Merge branch 'main' into readme

Portkey-AI · Oct 10, 2023 · 8cbf28e · 8cbf28e
2 parents c72bfcd + 4bf49c3
commit 8cbf28e
Show file tree

Hide file tree

Showing 20 changed files with 1,661 additions and 64 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,14 +1,33 @@
 ### Changelog
+#### Unreleased
 
-All notable changes to this project will be documented in this file. Dates are displayed in UTC.
+#### [v0.1.51](https://github.com/Portkey-AI/portkey-python-sdk/compare/v0.1.51...v0.1.52)
+> 21 September 2023
+- feat: added openschema for all integrations + anyscale integration
 
-Generated by [`auto-changelog`](https://github.com/CookPete/auto-changelog).
+#### [v0.1.50](https://github.com/Portkey-AI/portkey-python-sdk/compare/v0.1.50...v0.1.51)
+> 21 September 2023
+- fix: Issue with the str comparison on the api path
+
+#### [v0.1.50](https://github.com/Portkey-AI/portkey-python-sdk/compare/v0.1.49...v0.1.50)
+> 19 September 2023
+
+- feat: introducing a "Generate" API for prompt-based content generation with saved prompts
+
+<!-- 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)
 

diff --git a/README.md b/README.md
@@ -53,7 +53,8 @@ You can find a comprehensive [list of Portkey features here](#📔-list-of-portk
 Portkey is designed to be flexible. All the features you're familiar with from your LLM provider, like `top_p`, `top_k`, and `temperature`, can be used seamlessly. Check out the [complete list of provider features here](https://github.com/Portkey-AI/portkey-python-sdk/blob/af0814ebf4f1961b5dfed438918fe68b26ef5f1e/portkey/api_resources/utils.py#L137).
 
 **Setting the Prompt Input**:
-You can set the input in two ways. For models like Claude and GPT3, use `prompt` = `(str)`, and for models like GPT3.5 & GPT4, use `messages` = `[array]`.
+This param lets you override any prompt that is passed during the completion call - set a model-specific prompt here to optimise the model performance. You can set the input in two ways. For models like Claude and GPT3, use `prompt` = `(str)`, and for models like GPT3.5 & GPT4, use `messages` = `[array]`.
+
 
 Here's how you can combine everything:
 
@@ -73,7 +74,7 @@ temperature = 1
 messages = [{"role": "user", "content": "Who are you?"}]
 
 # Construct LLM
-llm = LLMOptions(provider=provider, virtual_key=virtual_key, trace_id=trace_id, model=model, temperature=temperature, messages=messages)
+llm = LLMOptions(provider=provider, virtual_key=virtual_key, trace_id=trace_id, model=model, temperature=temperature)
 ```
 
 ### **Steo 3️⃣ : Construct the Portkey Client**
@@ -101,8 +102,12 @@ The Portkey client can do `ChatCompletions` and `Completions`.
 Since our LLM is GPT4, we will use ChatCompletions:
 
 ```py
-response = portkey.ChatCompletions.create()
-
+response = portkey.ChatCompletions.create(
+    messages=[{
+      "role": "user",
+      "content": "Who are you ?"
+    }]
+)
 print(response.choices[0].message)
 ```
 

diff --git a/portkey/__init__.py b/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",

diff --git a/portkey/api_resources/__init__.py b/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",
 ]
diff --git a/portkey/api_resources/apis.py b/portkey/api_resources/apis.py
@@ -1,4 +1,4 @@
-from typing import Optional, Union, overload, Literal, List
+from typing import Optional, Union, overload, Literal, List, Mapping, Any
 from portkey.api_resources.base_client import APIClient
 from .utils import (
     Modes,
@@ -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 +244,29 @@ 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,
+        variables: Optional[Mapping[str, Any]] = None,
+    ) -> Union[GenericResponse, Stream[GenericResponse]]:
+        if config is None:
+            config = retrieve_config()
+        _client = APIClient(api_key=config.api_key, base_url=config.base_url)
+        body = {"variables": variables}
+        return cls(_client)._post(
+            f"/v1/prompts/{prompt_id}/generate",
+            body=body,
+            mode=None,
+            params=None,
+            cast_to=GenericResponse,
+            stream_cls=Stream[GenericResponse],
+            stream=False,
+        )