HabibAhmed commited on
Commit
92be409
·
verified ·
1 Parent(s): b7452f4

Upload 2 files

Browse files

My Evaluation results for Granite3.2-2B-Instruct fine tuned via QLoRA using FP16 LoRA Adapters with base precisions of: NF4, INT8, FP16, FP4, BF16.

Granite3.2-2B-Evaluation_Results-FP16-Coloured.xlsx ADDED
Binary file (10.1 kB). View file
 
Granite3.2-2B-Evaluation_Results-FP16.xlsx ADDED
Binary file (9.97 kB). View file