Upload dev+data-analyst r95 tiny BERT model

README.md

@@ -0,0 +1,145 @@
+---
+language:
+- ru
+- en
+license: mit
+library_name: transformers
+pipeline_tag: text-classification
+tags:
+- text-classification
+- bert
+- tiny-bert
+- rubert-tiny2
+- binary-classification
+- jobs
+- developer-classification
+- data-analyst-classification
+- dev-plus-da
+- r95
+- v1
+base_model: cointegrated/rubert-tiny2
+metrics:
+- precision
+- recall
+- roc_auc
+model-index:
+- name: dev_da_roles_1
+  results:
+  - task:
+      type: text-classification
+      name: Developer/Data Analyst vs Other Binary Classification
+    metrics:
+    - type: roc_auc
+      value: 0.9790
+    - type: precision
+      value: 0.8797
+    - type: recall
+      value: 0.9509
+---
+
+# dev_da_roles_1 - Developer + Data Analyst Classifier
+
+Binary job-vacancy classifier: detects **developer or Data Analyst** roles (`tech`) versus **other** roles (`other`).
+
+Built on top of [`cointegrated/rubert-tiny2`](https://huggingface.co/cointegrated/rubert-tiny2), a compact BERT model for Russian and English text.
+
+## Task Definition
+
+The positive class (`tech`) is defined as:
+
+> `role_category in TECH_CLASSES AND team_lead == 0`
+
+`TECH_CLASSES`:
+
+- Backend
+- Desktop / Systems
+- Embedded
+- Frontend
+- Fullstack
+- ML / AI / Data Scientist
+- Mobile
+- Data Analyst
+
+Team leads and management roles are intentionally excluded from the positive class.
+
+## Labels
+
+| id | label |
+|----|-------|
+| 0  | other |
+| 1  | tech  |
+
+## Validation Metrics
+
+| Metric | Value |
+|---|---:|
+| ROC AUC | 0.9790 |
+| Precision @ threshold | 0.8797 |
+| Recall @ threshold | 0.9509 |
+| Best threshold | 0.3765 |
+| Target recall | 0.95 |
+
+Best epoch: **8**. Training `pos_weight`: **2.4813**.
+
+## Inference Parameters
+
+- `max_length`: **256** tokens
+- Vacancy text is formed as `title + description`, with description truncated to **1200 characters**
+- Decision threshold for class `tech`: **0.3765**
+
+## Usage
+
+```python
+import torch
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+
+MODEL_ID = "AndreiTolmachev/dev_da_roles_1"
+THRESHOLD = 0.3765
+
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
+model = AutoModelForSequenceClassification.from_pretrained(MODEL_ID).eval()
+
+def is_developer_or_data_analyst(title: str, description: str = "") -> bool:
+    text = (title + " " + description[:1200]).strip()
+    enc = tokenizer(text, truncation=True, max_length=256, return_tensors="pt")
+    with torch.no_grad():
+        logits = model(**enc).logits
+    prob_tech = torch.softmax(logits, dim=-1)[0, 1].item()
+    return prob_tech >= THRESHOLD
+
+print(is_developer_or_data_analyst(
+    "Data Analyst",
+    "SQL, Python, dashboards, product metrics, A/B tests..."
+))
+```
+
+## Architecture
+
+- Model: `BertForSequenceClassification`
+- Base model: `cointegrated/rubert-tiny2`
+- Layers: 3, hidden size: 312, attention heads: 12
+- Vocab size: 83,828
+- Parameters: ~29M
+- `max_position_embeddings`: 2048
+
+## Training
+
+- Dataset: internal job-vacancy dataset (`vacancies_labeled.csv`), labeled by an LLM pipeline
+- Loss: weighted cross-entropy
+- Threshold selected for target recall = **0.95**
+- Positive class includes developer roles and Data Analyst, excluding team leads
+
+## Limitations
+
+- Trained primarily on Russian-language IT job vacancies; quality on other domains/languages is not guaranteed.
+- Team lead and management roles are treated as `other` by design.
+- Description is truncated to 1200 characters before tokenization.
+- The model intentionally groups developers and Data Analysts into one positive class; it does not distinguish between them.
+
+## Version
+
+Hub tag: `v1.0-dev-da-r95`
+
+## License
+
+MIT.

config.json

@@ -0,0 +1,39 @@
+{
+  "add_cross_attention": false,
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": null,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "emb_size": 312,
+  "eos_token_id": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 312,
+  "id2label": {
+    "0": "other",
+    "1": "tech"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 600,
+  "is_decoder": false,
+  "label2id": {
+    "other": 0,
+    "tech": 1
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 2048,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 3,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "tie_word_embeddings": true,
+  "transformers_version": "5.7.0",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 83828
+}

meta.json

@@ -0,0 +1,29 @@
+{
+  "task": "dev_da_vs_other_binary",
+  "positive_definition": "role_category in TECH_CLASSES AND team_lead==0",
+  "tech_classes": [
+    "Backend",
+    "Data Analyst",
+    "Desktop / Systems",
+    "Embedded",
+    "Frontend",
+    "Fullstack",
+    "ML / AI / Data Scientist",
+    "Mobile"
+  ],
+  "labels": [
+    "other",
+    "tech"
+  ],
+  "max_len": 256,
+  "description_chars": 1200,
+  "base_model": "cointegrated/rubert-tiny2",
+  "trained_on": "vacancies_labeled.csv",
+  "best_epoch": 8,
+  "best_threshold": 0.37654510140419006,
+  "best_precision_at_threshold": 0.8796791443850267,
+  "best_recall_at_threshold": 0.9508670520231214,
+  "best_roc_auc": 0.9790331422674761,
+  "target_recall": 0.95,
+  "pos_weight": 2.481304126337239
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d3eff74a0776fad17bbff6f6ac030ca9f884bb5e795fc7f1387f76bc2e815eb
+size 116784136

tokenizer_config.json

@@ -0,0 +1,24 @@
+{
+  "backend": "tokenizers",
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "[MASK]",
+  "max_length": 512,
+  "model_max_length": 2048,
+  "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]"
+}