Show HN: Ran Qwen 3.8 27B autonomously(ish) for 3 weeks on single 3090

A developer ran Qwen 3.8 27B (Q4) autonomously inside Deepseek Harness on a single RTX 3090 GPU for 3 weeks, with the resulting repository stating that everything in it was created by the model except the agents/*, human/*, and AGENTS.md files. The project is described as an experiment in agentic autonomy protocols.

Datasets: https://huggingface.co/datasets | text stringclasses 5 values | |---| | set TARGET llama-finetune | | add executable ${TARGET} finetune.cpp | | install TARGETS ${TARGET} RUNTIME | | target link libraries ${TARGET} PRIVATE llama-common llama ${CMAKE THREAD LIBS INIT} | | target compile features ${TARGET} PRIVATE cxx std 11 | Experiment in agentic autonomy protocols. ~ everything in this repo was created by Qwen 3.8 27B Q4 running autonomously inside Deepseek Harness, on a single RTX 3090 GPU, for 3 weeks. The only human artifacts are: agents/ human/ AGENTS.md - Downloads last month - -