|
| 1 | +""" |
| 2 | +Adaptive experiment generator for Parameter Golf autoresearch. |
| 3 | +
|
| 4 | +Reads completed results, analyzes what worked/didn't, and generates |
| 5 | +the next batch of experiments. Called by loop.sh after each batch. |
| 6 | +
|
| 7 | +Strategy: |
| 8 | +1. Rank completed experiments by BPB |
| 9 | +2. Identify which technique changes improved vs hurt |
| 10 | +3. Generate new experiments that: |
| 11 | + a. Combine top-2 individual winners |
| 12 | + b. Push winning techniques further (e.g., if XSA-6 beat XSA-4, try XSA-8) |
| 13 | + c. Sweep around the best hyperparameters |
| 14 | + d. Try removing the worst-performing changes (simplify) |
| 15 | +4. Always include 1 "wild card" experiment for exploration |
| 16 | +
|
| 17 | +All generated experiments must be: |
| 18 | +- Legal TTT (TTT_PASSES ≤ 1) |
| 19 | +- Within 16MB artifact budget (estimated) |
| 20 | +- Not duplicates of already-run experiments |
| 21 | +""" |
| 22 | + |
| 23 | +import json |
| 24 | +import sys |
| 25 | +import os |
| 26 | +from pathlib import Path |
| 27 | + |
| 28 | +sys.path.insert(0, str(Path(__file__).resolve().parent.parent)) |
| 29 | +from experiments import ( |
| 30 | + Experiment, ExperimentResult, estimate_artifact_bytes, |
| 31 | + MAX_ARTIFACT_BYTES, EXPERIMENTS, |
| 32 | +) |
| 33 | + |
| 34 | +STATE_FILE = Path(__file__).resolve().parent / "state.json" |
| 35 | +EXPERIMENTS_FILE = Path(__file__).resolve().parent.parent / "experiments.py" |
| 36 | + |
| 37 | +# Default env for legal experiments |
| 38 | +BASE_ENV = {"SEED": "1337", "TTT_PASSES": "1"} |
| 39 | + |
| 40 | + |
| 41 | +def load_state() -> dict: |
| 42 | + if STATE_FILE.exists(): |
| 43 | + return json.loads(STATE_FILE.read_text()) |
| 44 | + return {"results": [], "completed_experiments": []} |
| 45 | + |
| 46 | + |
| 47 | +def analyze_results(results: list[dict]) -> dict: |
| 48 | + """Analyze what worked and what didn't.""" |
| 49 | + if not results: |
| 50 | + return {"winners": [], "losers": [], "baseline_bpb": None} |
| 51 | + |
| 52 | + # Find baseline (no_ttt or legal_ttt_baseline) |
| 53 | + baseline = None |
| 54 | + for r in results: |
| 55 | + if r.get("experiment") in ("no_ttt_baseline", "legal_ttt_baseline"): |
| 56 | + if r.get("val_bpb"): |
| 57 | + baseline = r |
| 58 | + break |
| 59 | + |
| 60 | + if not baseline: |
| 61 | + # Use the first result with a BPB as reference |
| 62 | + for r in results: |
| 63 | + if r.get("val_bpb"): |
| 64 | + baseline = r |
| 65 | + break |
| 66 | + |
| 67 | + if not baseline: |
| 68 | + return {"winners": [], "losers": [], "baseline_bpb": None} |
| 69 | + |
| 70 | + base_bpb = baseline["val_bpb"] |
| 71 | + |
| 72 | + # Classify experiments |
| 73 | + winners = [] # Lower BPB = better |
| 74 | + losers = [] |
| 75 | + for r in results: |
| 76 | + if not r.get("val_bpb") or r["experiment"] == baseline["experiment"]: |
| 77 | + continue |
| 78 | + delta = r["val_bpb"] - base_bpb |
| 79 | + entry = {"experiment": r["experiment"], "val_bpb": r["val_bpb"], "delta": delta} |
| 80 | + if delta < -0.0005: # Improved by at least 0.0005 |
| 81 | + winners.append(entry) |
| 82 | + elif delta > 0.001: # Hurt by more than 0.001 |
| 83 | + losers.append(entry) |
| 84 | + |
| 85 | + winners.sort(key=lambda x: x["delta"]) |
| 86 | + losers.sort(key=lambda x: x["delta"], reverse=True) |
| 87 | + |
| 88 | + return { |
| 89 | + "winners": winners, |
| 90 | + "losers": losers, |
| 91 | + "baseline_bpb": base_bpb, |
| 92 | + "baseline_experiment": baseline["experiment"], |
| 93 | + } |
| 94 | + |
| 95 | + |
| 96 | +def extract_env_diff(experiment_name: str) -> dict: |
| 97 | + """Get the env overrides for a named experiment from EXPERIMENTS list.""" |
| 98 | + for exp in EXPERIMENTS: |
| 99 | + if exp.name == experiment_name: |
| 100 | + # Return only the non-default overrides |
| 101 | + diff = {} |
| 102 | + for k, v in exp.env.items(): |
| 103 | + if k == "SEED" or k == "TTT_PASSES": |
| 104 | + continue |
| 105 | + diff[k] = v |
| 106 | + return diff |
| 107 | + return {} |
| 108 | + |
| 109 | + |
| 110 | +def generate_next_batch(analysis: dict, completed: list[str], batch_size: int = 5) -> list[Experiment]: |
| 111 | + """Generate the next batch of experiments based on results analysis.""" |
| 112 | + new_experiments = [] |
| 113 | + used_names = set(completed) |
| 114 | + |
| 115 | + def _add(name, desc, env_extra=None, patches=None): |
| 116 | + if name in used_names or len(new_experiments) >= batch_size: |
| 117 | + return |
| 118 | + env = {**BASE_ENV} |
| 119 | + if env_extra: |
| 120 | + env.update(env_extra) |
| 121 | + exp = Experiment(name=name, description=desc, env=env, patches=patches or []) |
| 122 | + est = estimate_artifact_bytes(env) |
| 123 | + if est <= MAX_ARTIFACT_BYTES: |
| 124 | + new_experiments.append(exp) |
| 125 | + used_names.add(name) |
| 126 | + |
| 127 | + winners = analysis.get("winners", []) |
| 128 | + losers = analysis.get("losers", []) |
| 129 | + |
| 130 | + # Strategy 1: Combine top-2 winners |
| 131 | + if len(winners) >= 2: |
| 132 | + w1_env = extract_env_diff(winners[0]["experiment"]) |
| 133 | + w2_env = extract_env_diff(winners[1]["experiment"]) |
| 134 | + combined_env = {**w1_env, **w2_env} |
| 135 | + name = f"combo_{winners[0]['experiment']}_plus_{winners[1]['experiment']}"[:60] |
| 136 | + desc = f"Combine #{1} {winners[0]['experiment']} ({winners[0]['delta']:+.4f}) + #{2} {winners[1]['experiment']} ({winners[1]['delta']:+.4f})" |
| 137 | + _add(name, desc, combined_env) |
| 138 | + |
| 139 | + # Strategy 2: Combine top-3 winners |
| 140 | + if len(winners) >= 3: |
| 141 | + combined_env = {} |
| 142 | + for w in winners[:3]: |
| 143 | + combined_env.update(extract_env_diff(w["experiment"])) |
| 144 | + name = "combo_top3_winners" |
| 145 | + desc = f"Combine top 3: {', '.join(w['experiment'] for w in winners[:3])}" |
| 146 | + _add(name, desc, combined_env) |
| 147 | + |
| 148 | + # Strategy 3: Push winning hyperparameters further |
| 149 | + for w in winners[:3]: |
| 150 | + env_diff = extract_env_diff(w["experiment"]) |
| 151 | + for key, val in env_diff.items(): |
| 152 | + try: |
| 153 | + fval = float(val) |
| 154 | + # If this was an increase from baseline, try going further |
| 155 | + # If it was a decrease, try going even lower |
| 156 | + for mult, suffix in [(1.5, "more"), (0.5, "less")]: |
| 157 | + new_val = fval * mult |
| 158 | + name = f"{w['experiment']}_{suffix}" |
| 159 | + desc = f"Push {key}={new_val} ({suffix} than {val})" |
| 160 | + _add(name, desc, {key: str(new_val)}) |
| 161 | + except (ValueError, TypeError): |
| 162 | + pass |
| 163 | + |
| 164 | + # Strategy 4: Interpolate between winner and baseline |
| 165 | + for w in winners[:2]: |
| 166 | + env_diff = extract_env_diff(w["experiment"]) |
| 167 | + for key, val in env_diff.items(): |
| 168 | + try: |
| 169 | + fval = float(val) |
| 170 | + # Try halfway between baseline default and winning value |
| 171 | + # (We don't know the baseline default here, so skip this for now) |
| 172 | + pass |
| 173 | + except (ValueError, TypeError): |
| 174 | + pass |
| 175 | + |
| 176 | + # Strategy 5: Wild card — try something not yet tested |
| 177 | + wild_cards = [ |
| 178 | + ("seq_len_4096", "Longer sequence length (4096 vs 2048)", {"TRAIN_SEQ_LEN": "4096", "EVAL_SEQ_LEN": "4096"}), |
| 179 | + ("rope_base_50k", "Higher RoPE base (50000 vs 10000)", {"ROPE_BASE": "50000"}), |
| 180 | + ("softcap_50", "Higher logit softcap (50 vs 30)", {"LOGIT_SOFTCAP": "50.0"}), |
| 181 | + ("softcap_20", "Lower logit softcap (20 vs 30)", {"LOGIT_SOFTCAP": "20.0"}), |
| 182 | + ("qk_gain_2", "Higher QK gain init (2.0 vs 1.5)", {"QK_GAIN_INIT": "2.0"}), |
| 183 | + ("muon_momentum_095", "Lower Muon momentum (0.95 vs 0.99)", {"MUON_MOMENTUM": "0.95"}), |
| 184 | + ("embed_lr_08", "Higher embed LR (0.8 vs 0.6)", {"EMBED_LR": "0.8"}), |
| 185 | + ] |
| 186 | + for name, desc, env_extra in wild_cards: |
| 187 | + _add(name, desc, env_extra) |
| 188 | + if len(new_experiments) >= batch_size: |
| 189 | + break |
| 190 | + |
| 191 | + return new_experiments |
| 192 | + |
| 193 | + |
| 194 | +def main(): |
| 195 | + state = load_state() |
| 196 | + results = state.get("results", []) |
| 197 | + completed = state.get("completed_experiments", []) |
| 198 | + |
| 199 | + print(f"Completed experiments: {len(completed)}") |
| 200 | + print(f"Results with BPB: {sum(1 for r in results if r.get('val_bpb'))}") |
| 201 | + |
| 202 | + analysis = analyze_results(results) |
| 203 | + print(f"\nBaseline: {analysis.get('baseline_bpb', 'N/A')}") |
| 204 | + print(f"Winners ({len(analysis['winners'])}):") |
| 205 | + for w in analysis["winners"]: |
| 206 | + print(f" {w['delta']:+.4f} — {w['experiment']} ({w['val_bpb']:.4f})") |
| 207 | + print(f"Losers ({len(analysis['losers'])}):") |
| 208 | + for l in analysis["losers"]: |
| 209 | + print(f" {l['delta']:+.4f} — {l['experiment']} ({l['val_bpb']:.4f})") |
| 210 | + |
| 211 | + new_batch = generate_next_batch(analysis, completed) |
| 212 | + print(f"\nGenerated {len(new_batch)} new experiments:") |
| 213 | + for exp in new_batch: |
| 214 | + est = estimate_artifact_bytes(exp.env) |
| 215 | + print(f" {exp.name}: {exp.description} (~{est/1e6:.1f} MB)") |
| 216 | + |
| 217 | + if new_batch: |
| 218 | + # Append to experiments.py EXPERIMENTS list |
| 219 | + # Write them as a separate file that the provider can pick up |
| 220 | + out = Path(__file__).resolve().parent / "next_batch.json" |
| 221 | + batch_data = [] |
| 222 | + for exp in new_batch: |
| 223 | + batch_data.append({ |
| 224 | + "name": exp.name, |
| 225 | + "description": exp.description, |
| 226 | + "env": exp.env, |
| 227 | + "patches": exp.patches, |
| 228 | + }) |
| 229 | + out.write_text(json.dumps(batch_data, indent=2)) |
| 230 | + print(f"\nWrote {len(new_batch)} experiments to {out}") |
| 231 | + |
| 232 | + return new_batch |
| 233 | + |
| 234 | + |
| 235 | +if __name__ == "__main__": |
| 236 | + main() |
0 commit comments