File size: 664 Bytes
3d849b5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
models:
- model: tokyotech-llm/Swallow-7b-instruct-hf
# no parameters necessary for base model
- model: codellama/CodeLlama-7b-Python-hf # follow user intent
parameters:
density: 1
weight:
- filter: mlp.down_proj
value: [0.3, 0.25, 0.25, 0.15, 0.1]
- filter: mlp.gate_proj
value: [0.7, 0.25, 0.5, 0.45, 0.4]
- filter: mlp.up_proj
value: [0.7, 0.25, 0.5, 0.45, 0.4]
- filter: self_attn
value: [0.7, 0.25, 0.5, 0.45, 0.4]
- value: 0 # fallback for rest of tensors.
merge_method: dare_ties
base_model: tokyotech-llm/Swallow-7b-instruct-hf
dtype: bfloat16
tokenizer_source: union
|