data/exp_prune/EleutherAI--pythia-1b_prune_v2.json

{
  "model": "EleutherAI/pythia-1b",
  "n_layers": 16,
  "n_heads": 8,
  "d_head": 256,
  "n_pairs": 128,
  "k_dead": 91,
  "theta": 10000.0,
  "T_train": 2048,
  "L_crit_pred": 15,
  "alpha": 0.9374717358310781,
  "n_eval_chunks": 100,
  "chunk_size": 1024,
  "n_lambada": 200,
  "baseline": {
    "ppl": 14.973609132474365,
    "nll": 2.7062892603874205,
    "std": 0.2595318035858822,
    "lambada_acc": 0.57
  },
  "sweeps": [
    {
      "L_prune": 11,
      "n_pruned_entries": 12124160,
      "frac_qk_pruned": 0.09033203125,
      "ppl": 16.53646806228249,
      "nll": 2.805568127632141,
      "std": 0.2573620626824769,
      "delta_ppl": 1.562858929808126,
      "delta_nll": 0.09927886724472046,
      "lambada_acc": 0.405,
      "delta_lambada": -0.16499999999999992,
      "verdict": "DEGRADED"
    },
    {
      "L_prune": 13,
      "n_pruned_entries": 7274496,
      "frac_qk_pruned": 0.05419921875,
      "ppl": 15.548956277713987,
      "nll": 2.743993515968323,
      "std": 0.25814710014606357,
      "delta_ppl": 0.5753471452396219,
      "delta_nll": 0.0377042555809024,
      "lambada_acc": 0.55,
      "delta_lambada": -0.019999999999999907,
      "verdict": "DEGRADED"
    },
    {
      "L_prune": 15,
      "n_pruned_entries": 2424832,
      "frac_qk_pruned": 0.01806640625,
      "ppl": 15.16368330360712,
      "nll": 2.7189033126831053,
      "std": 0.2584823516819064,
      "delta_ppl": 0.19007417113275515,
      "delta_nll": 0.012614052295684797,
      "lambada_acc": 0.54,
      "delta_lambada": -0.029999999999999916,
      "verdict": "OK"
    },
    {
      "L_prune": 16,
      "n_pruned_entries": 0,
      "frac_qk_pruned": 0.0,
      "ppl": 14.973609132474365,
      "nll": 2.7062892603874205,
      "std": 0.2595318035858822,
      "delta_ppl": 0.0,
      "delta_nll": 0.0,
      "lambada_acc": 0.57,
      "delta_lambada": 0.0,
      "verdict": "OK"
    }
  ]
}