import sys, json from pathlib import Path sys.path.insert(0, "/root/ingot/src") from eval_service.run import load_tokenizer, _render tok = load_tokenizer(Path("/root/models/teacher")) rows = [json.loads(l) for l in open("/root/evalset.jsonl")] keep = [] for r in rows: try: p = _render(tok, r["messages"], r.get("tools")) except Exception: continue if len(p) <= 3000 and len(p) >= 8: keep.append(r) Path("/root/evalset.jsonl").write_text( "".join(json.dumps(r, ensure_ascii=False) + "\n" for r in keep)) print("kept", len(keep), "of", len(rows))