Whether you are a large corporation or a single person, getting started with machine learning is not easy, especially if you have little to no data. To overcome this cold-start problem, Kern AI shows you how to leverage state-of-the-art techniques to kickstart your next NLP project and get high-quality labels fast.
The talk will cover topics such as active learning, weak supervision as well as confident and zero-shot learning. Your hosts will be Moritz, Div, and Leo from the company Kern AI, who use these advanced techniques in their open-source data-centric IDE for NLP called “refinery”.
Kern AI is a startup focusing on NLP services such as building training data, email automation, web scraping, as well as incorporating tools into automatic pipelines that leverage NLP itself.
Homepage:
Github:
Twitter:
Resources mentioned in the talk:
GPT-3 to reduce your labeling cost:
Language models are few shot learners:
Pervasive Label Errors in Test Sets – Destabilize Machine Learning Benchmarks:
Fine-tuning for similarity learning:
The waluigi-effect:
About Munich NLP:
Munich🥨NLP is a community newly founded in May 2022 by LMU and TUM students focusing on NLP topics. Within just six months, the community has already grown to over 1000 members consisting not only of current students, but also including PhD students, professors, and industry practitioners. We host weekly workshops and/or paper-reading events, both to learn from guests and to gather inspiration for our own (research) projects, as well as to establish and keep going an active student NLP community in the Munich area. The goal is to promote NLP-related exchange between students, researchers, and practitioners inside and outside the university and to showcase paths and possibilities during and after university.
…
#nlp #ai #deeplearning #data #labeling #linguistics #machinelearning #learning #teaching #opensource #bert #artificialintelligence #lmu #munich #transformers #chatgpt #gpt3 #gpt4 #languagemodel #naturallanguageprocessing
