Adapt to cosine and update readme

3_SparseAutoEncoder/clearn_nfcorpus.pth

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:95fadff86747200fd73530b43fcb90fd9584b0c5862c7a3726522ef5b61bc27a
-size 268650896

README.md

@@ -25,39 +25,39 @@ model-index:
           - eng-Latn
       metrics:
         - type: ndcg@1
-          value: 0.55108
+          value: 0.43189
         - type: ndcg@3
-          value: 0.50489
+          value: 0.41132
         - type: ndcg@5
-          value: 0.48476
+          value: 0.40406
         - type: ndcg@10
-          value: 0.45184
+          value: 0.39624
         - type: ndcg@20
-          value: 0.42492
+          value: 0.38517
         - type: ndcg@100
-          value: 0.42262
+          value: 0.40068
         - type: ndcg@1000
-          value: 0.51072
+          value: 0.49126
         - type: map@10
-          value: 0.17877
+          value: 0.14342
         - type: map@100
-          value: 0.23468
+          value: 0.21866
         - type: map@1000
-          value: 0.2548
+          value: 0.2427
         - type: recall@10
-          value: 0.22986
+          value: 0.1968
         - type: recall@100
-          value: 0.43691
+          value: 0.45592
         - type: recall@1000
-          value: 0.76026
+          value: 0.78216
         - type: precision@1
-          value: 0.56347
+          value: 0.45511
         - type: precision@10
-          value: 0.33437
+          value: 0.32353
         - type: mrr@10
-          value: 0.657083
+          value: 0.537792
         - type: main_score
-          value: 0.45184
+          value: 0.39624
       task:
         type: Retrieval
 ---
@@ -75,25 +75,33 @@ We recommend using ``Transformers 4.47.0.``
 You can evaluate this model loaded by Sentence Transformers with the following code snippet:
 ```python
 import mteb
-from sentence_transformers import SentenceTransformer
-model = SentenceTransformer(
-    "Y-Research-Group/CSR-NV_Embed_v2-Retrieval-NFcorpus ",
-    trust_remote_code=True
-)
+from sentence_transformers import SparseEncoder
+
+model = SparseEncoder("Y-Research-Group/CSR-NV_Embed_v2-Retrieval-NFcorpus", trust_remote_code=True)
 model.prompts = {
-  "NFCorpus-query": "Instruct: Given a question, retrieve relevant documents that answer the question\nQuery:"
+    "NFCorpus-query": "Instruct: Given a question, retrieve relevant documents that answer the question\nQuery:"
 }
+
 task = mteb.get_tasks(tasks=["NFCorpus"])
 evaluation = mteb.MTEB(tasks=task)
-evaluation.run(model, eval_splits=["test"], output_folder="./results/NFCorpus", 
-               batch_size=32, show_progress_bar=True)
+evaluation.run(
+    model,
+    eval_splits=["test"],
+    output_folder="./results/NFCorpus",
+    show_progress_bar=True,
+    encode_kwargs={"convert_to_sparse_tensor": False, "batch_size": 8},
+)  # MTEB don't support sparse tensors yet, so we need to convert to dense tensors
 ```
 
 ## Citation
 ```bibtex
-@inproceedings{wenbeyond,
-  title={Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation},
-  author={Wen, Tiansheng and Wang, Yifei and Zeng, Zequn and Peng, Zhong and Su, Yudi and Liu, Xinyang and Chen, Bo and Liu, Hongwei and Jegelka, Stefanie and You, Chenyu},
-  booktitle={Forty-second International Conference on Machine Learning}
+@misc{wen2025matryoshkarevisitingsparsecoding,
+      title={Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation}, 
+      author={Tiansheng Wen and Yifei Wang and Zequn Zeng and Zhong Peng and Yudi Su and Xinyang Liu and Bo Chen and Hongwei Liu and Stefanie Jegelka and Chenyu You},
+      year={2025},
+      eprint={2503.01776},
+      archivePrefix={arXiv},
+      primaryClass={cs.LG},
+      url={https://arxiv.org/abs/2503.01776}, 
 }
 ```

config_sentence_transformers.json

@@ -1,27 +1,15 @@
 {
   "__version__": {
-    "sentence_transformers": "4.2.0",
+    "sentence_transformers": "5.0.0",
     "transformers": "4.47.0",
     "pytorch": "2.5.1+cu12"
   },
   "prompts": {
-    "Banking77Classification": "Instruct: Given a question, please describe the intent of this question. \n Question: ",
-    "MTOPIntentClassification": "Instruct: Given a question, please describe the intent of this question. \n Question: ",
-    "TweetSentimentClassification": "Classify the sentiment of a given tweet as either positive, negative, or neutral.",
-    "BiorxivClusteringP2P.v2": "Identify the main category of Biorxiv papers based on the titles and abstracts",
-    "BiorxivClusteringS2S.v2": "Identify the main category of Biorxiv papers based on the titles",
-    "TwentyNewsgroupsClustering": "Identify the topic or theme of the given news articles",
-    "FiQA2018": {
-      "query": "Given a financial question, retrieve relevant passages that answer the query"
-    },
-    "SciFact": {
-      "query": "Given a scientific claim, retrieve documents that support or refute the claim"
-    },
     "NFCorpus": {
-      "query": "Given a question, retrieve relevant documents that answer the question"
+      "query": "Instruct: Given a question, retrieve relevant documents that answer the question. \n Question: "
     }
   },
   "default_prompt_name": null,
   "model_type": "SparseEncoder",
-  "similarity_fn_name": "dot"
+  "similarity_fn_name": "cosine"
 }