Hermes-2-Pro-Mistral-10.7B-bpw4.8 / mergekit_config.yml
blockblockblock's picture
Upload folder using huggingface_hub
7528fe3 verified
raw
history blame contribute delete
475 Bytes
dtype: bfloat16
merge_method: passthrough
# Depth UpScaled (DUS) version of Mistral-7B-Instruct-v0.2
# where m = 8 (The number of layers to remove from the model)
# s = 48 (The number of layers the model will have after the DUS)
slices:
- sources:
- layer_range: [0, 24]
model: /Users/jsarnecki/opt/workspace/NousResearch/Hermes-2-Pro-Mistral-7B
- sources:
- layer_range: [8, 32]
model: /Users/jsarnecki/opt/workspace/NousResearch/Hermes-2-Pro-Mistral-7B