Mohamed682004/

qwen1.5-adversarial-eval

	Adversarial Evaluation of Qwen1.5-1.8B-Chat-GGUF




	
	Model Overview

This dataset contains failure points identified during an adversarial evaluation of the Qwen/Qwen1.5-1.8B-Chat-GGUF model.

Engine: llama-cpp-python Quantization: Q4_K_M Context Window: 2048 tokens

Logical Reasoning: The model struggles with syllogistic reasoning (e.g., 'all bloops are blips') and classic mathematical word problems, often… See the full description on the dataset page: https://huggingface.co/datasets/Mohamed682004/qwen1.5-adversarial-eval.