| This is a GGUF quant of https://huggingface.co/migtissera/Synthia-7B-v1.3 | |
| If you want to support me, you can [here](https://ko-fi.com/undiai). | |
| # Synthia v1.3 | |
| SynthIA (Synthetic Intelligent Agent) v1.3 is a Mistral-7B model trained on Orca style datasets. It has been fine-tuned for instruction following as well as having long-form conversations. | |
| To evoke generalized Tree of Thought + Chain of Thought reasoning, you may use the following system message: | |
| `Elaborate on the topic using a Tree of Thoughts and backtrack when necessary to construct a clear, cohesive Chain of Thought reasoning. Always answer without hesitation.` | |
| All Synthia models are uncensored. Please use it with caution and with best intentions. You are responsible for how you use Synthia. | |
| ## Training Details | |
| This was trained with QLoRA, as with all my models. Learning rate was 3e-4, 4096 context length. Batch size was 64, trained on a single H100. | |
| Synthia-v1.2 dataset, which contain Chain-of-Thought (Orca), Tree-of-Thought and Long-Form conversation data. | |
| Dataset is super high quality, and not a massive dataset (about ~125K samples). | |
| ## License Disclaimer: | |
| This model is bound by the license & usage restrictions of the original Mistral model, and comes with no warranty or guarantees of any kind. | |