Update Slide Deck 2

[calderonjesus]

…nts and references

What changes are you trying to make? (e.g. Adding or removing code, refactoring existing code, adding reports)

This pull request makes substantial improvements to the instructional slides on foundation models, focusing on clarity, accuracy, and up-to-date terminology. The changes include corrections to factual details, improved wording for readability, updates to references and dates, and the removal of redundant or outdated content. The slides now better reflect current understanding and best practices in the field of foundation models.

Key improvements and corrections:

Content accuracy and clarity:

• Refined explanations of key concepts such as scaling laws, model architecture, and post-training, including clearer descriptions of the Chinchilla scaling law, transformer architecture, and the role of transfer learning. [1] [2] [3] [4] [5] [6]
• Improved language for inclusivity and correctness, fixing typos (e.g., "misoginy" to "misogyny", "Russion" to "Russian", "immaging" to "imaging", "vlaue" to "value", etc.) and updating terminology for better readability. [1] [2] [3] [4] [5] [6]

Reference and factual updates:

• Corrected publication years and author references throughout the slides for accuracy (e.g., Bommasani et al. 2021, Huyen 2024). [1] [2] [3] [4] [5]

Structural and organizational improvements:

• Reorganized slide sections for better logical flow, including removing redundant process flow slides and clarifying section headings. [1] [2]
• Enhanced descriptions of dataset coverage, multilingual performance gaps, and bottlenecks in scaling, with updated insights on synthetic data and model collapse risks. [1] [2] [3]

Technical precision:

• Clarified technical details regarding model parameters, training tokens, and mixture-of-experts models, including more precise mathematical notation and explanations. [1] [2]

Post-training and transfer learning:

• Updated explanations and terminology for post-training methods, including RLHF, DPO, and RLAIF, and improved the description of transfer learning processes. [1] [2]

These changes collectively make the slides more accurate, readable, and aligned with current research and best practices.

What did you learn from the changes you have made?

N/A

Was there another approach you were thinking about making? If so, what approach(es) were you thinking of?

N/A

Were there any challenges? If so, what issue(s) did you face? How did you overcome it?

N/A

How were these changes tested?

N/A

A reference to a related issue in your repository (if applicable)

N/A

Checklist

• [ X] I can confirm that my changes are working as intended

[github-actions]

Hello, thank you for your contribution. If you are a participant, please close this pull request and open it in your own forked repository instead of here. Please read the instructions on your onboarding Assignment Submission Guide more carefully. If you are not a participant, please give us up to 72 hours to review your PR. Alternatively, you can reach out to us directly to expedite the review process.