folder is added

calibration.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f217d6037eac4d7c33b1d72b09f52bfab416ea26a81d2a775d505d20a65ff512
+size 884

category_model/1_Pooling/config.json

@@ -0,0 +1,10 @@
+{
+  "word_embedding_dimension": 384,
+  "pooling_mode_cls_token": false,
+  "pooling_mode_mean_tokens": true,
+  "pooling_mode_max_tokens": false,
+  "pooling_mode_mean_sqrt_len_tokens": false,
+  "pooling_mode_weightedmean_tokens": false,
+  "pooling_mode_lasttoken": false,
+  "include_prompt": true
+}

category_model/README.md

@@ -0,0 +1,279 @@
+---
+base_model: sentence-transformers/paraphrase-MiniLM-L3-v2
+library_name: setfit
+metrics:
+- accuracy
+pipeline_tag: text-classification
+tags:
+- setfit
+- sentence-transformers
+- text-classification
+- generated_from_setfit_trainer
+widget:
+- text: need help seeing the fucking status of order
+- text: i want to see how long the shipping takes, how can i do it?
+- text: i dont know how i can find my damn invoices from
+- text: where to sign up to the company nmewsletter
+- text: i have to notify of a payment error
+inference: true
+---
+
+# SetFit with sentence-transformers/paraphrase-MiniLM-L3-v2
+
+This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-MiniLM-L3-v2](https://huggingface.co/sentence-transformers/paraphrase-MiniLM-L3-v2) 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:
+
+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
+
+### Model Description
+- **Model Type:** SetFit
+- **Sentence Transformer body:** [sentence-transformers/paraphrase-MiniLM-L3-v2](https://huggingface.co/sentence-transformers/paraphrase-MiniLM-L3-v2)
+- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
+- **Maximum Sequence Length:** 128 tokens
+- **Number of Classes:** 4 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                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    |
+|:------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| 2     | <ul><li>"i'm having an issue with the . please assist. i was looking for to use for testing purposes, but it's not working, my mouse is too far from my keyboard and there's no way to turn i've tried clearing the cache and data for the app, but the issue persists."</li><li>"i'm having an issue with the . please assist. the is from an ad agency. we're not responsible for the item you bought without your consent, and we'll this problem started occurring after the recent software update. i haven't made any other changes to the device."</li><li>"i'm having an issue with the . please assist. i've performed a factory reset on my , hoping it would resolve the problem, but it didn't help."</li></ul> |
+| 1     | <ul><li>'i have to check when my item is gonnaarrive help me'</li><li>'can iu talk to customer support'</li><li>'i want help to earn a produxt'</li></ul>                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   |
+| 0     | <ul><li>'how can i give a quick look at my bill 00108?'</li><li>'where can i give a quick look at invoice 37777'</li><li>'could you help me list the allowed payment modalities?'</li></ul>                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 |
+| 3     | <ul><li>'how could i inform of a sign-up issue?'</li><li>'i need information about a pwd recovery'</li><li>'need to change the info on my user account'</li></ul>                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           |
+
+## Uses
+
+### Direct Use for Inference
+
+First install the SetFit library:
+
+```bash
+pip install setfit
+```
+
+Then you can load this model and run inference.
+
+```python
+from setfit import SetFitModel
+
+# Download from the 🤗 Hub
+model = SetFitModel.from_pretrained("setfit_model_id")
+# Run inference
+preds = model("i have to notify of a payment error")
+```
+
+<!--
+### 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   | 19.2405 | 61  |
+
+| Label | Training Sample Count |
+|:------|:----------------------|
+| 0     | 500                   |
+| 1     | 500                   |
+| 2     | 500                   |
+| 3     | 500                   |
+
+### Training Hyperparameters
+- 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
+- distance_metric: cosine_distance
+- margin: 0.25
+- end_to_end: False
+- use_amp: False
+- warmup_proportion: 0.1
+- seed: 42
+- eval_max_steps: -1
+- load_best_model_at_end: False
+
+### Training Results
+| Epoch  | Step | Training Loss | Validation Loss |
+|:------:|:----:|:-------------:|:---------------:|
+| 0.0002 | 1    | 0.3088        | -               |
+| 0.01   | 50   | 0.3099        | -               |
+| 0.02   | 100  | 0.2896        | -               |
+| 0.03   | 150  | 0.2212        | -               |
+| 0.04   | 200  | 0.2497        | -               |
+| 0.05   | 250  | 0.2114        | -               |
+| 0.06   | 300  | 0.2305        | -               |
+| 0.07   | 350  | 0.261         | -               |
+| 0.08   | 400  | 0.207         | -               |
+| 0.09   | 450  | 0.1719        | -               |
+| 0.1    | 500  | 0.1956        | -               |
+| 0.11   | 550  | 0.0833        | -               |
+| 0.12   | 600  | 0.2184        | -               |
+| 0.13   | 650  | 0.1225        | -               |
+| 0.14   | 700  | 0.136         | -               |
+| 0.15   | 750  | 0.1928        | -               |
+| 0.16   | 800  | 0.1396        | -               |
+| 0.17   | 850  | 0.0646        | -               |
+| 0.18   | 900  | 0.1592        | -               |
+| 0.19   | 950  | 0.0691        | -               |
+| 0.2    | 1000 | 0.1672        | -               |
+| 0.21   | 1050 | 0.0948        | -               |
+| 0.22   | 1100 | 0.0664        | -               |
+| 0.23   | 1150 | 0.1591        | -               |
+| 0.24   | 1200 | 0.0275        | -               |
+| 0.25   | 1250 | 0.0686        | -               |
+| 0.26   | 1300 | 0.2408        | -               |
+| 0.27   | 1350 | 0.1807        | -               |
+| 0.28   | 1400 | 0.0895        | -               |
+| 0.29   | 1450 | 0.0291        | -               |
+| 0.3    | 1500 | 0.0628        | -               |
+| 0.31   | 1550 | 0.0712        | -               |
+| 0.32   | 1600 | 0.0978        | -               |
+| 0.33   | 1650 | 0.0579        | -               |
+| 0.34   | 1700 | 0.0306        | -               |
+| 0.35   | 1750 | 0.0752        | -               |
+| 0.36   | 1800 | 0.12          | -               |
+| 0.37   | 1850 | 0.0958        | -               |
+| 0.38   | 1900 | 0.0436        | -               |
+| 0.39   | 1950 | 0.0735        | -               |
+| 0.4    | 2000 | 0.1054        | -               |
+| 0.41   | 2050 | 0.0921        | -               |
+| 0.42   | 2100 | 0.1           | -               |
+| 0.43   | 2150 | 0.075         | -               |
+| 0.44   | 2200 | 0.0374        | -               |
+| 0.45   | 2250 | 0.0739        | -               |
+| 0.46   | 2300 | 0.1083        | -               |
+| 0.47   | 2350 | 0.0388        | -               |
+| 0.48   | 2400 | 0.0492        | -               |
+| 0.49   | 2450 | 0.0524        | -               |
+| 0.5    | 2500 | 0.0367        | -               |
+| 0.51   | 2550 | 0.0653        | -               |
+| 0.52   | 2600 | 0.0967        | -               |
+| 0.53   | 2650 | 0.0872        | -               |
+| 0.54   | 2700 | 0.0664        | -               |
+| 0.55   | 2750 | 0.0403        | -               |
+| 0.56   | 2800 | 0.068         | -               |
+| 0.57   | 2850 | 0.0984        | -               |
+| 0.58   | 2900 | 0.0909        | -               |
+| 0.59   | 2950 | 0.086         | -               |
+| 0.6    | 3000 | 0.0158        | -               |
+| 0.61   | 3050 | 0.076         | -               |
+| 0.62   | 3100 | 0.0177        | -               |
+| 0.63   | 3150 | 0.0338        | -               |
+| 0.64   | 3200 | 0.0861        | -               |
+| 0.65   | 3250 | 0.0311        | -               |
+| 0.66   | 3300 | 0.0544        | -               |
+| 0.67   | 3350 | 0.0774        | -               |
+| 0.68   | 3400 | 0.0227        | -               |
+| 0.69   | 3450 | 0.0507        | -               |
+| 0.7    | 3500 | 0.0279        | -               |
+| 0.71   | 3550 | 0.0037        | -               |
+| 0.72   | 3600 | 0.0498        | -               |
+| 0.73   | 3650 | 0.0256        | -               |
+| 0.74   | 3700 | 0.0503        | -               |
+| 0.75   | 3750 | 0.0297        | -               |
+| 0.76   | 3800 | 0.0523        | -               |
+| 0.77   | 3850 | 0.019         | -               |
+| 0.78   | 3900 | 0.0733        | -               |
+| 0.79   | 3950 | 0.036         | -               |
+| 0.8    | 4000 | 0.024         | -               |
+| 0.81   | 4050 | 0.0808        | -               |
+| 0.82   | 4100 | 0.0838        | -               |
+| 0.83   | 4150 | 0.0362        | -               |
+| 0.84   | 4200 | 0.0796        | -               |
+| 0.85   | 4250 | 0.0664        | -               |
+| 0.86   | 4300 | 0.0229        | -               |
+| 0.87   | 4350 | 0.0461        | -               |
+| 0.88   | 4400 | 0.0125        | -               |
+| 0.89   | 4450 | 0.0153        | -               |
+| 0.9    | 4500 | 0.0483        | -               |
+| 0.91   | 4550 | 0.0371        | -               |
+| 0.92   | 4600 | 0.0305        | -               |
+| 0.93   | 4650 | 0.0481        | -               |
+| 0.94   | 4700 | 0.092         | -               |
+| 0.95   | 4750 | 0.0483        | -               |
+| 0.96   | 4800 | 0.0198        | -               |
+| 0.97   | 4850 | 0.0191        | -               |
+| 0.98   | 4900 | 0.0482        | -               |
+| 0.99   | 4950 | 0.0189        | -               |
+| 1.0    | 5000 | 0.0193        | -               |
+
+### Framework Versions
+- Python: 3.12.13
+- SetFit: 1.0.3
+- Sentence Transformers: 2.7.0
+- Transformers: 4.40.2
+- PyTorch: 2.11.0+cu128
+- Datasets: 3.2.0
+- Tokenizers: 0.19.1
+
+## 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
+
+*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
+-->

category_model/config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "sentence-transformers/paraphrase-MiniLM-L3-v2",
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 384,
+  "initializer_range": 0.02,
+  "intermediate_size": 1536,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 3,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.40.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

category_model/config_sentence_transformers.json

@@ -0,0 +1,9 @@
+{
+  "__version__": {
+    "sentence_transformers": "2.0.0",
+    "transformers": "4.7.0",
+    "pytorch": "1.9.0+cu102"
+  },
+  "prompts": {},
+  "default_prompt_name": null
+}

category_model/config_setfit.json

@@ -0,0 +1,4 @@
+{
+  "normalize_embeddings": false,
+  "labels": null
+}

category_model/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d47cb714e248b6092182b58af0aa38943ec53ab6d40aa4551af74ff305ea6d01
+size 69565312

category_model/model_head.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e0d576e5792dcbbfaaf6011efb00f1d189590e39a03487b4d2c79693df4f571
+size 13191

category_model/modules.json

@@ -0,0 +1,14 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  }
+]

category_model/sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+  "max_seq_length": 128,
+  "do_lower_case": false
+}

category_model/special_tokens_map.json

@@ -0,0 +1,37 @@
+{
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "[MASK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

category_model/tokenizer_config.json

@@ -0,0 +1,64 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "max_length": 128,
+  "model_max_length": 512,
+  "never_split": null,
+  "pad_to_multiple_of": null,
+  "pad_token": "[PAD]",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "stride": 0,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "[UNK]"
+}

label_mappings.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:862012dd3c4ee3360ad9f54c3b3731c5ef1c1b2b4967e8bd1012985144dfb2ab
+size 380

predict.py

@@ -0,0 +1,119 @@
+"""
+Support Ticket Classifier — Prediction Module
+Loads trained SetFit category model + calibration + keyword urgency rules.
+"""
+import re
+import pickle
+import numpy as np
+from setfit import SetFitModel
+
+
+# ── Text cleaning (must match Phase 1 exactly) ────────────────────────
+def clean_text(text: str) -> str:
+    if not isinstance(text, str):
+        return ""
+    text = text.lower()
+    text = re.sub(r"\{[^}]+\}", "", text)
+    text = re.sub(r"http\S+|www\.\S+", "", text)
+    text = re.sub(r"\S+@\S+", "", text)
+    text = re.sub(r"[^a-z0-9\s.,!?'\-]", " ", text)
+    text = re.sub(r"\s+", " ", text)
+    return text.strip()
+
+
+# ── Keyword-based urgency rules ───────────────────────────────────────
+HIGH_KEYWORDS = [
+    "urgent", "asap", "immediately", "emergency", "critical",
+    "can't access", "cannot access", "locked out", "account hacked",
+    "charged twice", "double charged", "unauthorized charge",
+    "service down", "not working", "completely broken", "data loss",
+    "refund immediately", "cancel immediately", "fraud", "security breach",
+    "down all day"
+]
+LOW_KEYWORDS = [
+    "just wondering", "question about", "when will", "how do i",
+    "could you explain", "i would like to know", "curious about",
+    "general question", "update my", "change my"
+]
+
+
+def get_urgency(text: str) -> str:
+    t = text.lower()
+    if any(kw in t for kw in HIGH_KEYWORDS):
+        return "high"
+    if any(kw in t for kw in LOW_KEYWORDS):
+        return "low"
+    return "medium"
+
+
+# ── Classifier (lazy-loaded singleton) ────────────────────────────────
+class TicketClassifier:
+    _instance = None
+
+    def __new__(cls):
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+            cls._instance._loaded = False
+        return cls._instance
+
+    def load(self):
+        if self._loaded:
+            return
+        print("Loading category model...")
+        self.cat_model = SetFitModel.from_pretrained("category_model")
+
+        print("Loading calibration...")
+        with open("calibration.pkl", "rb") as f:
+            self.platt_scaler = pickle.load(f)
+
+        print("Loading label mappings...")
+        with open("label_mappings.pkl", "rb") as f:
+            mappings = pickle.load(f)
+        self.cat_encoder = mappings["category"]
+
+        self._loaded = True
+        print("✅ Model ready.")
+
+    def predict(self, ticket_text: str) -> dict:
+        if not self._loaded:
+            self.load()
+
+        cleaned = clean_text(ticket_text)
+
+        # Category + calibrated confidence
+        raw_probs  = np.array(self.cat_model.predict_proba([cleaned]))
+        cal_probs  = self.platt_scaler.predict_proba(raw_probs)
+        cat_idx    = int(np.argmax(cal_probs[0]))
+        category   = self.cat_encoder.inverse_transform([cat_idx])[0]
+        confidence = round(float(cal_probs[0][cat_idx]), 4)
+
+        # Urgency from keyword rules
+        urgency = get_urgency(ticket_text)
+
+        return {
+            "category":   category,
+            "confidence": confidence,
+            "urgency":    urgency
+        }
+
+
+# Convenience function
+_classifier = TicketClassifier()
+
+def predict_ticket(text: str) -> dict:
+    return _classifier.predict(text)
+
+
+# Quick local test
+if __name__ == "__main__":
+    samples = [
+        "I was charged twice and need a refund immediately",
+        "How do I update my billing address?",
+        "The app is completely broken, I cannot access my account",
+        "Just wondering when my subscription renews",
+    ]
+    for s in samples:
+        result = predict_ticket(s)
+        print(f"\nTicket    : {s}")
+        print(f"Category  : {result['category']} ({result['confidence']*100:.1f}%)")
+        print(f"Urgency   : {result['urgency']}")

requirements.txt

@@ -0,0 +1,9 @@
+setfit==1.0.3
+sentence-transformers==2.7.0
+transformers==4.40.2
+huggingface_hub==0.23.5
+scikit-learn>=1.3.0
+numpy>=1.24.0
+fastapi>=0.110.0
+uvicorn[standard]>=0.27.0
+pydantic>=2.0.0