Collected molecules will appear here. Add from search or explore.
An end-to-end LLMOps pipeline for fine-tuning (SFT) and aligning (DPO) small language models like Phi-2 on low-cost hardware using QLoRA, vLLM, and Langfuse for feedback loops.
stars
0
forks
0
The project serves as a practical tutorial or reference architecture for a standard LLM alignment workflow. With 0 stars and no forks after 3 months, it lacks community traction and relies entirely on standard libraries (Hugging Face, vLLM, Langfuse). Frontier labs and established players like Hugging Face (TRL library) provide more robust, production-grade versions of this exact workflow.
TECH STACK
INTEGRATION
reference_implementation
READINESS