Views
No views yet
| Variant | Use Case | Download |
|---|---|---|
| base | The base model is suitable for applications where model size is not a concern, and high accuracy is required. It can be used for tasks such as text generation, language translation, and text summarization. This model is in FP16 format, providing a good balance between size and performance. | 📥 |
| q2_k | The q2_k variant is ideal for extremely constrained environments where model size is a significant concern, such as embedded systems or low-end mobile devices. Although it is highly compressed, it still maintains a reasonable level of accuracy, making it suitable for simple language tasks. This 2-bit quantized model is a good choice when storage space is limited. | 📥 |
| q3_k_m | The q3_k_m variant is designed for memory-limited devices that require a balance between model size and accuracy. This 3-bit quantized model is very compressed, making it suitable for mid-range mobile devices or systems with limited storage capacity. It is a good choice for applications where a moderate level of accuracy is required, such as language understanding or text classification. | 📥 |