2 of 2 Remote/Hybrid Permanent Qwen Jobs

model fine-tuning strategy, such as SFT, LoRA/QLoRA, and optimisation approaches Selecting and evaluating base models (e.g. Mistral, Qwen, Phi, Falcon) based on performance, cost, and use case Defining evaluation frameworks and standards to ensure models meet production-grade quality and reliability Building and scaling a suite ...

constrained hardware such as Jetson, Android, iPad, Raspberry Pi or microcontrollers. Hands-on with llama.cpp or similar, and evaluating small on-device models (e.g. Qwen, Gemma, MedGemma, custom). Some hardware exposure: anyone doing genuine edge AI tends to have this. Cloud experience (AWS), particularly running LLMs in a cloud ...