diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000000000000000000000000000000000000..7d3ed876292ce8d883c5257aefd3917e15024946 --- /dev/null +++ b/.gitattributes @@ -0,0 +1 @@ +*.ckpt filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..f1d3157643bd403ea9609c289e1abcb005f75b1a --- /dev/null +++ b/README.md @@ -0,0 +1,123 @@ +--- +license: apache-2.0 +library_name: pytorch +tags: +- jepa +- rectified-lp-jepa +- self-supervised-learning +- computer-vision +- pytorch-lightning +- representation-learning +--- + +# Rectified LP-JEPA Checkpoints + +This repository contains checkpoints for Rectified LP-JEPA models trained for 1000 epochs (unless otherwise noted). These models correspond to the results presented in Table 1 of our paper. + +For the full training code and implementation details, please refer to our main codebase: [rectified-lp-jepa](https://github.com/YilunKuang/rectified-lp-jepa). + +## Performance Card + +The following table summarizes the performance of the provided checkpoints on ImageNet-1k (val_acc1) and other metrics: + +| Name | val_acc1 | val_proj_acc1 | train_l1_sparsity | train_l0_sparsity | +| :--- | :--- | :--- | :--- | :--- | +| p2.0-mu1.0 | 85.28 | 79.64 | 0.6291 | 0.8678 | +| p2.0-mu0.5 | 85.40 | 80.02 | 0.4899 | 0.8024 | +| p2.0-mu0.75 | 85.10 | 80.96 | 0.5619 | 0.8360 | +| p2.0-mu0.25 | 84.94 | 80.10 | 0.4164 | 0.7685 | +| p2.0-mu0.0 | 84.58 | 80.30 | 0.3424 | 0.7301 | +| p2.0-mu-0.5 | 83.98 | 79.88 | 0.2079 | 0.6681 | +| p2.0-mu-0.25 | 84.70 | 79.24 | 0.2716 | 0.6955 | +| p2.0-mu-0.75 | 84.90 | 79.30 | 0.1529 | 0.6370 | +| p2.0-mu-1.0 | 84.48 | 79.54 | 0.1068 | 0.6042 | +| p2.0-mu-1.25 | 84.50 | 79.00 | 0.0717 | 0.4839 | +| p2.0-mu-1.50 | 84.58 | 77.84 | 0.0469 | 0.2807 | +| p2.0-mu-2.50 | 82.48 | 68.42 | 0.0126 | 0.0206 | +| p2.0-mu-2.0 | 83.06 | 72.38 | 0.0194 | 0.0420 | +| p2.0-mu-2.25 | 82.22 | 71.08 | 0.0158 | 0.0290 | +| p2.0-mu-2.75 | 81.14 | 64.02 | 0.0128 | 0.0189 | +| p1.0-mu-1.75 | 84.68 | 77.84 | 0.0300 | 0.1244 | +| p1.0-mu-2.0 | 84.06 | 76.00 | 0.0248 | 0.0742 | +| p1.0-mu-2.75 | 83.10 | 72.88 | 0.0068 | 0.0119 | +| p2.0-mu-1.75 | 83.62 | 76.10 | 0.0290 | 0.0898 | +| p1.0-mu0.0 | 84.76 | 80.14 | 0.2743 | 0.6943 | +| p1.0-mu0.25 | 85.80 | 80.40 | 0.3752 | 0.7444 | +| p1.0-mu-1.50 | 84.98 | 78.38 | 0.0397 | 0.2329 | +| p1.0-mu-0.25 | 85.02 | 79.88 | 0.1982 | 0.6461 | +| p1.0-mu-2.50 | 84.42 | 73.86 | 0.0134 | 0.0269 | +| p1.0-mu-2.25 | 84.20 | 74.50 | 0.0207 | 0.0507 | +| p1.0-mu-0.5 | 84.87 | 79.48 | 0.1430 | 0.6012 | +| p1.0-mu1.0 | 85.23 | 79.52 | 0.6474 | 0.8770 | +| p2.0-mu-3.0 | 78.44 | 56.40 | 0.0047 | 0.0060 | +| p1.0-mu-1.0 | 84.64 | 79.66 | 0.0745 | 0.4784 | +| p1.0-mu-1.25 | 84.54 | 78.60 | 0.0538 | 0.3675 | +| p1.0-mu-0.75 | 84.48 | 79.42 | 0.1036 | 0.5463 | +| p1.0-mu0.5 | 84.42 | 80.70 | 0.4774 | 0.7986 | +| p1.0-mu0.75 | 85.58 | 80.38 | 0.5703 | 0.8423 | +| p1.0-mu-3.0 | 82.66 | 72.32 | 0.0062 | 0.0109 | + +> **Note:** The `p1.0-mu-3.0` checkpoint is currently at 934 epochs. We plan to update it to the full 1000-epoch version soon. + +## Usage + +To use these checkpoints, first clone the original codebase: + +```bash +git clone https://github.com/YilunKuang/rectified-lp-jepa +cd rectified-lp-jepa +``` + +### Loading a Checkpoint for Inference + +Each model directory contains a `.ckpt` file and its corresponding `args.json`. You can load them as follows: + +```python +import json +import torch +from pathlib import Path +from omegaconf import OmegaConf +from solo.methods import METHODS + +# Setup paths (example for p2.0-mu-1.75) +ckpt_path = "p2.0-mu-1.75-1000epoch/p2.0-mu-1.75-1000epoch.ckpt" +args_path = "p2.0-mu-1.75-1000epoch/args.json" + +# 1. Load Config +with open(args_path) as f: + method_args = json.load(f) +cfg = OmegaConf.create(method_args) + +# 2. Load Model +print(f"Loading model from {ckpt_path}...") +model = METHODS[method_args["method"]].load_from_checkpoint(ckpt_path, strict=False, cfg=cfg) +model.cuda().eval() + +# 3. Inference +# Input should be a normalized 224x224 tensor +with torch.no_grad(): + features = model.encoder(input_tensor) + if hasattr(model, "projector"): + projected = model.projector(features) +``` + +## Acknowledgements + +This codebase is built upon the [solo-learn](https://github.com/vturrisi/solo-learn) framework. +We thank the solo-learn authors for releasing their code under the MIT license. + +## Citation + +Please cite our work if you find it helpful: + +```bibtex +@misc{kuang2026rectifiedlpjepajointembeddingpredictive, + title={Rectified LpJEPA: Joint-Embedding Predictive Architectures with Sparse and Maximum-Entropy Representations}, + author={Yilun Kuang and Yash Dagade and Tim G. J. Rudner and Randall Balestriero and Yann LeCun}, + year={2026}, + eprint={2602.01456}, + archivePrefix={arXiv}, + primaryClass={cs.LG}, + url={https://arxiv.org/abs/2602.01456}, +} +``` diff --git a/p1.0-mu-0.25-1000epoch/args.json b/p1.0-mu-0.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..dc37e28d81551195585075cfaf37924a39bb6db2 --- /dev/null +++ b/p1.0-mu-0.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-0.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -0.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "jtxvrrqi"} \ No newline at end of file diff --git a/p1.0-mu-0.25-1000epoch/p1.0-mu-0.25-1000epoch.ckpt b/p1.0-mu-0.25-1000epoch/p1.0-mu-0.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..6e9acef5ac98a585f7cbad3176a9aefcce6508de --- /dev/null +++ b/p1.0-mu-0.25-1000epoch/p1.0-mu-0.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b3e5e1bb93675c416a120c13228befb8d407a538a28660ee21ef0e64a76afdc0 +size 292535924 diff --git a/p1.0-mu-0.5-1000epoch/args.json b/p1.0-mu-0.5-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..0c0b3f879d138e20887e323a1fd2dbceca9e2a15 --- /dev/null +++ b/p1.0-mu-0.5-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-0.5", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -0.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "1dagp9ni"} \ No newline at end of file diff --git a/p1.0-mu-0.5-1000epoch/p1.0-mu-0.5-1000epoch.ckpt b/p1.0-mu-0.5-1000epoch/p1.0-mu-0.5-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..f8a6c606acbc1d2e891c2b453dbb830edc53074a --- /dev/null +++ b/p1.0-mu-0.5-1000epoch/p1.0-mu-0.5-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:839f613ee418185c9ee4a04fe221213200c1e3bae4db431ab429080ac2848973 +size 292535924 diff --git a/p1.0-mu-0.75-1000epoch/args.json b/p1.0-mu-0.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..4f58190a8668f4901095a75ece122fc39b388cdd --- /dev/null +++ b/p1.0-mu-0.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-0.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -0.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "nd5n979z"} \ No newline at end of file diff --git a/p1.0-mu-0.75-1000epoch/p1.0-mu-0.75-1000epoch.ckpt b/p1.0-mu-0.75-1000epoch/p1.0-mu-0.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..f4fe3f85a26cc5ac9e6682af0c6d59b91911b527 --- /dev/null +++ b/p1.0-mu-0.75-1000epoch/p1.0-mu-0.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:958b0b29326ed099fa313c9cb8d5e0b27419f0a468fe8bb373f4dea0bdd4569d +size 292535924 diff --git a/p1.0-mu-1.0-1000epoch/args.json b/p1.0-mu-1.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..cd65f21136bc5688d95b0e70f0877520a6ce67bd --- /dev/null +++ b/p1.0-mu-1.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-1.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "qn3vlwnv"} \ No newline at end of file diff --git a/p1.0-mu-1.0-1000epoch/p1.0-mu-1.0-1000epoch.ckpt b/p1.0-mu-1.0-1000epoch/p1.0-mu-1.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..88163696ff2f988288f8f2f8fde844197582234a --- /dev/null +++ b/p1.0-mu-1.0-1000epoch/p1.0-mu-1.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fae4bbd85990ade50b419a7adcd4c2b8082ab1a95b7c41ab7120e2414aca1291 +size 292535924 diff --git a/p1.0-mu-1.25-1000epoch/args.json b/p1.0-mu-1.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..6fb9f0d2e63c63439434dacdbf15b214d4d2cdd8 --- /dev/null +++ b/p1.0-mu-1.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-1.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "500vzrnu"} \ No newline at end of file diff --git a/p1.0-mu-1.25-1000epoch/p1.0-mu-1.25-1000epoch.ckpt b/p1.0-mu-1.25-1000epoch/p1.0-mu-1.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..274530020a3de418b288cb6a4c68789cb06fa2f8 --- /dev/null +++ b/p1.0-mu-1.25-1000epoch/p1.0-mu-1.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7064a07587888a03d73772846114e767e5fb5e16a09ca1f72478dd71c6025bcc +size 292535924 diff --git a/p1.0-mu-1.50-1000epoch/args.json b/p1.0-mu-1.50-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..114d5f49a601a82ffcfdea920b3aaa6cc4529a60 --- /dev/null +++ b/p1.0-mu-1.50-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-1.50", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "nkuszl74"} \ No newline at end of file diff --git a/p1.0-mu-1.50-1000epoch/p1.0-mu-1.50-1000epoch.ckpt b/p1.0-mu-1.50-1000epoch/p1.0-mu-1.50-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..3e6474a15df6f20819c0da2db13e0ae07edbac58 --- /dev/null +++ b/p1.0-mu-1.50-1000epoch/p1.0-mu-1.50-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df1939803b5c1fd962729c98282be8b91c9621c86f16710bcab1847d9f616454 +size 292535924 diff --git a/p1.0-mu-1.75-1000epoch/args.json b/p1.0-mu-1.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..9fe5d5a4ba1f988cb2152efe04e5cfa8fc57f30a --- /dev/null +++ b/p1.0-mu-1.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-1.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "dqvf43vf"} \ No newline at end of file diff --git a/p1.0-mu-1.75-1000epoch/p1.0-mu-1.75-1000epoch.ckpt b/p1.0-mu-1.75-1000epoch/p1.0-mu-1.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..6733c98b7aa2d5f88455348e84b67fa311ac9510 --- /dev/null +++ b/p1.0-mu-1.75-1000epoch/p1.0-mu-1.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0c5de8ca78feb58d379007f9d067769d1f25addfa3e2fadbf2065252364d017 +size 292535924 diff --git a/p1.0-mu-2.0-1000epoch/args.json b/p1.0-mu-2.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..b5ebe5238a3eb555e93285a749282040688ec891 --- /dev/null +++ b/p1.0-mu-2.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-2.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "owzi5nyb"} \ No newline at end of file diff --git a/p1.0-mu-2.0-1000epoch/p1.0-mu-2.0-1000epoch.ckpt b/p1.0-mu-2.0-1000epoch/p1.0-mu-2.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..db78d4b17f64e9f058637d031dd4b86175564178 --- /dev/null +++ b/p1.0-mu-2.0-1000epoch/p1.0-mu-2.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:deb6a858ae4d71cf3b2561211db6592a73693042933b3488141b01184adb00df +size 292535924 diff --git a/p1.0-mu-2.25-1000epoch/args.json b/p1.0-mu-2.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..bb237ce32af0df93457d1e6d8e40627447f032e2 --- /dev/null +++ b/p1.0-mu-2.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-2.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "6x160q84"} \ No newline at end of file diff --git a/p1.0-mu-2.25-1000epoch/p1.0-mu-2.25-1000epoch.ckpt b/p1.0-mu-2.25-1000epoch/p1.0-mu-2.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..d3dccad43a35594ee0a56bf938edc2bbe64517fa --- /dev/null +++ b/p1.0-mu-2.25-1000epoch/p1.0-mu-2.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:43df060e125310b91e2bd5f4405dfc5e1e88f1f0f4e81323693807f8a0010dc6 +size 292535924 diff --git a/p1.0-mu-2.50-1000epoch/args.json b/p1.0-mu-2.50-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..0143630e28898b5002f738d389b17f35a3fa00d8 --- /dev/null +++ b/p1.0-mu-2.50-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-2.50", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "anfm5w1c"} \ No newline at end of file diff --git a/p1.0-mu-2.50-1000epoch/p1.0-mu-2.50-1000epoch.ckpt b/p1.0-mu-2.50-1000epoch/p1.0-mu-2.50-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..778564e084dde6ea35120f84b5fed3ffd1e3064b --- /dev/null +++ b/p1.0-mu-2.50-1000epoch/p1.0-mu-2.50-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c64aaca7b5971bf607a8aa0a49c91751463f09c79643fbfa3e01511739a94954 +size 292535924 diff --git a/p1.0-mu-2.75-1000epoch/args.json b/p1.0-mu-2.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..178b5a6c56f0421c2b2778e6738ba03e8f535efe --- /dev/null +++ b/p1.0-mu-2.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-2.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "yvocnok8"} \ No newline at end of file diff --git a/p1.0-mu-2.75-1000epoch/p1.0-mu-2.75-1000epoch.ckpt b/p1.0-mu-2.75-1000epoch/p1.0-mu-2.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..f8941f864935ef13ccfe8a587654dd654d91eea0 --- /dev/null +++ b/p1.0-mu-2.75-1000epoch/p1.0-mu-2.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5e5e72a82f2d7f2b9f53040a6f334e9013b54cdbe021a47606955ba7e5bc7e7d +size 292535924 diff --git a/p1.0-mu-3.0-934epoch/args.json b/p1.0-mu-3.0-934epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..3b19a1a69b1532629f288796599829c0dfc856ab --- /dev/null +++ b/p1.0-mu-3.0-934epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu-3.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -3.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "7xsguytv"} \ No newline at end of file diff --git a/p1.0-mu-3.0-934epoch/p1.0-mu-3.0-934epoch.ckpt b/p1.0-mu-3.0-934epoch/p1.0-mu-3.0-934epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..76ec5c79082807a8120744dc01170246f114e92a --- /dev/null +++ b/p1.0-mu-3.0-934epoch/p1.0-mu-3.0-934epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a4357456db37e028cfaff170fed1bba13c1bdbcf5bb85be41ad94a77c68d07ee +size 292535924 diff --git a/p1.0-mu0.0-1000epoch/args.json b/p1.0-mu0.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..c71cc81e1cfbcc40b1fc43496b40658717627ef6 --- /dev/null +++ b/p1.0-mu0.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu0.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "sq6713ww"} \ No newline at end of file diff --git a/p1.0-mu0.0-1000epoch/p1.0-mu0.0-1000epoch.ckpt b/p1.0-mu0.0-1000epoch/p1.0-mu0.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..3fa2c87b38ef9cf5b4e4dfee5444760822a0836a --- /dev/null +++ b/p1.0-mu0.0-1000epoch/p1.0-mu0.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e726eb46ad137ffe509982cae2b9ffe6c8140f7a410f6312a6545784b2378895 +size 292535924 diff --git a/p1.0-mu0.25-1000epoch/args.json b/p1.0-mu0.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..3053d3458b4fa4ee0f70b19229dc1df0c5ee49fc --- /dev/null +++ b/p1.0-mu0.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu0.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "o6plm6ac"} \ No newline at end of file diff --git a/p1.0-mu0.25-1000epoch/p1.0-mu0.25-1000epoch.ckpt b/p1.0-mu0.25-1000epoch/p1.0-mu0.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..aaa9c2d329391228f8fb380011e66578031e3b49 --- /dev/null +++ b/p1.0-mu0.25-1000epoch/p1.0-mu0.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe88a951a31642b7dbea14c0e03b60c49daca6085fc0af0fab54772ee59f10be +size 292535924 diff --git a/p1.0-mu0.5-1000epoch/args.json b/p1.0-mu0.5-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..e666e2416c1f319a4b6ecbb94e919fc9aff145c0 --- /dev/null +++ b/p1.0-mu0.5-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu0.5", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "hayk6jvd"} \ No newline at end of file diff --git a/p1.0-mu0.5-1000epoch/p1.0-mu0.5-1000epoch.ckpt b/p1.0-mu0.5-1000epoch/p1.0-mu0.5-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..4e87f7d33f14195aa32f7faee97409f83eacb29c --- /dev/null +++ b/p1.0-mu0.5-1000epoch/p1.0-mu0.5-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47bbe37b02dcb14b36af6107eab04974834ac817faba0bbe87a88a0f198c4ef4 +size 292535924 diff --git a/p1.0-mu0.75-1000epoch/args.json b/p1.0-mu0.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..c9c922d86a411bacf6e7fe45fb8f82aa4a78af4a --- /dev/null +++ b/p1.0-mu0.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu0.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "854vyeds"} \ No newline at end of file diff --git a/p1.0-mu0.75-1000epoch/p1.0-mu0.75-1000epoch.ckpt b/p1.0-mu0.75-1000epoch/p1.0-mu0.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..2a5f5e0a2f7e4334b7de97fb8d8be5307086b9ce --- /dev/null +++ b/p1.0-mu0.75-1000epoch/p1.0-mu0.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:856f68a7bf7baf95d277ff9fc68ac17251827838203913f5d66ffbf50f19fd3f +size 292535924 diff --git a/p1.0-mu1.0-1000epoch/args.json b/p1.0-mu1.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..22f752be2737a01a001e1a9e27ea2f147feb8202 --- /dev/null +++ b/p1.0-mu1.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p1.0-mu1.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 1.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 1.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "2m1hlv99"} \ No newline at end of file diff --git a/p1.0-mu1.0-1000epoch/p1.0-mu1.0-1000epoch.ckpt b/p1.0-mu1.0-1000epoch/p1.0-mu1.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..d72fda6bd487cb305c6cd791f834df4f32bb5d0b --- /dev/null +++ b/p1.0-mu1.0-1000epoch/p1.0-mu1.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d00c430a865cd03abf8d2dd446c3fd236c0ada9d77619d02ada6d7cfd10472d +size 292535924 diff --git a/p2.0-mu-0.25-1000epoch/args.json b/p2.0-mu-0.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..199f22af630fb39c494795dd389b4a6a35485bea --- /dev/null +++ b/p2.0-mu-0.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-0.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -0.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "9w0sgkqy"} \ No newline at end of file diff --git a/p2.0-mu-0.25-1000epoch/p2.0-mu-0.25-1000epoch.ckpt b/p2.0-mu-0.25-1000epoch/p2.0-mu-0.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..53ced112d960f32790e64f60e614662f65970718 --- /dev/null +++ b/p2.0-mu-0.25-1000epoch/p2.0-mu-0.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4695872c63ea7d9a8411cad8c1c5ebdb10e1023868c83a78f84ccd77c219aa44 +size 292535924 diff --git a/p2.0-mu-0.5-1000epoch/args.json b/p2.0-mu-0.5-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..510d1064b4faf27ce5e9efa02c12e615fec074b7 --- /dev/null +++ b/p2.0-mu-0.5-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-0.5", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -0.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "8p02n68z"} \ No newline at end of file diff --git a/p2.0-mu-0.5-1000epoch/p2.0-mu-0.5-1000epoch.ckpt b/p2.0-mu-0.5-1000epoch/p2.0-mu-0.5-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..58f1c5f384d725abed76b0c011b35423300c570f --- /dev/null +++ b/p2.0-mu-0.5-1000epoch/p2.0-mu-0.5-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a3cb1fb79978df7a8001d5d73071ebb6ba98cd53f21f383404c7d2396d67b840 +size 292535924 diff --git a/p2.0-mu-0.75-1000epoch/args.json b/p2.0-mu-0.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..bab933021518f1f573b5e50f6efb15bb9b9eeb8c --- /dev/null +++ b/p2.0-mu-0.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-0.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -0.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 168}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "j72544ei"} \ No newline at end of file diff --git a/p2.0-mu-0.75-1000epoch/p2.0-mu-0.75-1000epoch.ckpt b/p2.0-mu-0.75-1000epoch/p2.0-mu-0.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..c8d860799411842dcf601b825df3fba3c15dcded --- /dev/null +++ b/p2.0-mu-0.75-1000epoch/p2.0-mu-0.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe9753bbad8d41edc3e7d9d2be14f2ef367ea79b70e5090848451f37507c1259 +size 292535924 diff --git a/p2.0-mu-1.0-1000epoch/args.json b/p2.0-mu-1.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..4fbfe49f2cca56a8ba8eec618865de89bb397551 --- /dev/null +++ b/p2.0-mu-1.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-1.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "d1oylwuw"} \ No newline at end of file diff --git a/p2.0-mu-1.0-1000epoch/p2.0-mu-1.0-1000epoch.ckpt b/p2.0-mu-1.0-1000epoch/p2.0-mu-1.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..afd5fba8ca75b750023c04205fb9450d9d4bd575 --- /dev/null +++ b/p2.0-mu-1.0-1000epoch/p2.0-mu-1.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c280c7f933692fe6aa9e8a8d11ae945ec789d3aac0fd6cf96b31febc711a34eb +size 292535924 diff --git a/p2.0-mu-1.25-1000epoch/args.json b/p2.0-mu-1.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..da35251f915854d951eba2dd7370911346d95618 --- /dev/null +++ b/p2.0-mu-1.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-1.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "3xbfmfdz"} \ No newline at end of file diff --git a/p2.0-mu-1.25-1000epoch/p2.0-mu-1.25-1000epoch.ckpt b/p2.0-mu-1.25-1000epoch/p2.0-mu-1.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..ba4c21deb07b2b3c1b2774324ba6415ea4cbec17 --- /dev/null +++ b/p2.0-mu-1.25-1000epoch/p2.0-mu-1.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d13e66b08236a0c1cb9dc6c40a76aa2b02014b8915344fdf20c947fab977df6 +size 292535924 diff --git a/p2.0-mu-1.50-1000epoch/args.json b/p2.0-mu-1.50-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..20bcac93397331fddf3ca188a03f420f8f8b7507 --- /dev/null +++ b/p2.0-mu-1.50-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-1.50", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "69iligvf"} \ No newline at end of file diff --git a/p2.0-mu-1.50-1000epoch/p2.0-mu-1.50-1000epoch.ckpt b/p2.0-mu-1.50-1000epoch/p2.0-mu-1.50-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..84710e39e7957c17df9d7f5b3261cdfec6dc181b --- /dev/null +++ b/p2.0-mu-1.50-1000epoch/p2.0-mu-1.50-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1ec723fb5745c20da6cb32e297904725a07f37e0cfe5d8500a70975e8098de26 +size 292535924 diff --git a/p2.0-mu-1.75-1000epoch/args.json b/p2.0-mu-1.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..ba6319884caae179d4b0cb9d3fa2fb1475119203 --- /dev/null +++ b/p2.0-mu-1.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-1.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -1.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "m6ag6dvn"} \ No newline at end of file diff --git a/p2.0-mu-1.75-1000epoch/p2.0-mu-1.75-1000epoch.ckpt b/p2.0-mu-1.75-1000epoch/p2.0-mu-1.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..fd1168c860c3e7edc6529ea487df60e7f17fa4ae --- /dev/null +++ b/p2.0-mu-1.75-1000epoch/p2.0-mu-1.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1b8f02183e72878c1f886d45f612583dae50c6c8cc3e30f9f9937a4f89fd798a +size 292535924 diff --git a/p2.0-mu-2.0-1000epoch/args.json b/p2.0-mu-2.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..b2c033c72a9a56de89e90fbf002b9e13a8048cc2 --- /dev/null +++ b/p2.0-mu-2.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-2.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "z7nziir8"} \ No newline at end of file diff --git a/p2.0-mu-2.0-1000epoch/p2.0-mu-2.0-1000epoch.ckpt b/p2.0-mu-2.0-1000epoch/p2.0-mu-2.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..fbc4f8fbefafa5853f8a6e5263a41b1a95635e1a --- /dev/null +++ b/p2.0-mu-2.0-1000epoch/p2.0-mu-2.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fccaaa9e2875e399259bb90e3ddc15ca62fe32221b843ddf5f694b073aabe6f2 +size 292535924 diff --git a/p2.0-mu-2.25-1000epoch/args.json b/p2.0-mu-2.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..7d55a0369490fa4550ac661d6b01208242376103 --- /dev/null +++ b/p2.0-mu-2.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-2.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "yyoak10r"} \ No newline at end of file diff --git a/p2.0-mu-2.25-1000epoch/p2.0-mu-2.25-1000epoch.ckpt b/p2.0-mu-2.25-1000epoch/p2.0-mu-2.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..a0515a77755ff73f66f5db9cb6f18c256b80766f --- /dev/null +++ b/p2.0-mu-2.25-1000epoch/p2.0-mu-2.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0b4b79d6af12bfcaae6e1133b04c7c176a2a43c2fc5c9c149ce617cc46d0886 +size 292535924 diff --git a/p2.0-mu-2.50-1000epoch/args.json b/p2.0-mu-2.50-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..e3c18005e601f50208d4a01affce1877d21785c1 --- /dev/null +++ b/p2.0-mu-2.50-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-2.50", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "ibmxkbyc"} \ No newline at end of file diff --git a/p2.0-mu-2.50-1000epoch/p2.0-mu-2.50-1000epoch.ckpt b/p2.0-mu-2.50-1000epoch/p2.0-mu-2.50-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..80b4995ac98562ae0658eb8b653aca0379941798 --- /dev/null +++ b/p2.0-mu-2.50-1000epoch/p2.0-mu-2.50-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:047f5d5b3d1da137ff7e500262e7220f2c4f9014cc8e599ee48a3a93761cec37 +size 292535924 diff --git a/p2.0-mu-2.75-1000epoch/args.json b/p2.0-mu-2.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..a43a34f5b1a3e0234f513cd7ad1f87eae3681759 --- /dev/null +++ b/p2.0-mu-2.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-2.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -2.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "vu17equn"} \ No newline at end of file diff --git a/p2.0-mu-2.75-1000epoch/p2.0-mu-2.75-1000epoch.ckpt b/p2.0-mu-2.75-1000epoch/p2.0-mu-2.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..751365c8a3b0acdfee8868187f72527c04edcca6 --- /dev/null +++ b/p2.0-mu-2.75-1000epoch/p2.0-mu-2.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b19aab16f9bb824dc240c890bfe19c11691573575724f6c67a8f212ea25af34b +size 292535924 diff --git a/p2.0-mu-3.0-1000epoch/args.json b/p2.0-mu-3.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..3370a54ea7a573f93dafbc36272f6f51284b1247 --- /dev/null +++ b/p2.0-mu-3.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu-3.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": -3.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "kf94kdlf"} \ No newline at end of file diff --git a/p2.0-mu-3.0-1000epoch/p2.0-mu-3.0-1000epoch.ckpt b/p2.0-mu-3.0-1000epoch/p2.0-mu-3.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..66d951c0547459c86a5d9afd098a16f8d2be6c6a --- /dev/null +++ b/p2.0-mu-3.0-1000epoch/p2.0-mu-3.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d27e96ae51bdf98ff087230ff6489c6cfebc70d2481670e905c322065df5104 +size 292535924 diff --git a/p2.0-mu0.0-1000epoch/args.json b/p2.0-mu0.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..4afde26392173845303f0b74415c88c353d2df66 --- /dev/null +++ b/p2.0-mu0.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu0.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "muczuwvv"} \ No newline at end of file diff --git a/p2.0-mu0.0-1000epoch/p2.0-mu0.0-1000epoch.ckpt b/p2.0-mu0.0-1000epoch/p2.0-mu0.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..95a06b66f3831dceb9e2ad5f0ffdbce7053e23e9 --- /dev/null +++ b/p2.0-mu0.0-1000epoch/p2.0-mu0.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8bcc7ffb02269992acfdaf8719f4669f2552c7568d9b67ffd53c9fcf194522e3 +size 292535924 diff --git a/p2.0-mu0.25-1000epoch/args.json b/p2.0-mu0.25-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..0577b673d7b1524b46d8cada9b450a9eb9029ec7 --- /dev/null +++ b/p2.0-mu0.25-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu0.25", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.25, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "noijmk92"} \ No newline at end of file diff --git a/p2.0-mu0.25-1000epoch/p2.0-mu0.25-1000epoch.ckpt b/p2.0-mu0.25-1000epoch/p2.0-mu0.25-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..4ccfa4f7c95a47d6630722933006dac2d577f428 --- /dev/null +++ b/p2.0-mu0.25-1000epoch/p2.0-mu0.25-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cc8b45b6c7bd47932903b09dbe3c2abfb0e905b2cc3ea28763bc4752c22094fe +size 292535924 diff --git a/p2.0-mu0.5-1000epoch/args.json b/p2.0-mu0.5-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..93fceee4014dde9d4260b365758bdd3e98ccad82 --- /dev/null +++ b/p2.0-mu0.5-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu0.5", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.5, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "6fnwy8sa"} \ No newline at end of file diff --git a/p2.0-mu0.5-1000epoch/p2.0-mu0.5-1000epoch.ckpt b/p2.0-mu0.5-1000epoch/p2.0-mu0.5-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..913c5de8806a2b5b1295c07a2ba522df16c4ea53 --- /dev/null +++ b/p2.0-mu0.5-1000epoch/p2.0-mu0.5-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7c8e6ac35318de3be45cbc385bfca81dc5b744e23e2f7d7c5ceed95a20dae7c9 +size 292535924 diff --git a/p2.0-mu0.75-1000epoch/args.json b/p2.0-mu0.75-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..94afcc16609e4fdbab066cb5fdfa0ee1aec02072 --- /dev/null +++ b/p2.0-mu0.75-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu0.75", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 0.75, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "jhvpwt21"} \ No newline at end of file diff --git a/p2.0-mu0.75-1000epoch/p2.0-mu0.75-1000epoch.ckpt b/p2.0-mu0.75-1000epoch/p2.0-mu0.75-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..9fb6aef62567f7c11fc3f36af24b1d1f8ab6679e --- /dev/null +++ b/p2.0-mu0.75-1000epoch/p2.0-mu0.75-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:67b9124966e14c27f12b9f7a9b1abde036890f632124a75dab6b271e7199ffa9 +size 292535924 diff --git a/p2.0-mu1.0-1000epoch/args.json b/p2.0-mu1.0-1000epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..8d7c9d66d2780ce5ad0bac49e70e35d16ec17dfe --- /dev/null +++ b/p2.0-mu1.0-1000epoch/args.json @@ -0,0 +1 @@ +{"name": "p2.0-mu1.0", "method": "rectified_lpjepa", "backbone": {"name": "resnet50", "kwargs": {}}, "method_kwargs": {"target_distribution": "rectified_lp_distribution", "lp_norm_parameter": 2.0, "proj_hidden_dim": 2048, "proj_output_dim": 2048, "invariance_loss_weight": 25.0, "rdm_reg_loss_weight": 125.0, "projector_type": "rectified_mlp", "num_projections": 8192, "projection_vectors_type": "random", "add_projector_classifier": true, "mean_shift_value": 1.0, "mode_of_sigma": "sigma_GN", "logging_interval": 50, "active_feature_threshold": 0.001}, "mlp_probe": {"enabled": false, "num_layers": 3}, "data": {"dataset": "imagenet100", "train_path": "/imagenet100_real/train", "val_path": "/imagenet100_real/val", "preload": true, "format": "image_folder", "num_workers": 8, "no_labels": false, "fraction": -1, "num_classes": 100, "num_large_crops": 2, "num_small_crops": 0}, "augmentations": [{"rrc": {"enabled": true, "crop_min_scale": 0.2, "crop_max_scale": 1.0}, "color_jitter": {"enabled": true, "brightness": 0.4, "contrast": 0.4, "saturation": 0.2, "hue": 0.1, "prob": 0.8}, "grayscale": {"enabled": true, "prob": 0.2}, "gaussian_blur": {"enabled": true, "prob": 0.5}, "solarization": {"enabled": true, "prob": 0.1}, "equalization": {"enabled": false, "prob": 0.0}, "horizontal_flip": {"enabled": true, "prob": 0.5}, "crop_size": 224, "num_crops": 2}], "optimizer": {"name": "lars", "batch_size": 128, "lr": 0.0825, "classifier_lr": 0.0275, "weight_decay": 0.0001, "kwargs": {"clip_lr": true, "eta": 0.02, "exclude_bias_n_norm": true, "momentum": 0.9}, "exclude_bias_n_norm_wd": false}, "scheduler": {"name": "warmup_cosine", "lr_decay_steps": null, "min_lr": 0.0, "warmup_start_lr": 3e-05, "warmup_epochs": 10, "interval": "step"}, "checkpoint": {"enabled": true, "dir": "trained_models", "frequency": 1, "keep_prev": false}, "auto_resume": {"enabled": true, "max_hours": 36}, "max_epochs": 1000, "devices": [0], "sync_batchnorm": true, "accelerator": "gpu", "strategy": "ddp", "precision": "16-mixed", "wandb": {"enabled": true, "entity": "yash21-dagade-duke-university", "project": "ICML-Table-1-Reproduction", "offline": false}, "auto_umap": {"enabled": false, "dir": "auto_umap", "frequency": 1}, "debug_augmentations": false, "seed": 5, "resume_from_checkpoint": null, "num_nodes": 1, "accumulate_grad_batches": 1, "knn_eval": {"enabled": false, "k": 20, "distance_func": "euclidean"}, "performance": {"disable_channel_last": false}, "wandb_run_id": "t5ujsb7u"} \ No newline at end of file diff --git a/p2.0-mu1.0-1000epoch/p2.0-mu1.0-1000epoch.ckpt b/p2.0-mu1.0-1000epoch/p2.0-mu1.0-1000epoch.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..fe72242320ceadf678482d70912438dc2fbd716d --- /dev/null +++ b/p2.0-mu1.0-1000epoch/p2.0-mu1.0-1000epoch.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c4acb1eeb279322878e8f030c064d0e7cfc2b8d38dca3df59ebb924e2dc10c58 +size 292535924