Add Support for Auto Retrival Evaluation of different RAG techniques

### Issue: Add `.evaluate` and `.generate_eval_dataset` Methods for RAG Pipelines

**Description:**
Enhance the RAG framework by implementing `.evaluate` and `.generate_eval_dataset` methods to enable easy and automated evaluation of different RAG pipelines.

**Requirements:**
1. **`.evaluate` Method:**
- Allow users to benchmark retrieval performance using standard metrics (e.g., Precision, Recall, F1).
- Support configuration for top-k retrieval and similarity scoring.

2. **`.generate_eval_dataset` Method:**
- Automatically create datasets for evaluation with annotated ground truth queries and responses.
- Include support for custom datasets.

3. **Integration:**
- Seamlessly integrate with existing RAG pipelines.

**Impact:**
Simplifies evaluation workflows and helps compare RAG pipeline performance effectively.

**Priority:** Medium