Examinando por Materia "Large Language Models (LLM)"

Mostrando 1 - 2 de 2
  • No hay miniatura disponible
    Publicación
    Ajuste fino de un modelo LLM para realizar reportes resumidos de expertos en trading, con integración de datos desde redes sociales
    (Universidad EAFIT, 2025) Restrepo Acevedo, Andrés Felipe; Martínez Vargas, Juan David
    The contemporary financial market is characterized by its high complexity and the massive volume of structured and unstructured data generated daily, posing significant challenges for individual investors in terms of analysis and informed decision making. This project proposes the fine-tuning of a Small Language Model (SLM) integrated into a tool capable of generating financial analysis reports similar to those produced by experts. For the proof of concept (PoC), transcripts from financial analysis videos published by experts on their YouTube channels are utilized. The SLM is fine-tuned using instruction-based techniques and the incorporation of the LoRa(Low-Rank Adapters) method, with the aim of extracting and summarizing key information relevant to individual investors. The main objective of this tool is to assist individual investors by generating efficient and accessible reports, facilitating access to valuable information in natural language, and enhancing their ability to make data-driven decisions from unstructured data, all with minimal investment of time and resources. Experimental results demonstrate the viability of using fine-tuned Small Language Models (SLMs) for the generation of high-quality financial reports. Specifically, the selected model, finetune qlora unsloth llama 3.1 8B Instruct bnb 4bit v2 Q8 0, achieved an average score of 5.67 out of 10 in the evaluation conducted by a judge LLM, with an average cosine distance of 0.159 compared to the reference summaries generated by the foundational pretrained model GPT-4.1. This improvement represents a 97.5% increase in performance compared to the same base model, Llama 3.1 8B Instruct, without fine-tuning. Qualitatively, the model exhibits high fidelity and coherence in the extraction and synthesis of key information in moderately long contexts, although it faces challenges in thematic interpretation when dealing with considerably lengthy transcripts. Additionally, implementation of this tool is projected to save 560 hours annually for individual investors, along with an estimated annual reduction in API costs ranging from 7.52 to 25 for the channels analyzed in the proof of concept.
  • No hay miniatura disponible
    Publicación
    Inteligencia del mercado laboral colombiano : detección automatizada de habilidades mediante modelos grandes de lenguaje (LLM) y recuperación aumentada (RAG)
    (Universidad EAFIT, 2025) Zapata Posada, Jorge Mario; Álvarez Barrera, Claudia Patricia; Padilla Buritica, Jorge Iván
    The demand for skills in the labor market has evolved significantly in recent decades, driven by changes in the economic environment and constant technological advances. In this context, the detailed description of each job offer, available on employment web portals, provides accurate information on the specific skills required by the market in real time. Labor Market Intelligence (LMI) research uses this data along with machine learning algorithms to anticipate trends and understand the evolution of talent demand. Despite advances in artificial intelligence and the availability of large data volumes, there remains a gap in adapting these technologies to local contexts. Regional markets, such as Colombia, require customized approaches to ensure that technological solutions respond to the specific needs of the labor market, effectively aligning talent supply and demand. This study analyzes data from the Talent.com employment platform for Colombia using a state-of-the-art approach based on Large Language Models (LLM) combined with Retrieval Augmented Generation (RAG) to identify emerging, traditional, technical, and soft skills. In the first stage, a multilingual LLM extracts skill mentions from job descriptions. In the second stage, a semantic retrieval module queries the European Commission’s open ESCO skills taxonomy to propose standardized candidate labels, the LLM then selects the most appropriate label and delivers validated, structured JSON outputs. Preliminary results show improvements in precision, coverage, and auditability compared to purely supervised approaches, reducing hallucinations through candidate-constrained selection and standardizing categories using ESCO skill classification. This framework provides valuable insights that, in future work, may support universities in designing academic programs aligned with labor market needs, thus facilitating strategic decision-making for employers, policymakers, and educators, and contributing to talent development and the reduction of unemployment in Colombia.