From a433b82cdfb012c5f35d0e6c12dd9b13ac25e190 Mon Sep 17 00:00:00 2001
From: ninehills <ninehills@users.noreply.github.com>
Date: Sat, 26 Oct 2024 13:43:49 +0000
Subject: [PATCH] [create-pull-request] automated change

---
 articles/118.md | 45 +--------------------------------------------
 1 file changed, 1 insertion(+), 44 deletions(-)

diff --git a/articles/118.md b/articles/118.md
index 1cb80bb..069e96b 100644
--- a/articles/118.md
+++ b/articles/118.md
@@ -205,47 +205,4 @@ TODO
 
 ## 7. NUDGE 微调
 
-参考 <https://github.com/szeighami/nudge> 进行 NUDGE 微调。
-
-```bash
-PYTHONPATH="." python eval/evaluate_nudge.py \
-    --dataset_path "./data/infgrad_retrieval_data_llm.json" \
-    --encoder "checkpoint/bge-small-zh-v1.5-sft" \
-    --query_instruction "为这个句子生成表示以用于检索相关文章：" \
-    --split "val" \
-    --search_top_k 10 \
-    --use_nudge_n True
-{
-    "ndcg_at_1": 0.61147,
-    "ndcg_at_3": 0.69472,
-    "ndcg_at_5": 0.71365,
-    "ndcg_at_10": 0.73091,
-    "map_at_1": 0.61147,
-    "map_at_3": 0.67397,
-    "map_at_5": 0.68447,
-    "map_at_10": 0.69166,
-    "recall_at_1": 0.61147,
-    "recall_at_3": 0.75487,
-    "recall_at_5": 0.80087,
-    "recall_at_10": 0.8539,
-    "precision_at_1": 0.61147,
-    "precision_at_3": 0.25162,
-    "precision_at_5": 0.16017,
-    "precision_at_10": 0.08539,
-    "mrr_at_1": 0.61147,
-    "mrr_at_3": 0.67397,
-    "mrr_at_5": 0.68447,
-    "mrr_at_10": 0.69166
-}
-```
-
-ndcg@10 从 0.67399 提升到 0.73091，提升 5.7pp。和全参数微调的效果差不多。
-
-NUDGE和全参数微调的对比：
-
-- 训练时间： NUDGE 的训练其实是训练 Embedding 变换参数。
-- NUDGE 针对 Embedding 后的数据，新增数据需要重复训练。且只影响 corpus embedding，query embedding 不变。
-- SFT 需要重新部署模型。
-- NUDGE 不需要 Negative samples，SFT 如果挖掘的难负样本不好，效果不是特别好。
-
-
+TODO