Inspired by https://huggingface.co/da-fr/Mistral-NeMo-Minitron-8B-ARChitects-Full-bnb-4bit, finetuned very similarly with the key exception that the ARC public evaluation set was excluded from the training data.

Downloads last month
11
Safetensors
Model size
3.79B params
Tensor type
F32
BF16
U8
Inference Providers NEW
This model isn't deployed by any Inference Provider. 馃檵 Ask for provider support

Model tree for paranke/Mistral-NeMo-Minitron-8B-arc-training

Quantized
(24)
this model
Adapters
1 model