https://github.com/aashish75/fine-tune-llama-trump
Science Score: 26.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (2.2%) to scientific vocabulary
Repository
Basic Info
- Host: GitHub
- Owner: Aashish75
- Language: Jupyter Notebook
- Default Branch: main
- Size: 95.7 KB
Statistics
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
- Releases: 0
Metadata Files
README.md
fine-tune-llama-trump
Fine Tuning of Llama-3.2-1B-Instruct to improve it's ability to tweet like Donald Trump.
Tweet Data from https://drive.google.com/file/d/1xRKHaP-QwACMydlDnyFPEaFdtskJuBa6/view?usp=sharing to fine tune the model.
The model was fine-tuned with instruction-tweet pairs where the instruction asked the model to talk like Donald Trump and the tweet was a sample from the data mentioned above.
LoRA(Low-Rank Adaptation) Config was used where tunable rank 16 matrices were induced instead of changing fine-tuning the Llama architecture itself.
Scaled training across multiple GPU instances using DeepSpeed on Amazon SageMaker; merged and pushed model to Hugging Face Hub.
Model available at - https://huggingface.co/Aashish75/llama-3.2-1b-trump0
Owner
- Name: Aashish M
- Login: Aashish75
- Kind: user
- Repositories: 1
- Profile: https://github.com/Aashish75
I code
GitHub Events
Total
- Push event: 10
- Create event: 2
Last Year
- Push event: 10
- Create event: 2