File size: 13,057 Bytes
d38a70f |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 |
"""
API Server for Mamba Swarm
FastAPI-based server for serving the distributed Mamba language model
"""
from fastapi import FastAPI, HTTPException, BackgroundTasks, Depends
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import StreamingResponse
from pydantic import BaseModel, Field
from typing import List, Optional, Dict, Any, AsyncGenerator
import asyncio
import json
import time
import logging
import torch
from contextlib import asynccontextmanager
import uvicorn
# Import your swarm components
from system.mambaSwarm import SwarmEngine
from system.inference import InferenceEngine
from routing.router import Router
from training.trainer import setup_logging
# Pydantic models for API
class GenerationRequest(BaseModel):
prompt: str = Field(..., description="Input text prompt")
max_length: int = Field(default=100, ge=1, le=2048, description="Maximum generation length")
temperature: float = Field(default=0.7, ge=0.1, le=2.0, description="Sampling temperature")
top_p: float = Field(default=0.9, ge=0.1, le=1.0, description="Top-p sampling")
top_k: int = Field(default=50, ge=1, le=100, description="Top-k sampling")
repetition_penalty: float = Field(default=1.1, ge=1.0, le=2.0, description="Repetition penalty")
stream: bool = Field(default=False, description="Enable streaming response")
domain: Optional[str] = Field(default=None, description="Specific domain for routing")
class GenerationResponse(BaseModel):
generated_text: str
prompt: str
generation_time: float
tokens_generated: int
model_info: Dict[str, Any]
class StreamingToken(BaseModel):
token: str
is_final: bool = False
metadata: Optional[Dict[str, Any]] = None
class HealthResponse(BaseModel):
status: str
swarm_status: Dict[str, Any]
system_info: Dict[str, Any]
timestamp: float
class ModelInfo(BaseModel):
total_parameters: int
active_encoders: int
total_encoders: int
memory_usage: Dict[str, float]
device_info: List[str]
# Global swarm engine instance
swarm_engine: Optional[SwarmEngine] = None
inference_engine: Optional[InferenceEngine] = None
@asynccontextmanager
async def lifespan(app: FastAPI):
"""Manage application lifespan"""
global swarm_engine, inference_engine
# Startup
logging.info("Initializing Mamba Swarm API Server...")
try:
# Initialize swarm engine
swarm_engine = SwarmEngine()
await asyncio.get_event_loop().run_in_executor(None, swarm_engine.initialize)
# Initialize inference engine
inference_engine = InferenceEngine(swarm_engine)
logging.info("Mamba Swarm API Server initialized successfully")
except Exception as e:
logging.error(f"Failed to initialize swarm: {e}")
raise
yield
# Shutdown
logging.info("Shutting down Mamba Swarm API Server...")
if swarm_engine:
swarm_engine.shutdown()
# Create FastAPI app
app = FastAPI(
title="Mamba Swarm API",
description="Distributed Mamba Language Model API with 100 encoder units",
version="1.0.0",
lifespan=lifespan
)
# Add CORS middleware
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
# Dependency to get swarm engine
async def get_swarm_engine() -> SwarmEngine:
if swarm_engine is None:
raise HTTPException(status_code=503, detail="Swarm engine not initialized")
return swarm_engine
async def get_inference_engine() -> InferenceEngine:
if inference_engine is None:
raise HTTPException(status_code=503, detail="Inference engine not initialized")
return inference_engine
@app.get("/health", response_model=HealthResponse)
async def health_check(swarm: SwarmEngine = Depends(get_swarm_engine)):
"""Health check endpoint"""
try:
swarm_status = swarm.get_status()
system_info = {
"cuda_available": torch.cuda.is_available(),
"cuda_device_count": torch.cuda.device_count() if torch.cuda.is_available() else 0,
"python_version": "3.8+",
}
return HealthResponse(
status="healthy",
swarm_status=swarm_status,
system_info=system_info,
timestamp=time.time()
)
except Exception as e:
raise HTTPException(status_code=500, detail=f"Health check failed: {str(e)}")
@app.get("/model/info", response_model=ModelInfo)
async def get_model_info(swarm: SwarmEngine = Depends(get_swarm_engine)):
"""Get model information"""
try:
info = swarm.get_model_info()
memory_stats = swarm.memory_manager.get_memory_stats()
return ModelInfo(
total_parameters=info.get("total_parameters", 7000000000), # 100 * 70M
active_encoders=info.get("active_encoders", 100),
total_encoders=info.get("total_encoders", 100),
memory_usage={
"system_memory_gb": memory_stats.used_memory,
"gpu_memory_gb": memory_stats.gpu_memory,
"cache_size_gb": memory_stats.cache_size
},
device_info=info.get("devices", ["cuda:0" if torch.cuda.is_available() else "cpu"])
)
except Exception as e:
raise HTTPException(status_code=500, detail=f"Failed to get model info: {str(e)}")
@app.post("/generate", response_model=GenerationResponse)
async def generate_text(
request: GenerationRequest,
inference: InferenceEngine = Depends(get_inference_engine)
):
"""Generate text from prompt"""
try:
start_time = time.time()
# Generate text
result = await asyncio.get_event_loop().run_in_executor(
None,
inference.generate,
request.prompt,
{
"max_length": request.max_length,
"temperature": request.temperature,
"top_p": request.top_p,
"top_k": request.top_k,
"repetition_penalty": request.repetition_penalty,
"domain": request.domain
}
)
generation_time = time.time() - start_time
return GenerationResponse(
generated_text=result["generated_text"],
prompt=request.prompt,
generation_time=generation_time,
tokens_generated=result.get("tokens_generated", 0),
model_info=result.get("model_info", {})
)
except Exception as e:
raise HTTPException(status_code=500, detail=f"Generation failed: {str(e)}")
@app.post("/generate/stream")
async def generate_text_stream(
request: GenerationRequest,
inference: InferenceEngine = Depends(get_inference_engine)
):
"""Generate text with streaming response"""
if not request.stream:
raise HTTPException(status_code=400, detail="Streaming not requested")
async def generate_stream() -> AsyncGenerator[str, None]:
try:
# Create generator for streaming
generator = inference.generate_stream(
request.prompt,
{
"max_length": request.max_length,
"temperature": request.temperature,
"top_p": request.top_p,
"top_k": request.top_k,
"repetition_penalty": request.repetition_penalty,
"domain": request.domain
}
)
for token_data in generator:
streaming_token = StreamingToken(
token=token_data.get("token", ""),
is_final=token_data.get("is_final", False),
metadata=token_data.get("metadata", {})
)
yield f"data: {streaming_token.json()}\n\n"
if streaming_token.is_final:
break
except Exception as e:
error_token = StreamingToken(
token="",
is_final=True,
metadata={"error": str(e)}
)
yield f"data: {error_token.json()}\n\n"
return StreamingResponse(
generate_stream(),
media_type="text/plain",
headers={"Cache-Control": "no-cache", "Connection": "keep-alive"}
)
@app.post("/generate/batch")
async def generate_batch(
requests: List[GenerationRequest],
inference: InferenceEngine = Depends(get_inference_engine)
):
"""Generate text for multiple prompts"""
if len(requests) > 10:
raise HTTPException(status_code=400, detail="Batch size too large (max 10)")
try:
# Process requests in parallel
tasks = []
for req in requests:
task = asyncio.get_event_loop().run_in_executor(
None,
inference.generate,
req.prompt,
{
"max_length": req.max_length,
"temperature": req.temperature,
"top_p": req.top_p,
"top_k": req.top_k,
"repetition_penalty": req.repetition_penalty,
"domain": req.domain
}
)
tasks.append(task)
results = await asyncio.gather(*tasks, return_exceptions=True)
responses = []
for i, (req, result) in enumerate(zip(requests, results)):
if isinstance(result, Exception):
responses.append({
"error": str(result),
"prompt": req.prompt,
"index": i
})
else:
responses.append(GenerationResponse(
generated_text=result["generated_text"],
prompt=req.prompt,
generation_time=result.get("generation_time", 0),
tokens_generated=result.get("tokens_generated", 0),
model_info=result.get("model_info", {})
))
return {"responses": responses}
except Exception as e:
raise HTTPException(status_code=500, detail=f"Batch generation failed: {str(e)}")
@app.get("/metrics")
async def get_metrics(swarm: SwarmEngine = Depends(get_swarm_engine)):
"""Get system metrics"""
try:
metrics = {
"memory_report": swarm.memory_manager.get_memory_report(),
"swarm_metrics": swarm.get_metrics(),
"inference_stats": swarm.get_inference_stats() if hasattr(swarm, 'get_inference_stats') else {},
"timestamp": time.time()
}
return metrics
except Exception as e:
raise HTTPException(status_code=500, detail=f"Failed to get metrics: {str(e)}")
@app.post("/admin/reload")
async def reload_model(
background_tasks: BackgroundTasks,
swarm: SwarmEngine = Depends(get_swarm_engine)
):
"""Reload the model (admin endpoint)"""
try:
background_tasks.add_task(swarm.reload_model)
return {"message": "Model reload initiated"}
except Exception as e:
raise HTTPException(status_code=500, detail=f"Failed to reload model: {str(e)}")
@app.post("/admin/cleanup")
async def cleanup_memory(swarm: SwarmEngine = Depends(get_swarm_engine)):
"""Force memory cleanup (admin endpoint)"""
try:
swarm.memory_manager.cleanup_memory(aggressive=True)
return {"message": "Memory cleanup completed"}
except Exception as e:
raise HTTPException(status_code=500, detail=f"Failed to cleanup memory: {str(e)}")
# Error handlers
@app.exception_handler(HTTPException)
async def http_exception_handler(request, exc):
return {
"error": exc.detail,
"status_code": exc.status_code,
"timestamp": time.time()
}
@app.exception_handler(Exception)
async def general_exception_handler(request, exc):
logging.error(f"Unhandled exception: {exc}")
return {
"error": "Internal server error",
"status_code": 500,
"timestamp": time.time()
}
def run_server(host: str = "0.0.0.0", port: int = 8000, workers: int = 1):
"""Run the API server"""
setup_logging()
config = uvicorn.Config(
app=app,
host=host,
port=port,
workers=workers,
log_level="info",
access_log=True,
reload=False # Set to True for development
)
server = uvicorn.Server(config)
server.run()
if __name__ == "__main__":
run_server() |