Instructions to use fumiama/ninus with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use fumiama/ninus with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("meta-llama/Meta-Llama-3.1-8B-Instruct") model = PeftModel.from_pretrained(base_model, "fumiama/ninus") - Transformers
How to use fumiama/ninus with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("fumiama/ninus", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Unsloth Desktop
Commit History
-
feat: change to new chat format
New chat format: ``` Below is an instruction that describes a chat task, paired with an JSON input that provides further context. The label `human` is from user and `gpt` is your response. Write a JSON response that appropriately completes the request. ### Instruction: 在QQ群与用户聊天,按自己的心情回复,不必完全跟随用户发送的内容。如果你愿意,也可以开启新话题。 ### Input: [{"from": "human", "value": "Steam昨晚\n真因为丝之歌挂了吗\n不应该吧"}] ### Response: {"from": "gpt", "value": "这玩意儿的开发者都没玩过吗"} ```d24c489
-
feat: add new version
93cb60a
-
Upload README.md
17abf29 -
fix: adjust format
812de39 -
fix: try to modify base model
739dd99 -
feat: add more files
ecc9aed -
Add 2 files
b6659b9 -
initial commit
b5c0e47