Flynntaggart26/qwen2.5-scholarship-rag-adapter
Text Generation • Updated
question stringclasses 2
values | context stringclasses 2
values | ground_truth stringclasses 2
values | category stringclasses 2
values |
|---|---|---|---|
What is Stanford University's financial aid policy for international students? | Stanford University provides Need-Blind admission for all international undergraduate applicants, fully covering 100% of demonstrated financial need. | Stanford offers need-blind admissions for international students and covers 100% of their demonstrated financial need. | factual_extraction |
How does MIT compare to Brown in terms of need-blind aid for internationals? | MIT is need-blind for all undergraduate applicants regardless of citizenship. Brown University transitioned to need-blind admissions for international students starting with the Class of 2029. | Both MIT and Brown offer need-blind admissions for international undergraduate applicants. | comparative_reasoning |
Learn-RAG Evaluation & Benchmarking QA Dataset 🧪
Dataset Summary
This synthetic evaluation dataset is constructed specifically for benchmarking Retrieval-Augmented Generation (RAG) systems, vector similarity search algorithms, and context grounding accuracy. It contains structured Question-Context-Answer triplets curated to evaluate Faithfulness, Answer Relevance, and Context Recall metrics.
Dataset Structure
The dataset contains the following schema:
| Column | Type | Description |
|---|---|---|
question |
string | The query posed to the RAG pipeline. |
context |
string | The retrieved document chunk required to answer the query. |
ground_truth |
string | The ideal, factual target answer for evaluation (e.g., via RAGAS / TruLens). |
category |
string | Query intent type (e.g., factual_extraction, comparative_reasoning). |
Intended Use
- RAG Framework Benchmarking: Evaluating vector search accuracy and LLM context adherence.
- Automated Evaluation Pipelines: Compatible with
RAGAS,TruLens, andDeepEvalframeworks. - LLM Fine-Tuning: Training light LLMs on context-grounded response generation.
License
Distributed under the MIT License.
- Downloads last month
- 17
Models trained or fine-tuned on Flynntaggart26/learn-rag-eval-qa