aixsatoshi's picture
Update README.md
18d3e6c verified
---
license: apache-2.0
---
Meta-Llama-3-70bのセルフマージにより120Bにパラメーター数を拡大したモデルの高性能化が報告されています
今回高品質な日本語LLMである、[karakuri-ai/karakuri-lm-8x7b-chat-v0.1](https://huggingface.co/karakuri-ai/karakuri-lm-8x7b-chat-v0.1)の精度を更に高めるために、"num_hidden_layers": 32、から、56への自己拡張マージを行いました。
マージに利用したスライスのインターバルから本モデル(Ex-karakuri-8x12B-chat-v2)が非マージ部分4層、[Ex-karakuri-8x12B-chat-v1](https://huggingface.co/aixsatoshi/Ex-karakuri-8x12B-chat-v1)は8層に設定しています
It was inspired by large merges like:
- [Meta-Llama-3-120B-Instruct](https://huggingface.co/mlabonne/Meta-Llama-3-120B-Instruct/)
- [alpindale/goliath-120b](https://huggingface.co/alpindale/goliath-120b)
- [nsfwthrowitaway69/Venus-120b-v1.0](https://huggingface.co/nsfwthrowitaway69/Venus-120b-v1.0)
- [cognitivecomputations/MegaDolphin-120b](https://huggingface.co/cognitivecomputations/MegaDolphin-120b)
- [wolfram/miquliz-120b-v2.0](https://huggingface.co/wolfram/miquliz-120b-v2.0).
```
slices:
- sources:
- layer_range: [0, 4]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [2, 6]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [4, 8]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [6, 10]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [8, 12]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [10, 14]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [12, 16]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [14, 18]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [16, 20]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [18, 22]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [20, 24]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [22, 26]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [24, 28]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [26, 30]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
- sources:
- layer_range: [28, 32]
model: karakuri-ai/karakuri-lm-8x7b-chat-v0.1
merge_method: passthrough
dtype: bfloat16
```