{ "model_class_name": "HookedViT", "model_name": "open-clip:laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K", "model_path": null, "hook_point_layer": 11, "layer_subtype": "hook_resid_post", "hook_point_head_index": null, "context_size": 50, "use_cached_activations": false, "use_patches_only": false, "cached_activations_path": "activations/_home_mila_p_praneet.suresh_prisma-v2_waterbird/openai_clip-vit-large-patch14/blocks.11.hook_resid_post", "d_in": 768, "activation_fn_str": "relu", "activation_fn_kwargs": {}, "cls_token_only": false, "max_grad_norm": 1.0, "initialization_method": "encoder_transpose_decoder", "normalize_activations": "layer_norm", "n_batches_in_buffer": 20, "store_batch_size": 32, "num_workers": 4, "num_epochs": 15, "total_training_images": 143880, "total_training_tokens": 7194000, "image_size": 224, "device": { "__type__": "torch.device", "value": "cuda" }, "seed": 42, "dtype": { "__type__": "torch.dtype", "value": "torch.float32" }, "architecture": "standard", "verbose": false, "b_dec_init_method": "geometric_median", "expansion_factor": 64, "from_pretrained_path": null, "d_sae": 49152, "l1_coefficient": 0.1413670469265696, "lp_norm": 1, "lr": 0.011813686729329077, "lr_scheduler_name": "cosineannealingwarmup", "lr_warm_up_steps": 500, "train_batch_size": 8192, "dataset_name": "waterbirds", "dataset_path": "/home/mila/p/praneet.suresh/prisma-v2/waterbird", "dataset_train_path": null, "dataset_val_path": null, "use_ghost_grads": true, "feature_sampling_window": 1000, "dead_feature_window": 5000, "dead_feature_threshold": 1e-08, "log_to_wandb": true, "wandb_project": "waterbirds_finetuning_sweep", "wandb_entity": null, "wandb_log_frequency": 1, "n_validation_runs": 100, "n_checkpoints": 10, "checkpoint_path": "/network/scratch/p/praneet.suresh/open_clip_waterbirds_checkpoints/a82fc967-waterbirds_finetuning_sweep" }