Add SetFit model

README.md

@@ -1,5 +1,5 @@
 ---
-base_model: akhooli/sbert_ar_nli_500k
+base_model: akhooli/Arabic-SBERT-100K
 library_name: setfit
 metrics:
 - accuracy
@@ -35,7 +35,7 @@ widget:
     فالفندق
 inference: true
 model-index:
-- name: SetFit with akhooli/sbert_ar_nli_500k
+- name: SetFit with akhooli/Arabic-SBERT-100K
   results:
   - task:
       type: text-classification
@@ -46,13 +46,13 @@ model-index:
       split: test
     metrics:
     - type: accuracy
-      value: 0.46636363636363637
+      value: 0.5005050505050505
       name: Accuracy
 ---
 
-# SetFit with akhooli/sbert_ar_nli_500k
+# SetFit with akhooli/Arabic-SBERT-100K
 
-This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [akhooli/sbert_ar_nli_500k](https://huggingface.co/akhooli/sbert_ar_nli_500k) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
+This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [akhooli/Arabic-SBERT-100K](https://huggingface.co/akhooli/Arabic-SBERT-100K) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
 
 The model has been trained using an efficient few-shot learning technique that involves:
 
@@ -63,7 +63,7 @@ The model has been trained using an efficient few-shot learning technique that i
 
 ### Model Description
 - **Model Type:** SetFit
-- **Sentence Transformer body:** [akhooli/sbert_ar_nli_500k](https://huggingface.co/akhooli/sbert_ar_nli_500k)
+- **Sentence Transformer body:** [akhooli/Arabic-SBERT-100K](https://huggingface.co/akhooli/Arabic-SBERT-100K)
 - **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
 - **Maximum Sequence Length:** 512 tokens
 - **Number of Classes:** 3 classes
@@ -89,7 +89,7 @@ The model has been trained using an efficient few-shot learning technique that i
 ### Metrics
 | Label   | Accuracy |
 |:--------|:---------|
-| **all** | 0.4664   |
+| **all** | 0.5005   |
 
 ## Uses
 
@@ -145,12 +145,12 @@ preds = model("مكان راحه البال . المكان نظيف جدا وم
 
 | Label    | Training Sample Count |
 |:---------|:----------------------|
+| Mixed    | 8                     |
 | Negative | 8                     |
 | Positive | 8                     |
-| Mixed    | 8                     |
 
 ### Training Hyperparameters
-- batch_size: (16, 16)
+- batch_size: (32, 32)
 - num_epochs: (4, 4)
 - max_steps: -1
 - sampling_strategy: oversampling
@@ -170,11 +170,11 @@ preds = model("مكان راحه البال . المكان نظيف جدا وم
 ### Training Results
 | Epoch  | Step | Training Loss | Validation Loss |
 |:------:|:----:|:-------------:|:---------------:|
-| 0.0833 | 1    | 0.2047        | -               |
-| 1.0    | 12   | -             | 0.2570          |
-| 2.0    | 24   | -             | 0.2721          |
-| 3.0    | 36   | -             | 0.2744          |
-| 4.0    | 48   | -             | 0.2741          |
+| 0.1667 | 1    | 0.2506        | -               |
+| 1.0    | 6    | -             | 0.2707          |
+| 2.0    | 12   | -             | 0.2555          |
+| 3.0    | 18   | -             | 0.2677          |
+| 4.0    | 24   | -             | 0.2754          |
 
 ### Framework Versions
 - Python: 3.10.14

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "akhooli/sbert_ar_nli_500k",
+  "_name_or_path": "akhooli/Arabic-SBERT-100K",
   "architectures": [
     "BertModel"
   ],

config_setfit.json

@@ -1,8 +1,8 @@
 {
   "normalize_embeddings": false,
   "labels": [
+    "Mixed",
     "Negative",
-    "Positive",
-    "Mixed"
+    "Positive"
   ]
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:371918891d5e1705e6e5bdf494f8d81a1f0cc8c8ff58cda6a03b30fd6b701668
+oid sha256:1a56c33d50e4ea6defd40d5c06e9feb367b7d6e473eb03e975cf65f772deab63
 size 540795752

model_head.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1419f33885c2a0aca451451040cd7f8c3fb8525745151bf49e5db609c8871107
+oid sha256:5f4e48c683cb7411d3312fb65609c7d3b5ee2bfba3b96357f2e84b85a38abad1
 size 19375