Spaces:
Running
Running
File size: 10,772 Bytes
99bd0b8 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 | """Simulation wrapper functions for the demo."""
import sys
from pathlib import Path
from typing import Any, Dict, List, Optional
# Ensure project root is on path (simulation.py is at utils/ under Space root)
PROJECT_ROOT = Path(__file__).resolve().parent.parent
if str(PROJECT_ROOT) not in sys.path:
sys.path.insert(0, str(PROJECT_ROOT))
from swarm.analysis.aggregation import MetricsAggregator # noqa: E402
from swarm.core.orchestrator import Orchestrator, OrchestratorConfig # noqa: E402
from swarm.governance.config import GovernanceConfig # noqa: E402
from swarm.scenarios.loader import ( # noqa: E402
build_orchestrator,
load_scenario,
)
SCENARIOS_DIR = PROJECT_ROOT / "scenarios"
# Safety limits to prevent excessive resource consumption
MAX_AGENTS_PER_TYPE = 10
MAX_TOTAL_AGENTS = 40
MAX_EPOCHS = 50
MAX_STEPS_PER_EPOCH = 30
def _requires_llm(data: dict) -> bool:
"""Return True if scenario YAML uses LLM-backed agents."""
for agent_spec in data.get("agents", []):
if agent_spec.get("type") == "llm":
return True
return False
def list_scenarios() -> List[Dict[str, str]]:
"""List available scenarios, excluding those that need LLM API keys."""
import yaml
scenarios = []
for yaml_file in sorted(SCENARIOS_DIR.glob("*.yaml")):
with open(yaml_file) as f:
data = yaml.safe_load(f)
if _requires_llm(data):
continue
scenarios.append(
{
"id": data.get("scenario_id", yaml_file.stem),
"description": data.get("description", ""),
"path": str(yaml_file),
"filename": yaml_file.name,
}
)
return scenarios
def run_scenario(scenario_path: str, seed: Optional[int] = None) -> Dict[str, Any]:
"""Run a scenario and return structured results.
Args:
scenario_path: Path to YAML scenario file (must be under scenarios/)
seed: Optional seed override
Returns:
Dict with epoch_metrics, agent_states, config info
Raises:
ValueError: If path is outside the scenarios directory
"""
# Path traversal protection: resolve and verify within scenarios dir
resolved = Path(scenario_path).resolve()
scenarios_resolved = SCENARIOS_DIR.resolve()
if not str(resolved).startswith(str(scenarios_resolved)):
raise ValueError(
f"Scenario path must be within {SCENARIOS_DIR}, got {scenario_path}"
)
# Reject scenarios that require LLM API keys
import yaml
with open(resolved) as f:
raw = yaml.safe_load(f)
if _requires_llm(raw):
raise ValueError("LLM-backed scenarios are not supported in the demo")
scenario = load_scenario(resolved)
if seed is not None:
scenario.orchestrator_config.seed = seed
# Disable file logging in demo mode to prevent disk writes
scenario.orchestrator_config.log_path = None
scenario.orchestrator_config.log_events = False
orchestrator = build_orchestrator(scenario)
# Attach aggregator for rich metrics
aggregator = MetricsAggregator()
aggregator.start_simulation(
simulation_id=scenario.scenario_id,
n_epochs=scenario.orchestrator_config.n_epochs,
steps_per_epoch=scenario.orchestrator_config.steps_per_epoch,
n_agents=len(orchestrator._agents),
seed=scenario.orchestrator_config.seed,
)
# Wire up interaction recording
def on_interaction(interaction, payoff_a, payoff_b):
aggregator.record_interaction(interaction)
aggregator.record_payoff(interaction.initiator, payoff_a)
aggregator.record_payoff(interaction.counterparty, payoff_b)
orchestrator._on_interaction_complete.append(on_interaction)
# Wire up epoch finalization
def on_epoch(epoch_metrics):
agent_states = {
aid: orchestrator.state.get_agent(aid) for aid in orchestrator._agents
}
aggregator.finalize_epoch(
epoch=orchestrator.state.current_epoch - 1,
agent_states=agent_states,
)
orchestrator._on_epoch_end.append(on_epoch)
# Run
epoch_metrics_list = orchestrator.run()
history = aggregator.end_simulation()
incoherence_series: List[float] = []
if history and getattr(history, "epoch_snapshots", None):
incoherence_series = [
float(getattr(snapshot, "incoherence_index", 0.0))
for snapshot in history.epoch_snapshots
]
# Extract agent final states
agent_states = []
for agent_id, _agent in orchestrator._agents.items():
state = orchestrator.state.get_agent(agent_id)
if state:
agent_states.append(
{
"agent_id": agent_id,
"agent_type": state.agent_type.value,
"reputation": round(state.reputation, 2),
"resources": round(state.resources, 2),
"interactions": state.interactions_initiated
+ state.interactions_received,
"total_payoff": round(state.total_payoff, 2),
}
)
return {
"scenario_id": scenario.scenario_id,
"description": scenario.description,
"epoch_metrics": epoch_metrics_list,
"agent_states": agent_states,
"history": history,
"incoherence_series": incoherence_series,
"n_epochs": scenario.orchestrator_config.n_epochs,
"n_agents": len(orchestrator._agents),
}
def run_custom(
n_honest: int = 3,
n_opportunistic: int = 1,
n_deceptive: int = 1,
n_adversarial: int = 0,
n_epochs: int = 20,
steps_per_epoch: int = 10,
tax_rate: float = 0.0,
reputation_decay: float = 1.0,
staking_enabled: bool = False,
min_stake: float = 0.0,
circuit_breaker_enabled: bool = False,
freeze_threshold: float = 0.7,
audit_enabled: bool = False,
audit_probability: float = 0.1,
seed: int = 42,
) -> Dict[str, Any]:
"""Run a custom simulation with specified parameters.
Returns:
Dict with epoch_metrics, agent_states, config info
Raises:
ValueError: If parameters exceed safety limits
"""
# Validate bounds to prevent resource exhaustion
total_agents = n_honest + n_opportunistic + n_deceptive + n_adversarial
if total_agents > MAX_TOTAL_AGENTS:
raise ValueError(
f"Total agents ({total_agents}) exceeds max ({MAX_TOTAL_AGENTS})"
)
if total_agents < 1:
raise ValueError("Must have at least 1 agent")
if n_epochs > MAX_EPOCHS:
raise ValueError(f"n_epochs ({n_epochs}) exceeds max ({MAX_EPOCHS})")
if steps_per_epoch > MAX_STEPS_PER_EPOCH:
raise ValueError(
f"steps_per_epoch ({steps_per_epoch}) exceeds max ({MAX_STEPS_PER_EPOCH})"
)
for name, val in [
("n_honest", n_honest),
("n_opportunistic", n_opportunistic),
("n_deceptive", n_deceptive),
("n_adversarial", n_adversarial),
]:
if val > MAX_AGENTS_PER_TYPE:
raise ValueError(f"{name} ({val}) exceeds max ({MAX_AGENTS_PER_TYPE})")
from swarm.agents.adversarial import AdversarialAgent
from swarm.agents.deceptive import DeceptiveAgent
from swarm.agents.honest import HonestAgent
from swarm.agents.opportunistic import OpportunisticAgent
governance_config = GovernanceConfig(
transaction_tax_rate=tax_rate,
reputation_decay_rate=reputation_decay,
staking_enabled=staking_enabled,
min_stake_to_participate=min_stake,
circuit_breaker_enabled=circuit_breaker_enabled,
freeze_threshold_toxicity=freeze_threshold,
audit_enabled=audit_enabled,
audit_probability=audit_probability,
)
config = OrchestratorConfig(
n_epochs=n_epochs,
steps_per_epoch=steps_per_epoch,
governance_config=governance_config,
seed=seed,
)
orchestrator = Orchestrator(config)
# Register agents
agent_specs = [
(HonestAgent, "honest", n_honest),
(OpportunisticAgent, "opportunistic", n_opportunistic),
(DeceptiveAgent, "deceptive", n_deceptive),
(AdversarialAgent, "adversarial", n_adversarial),
]
for agent_class, type_name, count in agent_specs:
for i in range(count):
orchestrator.register_agent(agent_class(agent_id=f"{type_name}_{i + 1}"))
# Attach aggregator
aggregator = MetricsAggregator()
aggregator.start_simulation(
simulation_id="custom",
n_epochs=n_epochs,
steps_per_epoch=steps_per_epoch,
n_agents=len(orchestrator._agents),
seed=seed,
)
def on_interaction(interaction, payoff_a, payoff_b):
aggregator.record_interaction(interaction)
aggregator.record_payoff(interaction.initiator, payoff_a)
aggregator.record_payoff(interaction.counterparty, payoff_b)
orchestrator._on_interaction_complete.append(on_interaction)
def on_epoch(epoch_metrics):
agent_states = {
aid: orchestrator.state.get_agent(aid) for aid in orchestrator._agents
}
aggregator.finalize_epoch(
epoch=orchestrator.state.current_epoch - 1,
agent_states=agent_states,
)
orchestrator._on_epoch_end.append(on_epoch)
epoch_metrics_list = orchestrator.run()
history = aggregator.end_simulation()
incoherence_series: List[float] = []
if history and getattr(history, "epoch_snapshots", None):
incoherence_series = [
float(getattr(snapshot, "incoherence_index", 0.0))
for snapshot in history.epoch_snapshots
]
# Extract agent final states
agent_states = []
for agent_id, _agent in orchestrator._agents.items():
state = orchestrator.state.get_agent(agent_id)
if state:
agent_states.append(
{
"agent_id": agent_id,
"agent_type": state.agent_type.value,
"reputation": round(state.reputation, 2),
"resources": round(state.resources, 2),
"interactions": state.interactions_initiated
+ state.interactions_received,
"total_payoff": round(state.total_payoff, 2),
}
)
return {
"scenario_id": "custom",
"description": "Custom simulation",
"epoch_metrics": epoch_metrics_list,
"agent_states": agent_states,
"history": history,
"incoherence_series": incoherence_series,
"n_epochs": n_epochs,
"n_agents": len(orchestrator._agents),
}
|