rev

prez/prez.html

@@ -1397,7 +1397,7 @@ <h2>Experimentation Results</h2>
 <li><p><strong>Report used</strong>: the <a href="https://www.unhcr.org/sites/default/files/legacy-pdf/5dd4f7d24.pdf" target="_blank">2019 Evaluation of UNHCR’s data use and information management approaches</a> with two test summary <a href="https://github.com/Edouard-Legoupil/rag_extraction/raw/main/generated/Evaluation_Brief_response_text_bert.docx" target="_blank">#1</a> &amp; <a href="https://github.com/Edouard-Legoupil/rag_extraction/raw/main/generated/Evaluation_Brief_response_mmr_recursivecharactertext_bge.docx" target="_blank">#2</a>.</p></li>
 <li><p><strong>Models Tested</strong>: Small large language model that can run out of a strong laptop: <a href="https://cohere.com/blog/command-r" target="_blank">Command-r</a> &amp; <a href="https://mistral.ai/news/mixtral-of-experts/" target="_blank">Mixtral</a> for the generation, <a href="https://huggingface.co/BAAI/bge-large-en-v1.5" target="_blank">bge-large-en-v1.5</a> for the embeddings</p></li>
 <li><p><strong>Integration &amp; Documentation</strong>: Use of <a href="https://python.langchain.com/v0.1/docs/use_cases/question_answering/" target="_blank">LangChain</a> for the orchestration. Code shared and documented in <a href="https://github.com/Edouard-Legoupil/rag_extraction/" target="_blank">Github</a></p></li>
-<li><p><strong>Human Validation</strong>: Ground truthing with <a href="https://labelstud.io/" target="_blank">labelStud.io</a>.</p></li>
+<li><p><strong>Human Validation</strong>: Ground truthing with <a href="https://labelstud.io/templates/generative-llm-ranker" target="_blank">labelStud.io</a>.</p></li>
 <li><p><strong>Evaluation</strong>: Assess accuracy, relevance, and efficiency using <a href="https://docs.ragas.io/en/stable/" target="_blank">RAGAS (Retrieval Augmented Generation Assessment)</a>.</p></li>
 </ol>
 </div>

prez/prez.qmd

@@ -222,7 +222,7 @@ See [full article here](https://edouard-legoupil.github.io/rag_extraction/){targ
 
 3.  **Integration & Documentation**: Use of [LangChain](https://python.langchain.com/v0.1/docs/use_cases/question_answering/){target="_blank"} for the orchestration. Code shared and documented in [Github](https://github.com/Edouard-Legoupil/rag_extraction/){target="_blank"}
 
-4.  **Human Validation**: Ground truthing with [labelStud.io](https://labelstud.io/){target="_blank"}.
+4.  **Human Validation**: Ground truthing with [labelStud.io](https://labelstud.io/templates/generative-llm-ranker){target="_blank"}.
 
 5.  **Evaluation**: Assess accuracy, relevance, and efficiency using [RAGAS (Retrieval Augmented Generation Assessment)](https://docs.ragas.io/en/stable/){target="_blank"}.
 :::