llama3-5.4b-instruct-unhealed

merged

This is a "merge" of pre-trained language models created using mergekit. It is a prune of Meta-Llama-3-8B-Instruct down to 20 layers, or about 5.4B parameter. Mostly, this is a test of pruning & healing an instruct-tuned model. THIS MODEL HAS NOT BEEN HEALED. It is presently unusable. The healed version will be in a different repository. This size should allow bf16 inference on 24GB VRAM, Q8 or Q6 inference on 6GB VRAM, Q5 inference on 4GB VRAM, and fine-tuning ... well, with less VRAM than an 8B model.

Merge Details

Merge Method

This model was merged using the passthrough merge method.

Models Merged

The following models were included in the merge:

• meta-llama/Meta-Llama-3-8B-Instruct

Configuration

The following YAML configuration was used to produce this model:

dtype: bfloat16
merge_method: passthrough
slices:
- sources:
  - layer_range: [0, 16]
    model: meta-llama/Meta-Llama-3-8B-Instruct
- sources:
  - layer_range: [20, 21]
    model: meta-llama/Meta-Llama-3-8B-Instruct
- sources:
  - layer_range: [29, 32]
    model: meta-llama/Meta-Llama-3-8B-Instruct