Scribd
Upload
6 views2 pages

4

Uploaded by

Siddhartha Negi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
6 views2 pages

4

Uploaded by

Siddhartha Negi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

I'll break down the time allocation based on 1-2 hours per day (roughly 8-14

hours per week).

Month 1: Foundation of Generative Models

Week 1-2: Autoregressive Models & Transformers (28 hours total)


- Theory (10 hours):
- Attention mechanisms (3 hours)
- Transformer architecture (4 hours)
- Tokenization and embeddings (3 hours)

- Implementation Practice (18 hours):


- Study and replicate basic transformer code (8 hours)
- Project 1: Character-level language model (10 hours)
Daily Schedule:
- Weekdays: 1.5 hours (30 min theory, 1 hour coding)
- Weekends: 2 hours (focused on implementation)

Week 3-4: Diffusion Models (28 hours total)


- Theory (12 hours):
- Forward/reverse diffusion (4 hours)
- U-Net architecture (4 hours)
- Score matching (4 hours)

- Implementation Practice (16 hours):


- Understanding existing implementations (6 hours)
- Project 2: Basic diffusion model (10 hours)
Daily Schedule:
- Weekdays: 1.5 hours (45 min theory, 45 min coding)
- Weekends: 2 hours (focused on project)

Month 2: Advanced Concepts & Implementation


Week 1-2: Advanced Language Models (28 hours total)
- Theory (14 hours):
- Prompt engineering (3 hours)
- Few-shot learning (3 hours)
- RLHF basics (4 hours)
- Fine-tuning concepts (4 hours)

- Implementation Practice (14 hours):


- Project 3: Fine-tuning a small LM
Daily Schedule:
- Weekdays: 1.5 hours (45 min theory, 45 min practice)
- Weekends: 2 hours (project work)

Week 3-4: Advanced Image Generation (28 hours total)


- Theory (12 hours):
- Latent diffusion (4 hours)
- Conditioning mechanisms (4 hours)
- Cross-attention (4 hours)

- Implementation Practice (16 hours):


- Project 4: Conditional generation
Daily Schedule:
- Weekdays: 1.5 hours (30 min theory, 1 hour coding)
- Weekends: 2.5 hours (project focus)

You might also like