Add SetFit model

README.md

@@ -11,15 +11,21 @@ datasets:
 metrics:
 - accuracy
 widget:
-- text: Are you mentally handicapped or this is your normal form? Both ways kill yourself
-    please.
+- text: WOW THIS GUY TALKS LIKE A RETARD HAHAHAHAHAH
 - text: American people are dumbass. They don't even know geography
-- text: Put a woman to abuse that fat woman and don’t put a kind guy
-- text: That's what we should do here in a Us survey take them to FEMA camps and put
-    the gas on these people are not coming from starvation destitute bad areas look
-    how they're dressed their well-dressed well fed it's like a trojan horse and it's
-    time we put them in the penal camps and get rid of them make use of the FEMA camp
-- text: LMAO while watching this church burn
+- text: Oh GOD YES!!!! Feel the shrapnel pierce through your skin, let the fire consume
+    your flesh and lay lifeless on the unforgiving sand that seeps into your wounds.
+    Your screams is what gives Murcia the fuck yeah in the most patriotic songs on
+    earth. BURN IN HELL YOU GOATS !!!
+- text: One needs to do a Pareto Analysis...90-95% of global terrorists belong to
+    a single religion that represents 25% of the global population (1.25 billion in
+    a 8 billion global population). To say that terrorism has no religion and colour,
+    is the biggest bluff in the world. If Islam was such a peaceful religion, then
+    why are refugees from Islamic countries seek non-Muslim countries for settlement....it
+    is because they want to Islamise the world.
+- text: It's so sick and sad that young people will think of the concept of clout
+    like rap niggers invented it because they aren't even properly taught about the
+    english language anymore.
 pipeline_tag: text-classification
 inference: false
 model-index:
@@ -34,7 +40,7 @@ model-index:
       split: test
     metrics:
     - type: accuracy
-      value: 0.44680851063829785
+      value: 0.39466666666666667
       name: Accuracy
 ---
 
@@ -70,7 +76,7 @@ The model has been trained using an efficient few-shot learning technique that i
 ### Metrics
 | Label   | Accuracy |
 |:--------|:---------|
-| **all** | 0.4468   |
+| **all** | 0.3947   |
 
 ## Uses
 
@@ -90,7 +96,7 @@ from setfit import SetFitModel
 # Download from the 🤗 Hub
 model = SetFitModel.from_pretrained("amitprgx/setfit-ethos-multilabel-example")
 # Run inference
-preds = model("LMAO while watching this church burn")
+preds = model("WOW THIS GUY TALKS LIKE A RETARD HAHAHAHAHAH")
 ```
 
 <!--
@@ -122,7 +128,7 @@ preds = model("LMAO while watching this church burn")
 ### Training Set Metrics
 | Training set | Min | Median  | Max |
 |:-------------|:----|:--------|:----|
-| Word count   | 3   | 18.1719 | 87  |
+| Word count   | 2   | 16.3594 | 72  |
 
 ### Training Hyperparameters
 - batch_size: (16, 16)
@@ -145,10 +151,10 @@ preds = model("LMAO while watching this church burn")
 ### Training Results
 | Epoch  | Step | Training Loss | Validation Loss |
 |:------:|:----:|:-------------:|:---------------:|
-| 0.0063 | 1    | 0.3377        | -               |
-| 0.3125 | 50   | 0.1288        | -               |
-| 0.625  | 100  | 0.0798        | -               |
-| 0.9375 | 150  | 0.1517        | -               |
+| 0.0063 | 1    | 0.2956        | -               |
+| 0.3125 | 50   | 0.1327        | -               |
+| 0.625  | 100  | 0.0963        | -               |
+| 0.9375 | 150  | 0.0525        | -               |
 
 ### Framework Versions
 - Python: 3.10.12

config_setfit.json

@@ -1,4 +1,4 @@
 {
-  "normalize_embeddings": false,
-  "labels": null
+  "labels": null,
+  "normalize_embeddings": false
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6ac0778714040e50528101b03ba629b3352c7c2f2d89b2b9dc4462572524606
+oid sha256:facd7f682af62404bfa4fadaf798e3517d9060603c9b66a38fb2fa43bbef7dd1
 size 437967672

model_head.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d2ddaa4e093425e1b6e6789ac250dffe6908931065ff61a6ec7d02efcbf6a1c5
+oid sha256:fcd73ab8a67bad84bcb9915a7f3bfd0fd3c3d51465c18a9fcdc7d0b4e8911b0f
 size 52836