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:111470
 - loss:MultipleNegativesRankingLoss
-base_model: Alibaba-NLP/gte-modernbert-base
+base_model: thenlper/gte-small
 widget:
 - source_sentence: why are some rocks radioactive
   sentences:
@@ -106,7 +106,7 @@ metrics:
 - cosine_mrr@10
 - cosine_map@100
 model-index:
-- name: SentenceTransformer based on Alibaba-NLP/gte-modernbert-base
+- name: SentenceTransformer based on thenlper/gte-small
   results:
   - task:
       type: information-retrieval
@@ -116,49 +116,49 @@ model-index:
       type: NanoMSMARCO
     metrics:
     - type: cosine_accuracy@1
-      value: 0.46
+      value: 0.36
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.66
+      value: 0.54
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.7
+      value: 0.68
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.82
+      value: 0.7
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.46
+      value: 0.36
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.22
+      value: 0.18
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.14
+      value: 0.136
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.08199999999999999
+      value: 0.07
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.46
+      value: 0.36
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.66
+      value: 0.54
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.7
+      value: 0.68
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.82
+      value: 0.7
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.6411634133079402
+      value: 0.5320160584932276
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.5842460317460317
+      value: 0.47755555555555546
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.5911749882485325
+      value: 0.4910009714157892
       name: Cosine Map@100
   - task:
       type: information-retrieval
@@ -168,49 +168,49 @@ model-index:
       type: NanoNQ
     metrics:
     - type: cosine_accuracy@1
-      value: 0.34
+      value: 0.46
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.58
+      value: 0.6
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.66
+      value: 0.62
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.78
+      value: 0.72
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.34
+      value: 0.46
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.19333333333333333
+      value: 0.2133333333333333
       name: Cosine Precision@3
     - type: cosine_precision@5
       value: 0.136
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.08399999999999999
+      value: 0.07800000000000001
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.32
+      value: 0.43
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.54
+      value: 0.58
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.62
+      value: 0.61
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.74
+      value: 0.7
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5340775600556903
+      value: 0.5714045958976129
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.4830238095238096
+      value: 0.5393809523809524
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.4703754816157255
+      value: 0.538206389483671
       name: Cosine Map@100
   - task:
       type: nano-beir
@@ -220,63 +220,63 @@ model-index:
       type: NanoBEIR_mean
     metrics:
     - type: cosine_accuracy@1
-      value: 0.4
+      value: 0.41000000000000003
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
-      value: 0.62
+      value: 0.5700000000000001
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.6799999999999999
+      value: 0.65
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.8
+      value: 0.71
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.4
+      value: 0.41000000000000003
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.20666666666666667
+      value: 0.19666666666666666
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.138
+      value: 0.136
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.08299999999999999
+      value: 0.07400000000000001
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.39
+      value: 0.395
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.6000000000000001
+      value: 0.56
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.6599999999999999
+      value: 0.645
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.78
+      value: 0.7
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5876204866818153
+      value: 0.5517103271954202
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.5336349206349207
+      value: 0.5084682539682539
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.5307752349321291
+      value: 0.5146036804497301
       name: Cosine Map@100
 ---
 
-# SentenceTransformer based on Alibaba-NLP/gte-modernbert-base
+# SentenceTransformer based on thenlper/gte-small
 
-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 [thenlper/gte-small](https://huggingface.co/thenlper/gte-small). 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:** [thenlper/gte-small](https://huggingface.co/thenlper/gte-small) <!-- at revision 17e1f347d17fe144873b1201da91788898c639cd -->
 - **Maximum Sequence Length:** 128 tokens
-- **Output Dimensionality:** 768 dimensions
+- **Output Dimensionality:** 384 dimensions
 - **Similarity Function:** Cosine Similarity
 <!-- - **Training Dataset:** Unknown -->
 <!-- - **Language:** Unknown -->
@@ -292,8 +292,9 @@ 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()
 )
 ```
 
@@ -321,14 +322,14 @@ 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, 1.0000, 0.2374],
-#         [1.0000, 1.0000, 0.2374],
-#         [0.2374, 0.2374, 1.0000]])
+# tensor([[1.0000, 1.0000, 0.9760],
+#         [1.0000, 1.0000, 0.9760],
+#         [0.9760, 0.9760, 1.0000]])
 ```
 
 <!--
@@ -366,21 +367,21 @@ You can finetune this model on your own dataset.
 
 | Metric              | NanoMSMARCO | NanoNQ     |
 |:--------------------|:------------|:-----------|
-| cosine_accuracy@1   | 0.46        | 0.34       |
-| cosine_accuracy@3   | 0.66        | 0.58       |
-| cosine_accuracy@5   | 0.7         | 0.66       |
-| cosine_accuracy@10  | 0.82        | 0.78       |
-| cosine_precision@1  | 0.46        | 0.34       |
-| cosine_precision@3  | 0.22        | 0.1933     |
-| cosine_precision@5  | 0.14        | 0.136      |
-| cosine_precision@10 | 0.082       | 0.084      |
-| cosine_recall@1     | 0.46        | 0.32       |
-| cosine_recall@3     | 0.66        | 0.54       |
-| cosine_recall@5     | 0.7         | 0.62       |
-| cosine_recall@10    | 0.82        | 0.74       |
-| **cosine_ndcg@10**  | **0.6412**  | **0.5341** |
-| cosine_mrr@10       | 0.5842      | 0.483      |
-| cosine_map@100      | 0.5912      | 0.4704     |
+| cosine_accuracy@1   | 0.36        | 0.46       |
+| cosine_accuracy@3   | 0.54        | 0.6        |
+| cosine_accuracy@5   | 0.68        | 0.62       |
+| cosine_accuracy@10  | 0.7         | 0.72       |
+| cosine_precision@1  | 0.36        | 0.46       |
+| cosine_precision@3  | 0.18        | 0.2133     |
+| cosine_precision@5  | 0.136       | 0.136      |
+| cosine_precision@10 | 0.07        | 0.078      |
+| cosine_recall@1     | 0.36        | 0.43       |
+| cosine_recall@3     | 0.54        | 0.58       |
+| cosine_recall@5     | 0.68        | 0.61       |
+| cosine_recall@10    | 0.7         | 0.7        |
+| **cosine_ndcg@10**  | **0.532**   | **0.5714** |
+| cosine_mrr@10       | 0.4776      | 0.5394     |
+| cosine_map@100      | 0.491       | 0.5382     |
 
 #### Nano BEIR
 
@@ -398,21 +399,21 @@ You can finetune this model on your own dataset.
 
 | Metric              | Value      |
 |:--------------------|:-----------|
-| cosine_accuracy@1   | 0.4        |
-| cosine_accuracy@3   | 0.62       |
-| cosine_accuracy@5   | 0.68       |
-| cosine_accuracy@10  | 0.8        |
-| cosine_precision@1  | 0.4        |
-| cosine_precision@3  | 0.2067     |
-| cosine_precision@5  | 0.138      |
-| cosine_precision@10 | 0.083      |
-| cosine_recall@1     | 0.39       |
-| cosine_recall@3     | 0.6        |
-| cosine_recall@5     | 0.66       |
-| cosine_recall@10    | 0.78       |
-| **cosine_ndcg@10**  | **0.5876** |
-| cosine_mrr@10       | 0.5336     |
-| cosine_map@100      | 0.5308     |
+| cosine_accuracy@1   | 0.41       |
+| cosine_accuracy@3   | 0.57       |
+| cosine_accuracy@5   | 0.65       |
+| cosine_accuracy@10  | 0.71       |
+| cosine_precision@1  | 0.41       |
+| cosine_precision@3  | 0.1967     |
+| cosine_precision@5  | 0.136      |
+| cosine_precision@10 | 0.074      |
+| cosine_recall@1     | 0.395      |
+| cosine_recall@3     | 0.56       |
+| cosine_recall@5     | 0.645      |
+| cosine_recall@10    | 0.7        |
+| **cosine_ndcg@10**  | **0.5517** |
+| cosine_mrr@10       | 0.5085     |
+| cosine_map@100      | 0.5146     |
 
 <!--
 ## Bias, Risks and Limitations
@@ -438,7 +439,7 @@ You can finetune this model on your own dataset.
   |         | anchor                                                                            | positive                                                                           | negative                                                                           |
   |:--------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
   | type    | string                                                                            | string                                                                             | string                                                                             |
-  | details | <ul><li>min: 4 tokens</li><li>mean: 11.17 tokens</li><li>max: 59 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 68.53 tokens</li><li>max: 128 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 67.56 tokens</li><li>max: 128 tokens</li></ul> |
+  | details | <ul><li>min: 4 tokens</li><li>mean: 10.95 tokens</li><li>max: 60 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 67.57 tokens</li><li>max: 128 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 66.64 tokens</li><li>max: 128 tokens</li></ul> |
 * Samples:
   | anchor                                                                | positive                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               | negative                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     |
   |:----------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
@@ -464,7 +465,7 @@ You can finetune this model on your own dataset.
   |         | anchor                                                                            | positive                                                                           | negative                                                                           |
   |:--------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
   | type    | string                                                                            | string                                                                             | string                                                                             |
-  | details | <ul><li>min: 4 tokens</li><li>mean: 11.35 tokens</li><li>max: 64 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 68.67 tokens</li><li>max: 128 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 67.03 tokens</li><li>max: 128 tokens</li></ul> |
+  | details | <ul><li>min: 4 tokens</li><li>mean: 11.11 tokens</li><li>max: 66 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 67.99 tokens</li><li>max: 128 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 66.08 tokens</li><li>max: 128 tokens</li></ul> |
 * Samples:
   | anchor                                                                                                    | positive                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                | negative                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      |
   |:----------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
@@ -629,19 +630,19 @@ 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    | -             | 2.5772          | 0.6530                     | 0.6552                | 0.6541                       |
-| 0.2874 | 250  | 1.4443        | 0.8286          | 0.6168                     | 0.5857                | 0.6012                       |
-| 0.5747 | 500  | 0.7965        | 0.7746          | 0.6253                     | 0.5651                | 0.5952                       |
-| 0.8621 | 750  | 0.7676        | 0.7581          | 0.6431                     | 0.5307                | 0.5869                       |
-| 1.1494 | 1000 | 0.7381        | 0.7493          | 0.6446                     | 0.5651                | 0.6048                       |
-| 1.4368 | 1250 | 0.7166        | 0.7433          | 0.6286                     | 0.5574                | 0.5930                       |
-| 1.7241 | 1500 | 0.7115        | 0.7380          | 0.6370                     | 0.5645                | 0.6008                       |
-| 2.0115 | 1750 | 0.7033        | 0.7349          | 0.6316                     | 0.5408                | 0.5862                       |
-| 2.2989 | 2000 | 0.678         | 0.7360          | 0.6380                     | 0.5270                | 0.5825                       |
-| 2.5862 | 2250 | 0.674         | 0.7345          | 0.6236                     | 0.5488                | 0.5862                       |
-| 2.8736 | 2500 | 0.6669        | 0.7330          | 0.6403                     | 0.5472                | 0.5937                       |
-| 3.1609 | 2750 | 0.6641        | 0.7331          | 0.6221                     | 0.5367                | 0.5794                       |
-| 3.4483 | 3000 | 0.6556        | 0.7334          | 0.6412                     | 0.5341                | 0.5876                       |
+| 0      | 0    | -             | 4.0678          | 0.6259                     | 0.6583                | 0.6421                       |
+| 0.2874 | 250  | 3.2728        | 1.0323          | 0.5233                     | 0.4878                | 0.5056                       |
+| 0.5747 | 500  | 1.1175        | 0.8830          | 0.5421                     | 0.5531                | 0.5476                       |
+| 0.8621 | 750  | 1.022         | 0.8583          | 0.5397                     | 0.5777                | 0.5587                       |
+| 1.1494 | 1000 | 0.9806        | 0.8444          | 0.5415                     | 0.5703                | 0.5559                       |
+| 1.4368 | 1250 | 0.966         | 0.8345          | 0.5315                     | 0.5775                | 0.5545                       |
+| 1.7241 | 1500 | 0.952         | 0.8284          | 0.5460                     | 0.5790                | 0.5625                       |
+| 2.0115 | 1750 | 0.9459        | 0.8220          | 0.5338                     | 0.5761                | 0.5549                       |
+| 2.2989 | 2000 | 0.9299        | 0.8186          | 0.5350                     | 0.5720                | 0.5535                       |
+| 2.5862 | 2250 | 0.9254        | 0.8156          | 0.5322                     | 0.5797                | 0.5559                       |
+| 2.8736 | 2500 | 0.9123        | 0.8136          | 0.5322                     | 0.5718                | 0.5520                       |
+| 3.1609 | 2750 | 0.9204        | 0.8120          | 0.5246                     | 0.5714                | 0.5480                       |
+| 3.4483 | 3000 | 0.9119        | 0.8116          | 0.5320                     | 0.5714                | 0.5517                       |
 
 
 ### Framework Versions

config_sentence_transformers.json

@@ -1,4 +1,5 @@
 {
+  "model_type": "SentenceTransformer",
   "__version__": {
     "sentence_transformers": "5.2.0",
     "transformers": "4.57.3",
@@ -9,6 +10,5 @@
     "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"
   }
 ]