icedduck/lab1-sft_ex

coding stage ②-B: SFT on the DeepSeek textbook, exercism modules from scratch with tests; pass@1 .329 with tests, .240 without (report §3.14, §3.15).

Part of LLM from scratch, to the aha moment, to a coding agent: fifteen experiments on four RTX 4090s. Code, report, the evaluation protocol and the scripts that produced this checkpoint: https://github.com/bethehand/lab1-llm-scratch-to-aha-moment-to-coding-agent

Evaluate it yourself from the repository root, for example python3 bench_observation_price.py --hf-dir icedduck/lab1-sft_ex for the coding-line checkpoints. Weights are derived from Qwen/Qwen2.5-Coder-1.5B and remain under the Qwen license.