Add new SentenceTransformer model

README.md

@@ -7,7 +7,7 @@ tags:
 - generated_from_trainer
 - dataset_size:111470
 - loss:MultipleNegativesRankingLoss
-base_model: sentence-transformers/all-MiniLM-L6-v2
+base_model: sentence-transformers/all-MiniLM-L12-v2
 widget:
 - source_sentence: when was the first elephant brought to america
   sentences:
@@ -132,7 +132,7 @@ metrics:
 - cosine_mrr@10
 - cosine_map@100
 model-index:
-- name: SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2
+- name: SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2
   results:
   - task:
       type: information-retrieval
@@ -142,49 +142,49 @@ model-index:
       type: NanoMSMARCO
     metrics:
     - type: cosine_accuracy@1
-      value: 0.32
+      value: 0.34
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
       value: 0.5
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.6
+      value: 0.66
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.76
+      value: 0.78
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.32
+      value: 0.34
       name: Cosine Precision@1
     - type: cosine_precision@3
       value: 0.16666666666666663
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.12000000000000002
+      value: 0.132
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.07600000000000001
+      value: 0.078
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.32
+      value: 0.34
       name: Cosine Recall@1
     - type: cosine_recall@3
       value: 0.5
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.6
+      value: 0.66
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.76
+      value: 0.78
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5174146339399069
+      value: 0.5446770528863051
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.4427063492063491
+      value: 0.4708571428571428
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.452501292753926
+      value: 0.47884258431632043
       name: Cosine Map@100
   - task:
       type: information-retrieval
@@ -194,49 +194,49 @@ model-index:
       type: NanoNQ
     metrics:
     - type: cosine_accuracy@1
-      value: 0.54
+      value: 0.5
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
       value: 0.66
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.68
+      value: 0.7
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.74
+      value: 0.78
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.54
+      value: 0.5
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.22
+      value: 0.22666666666666668
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.136
+      value: 0.14400000000000002
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.08
+      value: 0.08199999999999999
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.51
+      value: 0.48
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.62
+      value: 0.64
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.64
+      value: 0.67
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.72
+      value: 0.74
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.6171839770040762
+      value: 0.6136402968638738
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.6030555555555556
+      value: 0.5821666666666667
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.5845310002947148
+      value: 0.5768526974820034
       name: Cosine Map@100
   - task:
       type: nano-beir
@@ -246,61 +246,61 @@ model-index:
       type: NanoBEIR_mean
     metrics:
     - type: cosine_accuracy@1
-      value: 0.43000000000000005
+      value: 0.42000000000000004
       name: Cosine Accuracy@1
     - type: cosine_accuracy@3
       value: 0.5800000000000001
       name: Cosine Accuracy@3
     - type: cosine_accuracy@5
-      value: 0.64
+      value: 0.6799999999999999
       name: Cosine Accuracy@5
     - type: cosine_accuracy@10
-      value: 0.75
+      value: 0.78
       name: Cosine Accuracy@10
     - type: cosine_precision@1
-      value: 0.43000000000000005
+      value: 0.42000000000000004
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.1933333333333333
+      value: 0.19666666666666666
       name: Cosine Precision@3
     - type: cosine_precision@5
-      value: 0.128
+      value: 0.138
       name: Cosine Precision@5
     - type: cosine_precision@10
-      value: 0.07800000000000001
+      value: 0.07999999999999999
       name: Cosine Precision@10
     - type: cosine_recall@1
-      value: 0.41500000000000004
+      value: 0.41000000000000003
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.56
+      value: 0.5700000000000001
       name: Cosine Recall@3
     - type: cosine_recall@5
-      value: 0.62
+      value: 0.665
       name: Cosine Recall@5
     - type: cosine_recall@10
-      value: 0.74
+      value: 0.76
       name: Cosine Recall@10
     - type: cosine_ndcg@10
-      value: 0.5672993054719916
+      value: 0.5791586748750894
       name: Cosine Ndcg@10
     - type: cosine_mrr@10
-      value: 0.5228809523809523
+      value: 0.5265119047619048
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.5185161465243204
+      value: 0.5278476408991619
       name: Cosine Map@100
 ---
 
-# SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2
+# SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2
 
-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.
+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.
 
 ## Model Details
 
 ### Model Description
 - **Model Type:** Sentence Transformer
-- **Base model:** [sentence-transformers/all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2) <!-- at revision c9745ed1d9f207416be6d2e6f8de32d1f16199bf -->
+- **Base model:** [sentence-transformers/all-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L12-v2) <!-- at revision 936af83a2ecce5fe87a09109ff5cbcefe073173a -->
 - **Maximum Sequence Length:** 128 tokens
 - **Output Dimensionality:** 384 dimensions
 - **Similarity Function:** Cosine Similarity
@@ -353,9 +353,9 @@ print(embeddings.shape)
 # Get the similarity scores for the embeddings
 similarities = model.similarity(embeddings, embeddings)
 print(similarities)
-# tensor([[1.0000, 1.0000, 0.8824],
-#         [1.0000, 1.0000, 0.8824],
-#         [0.8824, 0.8824, 1.0000]])
+# tensor([[1.0000, 1.0000, 0.8845],
+#         [1.0000, 1.0000, 0.8845],
+#         [0.8845, 0.8845, 1.0000]])
 ```
 
 <!--
@@ -393,21 +393,21 @@ You can finetune this model on your own dataset.
 
 | Metric              | NanoMSMARCO | NanoNQ     |
 |:--------------------|:------------|:-----------|
-| cosine_accuracy@1   | 0.32        | 0.54       |
+| cosine_accuracy@1   | 0.34        | 0.5        |
 | cosine_accuracy@3   | 0.5         | 0.66       |
-| cosine_accuracy@5   | 0.6         | 0.68       |
-| cosine_accuracy@10  | 0.76        | 0.74       |
-| cosine_precision@1  | 0.32        | 0.54       |
-| cosine_precision@3  | 0.1667      | 0.22       |
-| cosine_precision@5  | 0.12        | 0.136      |
-| cosine_precision@10 | 0.076       | 0.08       |
-| cosine_recall@1     | 0.32        | 0.51       |
-| cosine_recall@3     | 0.5         | 0.62       |
-| cosine_recall@5     | 0.6         | 0.64       |
-| cosine_recall@10    | 0.76        | 0.72       |
-| **cosine_ndcg@10**  | **0.5174**  | **0.6172** |
-| cosine_mrr@10       | 0.4427      | 0.6031     |
-| cosine_map@100      | 0.4525      | 0.5845     |
+| cosine_accuracy@5   | 0.66        | 0.7        |
+| cosine_accuracy@10  | 0.78        | 0.78       |
+| cosine_precision@1  | 0.34        | 0.5        |
+| cosine_precision@3  | 0.1667      | 0.2267     |
+| cosine_precision@5  | 0.132       | 0.144      |
+| cosine_precision@10 | 0.078       | 0.082      |
+| cosine_recall@1     | 0.34        | 0.48       |
+| cosine_recall@3     | 0.5         | 0.64       |
+| cosine_recall@5     | 0.66        | 0.67       |
+| cosine_recall@10    | 0.78        | 0.74       |
+| **cosine_ndcg@10**  | **0.5447**  | **0.6136** |
+| cosine_mrr@10       | 0.4709      | 0.5822     |
+| cosine_map@100      | 0.4788      | 0.5769     |
 
 #### Nano BEIR
 
@@ -425,21 +425,21 @@ You can finetune this model on your own dataset.
 
 | Metric              | Value      |
 |:--------------------|:-----------|
-| cosine_accuracy@1   | 0.43       |
+| cosine_accuracy@1   | 0.42       |
 | cosine_accuracy@3   | 0.58       |
-| cosine_accuracy@5   | 0.64       |
-| cosine_accuracy@10  | 0.75       |
-| cosine_precision@1  | 0.43       |
-| cosine_precision@3  | 0.1933     |
-| cosine_precision@5  | 0.128      |
-| cosine_precision@10 | 0.078      |
-| cosine_recall@1     | 0.415      |
-| cosine_recall@3     | 0.56       |
-| cosine_recall@5     | 0.62       |
-| cosine_recall@10    | 0.74       |
-| **cosine_ndcg@10**  | **0.5673** |
-| cosine_mrr@10       | 0.5229     |
-| cosine_map@100      | 0.5185     |
+| cosine_accuracy@5   | 0.68       |
+| cosine_accuracy@10  | 0.78       |
+| cosine_precision@1  | 0.42       |
+| cosine_precision@3  | 0.1967     |
+| cosine_precision@5  | 0.138      |
+| cosine_precision@10 | 0.08       |
+| cosine_recall@1     | 0.41       |
+| cosine_recall@3     | 0.57       |
+| cosine_recall@5     | 0.665      |
+| cosine_recall@10    | 0.76       |
+| **cosine_ndcg@10**  | **0.5792** |
+| cosine_mrr@10       | 0.5265     |
+| cosine_map@100      | 0.5278     |
 
 <!--
 ## Bias, Risks and Limitations
@@ -513,9 +513,9 @@ You can finetune this model on your own dataset.
 - `eval_strategy`: steps
 - `per_device_train_batch_size`: 128
 - `per_device_eval_batch_size`: 128
-- `learning_rate`: 0.0001
-- `weight_decay`: 0.001
-- `max_steps`: 1687
+- `learning_rate`: 8e-05
+- `weight_decay`: 0.005
+- `max_steps`: 1125
 - `warmup_ratio`: 0.1
 - `fp16`: True
 - `dataloader_drop_last`: True
@@ -542,14 +542,14 @@ You can finetune this model on your own dataset.
 - `gradient_accumulation_steps`: 1
 - `eval_accumulation_steps`: None
 - `torch_empty_cache_steps`: None
-- `learning_rate`: 0.0001
-- `weight_decay`: 0.001
+- `learning_rate`: 8e-05
+- `weight_decay`: 0.005
 - `adam_beta1`: 0.9
 - `adam_beta2`: 0.999
 - `adam_epsilon`: 1e-08
 - `max_grad_norm`: 1.0
 - `num_train_epochs`: 3.0
-- `max_steps`: 1687
+- `max_steps`: 1125
 - `lr_scheduler_type`: linear
 - `lr_scheduler_kwargs`: {}
 - `warmup_ratio`: 0.1
@@ -656,13 +656,11 @@ 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    | -             | 0.1310          | 0.5540                     | 0.5931                | 0.5735                       |
-| 0.2874 | 250  | 0.1078        | 0.0793          | 0.5375                     | 0.5386                | 0.5380                       |
-| 0.5747 | 500  | 0.0893        | 0.0673          | 0.5031                     | 0.6009                | 0.5520                       |
-| 0.8621 | 750  | 0.081         | 0.0605          | 0.5414                     | 0.5786                | 0.5600                       |
-| 1.1494 | 1000 | 0.0593        | 0.0565          | 0.5158                     | 0.5786                | 0.5472                       |
-| 1.4368 | 1250 | 0.0422        | 0.0537          | 0.5300                     | 0.6107                | 0.5704                       |
-| 1.7241 | 1500 | 0.0402        | 0.0514          | 0.5174                     | 0.6172                | 0.5673                       |
+| 0      | 0    | -             | 0.1203          | 0.5887                     | 0.5786                | 0.5836                       |
+| 0.2874 | 250  | 0.094         | 0.0631          | 0.5536                     | 0.5611                | 0.5574                       |
+| 0.5747 | 500  | 0.0766        | 0.0586          | 0.5317                     | 0.5724                | 0.5521                       |
+| 0.8621 | 750  | 0.0674        | 0.0494          | 0.5357                     | 0.5675                | 0.5516                       |
+| 1.1494 | 1000 | 0.0491        | 0.0468          | 0.5447                     | 0.6136                | 0.5792                       |
 
 
 ### Framework Versions