Issue 54 mot ablation - #206
Open
H2Rhan wants to merge 3 commits into
Open
Conversation
feat(lora): LoRA rank sweep results for VisDrone and Brain Tumor datasets 犀牛鸟 Tencent#50 — 垂类场景 LoRA 高效微调适配 实验结果提交 ## Experiments (6 runs, SLURM job 89299, NVIDIA A40) ### VisDrone (航拍检测, 10 classes, 30 epochs) | Rank | mAP50 | mAP50-95 | Precision | Recall | Time | |------|-------|----------|-----------|--------|------| | r=4 | 0.045 | 0.019 | 0.421 | 0.066 | 5.06h | | r=8 | 0.049 | 0.022 | 0.423 | 0.072 | 5.06h | | r=16 | 0.064 | 0.029 | 0.334 | 0.088 | 5.25h | ### Brain Tumor (医疗检测, 2 classes, 40 epochs) | Rank | mAP50 | mAP50-95 | Precision | Recall | Time | |------|-------|----------|-----------|--------|------| | r=4 | 0.449 | 0.294 | 0.464 | 0.659 | 0.36h | | r=8 | 0.450 | 0.287 | 0.442 | 0.721 | 0.37h | | r=16 | 0.476 | 0.314 | 0.466 | 0.721 | 0.39h | ### Changes - Updated sweep shell scripts with proper headers and error handling - Modified LoRA YAML configs for rank sweep experiments - Updated README with experiment documentation - Added complete sweep results (yolo_master_lora_rank_sweep_results.csv) - Added detailed per-experiment metrics (yolo_master_lora_results.csv) Co-Authored-By: Claude <noreply@anthropic.com> @
… fix - tests/test_mot.py: 27→36 boundary tests (fp16 stability, hard routing gradient flow, routing determinism, top_k=E dense routing, scene-aware router, train/eval consistency, invalid top_k validation, multi-size localconv, shift_size=0 mode) - scripts/mot_routing_analysis.py: MoTBlock expert routing analysis script (router bias extraction, per-layer expert preference, scene- conditioned activation patterns, temperature annealing tracking) - block.py:220/232: fix dtype mismatch in _blend_experts (Half×Float→add .to(out.dtype) cast)
H2Rhan
force-pushed
the
issue-54-mot-ablation
branch
from
July 31, 2026 12:04
53e60ad to
570bd72
Compare
- experiments/issue54_mot_ablation/summary.csv: 4种变体汇总
- experiments/issue54_mot_ablation/{v10,v10_mot,v10_moa,v10_moa_mot}/results.csv
- experiments/issue54_mot_ablation/REPORT.md: 完整技术报告
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
变更内容
1. tests/test_mot.py: 27 → 36 边界测试
新增 9 个边界测试：fp16 稳定性、纯硬路由梯度流、top_k=E dense routing、路由确定性、train/eval 一致性、shift_size=0 模式、多尺寸兼容性、非法 top_k 校验、scene-aware routing
2. scripts/mot_routing_analysis.py (新文件)
MoT 路由可解释性分析脚本：Router bias 提取、分层 expert 偏好分析、温度退火追踪
3. block.py dtype 修复
_blend_experts 添加 .to(out.dtype) 防止 Half×Float 不匹配
4. experiments/issue54_mot_ablation/
消融实验训练结果：summary.csv + 4 变体 results.csv + 技术报告
消融实验核心结果
Router 分层分工