Add new SentenceTransformer model

1_Pooling/config.json

@@ -1,7 +1,7 @@
 {
-    "word_embedding_dimension": 768,
-    "pooling_mode_cls_token": true,
-    "pooling_mode_mean_tokens": false,
+    "word_embedding_dimension": 384,
+    "pooling_mode_cls_token": false,
+    "pooling_mode_mean_tokens": true,
     "pooling_mode_max_tokens": false,
     "pooling_mode_mean_sqrt_len_tokens": false,
     "pooling_mode_weightedmean_tokens": false,

README.md

@@ -7,7 +7,7 @@ tags:
 - generated_from_trainer
 - dataset_size:359997
 - loss:MultipleNegativesRankingLoss
-base_model: Alibaba-NLP/gte-modernbert-base
+base_model: sentence-transformers/all-MiniLM-L6-v2
 widget:
 - source_sentence: When do you use Ms. or Mrs.? Is one for a married woman and one
     for one that's not married? Which one is for what?
@@ -60,7 +60,7 @@ metrics:
 - cosine_mrr@10
 - cosine_map@100
 model-index:
-- name: SentenceTransformer based on Alibaba-NLP/gte-modernbert-base
+- name: SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2
   results:
   - task:
       type: information-retrieval
@@ -70,49 +70,49 @@ model-index:
       type: NanoMSMARCO
     metrics:
     - type: cosine_accuracy@1
-      value: 0.38
+      value: 0.22
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.62
+      value: 0.5
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.72
+      value: 0.62
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.78
+      value: 0.74
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.38
+      value: 0.22
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.20666666666666667
+      value: 0.16666666666666663
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.14400000000000002
+      value: 0.124
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.078
+      value: 0.07400000000000001
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.38
+      value: 0.22
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.62
+      value: 0.5
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.72
+      value: 0.62
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.78
+      value: 0.74
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5792677770404034
+      value: 0.47667177266958005
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.5150238095238094
+      value: 0.39240476190476187
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.5260186479155519
+      value: 0.406991563991564
       name: Cosine Map@100
   - task:
       type: information-retrieval
@@ -122,49 +122,49 @@ model-index:
       type: NanoNQ
     metrics:
     - type: cosine_accuracy@1
-      value: 0.38
+      value: 0.28
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.58
+      value: 0.46
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.66
+      value: 0.56
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.74
+      value: 0.64
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.38
+      value: 0.28
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.2
+      value: 0.15999999999999998
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.14
+      value: 0.11600000000000002
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.078
+      value: 0.066
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.36
+      value: 0.27
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.54
+      value: 0.45
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.62
+      value: 0.54
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.7
+      value: 0.61
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5417937853620868
+      value: 0.4442430372694745
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.5033571428571428
+      value: 0.39785714285714285
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.4942594774374801
+      value: 0.39869586832265574
       name: Cosine Map@100
   - task:
       type: nano-beir
@@ -174,63 +174,63 @@ model-index:
       type: NanoBEIR_mean
     metrics:
     - type: cosine_accuracy@1
-      value: 0.38
+      value: 0.25
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.6
+      value: 0.48
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.69
+      value: 0.5900000000000001
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.76
+      value: 0.69
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.38
+      value: 0.25
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.20333333333333334
+      value: 0.1633333333333333
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.14200000000000002
+      value: 0.12000000000000001
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.078
+      value: 0.07
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.37
+      value: 0.245
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.5800000000000001
+      value: 0.475
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.6699999999999999
+      value: 0.5800000000000001
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.74
+      value: 0.675
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5605307812012451
+      value: 0.46045740496952725
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.5091904761904762
+      value: 0.39513095238095236
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.510139062676516
+      value: 0.4028437161571099
       name: Cosine Map@100
 ---
 
-# SentenceTransformer based on Alibaba-NLP/gte-modernbert-base
+# SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2
 
-This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Alibaba-NLP/gte-modernbert-base](https://huggingface.co/Alibaba-NLP/gte-modernbert-base). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
+This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2). It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
 
 ## Model Details
 
 ### Model Description
 - **Model Type:** Sentence Transformer
-- **Base model:** [Alibaba-NLP/gte-modernbert-base](https://huggingface.co/Alibaba-NLP/gte-modernbert-base) <!-- at revision e7f32e3c00f91d699e8c43b53106206bcc72bb22 -->
+- **Base model:** [sentence-transformers/all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2) <!-- at revision c9745ed1d9f207416be6d2e6f8de32d1f16199bf -->
 - **Maximum Sequence Length:** 128 tokens
-- **Output Dimensionality:** 768 dimensions
+- **Output Dimensionality:** 384 dimensions
 - **Similarity Function:** Cosine Similarity
 <!-- - **Training Dataset:** Unknown -->
 <!-- - **Language:** Unknown -->
@@ -246,8 +246,9 @@ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [A
 
 ```
 SentenceTransformer(
-  (0): Transformer({'max_seq_length': 128, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
-  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
+  (0): Transformer({'max_seq_length': 128, 'do_lower_case': False, 'architecture': 'BertModel'})
+  (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
+  (2): Normalize()
 )
 ```
 
@@ -275,14 +276,14 @@ sentences = [
 ]
 embeddings = model.encode(sentences)
 print(embeddings.shape)
-# [3, 768]
+# [3, 384]
 
 # Get the similarity scores for the embeddings
 similarities = model.similarity(embeddings, embeddings)
 print(similarities)
-# tensor([[ 1.0000,  0.9926, -0.0086],
-#         [ 0.9926,  1.0000, -0.0135],
-#         [-0.0086, -0.0135,  1.0000]])
+# tensor([[1.0000, 0.9894, 0.0074],
+#         [0.9894, 1.0000, 0.0136],
+#         [0.0074, 0.0136, 1.0000]])
 ```
 
 <!--
@@ -320,21 +321,21 @@ You can finetune this model on your own dataset.
 
 | Metric              | NanoMSMARCO | NanoNQ     |
 |:--------------------|:------------|:-----------|
-| cosine_accuracy@1   | 0.38        | 0.38       |
-| cosine_accuracy@3   | 0.62        | 0.58       |
-| cosine_accuracy@5   | 0.72        | 0.66       |
-| cosine_accuracy@10  | 0.78        | 0.74       |
-| cosine_precision@1  | 0.38        | 0.38       |
-| cosine_precision@3  | 0.2067      | 0.2        |
-| cosine_precision@5  | 0.144       | 0.14       |
-| cosine_precision@10 | 0.078       | 0.078      |
-| cosine_recall@1     | 0.38        | 0.36       |
-| cosine_recall@3     | 0.62        | 0.54       |
-| cosine_recall@5     | 0.72        | 0.62       |
-| cosine_recall@10    | 0.78        | 0.7        |
-| **cosine_ndcg@10**  | **0.5793**  | **0.5418** |
-| cosine_mrr@10       | 0.515       | 0.5034     |
-| cosine_map@100      | 0.526       | 0.4943     |
+| cosine_accuracy@1   | 0.22        | 0.28       |
+| cosine_accuracy@3   | 0.5         | 0.46       |
+| cosine_accuracy@5   | 0.62        | 0.56       |
+| cosine_accuracy@10  | 0.74        | 0.64       |
+| cosine_precision@1  | 0.22        | 0.28       |
+| cosine_precision@3  | 0.1667      | 0.16       |
+| cosine_precision@5  | 0.124       | 0.116      |
+| cosine_precision@10 | 0.074       | 0.066      |
+| cosine_recall@1     | 0.22        | 0.27       |
+| cosine_recall@3     | 0.5         | 0.45       |
+| cosine_recall@5     | 0.62        | 0.54       |
+| cosine_recall@10    | 0.74        | 0.61       |
+| **cosine_ndcg@10**  | **0.4767**  | **0.4442** |
+| cosine_mrr@10       | 0.3924      | 0.3979     |
+| cosine_map@100      | 0.407       | 0.3987     |
 
 #### Nano BEIR
 
@@ -352,21 +353,21 @@ You can finetune this model on your own dataset.
 
 | Metric              | Value      |
 |:--------------------|:-----------|
-| cosine_accuracy@1   | 0.38       |
-| cosine_accuracy@3   | 0.6        |
-| cosine_accuracy@5   | 0.69       |
-| cosine_accuracy@10  | 0.76       |
-| cosine_precision@1  | 0.38       |
-| cosine_precision@3  | 0.2033     |
-| cosine_precision@5  | 0.142      |
-| cosine_precision@10 | 0.078      |
-| cosine_recall@1     | 0.37       |
-| cosine_recall@3     | 0.58       |
-| cosine_recall@5     | 0.67       |
-| cosine_recall@10    | 0.74       |
-| **cosine_ndcg@10**  | **0.5605** |
-| cosine_mrr@10       | 0.5092     |
-| cosine_map@100      | 0.5101     |
+| cosine_accuracy@1   | 0.25       |
+| cosine_accuracy@3   | 0.48       |
+| cosine_accuracy@5   | 0.59       |
+| cosine_accuracy@10  | 0.69       |
+| cosine_precision@1  | 0.25       |
+| cosine_precision@3  | 0.1633     |
+| cosine_precision@5  | 0.12       |
+| cosine_precision@10 | 0.07       |
+| cosine_recall@1     | 0.245      |
+| cosine_recall@3     | 0.475      |
+| cosine_recall@5     | 0.58       |
+| cosine_recall@10    | 0.675      |
+| **cosine_ndcg@10**  | **0.4605** |
+| cosine_mrr@10       | 0.3951     |
+| cosine_map@100      | 0.4028     |
 
 <!--
 ## Bias, Risks and Limitations
@@ -389,10 +390,10 @@ You can finetune this model on your own dataset.
 * Size: 359,997 training samples
 * Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
 * Approximate statistics based on the first 1000 samples:
-  |         | anchor                                                                           | positive                                                                          | negative                                                                          |
-  |:--------|:---------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
-  | type    | string                                                                           | string                                                                            | string                                                                            |
-  | details | <ul><li>min: 4 tokens</li><li>mean: 15.4 tokens</li><li>max: 47 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 15.47 tokens</li><li>max: 47 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 16.9 tokens</li><li>max: 125 tokens</li></ul> |
+  |         | anchor                                                                            | positive                                                                          | negative                                                                           |
+  |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
+  | type    | string                                                                            | string                                                                            | string                                                                             |
+  | details | <ul><li>min: 4 tokens</li><li>mean: 15.46 tokens</li><li>max: 49 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 15.52 tokens</li><li>max: 49 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 16.99 tokens</li><li>max: 128 tokens</li></ul> |
 * Samples:
   | anchor                                                                          | positive                                                                                             | negative                                                                                                                 |
   |:--------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------|
@@ -418,7 +419,7 @@ You can finetune this model on your own dataset.
   |         | anchor                                                                            | positive                                                                          | negative                                                                          |
   |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
   | type    | string                                                                            | string                                                                            | string                                                                            |
-  | details | <ul><li>min: 6 tokens</li><li>mean: 15.68 tokens</li><li>max: 72 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 15.75 tokens</li><li>max: 72 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 16.95 tokens</li><li>max: 78 tokens</li></ul> |
+  | details | <ul><li>min: 6 tokens</li><li>mean: 15.71 tokens</li><li>max: 65 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 15.79 tokens</li><li>max: 65 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 16.97 tokens</li><li>max: 78 tokens</li></ul> |
 * Samples:
   | anchor                                                                                                            | positive                                                                                                          | negative                                                                                                                                  |
   |:------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------|
@@ -583,27 +584,27 @@ You can finetune this model on your own dataset.
 ### Training Logs
 | Epoch  | Step | Training Loss | Validation Loss | NanoMSMARCO_cosine_ndcg@10 | NanoNQ_cosine_ndcg@10 | NanoBEIR_mean_cosine_ndcg@10 |
 |:------:|:----:|:-------------:|:---------------:|:--------------------------:|:---------------------:|:----------------------------:|
-| 0      | 0    | -             | 2.1886          | 0.6530                     | 0.6552                | 0.6541                       |
-| 0.0889 | 250  | 0.9475        | 0.4116          | 0.6233                     | 0.6439                | 0.6336                       |
-| 0.1778 | 500  | 0.3963        | 0.3836          | 0.6137                     | 0.6372                | 0.6254                       |
-| 0.2667 | 750  | 0.3776        | 0.3707          | 0.6243                     | 0.6259                | 0.6251                       |
-| 0.3556 | 1000 | 0.3675        | 0.3638          | 0.6250                     | 0.5981                | 0.6116                       |
-| 0.4445 | 1250 | 0.358         | 0.3581          | 0.6170                     | 0.6045                | 0.6108                       |
-| 0.5334 | 1500 | 0.3575        | 0.3544          | 0.6049                     | 0.5821                | 0.5935                       |
-| 0.6223 | 1750 | 0.3521        | 0.3513          | 0.5835                     | 0.5619                | 0.5727                       |
-| 0.7112 | 2000 | 0.3489        | 0.3486          | 0.5955                     | 0.5576                | 0.5765                       |
-| 0.8001 | 2250 | 0.3465        | 0.3463          | 0.6037                     | 0.5786                | 0.5911                       |
-| 0.8890 | 2500 | 0.3461        | 0.3440          | 0.5884                     | 0.5691                | 0.5788                       |
-| 0.9780 | 2750 | 0.3446        | 0.3428          | 0.5809                     | 0.5627                | 0.5718                       |
-| 1.0669 | 3000 | 0.328         | 0.3423          | 0.5701                     | 0.5599                | 0.5650                       |
-| 1.1558 | 3250 | 0.3235        | 0.3416          | 0.5691                     | 0.5419                | 0.5555                       |
-| 1.2447 | 3500 | 0.3221        | 0.3406          | 0.5694                     | 0.5534                | 0.5614                       |
-| 1.3336 | 3750 | 0.3221        | 0.3397          | 0.5736                     | 0.5519                | 0.5628                       |
-| 1.4225 | 4000 | 0.3196        | 0.3391          | 0.5811                     | 0.5416                | 0.5613                       |
-| 1.5114 | 4250 | 0.3201        | 0.3386          | 0.5525                     | 0.5538                | 0.5532                       |
-| 1.6003 | 4500 | 0.321         | 0.3384          | 0.5801                     | 0.5380                | 0.5591                       |
-| 1.6892 | 4750 | 0.3192        | 0.3382          | 0.5799                     | 0.5474                | 0.5636                       |
-| 1.7781 | 5000 | 0.3203        | 0.3379          | 0.5793                     | 0.5418                | 0.5605                       |
+| 0      | 0    | -             | 0.5501          | 0.5540                     | 0.5931                | 0.5735                       |
+| 0.0889 | 250  | 0.6218        | 0.4360          | 0.5499                     | 0.5725                | 0.5612                       |
+| 0.1778 | 500  | 0.557         | 0.4231          | 0.5414                     | 0.5239                | 0.5326                       |
+| 0.2667 | 750  | 0.5359        | 0.4146          | 0.5188                     | 0.5189                | 0.5188                       |
+| 0.3556 | 1000 | 0.5213        | 0.4095          | 0.4998                     | 0.5138                | 0.5068                       |
+| 0.4445 | 1250 | 0.51          | 0.4058          | 0.5021                     | 0.4988                | 0.5005                       |
+| 0.5334 | 1500 | 0.5086        | 0.4030          | 0.5040                     | 0.4970                | 0.5005                       |
+| 0.6223 | 1750 | 0.5031        | 0.4002          | 0.4963                     | 0.4997                | 0.4980                       |
+| 0.7112 | 2000 | 0.4964        | 0.3979          | 0.5033                     | 0.4880                | 0.4956                       |
+| 0.8001 | 2250 | 0.4927        | 0.3960          | 0.5077                     | 0.4881                | 0.4979                       |
+| 0.8890 | 2500 | 0.4925        | 0.3946          | 0.4939                     | 0.4826                | 0.4882                       |
+| 0.9780 | 2750 | 0.4889        | 0.3936          | 0.4953                     | 0.4778                | 0.4865                       |
+| 1.0669 | 3000 | 0.4819        | 0.3917          | 0.4838                     | 0.4723                | 0.4781                       |
+| 1.1558 | 3250 | 0.4798        | 0.3910          | 0.4900                     | 0.4587                | 0.4743                       |
+| 1.2447 | 3500 | 0.4773        | 0.3905          | 0.4888                     | 0.4557                | 0.4723                       |
+| 1.3336 | 3750 | 0.476         | 0.3899          | 0.4782                     | 0.4512                | 0.4647                       |
+| 1.4225 | 4000 | 0.4738        | 0.3891          | 0.4873                     | 0.4508                | 0.4691                       |
+| 1.5114 | 4250 | 0.4727        | 0.3887          | 0.4849                     | 0.4464                | 0.4657                       |
+| 1.6003 | 4500 | 0.4737        | 0.3887          | 0.4772                     | 0.4482                | 0.4627                       |
+| 1.6892 | 4750 | 0.4722        | 0.3884          | 0.4810                     | 0.4432                | 0.4621                       |
+| 1.7781 | 5000 | 0.4739        | 0.3883          | 0.4767                     | 0.4442                | 0.4605                       |
 
 
 ### Framework Versions

config_sentence_transformers.json

@@ -4,11 +4,11 @@
     "transformers": "4.57.3",
     "pytorch": "2.9.1+cu128"
   },
+  "model_type": "SentenceTransformer",
   "prompts": {
     "query": "",
     "document": ""
   },
   "default_prompt_name": null,
-  "similarity_fn_name": "cosine",
-  "model_type": "SentenceTransformer"
+  "similarity_fn_name": "cosine"
 }

modules.json

@@ -10,5 +10,11 @@
     "name": "1",
     "path": "1_Pooling",
     "type": "sentence_transformers.models.Pooling"
+  },
+  {
+    "idx": 2,
+    "name": "2",
+    "path": "2_Normalize",
+    "type": "sentence_transformers.models.Normalize"
   }
 ]