File size: 5,564 Bytes
3d1010f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
#!/usr/bin/env python3
"""
run_autonomous.py

One-shot autonomous paper generator for CAJAL-9B v2.
No human intervention required. Produces a P2PCLAW-ready paper.

Usage:
    python run_autonomous.py

Requirements:
    - Ollama running with cajal-9b-v2:latest loaded
    - Python 3.10+
    - requests package

Output:
    - Saves paper to papers/ directory
    - Prints paper stats
    - Optionally publishes to P2PCLAW if --publish flag is used
"""

import sys
import time
import subprocess
from pathlib import Path

# Import from v8 optimizer
sys.path.insert(0, str(Path(__file__).parent))
from q8_0_optimizer_v8 import (
    run_simulation, generate_paper, auto_structural_fixes,
    expand_paper_to_minimum, inject_code_and_bridge,
    extract_title, build_paper_prompt, SYSTEM_PROMPT,
    MODEL, PAPERS_DIR, complete_tribunal, publish_paper,
    poll_for_scores
)

DEFAULT_TOPIC = "Adaptive Timeout Calibration for Byzantine Fault-Tolerant Consensus"


def ensure_ollama_running():
    """Check if Ollama is accessible, try to start if not."""
    import requests
    try:
        r = requests.get("http://localhost:11434/api/tags", timeout=5)
        if r.status_code == 200:
            print("[OK] Ollama is running")
            return True
    except Exception:
        pass
    
    print("[START] Attempting to start Ollama...")
    try:
        subprocess.Popen([r"E:\Ollama\ollama.exe", "serve"], 
                        creationflags=subprocess.DETACHED_PROCESS)
        time.sleep(10)
        r = requests.get("http://localhost:11434/api/tags", timeout=5)
        if r.status_code == 200:
            print("[OK] Ollama started successfully")
            return True
    except Exception as e:
        print(f"[FAIL] Could not start Ollama: {e}")
    return False


def generate_autonomous_paper(topic: str = DEFAULT_TOPIC, publish: bool = False):
    print("=" * 70)
    print("  CAJAL-9B AUTONOMOUS PAPER GENERATOR v8")
    print("  100% Automated — No Human Intervention")
    print("=" * 70)
    
    if not ensure_ollama_running():
        print("[ERROR] Ollama is not available. Please start it manually.")
        return None
    
    print(f"[TOPIC] {topic}")
    
    # Run simulation
    sim_results = run_simulation()
    print(f"[SIM] Results: {sim_results}")
    
    # Generate paper
    prompt = build_paper_prompt(topic, sim_results, iteration=1)
    gen_opts = {
        "num_predict": 24000,
        "temperature": 0.4,
        "top_p": 0.90,
        "top_k": 50,
        "repeat_penalty": 1.18,
    }
    
    print("[GEN] Generating paper with CAJAL-9B Q8_0...")
    raw_paper = generate_paper(MODEL, prompt, SYSTEM_PROMPT, gen_opts)
    
    if not raw_paper or len(raw_paper) < 500:
        print("[FAIL] Paper generation failed or too short")
        return None
    
    # Apply all automated fixes
    print("[FIX] Applying structural corrections...")
    paper_text = inject_code_and_bridge(raw_paper, sim_results)
    
    word_count = len(paper_text.split())
    if word_count < 2600:
        print(f"[EXPAND] Paper too short ({word_count} words), expanding...")
        paper_text = expand_paper_to_minimum(paper_text, topic, target_words=2600)
        paper_text = auto_structural_fixes(paper_text)
    
    paper_text = auto_structural_fixes(paper_text)
    word_count = len(paper_text.split())
    
    title = extract_title(paper_text, topic)
    print(f"[DONE] Title: {title}")
    print(f"[DONE] Words: {word_count}")
    print(f"[DONE] Sections: Abstract, Introduction, Methodology, Results, Discussion, Conclusion, References")
    
    # Save
    from datetime import datetime
    ts = datetime.now().strftime("%Y%m%d_%H%M%S")
    filename = f"CAJAL_autonomous_{ts}.md"
    filepath = PAPERS_DIR / filename
    filepath.write_text(paper_text, encoding="utf-8")
    print(f"[SAVE] {filepath}")
    
    if publish:
        agent_id = f"cajal-9b-v2-autonomous-{ts}"
        print(f"[TRIBUNAL] Starting examination...")
        clearance = complete_tribunal(agent_id, topic)
        if clearance:
            print(f"[PUB] Publishing to P2PCLAW...")
            pub_result = publish_paper(title, paper_text, agent_id, clearance)
            paper_id = pub_result.get("paperId") or pub_result.get("id")
            if paper_id:
                print(f"[PUB] Published: {paper_id}")
                print("[WAIT] Waiting for scores (this may take 2-5 minutes)...")
                scores = poll_for_scores(paper_id, agent_id)
                if scores:
                    overall = scores.get("overall")
                    print(f"[SCORE] Overall: {overall}/10")
                    print(f"[SCORE] Reproducibility: {scores.get('reproducibility')}")
                    print(f"[SCORE] Citations: {scores.get('citation_quality')}")
                else:
                    print("[SCORE] Scores not yet available")
            else:
                print(f"[PUB] Failed: {pub_result}")
        else:
            print("[TRIBUNAL] Failed to pass examination")
    
    print("=" * 70)
    print("  AUTONOMOUS GENERATION COMPLETE")
    print("=" * 70)
    return filepath


if __name__ == "__main__":
    import argparse
    parser = argparse.ArgumentParser(description="Generate a P2PCLAW paper autonomously with CAJAL-9B")
    parser.add_argument("--topic", default=DEFAULT_TOPIC, help="Paper topic")
    parser.add_argument("--publish", action="store_true", help="Publish to P2PCLAW after generation")
    args = parser.parse_args()
    
    generate_autonomous_paper(topic=args.topic, publish=args.publish)