DeepSeek-7B-1M / mergekit_config.yml
yasserrmd's picture
Upload folder using huggingface_hub
4a00aad verified
raw
history blame contribute delete
203 Bytes
models:
- model: Qwen/Qwen2.5-7B-Instruct-1M
parameters:
weight: 0.5
- model: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B
parameters:
weight: 0.5
merge_method: linear
dtype: float16