LLM Reference

Cerebras GPT 2.7B

About

The Cerebras GPT 2.7B is a large language AI model developed by Cerebras Systems, featuring a GPT-3 style architecture with 32 layers and 2.7 billion parameters. It is designed to excel in text completion, summarization, and open-ended dialogue tasks. The model uses Cerebras' weight streaming technology for efficient training and has demonstrated superior performance on benchmarks compared to similar-sized models. Despite its strengths, it is best suited for English-only tasks and may exhibit biases from its training data. It supports long sequence lengths, allowing it to handle extensive text with ease. The model is openly available under the Apache 2.0 license.

Capabilities

MultimodalFunction CallingTool UseJSON Mode

Specifications

ArchitectureDecoder Only
Specializationgeneral