| license: apache-2.0 | |
| tags: | |
| - merge | |
| - mergekit | |
| - TencentARC/LLaMA-Pro-8B-Instruct | |
| - arcee-ai/Patent-Instruct-Extended | |
| # Patent-Instruct-Pro | |
| Patent-Instruct-Pro is a merge of the following models using [mergekit](https://github.com/cg123/mergekit): | |
| * [TencentARC/LLaMA-Pro-8B-Instruct](https://huggingface.co/TencentARC/LLaMA-Pro-8B-Instruct) | |
| * [arcee-ai/Patent-Instruct-Extended](https://huggingface.co/arcee-ai/Patent-Instruct-Extended) | |
| ## 🧩 Configuration | |
| ```yaml | |
| slices: | |
| - sources: | |
| - model: TencentARC/LLaMA-Pro-8B-Instruct | |
| layer_range: [0, 40] | |
| - model: arcee-ai/Patent-Instruct-Extended | |
| layer_range: [0, 40] | |
| merge_method: slerp | |
| base_model: TencentARC/LLaMA-Pro-8B-Instruct | |
| parameters: | |
| t: | |
| - filter: self_attn | |
| value: [0, 0.5, 0.3, 0.7, 1] | |
| - filter: mlp | |
| value: [1, 0.5, 0.7, 0.3, 0] | |
| - value: 0.5 | |
| dtype: bfloat16 | |
| ``` |