mirror of
https://github.com/rasbt/LLMs-from-scratch.git
synced 2025-10-27 15:59:49 +00:00
4f9775d91c
* updated requirements * fixes llama2 to llama3 * fixed llama 3.2 standalone * fixed typo * fixed rope formula * Update requirements-extra.txt * Update ch05/07_gpt_to_llama/converting-llama2-to-llama3.ipynb * Update ch05/07_gpt_to_llama/converting-llama2-to-llama3.ipynb * Update ch05/07_gpt_to_llama/standalone-llama32.ipynb --------- Co-authored-by: Sebastian Raschka <mail@sebastianraschka.com>
Chapter 5: Pretraining on Unlabeled Data
Main Chapter Code
- 01_main-chapter-code contains the main chapter code
Bonus Materials
- 02_alternative_weight_loading contains code to load the GPT model weights from alternative places in case the model weights become unavailable from OpenAI
- 03_bonus_pretraining_on_gutenberg contains code to pretrain the LLM longer on the whole corpus of books from Project Gutenberg
- 04_learning_rate_schedulers contains code implementing a more sophisticated training function including learning rate schedulers and gradient clipping
- 05_bonus_hparam_tuning contains an optional hyperparameter tuning script
- 06_user_interface implements an interactive user interface to interact with the pretrained LLM
- 07_gpt_to_llama contains a step-by-step guide for converting a GPT architecture implementation to Llama 3.2 and loads pretrained weights from Meta AI