Add new SentenceTransformer model

README.md

@@ -286,9 +286,9 @@ print(embeddings.shape)
 # Get the similarity scores for the embeddings
 similarities = model.similarity(embeddings, embeddings)
 print(similarities)
-# tensor([[1.0000, 0.9020, 0.5829],
-#         [0.9020, 1.0000, 0.6243],
-#         [0.5829, 0.6243, 1.0000]])
+# tensor([[1.0000, 0.9031, 0.5886],
+#         [0.9031, 1.0000, 0.6318],
+#         [0.5886, 0.6318, 1.0000]])
 ```
 
 <!--
@@ -393,7 +393,7 @@ You can finetune this model on your own dataset.
 - `learning_rate`: 2e-05
 - `num_train_epochs`: 5
 - `lr_scheduler_type`: cosine
-- `warmup_ratio`: 0.1
+- `warmup_ratio`: 0.2
 - `bf16`: True
 - `tf32`: True
 - `load_best_model_at_end`: True
@@ -423,7 +423,7 @@ You can finetune this model on your own dataset.
 - `max_steps`: -1
 - `lr_scheduler_type`: cosine
 - `lr_scheduler_kwargs`: {}
-- `warmup_ratio`: 0.1
+- `warmup_ratio`: 0.2
 - `warmup_steps`: 0
 - `log_level`: passive
 - `log_level_replica`: warning
@@ -525,11 +525,11 @@ You can finetune this model on your own dataset.
 ### Training Logs
 | Epoch   | Step   | Training Loss | Validation Loss |
 |:-------:|:------:|:-------------:|:---------------:|
-| 1.0     | 6      | 0.1654        | 0.0145          |
-| 2.0     | 12     | 0.0099        | 0.0039          |
-| 3.0     | 18     | 0.0055        | 0.0026          |
-| 4.0     | 24     | 0.0046        | 0.0022          |
-| **5.0** | **30** | **0.0042**    | **0.0022**      |
+| 1.0     | 6      | 0.1973        | 0.0280          |
+| 2.0     | 12     | 0.0151        | 0.0046          |
+| 3.0     | 18     | 0.0061        | 0.0030          |
+| 4.0     | 24     | 0.0049        | 0.0025          |
+| **5.0** | **30** | **0.0045**    | **0.0024**      |
 
 * The bold row denotes the saved checkpoint.
 

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:220dd60bcc9208c6d7e5f0f926c27bef36e6ff2bd8f0cf26a1414ce3ded20381
+oid sha256:441c9ad57a4f568fc917d596c416bf4832dd261cbee90b1d02507ba544c66991
 size 596070136