Update main.py

main.py

@@ -161,32 +161,33 @@ def generate_textAsync(request: Request, input_data: InputDataAsync):
     result_data = asyncio.run(GeneraTestoAsync("https://matteoscript-fastapi.hf.space/Genera", input_data))
     return {"response": result_data}
 
-async def make_request(session, token, data, url, index, max_retries=3):
-    headers = {
-        'Content-Type': 'application/json',
-        'Authorization': 'Bearer ' + token
-    }
-    if (int(index)+1) % 3 == 1:
-        data['max_new_tokens'] = data['max_new_tokens'] 
-    elif (int(index)+1) % 3 == 2:
-        data['max_new_tokens'] = max(200, data['max_new_tokens'] - 200)
-    else:
-        data['max_new_tokens'] = data['max_new_tokens'] + 200 
-    for _ in range(max_retries):
-        try:
-            async with session.post(url, headers=headers, json=data) as response:
-                response.raise_for_status()
-                try:
-                    result_data = await response.json()
-                except aiohttp.ContentTypeError:
-                    result_data = await response.text()                
-                return result_data
-        except (asyncio.TimeoutError, aiohttp.ClientError, requests.exceptions.HTTPError) as e:
-            LoggaTesto("ERRORE ASYNC", {e}, False)
-            if isinstance(e, (asyncio.TimeoutError, requests.exceptions.HTTPError)) and e.response.status in [502, 504]:
-                break
-            await asyncio.sleep(3)  
-    raise Exception("Max retries reached or skipping retries. Unable to make the request.")
+async def make_request(session, token, data, url, index, semaphore, max_retries=3):
+    async with semaphore:
+        headers = {
+            'Content-Type': 'application/json',
+            'Authorization': 'Bearer ' + token
+        }
+        if (int(index)+1) % 3 == 1:
+            data['max_new_tokens'] = data['max_new_tokens'] 
+        elif (int(index)+1) % 3 == 2:
+            data['max_new_tokens'] = max(200, data['max_new_tokens'] - 200)
+        else:
+            data['max_new_tokens'] = data['max_new_tokens'] + 200 
+        for _ in range(max_retries):
+            try:
+                async with session.post(url, headers=headers, json=data) as response:
+                    response.raise_for_status()
+                    try:
+                        result_data = await response.json()
+                    except aiohttp.ContentTypeError:
+                        result_data = await response.text()                
+                    return result_data
+            except (asyncio.TimeoutError, aiohttp.ClientError, requests.exceptions.HTTPError) as e:
+                LoggaTesto("ERRORE ASYNC", {e}, False)
+                if isinstance(e, (asyncio.TimeoutError, requests.exceptions.HTTPError)) and e.response.status in [502, 504]:
+                    break
+                await asyncio.sleep(3)  
+        raise Exception("Max retries reached or skipping retries. Unable to make the request.")
 
 async def CreaListaInput(input_data):
     if input_data.instruction.startswith("http"):
@@ -287,12 +288,13 @@ def DividiInstructionText(input_data):
 
 async def GeneraTestoAsync(url, input_data):
     token = os.getenv('TOKEN')
+    semaphore = asyncio.Semaphore(20) 
     async with aiohttp.ClientSession() as session:
         tasks = []
         ListaInput = await CreaListaInput(input_data)
         for data in ListaInput:
             LoggaTesto("RICHIESTA ASINCRONA", data)
-            tasks.extend([make_request(session, token, data, url, index) for index in range(input_data.NumeroGenerazioni)])
+            tasks.extend([make_request(session, token, data, url, index, semaphore) for index in range(input_data.NumeroGenerazioni)])
             await asyncio.sleep(0.1)
         return await asyncio.gather(*tasks)