Model Base Change:
- Switched from GPT-2 to DistilGPT-2 due to resource limitations on ChromeOS (Crostini)
Step 1 Completion: ✅
- Successful installation of key packages:
- PyTorch
- Hugging Face Transformers
- OpenAI DistilGPT-2 datasets
Jump to Step 3 (Fine-Tuning): 💬
- Fine-tuned DistilGPT-2 using the OpenWebText dataset.
System Improvements:
- Added extra storage, routed microSD card to Linux environment to resolve temp storage issues.
- Addressed storage limitations that initially prevented testing fine-tuned model.
Repository Updates:
- Began editing Juniper2.0 repository.
Collect Additional Datasets: 💬
- Explore more datasets for model fine-tuning.
Model Integration:
- Build either CLI or UI for user interaction
- Options incl Flask or Django.
Model Refinement:
- Cont improving fine-tuned model's performance.
Custom Dataset Creation:
- Develop custom samples from Juniper (GPT-4) dataset.
Further Testing:
- Begin testing more input/output scenarios with the fine-tuned model.
System Optimization:
- Apply additional sys optimizations to improve efficiency as necessary.