Update README.md

The model was fine tuned on dataset
LLAMA2_Legal_Dataset_4k_Instructions_train.jsonl

taken from
https://huggingface.co/datasets/nisaar/LLAMA2_Legal_Dataset_4.4k_Instructions/blob/main/LLAMA2_Legal_Dataset_4k_Instructions_train.jsonl

on 4000 prompt text values using SFTrainer() class.
The results of training attached as reference


README.md

@@ -19,3 +19,5 @@ The following `bitsandbytes` quantization config was used during training:
 
 
 - PEFT 0.6.0.dev0
+
+-