File size: 7,548 Bytes
d9d0a60
 
 
 
 
 
 
 
 
 
a6df84a
d9d0a60
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a6df84a
 
d9d0a60
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
"""
Promptfoo Evaluation Runner

Python wrapper for running Promptfoo evaluations and processing results.
"""

import subprocess
import json
import logging
from pathlib import Path
from typing import Dict, Any, Optional
from datetime import datetime

logger = logging.getLogger(__name__)


class PromptfooRunner:
    """
    Wrapper for running Promptfoo evaluations.
    
    Provides Python interface for:
    - Running evaluations
    - Parsing results
    - Generating reports
    """
    
    def __init__(self, config_path: str = "eval/promptfooconfig.yaml"):
        """
        Initialize the runner.
        
        Args:
            config_path: Path to promptfoo config file
        """
        self.config_path = Path(config_path)
        self.results_dir = Path("eval/results")
        self.results_dir.mkdir(parents=True, exist_ok=True)
    
    def run_eval(
        self,
        output_path: Optional[str] = None,
        no_cache: bool = False,
        verbose: bool = True,
    ) -> Dict[str, Any]:
        """
        Run Promptfoo evaluation.
        
        Args:
            output_path: Custom output path for results
            no_cache: Disable caching
            verbose: Show detailed output
            
        Returns:
            Evaluation results dictionary
        """
        if not output_path:
            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
            output_path = self.results_dir / f"eval_{timestamp}.json"
        
        cmd = [
            "npx", "promptfoo", "eval",
            "-c", str(self.config_path),
            "-o", str(output_path),
            "--json",
        ]
        
        if no_cache:
            cmd.append("--no-cache")
        
        logger.info(f"Running Promptfoo evaluation: {' '.join(cmd)}")
        
        try:
            result = subprocess.run(
                cmd,
                capture_output=True,
                text=True,
                timeout=300,  # 5 minute timeout
            )
            
            if result.returncode != 0:
                logger.error(f"Promptfoo failed: {result.stderr}")
                return {"error": result.stderr}
            
            # Parse results
            with open(output_path, "r") as f:
                results = json.load(f)
            
            return results
            
        except subprocess.TimeoutExpired:
            logger.error("Promptfoo evaluation timed out")
            return {"error": "Evaluation timed out"}
        except Exception as e:
            logger.error(f"Failed to run Promptfoo: {e}")
            return {"error": str(e)}
    
    def parse_results(self, results: Dict[str, Any]) -> Dict[str, Any]:
        """
        Parse and summarize evaluation results.
        
        Args:
            results: Raw Promptfoo results
            
        Returns:
            Summary dictionary
        """
        if "error" in results:
            return results
        
        summary = {
            "total_tests": 0,
            "passed": 0,
            "failed": 0,
            "pass_rate": 0.0,
            "avg_latency_ms": 0,
            "total_cost": 0.0,
            "failures": [],
        }
        
        tests = results.get("results", [])
        summary["total_tests"] = len(tests)
        
        latencies = []
        costs = []
        
        for test in tests:
            if test.get("success"):
                summary["passed"] += 1
            else:
                summary["failed"] += 1
                summary["failures"].append({
                    "description": test.get("description", "Unknown"),
                    "error": test.get("error", "Assertion failed"),
                })
            
            if "latencyMs" in test:
                latencies.append(test["latencyMs"])
            if "cost" in test:
                costs.append(test["cost"])
        
        if summary["total_tests"] > 0:
            summary["pass_rate"] = summary["passed"] / summary["total_tests"]
        
        if latencies:
            summary["avg_latency_ms"] = sum(latencies) / len(latencies)
        
        if costs:
            summary["total_cost"] = sum(costs)
        
        return summary
    
    def generate_report(
        self,
        results: Dict[str, Any],
        output_format: str = "markdown",
    ) -> str:
        """
        Generate human-readable report.
        
        Args:
            results: Evaluation results
            output_format: Output format (markdown, text)
            
        Returns:
            Formatted report string
        """
        summary = self.parse_results(results)
        
        if output_format == "markdown":
            return self._generate_markdown_report(summary)
        else:
            return self._generate_text_report(summary)
    
    def _generate_markdown_report(self, summary: Dict[str, Any]) -> str:
        """Generate Markdown report."""
        report = [
            "# RAG Evaluation Report",
            "",
            f"**Date:** {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}",
            "",
            "## Summary",
            "",
            "| Metric | Value |",
            "|--------|-------|",
            f"| Total Tests | {summary['total_tests']} |",
            f"| Passed | {summary['passed']} |",
            f"| Failed | {summary['failed']} |",
            f"| Pass Rate | {summary['pass_rate']:.1%} |",
            f"| Avg Latency | {summary['avg_latency_ms']:.0f}ms |",
            f"| Total Cost | ${summary['total_cost']:.4f} |",
            "",
        ]
        
        if summary["failures"]:
            report.extend([
                "## Failures",
                "",
            ])
            for failure in summary["failures"]:
                report.append(f"- **{failure['description']}**: {failure['error']}")
            report.append("")
        
        return "\n".join(report)
    
    def _generate_text_report(self, summary: Dict[str, Any]) -> str:
        """Generate plain text report."""
        lines = [
            "=== RAG Evaluation Report ===",
            "",
            f"Total Tests: {summary['total_tests']}",
            f"Passed: {summary['passed']}",
            f"Failed: {summary['failed']}",
            f"Pass Rate: {summary['pass_rate']:.1%}",
            f"Avg Latency: {summary['avg_latency_ms']:.0f}ms",
            f"Total Cost: ${summary['total_cost']:.4f}",
        ]
        
        if summary["failures"]:
            lines.extend(["", "Failures:"])
            for failure in summary["failures"]:
                lines.append(f"  - {failure['description']}: {failure['error']}")
        
        return "\n".join(lines)


def run_evaluation(
    config_path: str = "eval/promptfooconfig.yaml",
    verbose: bool = True,
) -> Dict[str, Any]:
    """
    Convenience function to run evaluation.
    
    Args:
        config_path: Path to config file
        verbose: Show detailed output
        
    Returns:
        Evaluation summary
    """
    runner = PromptfooRunner(config_path)
    results = runner.run_eval(verbose=verbose)
    return runner.parse_results(results)


if __name__ == "__main__":
    # Run evaluation when executed directly
    import sys
    
    logging.basicConfig(level=logging.INFO)
    
    runner = PromptfooRunner()
    results = runner.run_eval()
    
    if "error" in results:
        print(f"Evaluation failed: {results['error']}")
        sys.exit(1)
    
    report = runner.generate_report(results)
    print(report)