Samuel Lima Braz PRO

samuellimabraz

AI & ML interests

None yet

Recent Activity

Organizations

Tech4Humans's profile picture Hugging Face Discord Community's profile picture

samuellimabraz's activity

posted an update 1 day ago
view post
Post
477
Sharing my new article on an open-source project for automated signature detection in document processing. The article details:

- Dataset Engineering: Combining two public collections to create a hybrid dataset.
- Architecture Benchmarking: Evaluating sota models such as YOLO series, DETR variants, and YOLOS for accuracy and efficiency.
- Model Optimization: Using Optuna for hyperparameter tuning, achieving a 7.94% F1-score improvement.
- Production Deployment: Implementing Triton Inference Server with an OpenVINO CPU backend for optimized inference.

It's not such a complex project, but I explore the training of the best current architectures for object detection and share all notebooks, data, models, and the repo with deployment and benchmarking details.

Thanks @SkalskiP , @nielsr , @sergiopaniego for the notebooks and resources that have been very helpful.

- https://huggingface.co/blog/samuellimabraz/signature-detection-model
- tech4humans/signature-detection-678b087d8b0ce22ae8c3f60e
posted an update about 2 months ago
view post
Post
458
I wrote a article on Parameter-Efficient Fine-Tuning (PEFT), exploring techniques for efficient fine-tuning in LLMs, their implementations, and variations.

The study is based on the article "Scaling Down to Scale Up: A Guide to Parameter-Efficient Fine-Tuning" and the PEFT library integrated with Hugging Face's Transformers.

Article: https://huggingface.co/blog/samuellimabraz/peft-methods
Notebook: https://colab.research.google.com/drive/1B9RsKLMa8SwTxLsxRT8g9OedK10zfBEP?usp=sharing
Collection: samuellimabraz/service-summary-6793ccfe774073328ea9f8df

Analyzed methods:
- Adapters: Soft Prompts (Prompt Tuning, Prefix Tuning, P-tuning), IAΒ³.
- Reparameterization: LoRA, QLoRA, LoHa, LoKr, X-LoRA, Intrinsic SAID, and variations of initializations (PiSSA, OLoRA, rsLoRA, DoRA).
- Selective Tuning: BitFit, DiffPruning, FAR, FishMask.

I'm starting out in generative AI, I have more experience with computer vision and robotics. Just sharing here πŸ€—