File size: 20,062 Bytes
31f0e50 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 | """
Groq API Client Module with Rate Limiting and Retry Logic.
Implements Task 6.1 requirements:
- Rate limiting for Groq API (25 calls/minute with buffer)
- Retry logic with exponential backoff
- Response time measurement
- Support for Hindi and English prompts
Acceptance Criteria:
- Rate limiting prevents API errors
- Retry logic handles transient failures
- Response time <2s per call
"""
import time
import threading
from functools import wraps
from typing import Any, Callable, Dict, List, Optional, TypeVar, Union
from datetime import datetime
from app.config import settings
from app.utils.logger import get_logger
from app.utils.metrics import track_response_time, track_error
logger = get_logger(__name__)
# Type variable for generic function return
T = TypeVar("T")
# Default rate limiting configuration
DEFAULT_MAX_CALLS_PER_MINUTE = 25 # Buffer below Groq's 30 limit
DEFAULT_MAX_RETRIES = 3
DEFAULT_BASE_DELAY = 1.0 # seconds
DEFAULT_MAX_DELAY = 30.0 # seconds
DEFAULT_TIMEOUT = 30.0 # seconds
# Response time target
TARGET_RESPONSE_TIME = 2.0 # seconds
class RateLimitError(Exception):
"""Exception raised when rate limit is exceeded and waiting is not desired."""
def __init__(self, message: str, retry_after: float = 0.0):
super().__init__(message)
self.retry_after = retry_after
class GroqAPIError(Exception):
"""Exception raised for Groq API errors."""
def __init__(self, message: str, status_code: Optional[int] = None, retryable: bool = False):
super().__init__(message)
self.status_code = status_code
self.retryable = retryable
class RateLimiter:
"""
Thread-safe rate limiter using sliding window algorithm.
Limits API calls to a maximum number per minute window.
When limit is reached, blocks until a slot becomes available.
Attributes:
max_calls: Maximum calls allowed per minute
calls: List of call timestamps
lock: Thread lock for synchronization
"""
def __init__(self, max_calls_per_minute: int = DEFAULT_MAX_CALLS_PER_MINUTE):
"""
Initialize the rate limiter.
Args:
max_calls_per_minute: Maximum API calls allowed per minute
"""
self.max_calls = max_calls_per_minute
self.calls: List[float] = []
self.lock = threading.Lock()
logger.info(f"RateLimiter initialized: {max_calls_per_minute} calls/minute")
def acquire(self, block: bool = True) -> bool:
"""
Acquire a rate limit slot.
Args:
block: Whether to block until slot is available
Returns:
True if slot acquired, False if non-blocking and limit exceeded
Raises:
RateLimitError: If block=False and limit exceeded
"""
with self.lock:
now = time.time()
# Remove calls older than 60 seconds
self.calls = [c for c in self.calls if c > now - 60]
if len(self.calls) >= self.max_calls:
if not block:
wait_time = 60 - (now - self.calls[0])
raise RateLimitError(
f"Rate limit exceeded: {len(self.calls)}/{self.max_calls}",
retry_after=wait_time,
)
# Calculate sleep time
sleep_time = 60 - (now - self.calls[0]) + 0.1 # Small buffer
logger.warning(
f"Rate limit reached ({len(self.calls)}/{self.max_calls}), "
f"sleeping for {sleep_time:.2f}s"
)
# Release lock while sleeping
self.lock.release()
try:
time.sleep(sleep_time)
finally:
self.lock.acquire()
# Update now and clean up again
now = time.time()
self.calls = [c for c in self.calls if c > now - 60]
# Record this call
self.calls.append(now)
return True
def get_current_usage(self) -> Dict[str, Any]:
"""
Get current rate limit usage.
Returns:
Dict with current usage statistics
"""
with self.lock:
now = time.time()
active_calls = [c for c in self.calls if c > now - 60]
return {
"current_calls": len(active_calls),
"max_calls": self.max_calls,
"remaining": self.max_calls - len(active_calls),
"reset_in_seconds": 60 - (now - active_calls[0]) if active_calls else 0,
}
def reset(self) -> None:
"""Reset the rate limiter (for testing)."""
with self.lock:
self.calls = []
def __call__(self, func: Callable[..., T]) -> Callable[..., T]:
"""
Use as decorator to rate limit function calls.
Args:
func: Function to decorate
Returns:
Decorated function
"""
@wraps(func)
def wrapper(*args, **kwargs) -> T:
self.acquire(block=True)
return func(*args, **kwargs)
return wrapper
def exponential_backoff(
attempt: int,
base_delay: float = DEFAULT_BASE_DELAY,
max_delay: float = DEFAULT_MAX_DELAY,
) -> float:
"""
Calculate exponential backoff delay.
Uses exponential backoff with jitter for optimal retry timing.
Args:
attempt: Current attempt number (0-indexed)
base_delay: Base delay in seconds
max_delay: Maximum delay in seconds
Returns:
Delay in seconds
"""
import random
# Calculate exponential delay with jitter
delay = min(base_delay * (2 ** attempt), max_delay)
# Add jitter (up to 25% of delay)
jitter = delay * 0.25 * random.random()
return delay + jitter
def is_retryable_error(error: Exception) -> bool:
"""
Check if an error is retryable.
Args:
error: Exception to check
Returns:
True if error should be retried
"""
error_str = str(error).lower()
# Rate limit errors are retryable
if "rate_limit" in error_str or "rate limit" in error_str:
return True
# Timeout errors are retryable
if "timeout" in error_str:
return True
# Connection errors are retryable
if "connection" in error_str or "network" in error_str:
return True
# Server errors (5xx) are generally retryable
if "500" in error_str or "502" in error_str or "503" in error_str or "504" in error_str:
return True
# Check for GroqAPIError
if isinstance(error, GroqAPIError):
return error.retryable
return False
def retry_with_backoff(
max_retries: int = DEFAULT_MAX_RETRIES,
base_delay: float = DEFAULT_BASE_DELAY,
max_delay: float = DEFAULT_MAX_DELAY,
) -> Callable[[Callable[..., T]], Callable[..., T]]:
"""
Decorator factory for retry with exponential backoff.
Args:
max_retries: Maximum number of retry attempts
base_delay: Base delay between retries
max_delay: Maximum delay between retries
Returns:
Decorator function
"""
def decorator(func: Callable[..., T]) -> Callable[..., T]:
@wraps(func)
def wrapper(*args, **kwargs) -> T:
last_error: Optional[Exception] = None
for attempt in range(max_retries):
try:
return func(*args, **kwargs)
except Exception as e:
last_error = e
if not is_retryable_error(e):
logger.error(f"Non-retryable error on attempt {attempt + 1}: {e}")
raise
if attempt < max_retries - 1:
delay = exponential_backoff(attempt, base_delay, max_delay)
logger.warning(
f"Retryable error on attempt {attempt + 1}/{max_retries}: {e}. "
f"Retrying in {delay:.2f}s"
)
time.sleep(delay)
else:
logger.error(f"Max retries ({max_retries}) exceeded: {e}")
# Raise the last error
if last_error:
raise last_error
raise RuntimeError("Unexpected state in retry logic")
return wrapper
return decorator
class GroqClient:
"""
High-level Groq API client with rate limiting and retry logic.
Provides a robust interface to the Groq LLM API with:
- Automatic rate limiting
- Exponential backoff retry
- Response time tracking
- Multi-language support
Attributes:
llm: Underlying LangChain Groq LLM
rate_limiter: Rate limiter instance
max_retries: Maximum retry attempts
"""
def __init__(
self,
api_key: Optional[str] = None,
model: Optional[str] = None,
temperature: Optional[float] = None,
max_tokens: Optional[int] = None,
max_calls_per_minute: int = DEFAULT_MAX_CALLS_PER_MINUTE,
max_retries: int = DEFAULT_MAX_RETRIES,
):
"""
Initialize the Groq client.
Args:
api_key: Groq API key (defaults to settings)
model: Model name (defaults to settings)
temperature: Generation temperature (defaults to settings)
max_tokens: Maximum tokens (defaults to settings)
max_calls_per_minute: Rate limit
max_retries: Maximum retry attempts
"""
self.api_key = api_key or settings.GROQ_API_KEY
self.model = model or settings.GROQ_MODEL
self.temperature = temperature if temperature is not None else settings.GROQ_TEMPERATURE
self.max_tokens = max_tokens or settings.GROQ_MAX_TOKENS
self.rate_limiter = RateLimiter(max_calls_per_minute)
self.max_retries = max_retries
self.llm = None
self._initialized = False
if self.api_key:
self._initialize_llm()
def _initialize_llm(self) -> None:
"""Initialize the LangChain Groq LLM."""
try:
from langchain_groq import ChatGroq
self.llm = ChatGroq(
model=self.model,
api_key=self.api_key,
temperature=self.temperature,
max_tokens=self.max_tokens,
)
self._initialized = True
logger.info(f"GroqClient initialized: model={self.model}")
except ImportError as e:
logger.error(f"Failed to import langchain_groq: {e}")
raise
except Exception as e:
logger.error(f"Failed to initialize GroqClient: {e}")
raise
@property
def is_available(self) -> bool:
"""Check if the client is available for calls."""
return self._initialized and self.llm is not None
def invoke(
self,
messages: List[Dict[str, str]],
language: str = "en",
) -> str:
"""
Invoke the LLM with rate limiting and retry.
Args:
messages: List of message dicts with 'role' and 'content'
language: Expected response language ('en' or 'hi')
Returns:
Generated response text
Raises:
GroqAPIError: If API call fails after retries
RateLimitError: If rate limit exceeded and non-blocking
"""
if not self.is_available:
raise GroqAPIError("GroqClient not initialized", retryable=False)
return self._invoke_with_rate_limit_and_retry(messages, language)
def _invoke_with_rate_limit_and_retry(
self,
messages: List[Dict[str, str]],
language: str,
) -> str:
"""
Internal invoke with rate limiting and retry.
Args:
messages: LLM messages
language: Response language
Returns:
Generated text
"""
last_error: Optional[Exception] = None
for attempt in range(self.max_retries):
try:
# Acquire rate limit slot
self.rate_limiter.acquire(block=True)
# Measure response time
start_time = time.time()
try:
# Convert messages to LangChain format
from langchain_core.messages import SystemMessage, HumanMessage, AIMessage
lc_messages = []
for msg in messages:
role = msg.get("role", "user")
content = msg.get("content", "")
if role == "system":
lc_messages.append(SystemMessage(content=content))
elif role == "assistant":
lc_messages.append(AIMessage(content=content))
else:
lc_messages.append(HumanMessage(content=content))
# Make API call
response = self.llm.invoke(lc_messages)
# Track response time
elapsed = time.time() - start_time
track_response_time(elapsed)
# Log if response time exceeds target
if elapsed > TARGET_RESPONSE_TIME:
logger.warning(
f"Response time {elapsed:.2f}s exceeds target {TARGET_RESPONSE_TIME}s"
)
else:
logger.debug(f"Groq API response in {elapsed:.2f}s")
# Extract content
if hasattr(response, "content"):
return response.content
return str(response)
except Exception as e:
elapsed = time.time() - start_time
track_response_time(elapsed)
raise
except Exception as e:
last_error = e
track_error("groq_api_error")
if not is_retryable_error(e):
logger.error(f"Non-retryable Groq error: {e}")
raise GroqAPIError(str(e), retryable=False)
if attempt < self.max_retries - 1:
delay = exponential_backoff(attempt)
logger.warning(
f"Groq API error (attempt {attempt + 1}/{self.max_retries}): {e}. "
f"Retrying in {delay:.2f}s"
)
time.sleep(delay)
else:
logger.error(f"Groq API failed after {self.max_retries} attempts: {e}")
if last_error:
raise GroqAPIError(str(last_error), retryable=True)
raise GroqAPIError("Unknown error in Groq API call", retryable=False)
def generate_response(
self,
system_prompt: str,
user_message: str,
language: str = "en",
conversation_history: Optional[List[Dict[str, str]]] = None,
) -> str:
"""
Generate a response with proper message formatting.
Args:
system_prompt: System prompt text
user_message: User message text
language: Response language
conversation_history: Optional previous messages
Returns:
Generated response text
"""
messages = [{"role": "system", "content": system_prompt}]
# Add conversation history if provided
if conversation_history:
for msg in conversation_history:
role = msg.get("sender", "user")
if role == "scammer":
role = "user"
elif role == "agent":
role = "assistant"
messages.append({
"role": role,
"content": msg.get("message", msg.get("content", "")),
})
# Add current user message
messages.append({"role": "user", "content": user_message})
return self.invoke(messages, language=language)
def get_rate_limit_status(self) -> Dict[str, Any]:
"""Get current rate limit status."""
return self.rate_limiter.get_current_usage()
def reset_rate_limiter(self) -> None:
"""Reset the rate limiter (for testing)."""
self.rate_limiter.reset()
# Rate limiter decorator for use with existing LLM instances
_default_rate_limiter = RateLimiter(DEFAULT_MAX_CALLS_PER_MINUTE)
@_default_rate_limiter
@retry_with_backoff(max_retries=DEFAULT_MAX_RETRIES)
def call_groq_with_retry(
llm: Any,
messages: List[Any],
timeout: float = DEFAULT_TIMEOUT,
) -> Any:
"""
Call Groq API with rate limiting and retry logic.
This is a convenience function for calling existing LLM instances
with proper rate limiting and retry.
Args:
llm: LangChain Groq LLM instance
messages: List of LangChain messages
timeout: Request timeout in seconds
Returns:
LLM response object
Raises:
GroqAPIError: If call fails after retries
"""
start_time = time.time()
try:
response = llm.invoke(messages)
elapsed = time.time() - start_time
track_response_time(elapsed)
if elapsed > TARGET_RESPONSE_TIME:
logger.warning(f"Response time {elapsed:.2f}s exceeds {TARGET_RESPONSE_TIME}s target")
return response
except Exception as e:
elapsed = time.time() - start_time
track_response_time(elapsed)
track_error("groq_api_error")
raise
def reset_rate_limiter() -> None:
"""Reset the default rate limiter (for testing)."""
_default_rate_limiter.reset()
def get_rate_limit_status() -> Dict[str, Any]:
"""Get default rate limiter status."""
return _default_rate_limiter.get_current_usage()
# Singleton client instance
_groq_client: Optional[GroqClient] = None
def get_groq_client() -> GroqClient:
"""
Get singleton Groq client instance.
Returns:
GroqClient instance
"""
global _groq_client
if _groq_client is None:
_groq_client = GroqClient()
return _groq_client
def reset_groq_client() -> None:
"""Reset the singleton client (for testing)."""
global _groq_client
_groq_client = None
|