Layer Addition Models
Collection
Strategic merging of language models through layer-level architecture optimization.
•
3 items
•
Updated
This is a merge of pre-trained language models created using mergekit.
This model was merged using the passthrough merge method.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
dtype: bfloat16
merge_method: passthrough
slices:
- sources:
- layer_range:
- 0
- 55
model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
- sources:
- layer_range:
- 50
- 60
model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
- sources:
- layer_range:
- 55
- 65
model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
- sources:
- layer_range:
- 60
- 70
model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
- sources:
- layer_range:
- 65
- 75
model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
- sources:
- layer_range:
- 70
- 80
model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF