Update README.md

README.md

@@ -3,7 +3,8 @@ tags:
 - setfit
 - sentence-transformers
 - text-classification
-- generated_from_setfit_trainer
+- geospatial
+- spatial-queries
 widget:
 - text: hotel in geneva airport
 - text: what payroll deduction is mpp
@@ -12,197 +13,41 @@ widget:
 - text: what's the weather in roseburg
 metrics:
 - accuracy
+- f1
 pipeline_tag: text-classification
 library_name: setfit
 inference: true
 base_model: BAAI/bge-small-en-v1.5
 ---
 
-# SetFit with BAAI/bge-small-en-v1.5
+# Spatial Web Search Query Classifier
 
-This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [BAAI/bge-small-en-v1.5](https://huggingface.co/BAAI/bge-small-en-v1.5) 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.
+A binary [SetFit](https://github.com/huggingface/setfit) classifier that distinguishes spatial from non-spatial web search queries. Trained on a gold-annotated sample of [MS MARCO](https://microsoft.github.io/msmarco/) and used to identify 104,288 spatial queries (10.3%) across the full 1.01M-query corpus.
 
-The model has been trained using an efficient few-shot learning technique that involves:
+**Accuracy / F1: 0.986** on a held-out balanced test set (76 negative, 72 positive).
 
-1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
-2. Training a classification head with features from the fine-tuned Sentence Transformer.
 
-## Model Details
+## What counts as spatial?
 
-### Model Description
-- **Model Type:** SetFit
-- **Sentence Transformer body:** [BAAI/bge-small-en-v1.5](https://huggingface.co/BAAI/bge-small-en-v1.5)
-- **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:** 2 classes
-<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
-<!-- - **Language:** Unknown -->
-<!-- - **License:** Unknown -->
-
-### Model Sources
-
-- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
-- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
-- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
-
-### Model Labels
-| Label | Examples                                                                                                                                             |
-|:------|:-----------------------------------------------------------------------------------------------------------------------------------------------------|
-| 1     | <ul><li>'how far is palms casino from the airport in las vegas'</li><li>'anarkali bazar lahore'</li><li>'what county is alma nebraska in?'</li></ul> |
-| 0     | <ul><li>'what is symptom of bipolar disorder'</li><li>'early symptoms of shingles outbreak'</li><li>'bnsf total employees'</li></ul>                 |
+A query is spatial if its answer is geographically variant and requires reasoning about geographic primitives (location, distance, or direction) or topological relationships (adjacency, containment, or connectivity). This includes implicitly spatial queries such as costs and prices in a specific area — not just those containing a toponym.
 
-## Uses
+## Model details
 
-### Direct Use for Inference
-
-First install the SetFit library:
-
-```bash
-pip install setfit
-```
+- **Sentence Transformer body:** [BAAI/bge-small-en-v1.5](https://huggingface.co/BAAI/bge-small-en-v1.5)
+- **Classification head:** LogisticRegression
+- **Training data:** 1,473 gold-labelled MS MARCO queries (755 non-spatial, 718 spatial), sampled via K-means centroids across the full embedding space for representativeness
+- **Labels:** `1` = spatial, `0` = non-spatial
 
-Then you can load this model and run inference.
+## Usage
 
 ```python
 from setfit import SetFitModel
 
-# Download from the 🤗 Hub
-model = SetFitModel.from_pretrained("setfit_model_id")
-# Run inference
-preds = model("weather in erlanger ky")
+model = SetFitModel.from_pretrained("TODO")
+preds = model(["weather in erlanger ky", "what is symptom of bipolar disorder"])
+# => [1, 0]
 ```
 
-<!--
-### Downstream Use
-
-*List how someone could finetune this model on their own dataset.*
--->
-
-<!--
-### Out-of-Scope Use
-
-*List how the model may foreseeably be misused and address what users ought not to do with the model.*
--->
-
-<!--
-## Bias, Risks and Limitations
-
-*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
--->
-
-<!--
-### Recommendations
-
-*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
--->
-
-## Training Details
-
-### Training Set Metrics
-| Training set | Min | Median | Max |
-|:-------------|:----|:-------|:----|
-| Word count   | 2   | 6.3028 | 21  |
-
-| Label | Training Sample Count |
-|:------|:----------------------|
-| 0     | 755                   |
-| 1     | 718                   |
-
-### Training Hyperparameters
-- batch_size: (64, 64)
-- num_epochs: (1, 1)
-- max_steps: -1
-- sampling_strategy: oversampling
-- body_learning_rate: (1e-05, 1e-05)
-- head_learning_rate: 0.01
-- loss: CosineSimilarityLoss
-- distance_metric: cosine_distance
-- margin: 0.25
-- end_to_end: False
-- use_amp: False
-- warmup_proportion: 0.1
-- l2_weight: 0.01
-- seed: 42
-- eval_max_steps: -1
-- load_best_model_at_end: False
-
-### Training Results
-| Epoch  | Step  | Training Loss | Validation Loss |
-|:------:|:-----:|:-------------:|:---------------:|
-| 0.0001 | 1     | 0.2507        | -               |
-| 0.0294 | 500   | 0.1803        | -               |
-| 0.0589 | 1000  | 0.0135        | -               |
-| 0.0883 | 1500  | 0.0021        | -               |
-| 0.1178 | 2000  | 0.001         | -               |
-| 0.1472 | 2500  | 0.0007        | -               |
-| 0.1766 | 3000  | 0.0005        | -               |
-| 0.2061 | 3500  | 0.0004        | -               |
-| 0.2355 | 4000  | 0.0004        | -               |
-| 0.2649 | 4500  | 0.0003        | -               |
-| 0.2944 | 5000  | 0.0003        | -               |
-| 0.3238 | 5500  | 0.0003        | -               |
-| 0.3533 | 6000  | 0.0003        | -               |
-| 0.3827 | 6500  | 0.0002        | -               |
-| 0.4121 | 7000  | 0.0003        | -               |
-| 0.4416 | 7500  | 0.0002        | -               |
-| 0.4710 | 8000  | 0.0002        | -               |
-| 0.5004 | 8500  | 0.0002        | -               |
-| 0.5299 | 9000  | 0.0002        | -               |
-| 0.5593 | 9500  | 0.0002        | -               |
-| 0.5888 | 10000 | 0.0002        | -               |
-| 0.6182 | 10500 | 0.0002        | -               |
-| 0.6476 | 11000 | 0.0001        | -               |
-| 0.6771 | 11500 | 0.0001        | -               |
-| 0.7065 | 12000 | 0.0001        | -               |
-| 0.7359 | 12500 | 0.0001        | -               |
-| 0.7654 | 13000 | 0.0001        | -               |
-| 0.7948 | 13500 | 0.0001        | -               |
-| 0.8243 | 14000 | 0.0001        | -               |
-| 0.8537 | 14500 | 0.0001        | -               |
-| 0.8831 | 15000 | 0.0001        | -               |
-| 0.9126 | 15500 | 0.0001        | -               |
-| 0.9420 | 16000 | 0.0001        | -               |
-| 0.9714 | 16500 | 0.0001        | -               |
-
-### Framework Versions
-- Python: 3.11.5
-- SetFit: 1.1.2
-- Sentence Transformers: 4.0.2
-- Transformers: 4.55.2
-- PyTorch: 2.8.0
-- Datasets: 2.15.0
-- Tokenizers: 0.21.1
-
-## Citation
-
-### BibTeX
-```bibtex
-@article{https://doi.org/10.48550/arxiv.2209.11055,
-    doi = {10.48550/ARXIV.2209.11055},
-    url = {https://arxiv.org/abs/2209.11055},
-    author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
-    keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
-    title = {Efficient Few-Shot Learning Without Prompts},
-    publisher = {arXiv},
-    year = {2022},
-    copyright = {Creative Commons Attribution 4.0 International}
-}
-```
-
-<!--
-## Glossary
-
-*Clearly define terms in order to be accessible across audiences.*
--->
-
-<!--
-## Model Card Authors
-
-*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
--->
-
-<!--
-## Model Card Contact
+## Training
 
-*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
--->
+Weak labels were generated by running Llama 3.1 five times per query at temperature 0.2, then manually verified. The SetFit model was trained for one epoch with batch size 64 and learning rate 1e-5, then retrained on the full gold dataset for production inference.