|
|
--- |
|
|
license: apache-2.0 |
|
|
tags: |
|
|
- biology |
|
|
--- |
|
|
## Genos-10B-v2 |
|
|
|
|
|
This repository hosts the model weights for **Genos-10B-v2**. For instructions and details, please refer to the [Genos GitHub](https://github.com/zhejianglab/Genos). |
|
|
|
|
|
The Genos collection of models are foundation models for the human genome, achieving million-base-pair context modeling and single-nucleotide resolution learning capability for the human genome sequence. Genos adheres to the open science principles of collaboration, sharing, and co-construction, with a strong commitment to supporting the global genomics research community. To this end, we have openly released three models: |
|
|
|
|
|
- [Genos-1.2B](https://huggingface.co/ZhejiangLab/Genos-1.2B): a 1.2 billion parameter model designed for high-efficiency genome sequence analysis. |
|
|
- [Genos-10B](https://huggingface.co/ZhejiangLab/Genos-10B): a larger model with 10 billion parameters, offering enhanced performance for complex genomic tasks. |
|
|
- [Genos-10B-v2](https://huggingface.co/ZhejiangLab/Genos-10B-v2): an enhanced version trained with additional human and non-human genomes to expand evolutionary context and sequence diversity. |
|
|
|
|
|
We also provide checkpoints trained under the Megatron-LM framework: |
|
|
|
|
|
- [Genos-Megatron-1.2B](https://huggingface.co/ZhejiangLab/Genos-Megatron-1.2B) |
|
|
- [Genos-Megatron-10B](https://huggingface.co/ZhejiangLab/Genos-Megatron-10B) |
|
|
- [Genos-Megatron-10B-v2](https://huggingface.co/ZhejiangLab/Genos-Megatron-10B-v2) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
