Add SetFit ABSA model

README.md

@@ -6,14 +6,14 @@ tags:
 - text-classification
 - generated_from_setfit_trainer
 widget:
-- text: battery life is fantastic,:battery life is fantastic, screen is nice and large,
-    mouse pad is perfect and the unit is very light.
-- text: no smell, clear photos and clear:good texture, no smell, clear photos and
-    clear picture quality.
-- text: quality on the lens are pretty great:the pictures quality on the lens are
-    pretty great
-- text: it is not worth your money.:it is not worth your money.
-- text: have all the settings it should.:it doesnt have all the settings it should.
+- text: picture quality, easy to use.:loads of features, great picture quality, easy
+    to use.
+- text: very good auto focus, and a:has a timer that keeps its setting for multiple
+    pictures, has very good auto focus, and a large view screen.
+- text: takes horrible photos and not easy:takes horrible photos and not easy to use.
+- text: pictures are very clear and precise.:the pictures are very clear and precise.
+- text: compact and the images are sharp and:it is compact and the images are sharp
+    and clear.
 metrics:
 - accuracy
 pipeline_tag: text-classification
@@ -59,10 +59,10 @@ This model was trained within the context of a larger system for ABSA, which loo
 - **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
 
 ### Model Labels
-| Label    | Examples                                                                                                                                                                                                                                                     |
-|:---------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
-| negative | <ul><li>'like out of focus.:the color was not clear almost like out of focus.'</li><li>'the color was not clear:the color was not clear almost like out of focus.'</li><li>"the zoom was horrible and:the zoom was horrible and didn't focus."</li></ul>     |
-| positive | <ul><li>'the setup was easy and she was:the setup was easy and she was taking pictures almost right away.'</li><li>'easy to operate even:easy to operate even for a novice!'</li><li>'it is easy to use and:it is easy to use and super reliable.'</li></ul> |
+| Label    | Examples                                                                                                                                                                                                                                                                                                                                |
+|:---------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| positive | <ul><li>',  auto focus excellent:good starter,  easy use,  auto focus excellent'</li><li>'zoom works very well:zoom works very well.'</li><li>'auto focus is great and:auto focus is great and take superb close ups.'</li></ul>                                                                                                        |
+| negative | <ul><li>"lens that it comes:lens that it comes with definitely ain't for long distances, and the picture quality isn't the best."</li><li>'the zoom on the camera:the zoom on the camera seemed to have a problem, too.'</li><li>'focus is slow,:focus is slow, no continuous options, no touch screen like my older camera.'</li></ul> |
 
 ## Uses
 
@@ -117,12 +117,12 @@ preds = model("The food was great, but the venue is just way too busy.")
 ### Training Set Metrics
 | Training set | Min | Median  | Max |
 |:-------------|:----|:--------|:----|
-| Word count   | 5   | 13.7379 | 31  |
+| Word count   | 5   | 13.8511 | 31  |
 
 | Label    | Training Sample Count |
 |:---------|:----------------------|
-| negative | 51                    |
-| positive | 52                    |
+| negative | 47                    |
+| positive | 47                    |
 
 ### Training Hyperparameters
 - batch_size: (128, 128)
@@ -145,11 +145,10 @@ preds = model("The food was great, but the venue is just way too busy.")
 ### Training Results
 | Epoch  | Step | Training Loss | Validation Loss |
 |:------:|:----:|:-------------:|:---------------:|
-| 0.0233 | 1    | 0.2709        | -               |
-| 1.1628 | 50   | 0.0951        | 0.0004          |
-| 2.3256 | 100  | 0.0004        | 0.0002          |
-| 3.4884 | 150  | 0.0002        | 0.0001          |
-| 4.6512 | 200  | 0.0002        | 0.0001          |
+| 0.0278 | 1    | 0.3158        | -               |
+| 1.3889 | 50   | 0.0839        | 0.0013          |
+| 2.7778 | 100  | 0.0004        | 0.0006          |
+| 4.1667 | 150  | 0.0002        | 0.0005          |
 
 ### Framework Versions
 - Python: 3.11.12

config_setfit.json

@@ -1,9 +1,9 @@
 {
+  "normalize_embeddings": false,
   "labels": [
     "negative",
     "positive"
   ],
   "span_context": 3,
-  "normalize_embeddings": false,
   "spacy_model": "en_core_web_sm"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fbfa426ed433738367083d74c906b34090e3dd2b2e4c47d938b54ccfef687dbb
+oid sha256:d4de874438f44f92a7e655ee31606c85290afa6e3da1a3a83b4f0f35cf5ea095
 size 437967672

model_head.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:50d3e9d719b99e389850490d66757e2c8ec0c1579dd7554ac843bf5754a6be04
+oid sha256:2d1bdc66d165b33f1b4e395e179d385a3d8c287de4518a00302072cdd93d6f7f
 size 7055