File size: 9,698 Bytes
8b7b267 |
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 |
#!/usr/bin/env python3
"""
ML Training Service
===================
سرویس آموزش مدلهای یادگیری ماشین با قابلیت پیگیری پیشرفت و ذخیره checkpoint
"""
from typing import Optional, List, Dict, Any
from datetime import datetime
from sqlalchemy.orm import Session
from sqlalchemy import and_, desc
import uuid
import logging
import json
from database.models import (
Base, MLTrainingJob, TrainingStep, TrainingStatus
)
logger = logging.getLogger(__name__)
class MLTrainingService:
"""سرویس اصلی آموزش مدلهای ML"""
def __init__(self, db_session: Session):
"""
Initialize the ML training service.
Args:
db_session: SQLAlchemy database session
"""
self.db = db_session
def start_training(
self,
model_name: str,
training_data_start: datetime,
training_data_end: datetime,
batch_size: int = 32,
learning_rate: Optional[float] = None,
config: Optional[Dict[str, Any]] = None
) -> Dict[str, Any]:
"""
Start training a model.
Args:
model_name: Name of the model to train
training_data_start: Start date for training data
training_data_end: End date for training data
batch_size: Training batch size
learning_rate: Learning rate (optional)
config: Additional training configuration
Returns:
Dict containing training job details
"""
try:
# Generate job ID
job_id = f"TR-{uuid.uuid4().hex[:12].upper()}"
# Create training job
job = MLTrainingJob(
job_id=job_id,
model_name=model_name,
model_version="1.0.0",
status=TrainingStatus.PENDING,
training_data_start=training_data_start,
training_data_end=training_data_end,
batch_size=batch_size,
learning_rate=learning_rate or 0.001,
config=json.dumps(config) if config else None
)
self.db.add(job)
self.db.commit()
self.db.refresh(job)
logger.info(f"Created training job {job_id} for model {model_name}")
# In production, this would start training in background
# For now, we just return the job details
return self._job_to_dict(job)
except Exception as e:
self.db.rollback()
logger.error(f"Error starting training: {e}", exc_info=True)
raise
def execute_training_step(
self,
job_id: str,
step_number: int,
loss: Optional[float] = None,
accuracy: Optional[float] = None,
learning_rate: Optional[float] = None,
metrics: Optional[Dict[str, Any]] = None
) -> Dict[str, Any]:
"""
Execute a single training step.
Args:
job_id: Training job ID
step_number: Step number
loss: Training loss
accuracy: Training accuracy
learning_rate: Current learning rate
metrics: Additional metrics
Returns:
Dict containing step details
"""
try:
# Get training job
job = self.db.query(MLTrainingJob).filter(
MLTrainingJob.job_id == job_id
).first()
if not job:
raise ValueError(f"Training job {job_id} not found")
if job.status != TrainingStatus.RUNNING:
raise ValueError(f"Training job {job_id} is not in RUNNING status")
# Create training step
step = TrainingStep(
job_id=job_id,
step_number=step_number,
loss=loss,
accuracy=accuracy,
learning_rate=learning_rate,
metrics=json.dumps(metrics) if metrics else None
)
self.db.add(step)
# Update job
job.current_step = step_number
if loss is not None:
job.loss = loss
if accuracy is not None:
job.accuracy = accuracy
if learning_rate is not None:
job.learning_rate = learning_rate
self.db.commit()
self.db.refresh(step)
logger.info(f"Training step {step_number} executed for job {job_id}")
return self._step_to_dict(step)
except Exception as e:
self.db.rollback()
logger.error(f"Error executing training step: {e}", exc_info=True)
raise
def get_training_status(self, job_id: str) -> Dict[str, Any]:
"""
Get the current training status.
Args:
job_id: Training job ID
Returns:
Dict containing training status
"""
try:
job = self.db.query(MLTrainingJob).filter(
MLTrainingJob.job_id == job_id
).first()
if not job:
raise ValueError(f"Training job {job_id} not found")
return self._job_to_dict(job)
except Exception as e:
logger.error(f"Error getting training status: {e}", exc_info=True)
raise
def get_training_history(
self,
model_name: Optional[str] = None,
limit: int = 100
) -> List[Dict[str, Any]]:
"""
Get training history.
Args:
model_name: Filter by model name (optional)
limit: Maximum number of jobs to return
Returns:
List of training job dictionaries
"""
try:
query = self.db.query(MLTrainingJob)
if model_name:
query = query.filter(MLTrainingJob.model_name == model_name)
jobs = query.order_by(desc(MLTrainingJob.created_at)).limit(limit).all()
return [self._job_to_dict(job) for job in jobs]
except Exception as e:
logger.error(f"Error retrieving training history: {e}", exc_info=True)
raise
def update_training_status(
self,
job_id: str,
status: str,
checkpoint_path: Optional[str] = None,
error_message: Optional[str] = None
) -> Dict[str, Any]:
"""
Update training job status.
Args:
job_id: Training job ID
status: New status
checkpoint_path: Path to checkpoint (optional)
error_message: Error message if failed (optional)
Returns:
Dict containing updated job details
"""
try:
job = self.db.query(MLTrainingJob).filter(
MLTrainingJob.job_id == job_id
).first()
if not job:
raise ValueError(f"Training job {job_id} not found")
job.status = TrainingStatus[status.upper()]
if status.upper() == "RUNNING" and not job.started_at:
job.started_at = datetime.utcnow()
if status.upper() in ["COMPLETED", "FAILED", "CANCELLED"]:
job.completed_at = datetime.utcnow()
if checkpoint_path:
job.checkpoint_path = checkpoint_path
if error_message:
job.error_message = error_message
self.db.commit()
self.db.refresh(job)
return self._job_to_dict(job)
except Exception as e:
self.db.rollback()
logger.error(f"Error updating training status: {e}", exc_info=True)
raise
def _job_to_dict(self, job: MLTrainingJob) -> Dict[str, Any]:
"""Convert job model to dictionary."""
config = json.loads(job.config) if job.config else {}
return {
"job_id": job.job_id,
"model_name": job.model_name,
"model_version": job.model_version,
"status": job.status.value if job.status else None,
"training_data_start": job.training_data_start.isoformat() if job.training_data_start else None,
"training_data_end": job.training_data_end.isoformat() if job.training_data_end else None,
"total_steps": job.total_steps,
"current_step": job.current_step,
"batch_size": job.batch_size,
"learning_rate": job.learning_rate,
"loss": job.loss,
"accuracy": job.accuracy,
"checkpoint_path": job.checkpoint_path,
"config": config,
"error_message": job.error_message,
"created_at": job.created_at.isoformat() if job.created_at else None,
"started_at": job.started_at.isoformat() if job.started_at else None,
"completed_at": job.completed_at.isoformat() if job.completed_at else None,
"updated_at": job.updated_at.isoformat() if job.updated_at else None
}
def _step_to_dict(self, step: TrainingStep) -> Dict[str, Any]:
"""Convert step model to dictionary."""
metrics = json.loads(step.metrics) if step.metrics else {}
return {
"id": step.id,
"job_id": step.job_id,
"step_number": step.step_number,
"loss": step.loss,
"accuracy": step.accuracy,
"learning_rate": step.learning_rate,
"metrics": metrics,
"timestamp": step.timestamp.isoformat() if step.timestamp else None
}
|