Hands-on Qwen3-1.7B code post-training lab: SFT, LoRA, DPO, PPO, GRPO, and reproducible EvalPlus benchmarks.
-
Updated
Oct 7, 2026 - HTML
#
evalplus
Here are 7 public repositories matching this topic...
NTU final-year project: can an LLM agent tell whether generated code is correct without ground truth? Verifier reliability measured on EvalPlus HumanEval+
-
Updated
Oct 7, 2026 - Python
Cluster-Aware Sequential Refinement (CASR) for test-free multi-model code generation and model-pool selection.
machine-learning program-synthesis python3 refinement ensemble-learning code-generation large-language-models llm code-llms evalplus apps-dataset
-
Updated
Aug 25, 2026 - Python
Compute-efficient QLoRA specialization of Qwen3-4B for Python, with executable EvalPlus evidence and reproducible local training.
-
Updated
Sep 11, 2026 - Python
Research companion for LoRA instruction fine-tuning of Qwen2.5-Coder, with benchmark results from the manuscript, comparisons across data budgets, and documented reproducibility limits.
-
Updated
Nov 1, 2026 - Python
Executable-code evaluation and verifier-guided inference, with reproducible benchmarks and separate model-weight retention audits.
code-generation reproducibility verifier post-training l20 qwen llm-evaluation coding-agents rlvr livecodebench evalplus
-
Updated
Sep 6, 2026 - Python
Research code and artifacts for How LLMs Fail in Code Generation: a cross-benchmark, fine-grained error taxonomy (SDDT 2026).
taxonomy reproducible-research code-generation error-analysis large-language-models llm-evaluation bigcodebench evalplus
-
Updated
Oct 3, 2026 - Python
Add this topic to your repo
To associate your repository with the evalplus topic, visit your repo's landing page and select "manage topics."