synth-ai 0.2.14__py3-none-any.whl → 0.4.4__py3-none-any.whl

synth_ai/tracing_v3/utils.py

@@ -1,108 +0,0 @@
-"""Utility functions for tracing v3."""
-
-from __future__ import annotations
-
-import hashlib
-import json
-import uuid
-from datetime import datetime, timezone
-from typing import Any
-
-
-def iso_now() -> str:
-    """Get current timezone.utc time as ISO format string."""
-    return datetime.now(timezone.utc).isoformat()
-
-
-def json_dumps(obj: Any) -> str:
-    """JSON dump with consistent formatting and datetime handling."""
-    return json.dumps(obj, default=str, separators=(",", ":"))
-
-
-def json_loads(s: str) -> Any:
-    """Safe JSON load with error handling."""
-    try:
-        return json.loads(s)
-    except (json.JSONDecodeError, TypeError):
-        return {}
-
-
-def generate_session_id(prefix: str = "session") -> str:
-    """Generate a unique session ID."""
-    return f"{prefix}_{uuid.uuid4().hex[:12]}"
-
-
-def generate_experiment_id(name: str) -> str:
-    """Generate experiment ID from name."""
-    hash_obj = hashlib.sha256(name.encode())
-    return f"exp_{hash_obj.hexdigest()[:12]}"
-
-
-def detect_provider(model_name: str | None) -> str:
-    """Detect LLM provider from model name."""
-    if not model_name:
-        return "unknown"
-
-    model_lower = model_name.lower()
-
-    if any(
-        x in model_lower for x in ["gpt-", "text-davinci", "text-curie", "text-babbage", "text-ada"]
-    ):
-        return "openai"
-    elif any(x in model_lower for x in ["claude", "anthropic"]):
-        return "anthropic"
-    elif any(x in model_lower for x in ["palm", "gemini", "bard"]):
-        return "google"
-    elif "azure" in model_lower:
-        return "azure"
-    elif any(x in model_lower for x in ["llama", "mistral", "mixtral", "local"]):
-        return "local"
-    else:
-        return "unknown"
-
-
-def calculate_cost(model_name: str, input_tokens: int, output_tokens: int) -> float | None:
-    """Calculate cost in USD based on model and token counts."""
-    # This is a simplified version - in production you'd want a proper pricing table
-    pricing = {
-        "gpt-4": {"input": 0.03, "output": 0.06},  # per 1K tokens
-        "gpt-4-turbo": {"input": 0.01, "output": 0.03},
-        "gpt-3.5-turbo": {"input": 0.0005, "output": 0.0015},
-        "claude-3-opus": {"input": 0.015, "output": 0.075},
-        "claude-3-sonnet": {"input": 0.003, "output": 0.015},
-        "claude-3-haiku": {"input": 0.00025, "output": 0.00125},
-    }
-
-    for model_prefix, prices in pricing.items():
-        if model_prefix in model_name.lower():
-            input_cost = (input_tokens / 1000) * prices["input"]
-            output_cost = (output_tokens / 1000) * prices["output"]
-            return input_cost + output_cost
-
-    return None
-
-
-def truncate_content(content: str, max_length: int = 10000) -> str:
-    """Truncate content to maximum length."""
-    if len(content) <= max_length:
-        return content
-    return content[: max_length - 3] + "..."
-
-
-def format_duration(start: datetime, end: datetime) -> str:
-    """Format duration between two timestamps."""
-    delta = end - start
-    total_seconds = delta.total_seconds()
-
-    if total_seconds < 1:
-        return f"{int(total_seconds * 1000)}ms"
-    elif total_seconds < 60:
-        return f"{total_seconds:.1f}s"
-    elif total_seconds < 3600:
-        minutes = int(total_seconds // 60)
-        seconds = int(total_seconds % 60)
-        return f"{minutes}m {seconds}s"
-    else:
-        hours = int(total_seconds // 3600)
-        minutes = int((total_seconds % 3600) // 60)
-        return f"{hours}h {minutes}m"

synth_ai/tui/__init__.py

@@ -1,5 +0,0 @@
-"""Text User Interface utilities for synth-ai."""
-
-from .dashboard import main
-
-__all__ = ["main"]

synth_ai/tui/__main__.py

@@ -1,13 +0,0 @@
-#!/usr/bin/env python3
-"""
-Entry point for Synth AI TUI dashboard.
-
-Usage:
-    python -m synth_ai.tui
-    python -m synth_ai.tui --url sqlite+aiosqlite:///path/to/db
-"""
-
-from .dashboard import main
-
-if __name__ == "__main__":
-    main()

synth_ai/tui/cli/__init__.py

@@ -1 +0,0 @@
-"""Command Line Interface tools for synth-ai."""

synth_ai/tui/cli/query_experiments.py

@@ -1,164 +0,0 @@
-#!/usr/bin/env python3
-"""
-Query experiments and sessions from Turso/sqld using v3 tracing.
-"""
-
-import argparse
-import asyncio
-
-from synth_ai.tracing_v3.turso.manager import AsyncSQLTraceManager
-
-
-async def list_experiments(db_url: str):
-    """List all experiments in the database."""
-    db = AsyncSQLTraceManager(db_url)
-    await db.initialize()
-
-    try:
-        df = await db.query_traces("""
-            SELECT 
-                e.experiment_id,
-                e.name,
-                e.description,
-                e.created_at,
-                COUNT(DISTINCT st.session_id) as num_sessions,
-                COUNT(DISTINCT ev.id) as num_events,
-                SUM(CASE WHEN ev.event_type = 'cais' THEN ev.cost_usd ELSE 0 END) / 100.0 as total_cost,
-                SUM(CASE WHEN ev.event_type = 'cais' THEN ev.total_tokens ELSE 0 END) as total_tokens
-            FROM experiments e
-            LEFT JOIN session_traces st ON e.experiment_id = st.experiment_id
-            LEFT JOIN events ev ON st.session_id = ev.session_id
-            GROUP BY e.experiment_id, e.name, e.description, e.created_at
-            ORDER BY e.created_at DESC
-        """)
-
-        if df.empty:
-            print("No experiments found in database.")
-            return
-
-        print(f"\n{'=' * 100}")
-        print(f"{'Experiments in ' + db_url:^100}")
-        print(f"{'=' * 100}\n")
-
-        for _, row in df.iterrows():
-            print(f"🧪 {row['name']} (id: {row['experiment_id'][:8]}...)")
-            print(f"   Created: {row['created_at']}")
-            print(f"   Description: {row['description']}")
-            print(f"   Sessions: {row['num_sessions']}")
-            print(f"   Events: {row['num_events']:,}")
-            if row["total_cost"] and row["total_cost"] > 0:
-                print(f"   Cost: ${row['total_cost']:.4f}")
-            if row["total_tokens"] and row["total_tokens"] > 0:
-                print(f"   Tokens: {int(row['total_tokens']):,}")
-            print()
-    finally:
-        await db.close()
-
-
-async def show_experiment_details(db_url: str, experiment_id: str):
-    """Show detailed information about a specific experiment."""
-    db = AsyncSQLTraceManager(db_url)
-    await db.initialize()
-
-    try:
-        # Get experiment info
-        exp_df = await db.query_traces(
-            """
-            SELECT * FROM experiments WHERE experiment_id LIKE :exp_id
-        """,
-            {"exp_id": f"{experiment_id}%"},
-        )
-
-        if exp_df.empty:
-            print(f"No experiment found matching ID: {experiment_id}")
-            return
-
-        exp = exp_df.iloc[0]
-        print(f"\n{'=' * 100}")
-        print(f"Experiment: {exp['name']} ({exp['experiment_id']})")
-        print(f"{'=' * 100}\n")
-
-        # Get session statistics
-        sessions_df = await db.get_sessions_by_experiment(exp["experiment_id"])
-
-        if sessions_df:
-            print(f"Sessions: {len(sessions_df)}")
-
-            # Get aggregated stats
-            stats_df = await db.query_traces(
-                """
-                SELECT 
-                    COUNT(DISTINCT ev.id) as total_events,
-                    COUNT(DISTINCT m.id) as total_messages,
-                    SUM(CASE WHEN ev.event_type = 'cais' THEN ev.cost_usd ELSE 0 END) / 100.0 as total_cost,
-                    SUM(CASE WHEN ev.event_type = 'cais' THEN ev.total_tokens ELSE 0 END) as total_tokens
-                FROM session_traces st
-                LEFT JOIN events ev ON st.session_id = ev.session_id
-                LEFT JOIN messages m ON st.session_id = m.session_id
-                WHERE st.experiment_id = :exp_id
-            """,
-                {"exp_id": exp["experiment_id"]},
-            )
-
-            if not stats_df.empty:
-                stats = stats_df.iloc[0]
-                print(f"Total events: {int(stats['total_events']):,}")
-                print(f"Total messages: {int(stats['total_messages']):,}")
-                print(f"Total cost: ${stats['total_cost']:.4f}")
-                print(f"Total tokens: {int(stats['total_tokens']):,}")
-
-            # Show session list
-            print("\nSession list:")
-            for sess in sessions_df:
-                print(f"  - {sess['session_id']} ({sess['created_at']})")
-                print(
-                    f"    Timesteps: {sess['num_timesteps']}, Events: {sess['num_events']}, Messages: {sess['num_messages']}"
-                )
-    finally:
-        await db.close()
-
-
-async def show_model_usage(db_url: str, model_name: str | None = None):
-    """Show model usage statistics."""
-    db = AsyncSQLTraceManager(db_url)
-    await db.initialize()
-
-    try:
-        df = await db.get_model_usage(model_name=model_name)
-
-        if df.empty:
-            print("No model usage data found.")
-            return
-
-        print(f"\n{'=' * 100}")
-        print(f"{'Model Usage Statistics':^100}")
-        print(f"{'=' * 100}\n")
-
-        print(df.to_string(index=False))
-    finally:
-        await db.close()
-
-
-async def main():
-    parser = argparse.ArgumentParser(description="Query experiments from Turso/sqld (v3)")
-    parser.add_argument(
-        "-u", "--url", default="sqlite+libsql://http://127.0.0.1:8080", help="Turso database URL"
-    )
-    parser.add_argument(
-        "-e", "--experiment", help="Show details for specific experiment ID (can be partial)"
-    )
-    parser.add_argument("-m", "--model", help="Show usage for specific model")
-    parser.add_argument("--usage", action="store_true", help="Show model usage statistics")
-
-    args = parser.parse_args()
-
-    if args.usage or args.model:
-        await show_model_usage(args.url, args.model)
-    elif args.experiment:
-        await show_experiment_details(args.url, args.experiment)
-    else:
-        await list_experiments(args.url)
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

synth_ai/tui/cli/query_experiments_v3.py

@@ -1,164 +0,0 @@
-#!/usr/bin/env python3
-"""
-Query experiments and sessions from Turso/sqld using v3 tracing.
-"""
-
-import argparse
-import asyncio
-
-from synth_ai.tracing_v3.turso.manager import AsyncSQLTraceManager
-
-
-async def list_experiments(db_url: str):
-    """List all experiments in the database."""
-    db = AsyncSQLTraceManager(db_url)
-    await db.initialize()
-
-    try:
-        df = await db.query_traces("""
-            SELECT 
-                e.experiment_id,
-                e.name,
-                e.description,
-                e.created_at,
-                COUNT(DISTINCT st.session_id) as num_sessions,
-                COUNT(DISTINCT ev.id) as num_events,
-                SUM(CASE WHEN ev.event_type = 'cais' THEN ev.cost_usd ELSE 0 END) / 100.0 as total_cost,
-                SUM(CASE WHEN ev.event_type = 'cais' THEN ev.total_tokens ELSE 0 END) as total_tokens
-            FROM experiments e
-            LEFT JOIN session_traces st ON e.experiment_id = st.experiment_id
-            LEFT JOIN events ev ON st.session_id = ev.session_id
-            GROUP BY e.experiment_id, e.name, e.description, e.created_at
-            ORDER BY e.created_at DESC
-        """)
-
-        if df.empty:
-            print("No experiments found in database.")
-            return
-
-        print(f"\n{'=' * 100}")
-        print(f"{'Experiments in ' + db_url:^100}")
-        print(f"{'=' * 100}\n")
-
-        for _, row in df.iterrows():
-            print(f"🧪 {row['name']} (id: {row['experiment_id'][:8]}...)")
-            print(f"   Created: {row['created_at']}")
-            print(f"   Description: {row['description']}")
-            print(f"   Sessions: {row['num_sessions']}")
-            print(f"   Events: {row['num_events']:,}")
-            if row["total_cost"] and row["total_cost"] > 0:
-                print(f"   Cost: ${row['total_cost']:.4f}")
-            if row["total_tokens"] and row["total_tokens"] > 0:
-                print(f"   Tokens: {int(row['total_tokens']):,}")
-            print()
-    finally:
-        await db.close()
-
-
-async def show_experiment_details(db_url: str, experiment_id: str):
-    """Show detailed information about a specific experiment."""
-    db = AsyncSQLTraceManager(db_url)
-    await db.initialize()
-
-    try:
-        # Get experiment info
-        exp_df = await db.query_traces(
-            """
-            SELECT * FROM experiments WHERE experiment_id LIKE :exp_id
-        """,
-            {"exp_id": f"{experiment_id}%"},
-        )
-
-        if exp_df.empty:
-            print(f"No experiment found matching ID: {experiment_id}")
-            return
-
-        exp = exp_df.iloc[0]
-        print(f"\n{'=' * 100}")
-        print(f"Experiment: {exp['name']} ({exp['experiment_id']})")
-        print(f"{'=' * 100}\n")
-
-        # Get session statistics
-        sessions_df = await db.get_sessions_by_experiment(exp["experiment_id"])
-
-        if sessions_df:
-            print(f"Sessions: {len(sessions_df)}")
-
-            # Get aggregated stats
-            stats_df = await db.query_traces(
-                """
-                SELECT 
-                    COUNT(DISTINCT ev.id) as total_events,
-                    COUNT(DISTINCT m.id) as total_messages,
-                    SUM(CASE WHEN ev.event_type = 'cais' THEN ev.cost_usd ELSE 0 END) / 100.0 as total_cost,
-                    SUM(CASE WHEN ev.event_type = 'cais' THEN ev.total_tokens ELSE 0 END) as total_tokens
-                FROM session_traces st
-                LEFT JOIN events ev ON st.session_id = ev.session_id
-                LEFT JOIN messages m ON st.session_id = m.session_id
-                WHERE st.experiment_id = :exp_id
-            """,
-                {"exp_id": exp["experiment_id"]},
-            )
-
-            if not stats_df.empty:
-                stats = stats_df.iloc[0]
-                print(f"Total events: {int(stats['total_events']):,}")
-                print(f"Total messages: {int(stats['total_messages']):,}")
-                print(f"Total cost: ${stats['total_cost']:.4f}")
-                print(f"Total tokens: {int(stats['total_tokens']):,}")
-
-            # Show session list
-            print("\nSession list:")
-            for sess in sessions_df:
-                print(f"  - {sess['session_id']} ({sess['created_at']})")
-                print(
-                    f"    Timesteps: {sess['num_timesteps']}, Events: {sess['num_events']}, Messages: {sess['num_messages']}"
-                )
-    finally:
-        await db.close()
-
-
-async def show_model_usage(db_url: str, model_name: str | None = None):
-    """Show model usage statistics."""
-    db = AsyncSQLTraceManager(db_url)
-    await db.initialize()
-
-    try:
-        df = await db.get_model_usage(model_name=model_name)
-
-        if df.empty:
-            print("No model usage data found.")
-            return
-
-        print(f"\n{'=' * 100}")
-        print(f"{'Model Usage Statistics':^100}")
-        print(f"{'=' * 100}\n")
-
-        print(df.to_string(index=False))
-    finally:
-        await db.close()
-
-
-async def main():
-    parser = argparse.ArgumentParser(description="Query experiments from Turso/sqld (v3)")
-    parser.add_argument(
-        "-u", "--url", default="sqlite+libsql://http://127.0.0.1:8080", help="Turso database URL"
-    )
-    parser.add_argument(
-        "-e", "--experiment", help="Show details for specific experiment ID (can be partial)"
-    )
-    parser.add_argument("-m", "--model", help="Show usage for specific model")
-    parser.add_argument("--usage", action="store_true", help="Show model usage statistics")
-
-    args = parser.parse_args()
-
-    if args.usage or args.model:
-        await show_model_usage(args.url, args.model)
-    elif args.experiment:
-        await show_experiment_details(args.url, args.experiment)
-    else:
-        await list_experiments(args.url)
-
-
-if __name__ == "__main__":
-    asyncio.run(main())