| |
| """Compare opt-in layer dumps. All-row maxima plus last-token cosine.""" |
| import argparse |
| import json |
| from pathlib import Path |
| import numpy as np |
|
|
| parser = argparse.ArgumentParser(description=__doc__) |
| parser.add_argument("baseline", type=Path) |
| parser.add_argument("candidate", type=Path) |
| args = parser.parse_args() |
| records = [] |
| for position in (0, 128, 129): |
| for layer in range(24): |
| row = {"position": position, "layer": layer} |
| for stage in ("attention", "post_attention", "ffn", "output", "routes"): |
| name = f"{position}_layer{layer}_{stage}.f32" |
| a_path, b_path = args.baseline / name, args.candidate / name |
| if not a_path.exists() or not b_path.exists(): |
| continue |
| a = np.fromfile(a_path, dtype=np.float32) |
| b = np.fromfile(b_path, dtype=np.float32) |
| if a.shape != b.shape: |
| raise ValueError(name) |
| if stage == "routes": |
| a, b = a.reshape(-1, 16)[:, :8], b.reshape(-1, 16)[:, :8] |
| row[stage] = {"changed_sets": int(np.any(np.sort(a) != np.sort(b), axis=1).sum()), |
| "total": len(a), "last_old": a[-1].tolist(), "last_new": b[-1].tolist()} |
| continue |
| last_a, last_b = a[-1536:].astype(np.float64), b[-1536:].astype(np.float64) |
| cosine = np.dot(last_a, last_b) / (np.linalg.norm(last_a) * np.linalg.norm(last_b)) |
| row[stage] = {"last_cosine": float(cosine), "last_max": float(np.abs(last_a-last_b).max()), |
| "all_max": float(np.abs(a-b).max()), "different": int((a != b).sum())} |
| if len(row) > 2: |
| records.append(row) |
| print(json.dumps(records, indent=2)) |
|
|
