hugging-science/CodVa-2-SFT
2
1{2 "vocab_size": 50304,3 "d_model": 896,4 "n_layers": 18,5 "n_heads": 14,6 "n_kv_heads": 2,7 "max_len": 2048,8 "rope_theta": 500000.0,9 "window_size": 512,10 "pattern_mult": 2.5,11 "reason_mult": 0.75,12 "reason_depth": 2,13 "gate_hidden": 64,14 "gate_init": 0.0,15 "diff_lambda_init": 0.8,16 "tie_embeddings": true,17 "muon_lr": 0.002,18 "muon_min_lr": 0.0002,19 "muon_mom": 0.95,20 "adamw_lr": 5e-05,21 "adamw_min_lr": 5e-06,22 "beta1": 0.9,23 "beta2": 0.95,24 "wd": 0.05,25 "grad_clip": 1.0,26 "warmup": 100,27 "total_steps": 8000,28 "micro_batch": 8,29 "grad_accum": 8,30 "log_every": 10,31 "eval_every": 200,32 "ckpt_every": 1000,33 "step": 8000,34 "epoch": 0,35 "loss": 1.4803357672691346,36 "arch": "CodVa-2-SFT"37}