emirhanboge/mnli_llama1b_modified

MNLI (Modified for LLaMA 1B)

This dataset is a modified version of the Multi-Genre Natural Language Inference (MNLI) dataset.

Modifications:

• Labels were originally "entailment", "contradiction", and "neutral", now converted to integers (0, 1, 2).
• The premise and hypothesis columns were renamed to text1 and text2.
• Each example includes a task prefix: Task: MNLI | Premise: ... Hypothesis: ...
• Tokenized using the LLaMA-1B tokenizer.
• Maximum sequence length is 128 tokens.

Dataset Usage:

from datasets import load_dataset
dataset = load_dataset("emirhanboge/mnli_llama1b_modified")