Spaces:
Running
on
Zero
Running
on
Zero
File size: 9,015 Bytes
ca80d1d |
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 |
import logging
import gc
import time
from typing import Dict, Any, Optional, Callable
from dataclasses import dataclass, field
from threading import Lock
import torch
logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)
@dataclass
class ModelInfo:
"""Information about a registered model."""
name: str
loader: Callable[[], Any]
is_critical: bool = False # Critical models are not unloaded under memory pressure
estimated_memory_gb: float = 0.0
is_loaded: bool = False
last_used: float = 0.0
model_instance: Any = None
class ModelManager:
"""
Singleton model manager for unified model lifecycle management.
Handles lazy loading, caching, and intelligent memory management.
"""
_instance = None
_lock = Lock()
def __new__(cls):
if cls._instance is None:
with cls._lock:
if cls._instance is None:
cls._instance = super().__new__(cls)
cls._instance._initialized = False
return cls._instance
def __init__(self):
if self._initialized:
return
self._models: Dict[str, ModelInfo] = {}
self._memory_threshold = 0.80 # Trigger cleanup at 80% GPU memory usage
self._device = self._detect_device()
logger.info(f"🔧 ModelManager initialized on {self._device}")
self._initialized = True
def _detect_device(self) -> str:
"""Detect best available device."""
if torch.cuda.is_available():
return "cuda"
elif hasattr(torch.backends, 'mps') and torch.backends.mps.is_available():
return "mps"
return "cpu"
def register_model(
self,
name: str,
loader: Callable[[], Any],
is_critical: bool = False,
estimated_memory_gb: float = 0.0
):
"""
Register a model for managed loading.
Args:
name: Unique model identifier
loader: Callable that returns the loaded model
is_critical: If True, model won't be unloaded under memory pressure
estimated_memory_gb: Estimated GPU memory usage in GB
"""
if name in self._models:
logger.warning(f"⚠️ Model '{name}' already registered, updating")
self._models[name] = ModelInfo(
name=name,
loader=loader,
is_critical=is_critical,
estimated_memory_gb=estimated_memory_gb,
is_loaded=False,
last_used=0.0,
model_instance=None
)
logger.info(f"📝 Registered model: {name} (critical={is_critical}, ~{estimated_memory_gb:.1f}GB)")
def load_model(self, name: str) -> Any:
"""
Load a model by name. Returns cached instance if already loaded.
Args:
name: Model identifier
Returns:
Loaded model instance
Raises:
KeyError: If model not registered
RuntimeError: If loading fails
"""
if name not in self._models:
raise KeyError(f"Model '{name}' not registered")
model_info = self._models[name]
# Return cached instance
if model_info.is_loaded and model_info.model_instance is not None:
model_info.last_used = time.time()
logger.debug(f"📦 Using cached model: {name}")
return model_info.model_instance
# Check memory pressure before loading
self.check_memory_pressure()
# Load the model
try:
logger.info(f"📥 Loading model: {name}")
start_time = time.time()
model_instance = model_info.loader()
model_info.model_instance = model_instance
model_info.is_loaded = True
model_info.last_used = time.time()
load_time = time.time() - start_time
logger.info(f"✅ Model '{name}' loaded in {load_time:.1f}s")
return model_instance
except Exception as e:
logger.error(f"❌ Failed to load model '{name}': {e}")
raise RuntimeError(f"Model loading failed: {e}")
def unload_model(self, name: str):
"""
Unload a specific model to free memory.
Args:
name: Model identifier
"""
if name not in self._models:
return
model_info = self._models[name]
if not model_info.is_loaded:
return
try:
logger.info(f"🗑️ Unloading model: {name}")
# Delete model instance
if model_info.model_instance is not None:
del model_info.model_instance
model_info.model_instance = None
model_info.is_loaded = False
# Cleanup
gc.collect()
if torch.cuda.is_available():
torch.cuda.empty_cache()
logger.info(f"✅ Model '{name}' unloaded")
except Exception as e:
logger.error(f"❌ Error unloading model '{name}': {e}")
def check_memory_pressure(self) -> bool:
"""
Check GPU memory usage and unload least-used non-critical models if needed.
Returns:
True if cleanup was performed
"""
if not torch.cuda.is_available():
return False
allocated = torch.cuda.memory_allocated() / 1024**3
total = torch.cuda.get_device_properties(0).total_memory / 1024**3
usage_ratio = allocated / total
if usage_ratio < self._memory_threshold:
return False
logger.warning(f"⚠️ Memory pressure detected: {usage_ratio:.1%} used")
# Find non-critical models sorted by last used time
unloadable = [
(name, info) for name, info in self._models.items()
if info.is_loaded and not info.is_critical
]
unloadable.sort(key=lambda x: x[1].last_used)
# Unload oldest non-critical models
cleaned = False
for name, info in unloadable:
self.unload_model(name)
cleaned = True
# Re-check memory
new_ratio = torch.cuda.memory_allocated() / torch.cuda.get_device_properties(0).total_memory
if new_ratio < self._memory_threshold * 0.7: # Target 70% of threshold
break
return cleaned
def force_cleanup(self):
"""Force cleanup all non-critical models and clear caches."""
logger.info("🧹 Force cleanup initiated")
# Unload all non-critical models
for name, info in self._models.items():
if info.is_loaded and not info.is_critical:
self.unload_model(name)
# Aggressive garbage collection
for _ in range(5):
gc.collect()
if torch.cuda.is_available():
torch.cuda.empty_cache()
torch.cuda.ipc_collect()
torch.cuda.synchronize()
logger.info("✅ Force cleanup completed")
def get_memory_status(self) -> Dict[str, Any]:
"""
Get detailed memory status.
Returns:
Dictionary with memory statistics
"""
status = {
"device": self._device,
"models": {},
"total_estimated_gb": 0.0
}
# Model status
for name, info in self._models.items():
status["models"][name] = {
"loaded": info.is_loaded,
"critical": info.is_critical,
"estimated_gb": info.estimated_memory_gb,
"last_used": info.last_used
}
if info.is_loaded:
status["total_estimated_gb"] += info.estimated_memory_gb
# GPU memory
if torch.cuda.is_available():
allocated = torch.cuda.memory_allocated() / 1024**3
total = torch.cuda.get_device_properties(0).total_memory / 1024**3
cached = torch.cuda.memory_reserved() / 1024**3
status["gpu"] = {
"allocated_gb": round(allocated, 2),
"total_gb": round(total, 2),
"cached_gb": round(cached, 2),
"free_gb": round(total - allocated, 2),
"usage_percent": round((allocated / total) * 100, 1)
}
return status
def get_loaded_models(self) -> list:
"""Get list of currently loaded model names."""
return [name for name, info in self._models.items() if info.is_loaded]
def is_model_loaded(self, name: str) -> bool:
"""Check if a specific model is loaded."""
if name not in self._models:
return False
return self._models[name].is_loaded
# Global singleton instance
_model_manager = None
def get_model_manager() -> ModelManager:
"""Get the global ModelManager singleton instance."""
global _model_manager
if _model_manager is None:
_model_manager = ModelManager()
return _model_manager
|