metadata
license: apache-2.0
language:
- ko
base_model:
- trillionlabs/Tri-7B
This is a 4-bit quantized version of trillionlabs/Tri-7B using Intel AutoRound.
Model Details
Base Model: trillionlabs/Tri-7B
Quantization Method: Intel AutoRound (Best Configuration)
Precision: 4-bit
Group Size: 128
Symmetric: True
Calibration Samples: 512
Tuning Iterations: 1000
Requirements
pip install transformers accelerate auto-round
License
This model inherits the Apache 2.0 license from the original Tri-7B model.