Skip to main content

Python client library for the Fireworks.ai Generative AI Platform

Project description

Fireworks.ai Python library

Fireworks.ai Python Library provides a convenient API for accessing Fireworks supported LLMs. We are targeting our API to be very similar to OpenAI's API so you can replace OpenAI usage with minimal modifications

Installation

pip install --upgrade fireworks-ai

API definitions

Please check our completion and chat completion API reference for the arguments we support and the meaning of each arguments.

Example code

List

import fireworks.client
fireworks.client.api_key = "your-key"
print(fireworks.client.Models.list())
object='list' data=[Model(id="accounts/fireworks/models/llama-v2-7b", object="model", created=0), ...]

Completion

import fireworks.client
fireworks.client.api_key = "your-key"
completion = fireworks.client.Completion.create("accounts/fireworks/models/llama-v2-7b", "Once upon a time", temperature=0.1, n=2, max_tokens=16)
print(completion)
id='cmpl-988e179fa14fbaebdf17c713' object='text_completion' created=1691602259 model='accounts/fireworks/models/llama-v2-7b' choices=[Choice(text=', there was an emperor who reigned over all the kingdoms of the', index=0, finish_reason='length'), Choice(text=', a boy lived in a small house with his mom and dad. His', index=1, finish_reason='length')]

Streaming completion

import fireworks.client
fireworks.client.api_key = "your-key"
for completion in fireworks.client.Completion.create(
    "accounts/fireworks/models/llama-v2-7b",
    prompt="Once upon a time",
    temperature=0.1,
    n=2,
    max_tokens=16
):
    print(completion)

Async completion

import asyncio
import fireworks.client
fireworks.client.api_key = "your-key"
async def main():
    response = await fireworks.client.Completion.acreate("accounts/fireworks/models/llama-v2-7b", "Once upon a time", echo=True, max_tokens=16)
    print(response.choices[0].text)
asyncio.run(main())

then run the script

$ python test.py
Once upon a time, there used to be a huge mountain that was the most famous mou

ChatCompletion

import fireworks.client
fireworks.client.api_key = "your-key"
completion = fireworks.client.ChatCompletion.create(
    "accounts/fireworks/models/llama-v2-7b-chat",
    messages=[{"role": "user", "content": "Hello there!"}],
    temperature=0.7,
    n=2,
    max_tokens=16
)
print(completion)
id='cmpl-ec241c8f5b8d50bcf792f2df' object='chat.completion' created=1691896960 model='accounts/fireworks/models/llama-v2-7b-chat' choices=[ChatCompletionResponseChoice(index=0, message=ChatMessage(role='assistant', content=" Hello! It's nice to meet you. Is there something I can"), finish_reason='length'), ChatCompletionResponseChoice(index=1, message=ChatMessage(role='assistant', content=" Hello! It's nice to meet you. Is there something I can"), finish_reason='length')] usage=UsageInfo(prompt_tokens=23, total_tokens=55, completion_tokens=32)

Requirements

  • Python 3.7

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

fireworks-ai-0.11.0.tar.gz (62.2 kB view details)

Uploaded Source

Built Distribution

fireworks_ai-0.11.0-py3-none-any.whl (78.0 kB view details)

Uploaded Python 3

File details

Details for the file fireworks-ai-0.11.0.tar.gz.

File metadata

  • Download URL: fireworks-ai-0.11.0.tar.gz
  • Upload date:
  • Size: 62.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.12

File hashes

Hashes for fireworks-ai-0.11.0.tar.gz
Algorithm Hash digest
SHA256 5b7fe1a002fd935f0df36bd280f2104b262533227f7401ef61b6b750041d290b
MD5 bb2c8d0a1c7844a0e1dd36cbbe1c648b
BLAKE2b-256 21326901a9f38370468c85dcfcc7817fda6082957c7140f2c68e74a94107cf6c

See more details on using hashes here.

File details

Details for the file fireworks_ai-0.11.0-py3-none-any.whl.

File metadata

  • Download URL: fireworks_ai-0.11.0-py3-none-any.whl
  • Upload date:
  • Size: 78.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.12

File hashes

Hashes for fireworks_ai-0.11.0-py3-none-any.whl
Algorithm Hash digest
SHA256 a9d5ee9e199a929c9873a3378880daaeeff47880bd53fce793d585effa089198
MD5 709ce98f085de53f9c5bca41b40ec5d5
BLAKE2b-256 b68b43ed09ec1c86aaf4569c02b30965c7567b644e3dd5f23c01a573b39b0b11

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page