{ "arm": "latent", "data": "data/world-a,data/pool/w3-s0,data/pool/w3-s1,data/pool/w3-s2,data/pool/w4-s0,data/pool/w4-s1,data/pool/w4-s2,data/pool/w5-s0,data/pool/w5-s1,data/pool/w5-s2,data/pool/w6-s0,data/pool/w6-s1,data/pool/w6-s2,data/pool/w7-s0,data/pool/w7-s1,data/pool/w7-s2", "out": "runs/latent-multi", "edits": 0, "mix": [ [ "bios", 0.5 ], [ "qa", 0.5 ] ], "init": null, "init_checkpoint": "final", "size": "small", "dtype": "float32", "steps": 10000, "batch_size": 64, "seq_len": 256, "lr": 0.001, "min_lr_frac": 0.1, "warmup": 500, "weight_decay": 0.1, "grad_clip": 1.0, "log_every": 100, "eval_every": 2000, "eval_sets": "test_id,test_ood,test_1hop_ood", "eval_n": 200, "save_every": 2000, "seed": 0, "n_reads": 3, "top_k": 4, "hop_weight": 0.5, "max_supervision": 1024, "supervise": "all", "model_type": "latent", "model": { "vocab_size": 576, "n_layer": 8, "n_head": 8, "d_model": 512, "mlp_ratio": 4, "rope_base": 10000.0, "n_reads": 3, "top_k": 4, "key_dim": 128, "retrieval_chunk": 2048, "retrieval_group": 256 }, "params": 29157915, "vocab": 544 }