This release accompanies the release of Llama 3.2 which included new versions of the Llama models in sizes of 1B, 3B, 11B and 90B. To get started with the new models you can find information in the official documentation or the on the HuggingFace hub. Further details can also be found in the model card and the The Llama 3 Herd of Models paper. For this release we updated the documentation and made sure all components work with the new models including multimodal finetuning.

What's Changed

Integrate Llama 3.2

• Upstream merge by @albertodepaola in https://github.com/meta-llama/llama-recipes/pull/677

New and updated recipes

• Adding end-to-end llama chatbot recipe using Retrieval Augmented Fine Tuning (RAFT) by @wukaixingxp in https://github.com/meta-llama/llama-recipes/pull/569
• [WIP] adding chatbot-e2e by @HamidShojanazeri in https://github.com/meta-llama/llama-recipes/pull/462
• [Azure] Update Azure API usage example to 3.1 by @WuhanMonkey in https://github.com/meta-llama/llama-recipes/pull/615
• Corrected wrong order of commands by @BakungaBronson in https://github.com/meta-llama/llama-recipes/pull/602
• Fill in one sentence in the prompt guard tutorial. by @cynikolai in https://github.com/meta-llama/llama-recipes/pull/609
• Llamaguard notebook colab link fix by @tryrobbo in https://github.com/meta-llama/llama-recipes/pull/619
• Updating llama 3 references to 3.1 model by @init27 in https://github.com/meta-llama/llama-recipes/pull/632
• recipes/quickstart/Getting_to_know_Llama.ipynb, typo fix lama -> llama line 127 by @cselip in https://github.com/meta-llama/llama-recipes/pull/635
• Update hello_llama_cloud.ipynb by @MrDlt in https://github.com/meta-llama/llama-recipes/pull/584
• Update hello_llama_cloud.ipynb by @MrDlt in https://github.com/meta-llama/llama-recipes/pull/638
• Add preprocessor to patch PromptGuard scores for inserted characters by @cynikolai in https://github.com/meta-llama/llama-recipes/pull/636
• Eval reproduce recipe using lm-evaluation-harness and our 3.1 evals datasets by @wukaixingxp in https://github.com/meta-llama/llama-recipes/pull/627

Documentation update

• Update readme text to be version-agnostic by @subramen in https://github.com/meta-llama/llama-recipes/pull/614
• Move supported features table to main README by @subramen in https://github.com/meta-llama/llama-recipes/pull/616
• document less obvious training config parameters by @kjslag in https://github.com/meta-llama/llama-recipes/pull/522

Misc fixes

• Enable users to trust remote code in samsum dataset by @mreso in https://github.com/meta-llama/llama-recipes/pull/628
• Use new get_model_state_dict api for save_pretrained peft model by @mreso in https://github.com/meta-llama/llama-recipes/pull/629
• Fix version number in Python example by @wstnmssr in https://github.com/meta-llama/llama-recipes/pull/643
• Fix checkpoint saving by @mreso in https://github.com/meta-llama/llama-recipes/pull/650
• Adding custom dataset file by @goswamig in https://github.com/meta-llama/llama-recipes/pull/659
• Make gradio and langchain optional dependencies by @mreso in https://github.com/meta-llama/llama-recipes/pull/676
• Update get_default_finetune_args.py by @edamamez in https://github.com/meta-llama/llama-recipes/pull/662
• Fix/custom dataset chat template by @mreso in https://github.com/meta-llama/llama-recipes/pull/665
• Create v0.0.4 release by @mreso in https://github.com/meta-llama/llama-recipes/pull/678

New Contributors

• @cynikolai made their first contribution in https://github.com/meta-llama/llama-recipes/pull/609
• @BakungaBronson made their first contribution in https://github.com/meta-llama/llama-recipes/pull/602
• @init27 made their first contribution in https://github.com/meta-llama/llama-recipes/pull/632
• @cselip made their first contribution in https://github.com/meta-llama/llama-recipes/pull/635
• @MrDlt made their first contribution in https://github.com/meta-llama/llama-recipes/pull/584
• @wstnmssr made their first contribution in https://github.com/meta-llama/llama-recipes/pull/643
• @goswamig made their first contribution in https://github.com/meta-llama/llama-recipes/pull/659
• @edamamez made their first contribution in https://github.com/meta-llama/llama-recipes/pull/662

Full Changelog: https://github.com/meta-llama/llama-recipes/compare/v0.0.3...v0.0.4

We would like to thank all who contributed to this release and are looking forward to future contributions!