Deep Learning and LLM
Copyright Elephant Scale May 16, 2023
Course Description
- This course concentrates on open-source models other than ChatGPT.
- Large Language Models (LLM) are taking the world by storm.
- HuggingFace provides libraries and a place to put LLMs in production.
- This course introduces the students to AI, Neural Networks, and LLMs.
After the course, you will be able to do the following tasks
- Evaluate LLM models
- Put the models into production
- Use HuggingFace as a possible implementation platform
Course objectives
- By the end of this course, students will know…
- How to understand the current state of the art in Deep Learning and AI
- How to put the claims of AI to the test
- How to utilize the existing results through transfer learning, pre-training, and fine-tuning.
- How to package your models for deployment.
- How to create machine learning pipelines and improve them in production.
Audience
- Developers, data scientists, team leads, project managers
Skill Level
- Intermediate
Duration
- Two days
Prerequisites
- General familiarity with machine learning
Format
- Lectures and hands-on labs. (50% – 50%)
Lab environment
- Zero Install: There is no need to install software on students’ machines!
- A lab environment in the cloud will be provided for students.
Students will need the following
- A reasonably modern laptop with unrestricted connection to the Internet. Laptops with overly restrictive VPNs or firewalls may not work properly.
- A checklist to verify connectivity will be provided
- Chrome browser
Detailed outline
Introduction to Deep Learning
- Understanding Deep Learning Use Cases
- Overview of Neural Networks: NN, CNN, RNN.
Prompt engineering for LLMs
- Best practices
- Practical advice
HuggingFace offering
- Transformers library
- Models
- Putting it all together
Fine-tuning a pre-trained model
- Processing the data
- Fine-tuning a model with the Trainer API or Keras
- A full training
Main NLP tasks
- Token classification
- Fine-tuning a masked language model
- Translation
- Summarization
- Training a causal language model from scratch
- Question answering
- Mastering NLP
Open LLM
- Overview of LLMs available
- Comparison of capabilities
- Evaluating and fine-tuning an LLM
- Alpaca from Stanford
- LLama from Facebook
- Dolly from Databricks
- Nomic
- Vicuna