Add new SentenceTransformer model.

README.md

@@ -21,6 +21,7 @@ tags:
 - loss:MSELoss
 - dataset_size:5000
 - dataset_size:8000
+- dataset_size:100000
 widget:
 - source_sentence: 'The aggressive semi-employed religion workshop of Razzak, (EFP).
 
@@ -112,7 +113,7 @@ model-index:
       type: unknown
     metrics:
     - type: negative_mse
-      value: -0.37439612206071615
+      value: -0.32407890539616346
       name: Negative Mse
   - task:
       type: translation
@@ -122,13 +123,13 @@ model-index:
       type: unknown
     metrics:
     - type: src2trg_accuracy
-      value: 0.0186
+      value: 0.05445
       name: Src2Trg Accuracy
     - type: trg2src_accuracy
-      value: 0.00835
+      value: 0.02105
       name: Trg2Src Accuracy
     - type: mean_accuracy
-      value: 0.013474999999999999
+      value: 0.03775
       name: Mean Accuracy
 ---
 
@@ -231,7 +232,7 @@ You can finetune this model on your own dataset.
 
 | Metric           | Value       |
 |:-----------------|:------------|
-| **negative_mse** | **-0.3744** |
+| **negative_mse** | **-0.3241** |
 
 #### Translation
 
@@ -239,9 +240,9 @@ You can finetune this model on your own dataset.
 
 | Metric            | Value      |
 |:------------------|:-----------|
-| src2trg_accuracy  | 0.0186     |
-| trg2src_accuracy  | 0.0083     |
-| **mean_accuracy** | **0.0135** |
+| src2trg_accuracy  | 0.0544     |
+| trg2src_accuracy  | 0.021      |
+| **mean_accuracy** | **0.0377** |
 
 <!--
 ## Bias, Risks and Limitations
@@ -262,7 +263,7 @@ You can finetune this model on your own dataset.
 #### momo22/eng2nep
 
 * Dataset: [momo22/eng2nep](https://huggingface.co/datasets/momo22/eng2nep) at [57da8d4](https://huggingface.co/datasets/momo22/eng2nep/tree/57da8d44266896e334c1d8f2528cbbf666fbd0ca)
-* Size: 8,000 training samples
+* Size: 100,000 training samples
 * Columns: <code>English</code>, <code>Nepali</code>, and <code>label</code>
 * Approximate statistics based on the first 1000 samples:
   |         | English                                                                            | Nepali                                                                             | label                                |
@@ -282,13 +283,13 @@ You can finetune this model on your own dataset.
 #### momo22/eng2nep
 
 * Dataset: [momo22/eng2nep](https://huggingface.co/datasets/momo22/eng2nep) at [57da8d4](https://huggingface.co/datasets/momo22/eng2nep/tree/57da8d44266896e334c1d8f2528cbbf666fbd0ca)
-* Size: 500 evaluation samples
+* Size: 8,000 evaluation samples
 * Columns: <code>English</code>, <code>Nepali</code>, and <code>label</code>
 * Approximate statistics based on the first 1000 samples:
-  |         | English                                                                            | Nepali                                                                            | label                                |
-  |:--------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:-------------------------------------|
-  | type    | string                                                                             | string                                                                            | list                                 |
-  | details | <ul><li>min: 4 tokens</li><li>mean: 26.71 tokens</li><li>max: 213 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 64.1 tokens</li><li>max: 256 tokens</li></ul> | <ul><li>size: 384 elements</li></ul> |
+  |         | English                                                                            | Nepali                                                                             | label                                |
+  |:--------|:-----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:-------------------------------------|
+  | type    | string                                                                             | string                                                                             | list                                 |
+  | details | <ul><li>min: 4 tokens</li><li>mean: 26.48 tokens</li><li>max: 213 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 63.73 tokens</li><li>max: 256 tokens</li></ul> | <ul><li>size: 384 elements</li></ul> |
 * Samples:
   | English                                                                                                                                                                                                        | Nepali                                                                                                                                                                          | label                                                                                                                              |
   |:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------|
@@ -304,7 +305,6 @@ You can finetune this model on your own dataset.
 - `per_device_train_batch_size`: 64
 - `per_device_eval_batch_size`: 64
 - `learning_rate`: 2e-05
-- `num_train_epochs`: 1
 - `warmup_ratio`: 0.1
 - `bf16`: True
 - `push_to_hub`: True
@@ -330,7 +330,7 @@ You can finetune this model on your own dataset.
 - `adam_beta2`: 0.999
 - `adam_epsilon`: 1e-08
 - `max_grad_norm`: 1.0
-- `num_train_epochs`: 1
+- `num_train_epochs`: 3
 - `max_steps`: -1
 - `lr_scheduler_type`: linear
 - `lr_scheduler_kwargs`: {}
@@ -427,12 +427,21 @@ You can finetune this model on your own dataset.
 </details>
 
 ### Training Logs
-| Epoch | Step | Training Loss | loss   | mean_accuracy | negative_mse |
-|:-----:|:----:|:-------------:|:------:|:-------------:|:------------:|
-| 0.4   | 50   | 0.0021        | 0.0019 | 0.0111        | -0.3837      |
-| 0.8   | 100  | 0.002         | 0.0019 | 0.0123        | -0.3794      |
-| 0.4   | 50   | 0.002         | 0.0019 | 0.0130        | -0.3773      |
-| 0.8   | 100  | 0.002         | 0.0019 | 0.0135        | -0.3744      |
+| Epoch  | Step | Training Loss | loss   | mean_accuracy | negative_mse |
+|:------:|:----:|:-------------:|:------:|:-------------:|:------------:|
+| 0.4    | 50   | 0.0021        | 0.0019 | 0.0111        | -0.3837      |
+| 0.8    | 100  | 0.002         | 0.0019 | 0.0123        | -0.3794      |
+| 0.4    | 50   | 0.002         | 0.0019 | 0.0130        | -0.3773      |
+| 0.8    | 100  | 0.002         | 0.0019 | 0.0135        | -0.3744      |
+| 0.3199 | 500  | 0.002         | 0.0018 | 0.0166        | -0.3597      |
+| 0.6398 | 1000 | 0.0019        | 0.0018 | 0.0204        | -0.3461      |
+| 0.9597 | 1500 | 0.0018        | 0.0017 | 0.0241        | -0.3389      |
+| 1.2796 | 2000 | 0.0018        | 0.0017 | 0.0273        | -0.3351      |
+| 1.5995 | 2500 | 0.0018        | 0.0017 | 0.0312        | -0.3302      |
+| 1.9194 | 3000 | 0.0018        | 0.0017 | 0.0328        | -0.3284      |
+| 2.2393 | 3500 | 0.0018        | 0.0017 | 0.0353        | -0.3264      |
+| 2.5592 | 4000 | 0.0018        | 0.0016 | 0.0374        | -0.3246      |
+| 2.8791 | 4500 | 0.0018        | 0.0016 | 0.0377        | -0.3241      |
 
 
 ### Framework Versions