This is a Lite series model based on the Mistral architecture, comprising approximately 157 million parameters.

70 Pulls Updated 2 months ago

Readme

Lite-Mistral-150M-Instruct

This is a Lite series model based on the Mistral architecture, comprising approximately 157 million parameters.

Lite-Mistral-150M-Instruct-GGUF

Lite-Mistral-150M-Instruct

Lite-Mistral-150M-GGUF

Lite-Mistral-150M

Use Cases

This lightweight model is designed for applications where computational resources are limited.
Due to its smaller size, this model may be more amenable to fine-tuning on specific tasks with limited computational resources.

Benchmarks

Benchmark Lite-Mistral-150M Lite-Mistral-150M-Instruct
hellaswag (5-shot) 27.14 27.14
openbookqa (5-shot) 13.20 14.00
piqa (5-shot) 58.27 59.03
hellaswag (0-shot) 26.78 26.78
openbookqa (0-shot) 11.80 13.80
piqa (0-shot) 57.73 58.60

Chat Template

This model uses the following chat template:

<|system|>
{{system}}<|end|>
<|user|>
{{user}}<|end|>
<|assistant|>
{{assistant}}<|end|>

Risk Disclaimer

By using this model, you acknowledge that you understand and assume the risks associated with its use. You are solely responsible for ensuring compliance with all applicable laws and regulations. We disclaim any liability for problems arising from the use of this open-source model, including but not limited to direct, indirect, incidental, consequential, or punitive damages. We make no warranties, express or implied, regarding the model’s performance, accuracy, or fitness for a particular purpose. Your use of this model is at your own risk, and you agree to hold harmless and indemnify us, our affiliates, and our contributors from any claims, damages, or expenses arising from your use of the model.

Hugging Face Page