| {"sae": {"expansion_factor": 8, "normalize_decoder": true, "num_latents": 0, "k": 192, "multi_topk": false}, "batch_size": 24, "grad_acc_steps": 1, "micro_acc_steps": 1, "lr": null, "lr_warmup_steps": 1000, "auxk_alpha": 0.0, "dead_feature_threshold": 10000000, "hookpoints": ["layers.6", "layers.8", "layers.10", "layers.12", "layers.14", "layers.16", "layers.18", "layers.20", "layers.22"], "layers": [6, 8, 10, 12, 14, 16, 18, 20, 22], "layer_stride": 1, "distribute_modules": false, "save_every": 1000, "log_to_wandb": true, "run_name": null, "wandb_log_frequency": 1, "model": "Qwen/Qwen2.5-Coder-7B-Instruct", "dataset": "nguyentruong-ins/codeforces_cpp_cleaned_scaled_class", "split": "train", "ctx_len": 756, "hf_token": null, "load_in_8bit": false, "max_examples": null, "resume": false, "seed": 42, "data_preprocessing_num_proc": 24} |