llama-3-merge-virt-req-8B
This is a merge of pre-trained language models created using mergekit.
In light testing (temperature=1.0, minP=0.02, provisional Llama 3 Instruct prompts), there appear to be occasional lapses in reasoning, but the model appears to follow formatting well overall.
Built with Meta Llama 3.
Merge Details
Merge Method
This model was merged using the SLERP merge method.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
slices:
- sources:
- model: grimjim/llama-3-merge-pp-instruct-8B
layer_range: [0,32]
- model: openlynn/Llama-3-Soliloquy-8B-v2
layer_range: [0,32]
merge_method: slerp
base_model: grimjim/llama-3-merge-pp-instruct-8B
parameters:
t:
- value: 0.5
dtype: bfloat16