Full Stack Slm Development with Advanced Ml/Dl and Optimization

Descripción de la vacante

We are seeking a highly skilled full stack developer or team to build a Small Language Model (SLM) with comprehensive capabilities. The project requires expertise across various domains, from core programming to advanced machine learning, deep learning, and robust deployment strategies. The ideal candidate will be proficient in the specified technologies and able to deliver a high-performance, scalable solution.

Key areas of development include:

Core Languages: Development will primarily utilize Python, C++, SQL, and Shell/Bash scripting for various components of the SLM and its infrastructure.

ML & Deep Learning: Implementation will involve Py Torch for model training and leveraging Hugging Face libraries, specifically Transformers, for advanced model architectures. Experience with distributed training techniques is a plus.

Optimization & Fine-Tuning: Expertise in model optimization techniques such as quantization (gguf, awq, gptq) and efficient fine-tuning methods like lora / qlora is essential to ensure optimal performance and resource utilization.

Inference Engines & Runtimes: The project requires integration with high-performance inference engines and runtimes to ensure efficient model serving.

Data & Backend: Development of the data pipeline and backend services will involve tools like Polars, Pandas, and Ray for data processing, Fastapi for api development, and integration with rest apis. Knowledge of Vector Databases such as Pinecone, Qdrant, or Milvus is also required.

Dev Ops & Cloud: The solution must be deployable and manageable in cloud environments. Experience with Docker for containerization, Kubernetes for orchestration, and cloud platforms like AWS (Sage Maker/EC2) and GCP is necessary. Strong understanding of Git and ci/cd practices is expected for efficient development workflows.

Category: IT & Programming

Subcategory: Artificial Intelligence

Project size: Small