L3-DeepDolph-R1-8B / mergekit_config.yml
DopeyGay's picture
Upload folder using huggingface_hub
a2f755d verified
raw
history blame contribute delete
317 Bytes
models:
- model: cognitivecomputations/dolphin-2.9-llama3-8b
- model: deepseek-ai/DeepSeek-R1-Distill-Llama-8B
merge_method: slerp
base_model: deepseek-ai/DeepSeek-R1-Distill-Llama-8B
dtype: bfloat16
parameters:
t: [0, 0.5, 1, 0.5, 0] # V shaped curve: Deepseek for input & output, Dolphin in the middle layers