Zephyr is a series of language models that are trained to act as helpful assistants. Zephyr-7B-β is the second model in the series, and is a fine-tuned version of mistralai/Mistral-7B-v0.1 that was trained on on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO). We found that removing the in-built alignment of these datasets boosted performance on MT Bench and made the model more helpful. However, this means that model is likely to generate problematic text when prompted to do so. You can find more details in the technical report.

Get started with huggingfaceh4_zephyr-7b-beta

The model used in the example below is the Zephyr-7b-beta model, with 7b parameters, which is a chat model.

API

1. Start Ollama server (Run ollama serve)
2. Run the model

  curl -X POST http://localhost:11434/api/generate -d '{
    "model": "mattw/huggingfaceh4_zephyr-7b-beta:latest",
    "prompt":"Why is the sky blue?"
   }'

CLI

1. Install Ollama
2. Open the terminal and run ollama run mattw/huggingfaceh4_zephyr-7b-beta:latest

Note: The ollama run command performs an ollama pull if the model is not already downloaded. To download the model without running it, use ollama pull mattw/huggingfaceh4_zephyr-7b-beta:latest

Memory requirements

• 7b models generally require at least 8GB of RAM

If you run into issues with higher quantization levels, try using the q4 model or shut down any other programs that are using a lot of memory.

Model variants

By default, Ollama uses 4-bit quantization. To try other quantization levels, please try the other tags. The number after the q represents the number of bits used for quantization (i.e. q4 means 4-bit quantization). The higher the number, the more accurate the model is, but the slower it runs, and the more memory it requires.

Model source

Zephyr-7b-beta source on Ollama

7b parameters source: HuggingFace