Machine Learning Expert for Fine-Tuning & Weight Merging
Budget: $250 – $750 USD
I'm looking for an experienced Machine Learning professional to develop a fine-tuning and weight merging script. The goal is to use a dataset specific to my business to create a smaller model that will merge with the original one. The base models are:
- DeepSeek-R1-Distill-Qwen-14B
- DeepSeek-R1-Distill-Qwen-7B
Technical Requirements:
- Proficient in PyTorch, as this is the preferred framework for the project.
- Skilled in weight merging techniques, particularly Averaging and Stochastic Weight Averaging (SWA).
- Comfortable working in a Linux environment, as this is the chosen development platform.
- Experience with the DS-r1-14B/7B models is a plus.
Available Hardware:
- CPU: 8 cores / 16 threads (AMD Ryzen 7 6800H)
- RAM: 32 GB DDR5 (4800MHz)
- GPU: NVIDIA 3070Ti (8GB)
Deliverables:
- A fully functional script with clear documentation.
- Support for adjustments and integration throughout the development process.
- DeepSeek-R1-Distill-Qwen-14B
- DeepSeek-R1-Distill-Qwen-7B
Technical Requirements:
- Proficient in PyTorch, as this is the preferred framework for the project.
- Skilled in weight merging techniques, particularly Averaging and Stochastic Weight Averaging (SWA).
- Comfortable working in a Linux environment, as this is the chosen development platform.
- Experience with the DS-r1-14B/7B models is a plus.
Available Hardware:
- CPU: 8 cores / 16 threads (AMD Ryzen 7 6800H)
- RAM: 32 GB DDR5 (4800MHz)
- GPU: NVIDIA 3070Ti (8GB)
Deliverables:
- A fully functional script with clear documentation.
- Support for adjustments and integration throughout the development process.