Comparing albert-small model with apertus-small model alone, and in a RAG setting with service-public + travail-emploi sheets, on a french administration Q/A datasets
This dataset contains 8 experiments
from the EvalAP evaluation platform.
Datasets: Assistant IA - QA, MFS_questions_v01
Models evaluated: meta-llama/Llama-3.1-8B-Instruct, swiss-ai/Apertus-8B-Instruct-2509
Metrics: generation_time… See the full description on the dataset page:
https://huggingface.co/datasets/AgentPublic/evalap-compare-albert-small-with-apertus-small-models-82.
