metadata
license: apache-2.0
base_model:
- ibm-granite/granite-4.0-micro
Neuron 3B v0
This is my finetune of granite 4.0 micro, I have optimized it for more natural responses while staying general purpose. I have also trained it on tool-use. This is still very experimental and may be buggy in it's current state, an updated model will be released soon.
Training data
A good portion of the training data is from the Hermes3 dataset, Google's mobile tool use, and the Dolphin R1 (non-reasoning) data. I also used some modified and fully custom data as well.