| {"sae": {"expansion_factor": 16, "normalize_decoder": true, "num_latents": 0, "k": 64, "multi_topk": false}, "batch_size": 8, "grad_acc_steps": 1, "micro_acc_steps": 1, "lr": null, "lr_warmup_steps": 1000, "auxk_alpha": 0.0, "dead_feature_threshold": 10000000, "hookpoints": ["layers.12", "layers.24"], "layers": [12, 24], "layer_stride": 1, "distribute_modules": false, "save_every": 1000, "log_to_wandb": true, "run_name": null, "wandb_log_frequency": 1, "model": "Qwen/Qwen2.5-14B-Instruct", "dataset": "huypn16/MetaMath-SAE", "split": "train", "ctx_len": 1024, "hf_token": null, "load_in_8bit": false, "max_examples": null, "resume": false, "seed": 42, "data_preprocessing_num_proc": 24} |