Yeah, another L3 model. Perform really well from my testing and scores pretty high on Chai leaderboard (known as slerp8bv2 there). Feedback would be really appreciated.

I'm using default Llama 3 Instruct context template and instruct mode for my testing.

Temp: 0.9
Min P: 0.1
Top K: 40

Others are off, ajust as you need.

Downloads last month: 11

Safetensors

Model size

8.03B params

Tensor type

BF16

Inference Providers NEW

Text Generation

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for R136a1/Bungo-L3-8B

Merges

1 model

Quantizations

4 models