Skip to content

Week6_복습과제_최지희#99

Open
SweetFriedPotato wants to merge 3 commits intoEwha-Euron:Week_6from
SweetFriedPotato:Week_6
Open

Week6_복습과제_최지희#99
SweetFriedPotato wants to merge 3 commits intoEwha-Euron:Week_6from
SweetFriedPotato:Week_6

Conversation

@SweetFriedPotato
Copy link
Copy Markdown

No description provided.

Add a comprehensive overview of the research paper, including its contributions, limitations, and experimental results in natural language processing and visual object recognition.
Copilot AI review requested due to automatic review settings May 4, 2026 12:28
Copy link
Copy Markdown

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Adds Week 6 assignment deliverables to the repository: a GPT-1 implementation notebook (pre-training + fine-tuning workflow) and a markdown write-up of the GPT-1 paper notes.

Changes:

  • Add a Jupyter notebook implementing a Transformer decoder/GPT(-1)-style model and demonstrating training setup (WikiText2 pre-train, IMDB fine-tune).
  • Add a markdown document containing the Week 6 “예습과제” write-up and an external Notion link.

Review notes (blocking):

  • Week6_복습과제_최지희.ipynb contains a hard-coded os.chdir('/home/jiheechoi0102/jihi') and committed cell outputs/metadata (e.g., pip install logs) that leak local environment details and reduce reproducibility.
  • In MHA, self.scale = torch.sqrt(torch.tensor(d_model/n_heads)) creates a CPU tensor that will not move with .to(device) and will break on CUDA with a device-mismatch error when dividing attention_score on GPU.
  • Week6_예습과제_최지희.md uses attachment: image links exported from Notion that won’t render on GitHub unless the images are added to the repo or replaced with accessible URLs.

Reviewed changes

Copilot reviewed 1 out of 2 changed files in this pull request and generated no comments.

File Description
Week6_복습과제_최지희.ipynb GPT-1-style model + training notebook added (includes env-specific paths/outputs and a CUDA-breaking scale tensor).
Week6_예습과제_최지희.md Week 6 paper notes and Notion link added (images currently reference non-GitHub attachment: URIs).

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants