Skip to content

Drop in a screenshot and convert it to clean code (HTML/Tailwind/React/Vue)

Python

Limit concurrent Replicate requests to 20 per tool call

Claudeb18302fe3494
diff --git a/backend/agent/tools/runtime.py b/backend/agent/tools/runtime.py
index ba031c83..c8b5e5c2 100644
--- a/backend/agent/tools/runtime.py
+++ b/backend/agent/tools/runtime.py
@@ -261,9 +261,14 @@ class AgentToolRuntime:
summary={"error": "No valid image_urls"},
)
- tasks = [
- remove_background(url, REPLICATE_API_KEY) for url in unique_urls
- ]
+ max_concurrent = 20
+ semaphore = asyncio.Semaphore(max_concurrent)
+
+ async def _limited(url: str) -> str:
+ async with semaphore:
+ return await remove_background(url, REPLICATE_API_KEY)
+
+ tasks = [_limited(url) for url in unique_urls]
raw_results = await asyncio.gather(*tasks, return_exceptions=True)
results: List[Dict[str, Any]] = []
diff --git a/backend/image_generation/generation.py b/backend/image_generation/generation.py
index 008ae779..9219088b 100644
--- a/backend/image_generation/generation.py
+++ b/backend/image_generation/generation.py
@@ -7,6 +7,9 @@ from openai import AsyncOpenAI
from image_generation.replicate import call_replicate
+MAX_CONCURRENT_REPLICATE_REQUESTS = 20
+
+
async def process_tasks(
prompts: List[str],
api_key: str,
@@ -16,9 +19,16 @@ async def process_tasks(
start_time = time.time()
if model == "dalle3":
tasks = [generate_image_dalle(prompt, api_key, base_url) for prompt in prompts]
+ results = await asyncio.gather(*tasks, return_exceptions=True)
else:
- tasks = [generate_image_replicate(prompt, api_key) for prompt in prompts]
- results = await asyncio.gather(*tasks, return_exceptions=True)
+ semaphore = asyncio.Semaphore(MAX_CONCURRENT_REPLICATE_REQUESTS)
+
+ async def _limited(prompt: str) -> str:
+ async with semaphore:
+ return await generate_image_replicate(prompt, api_key)
+
+ tasks = [_limited(prompt) for prompt in prompts]
+ results = await asyncio.gather(*tasks, return_exceptions=True)
end_time = time.time()
generation_time = end_time - start_time
print(f"Image generation time: {generation_time:.2f} seconds")
diff --git a/backend/image_generation/replicate.py b/backend/image_generation/replicate.py
index 97114001..843eade9 100644
--- a/backend/image_generation/replicate.py
+++ b/backend/image_generation/replicate.py
@@ -10,9 +10,6 @@ REMOVE_BACKGROUND_VERSION = (
)
POLL_INTERVAL_SECONDS = 0.1
MAX_POLLS = 100
-MAX_CONCURRENT_REQUESTS = 20
-
-_replicate_semaphore = asyncio.Semaphore(MAX_CONCURRENT_REQUESTS)
def _build_headers(api_token: str) -> dict[str, str]:
@@ -60,7 +57,7 @@ async def _run_prediction(
) -> Any:
headers = _build_headers(api_token)
- async with _replicate_semaphore, httpx.AsyncClient() as client:
+ async with httpx.AsyncClient() as client:
try:
response = await client.post(endpoint_url, headers=headers, json=payload)
response.raise_for_status()