lm-deluge 0.0.18__tar.gz → 0.0.20__tar.gz

{lm_deluge-0.0.18/src/lm_deluge.egg-info → lm_deluge-0.0.20}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lm_deluge
-Version: 0.0.18
+Version: 0.0.20
 Summary: Python utility for using LLM API models.
 Author-email: Benjamin Anderson <ben@trytaylor.ai>
 Requires-Python: >=3.10

{lm_deluge-0.0.18 → lm_deluge-0.0.20}/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["setuptools", "wheel"]
 
 [project]
 name = "lm_deluge"
-version = "0.0.18"
+version = "0.0.20"
 authors = [{ name = "Benjamin Anderson", email = "ben@trytaylor.ai" }]
 description = "Python utility for using LLM API models."
 readme = "README.md"

{lm_deluge-0.0.18 → lm_deluge-0.0.20}/src/lm_deluge/api_requests/anthropic.py

@@ -185,6 +185,7 @@ class AnthropicRequest(APIRequestBase):
             error_message = text
 
         # handle special kinds of errors. TODO: make sure these are correct for anthropic
+        retry_with_different_model = status_code in [529, 429, 400, 401, 403, 413]
         if is_error and error_message is not None:
             if (
                 "rate limit" in error_message.lower()
@@ -195,6 +196,7 @@ class AnthropicRequest(APIRequestBase):
             if "context length" in error_message:
                 error_message += " (Context length exceeded, set retries to 0.)"
                 self.context.attempts_left = 0
+            retry_with_different_model = True
 
         return APIResponse(
             id=self.context.task_id,
@@ -208,4 +210,5 @@ class AnthropicRequest(APIRequestBase):
             sampling_params=self.context.sampling_params,
             usage=usage,
             raw_response=data,
+            retry_with_different_model=retry_with_different_model,
         )

{lm_deluge-0.0.18 → lm_deluge-0.0.20}/src/lm_deluge/api_requests/openai.py

@@ -159,6 +159,7 @@ class OpenAIRequest(APIRequestBase):
             error_message = text
 
         # handle special kinds of errors
+        retry_with_different_model = status_code in [529, 429, 400, 401, 403, 413]
         if is_error and error_message is not None:
             if "rate limit" in error_message.lower() or status_code == 429:
                 error_message += " (Rate limit error, triggering cooldown.)"
@@ -166,6 +167,7 @@ class OpenAIRequest(APIRequestBase):
             if "context length" in error_message:
                 error_message += " (Context length exceeded, set retries to 0.)"
                 self.context.attempts_left = 0
+            retry_with_different_model = True
 
         return APIResponse(
             id=self.context.task_id,
@@ -181,6 +183,7 @@ class OpenAIRequest(APIRequestBase):
             usage=usage,
             raw_response=data,
             finish_reason=finish_reason,
+            retry_with_different_model=retry_with_different_model,
         )
 
 

{lm_deluge-0.0.18 → lm_deluge-0.0.20}/src/lm_deluge/client.py

@@ -104,6 +104,8 @@ class LLMClient(BaseModel):
                     top_logprobs=data.get("top_logprobs", None),
                 )
             ]
+        if len(data["model_names"]) > 1 and len(data["sampling_params"]) == 1:
+            data["sampling_params"] = data["sampling_params"] * len(data["model_names"])
         return data
 
     @model_validator(mode="after")

{lm_deluge-0.0.18 → lm_deluge-0.0.20/src/lm_deluge.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lm_deluge
-Version: 0.0.18
+Version: 0.0.20
 Summary: Python utility for using LLM API models.
 Author-email: Benjamin Anderson <ben@trytaylor.ai>
 Requires-Python: >=3.10