Add new SentenceTransformer model

1_Pooling/config.json

@@ -1,7 +1,7 @@
 {
-    "word_embedding_dimension": 384,
-    "pooling_mode_cls_token": false,
-    "pooling_mode_mean_tokens": true,
+    "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,

README.md

@@ -7,7 +7,7 @@ tags:
 - generated_from_trainer
 - dataset_size:89998
 - loss:MultipleNegativesRankingLoss
-base_model: sentence-transformers/all-MiniLM-L12-v2
+base_model: Alibaba-NLP/gte-modernbert-base
 widget:
 - source_sentence: Indian university which follow" international education "type system?
   sentences:
@@ -57,7 +57,7 @@ metrics:
 - cosine_mrr@10
 - cosine_map@100
 model-index:
-- name: SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2
+- name: SentenceTransformer based on Alibaba-NLP/gte-modernbert-base
   results:
   - task:
       type: information-retrieval
@@ -67,49 +67,49 @@ model-index:
       type: NanoMSMARCO
     metrics:
     - type: cosine_accuracy@1
-      value: 0.32
+      value: 0.42
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.56
+      value: 0.66
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.72
+      value: 0.68
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.82
+      value: 0.74
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.32
+      value: 0.42
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.18666666666666665
+      value: 0.22
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.14400000000000002
+      value: 0.136
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.08199999999999999
+      value: 0.07400000000000001
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.32
+      value: 0.42
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.56
+      value: 0.66
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.72
+      value: 0.68
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.82
+      value: 0.74
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5574382679738011
+      value: 0.593377048050137
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.4747460317460317
+      value: 0.5453888888888889
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.4820380014583416
+      value: 0.5590741871418341
       name: Cosine Map@100
   - task:
       type: information-retrieval
@@ -119,7 +119,7 @@ model-index:
       type: NanoNQ
     metrics:
     - type: cosine_accuracy@1
-      value: 0.32
+      value: 0.38
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
       value: 0.54
@@ -128,40 +128,40 @@ model-index:
       value: 0.62
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.68
+      value: 0.72
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.32
+      value: 0.38
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.19333333333333333
+      value: 0.18666666666666665
       name: Cosine Precision@3
     - type: cosine_precision@5
       value: 0.132
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.07200000000000001
+      value: 0.08
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.31
+      value: 0.34
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.53
+      value: 0.51
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.6
+      value: 0.58
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.66
+      value: 0.7
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.492580214786822
+      value: 0.5235400236111211
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.4418809523809524
+      value: 0.4836031746031746
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.4462290155539738
+      value: 0.4659949769889572
       name: Cosine Map@100
   - task:
       type: nano-beir
@@ -171,63 +171,63 @@ model-index:
       type: NanoBEIR_mean
     metrics:
     - type: cosine_accuracy@1
-      value: 0.32
+      value: 0.4
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.55
+      value: 0.6000000000000001
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.6699999999999999
+      value: 0.65
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.75
+      value: 0.73
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.32
+      value: 0.4
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.19
+      value: 0.2033333333333333
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.138
+      value: 0.134
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.077
+      value: 0.07700000000000001
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.315
+      value: 0.38
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.545
+      value: 0.585
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.6599999999999999
+      value: 0.63
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.74
+      value: 0.72
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5250092413803116
+      value: 0.5584585358306291
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.45831349206349203
+      value: 0.5144960317460318
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.4641335085061577
+      value: 0.5125345820653957
       name: Cosine Map@100
 ---
 
-# SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2
+# SentenceTransformer based on Alibaba-NLP/gte-modernbert-base
 
-This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/all-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L12-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.
+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.
 
 ## Model Details
 
 ### Model Description
 - **Model Type:** Sentence Transformer
-- **Base model:** [sentence-transformers/all-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L12-v2) <!-- at revision 936af83a2ecce5fe87a09109ff5cbcefe073173a -->
+- **Base model:** [Alibaba-NLP/gte-modernbert-base](https://huggingface.co/Alibaba-NLP/gte-modernbert-base) <!-- at revision e7f32e3c00f91d699e8c43b53106206bcc72bb22 -->
 - **Maximum Sequence Length:** 128 tokens
-- **Output Dimensionality:** 384 dimensions
+- **Output Dimensionality:** 768 dimensions
 - **Similarity Function:** Cosine Similarity
 <!-- - **Training Dataset:** Unknown -->
 <!-- - **Language:** Unknown -->
@@ -243,9 +243,8 @@ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [s
 
 ```
 SentenceTransformer(
-  (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()
+  (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})
 )
 ```
 
@@ -273,14 +272,14 @@ sentences = [
 ]
 embeddings = model.encode(sentences)
 print(embeddings.shape)
-# [3, 384]
+# [3, 768]
 
 # Get the similarity scores for the embeddings
 similarities = model.similarity(embeddings, embeddings)
 print(similarities)
-# tensor([[ 1.0000,  0.8663,  0.0078],
-#         [ 0.8663,  1.0000, -0.0501],
-#         [ 0.0078, -0.0501,  1.0000]])
+# tensor([[1.0000, 0.5622, 0.1084],
+#         [0.5622, 1.0000, 0.0196],
+#         [0.1084, 0.0196, 1.0000]])
 ```
 
 <!--
@@ -318,21 +317,21 @@ You can finetune this model on your own dataset.
 
 | Metric              | NanoMSMARCO | NanoNQ     |
 |:--------------------|:------------|:-----------|
-| cosine_accuracy@1   | 0.32        | 0.32       |
-| cosine_accuracy@3   | 0.56        | 0.54       |
-| cosine_accuracy@5   | 0.72        | 0.62       |
-| cosine_accuracy@10  | 0.82        | 0.68       |
-| cosine_precision@1  | 0.32        | 0.32       |
-| cosine_precision@3  | 0.1867      | 0.1933     |
-| cosine_precision@5  | 0.144       | 0.132      |
-| cosine_precision@10 | 0.082       | 0.072      |
-| cosine_recall@1     | 0.32        | 0.31       |
-| cosine_recall@3     | 0.56        | 0.53       |
-| cosine_recall@5     | 0.72        | 0.6        |
-| cosine_recall@10    | 0.82        | 0.66       |
-| **cosine_ndcg@10**  | **0.5574**  | **0.4926** |
-| cosine_mrr@10       | 0.4747      | 0.4419     |
-| cosine_map@100      | 0.482       | 0.4462     |
+| cosine_accuracy@1   | 0.42        | 0.38       |
+| cosine_accuracy@3   | 0.66        | 0.54       |
+| cosine_accuracy@5   | 0.68        | 0.62       |
+| cosine_accuracy@10  | 0.74        | 0.72       |
+| cosine_precision@1  | 0.42        | 0.38       |
+| cosine_precision@3  | 0.22        | 0.1867     |
+| cosine_precision@5  | 0.136       | 0.132      |
+| cosine_precision@10 | 0.074       | 0.08       |
+| cosine_recall@1     | 0.42        | 0.34       |
+| cosine_recall@3     | 0.66        | 0.51       |
+| cosine_recall@5     | 0.68        | 0.58       |
+| cosine_recall@10    | 0.74        | 0.7        |
+| **cosine_ndcg@10**  | **0.5934**  | **0.5235** |
+| cosine_mrr@10       | 0.5454      | 0.4836     |
+| cosine_map@100      | 0.5591      | 0.466      |
 
 #### Nano BEIR
 
@@ -348,23 +347,23 @@ You can finetune this model on your own dataset.
   }
   ```
 
-| Metric              | Value     |
-|:--------------------|:----------|
-| cosine_accuracy@1   | 0.32      |
-| cosine_accuracy@3   | 0.55      |
-| cosine_accuracy@5   | 0.67      |
-| cosine_accuracy@10  | 0.75      |
-| cosine_precision@1  | 0.32      |
-| cosine_precision@3  | 0.19      |
-| cosine_precision@5  | 0.138     |
-| cosine_precision@10 | 0.077     |
-| cosine_recall@1     | 0.315     |
-| cosine_recall@3     | 0.545     |
-| cosine_recall@5     | 0.66      |
-| cosine_recall@10    | 0.74      |
-| **cosine_ndcg@10**  | **0.525** |
-| cosine_mrr@10       | 0.4583    |
-| cosine_map@100      | 0.4641    |
+| Metric              | Value      |
+|:--------------------|:-----------|
+| cosine_accuracy@1   | 0.4        |
+| cosine_accuracy@3   | 0.6        |
+| cosine_accuracy@5   | 0.65       |
+| cosine_accuracy@10  | 0.73       |
+| cosine_precision@1  | 0.4        |
+| cosine_precision@3  | 0.2033     |
+| cosine_precision@5  | 0.134      |
+| cosine_precision@10 | 0.077      |
+| cosine_recall@1     | 0.38       |
+| cosine_recall@3     | 0.585      |
+| cosine_recall@5     | 0.63       |
+| cosine_recall@10    | 0.72       |
+| **cosine_ndcg@10**  | **0.5585** |
+| cosine_mrr@10       | 0.5145     |
+| cosine_map@100      | 0.5125     |
 
 <!--
 ## Bias, Risks and Limitations
@@ -390,7 +389,7 @@ You can finetune this model on your own dataset.
   |         | anchor                                                                            | positive                                                                          | negative                                                                          |
   |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
   | type    | string                                                                            | string                                                                            | string                                                                            |
-  | details | <ul><li>min: 5 tokens</li><li>mean: 15.61 tokens</li><li>max: 71 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 15.72 tokens</li><li>max: 71 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 16.55 tokens</li><li>max: 67 tokens</li></ul> |
+  | details | <ul><li>min: 5 tokens</li><li>mean: 15.58 tokens</li><li>max: 67 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 15.69 tokens</li><li>max: 67 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 16.51 tokens</li><li>max: 71 tokens</li></ul> |
 * Samples:
   | anchor                                                                                                     | positive                                                                                                   | negative                                                                                                           |
   |:-----------------------------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------|
@@ -416,7 +415,7 @@ You can finetune this model on your own dataset.
   |         | anchor                                                                            | positive                                                                          | negative                                                                          |
   |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
   | type    | string                                                                            | string                                                                            | string                                                                            |
-  | details | <ul><li>min: 3 tokens</li><li>mean: 15.75 tokens</li><li>max: 65 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 15.86 tokens</li><li>max: 65 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 16.66 tokens</li><li>max: 74 tokens</li></ul> |
+  | details | <ul><li>min: 4 tokens</li><li>mean: 15.73 tokens</li><li>max: 64 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 15.83 tokens</li><li>max: 64 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 16.59 tokens</li><li>max: 72 tokens</li></ul> |
 * Samples:
   | anchor                                                                    | positive                                                                  | negative                                                            |
   |:--------------------------------------------------------------------------|:--------------------------------------------------------------------------|:--------------------------------------------------------------------|
@@ -581,19 +580,19 @@ 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    | -             | 0.5972          | 0.5887                     | 0.5786                | 0.5836                       |
-| 0.3556 | 250  | 0.5902        | 0.4140          | 0.5596                     | 0.5395                | 0.5495                       |
-| 0.7112 | 500  | 0.5168        | 0.4000          | 0.5798                     | 0.5206                | 0.5502                       |
-| 1.0669 | 750  | 0.4977        | 0.3934          | 0.5722                     | 0.5079                | 0.5401                       |
-| 1.4225 | 1000 | 0.4825        | 0.3875          | 0.5612                     | 0.5129                | 0.5370                       |
-| 1.7781 | 1250 | 0.4764        | 0.3843          | 0.5734                     | 0.5179                | 0.5457                       |
-| 2.1337 | 1500 | 0.4672        | 0.3821          | 0.5740                     | 0.5065                | 0.5402                       |
-| 2.4893 | 1750 | 0.4612        | 0.3804          | 0.5721                     | 0.4950                | 0.5335                       |
-| 2.8450 | 2000 | 0.4576        | 0.3791          | 0.5588                     | 0.4836                | 0.5212                       |
-| 3.2006 | 2250 | 0.4533        | 0.3775          | 0.5550                     | 0.5005                | 0.5278                       |
-| 3.5562 | 2500 | 0.4491        | 0.3770          | 0.5604                     | 0.4919                | 0.5262                       |
-| 3.9118 | 2750 | 0.4483        | 0.3763          | 0.5569                     | 0.4897                | 0.5233                       |
-| 4.2674 | 3000 | 0.446         | 0.3760          | 0.5574                     | 0.4926                | 0.5250                       |
+| 0      | 0    | -             | 2.1820          | 0.6530                     | 0.6552                | 0.6541                       |
+| 0.3556 | 250  | 0.8291        | 0.3946          | 0.6117                     | 0.6214                | 0.6165                       |
+| 0.7112 | 500  | 0.3866        | 0.3722          | 0.5998                     | 0.6314                | 0.6156                       |
+| 1.0669 | 750  | 0.3687        | 0.3623          | 0.6157                     | 0.6044                | 0.6100                       |
+| 1.4225 | 1000 | 0.3451        | 0.3579          | 0.6192                     | 0.5948                | 0.6070                       |
+| 1.7781 | 1250 | 0.3418        | 0.3542          | 0.6013                     | 0.5955                | 0.5984                       |
+| 2.1337 | 1500 | 0.3303        | 0.3567          | 0.6080                     | 0.5532                | 0.5806                       |
+| 2.4893 | 1750 | 0.3158        | 0.3548          | 0.6038                     | 0.5440                | 0.5739                       |
+| 2.8450 | 2000 | 0.3136        | 0.3532          | 0.6015                     | 0.5497                | 0.5756                       |
+| 3.2006 | 2250 | 0.3056        | 0.3571          | 0.6015                     | 0.5356                | 0.5686                       |
+| 3.5562 | 2500 | 0.2983        | 0.3575          | 0.6052                     | 0.5321                | 0.5686                       |
+| 3.9118 | 2750 | 0.2973        | 0.3572          | 0.5934                     | 0.5231                | 0.5582                       |
+| 4.2674 | 3000 | 0.2933        | 0.3596          | 0.5934                     | 0.5235                | 0.5585                       |
 
 
 ### 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"
+  "similarity_fn_name": "cosine",
+  "model_type": "SentenceTransformer"
 }

modules.json

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