Merge branch 'main' into watch

aider/sendchat.py

@@ -1,5 +1,6 @@
 import hashlib
 import json
+import time
 
 import backoff
 
@@ -18,26 +19,32 @@ RETRY_TIMEOUT = 60
 
 def retry_exceptions():
     import httpx
+    import openai
 
     return (
         # httpx
         httpx.ConnectError,
         httpx.RemoteProtocolError,
         httpx.ReadTimeout,
-        # litellm
-        litellm.exceptions.BadRequestError,
-        litellm.exceptions.AuthenticationError,
-        litellm.exceptions.PermissionDeniedError,
-        litellm.exceptions.NotFoundError,
-        litellm.exceptions.UnprocessableEntityError,
-        litellm.exceptions.RateLimitError,
-        litellm.exceptions.InternalServerError,
-        litellm.exceptions.ContextWindowExceededError,
-        litellm.exceptions.ContentPolicyViolationError,
-        litellm.exceptions.APIConnectionError,
-        litellm.exceptions.APIError,
-        litellm.exceptions.ServiceUnavailableError,
-        litellm.exceptions.Timeout,
+        #
+        # litellm exceptions inherit from openai exceptions
+        # https://docs.litellm.ai/docs/exception_mapping
+        #
+        # openai.BadRequestError,
+        # litellm.ContextWindowExceededError,
+        # litellm.ContentPolicyViolationError,
+        #
+        # openai.AuthenticationError,
+        # openai.PermissionDeniedError,
+        # openai.NotFoundError,
+        #
+        openai.APITimeoutError,
+        openai.UnprocessableEntityError,
+        openai.RateLimitError,
+        openai.APIConnectionError,
+        openai.APIError,
+        openai.APIStatusError,
+        openai.InternalServerError,
     )
 
 
@@ -64,8 +71,6 @@ def send_completion(
     temperature=0,
     extra_params=None,
 ):
-    from aider.llm import litellm
-
     kwargs = dict(
         model=model_name,
         messages=messages,
@@ -98,18 +103,27 @@ def send_completion(
     return hash_object, res
 
 
-@lazy_litellm_retry_decorator
 def simple_send_with_retries(model_name, messages, extra_params=None):
-    try:
-        kwargs = {
-            "model_name": model_name,
-            "messages": messages,
-            "functions": None,
-            "stream": False,
-            "extra_params": extra_params,
-        }
+    retry_delay = 0.125
+    while True:
+        try:
+            kwargs = {
+                "model_name": model_name,
+                "messages": messages,
+                "functions": None,
+                "stream": False,
+                "extra_params": extra_params,
+            }
 
-        _hash, response = send_completion(**kwargs)
-        return response.choices[0].message.content
-    except (AttributeError, litellm.exceptions.BadRequestError):
-        return
+            _hash, response = send_completion(**kwargs)
+            return response.choices[0].message.content
+        except retry_exceptions() as err:
+            print(str(err))
+            retry_delay *= 2
+            if retry_delay > RETRY_TIMEOUT:
+                break
+            print(f"Retrying in {retry_delay:.1f} seconds...")
+            time.sleep(retry_delay)
+            continue
+        except AttributeError:
+            return

aider/website/docs/more/infinite-output.md

@@ -55,6 +55,7 @@ model_list = "\n".join(f"- {model}" for model in sorted(prefill_models))
 
 cog.out(model_list)
 ]]]-->
+- anthropic.claude-3-5-sonnet-20241022-v2:0
 - anthropic/claude-3-5-sonnet-20241022
 - claude-3-5-sonnet-20240620
 - claude-3-5-sonnet-20241022
@@ -65,6 +66,7 @@ cog.out(model_list)
 - codestral/codestral-latest
 - deepseek-chat
 - deepseek-coder
+- eu.anthropic.claude-3-5-sonnet-20241022-v2:0
 - mistral/codestral-2405
 - mistral/codestral-latest
 - mistral/codestral-mamba-latest
@@ -85,6 +87,7 @@ cog.out(model_list)
 - mistral/open-mixtral-8x7b
 - mistral/pixtral-12b-2409
 - openrouter/anthropic/claude-3.5-sonnet
+- us.anthropic.claude-3-5-sonnet-20241022-v2:0
 - vertex_ai/claude-3-5-sonnet-v2@20241022
 - vertex_ai/claude-3-5-sonnet@20240620
 - vertex_ai/claude-3-haiku@20240307

tests/basic/test_sendchat.py

@@ -4,7 +4,7 @@ from unittest.mock import MagicMock, patch
 import httpx
 
 from aider.llm import litellm
-from aider.sendchat import simple_send_with_retries
+from aider.sendchat import retry_exceptions, simple_send_with_retries
 
 
 class PrintCalled(Exception):
@@ -12,6 +12,10 @@ class PrintCalled(Exception):
 
 
 class TestSendChat(unittest.TestCase):
+    def test_retry_exceptions(self):
+        """Test that retry_exceptions() can be called without raising errors"""
+        retry_exceptions()  # Should not raise any exceptions
+
     @patch("litellm.completion")
     @patch("builtins.print")
     def test_simple_send_with_retries_rate_limit_error(self, mock_print, mock_completion):
@@ -31,7 +35,7 @@ class TestSendChat(unittest.TestCase):
 
         # Call the simple_send_with_retries method
         simple_send_with_retries("model", ["message"])
-        mock_print.assert_called_once()
+        assert mock_print.call_count == 2
 
     @patch("litellm.completion")
     @patch("builtins.print")
@@ -44,4 +48,4 @@ class TestSendChat(unittest.TestCase):
 
         # Call the simple_send_with_retries method
         simple_send_with_retries("model", ["message"])
-        mock_print.assert_called_once()
+        assert mock_print.call_count == 2