v0.14.0

What's Changed

• Add stop param by @jackmpcollins and @mnicstruwig in #80

Full Changelog: v0.13.0...v0.14.0

v0.13.0

What's Changed

• Bump jupyter-server from 2.7.2 to 2.11.2 by @dependabot in #75
• Allow setting api_key in OpenaiChatModel by @jackmpcollins in #76

Full Changelog: v0.12.0...v0.13.0

v0.12.0

What's Changed

• Bump aiohttp from 3.8.6 to 3.9.0 by @dependabot in #70
• Add OpenAI seed param for deterministic sampling by @jackmpcollins in #71

Full Changelog: v0.11.1...v0.12.0

v0.11.1

What's Changed

• Fix: OpenAI base url by @jackmpcollins in #67

Full Changelog: v0.11.0...v0.11.1

v0.11.0

What's Changed

• Add support for Azure via OpenaiChatModel by @jackmpcollins in #65

Full Changelog: v0.10.0...v0.11.0

v0.10.0

What's Changed

• Update openai to v1 by @jackmpcollins in #59

Full Changelog: v0.9.1...v0.10.0

v0.9.1

Bound openai version <1.0

Full Changelog: v0.9.0...v0.9.1

v0.9.0

What's Changed

• Add LiteLLM backend by @jackmpcollins in #54

Full Changelog: v0.8.0...v0.9.0

---

Example of LiteLLM backend

from magentic import prompt
from magentic.chat_model.litellm_chat_model import LitellmChatModel


@prompt(
    "Talk to me! ",
    model=LitellmChatModel("ollama/llama2"),
)
def say_hello() -> str:
    ...


say_hello()

See the Backend/LLM Configuration section of the README for how to set the LiteLLM backend as the default.

v0.8.0

What's Changed

• Make backend configurable by @jackmpcollins in #46
• Bump urllib3 from 2.0.6 to 2.0.7 by @dependabot in #47
• Replace black with ruff formatter by @jackmpcollins in #48
• Handle pydantic generic BaseModel in name_type and function schema by @jackmpcollins in #52
• Allow ChatModel to be set with context manager by @jackmpcollins in #53

Full Changelog: v0.7.2...v0.8.0

---

Allow the chat_model/LLM to be set using a context manager. This allows the same prompt-function to easily be reused with different models, and also makes it neater to dynamically set the model.

from magentic import OpenaiChatModel, prompt


@prompt("Say hello")
def say_hello() -> str:
    ...


@prompt(
    "Say hello",
    model=OpenaiChatModel("gpt-4", temperature=1),
)
def say_hello_gpt4() -> str:
    ...


say_hello()  # Uses env vars or default settings

with OpenaiChatModel("gpt-3.5-turbo"):
    say_hello()  # Uses gpt-3.5-turbo due to context manager
    say_hello_gpt4()  # Uses gpt-4 with temperature=1 because explicitly configured

v0.7.2

What's Changed

• Allow setting max_tokens param by @jackmpcollins in #45

Full Changelog: v0.7.1...v0.7.2

---

Allow setting max_tokens param in OpenaiChatModel. The default value for this can also be set using an environment variable MAGENTIC_OPENAI_MAX_TOKENS.

Example

from magentic import prompt
from magentic.chat_model.openai_chat_model import OpenaiChatModel

@prompt("Hello, how are you?", model=OpenaiChatModel(max_tokens=3))
def test() -> str:
    ...

test()
# 'Hello! I'