Michelle0574 aeddf8e1ee feat: 支持固定超参训练并部署 4 专家稀疏 MoE 模型

2026-08-15 12:22:21 +00:00

..

check_prompt_len.py

feat: final_train_optuna_cv 支持 MoE + Qwen2.5-7B QLoRA + soft-RAG，产出可部署 checkpoint

2026-08-13 10:05:00 +00:00

evaluate_external.py

增加LiON的评估指标

2026-01-21 16:20:10 +08:00

fixed_hparams.json

feat: final_train_optuna_cv 支持 MoE + Qwen2.5-7B QLoRA + soft-RAG，产出可部署 checkpoint

2026-08-13 10:05:00 +00:00

hparams_sparse_moe.json

feat: 支持固定超参训练并部署 4 专家稀疏 MoE 模型

2026-08-15 12:22:21 +00:00

measure_cost.py

feat: final_train_optuna_cv 支持 MoE + Qwen2.5-7B QLoRA + soft-RAG，产出可部署 checkpoint

2026-08-13 10:05:00 +00:00

precompute_chemeleon.py

Add full ablation results and CheMeleon/UniMol encoders

2026-07-24 17:56:49 +00:00

precompute_mole.py

Add MoleculeSTM/MolE/RF/KNN baselines: encoders, scripts, and CV results

2026-07-28 14:14:10 +00:00

precompute_moleculestm.py

Add MoleculeSTM/MolE/RF/KNN baselines: encoders, scripts, and CV results

2026-07-28 14:14:10 +00:00

precompute_unimol.py

Add full ablation results and CheMeleon/UniMol encoders

2026-07-24 17:56:49 +00:00

preprocess_internal.py

fix: 修正 size 预测的反归一化并将 MoE+LLM 模型接入应用

2026-08-14 11:24:02 +00:00

process_benchmark_data.py

整合与baseline比对的相关脚本

2026-02-26 18:31:22 +08:00

process_external.py

移除其他无用脚本

2026-02-26 19:31:48 +08:00

run_classical_baselines.py

Fix task-specific Random Forest configuration and results

2026-08-07 17:07:55 +08:00

summarize_ablation.py

feat(moe): MoE + MolT5 LLM 融合层、消融流水线

2026-06-18 08:51:30 +00:00