feat: introducing a "Generate" API for prompt-based content generation with saved prompts

CHANGELOG.md

@@ -1,14 +1,22 @@
 ### Changelog
+#### Unreleased
 
-All notable changes to this project will be documented in this file. Dates are displayed in UTC.
+- feat: introducing a "Generate" API for prompt-based content generation with saved prompts
 
-Generated by [`auto-changelog`](https://github.com/CookPete/auto-changelog).
+<!-- auto-changelog-below -->
+
+#### [v0.1.49](https://github.com/Portkey-AI/portkey-python-sdk/compare/v0.1.48...v0.1.49)
+
+> 16 September 2023
+
+- fix: Fixing the issue with the capturing of env variables [`#13`](https://github.com/Portkey-AI/portkey-python-sdk/pull/13)
 
 #### [v0.1.48](https://github.com/Portkey-AI/portkey-python-sdk/compare/v0.1.45...v0.1.48)
 
 > 14 September 2023
 
 - fix: cache age type fix [`#11`](https://github.com/Portkey-AI/portkey-python-sdk/pull/11)
+- docs: updated change log [`#10`](https://github.com/Portkey-AI/portkey-python-sdk/pull/10)
 
 #### [v0.1.45](https://github.com/Portkey-AI/portkey-python-sdk/compare/v0.1.44...v0.1.45)
 

README.md

@@ -68,7 +68,7 @@ All of your LLM provider features work as they would in their own SDKs. For exam
 * Set `messages` = `[array]` to set input for models like GPT3.5 & GPT4
 
 Let's see it in action.
-```
+```python
 from portkey import LLMOptions
 llm = LLMOptions(
   provider="openai", 

portkey/__init__.py

@@ -19,6 +19,7 @@
     ChatCompletionChunk,
     TextCompletion,
     TextCompletionChunk,
+    Generations,
 )
 from portkey.version import VERSION
 from portkey.api_resources.global_constants import (
@@ -50,6 +51,7 @@
     "ChatCompletionChunk",
     "TextCompletion",
     "TextCompletionChunk",
+    "Generations",
     "Config",
     "api_key",
     "base_url",

portkey/api_resources/__init__.py

@@ -1,5 +1,5 @@
 """"""
-from .apis import ChatCompletions, Completions
+from .apis import ChatCompletions, Completions, Generations
 from .utils import (
     Modes,
     ModesLiteral,
@@ -41,4 +41,5 @@
     "ChatCompletionChunk",
     "TextCompletion",
     "TextCompletionChunk",
+    "Generations",
 ]

portkey/api_resources/apis.py

@@ -10,7 +10,7 @@
     ChatCompletion,
     TextCompletion,
     TextCompletionChunk,
-    ApiType,
+    GenericResponse,
 )
 
 from .streaming import Stream
@@ -49,7 +49,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> Stream[TextCompletionChunk]:
         ...
 
@@ -65,7 +65,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> TextCompletion:
         ...
 
@@ -81,7 +81,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> Union[TextCompletion, Stream[TextCompletionChunk]]:
         ...
 
@@ -96,7 +96,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> Union[TextCompletion, Stream[TextCompletionChunk]]:
         if config is None:
             config = retrieve_config()
@@ -107,7 +107,7 @@ def create(
             max_tokens=max_tokens,
             top_k=top_k,
             top_p=top_p,
-            **kwargs
+            **kwargs,
         )
         if config.mode == Modes.SINGLE.value:
             return cls(_client)._post(
@@ -118,7 +118,6 @@ def create(
                 cast_to=TextCompletion,
                 stream_cls=Stream[TextCompletionChunk],
                 stream=stream,
-                _type=ApiType.COMPLETIONS,
             )
         if config.mode == Modes.FALLBACK.value:
             return cls(_client)._post(
@@ -129,7 +128,6 @@ def create(
                 cast_to=TextCompletion,
                 stream_cls=Stream[TextCompletionChunk],
                 stream=stream,
-                _type=ApiType.COMPLETIONS,
             )
         if config.mode == Modes.AB_TEST.value:
             return cls(_client)._post(
@@ -140,7 +138,6 @@ def create(
                 cast_to=TextCompletion,
                 stream_cls=Stream[TextCompletionChunk],
                 stream=stream,
-                _type=ApiType.COMPLETIONS,
             )
         raise NotImplementedError("Mode not implemented.")
 
@@ -158,7 +155,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> Stream[ChatCompletionChunk]:
         ...
 
@@ -174,7 +171,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> ChatCompletion:
         ...
 
@@ -190,7 +187,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> Union[ChatCompletion, Stream[ChatCompletionChunk]]:
         ...
 
@@ -205,7 +202,7 @@ def create(
         max_tokens: Optional[int] = None,
         top_k: Optional[int] = None,
         top_p: Optional[float] = None,
-        **kwargs
+        **kwargs,
     ) -> Union[ChatCompletion, Stream[ChatCompletionChunk]]:
         if config is None:
             config = retrieve_config()
@@ -216,7 +213,7 @@ def create(
             max_tokens=max_tokens,
             top_k=top_k,
             top_p=top_p,
-            **kwargs
+            **kwargs,
         )
         if config.mode == Modes.SINGLE.value:
             return cls(_client)._post(
@@ -227,7 +224,6 @@ def create(
                 cast_to=ChatCompletion,
                 stream_cls=Stream[ChatCompletionChunk],
                 stream=stream,
-                _type=ApiType.CHAT_COMPLETION,
             )
         if config.mode == Modes.FALLBACK.value:
             return cls(_client)._post(
@@ -238,7 +234,6 @@ def create(
                 cast_to=ChatCompletion,
                 stream_cls=Stream[ChatCompletionChunk],
                 stream=stream,
-                _type=ApiType.CHAT_COMPLETION,
             )
         if config.mode == Modes.AB_TEST.value:
             return cls(_client)._post(
@@ -249,6 +244,24 @@ def create(
                 cast_to=ChatCompletion,
                 stream_cls=Stream[ChatCompletionChunk],
                 stream=stream,
-                _type=ApiType.CHAT_COMPLETION,
             )
         raise NotImplementedError("Mode not implemented.")
+
+
+class Generations(APIResource):
+    @classmethod
+    def create(
+        cls, *, prompt_id: str, config: Optional[Config] = None, **kwargs
+    ) -> Union[GenericResponse, Stream[GenericResponse]]:
+        if config is None:
+            config = retrieve_config()
+        _client = APIClient(api_key=config.api_key, base_url=config.base_url)
+        return cls(_client)._post(
+            f"/v1/prompts/{prompt_id}/generate",
+            body=config.llms,
+            mode=None,
+            params=None,
+            cast_to=GenericResponse,
+            stream_cls=Stream[GenericResponse],
+            stream=False,
+        )