File size: 3,295 Bytes
0b504c6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
caa90b0
 
 
 
 
0b504c6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
#!/usr/bin/env python3
"""LLM agents on procedural text mazes: one-shot answer, then LLM-Modulo repair with exact validator feedback.

  GROQ_API_KEY=... CEREBRAS_API_KEY=... GEMINI_API_KEY=... python scripts/maze_llm_agents.py [agent ...]
Resumable: results/maze_llm/<agent>.jsonl
"""
import json
import re
import sys
import threading
import time
from pathlib import Path

sys.path.insert(0, str(Path(__file__).resolve().parents[1] / "src"))
from robotics.benchmarks import maze_text as mt  # noqa: E402
from robotics.llm.providers import LLMError, chat  # noqa: E402

AGENTS = {
    "gpt-oss-120b@groq": ("groq", "openai/gpt-oss-120b", 3.0),
    "qwen3.8-27b@groq": ("groq", "qwen/qwen3.8-27b", 2.0),
    "qwen3.8-27b@cerebras": ("cerebras", "qwen-3.8-27b", 2.0),
    "gemini-3.1-flash-lite": ("gemini", "gemini-3.1-flash-lite", 4.5),
}
REPAIRS = 3


def run_agent(name, insts, outdir):
    prov, model, pace = AGENTS[name]
    path = Path(outdir) / (re.sub(r"[^A-Za-z0-9.@-]+", "_", name) + ".jsonl")
    done = set()
    if path.exists():
        done = {json.loads(line)["id"] for line in open(path) if not json.loads(line).get("error_api")}
    for inst in insts:
        if inst["id"] in done:
            continue
        msgs = [{"role": "user", "content": mt.prompt(inst)}]
        attempts, err_api, t = [], None, time.perf_counter()
        for _ in range(REPAIRS + 1):
            try:
                text, _ = chat(msgs, providers=[prov], model=model, max_tokens=8000, retries_per_provider=6,
                               backoff=8.0, timeout=240)
            except LLMError as e:
                if "empty content" in str(e):       # the model spent its whole output budget without answering:
                    text = ""                       # that is a failed attempt, not an infrastructure error
                else:
                    err_api = str(e)[:150]
                    break
            p = mt.parse(text)
            v = mt.validate(inst, p)
            attempts.append({"path": p, "solved": v["solved"], "error": v["error"], "n_moves": v["n_moves"]})
            if v["solved"]:
                break
            msgs += [{"role": "assistant", "content": text[-1500:]},
                     {"role": "user", "content": f"That path is wrong: {v['error']}. Give a corrected complete path "
                                                 "from S, ending with a line 'PATH: <moves>'."}]
            time.sleep(pace)
        rec = {"id": inst["id"], "agent": name, "cells": inst["cells"], "optimal": inst["optimal"], "attempts": attempts,
               "oneshot": bool(attempts and attempts[0]["solved"]), "modulo": bool(attempts and attempts[-1]["solved"]),
               "calls": len(attempts), "latency_s": round(time.perf_counter() - t, 1), "error_api": err_api}
        with path.open("a") as f:
            f.write(json.dumps(rec) + "\n")
        time.sleep(pace)
    print("done", name, flush=True)


def main():
    names = sys.argv[1:] or list(AGENTS)
    insts = mt.suite()
    Path("results/maze_llm").mkdir(parents=True, exist_ok=True)
    ths = [threading.Thread(target=run_agent, args=(n, insts, "results/maze_llm")) for n in names]
    for t in ths:
        t.start()
    for t in ths:
        t.join()


if __name__ == "__main__":
    main()