Downloads · 30 days
11
7% of all-time downloads
Malaysian_Gemma3_270M_F16
Faris-Faiz/Malaysian_Gemma3_270M_F16
Malaysian_Gemma3_270M_F16 is a text generation model from Faris-Faiz. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as apache-2.0.
- Developed by: Faris-Faiz - License: apache-2.0 - Finetuned from model : unsloth/gemma-3-270m-it
All-time downloads
147
Public
Parameters
268M
574 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors536 MB · 93%
Parameter types
How the weights are stored.
BF16168M · 63%
Try a prompt
Base models
Datasets
Model card
From the Hugging Face model README
MMLU Benchmark
Model Accuracy shot by_letter category
0 Malaysian_Gemma3_270M_F16 43.143676 0shot True STEM
1 Malaysian_Gemma3_270M_F16 43.861323 0shot True Language
2 Malaysian_Gemma3_270M_F16 44.246892 0shot True Social science
3 Malaysian_Gemma3_270M_F16 43.367714 0shot True Others
4 Malaysian_Gemma3_270M_F16 49.260523 0shot True Humanities
{'Social science': np.int64(6918), 'Language': np.int64(6288), 'Humanities': np.int64(4395), 'Others': np.int64(4169), 'STEM': np.int64(2443)}
Model : Malaysian_Gemma3_270M_F16
Metric : first
Shot : 0shot
average accuracy 44.794118861768474
accuracy for STEM 43.143675808432256
accuracy for Language 43.86132315521628
accuracy for Social science 44.246892165365715
accuracy for Others 43.367714080115135
accuracy for Humanities 49.26052332195677
Uploaded finetuned model
- Developed by: Faris-Faiz
- License: apache-2.0
- Finetuned from model : unsloth/gemma-3-270m-it
dataset was fine-tuned on split (split="force_malay")
This gemma3_text model was trained 2x faster with Unsloth and Huggingface's TRL library.
- Task
- Text Generation
- Library
- transformers
- Type
- gemma3_text
- License
- apache-2.0
- Languages
- en
- Created
- Aug 21, 2025
- Updated
- Aug 28, 2025