Using Zephyr 7B Beta on Replicate API

Quick Start

1. 1
   Create an account at Replicate API and generate an API key.
2. 2
   Use the Replicate API SDK or REST API to call zephyr-7b-beta — see the documentation for request format.
3. 3
   You'll be billed $0.05/1M input, $0.25/1M output tokens. See full pricing.

Code Examples

import replicate

# reads REPLICATE_API_TOKEN from env
# zephyr-7b-beta format: "owner/model-name" (latest version) or "owner/model-name:version-hash"
output = replicate.run(
    "zephyr-7b-beta",
    input={"prompt": "Hello"}
)
# Output is a list or generator depending on the model
print("".join(output))

About Replicate API

Replicate offers a cloud-based AI platform that simplifies the deployment and integration of machine learning models. The platform provides an extensive library of open-source models that users can run with minimal coding, enabling easy access to advanced AI functionalities such as text generation, image creation, and video production. With automatic API generation, users can effortlessly deploy custom models on a large GPU cluster. The platform also supports the "Cog" tool, which packages models into production-ready containers, streamlining the management and scaling of AI applications. The platform's scalability is a key feature, automatically adjusting resources based on demand to ensure optimal performance during peak usage times. Users benefit from a cost-effective pricing model, paying only for the active time their code runs. Replicate fosters collaboration by allowing users to share their models publicly or keep them private, promoting innovation and knowledge sharing within the developer community. The platform's focus on accessibility and ease of use makes it an ideal solution for developers looking to integrate AI into their projects without the complexities typically associated with machine learning.

Replicate is a cloud-based platform that enables users to run machine learning models easily and efficiently. The company specializes in providing a streamlined environment for deploying, scaling, and managing AI models, making advanced machine learning capabilities accessible to developers and researchers. Replicate's platform allows users to run a wide variety of pre-trained models or deploy their own custom models, facilitating rapid experimentation and development in AI projects. The service is designed to handle the complexities of infrastructure management, allowing users to focus on their core AI tasks rather than worrying about the underlying technical details of model deployment and scaling. By offering a user-friendly interface and robust cloud infrastructure, Replicate aims to democratize access to cutting-edge AI technologies, enabling both individuals and organizations to leverage powerful machine learning models without the need for extensive in-house resources or expertise.

Pricing on Replicate API

Capabilities

No model capability flags are currently sourced.

About Zephyr 7B Beta

Zephyr 7B Beta is a 7-billion parameter large language model, fine-tuned from the Mistral-7B-v0.1 model. It is tailored to serve as an effective virtual assistant, performing well in generating human-like responses. The model's training involved Direct Preference Optimization (DPO) on a combination of publicly available and synthetic datasets, achieving strong performance on benchmarks like MT-Bench and AlpacaEval, especially for conversational tasks. However, its complexity falls short when compared to proprietary models, especially in tasks involving coding and mathematics. A notable limitation is its insufficient alignment to human safety preferences and the absence of in-the-loop filtering to prevent problematic outputs. Zephyr 7B Beta is English-based and carries an MIT license.