Add SetFit model

README.md

@@ -1,37 +1,41 @@
 ---
-library_name: setfit
-tags:
-- setfit
-- sentence-transformers
-- text-classification
-- generated_from_setfit_trainer
 base_model: sentence-transformers/paraphrase-mpnet-base-v2
+library_name: setfit
 metrics:
 - accuracy
 - precision
 - recall
 - f1
-widget:
-- text: table drinking water threatsresource5c5007c2 56ef 4fb8 a37b 4f6dc9e37295
-- text: physical profile data collected oregon coast provide observation seasoar tow
-    support globec northeast pacific mesoscale survey august 2000 ncei accession 0001050
-- text: wood production intensification area aipl area intensification wood production
-    aipl territory mainly intended wood production silvicultural work aim increase
-    value per unit area volume per stem quality stem production desired specie combination
-    various production objective official register aipl applicable according article
-    69 sustainable forest management act ladtfthis third party metadata element translated
-    using automated translation tool amazon translate formdescriptors natureandenvironment
-    scienceandtechnology aipl area high forest potential sustainable forest management
-    forest development public forest silvicultural investment forest territorial limit
-    wood production stf forest territorial subdivision government information
-- text: shelter profile information
-- text: incident cerregulated pipeline facility canada energy regulator provides data
-    incident cerregulated pipeline facility defined onshore pipeline regulation processing
-    plant regulation data updated quarterlythe data provided 2008 current used visualization
-    tool canada energy regulator website source code visualization tool also available
-    economicsandindustry healthandsafety natureandenvironment cer pipeline safety
-    incident incident pipeline incident pipeline spill
 pipeline_tag: text-classification
+tags:
+- setfit
+- sentence-transformers
+- text-classification
+- generated_from_setfit_trainer
+widget:
+- text: topographic mapping physical location tree
+- text: forest tenure harvesting authority polygon spatial layer reflects operational
+    activity harvesting authority harvesting authority legal area cleared ministry
+    forest range claim land exist harvesting purpose corresponds outlined area exhibit
+    map forest tenure section ft responsible creation maintenance digital forest atlas
+    file province british columbia encompassing forest range act tenure also support
+    forest resource program delivered mofr feature contains ministry forest range
+    mofr featureclassskey number column defines type feature layer contains harvesting
+    authority boundary following feature class christmas tree permit 489 forest licence
+    cutting permit 556 licence cut 615 timber licence cutting permit 811 timber sale
+    licence major cutting permit 818 timber sale licence minor 819 tree farm licence
+    cutting permit 834 woodlot licence cutting permit 864 timber sale licence minor
+    cutting permit 917 community forest cutting permit 2401 harvesting authority life
+    cycle status code either pendingthe harvesting authority submitted new harvesting
+    authority amendment yet approved rejected activethe harvesting authority approved
+    activity may taking place harvesting authority retiredall activity completed harvesting
+    authority formdescriptors natureandenvironment scienceandtechnology canada fta
+    cut permit forest forest tenure harvest government information
+- text: concentration ultrapure water soluble aerosol trace element collected bulk
+    aerosol sample 2015 u geotraces western arctic transect uscgc healy hly1502 august
+    october 2015 ncei accession 0277333
+- text: death registration ontario location
+- text: aquatic resource area survey point
 inference: false
 model-index:
 - name: SetFit with sentence-transformers/paraphrase-mpnet-base-v2
@@ -45,16 +49,16 @@ model-index:
       split: test
     metrics:
     - type: accuracy
-      value: 0.6
+      value: 0.125
       name: Accuracy
     - type: precision
-      value: 1.0
+      value: 0.6666666666666666
       name: Precision
     - type: recall
-      value: 0.6666666666666666
+      value: 0.2222222222222222
       name: Recall
     - type: f1
-      value: 0.8
+      value: 0.3333333333333333
       name: F1
 ---
 
@@ -88,9 +92,9 @@ The model has been trained using an efficient few-shot learning technique that i
 ## Evaluation
 
 ### Metrics
-| Label   | Accuracy | Precision | Recall | F1  |
-|:--------|:---------|:----------|:-------|:----|
-| **all** | 0.6      | 1.0       | 0.6667 | 0.8 |
+| Label   | Accuracy | Precision | Recall | F1     |
+|:--------|:---------|:----------|:-------|:-------|
+| **all** | 0.125    | 0.6667    | 0.2222 | 0.3333 |
 
 ## Uses
 
@@ -110,7 +114,7 @@ from setfit import SetFitModel
 # Download from the 🤗 Hub
 model = SetFitModel.from_pretrained("lgd/setfit-multilabel")
 # Run inference
-preds = model("shelter profile information")
+preds = model("aquatic resource area survey point")
 ```
 
 <!--
@@ -142,13 +146,14 @@ preds = model("shelter profile information")
 ### Training Set Metrics
 | Training set | Min | Median | Max |
 |:-------------|:----|:-------|:----|
-| Word count   | 2   | 5.3    | 7   |
+| Word count   | 3   | 5.75   | 7   |
 
 ### Training Hyperparameters
-- batch_size: (8, 8)
-- num_epochs: (12, 12)
+- batch_size: (16, 16)
+- num_epochs: (1, 1)
 - max_steps: -1
 - sampling_strategy: oversampling
+- num_iterations: 20
 - body_learning_rate: (2e-05, 1e-05)
 - head_learning_rate: 0.01
 - loss: CosineSimilarityLoss
@@ -164,12 +169,7 @@ preds = model("shelter profile information")
 ### Training Results
 | Epoch | Step | Training Loss | Validation Loss |
 |:-----:|:----:|:-------------:|:---------------:|
-| 0.125 | 1    | 0.3106        | -               |
-| 6.25  | 50   | 0.0408        | -               |
-| 0.125 | 1    | 0.0929        | -               |
-| 6.25  | 50   | 0.0393        | -               |
-| 0.125 | 1    | 0.0134        | -               |
-| 6.25  | 50   | 0.0165        | -               |
+| 0.05  | 1    | 0.0058        | -               |
 
 ### Framework Versions
 - Python: 3.10.12

config_setfit.json

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

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1a4d4d0460bb696aa6144b45d6cfc56e3f10ed447d2b3e43fb620cba475d2ef9
+oid sha256:5f76e6d6812f4692b2394f819983598b48718cd9713ce9bc4bc704ddc0e65c9a
 size 437967672

model_head.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ed201ba19a800fa208d63db7b3c9bb6b3160e4069303e2bd4b9115823dfac0fc
+oid sha256:092c0421b2b598ae0eae80b8867445970f46b4bc4448e929b8aba5d135959d78
 size 26916