willow_tree", + "97": "wolf", + "98": "woman", + "99": "worm" + }, + "label2id": { + "apple": 0, + "aquarium_fish": 1, + "baby": 2, + "bear": 3, + "beaver": 4, + "bed": 5, + "bee": 6, + "beetle": 7, + "bicycle": 8, + "bottle": 9, + "bowl": 10, + "boy": 11, + "bridge": 12, + "bus": 13, + "butterfly": 14, + "camel": 15, + "can": 16, + "castle": 17, + "caterpillar": 18, + "cattle": 19, + "chair": 20, + "chimpanzee": 21, + "clock": 22, + "cloud": 23, + "cockroach": 24, + "couch": 25, + "cra": 26, + "crocodile": 27, + "cup": 28, + "dinosaur": 29, + "dolphin": 30, + "elephant": 31, + "flatfish": 32, + "forest": 33, + "fox": 34, + "girl": 35, + "hamster": 36, + "house": 37, + "kangaroo": 38, + "keyboard": 39, + "lamp": 40, + "lawn_mower": 41, + "leopard": 42, + "lion": 43, + "lizard": 44, + "lobster": 45, + "man": 46, + "maple_tree": 47, + "motorcycle": 48, + "mountain": 49, + "mouse": 50, + "mushroom": 51, + "oak_tree": 52, + "orange": 53, + "orchid": 54, + "otter": 55, + "palm_tree": 56, + "pear": 57, + "pickup_truck": 58, + "pine_tree": 59, + "plain": 60, + "plate": 61, + "poppy": 62, + "porcupine": 63, + "possum": 64, + "rabbit": 65, + "raccoon": 66, + "ray": 67, + "road": 68, + "rocket": 69, + "rose": 70, + "sea": 71, + "seal": 72, + "shark": 73, + "shrew": 74, + "skunk": 75, + "skyscraper": 76, + "snail": 77, + "snake": 78, + "spider": 79, + "squirrel": 80, + "streetcar": 81, + "sunflower": 82, + "sweet_pepper": 83, + "table": 84, + "tank": 85, + "telephone": 86, + "television": 87, + "tiger": 88, + "tractor": 89, + "train": 90, + "trout": 91, + "tulip": 92, + "turtle": 93, + "wardrobe": 94, + "whale": 95, + "willow_tree": 96, + "wolf": 97, + "woman": 98, + "worm": 99 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-cifar100-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/README.md b/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e02b6eb0be10204aeb36592636562825161033de --- /dev/null +++ b/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/config.json b/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9be9d43cf554cd4001bc241de10a3998fe29da7d --- /dev/null +++ b/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-dtd-batch_size_256_lr_0.001_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..135a26e1f5912d24cf6edca44c237898cdd07ec7 --- /dev/null +++ b/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset dtd. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: dtd + pretrained: true +test_datasets: null +train_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: train +type: transformers +val_datasets: + dtd: + _target_: datasets.load_dataset + path: tanganke/dtd + split: test +``` + diff --git a/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9be9d43cf554cd4001bc241de10a3998fe29da7d --- /dev/null +++ b/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,137 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "banded", + "1": "blotchy", + "2": "braided", + "3": "bubbly", + "4": "bumpy", + "5": "chequered", + "6": "cobwebbed", + "7": "cracked", + "8": "crosshatched", + "9": "crystalline", + "10": "dotted", + "11": "fibrous", + "12": "flecked", + "13": "freckled", + "14": "frilly", + "15": "gauzy", + "16": "grid", + "17": "grooved", + "18": "honeycombed", + "19": "interlaced", + "20": "knitted", + "21": "lacelike", + "22": "lined", + "23": "marbled", + "24": "matted", + "25": "meshed", + "26": "paisley", + "27": "perforated", + "28": "pitted", + "29": "pleated", + "30": "polka-dotted", + "31": "porous", + "32": "potholed", + "33": "scaly", + "34": "smeared", + "35": "spiralled", + "36": "sprinkled", + "37": "stained", + "38": "stratified", + "39": "striped", + "40": "studded", + "41": "swirly", + "42": "veined", + "43": "waffled", + "44": "woven", + "45": "wrinkled", + "46": "zigzagged" + }, + "label2id": { + "banded": 0, + "blotchy": 1, + "braided": 2, + "bubbly": 3, + "bumpy": 4, + "chequered": 5, + "cobwebbed": 6, + "cracked": 7, + "crosshatched": 8, + "crystalline": 9, + "dotted": 10, + "fibrous": 11, + "flecked": 12, + "freckled": 13, + "frilly": 14, + "gauzy": 15, + "grid": 16, + "grooved": 17, + "honeycombed": 18, + "interlaced": 19, + "knitted": 20, + "lacelike": 21, + "lined": 22, + "marbled": 23, + "matted": 24, + "meshed": 25, + "paisley": 26, + "perforated": 27, + "pitted": 28, + "pleated": 29, + "polka-dotted": 30, + "porous": 31, + "potholed": 32, + "scaly": 33, + "smeared": 34, + "spiralled": 35, + "sprinkled": 36, + "stained": 37, + "stratified": 38, + "striped": 39, + "studded": 40, + "swirly": 41, + "veined": 42, + "waffled": 43, + "woven": 44, + "wrinkled": 45, + "zigzagged": 46 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-dtd-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4e1848d914642ede35edc7a3ec788e4931031406 --- /dev/null +++ b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9d1862748747d6d484a29b20c512dd46535c93b1 --- /dev/null +++ b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4e1848d914642ede35edc7a3ec788e4931031406 --- /dev/null +++ b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9d1862748747d6d484a29b20c512dd46535c93b1 --- /dev/null +++ b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-emnist_letters-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md b/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..78e7798ed11d9beb29887e37221bbe8060cd3c07 --- /dev/null +++ b/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json b/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9d1862748747d6d484a29b20c512dd46535c93b1 --- /dev/null +++ b/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-emnist_letters-batch_size_64_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..579a6d207e863510504a704f130d146a942e96ea --- /dev/null +++ b/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset emnist_letters. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: emnist_letters + pretrained: true +test_datasets: null +train_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: train +type: transformers +val_datasets: + emnist_letters: + _target_: datasets.load_dataset + path: tanganke/emnist_letters + split: test +``` + diff --git a/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9d1862748747d6d484a29b20c512dd46535c93b1 --- /dev/null +++ b/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,95 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "A", + "1": "B", + "2": "C", + "3": "D", + "4": "E", + "5": "F", + "6": "G", + "7": "H", + "8": "I", + "9": "J", + "10": "K", + "11": "L", + "12": "M", + "13": "N", + "14": "O", + "15": "P", + "16": "Q", + "17": "R", + "18": "S", + "19": "T", + "20": "U", + "21": "V", + "22": "W", + "23": "X", + "24": "Y", + "25": "Z" + }, + "label2id": { + "A": 0, + "B": 1, + "C": 2, + "D": 3, + "E": 4, + "F": 5, + "G": 6, + "H": 7, + "I": 8, + "J": 9, + "K": 10, + "L": 11, + "M": 12, + "N": 13, + "O": 14, + "P": 15, + "Q": 16, + "R": 17, + "S": 18, + "T": 19, + "U": 20, + "V": 21, + "W": 22, + "X": 23, + "Y": 24, + "Z": 25 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-emnist_letters-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3a4c815ae1ad7b99ab79f03f0d7418352a8cb6af --- /dev/null +++ b/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b8f5a722ee7ec6f4a2d02ac651ae80260967eae --- /dev/null +++ b/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-eurosat-batch_size_128_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md b/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..5a4430c99d15c3990f540e1f1e0f4c5b5d5d406b --- /dev/null +++ b/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json b/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b8f5a722ee7ec6f4a2d02ac651ae80260967eae --- /dev/null +++ b/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "annual crop land", + "1": "forest", + "2": "brushland or shrubland", + "3": "highway or road", + "4": "industrial buildings or commercial buildings", + "5": "pasture land", + "6": "permanent crop land", + "7": "residential buildings or homes or apartments", + "8": "river", + "9": "lake or sea" + }, + "label2id": { + "annual crop land": 0, + "brushland or shrubland": 2, + "forest": 1, + "highway or road": 3, + "industrial buildings or commercial buildings": 4, + "lake or sea": 9, + "pasture land": 5, + "permanent crop land": 6, + "residential buildings or homes or apartments": 7, + "river": 8 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-eurosat-batch_size_256_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-eurosat-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet50-eurosat-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4ae92c18343867154602f143c4fea7c0e630255a --- /dev/null +++ b/resnet50-eurosat-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset eurosat. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: eurosat + pretrained: true +test_datasets: null +train_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: train +type: transformers +val_datasets: + eurosat: + _target_: datasets.load_dataset + path: tanganke/eurosat + split: test +``` + diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..135116bef15d671656a6628252c41f2b56024617 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55bcbf447ad396e78893f8c1d15e0049595b04f9 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2532f7b431bcffe119e09ccb9b2a2d64c8a35f12 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55bcbf447ad396e78893f8c1d15e0049595b04f9 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/README.md b/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9aa939c555af3fd3b7132b040f644d86373aa9b8 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/config.json b/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55bcbf447ad396e78893f8c1d15e0049595b04f9 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_128_lr_0.005_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/README.md b/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..76a046a321cd27a722589f73ad4714406209cb2a --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/config.json b/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55bcbf447ad396e78893f8c1d15e0049595b04f9 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md b/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..09eedade45f7f30700b9340b37d15c42c318ccd7 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json b/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55bcbf447ad396e78893f8c1d15e0049595b04f9 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..46e470493849fde4fdb11b43455f3ce5f43069c1 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fashion_mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fashion_mnist + pretrained: true +test_datasets: null +train_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: train +type: transformers +val_datasets: + fashion_mnist: + _target_: datasets.load_dataset + path: zalando-datasets/fashion_mnist + split: test +``` + diff --git a/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..55bcbf447ad396e78893f8c1d15e0049595b04f9 --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "T - shirt / top", + "1": "Trouser", + "2": "Pullover", + "3": "Dress", + "4": "Coat", + "5": "Sandal", + "6": "Shirt", + "7": "Sneaker", + "8": "Bag", + "9": "Ankle boot" + }, + "label2id": { + "Ankle boot": 9, + "Bag": 8, + "Coat": 4, + "Dress": 3, + "Pullover": 2, + "Sandal": 5, + "Shirt": 6, + "Sneaker": 7, + "T - shirt / top": 0, + "Trouser": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fashion_mnist-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8eb70e05bb95be714f21c6ef613f7fe41ab8cd45 --- /dev/null +++ b/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f25f031e5d55136af7a32fa76d4d09b952d788fc --- /dev/null +++ b/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fer2013-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4e1c43b41971aec8ceda60a954018d1df206972e --- /dev/null +++ b/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,74 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset fer2013. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: fer2013 + pretrained: true +test_datasets: null +train_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: train +type: transformers +val_datasets: + fer2013: + _target_: fusion_bench.dataset.fer2013.load_fer2013 + split: test +``` + diff --git a/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f25f031e5d55136af7a32fa76d4d09b952d788fc --- /dev/null +++ b/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,57 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "angry", + "1": "disgusted", + "2": "fearful", + "3": "happy", + "4": "neutral", + "5": "sad", + "6": "surprised" + }, + "label2id": { + "angry": 0, + "disgusted": 1, + "fearful": 2, + "happy": 3, + "neutral": 4, + "sad": 5, + "surprised": 6 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-fer2013-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7e2eb2bccd78ad0d29e90ecd9e2f09021526bfcf --- /dev/null +++ b/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..d9f442b44934b43685324cae134e295127e8bfb3 --- /dev/null +++ b/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-food101-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4927ff3eddefd53aacdff52fe1f1e1a126797810 --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..d9f442b44934b43685324cae134e295127e8bfb3 --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9ac788616883574ac2788eb8ebf5e623cb2db132 --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset food101. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: food101 + pretrained: true +test_datasets: null +train_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: train +type: transformers +val_datasets: + food101: + _target_: datasets.load_dataset + path: ethz/food101 + split: validation +``` + diff --git a/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..d9f442b44934b43685324cae134e295127e8bfb3 --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-food101-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet50-food101-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..d9f442b44934b43685324cae134e295127e8bfb3 --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,245 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "apple pie", + "1": "baby back ribs", + "2": "baklava", + "3": "beef carpaccio", + "4": "beef tartare", + "5": "beet salad", + "6": "beignets", + "7": "bibimbap", + "8": "bread pudding", + "9": "breakfast burrito", + "10": "bruschetta", + "11": "caesar salad", + "12": "cannoli", + "13": "caprese salad", + "14": "carrot cake", + "15": "ceviche", + "16": "cheese plate", + "17": "cheesecake", + "18": "chicken curry", + "19": "chicken quesadilla", + "20": "chicken wings", + "21": "chocolate cake", + "22": "chocolate mousse", + "23": "churros", + "24": "clam chowder", + "25": "club sandwich", + "26": "crab cakes", + "27": "creme brulee", + "28": "croque madame", + "29": "cup cakes", + "30": "deviled eggs", + "31": "donuts", + "32": "dumplings", + "33": "edamame", + "34": "eggs benedict", + "35": "escargots", + "36": "falafel", + "37": "filet mignon", + "38": "fish and chips", + "39": "foie gras", + "40": "french fries", + "41": "french onion soup", + "42": "french toast", + "43": "fried calamari", + "44": "fried rice", + "45": "frozen yogurt", + "46": "garlic bread", + "47": "gnocchi", + "48": "greek salad", + "49": "grilled cheese sandwich", + "50": "grilled salmon", + "51": "guacamole", + "52": "gyoza", + "53": "hamburger", + "54": "hot and sour soup", + "55": "hot dog", + "56": "huevos rancheros", + "57": "hummus", + "58": "ice cream", + "59": "lasagna", + "60": "lobster bisque", + "61": "lobster roll sandwich", + "62": "macaroni and cheese", + "63": "macarons", + "64": "miso soup", + "65": "mussels", + "66": "nachos", + "67": "omelette", + "68": "onion rings", + "69": "oysters", + "70": "pad thai", + "71": "paella", + "72": "pancakes", + "73": "panna cotta", + "74": "peking duck", + "75": "pho", + "76": "pizza", + "77": "pork chop", + "78": "poutine", + "79": "prime rib", + "80": "pulled pork sandwich", + "81": "ramen", + "82": "ravioli", + "83": "red velvet cake", + "84": "risotto", + "85": "samosa", + "86": "sashimi", + "87": "scallops", + "88": "seaweed salad", + "89": "shrimp and grits", + "90": "spaghetti bolognese", + "91": "spaghetti carbonara", + "92": "spring rolls", + "93": "steak", + "94": "strawberry shortcake", + "95": "sushi", + "96": "tacos", + "97": "takoyaki", + "98": "tiramisu", + "99": "tuna tartare", + "100": "waffles" + }, + "label2id": { + "apple pie": 0, + "baby back ribs": 1, + "baklava": 2, + "beef carpaccio": 3, + "beef tartare": 4, + "beet salad": 5, + "beignets": 6, + "bibimbap": 7, + "bread pudding": 8, + "breakfast burrito": 9, + "bruschetta": 10, + "caesar salad": 11, + "cannoli": 12, + "caprese salad": 13, + "carrot cake": 14, + "ceviche": 15, + "cheese plate": 16, + "cheesecake": 17, + "chicken curry": 18, + "chicken quesadilla": 19, + "chicken wings": 20, + "chocolate cake": 21, + "chocolate mousse": 22, + "churros": 23, + "clam chowder": 24, + "club sandwich": 25, + "crab cakes": 26, + "creme brulee": 27, + "croque madame": 28, + "cup cakes": 29, + "deviled eggs": 30, + "donuts": 31, + "dumplings": 32, + "edamame": 33, + "eggs benedict": 34, + "escargots": 35, + "falafel": 36, + "filet mignon": 37, + "fish and chips": 38, + "foie gras": 39, + "french fries": 40, + "french onion soup": 41, + "french toast": 42, + "fried calamari": 43, + "fried rice": 44, + "frozen yogurt": 45, + "garlic bread": 46, + "gnocchi": 47, + "greek salad": 48, + "grilled cheese sandwich": 49, + "grilled salmon": 50, + "guacamole": 51, + "gyoza": 52, + "hamburger": 53, + "hot and sour soup": 54, + "hot dog": 55, + "huevos rancheros": 56, + "hummus": 57, + "ice cream": 58, + "lasagna": 59, + "lobster bisque": 60, + "lobster roll sandwich": 61, + "macaroni and cheese": 62, + "macarons": 63, + "miso soup": 64, + "mussels": 65, + "nachos": 66, + "omelette": 67, + "onion rings": 68, + "oysters": 69, + "pad thai": 70, + "paella": 71, + "pancakes": 72, + "panna cotta": 73, + "peking duck": 74, + "pho": 75, + "pizza": 76, + "pork chop": 77, + "poutine": 78, + "prime rib": 79, + "pulled pork sandwich": 80, + "ramen": 81, + "ravioli": 82, + "red velvet cake": 83, + "risotto": 84, + "samosa": 85, + "sashimi": 86, + "scallops": 87, + "seaweed salad": 88, + "shrimp and grits": 89, + "spaghetti bolognese": 90, + "spaghetti carbonara": 91, + "spring rolls": 92, + "steak": 93, + "strawberry shortcake": 94, + "sushi": 95, + "tacos": 96, + "takoyaki": 97, + "tiramisu": 98, + "tuna tartare": 99, + "waffles": 100 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-food101-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-food101-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-food101-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4ea0f963ebc7c3b7792b9f2b43eab37e16daea30 --- /dev/null +++ b/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e5aa2e6ca1e8845fcf4a68dc7cd82ec3c90ca4 --- /dev/null +++ b/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-gtsrb-batch_size_128_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f23eceaab0e22dd25f390bbe9a9fd701044b0689 --- /dev/null +++ b/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e5aa2e6ca1e8845fcf4a68dc7cd82ec3c90ca4 --- /dev/null +++ b/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-gtsrb-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md b/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f61314be74c0f29eb9c88777a5bb2687276f441c --- /dev/null +++ b/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json b/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e5aa2e6ca1e8845fcf4a68dc7cd82ec3c90ca4 --- /dev/null +++ b/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-gtsrb-batch_size_256_lr_0.001_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..335084a4fb3472ac81560f51615cbb5fdee3204d --- /dev/null +++ b/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e5aa2e6ca1e8845fcf4a68dc7cd82ec3c90ca4 --- /dev/null +++ b/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-gtsrb-batch_size_256_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..21293688b65f01de752907ce3656a4ee161d5106 --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e5aa2e6ca1e8845fcf4a68dc7cd82ec3c90ca4 --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ccf36cfb19da52a4ad800430c319502c82aa0534 --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e5aa2e6ca1e8845fcf4a68dc7cd82ec3c90ca4 --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ccf36cfb19da52a4ad800430c319502c82aa0534 --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset gtsrb. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: gtsrb + pretrained: true +test_datasets: null +train_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: train +type: transformers +val_datasets: + gtsrb: + _target_: datasets.load_dataset + path: tanganke/gtsrb + split: test +``` + diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e5aa2e6ca1e8845fcf4a68dc7cd82ec3c90ca4 --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,129 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "red and white circle 20 kph speed limit", + "1": "red and white circle 30 kph speed limit", + "2": "red and white circle 50 kph speed limit", + "3": "red and white circle 60 kph speed limit", + "4": "red and white circle 70 kph speed limit", + "5": "red and white circle 80 kph speed limit", + "6": "end / de-restriction of 80 kph speed limit", + "7": "red and white circle 100 kph speed limit", + "8": "red and white circle 120 kph speed limit", + "9": "red and white circle red car and black car no passing", + "10": "red and white circle red truck and black car no passing", + "11": "red and white triangle road intersection warning", + "12": "white and yellow diamond priority road", + "13": "red and white upside down triangle yield right-of-way", + "14": "stop", + "15": "empty red and white circle", + "16": "red and white circle no truck entry", + "17": "red circle with white horizonal stripe no entry", + "18": "red and white triangle with exclamation mark warning", + "19": "red and white triangle with black left curve approaching warning", + "20": "red and white triangle with black right curve approaching warning", + "21": "red and white triangle with black double curve approaching warning", + "22": "red and white triangle rough / bumpy road warning", + "23": "red and white triangle car skidding / slipping warning", + "24": "red and white triangle with merging / narrow lanes warning", + "25": "red and white triangle with person digging / construction / road work warning", + "26": "red and white triangle with traffic light approaching warning", + "27": "red and white triangle with person walking warning", + "28": "red and white triangle with child and person walking warning", + "29": "red and white triangle with bicyle warning", + "30": "red and white triangle with snowflake / ice warning", + "31": "red and white triangle with deer warning", + "32": "white circle with gray strike bar no speed limit", + "33": "blue circle with white right turn arrow mandatory", + "34": "blue circle with white left turn arrow mandatory", + "35": "blue circle with white forward arrow mandatory", + "36": "blue circle with white forward or right turn arrow mandatory", + "37": "blue circle with white forward or left turn arrow mandatory", + "38": "blue circle with white keep right arrow mandatory", + "39": "blue circle with white keep left arrow mandatory", + "40": "blue circle with white arrows indicating a traffic circle", + "41": "white circle with gray strike bar indicating no passing for cars has ended", + "42": "white circle with gray strike bar indicating no passing for trucks has ended" + }, + "label2id": { + "blue circle with white arrows indicating a traffic circle": 40, + "blue circle with white forward arrow mandatory": 35, + "blue circle with white forward or left turn arrow mandatory": 37, + "blue circle with white forward or right turn arrow mandatory": 36, + "blue circle with white keep left arrow mandatory": 39, + "blue circle with white keep right arrow mandatory": 38, + "blue circle with white left turn arrow mandatory": 34, + "blue circle with white right turn arrow mandatory": 33, + "empty red and white circle": 15, + "end / de-restriction of 80 kph speed limit": 6, + "red and white circle 100 kph speed limit": 7, + "red and white circle 120 kph speed limit": 8, + "red and white circle 20 kph speed limit": 0, + "red and white circle 30 kph speed limit": 1, + "red and white circle 50 kph speed limit": 2, + "red and white circle 60 kph speed limit": 3, + "red and white circle 70 kph speed limit": 4, + "red and white circle 80 kph speed limit": 5, + "red and white circle no truck entry": 16, + "red and white circle red car and black car no passing": 9, + "red and white circle red truck and black car no passing": 10, + "red and white triangle car skidding / slipping warning": 23, + "red and white triangle road intersection warning": 11, + "red and white triangle rough / bumpy road warning": 22, + "red and white triangle with bicyle warning": 29, + "red and white triangle with black double curve approaching warning": 21, + "red and white triangle with black left curve approaching warning": 19, + "red and white triangle with black right curve approaching warning": 20, + "red and white triangle with child and person walking warning": 28, + "red and white triangle with deer warning": 31, + "red and white triangle with exclamation mark warning": 18, + "red and white triangle with merging / narrow lanes warning": 24, + "red and white triangle with person digging / construction / road work warning": 25, + "red and white triangle with person walking warning": 27, + "red and white triangle with snowflake / ice warning": 30, + "red and white triangle with traffic light approaching warning": 26, + "red and white upside down triangle yield right-of-way": 13, + "red circle with white horizonal stripe no entry": 17, + "stop": 14, + "white and yellow diamond priority road": 12, + "white circle with gray strike bar indicating no passing for cars has ended": 41, + "white circle with gray strike bar indicating no passing for trucks has ended": 42, + "white circle with gray strike bar no speed limit": 32 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-gtsrb-batch_size_64_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..61ab6d70e77853299d65c3d7471f23ee8c88a49d --- /dev/null +++ b/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7cb9e38ace4738e265b487d7c814cd0f7cdbeae1 --- /dev/null +++ b/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-kmnist-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4aa8e4fde60e5cb1d7947c932b8b70f1ca862888 --- /dev/null +++ b/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset kmnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: kmnist + pretrained: true +test_datasets: null +train_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: train +type: transformers +val_datasets: + kmnist: + _target_: datasets.load_dataset + path: tanganke/kmnist + split: test +``` + diff --git a/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7cb9e38ace4738e265b487d7c814cd0f7cdbeae1 --- /dev/null +++ b/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "\u304a", + "1": "\u304d", + "2": "\u3059", + "3": "\u3064", + "4": "\u306a", + "5": "\u306f", + "6": "\u307e", + "7": "\u3084", + "8": "\u308c", + "9": "\u3092" + }, + "label2id": { + "\u304a": 0, + "\u304d": 1, + "\u3059": 2, + "\u3064": 3, + "\u306a": 4, + "\u306f": 5, + "\u307e": 6, + "\u3084": 7, + "\u308c": 8, + "\u3092": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-kmnist-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/README.md b/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..99378f55eeed6b56d5d0d620c05e263fee24138d --- /dev/null +++ b/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: mnist + pretrained: true +test_datasets: null +train_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: train +type: transformers +val_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: test +``` + diff --git a/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/config.json b/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-mnist-batch_size_128_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/README.md b/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1aaf5825ebdc8c02ff23a5c8152e93d872b111d7 --- /dev/null +++ b/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: mnist + pretrained: true +test_datasets: null +train_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: train +type: transformers +val_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: test +``` + diff --git a/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/config.json b/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-mnist-batch_size_256_lr_0.005_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e5a3cb593dd0b661e6b95332e06f77f8a20a5fa7 --- /dev/null +++ b/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset mnist. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: mnist + pretrained: true +test_datasets: null +train_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: train +type: transformers +val_datasets: + mnist: + _target_: datasets.load_dataset + path: mnist + split: test +``` + diff --git a/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-mnist-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ddb76d40e6b746f2a6cd30af4763bf4c668f75a2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6aa4e1b5c78beb32d8892e1606c0114e6f11cbec --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_128_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..28ee5ded9917d8c55447de14d4cb40cefca4997a --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..399b853affc72cd331a0538f0eaaae20179a092f --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..892e0690e6304189a75bfaa9192c08aa76fbf69f --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3213d125b594373ab751aa76814150667df159c2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..911bbd7813928e557b5f6db6048cf173aacaa799 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.001_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..23549ae15ab509b380fe54c9fd0036ee520e921b --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.005_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..aeaaea9f2c8e48c4e353999bf20edbc9f67d7f65 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford-iiit-pet. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford-iiit-pet + pretrained: true +test_datasets: null +train_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: train +type: transformers +val_datasets: + oxford-iiit-pet: + _target_: datasets.load_dataset + path: timm/oxford-iiit-pet + split: test +``` + diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..70c303978f48704b0d2ed76247d55f7c0a3fe6b2 --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,117 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abyssinian", + "1": "american_bulldog", + "2": "american_pit_bull_terrier", + "3": "basset_hound", + "4": "beagle", + "5": "bengal", + "6": "birman", + "7": "bombay", + "8": "boxer", + "9": "british_shorthair", + "10": "chihuahua", + "11": "egyptian_mau", + "12": "english_cocker_spaniel", + "13": "english_setter", + "14": "german_shorthaired", + "15": "great_pyrenees", + "16": "havanese", + "17": "japanese_chin", + "18": "keeshond", + "19": "leonberger", + "20": "maine_coon", + "21": "miniature_pinscher", + "22": "newfoundland", + "23": "persian", + "24": "pomeranian", + "25": "pug", + "26": "ragdoll", + "27": "russian_blue", + "28": "saint_bernard", + "29": "samoyed", + "30": "scottish_terrier", + "31": "shiba_inu", + "32": "siamese", + "33": "sphynx", + "34": "staffordshire_bull_terrier", + "35": "wheaten_terrier", + "36": "yorkshire_terrier" + }, + "label2id": { + "abyssinian": 0, + "american_bulldog": 1, + "american_pit_bull_terrier": 2, + "basset_hound": 3, + "beagle": 4, + "bengal": 5, + "birman": 6, + "bombay": 7, + "boxer": 8, + "british_shorthair": 9, + "chihuahua": 10, + "egyptian_mau": 11, + "english_cocker_spaniel": 12, + "english_setter": 13, + "german_shorthaired": 14, + "great_pyrenees": 15, + "havanese": 16, + "japanese_chin": 17, + "keeshond": 18, + "leonberger": 19, + "maine_coon": 20, + "miniature_pinscher": 21, + "newfoundland": 22, + "persian": 23, + "pomeranian": 24, + "pug": 25, + "ragdoll": 26, + "russian_blue": 27, + "saint_bernard": 28, + "samoyed": 29, + "scottish_terrier": 30, + "shiba_inu": 31, + "siamese": 32, + "sphynx": 33, + "staffordshire_bull_terrier": 34, + "wheaten_terrier": 35, + "yorkshire_terrier": 36 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford-iiit-pet-batch_size_64_lr_0.01_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/README.md b/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1b439fb887b5bca3e60187afcfeb82a5cd56afa3 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/config.json b/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3265e96f4c38c2fe5a41efea96664e37da81a985 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..667c145990791f8bb15a5cdc0ac03165f56e715b --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md b/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a0e24c24d32dfba60fc9bfa02ddd912c9f138888 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json b/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_256_lr_0.01_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md b/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7c2e0e1ed79bdcc14d9da2028357def9b815e8f0 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.001_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d65e257b165be652faaf60854f1a1a52d5131154 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b25595bb3f703270064b56c5425a5b43d9281b03 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8e96efe79962177d852017adb8d86ff415cbf551 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset oxford_flowers102. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: oxford_flowers102 + pretrained: true +test_datasets: null +train_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: train +type: transformers +val_datasets: + oxford_flowers102: + _target_: datasets.load_dataset + path: dpdl-benchmark/oxford_flowers102 + split: test +``` + diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ee05e378f6d60c8217f7b8db60be10c17bce37e2 --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,247 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "pink primrose", + "1": "hard-leaved pocket orchid", + "2": "canterbury bells", + "3": "sweet pea", + "4": "english marigold", + "5": "tiger lily", + "6": "moon orchid", + "7": "bird of paradise", + "8": "monkshood", + "9": "globe thistle", + "10": "snapdragon", + "11": "colt's foot", + "12": "king protea", + "13": "spear thistle", + "14": "yellow iris", + "15": "globe-flower", + "16": "purple coneflower", + "17": "peruvian lily", + "18": "balloon flower", + "19": "giant white arum lily", + "20": "fire lily", + "21": "pincushion flower", + "22": "fritillary", + "23": "red ginger", + "24": "grape hyacinth", + "25": "corn poppy", + "26": "prince of wales feathers", + "27": "stemless gentian", + "28": "artichoke", + "29": "sweet william", + "30": "carnation", + "31": "garden phlox", + "32": "love in the mist", + "33": "mexican aster", + "34": "alpine sea holly", + "35": "ruby-lipped cattleya", + "36": "cape flower", + "37": "great masterwort", + "38": "siam tulip", + "39": "lenten rose", + "40": "barbeton daisy", + "41": "daffodil", + "42": "sword lily", + "43": "poinsettia", + "44": "bolero deep blue", + "45": "wallflower", + "46": "marigold", + "47": "buttercup", + "48": "oxeye daisy", + "49": "common dandelion", + "50": "petunia", + "51": "wild pansy", + "52": "primula", + "53": "sunflower", + "54": "pelargonium", + "55": "bishop of llandaff", + "56": "gaura", + "57": "geranium", + "58": "orange dahlia", + "59": "pink-yellow dahlia?", + "60": "cautleya spicata", + "61": "japanese anemone", + "62": "black-eyed susan", + "63": "silverbush", + "64": "californian poppy", + "65": "osteospermum", + "66": "spring crocus", + "67": "bearded iris", + "68": "windflower", + "69": "tree poppy", + "70": "gazania", + "71": "azalea", + "72": "water lily", + "73": "rose", + "74": "thorn apple", + "75": "morning glory", + "76": "passion flower", + "77": "lotus", + "78": "toad lily", + "79": "anthurium", + "80": "frangipani", + "81": "clematis", + "82": "hibiscus", + "83": "columbine", + "84": "desert-rose", + "85": "tree mallow", + "86": "magnolia", + "87": "cyclamen", + "88": "watercress", + "89": "canna lily", + "90": "hippeastrum", + "91": "bee balm", + "92": "ball moss", + "93": "foxglove", + "94": "bougainvillea", + "95": "camellia", + "96": "mallow", + "97": "mexican petunia", + "98": "bromelia", + "99": "blanket flower", + "100": "trumpet creeper", + "101": "blackberry lily" + }, + "label2id": { + "alpine sea holly": 34, + "anthurium": 79, + "artichoke": 28, + "azalea": 71, + "ball moss": 92, + "balloon flower": 18, + "barbeton daisy": 40, + "bearded iris": 67, + "bee balm": 91, + "bird of paradise": 7, + "bishop of llandaff": 55, + "black-eyed susan": 62, + "blackberry lily": 101, + "blanket flower": 99, + "bolero deep blue": 44, + "bougainvillea": 94, + "bromelia": 98, + "buttercup": 47, + "californian poppy": 64, + "camellia": 95, + "canna lily": 89, + "canterbury bells": 2, + "cape flower": 36, + "carnation": 30, + "cautleya spicata": 60, + "clematis": 81, + "colt's foot": 11, + "columbine": 83, + "common dandelion": 49, + "corn poppy": 25, + "cyclamen": 87, + "daffodil": 41, + "desert-rose": 84, + "english marigold": 4, + "fire lily": 20, + "foxglove": 93, + "frangipani": 80, + "fritillary": 22, + "garden phlox": 31, + "gaura": 56, + "gazania": 70, + "geranium": 57, + "giant white arum lily": 19, + "globe thistle": 9, + "globe-flower": 15, + "grape hyacinth": 24, + "great masterwort": 37, + "hard-leaved pocket orchid": 1, + "hibiscus": 82, + "hippeastrum": 90, + "japanese anemone": 61, + "king protea": 12, + "lenten rose": 39, + "lotus": 77, + "love in the mist": 32, + "magnolia": 86, + "mallow": 96, + "marigold": 46, + "mexican aster": 33, + "mexican petunia": 97, + "monkshood": 8, + "moon orchid": 6, + "morning glory": 75, + "orange dahlia": 58, + "osteospermum": 65, + "oxeye daisy": 48, + "passion flower": 76, + "pelargonium": 54, + "peruvian lily": 17, + "petunia": 50, + "pincushion flower": 21, + "pink primrose": 0, + "pink-yellow dahlia?": 59, + "poinsettia": 43, + "primula": 52, + "prince of wales feathers": 26, + "purple coneflower": 16, + "red ginger": 23, + "rose": 73, + "ruby-lipped cattleya": 35, + "siam tulip": 38, + "silverbush": 63, + "snapdragon": 10, + "spear thistle": 13, + "spring crocus": 66, + "stemless gentian": 27, + "sunflower": 53, + "sweet pea": 3, + "sweet william": 29, + "sword lily": 42, + "thorn apple": 74, + "tiger lily": 5, + "toad lily": 78, + "tree mallow": 85, + "tree poppy": 69, + "trumpet creeper": 100, + "wallflower": 45, + "water lily": 72, + "watercress": 88, + "wild pansy": 51, + "windflower": 68, + "yellow iris": 14 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-oxford_flowers102-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae671b3d741eb073f2f7ff6886fb9c7516ee09cc --- /dev/null +++ b/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b0ac74d3d8fd86ac59d6ac1791d9e00ac89d34c --- /dev/null +++ b/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-pcam-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/README.md b/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a4fe748a58e3562b871170a0f752b52ee8b9ac01 --- /dev/null +++ b/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/config.json b/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b0ac74d3d8fd86ac59d6ac1791d9e00ac89d34c --- /dev/null +++ b/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json b/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-pcam-batch_size_128_lr_0.005_training_data_ratio_0.8-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0c4347a53e13f4f213f62d16b5388363efd12e72 --- /dev/null +++ b/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b0ac74d3d8fd86ac59d6ac1791d9e00ac89d34c --- /dev/null +++ b/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-pcam-batch_size_256_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md b/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..313ef3805b96f886cc9bbe4608b33554ebe4d4cc --- /dev/null +++ b/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json b/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b0ac74d3d8fd86ac59d6ac1791d9e00ac89d34c --- /dev/null +++ b/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-pcam-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..99408da3c9dd6b52013532f248f3bcffbe126a15 --- /dev/null +++ b/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset pcam. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: pcam + pretrained: true +test_datasets: null +train_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: train +type: transformers +val_datasets: + pcam: + _target_: datasets.load_dataset + path: 1aurent/PatchCamelyon + split: test +``` + diff --git a/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b0ac74d3d8fd86ac59d6ac1791d9e00ac89d34c --- /dev/null +++ b/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "lymph node", + "1": "lymph node containing metastatic tumor tissue" + }, + "label2id": { + "lymph node": 0, + "lymph node containing metastatic tumor tissue": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-pcam-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..549a16e02bad0ccfbc7f8a9b28aeb1f706b14d08 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8e2d395685cee2a9c76c2b7224752720ae236c37 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_128_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md b/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..fe43d624e1db2ff7d47e4dc87de5a3243e3f8564 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json b/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8e2d395685cee2a9c76c2b7224752720ae236c37 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json b/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_256_lr_0.001_training_data_ratio_0.5-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md b/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..99257b0d9874a82491edb378bace46d944b7eac6 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json b/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8e2d395685cee2a9c76c2b7224752720ae236c37 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_256_lr_0.01_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md b/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c14967019f50d543e692db441d6f2b01432f12a1 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json b/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8e2d395685cee2a9c76c2b7224752720ae236c37 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_64_lr_0.005_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..43f77cbfbc67fdf72f26fc917f9993dcb6c4f3a6 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset rendered-sst2. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: rendered-sst2 + pretrained: true +test_datasets: null +train_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: train +type: transformers +val_datasets: + rendered-sst2: + _target_: datasets.load_dataset + path: nateraw/rendered-sst2 + split: test +``` + diff --git a/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8e2d395685cee2a9c76c2b7224752720ae236c37 --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,47 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "negative", + "1": "positive" + }, + "label2id": { + "negative": 0, + "positive": 1 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-rendered-sst2-batch_size_64_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8a4a1bcd014a81ca3af0c9ce0df0dc463bbce5f8 --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f19cd9d15140683527543a4e276977d9a205e4b4 --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8a4a1bcd014a81ca3af0c9ce0df0dc463bbce5f8 --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f19cd9d15140683527543a4e276977d9a205e4b4 --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.001_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md b/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..831e432c6062fa5d7d5f14dc97551f6b1c184448 --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json b/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f19cd9d15140683527543a4e276977d9a205e4b4 --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-resisc45-batch_size_128_lr_0.01_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/README.md b/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e0dad3bb5ca537116f7403e770b5ed63102727fe --- /dev/null +++ b/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/config.json b/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f19cd9d15140683527543a4e276977d9a205e4b4 --- /dev/null +++ b/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-resisc45-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/README.md b/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9f704e850026a7453903d093cd07a6c95679dc3a --- /dev/null +++ b/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/config.json b/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f19cd9d15140683527543a4e276977d9a205e4b4 --- /dev/null +++ b/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-resisc45-batch_size_64_lr_0.005_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d5c86c85ac560c7e7df7d7069b8fabf29d84d273 --- /dev/null +++ b/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset resisc45. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: resisc45 + pretrained: true +test_datasets: null +train_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: train +type: transformers +val_datasets: + resisc45: + _target_: datasets.load_dataset + path: tanganke/resisc45 + split: test +``` + diff --git a/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f19cd9d15140683527543a4e276977d9a205e4b4 --- /dev/null +++ b/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,133 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "airport", + "2": "baseball diamond", + "3": "basketball court", + "4": "beach", + "5": "bridge", + "6": "chaparral", + "7": "church", + "8": "circular farmland", + "9": "cloud", + "10": "commercial area", + "11": "dense residential", + "12": "desert", + "13": "forest", + "14": "freeway", + "15": "golf course", + "16": "ground track field", + "17": "harbor", + "18": "industrial area", + "19": "intersection", + "20": "island", + "21": "lake", + "22": "meadow", + "23": "medium residential", + "24": "mobile home park", + "25": "mountain", + "26": "overpass", + "27": "palace", + "28": "parking lot", + "29": "railway", + "30": "railway station", + "31": "rectangular farmland", + "32": "river", + "33": "roundabout", + "34": "runway", + "35": "sea ice", + "36": "ship", + "37": "snowberg", + "38": "sparse residential", + "39": "stadium", + "40": "storage tank", + "41": "tennis court", + "42": "terrace", + "43": "thermal power station", + "44": "wetland" + }, + "label2id": { + "airplane": 0, + "airport": 1, + "baseball diamond": 2, + "basketball court": 3, + "beach": 4, + "bridge": 5, + "chaparral": 6, + "church": 7, + "circular farmland": 8, + "cloud": 9, + "commercial area": 10, + "dense residential": 11, + "desert": 12, + "forest": 13, + "freeway": 14, + "golf course": 15, + "ground track field": 16, + "harbor": 17, + "industrial area": 18, + "intersection": 19, + "island": 20, + "lake": 21, + "meadow": 22, + "medium residential": 23, + "mobile home park": 24, + "mountain": 25, + "overpass": 26, + "palace": 27, + "parking lot": 28, + "railway": 29, + "railway station": 30, + "rectangular farmland": 31, + "river": 32, + "roundabout": 33, + "runway": 34, + "sea ice": 35, + "ship": 36, + "snowberg": 37, + "sparse residential": 38, + "stadium": 39, + "storage tank": 40, + "tennis court": 41, + "terrace": 42, + "thermal power station": 43, + "wetland": 44 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-resisc45-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3880b16bd40ae00fb440771211cfa145775baeb5 --- /dev/null +++ b/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4ab8a7f592f8381989e1ca00e428926f91d8a1a5 --- /dev/null +++ b/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stanford-cars-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/README.md b/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..69bbbc1753f865cfb0b8d68571cc0e178b05505d --- /dev/null +++ b/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/config.json b/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4ab8a7f592f8381989e1ca00e428926f91d8a1a5 --- /dev/null +++ b/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stanford-cars-batch_size_256_lr_0.001_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ceb38d117cd28e4ece183f72a09e3a0a672ebf59 --- /dev/null +++ b/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4ab8a7f592f8381989e1ca00e428926f91d8a1a5 --- /dev/null +++ b/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stanford-cars-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md b/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..658b019f60eab1fa5dcf7471b56c70088530e184 --- /dev/null +++ b/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stanford-cars. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stanford-cars + pretrained: true +test_datasets: null +train_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: train +type: transformers +val_datasets: + stanford-cars: + _target_: datasets.load_dataset + path: tanganke/stanford_cars + split: test +``` + diff --git a/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json b/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4ab8a7f592f8381989e1ca00e428926f91d8a1a5 --- /dev/null +++ b/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,435 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "AM General Hummer SUV 2000", + "1": "Acura RL Sedan 2012", + "2": "Acura TL Sedan 2012", + "3": "Acura TL Type-S 2008", + "4": "Acura TSX Sedan 2012", + "5": "Acura Integra Type R 2001", + "6": "Acura ZDX Hatchback 2012", + "7": "Aston Martin V8 Vantage Convertible 2012", + "8": "Aston Martin V8 Vantage Coupe 2012", + "9": "Aston Martin Virage Convertible 2012", + "10": "Aston Martin Virage Coupe 2012", + "11": "Audi RS 4 Convertible 2008", + "12": "Audi A5 Coupe 2012", + "13": "Audi TTS Coupe 2012", + "14": "Audi R8 Coupe 2012", + "15": "Audi V8 Sedan 1994", + "16": "Audi 100 Sedan 1994", + "17": "Audi 100 Wagon 1994", + "18": "Audi TT Hatchback 2011", + "19": "Audi S6 Sedan 2011", + "20": "Audi S5 Convertible 2012", + "21": "Audi S5 Coupe 2012", + "22": "Audi S4 Sedan 2012", + "23": "Audi S4 Sedan 2007", + "24": "Audi TT RS Coupe 2012", + "25": "BMW ActiveHybrid 5 Sedan 2012", + "26": "BMW 1 Series Convertible 2012", + "27": "BMW 1 Series Coupe 2012", + "28": "BMW 3 Series Sedan 2012", + "29": "BMW 3 Series Wagon 2012", + "30": "BMW 6 Series Convertible 2007", + "31": "BMW X5 SUV 2007", + "32": "BMW X6 SUV 2012", + "33": "BMW M3 Coupe 2012", + "34": "BMW M5 Sedan 2010", + "35": "BMW M6 Convertible 2010", + "36": "BMW X3 SUV 2012", + "37": "BMW Z4 Convertible 2012", + "38": "Bentley Continental Supersports Conv. Convertible 2012", + "39": "Bentley Arnage Sedan 2009", + "40": "Bentley Mulsanne Sedan 2011", + "41": "Bentley Continental GT Coupe 2012", + "42": "Bentley Continental GT Coupe 2007", + "43": "Bentley Continental Flying Spur Sedan 2007", + "44": "Bugatti Veyron 16.4 Convertible 2009", + "45": "Bugatti Veyron 16.4 Coupe 2009", + "46": "Buick Regal GS 2012", + "47": "Buick Rainier SUV 2007", + "48": "Buick Verano Sedan 2012", + "49": "Buick Enclave SUV 2012", + "50": "Cadillac CTS-V Sedan 2012", + "51": "Cadillac SRX SUV 2012", + "52": "Cadillac Escalade EXT Crew Cab 2007", + "53": "Chevrolet Silverado 1500 Hybrid Crew Cab 2012", + "54": "Chevrolet Corvette Convertible 2012", + "55": "Chevrolet Corvette ZR1 2012", + "56": "Chevrolet Corvette Ron Fellows Edition Z06 2007", + "57": "Chevrolet Traverse SUV 2012", + "58": "Chevrolet Camaro Convertible 2012", + "59": "Chevrolet HHR SS 2010", + "60": "Chevrolet Impala Sedan 2007", + "61": "Chevrolet Tahoe Hybrid SUV 2012", + "62": "Chevrolet Sonic Sedan 2012", + "63": "Chevrolet Express Cargo Van 2007", + "64": "Chevrolet Avalanche Crew Cab 2012", + "65": "Chevrolet Cobalt SS 2010", + "66": "Chevrolet Malibu Hybrid Sedan 2010", + "67": "Chevrolet TrailBlazer SS 2009", + "68": "Chevrolet Silverado 2500HD Regular Cab 2012", + "69": "Chevrolet Silverado 1500 Classic Extended Cab 2007", + "70": "Chevrolet Express Van 2007", + "71": "Chevrolet Monte Carlo Coupe 2007", + "72": "Chevrolet Malibu Sedan 2007", + "73": "Chevrolet Silverado 1500 Extended Cab 2012", + "74": "Chevrolet Silverado 1500 Regular Cab 2012", + "75": "Chrysler Aspen SUV 2009", + "76": "Chrysler Sebring Convertible 2010", + "77": "Chrysler Town and Country Minivan 2012", + "78": "Chrysler 300 SRT-8 2010", + "79": "Chrysler Crossfire Convertible 2008", + "80": "Chrysler PT Cruiser Convertible 2008", + "81": "Daewoo Nubira Wagon 2002", + "82": "Dodge Caliber Wagon 2012", + "83": "Dodge Caliber Wagon 2007", + "84": "Dodge Caravan Minivan 1997", + "85": "Dodge Ram Pickup 3500 Crew Cab 2010", + "86": "Dodge Ram Pickup 3500 Quad Cab 2009", + "87": "Dodge Sprinter Cargo Van 2009", + "88": "Dodge Journey SUV 2012", + "89": "Dodge Dakota Crew Cab 2010", + "90": "Dodge Dakota Club Cab 2007", + "91": "Dodge Magnum Wagon 2008", + "92": "Dodge Challenger SRT8 2011", + "93": "Dodge Durango SUV 2012", + "94": "Dodge Durango SUV 2007", + "95": "Dodge Charger Sedan 2012", + "96": "Dodge Charger SRT-8 2009", + "97": "Eagle Talon Hatchback 1998", + "98": "FIAT 500 Abarth 2012", + "99": "FIAT 500 Convertible 2012", + "100": "Ferrari FF Coupe 2012", + "101": "Ferrari California Convertible 2012", + "102": "Ferrari 458 Italia Convertible 2012", + "103": "Ferrari 458 Italia Coupe 2012", + "104": "Fisker Karma Sedan 2012", + "105": "Ford F-450 Super Duty Crew Cab 2012", + "106": "Ford Mustang Convertible 2007", + "107": "Ford Freestar Minivan 2007", + "108": "Ford Expedition EL SUV 2009", + "109": "Ford Edge SUV 2012", + "110": "Ford Ranger SuperCab 2011", + "111": "Ford GT Coupe 2006", + "112": "Ford F-150 Regular Cab 2012", + "113": "Ford F-150 Regular Cab 2007", + "114": "Ford Focus Sedan 2007", + "115": "Ford E-Series Wagon Van 2012", + "116": "Ford Fiesta Sedan 2012", + "117": "GMC Terrain SUV 2012", + "118": "GMC Savana Van 2012", + "119": "GMC Yukon Hybrid SUV 2012", + "120": "GMC Acadia SUV 2012", + "121": "GMC Canyon Extended Cab 2012", + "122": "Geo Metro Convertible 1993", + "123": "HUMMER H3T Crew Cab 2010", + "124": "HUMMER H2 SUT Crew Cab 2009", + "125": "Honda Odyssey Minivan 2012", + "126": "Honda Odyssey Minivan 2007", + "127": "Honda Accord Coupe 2012", + "128": "Honda Accord Sedan 2012", + "129": "Hyundai Veloster Hatchback 2012", + "130": "Hyundai Santa Fe SUV 2012", + "131": "Hyundai Tucson SUV 2012", + "132": "Hyundai Veracruz SUV 2012", + "133": "Hyundai Sonata Hybrid Sedan 2012", + "134": "Hyundai Elantra Sedan 2007", + "135": "Hyundai Accent Sedan 2012", + "136": "Hyundai Genesis Sedan 2012", + "137": "Hyundai Sonata Sedan 2012", + "138": "Hyundai Elantra Touring Hatchback 2012", + "139": "Hyundai Azera Sedan 2012", + "140": "Infiniti G Coupe IPL 2012", + "141": "Infiniti QX56 SUV 2011", + "142": "Isuzu Ascender SUV 2008", + "143": "Jaguar XK XKR 2012", + "144": "Jeep Patriot SUV 2012", + "145": "Jeep Wrangler SUV 2012", + "146": "Jeep Liberty SUV 2012", + "147": "Jeep Grand Cherokee SUV 2012", + "148": "Jeep Compass SUV 2012", + "149": "Lamborghini Reventon Coupe 2008", + "150": "Lamborghini Aventador Coupe 2012", + "151": "Lamborghini Gallardo LP 570-4 Superleggera 2012", + "152": "Lamborghini Diablo Coupe 2001", + "153": "Land Rover Range Rover SUV 2012", + "154": "Land Rover LR2 SUV 2012", + "155": "Lincoln Town Car Sedan 2011", + "156": "MINI Cooper Roadster Convertible 2012", + "157": "Maybach Landaulet Convertible 2012", + "158": "Mazda Tribute SUV 2011", + "159": "McLaren MP4-12C Coupe 2012", + "160": "Mercedes-Benz 300-Class Convertible 1993", + "161": "Mercedes-Benz C-Class Sedan 2012", + "162": "Mercedes-Benz SL-Class Coupe 2009", + "163": "Mercedes-Benz E-Class Sedan 2012", + "164": "Mercedes-Benz S-Class Sedan 2012", + "165": "Mercedes-Benz Sprinter Van 2012", + "166": "Mitsubishi Lancer Sedan 2012", + "167": "Nissan Leaf Hatchback 2012", + "168": "Nissan NV Passenger Van 2012", + "169": "Nissan Juke Hatchback 2012", + "170": "Nissan 240SX Coupe 1998", + "171": "Plymouth Neon Coupe 1999", + "172": "Porsche Panamera Sedan 2012", + "173": "Ram C/V Cargo Van Minivan 2012", + "174": "Rolls-Royce Phantom Drophead Coupe Convertible 2012", + "175": "Rolls-Royce Ghost Sedan 2012", + "176": "Rolls-Royce Phantom Sedan 2012", + "177": "Scion xD Hatchback 2012", + "178": "Spyker C8 Convertible 2009", + "179": "Spyker C8 Coupe 2009", + "180": "Suzuki Aerio Sedan 2007", + "181": "Suzuki Kizashi Sedan 2012", + "182": "Suzuki SX4 Hatchback 2012", + "183": "Suzuki SX4 Sedan 2012", + "184": "Tesla Model S Sedan 2012", + "185": "Toyota Sequoia SUV 2012", + "186": "Toyota Camry Sedan 2012", + "187": "Toyota Corolla Sedan 2012", + "188": "Toyota 4Runner SUV 2012", + "189": "Volkswagen Golf Hatchback 2012", + "190": "Volkswagen Golf Hatchback 1991", + "191": "Volkswagen Beetle Hatchback 2012", + "192": "Volvo C30 Hatchback 2012", + "193": "Volvo 240 Sedan 1993", + "194": "Volvo XC90 SUV 2007", + "195": "smart fortwo Convertible 2012" + }, + "label2id": { + "AM General Hummer SUV 2000": 0, + "Acura Integra Type R 2001": 5, + "Acura RL Sedan 2012": 1, + "Acura TL Sedan 2012": 2, + "Acura TL Type-S 2008": 3, + "Acura TSX Sedan 2012": 4, + "Acura ZDX Hatchback 2012": 6, + "Aston Martin V8 Vantage Convertible 2012": 7, + "Aston Martin V8 Vantage Coupe 2012": 8, + "Aston Martin Virage Convertible 2012": 9, + "Aston Martin Virage Coupe 2012": 10, + "Audi 100 Sedan 1994": 16, + "Audi 100 Wagon 1994": 17, + "Audi A5 Coupe 2012": 12, + "Audi R8 Coupe 2012": 14, + "Audi RS 4 Convertible 2008": 11, + "Audi S4 Sedan 2007": 23, + "Audi S4 Sedan 2012": 22, + "Audi S5 Convertible 2012": 20, + "Audi S5 Coupe 2012": 21, + "Audi S6 Sedan 2011": 19, + "Audi TT Hatchback 2011": 18, + "Audi TT RS Coupe 2012": 24, + "Audi TTS Coupe 2012": 13, + "Audi V8 Sedan 1994": 15, + "BMW 1 Series Convertible 2012": 26, + "BMW 1 Series Coupe 2012": 27, + "BMW 3 Series Sedan 2012": 28, + "BMW 3 Series Wagon 2012": 29, + "BMW 6 Series Convertible 2007": 30, + "BMW ActiveHybrid 5 Sedan 2012": 25, + "BMW M3 Coupe 2012": 33, + "BMW M5 Sedan 2010": 34, + "BMW M6 Convertible 2010": 35, + "BMW X3 SUV 2012": 36, + "BMW X5 SUV 2007": 31, + "BMW X6 SUV 2012": 32, + "BMW Z4 Convertible 2012": 37, + "Bentley Arnage Sedan 2009": 39, + "Bentley Continental Flying Spur Sedan 2007": 43, + "Bentley Continental GT Coupe 2007": 42, + "Bentley Continental GT Coupe 2012": 41, + "Bentley Continental Supersports Conv. Convertible 2012": 38, + "Bentley Mulsanne Sedan 2011": 40, + "Bugatti Veyron 16.4 Convertible 2009": 44, + "Bugatti Veyron 16.4 Coupe 2009": 45, + "Buick Enclave SUV 2012": 49, + "Buick Rainier SUV 2007": 47, + "Buick Regal GS 2012": 46, + "Buick Verano Sedan 2012": 48, + "Cadillac CTS-V Sedan 2012": 50, + "Cadillac Escalade EXT Crew Cab 2007": 52, + "Cadillac SRX SUV 2012": 51, + "Chevrolet Avalanche Crew Cab 2012": 64, + "Chevrolet Camaro Convertible 2012": 58, + "Chevrolet Cobalt SS 2010": 65, + "Chevrolet Corvette Convertible 2012": 54, + "Chevrolet Corvette Ron Fellows Edition Z06 2007": 56, + "Chevrolet Corvette ZR1 2012": 55, + "Chevrolet Express Cargo Van 2007": 63, + "Chevrolet Express Van 2007": 70, + "Chevrolet HHR SS 2010": 59, + "Chevrolet Impala Sedan 2007": 60, + "Chevrolet Malibu Hybrid Sedan 2010": 66, + "Chevrolet Malibu Sedan 2007": 72, + "Chevrolet Monte Carlo Coupe 2007": 71, + "Chevrolet Silverado 1500 Classic Extended Cab 2007": 69, + "Chevrolet Silverado 1500 Extended Cab 2012": 73, + "Chevrolet Silverado 1500 Hybrid Crew Cab 2012": 53, + "Chevrolet Silverado 1500 Regular Cab 2012": 74, + "Chevrolet Silverado 2500HD Regular Cab 2012": 68, + "Chevrolet Sonic Sedan 2012": 62, + "Chevrolet Tahoe Hybrid SUV 2012": 61, + "Chevrolet TrailBlazer SS 2009": 67, + "Chevrolet Traverse SUV 2012": 57, + "Chrysler 300 SRT-8 2010": 78, + "Chrysler Aspen SUV 2009": 75, + "Chrysler Crossfire Convertible 2008": 79, + "Chrysler PT Cruiser Convertible 2008": 80, + "Chrysler Sebring Convertible 2010": 76, + "Chrysler Town and Country Minivan 2012": 77, + "Daewoo Nubira Wagon 2002": 81, + "Dodge Caliber Wagon 2007": 83, + "Dodge Caliber Wagon 2012": 82, + "Dodge Caravan Minivan 1997": 84, + "Dodge Challenger SRT8 2011": 92, + "Dodge Charger SRT-8 2009": 96, + "Dodge Charger Sedan 2012": 95, + "Dodge Dakota Club Cab 2007": 90, + "Dodge Dakota Crew Cab 2010": 89, + "Dodge Durango SUV 2007": 94, + "Dodge Durango SUV 2012": 93, + "Dodge Journey SUV 2012": 88, + "Dodge Magnum Wagon 2008": 91, + "Dodge Ram Pickup 3500 Crew Cab 2010": 85, + "Dodge Ram Pickup 3500 Quad Cab 2009": 86, + "Dodge Sprinter Cargo Van 2009": 87, + "Eagle Talon Hatchback 1998": 97, + "FIAT 500 Abarth 2012": 98, + "FIAT 500 Convertible 2012": 99, + "Ferrari 458 Italia Convertible 2012": 102, + "Ferrari 458 Italia Coupe 2012": 103, + "Ferrari California Convertible 2012": 101, + "Ferrari FF Coupe 2012": 100, + "Fisker Karma Sedan 2012": 104, + "Ford E-Series Wagon Van 2012": 115, + "Ford Edge SUV 2012": 109, + "Ford Expedition EL SUV 2009": 108, + "Ford F-150 Regular Cab 2007": 113, + "Ford F-150 Regular Cab 2012": 112, + "Ford F-450 Super Duty Crew Cab 2012": 105, + "Ford Fiesta Sedan 2012": 116, + "Ford Focus Sedan 2007": 114, + "Ford Freestar Minivan 2007": 107, + "Ford GT Coupe 2006": 111, + "Ford Mustang Convertible 2007": 106, + "Ford Ranger SuperCab 2011": 110, + "GMC Acadia SUV 2012": 120, + "GMC Canyon Extended Cab 2012": 121, + "GMC Savana Van 2012": 118, + "GMC Terrain SUV 2012": 117, + "GMC Yukon Hybrid SUV 2012": 119, + "Geo Metro Convertible 1993": 122, + "HUMMER H2 SUT Crew Cab 2009": 124, + "HUMMER H3T Crew Cab 2010": 123, + "Honda Accord Coupe 2012": 127, + "Honda Accord Sedan 2012": 128, + "Honda Odyssey Minivan 2007": 126, + "Honda Odyssey Minivan 2012": 125, + "Hyundai Accent Sedan 2012": 135, + "Hyundai Azera Sedan 2012": 139, + "Hyundai Elantra Sedan 2007": 134, + "Hyundai Elantra Touring Hatchback 2012": 138, + "Hyundai Genesis Sedan 2012": 136, + "Hyundai Santa Fe SUV 2012": 130, + "Hyundai Sonata Hybrid Sedan 2012": 133, + "Hyundai Sonata Sedan 2012": 137, + "Hyundai Tucson SUV 2012": 131, + "Hyundai Veloster Hatchback 2012": 129, + "Hyundai Veracruz SUV 2012": 132, + "Infiniti G Coupe IPL 2012": 140, + "Infiniti QX56 SUV 2011": 141, + "Isuzu Ascender SUV 2008": 142, + "Jaguar XK XKR 2012": 143, + "Jeep Compass SUV 2012": 148, + "Jeep Grand Cherokee SUV 2012": 147, + "Jeep Liberty SUV 2012": 146, + "Jeep Patriot SUV 2012": 144, + "Jeep Wrangler SUV 2012": 145, + "Lamborghini Aventador Coupe 2012": 150, + "Lamborghini Diablo Coupe 2001": 152, + "Lamborghini Gallardo LP 570-4 Superleggera 2012": 151, + "Lamborghini Reventon Coupe 2008": 149, + "Land Rover LR2 SUV 2012": 154, + "Land Rover Range Rover SUV 2012": 153, + "Lincoln Town Car Sedan 2011": 155, + "MINI Cooper Roadster Convertible 2012": 156, + "Maybach Landaulet Convertible 2012": 157, + "Mazda Tribute SUV 2011": 158, + "McLaren MP4-12C Coupe 2012": 159, + "Mercedes-Benz 300-Class Convertible 1993": 160, + "Mercedes-Benz C-Class Sedan 2012": 161, + "Mercedes-Benz E-Class Sedan 2012": 163, + "Mercedes-Benz S-Class Sedan 2012": 164, + "Mercedes-Benz SL-Class Coupe 2009": 162, + "Mercedes-Benz Sprinter Van 2012": 165, + "Mitsubishi Lancer Sedan 2012": 166, + "Nissan 240SX Coupe 1998": 170, + "Nissan Juke Hatchback 2012": 169, + "Nissan Leaf Hatchback 2012": 167, + "Nissan NV Passenger Van 2012": 168, + "Plymouth Neon Coupe 1999": 171, + "Porsche Panamera Sedan 2012": 172, + "Ram C/V Cargo Van Minivan 2012": 173, + "Rolls-Royce Ghost Sedan 2012": 175, + "Rolls-Royce Phantom Drophead Coupe Convertible 2012": 174, + "Rolls-Royce Phantom Sedan 2012": 176, + "Scion xD Hatchback 2012": 177, + "Spyker C8 Convertible 2009": 178, + "Spyker C8 Coupe 2009": 179, + "Suzuki Aerio Sedan 2007": 180, + "Suzuki Kizashi Sedan 2012": 181, + "Suzuki SX4 Hatchback 2012": 182, + "Suzuki SX4 Sedan 2012": 183, + "Tesla Model S Sedan 2012": 184, + "Toyota 4Runner SUV 2012": 188, + "Toyota Camry Sedan 2012": 186, + "Toyota Corolla Sedan 2012": 187, + "Toyota Sequoia SUV 2012": 185, + "Volkswagen Beetle Hatchback 2012": 191, + "Volkswagen Golf Hatchback 1991": 190, + "Volkswagen Golf Hatchback 2012": 189, + "Volvo 240 Sedan 1993": 193, + "Volvo C30 Hatchback 2012": 192, + "Volvo XC90 SUV 2007": 194, + "smart fortwo Convertible 2012": 195 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stanford-cars-batch_size_64_lr_0.01_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8327291f1c1d152164abb591813ba4fbc6b1d731 --- /dev/null +++ b/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..589b71bb15675afba14a3d018d8e1b4a07ab7182 --- /dev/null +++ b/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stl10-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c14ef6469423761a4f263cc27eb380b524b2d7a0 --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..589b71bb15675afba14a3d018d8e1b4a07ab7182 --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c14ef6469423761a4f263cc27eb380b524b2d7a0 --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..589b71bb15675afba14a3d018d8e1b4a07ab7182 --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.001_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/README.md b/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f4b1032e942bd1658fbfe5f08c842ad429815949 --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/config.json b/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..589b71bb15675afba14a3d018d8e1b4a07ab7182 --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stl10-batch_size_256_lr_0.005_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md b/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..17af393e6f54a2e5a6817e10d91703d1d0c8e917 --- /dev/null +++ b/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset stl10. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: stl10 + pretrained: true +test_datasets: null +train_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: train +type: transformers +val_datasets: + stl10: + _target_: datasets.load_dataset + path: tanganke/stl10 + split: test +``` + diff --git a/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json b/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..589b71bb15675afba14a3d018d8e1b4a07ab7182 --- /dev/null +++ b/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "airplane", + "1": "bird", + "2": "car", + "3": "cat", + "4": "deer", + "5": "dog", + "6": "horse", + "7": "monkey", + "8": "ship", + "9": "truck" + }, + "label2id": { + "airplane": 0, + "bird": 1, + "car": 2, + "cat": 3, + "deer": 4, + "dog": 5, + "horse": 6, + "monkey": 7, + "ship": 8, + "truck": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-stl10-batch_size_64_lr_0.001_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b5e6659084691570db375b007d87e993663cc031 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..475764883d5c42c7cf491ba91c6f9055d9008965 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..475764883d5c42c7cf491ba91c6f9055d9008965 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..475764883d5c42c7cf491ba91c6f9055d9008965 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 128 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_128_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md b/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4c47a6a6273e65ed58f544e5cc5c5110af378ea5 --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.001 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json b/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.001_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1a9003d1e5701486a6671964c073eae82df1c183 --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.5-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..5cc2c4a27fa94375ec36c01b40aad24c7134c14d --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_256_lr_0.01_training_data_ratio_0.8-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..87f4cfdc618f8f846bc4c80ca769a8f312a9ddae --- /dev/null +++ b/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,76 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset sun397. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: sun397 + pretrained: true +test_datasets: null +train_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: train +type: transformers +val_datasets: + sun397: + _target_: datasets.load_dataset + path: tanganke/sun397 + split: test +``` + diff --git a/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7b3f0b16022576c1d5cf54a17e75ff9088c89b17 --- /dev/null +++ b/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,837 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "abbey", + "1": "airplane cabin", + "2": "airport terminal", + "3": "alley", + "4": "amphitheater", + "5": "amusement arcade", + "6": "amusement park", + "7": "anechoic chamber", + "8": "apartment building outdoor", + "9": "apse indoor", + "10": "aquarium", + "11": "aqueduct", + "12": "arch", + "13": "archive", + "14": "arrival gate outdoor", + "15": "art gallery", + "16": "art school", + "17": "art studio", + "18": "assembly line", + "19": "athletic field outdoor", + "20": "atrium public", + "21": "attic", + "22": "auditorium", + "23": "auto factory", + "24": "badlands", + "25": "badminton court indoor", + "26": "baggage claim", + "27": "bakery shop", + "28": "balcony exterior", + "29": "balcony interior", + "30": "ball pit", + "31": "ballroom", + "32": "bamboo forest", + "33": "banquet hall", + "34": "bar", + "35": "barn", + "36": "barndoor", + "37": "baseball field", + "38": "basement", + "39": "basilica", + "40": "basketball court outdoor", + "41": "bathroom", + "42": "batters box", + "43": "bayou", + "44": "bazaar indoor", + "45": "bazaar outdoor", + "46": "beach", + "47": "beauty salon", + "48": "bedroom", + "49": "berth", + "50": "biology laboratory", + "51": "bistro indoor", + "52": "boardwalk", + "53": "boat deck", + "54": "boathouse", + "55": "bookstore", + "56": "booth indoor", + "57": "botanical garden", + "58": "bow window indoor", + "59": "bow window outdoor", + "60": "bowling alley", + "61": "boxing ring", + "62": "brewery indoor", + "63": "bridge", + "64": "building facade", + "65": "bullring", + "66": "burial chamber", + "67": "bus interior", + "68": "butchers shop", + "69": "butte", + "70": "cabin outdoor", + "71": "cafeteria", + "72": "campsite", + "73": "campus", + "74": "canal natural", + "75": "canal urban", + "76": "candy store", + "77": "canyon", + "78": "car interior backseat", + "79": "car interior frontseat", + "80": "carrousel", + "81": "casino indoor", + "82": "castle", + "83": "catacomb", + "84": "cathedral indoor", + "85": "cathedral outdoor", + "86": "cavern indoor", + "87": "cemetery", + "88": "chalet", + "89": "cheese factory", + "90": "chemistry lab", + "91": "chicken coop indoor", + "92": "chicken coop outdoor", + "93": "childs room", + "94": "church indoor", + "95": "church outdoor", + "96": "classroom", + "97": "clean room", + "98": "cliff", + "99": "cloister indoor", + "100": "closet", + "101": "clothing store", + "102": "coast", + "103": "cockpit", + "104": "coffee shop", + "105": "computer room", + "106": "conference center", + "107": "conference room", + "108": "construction site", + "109": "control room", + "110": "control tower outdoor", + "111": "corn field", + "112": "corral", + "113": "corridor", + "114": "cottage garden", + "115": "courthouse", + "116": "courtroom", + "117": "courtyard", + "118": "covered bridge exterior", + "119": "creek", + "120": "crevasse", + "121": "crosswalk", + "122": "cubicle office", + "123": "dam", + "124": "delicatessen", + "125": "dentists office", + "126": "desert sand", + "127": "desert vegetation", + "128": "diner indoor", + "129": "diner outdoor", + "130": "dinette home", + "131": "dinette vehicle", + "132": "dining car", + "133": "dining room", + "134": "discotheque", + "135": "dock", + "136": "doorway outdoor", + "137": "dorm room", + "138": "driveway", + "139": "driving range outdoor", + "140": "drugstore", + "141": "electrical substation", + "142": "elevator door", + "143": "elevator interior", + "144": "elevator shaft", + "145": "engine room", + "146": "escalator indoor", + "147": "excavation", + "148": "factory indoor", + "149": "fairway", + "150": "fastfood restaurant", + "151": "field cultivated", + "152": "field wild", + "153": "fire escape", + "154": "fire station", + "155": "firing range indoor", + "156": "fishpond", + "157": "florist shop indoor", + "158": "food court", + "159": "forest broadleaf", + "160": "forest needleleaf", + "161": "forest path", + "162": "forest road", + "163": "formal garden", + "164": "fountain", + "165": "galley", + "166": "game room", + "167": "garage indoor", + "168": "garbage dump", + "169": "gas station", + "170": "gazebo exterior", + "171": "general store indoor", + "172": "general store outdoor", + "173": "gift shop", + "174": "golf course", + "175": "greenhouse indoor", + "176": "greenhouse outdoor", + "177": "gymnasium indoor", + "178": "hangar indoor", + "179": "hangar outdoor", + "180": "harbor", + "181": "hayfield", + "182": "heliport", + "183": "herb garden", + "184": "highway", + "185": "hill", + "186": "home office", + "187": "hospital", + "188": "hospital room", + "189": "hot spring", + "190": "hot tub outdoor", + "191": "hotel outdoor", + "192": "hotel room", + "193": "house", + "194": "hunting lodge outdoor", + "195": "ice cream parlor", + "196": "ice floe", + "197": "ice shelf", + "198": "ice skating rink indoor", + "199": "ice skating rink outdoor", + "200": "iceberg", + "201": "igloo", + "202": "industrial area", + "203": "inn outdoor", + "204": "islet", + "205": "jacuzzi indoor", + "206": "jail cell", + "207": "jail indoor", + "208": "jewelry shop", + "209": "kasbah", + "210": "kennel indoor", + "211": "kennel outdoor", + "212": "kindergarden classroom", + "213": "kitchen", + "214": "kitchenette", + "215": "labyrinth outdoor", + "216": "lake natural", + "217": "landfill", + "218": "landing deck", + "219": "laundromat", + "220": "lecture room", + "221": "library indoor", + "222": "library outdoor", + "223": "lido deck outdoor", + "224": "lift bridge", + "225": "lighthouse", + "226": "limousine interior", + "227": "living room", + "228": "lobby", + "229": "lock chamber", + "230": "locker room", + "231": "mansion", + "232": "manufactured home", + "233": "market indoor", + "234": "market outdoor", + "235": "marsh", + "236": "martial arts gym", + "237": "mausoleum", + "238": "medina", + "239": "moat water", + "240": "monastery outdoor", + "241": "mosque indoor", + "242": "mosque outdoor", + "243": "motel", + "244": "mountain", + "245": "mountain snowy", + "246": "movie theater indoor", + "247": "museum indoor", + "248": "music store", + "249": "music studio", + "250": "nuclear power plant outdoor", + "251": "nursery", + "252": "oast house", + "253": "observatory outdoor", + "254": "ocean", + "255": "office", + "256": "office building", + "257": "oil refinery outdoor", + "258": "oilrig", + "259": "operating room", + "260": "orchard", + "261": "outhouse outdoor", + "262": "pagoda", + "263": "palace", + "264": "pantry", + "265": "park", + "266": "parking garage indoor", + "267": "parking garage outdoor", + "268": "parking lot", + "269": "parlor", + "270": "pasture", + "271": "patio", + "272": "pavilion", + "273": "pharmacy", + "274": "phone booth", + "275": "physics laboratory", + "276": "picnic area", + "277": "pilothouse indoor", + "278": "planetarium outdoor", + "279": "playground", + "280": "playroom", + "281": "plaza", + "282": "podium indoor", + "283": "podium outdoor", + "284": "pond", + "285": "poolroom establishment", + "286": "poolroom home", + "287": "power plant outdoor", + "288": "promenade deck", + "289": "pub indoor", + "290": "pulpit", + "291": "putting green", + "292": "racecourse", + "293": "raceway", + "294": "raft", + "295": "railroad track", + "296": "rainforest", + "297": "reception", + "298": "recreation room", + "299": "residential neighborhood", + "300": "restaurant", + "301": "restaurant kitchen", + "302": "restaurant patio", + "303": "rice paddy", + "304": "riding arena", + "305": "river", + "306": "rock arch", + "307": "rope bridge", + "308": "ruin", + "309": "runway", + "310": "sandbar", + "311": "sandbox", + "312": "sauna", + "313": "schoolhouse", + "314": "sea cliff", + "315": "server room", + "316": "shed", + "317": "shoe shop", + "318": "shopfront", + "319": "shopping mall indoor", + "320": "shower", + "321": "skatepark", + "322": "ski lodge", + "323": "ski resort", + "324": "ski slope", + "325": "sky", + "326": "skyscraper", + "327": "slum", + "328": "snowfield", + "329": "squash court", + "330": "stable", + "331": "stadium baseball", + "332": "stadium football", + "333": "stage indoor", + "334": "staircase", + "335": "street", + "336": "subway interior", + "337": "subway station platform", + "338": "supermarket", + "339": "sushi bar", + "340": "swamp", + "341": "swimming pool indoor", + "342": "swimming pool outdoor", + "343": "synagogue indoor", + "344": "synagogue outdoor", + "345": "television studio", + "346": "temple east asia", + "347": "temple south asia", + "348": "tennis court indoor", + "349": "tennis court outdoor", + "350": "tent outdoor", + "351": "theater indoor procenium", + "352": "theater indoor seats", + "353": "thriftshop", + "354": "throne room", + "355": "ticket booth", + "356": "toll plaza", + "357": "topiary garden", + "358": "tower", + "359": "toyshop", + "360": "track outdoor", + "361": "train railway", + "362": "train station platform", + "363": "tree farm", + "364": "tree house", + "365": "trench", + "366": "underwater coral reef", + "367": "utility room", + "368": "valley", + "369": "van interior", + "370": "vegetable garden", + "371": "veranda", + "372": "veterinarians office", + "373": "viaduct", + "374": "videostore", + "375": "village", + "376": "vineyard", + "377": "volcano", + "378": "volleyball court indoor", + "379": "volleyball court outdoor", + "380": "waiting room", + "381": "warehouse indoor", + "382": "water tower", + "383": "waterfall block", + "384": "waterfall fan", + "385": "waterfall plunge", + "386": "watering hole", + "387": "wave", + "388": "wet bar", + "389": "wheat field", + "390": "wind farm", + "391": "windmill", + "392": "wine cellar barrel storage", + "393": "wine cellar bottle storage", + "394": "wrestling ring indoor", + "395": "yard", + "396": "youth hostel" + }, + "label2id": { + "abbey": 0, + "airplane cabin": 1, + "airport terminal": 2, + "alley": 3, + "amphitheater": 4, + "amusement arcade": 5, + "amusement park": 6, + "anechoic chamber": 7, + "apartment building outdoor": 8, + "apse indoor": 9, + "aquarium": 10, + "aqueduct": 11, + "arch": 12, + "archive": 13, + "arrival gate outdoor": 14, + "art gallery": 15, + "art school": 16, + "art studio": 17, + "assembly line": 18, + "athletic field outdoor": 19, + "atrium public": 20, + "attic": 21, + "auditorium": 22, + "auto factory": 23, + "badlands": 24, + "badminton court indoor": 25, + "baggage claim": 26, + "bakery shop": 27, + "balcony exterior": 28, + "balcony interior": 29, + "ball pit": 30, + "ballroom": 31, + "bamboo forest": 32, + "banquet hall": 33, + "bar": 34, + "barn": 35, + "barndoor": 36, + "baseball field": 37, + "basement": 38, + "basilica": 39, + "basketball court outdoor": 40, + "bathroom": 41, + "batters box": 42, + "bayou": 43, + "bazaar indoor": 44, + "bazaar outdoor": 45, + "beach": 46, + "beauty salon": 47, + "bedroom": 48, + "berth": 49, + "biology laboratory": 50, + "bistro indoor": 51, + "boardwalk": 52, + "boat deck": 53, + "boathouse": 54, + "bookstore": 55, + "booth indoor": 56, + "botanical garden": 57, + "bow window indoor": 58, + "bow window outdoor": 59, + "bowling alley": 60, + "boxing ring": 61, + "brewery indoor": 62, + "bridge": 63, + "building facade": 64, + "bullring": 65, + "burial chamber": 66, + "bus interior": 67, + "butchers shop": 68, + "butte": 69, + "cabin outdoor": 70, + "cafeteria": 71, + "campsite": 72, + "campus": 73, + "canal natural": 74, + "canal urban": 75, + "candy store": 76, + "canyon": 77, + "car interior backseat": 78, + "car interior frontseat": 79, + "carrousel": 80, + "casino indoor": 81, + "castle": 82, + "catacomb": 83, + "cathedral indoor": 84, + "cathedral outdoor": 85, + "cavern indoor": 86, + "cemetery": 87, + "chalet": 88, + "cheese factory": 89, + "chemistry lab": 90, + "chicken coop indoor": 91, + "chicken coop outdoor": 92, + "childs room": 93, + "church indoor": 94, + "church outdoor": 95, + "classroom": 96, + "clean room": 97, + "cliff": 98, + "cloister indoor": 99, + "closet": 100, + "clothing store": 101, + "coast": 102, + "cockpit": 103, + "coffee shop": 104, + "computer room": 105, + "conference center": 106, + "conference room": 107, + "construction site": 108, + "control room": 109, + "control tower outdoor": 110, + "corn field": 111, + "corral": 112, + "corridor": 113, + "cottage garden": 114, + "courthouse": 115, + "courtroom": 116, + "courtyard": 117, + "covered bridge exterior": 118, + "creek": 119, + "crevasse": 120, + "crosswalk": 121, + "cubicle office": 122, + "dam": 123, + "delicatessen": 124, + "dentists office": 125, + "desert sand": 126, + "desert vegetation": 127, + "diner indoor": 128, + "diner outdoor": 129, + "dinette home": 130, + "dinette vehicle": 131, + "dining car": 132, + "dining room": 133, + "discotheque": 134, + "dock": 135, + "doorway outdoor": 136, + "dorm room": 137, + "driveway": 138, + "driving range outdoor": 139, + "drugstore": 140, + "electrical substation": 141, + "elevator door": 142, + "elevator interior": 143, + "elevator shaft": 144, + "engine room": 145, + "escalator indoor": 146, + "excavation": 147, + "factory indoor": 148, + "fairway": 149, + "fastfood restaurant": 150, + "field cultivated": 151, + "field wild": 152, + "fire escape": 153, + "fire station": 154, + "firing range indoor": 155, + "fishpond": 156, + "florist shop indoor": 157, + "food court": 158, + "forest broadleaf": 159, + "forest needleleaf": 160, + "forest path": 161, + "forest road": 162, + "formal garden": 163, + "fountain": 164, + "galley": 165, + "game room": 166, + "garage indoor": 167, + "garbage dump": 168, + "gas station": 169, + "gazebo exterior": 170, + "general store indoor": 171, + "general store outdoor": 172, + "gift shop": 173, + "golf course": 174, + "greenhouse indoor": 175, + "greenhouse outdoor": 176, + "gymnasium indoor": 177, + "hangar indoor": 178, + "hangar outdoor": 179, + "harbor": 180, + "hayfield": 181, + "heliport": 182, + "herb garden": 183, + "highway": 184, + "hill": 185, + "home office": 186, + "hospital": 187, + "hospital room": 188, + "hot spring": 189, + "hot tub outdoor": 190, + "hotel outdoor": 191, + "hotel room": 192, + "house": 193, + "hunting lodge outdoor": 194, + "ice cream parlor": 195, + "ice floe": 196, + "ice shelf": 197, + "ice skating rink indoor": 198, + "ice skating rink outdoor": 199, + "iceberg": 200, + "igloo": 201, + "industrial area": 202, + "inn outdoor": 203, + "islet": 204, + "jacuzzi indoor": 205, + "jail cell": 206, + "jail indoor": 207, + "jewelry shop": 208, + "kasbah": 209, + "kennel indoor": 210, + "kennel outdoor": 211, + "kindergarden classroom": 212, + "kitchen": 213, + "kitchenette": 214, + "labyrinth outdoor": 215, + "lake natural": 216, + "landfill": 217, + "landing deck": 218, + "laundromat": 219, + "lecture room": 220, + "library indoor": 221, + "library outdoor": 222, + "lido deck outdoor": 223, + "lift bridge": 224, + "lighthouse": 225, + "limousine interior": 226, + "living room": 227, + "lobby": 228, + "lock chamber": 229, + "locker room": 230, + "mansion": 231, + "manufactured home": 232, + "market indoor": 233, + "market outdoor": 234, + "marsh": 235, + "martial arts gym": 236, + "mausoleum": 237, + "medina": 238, + "moat water": 239, + "monastery outdoor": 240, + "mosque indoor": 241, + "mosque outdoor": 242, + "motel": 243, + "mountain": 244, + "mountain snowy": 245, + "movie theater indoor": 246, + "museum indoor": 247, + "music store": 248, + "music studio": 249, + "nuclear power plant outdoor": 250, + "nursery": 251, + "oast house": 252, + "observatory outdoor": 253, + "ocean": 254, + "office": 255, + "office building": 256, + "oil refinery outdoor": 257, + "oilrig": 258, + "operating room": 259, + "orchard": 260, + "outhouse outdoor": 261, + "pagoda": 262, + "palace": 263, + "pantry": 264, + "park": 265, + "parking garage indoor": 266, + "parking garage outdoor": 267, + "parking lot": 268, + "parlor": 269, + "pasture": 270, + "patio": 271, + "pavilion": 272, + "pharmacy": 273, + "phone booth": 274, + "physics laboratory": 275, + "picnic area": 276, + "pilothouse indoor": 277, + "planetarium outdoor": 278, + "playground": 279, + "playroom": 280, + "plaza": 281, + "podium indoor": 282, + "podium outdoor": 283, + "pond": 284, + "poolroom establishment": 285, + "poolroom home": 286, + "power plant outdoor": 287, + "promenade deck": 288, + "pub indoor": 289, + "pulpit": 290, + "putting green": 291, + "racecourse": 292, + "raceway": 293, + "raft": 294, + "railroad track": 295, + "rainforest": 296, + "reception": 297, + "recreation room": 298, + "residential neighborhood": 299, + "restaurant": 300, + "restaurant kitchen": 301, + "restaurant patio": 302, + "rice paddy": 303, + "riding arena": 304, + "river": 305, + "rock arch": 306, + "rope bridge": 307, + "ruin": 308, + "runway": 309, + "sandbar": 310, + "sandbox": 311, + "sauna": 312, + "schoolhouse": 313, + "sea cliff": 314, + "server room": 315, + "shed": 316, + "shoe shop": 317, + "shopfront": 318, + "shopping mall indoor": 319, + "shower": 320, + "skatepark": 321, + "ski lodge": 322, + "ski resort": 323, + "ski slope": 324, + "sky": 325, + "skyscraper": 326, + "slum": 327, + "snowfield": 328, + "squash court": 329, + "stable": 330, + "stadium baseball": 331, + "stadium football": 332, + "stage indoor": 333, + "staircase": 334, + "street": 335, + "subway interior": 336, + "subway station platform": 337, + "supermarket": 338, + "sushi bar": 339, + "swamp": 340, + "swimming pool indoor": 341, + "swimming pool outdoor": 342, + "synagogue indoor": 343, + "synagogue outdoor": 344, + "television studio": 345, + "temple east asia": 346, + "temple south asia": 347, + "tennis court indoor": 348, + "tennis court outdoor": 349, + "tent outdoor": 350, + "theater indoor procenium": 351, + "theater indoor seats": 352, + "thriftshop": 353, + "throne room": 354, + "ticket booth": 355, + "toll plaza": 356, + "topiary garden": 357, + "tower": 358, + "toyshop": 359, + "track outdoor": 360, + "train railway": 361, + "train station platform": 362, + "tree farm": 363, + "tree house": 364, + "trench": 365, + "underwater coral reef": 366, + "utility room": 367, + "valley": 368, + "van interior": 369, + "vegetable garden": 370, + "veranda": 371, + "veterinarians office": 372, + "viaduct": 373, + "videostore": 374, + "village": 375, + "vineyard": 376, + "volcano": 377, + "volleyball court indoor": 378, + "volleyball court outdoor": 379, + "waiting room": 380, + "warehouse indoor": 381, + "water tower": 382, + "waterfall block": 383, + "waterfall fan": 384, + "waterfall plunge": 385, + "watering hole": 386, + "wave": 387, + "wet bar": 388, + "wheat field": 389, + "wind farm": 390, + "windmill": 391, + "wine cellar barrel storage": 392, + "wine cellar bottle storage": 393, + "wrestling ring indoor": 394, + "yard": 395, + "youth hostel": 396 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-sun397-batch_size_64_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/README.md b/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..fa3c01dcc845c39243e6134cebf621a04710fbfb --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/config.json b/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json b/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.005_training_data_ratio_0.8-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2c8c5072376ce90afd118fb388ca83b31896893e --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2c8c5072376ce90afd118fb388ca83b31896893e --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 256 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_256_lr_0.01_training_data_ratio_1.0-4000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..289ab4ebbcf9bd9b056227211b3d07ef189e6e9f --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.5-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4a41899e5482a6d99aa57979744b9540503e51fd --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.005 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: true +save_top_k: -1 +training_data_ratio: 0.8 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.005_training_data_ratio_0.8-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4694698da5612c3ebd78d52832ba5fc522ba860e --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 0.5 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_0.5-3000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..70a88ccd1a42d4e5cc103ba27d6ac26b60dbc392 --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-1000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +} diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..70a88ccd1a42d4e5cc103ba27d6ac26b60dbc392 --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/README.md @@ -0,0 +1,80 @@ +--- +base_model: microsoft/resnet-50 +library_name: transformers +tags: +- fusion-bench +- merge +--- +# Deep Model Fusion + +Fine-tuned ResNet model on dataset svhn. + +## Models Merged + +This is a merged model created using [fusion-bench](https://github.com/tanganke/fusion_bench). + +The following models were included in the merge: + + + + +## Configuration + +The following YAML configuration was used to produce this model: + +### Algorithm Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.method.classification.image_classification_finetune.ImageClassificationFineTuning +_usage_: null +_version_: 0.2.25.dev0 +dataloader_kwargs: + batch_size: 64 + num_workers: 8 + pin_memory: true +label_smoothing: 0 +lr_scheduler: null +max_epochs: -1 +max_steps: 4000 +optimizer: + _target_: torch.optim.SGD + lr: 0.01 + momentum: 0.9 + weight_decay: 0.0001 +save_interval: 1000 +save_on_train_epoch_end: false +save_top_k: -1 +training_data_ratio: 1.0 +``` + +### Model Pool Configuration + +```yaml +_recursive_: false +_target_: fusion_bench.modelpool.resnet_for_image_classification.ResNetForImageClassificationPool +_usage_: null +_version_: 0.2.25.dev0 +models: + _pretrained_: + config_path: microsoft/resnet-50 + dataset_name: svhn + pretrained: true +test_datasets: null +train_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: train +type: transformers +val_datasets: + svhn: + _args_: + - svhn + - cropped_digits + _target_: datasets.load_dataset + split: test +``` + diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1edcdd247f31547feb0c0bebd29bf1cd24a4532f --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/config.json @@ -0,0 +1,63 @@ +{ + "architectures": [ + "ResNetForImageClassification" + ], + "depths": [ + 3, + 4, + 6, + 3 + ], + "downsample_in_bottleneck": false, + "downsample_in_first_stage": false, + "dtype": "float32", + "embedding_size": 64, + "hidden_act": "relu", + "hidden_sizes": [ + 256, + 512, + 1024, + 2048 + ], + "id2label": { + "0": "0", + "1": "1", + "2": "2", + "3": "3", + "4": "4", + "5": "5", + "6": "6", + "7": "7", + "8": "8", + "9": "9" + }, + "label2id": { + "0": 0, + "1": 1, + "2": 2, + "3": 3, + "4": 4, + "5": 5, + "6": 6, + "7": 7, + "8": 8, + "9": 9 + }, + "layer_type": "bottleneck", + "model_type": "resnet", + "num_channels": 3, + "out_features": [ + "stage4" + ], + "out_indices": [ + 4 + ], + "stage_names": [ + "stem", + "stage1", + "stage2", + "stage3", + "stage4" + ], + "transformers_version": "4.56.1" +} diff --git a/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8228c2b6b7368920da26fa97ecf6b616791fd03d --- /dev/null +++ b/resnet50-svhn-batch_size_64_lr_0.01_training_data_ratio_1.0-2000/preprocessor_config.json @@ -0,0 +1,22 @@ +{ + "crop_pct": 0.875, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.485, + 0.456, + 0.406 + ], + "image_processor_type": "ConvNextImageProcessor", + "image_std": [ + 0.229, + 0.224, + 0.225 + ], + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "shortest_edge": 224 + } +}