Upload Italian PII detection model OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1

README.md

@@ -0,0 +1,290 @@
+---
+language:
+  - it
+license: apache-2.0
+base_model: microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext
+tags:
+  - token-classification
+  - ner
+  - pii
+  - pii-detection
+  - de-identification
+  - privacy
+  - healthcare
+  - medical
+  - clinical
+  - phi
+  - italian
+  - pytorch
+  - transformers
+  - openmed
+pipeline_tag: token-classification
+library_name: transformers
+metrics:
+  - f1
+  - precision
+  - recall
+model-index:
+  - name: OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1
+    results:
+      - task:
+          type: token-classification
+          name: Named Entity Recognition
+        dataset:
+          name: AI4Privacy (Italian subset)
+          type: ai4privacy/pii-masking-400k
+          split: test
+        metrics:
+          - type: f1
+            value: 0.9255
+            name: F1 (micro)
+          - type: precision
+            value: 0.9197
+            name: Precision
+          - type: recall
+            value: 0.9313
+            name: Recall
+widget:
+  - text: "Dr. Marco Rossi (Codice Fiscale: RSSMRC85C15H501Z) può essere contattato a marco.rossi@ospedale.it o al +39 333 123 4567. Abita in Via Roma 25, 00184 Roma."
+    example_title: Clinical Note with PII (Italian)
+---
+
+# OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1
+
+**Italian PII Detection Model** | 110M Parameters | Open Source
+
+[![F1 Score](https://img.shields.io/badge/F1-92.55%25-brightgreen)]() [![Precision](https://img.shields.io/badge/Precision-91.97%25-blue)]() [![Recall](https://img.shields.io/badge/Recall-93.13%25-orange)]()
+
+## Model Description
+
+**OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1** is a transformer-based token classification model fine-tuned for **Personally Identifiable Information (PII) detection in Italian text**. This model identifies and classifies **54 types of sensitive information** including names, addresses, social security numbers, medical record numbers, and more.
+
+### Key Features
+
+- **Italian-Optimized**: Specifically trained on Italian text for optimal performance
+- **High Accuracy**: Achieves strong F1 scores across diverse PII categories
+- **Comprehensive Coverage**: Detects 55+ entity types spanning personal, financial, medical, and contact information
+- **Privacy-Focused**: Designed for de-identification and compliance with GDPR and other privacy regulations
+- **Production-Ready**: Optimized for real-world text processing pipelines
+
+## Performance
+
+Evaluated on the Italian subset of AI4Privacy dataset:
+
+| Metric | Score |
+|:---|:---:|
+| **Micro F1** | **0.9255** |
+| Precision | 0.9197 |
+| Recall | 0.9313 |
+| Macro F1 | 0.9105 |
+| Weighted F1 | 0.9210 |
+| Accuracy | 0.9893 |
+
+## Supported Entity Types
+
+This model detects **54 PII entity types** organized into categories:
+
+<details>
+<summary><strong>Identifiers</strong> (22 types)</summary>
+
+| Entity | Description |
+|:---|:---|
+| `ACCOUNTNAME` | Accountname |
+| `BANKACCOUNT` | Bankaccount |
+| `BIC` | Bic |
+| `BITCOINADDRESS` | Bitcoinaddress |
+| `CREDITCARD` | Creditcard |
+| `CREDITCARDISSUER` | Creditcardissuer |
+| `CVV` | Cvv |
+| `ETHEREUMADDRESS` | Ethereumaddress |
+| `IBAN` | Iban |
+| `IMEI` | Imei |
+| ... | *and 12 more* |
+
+</details>
+
+<details>
+<summary><strong>Personal Info</strong> (11 types)</summary>
+
+| Entity | Description |
+|:---|:---|
+| `AGE` | Age |
+| `DATEOFBIRTH` | Dateofbirth |
+| `EYECOLOR` | Eyecolor |
+| `FIRSTNAME` | Firstname |
+| `GENDER` | Gender |
+| `HEIGHT` | Height |
+| `LASTNAME` | Lastname |
+| `MIDDLENAME` | Middlename |
+| `OCCUPATION` | Occupation |
+| `PREFIX` | Prefix |
+| ... | *and 1 more* |
+
+</details>
+
+<details>
+<summary><strong>Contact Info</strong> (2 types)</summary>
+
+| Entity | Description |
+|:---|:---|
+| `EMAIL` | Email |
+| `PHONE` | Phone |
+
+</details>
+
+<details>
+<summary><strong>Location</strong> (9 types)</summary>
+
+| Entity | Description |
+|:---|:---|
+| `BUILDINGNUMBER` | Buildingnumber |
+| `CITY` | City |
+| `COUNTY` | County |
+| `GPSCOORDINATES` | Gpscoordinates |
+| `ORDINALDIRECTION` | Ordinaldirection |
+| `SECONDARYADDRESS` | Secondaryaddress |
+| `STATE` | State |
+| `STREET` | Street |
+| `ZIPCODE` | Zipcode |
+
+</details>
+
+<details>
+<summary><strong>Organization</strong> (3 types)</summary>
+
+| Entity | Description |
+|:---|:---|
+| `JOBDEPARTMENT` | Jobdepartment |
+| `JOBTITLE` | Jobtitle |
+| `ORGANIZATION` | Organization |
+
+</details>
+
+<details>
+<summary><strong>Financial</strong> (5 types)</summary>
+
+| Entity | Description |
+|:---|:---|
+| `AMOUNT` | Amount |
+| `CURRENCY` | Currency |
+| `CURRENCYCODE` | Currencycode |
+| `CURRENCYNAME` | Currencyname |
+| `CURRENCYSYMBOL` | Currencysymbol |
+
+</details>
+
+<details>
+<summary><strong>Temporal</strong> (2 types)</summary>
+
+| Entity | Description |
+|:---|:---|
+| `DATE` | Date |
+| `TIME` | Time |
+
+</details>
+
+## Usage
+
+### Quick Start
+
+```python
+from transformers import pipeline
+
+# Load the PII detection pipeline
+ner = pipeline("ner", model="OpenMed/OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1", aggregation_strategy="simple")
+
+text = """
+Paziente Marco Bianchi (nato il 15/03/1985, CF: BNCMRC85C15H501Z) è stato visitato oggi.
+Contatto: marco.bianchi@email.it, Telefono: +39 333 123 4567.
+Indirizzo: Via Garibaldi 42, 20121 Milano.
+"""
+
+entities = ner(text)
+for entity in entities:
+    print(f"{entity['entity_group']}: {entity['word']} (score: {entity['score']:.3f})")
+```
+
+### De-identification Example
+
+```python
+def redact_pii(text, entities, placeholder='[REDACTED]'):
+    """Replace detected PII with placeholders."""
+    # Sort entities by start position (descending) to preserve offsets
+    sorted_entities = sorted(entities, key=lambda x: x['start'], reverse=True)
+    redacted = text
+    for ent in sorted_entities:
+        redacted = redacted[:ent['start']] + f"[{ent['entity_group']}]" + redacted[ent['end']:]
+    return redacted
+
+# Apply de-identification
+redacted_text = redact_pii(text, entities)
+print(redacted_text)
+```
+
+### Batch Processing
+
+```python
+from transformers import AutoModelForTokenClassification, AutoTokenizer
+import torch
+
+model_name = "OpenMed/OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1"
+model = AutoModelForTokenClassification.from_pretrained(model_name)
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+
+texts = [
+    "Paziente Marco Bianchi (nato il 15/03/1985, CF: BNCMRC85C15H501Z) è stato visitato oggi.",
+    "Contatto: marco.bianchi@email.it, Telefono: +39 333 123 4567.",
+]
+
+inputs = tokenizer(texts, return_tensors='pt', padding=True, truncation=True)
+with torch.no_grad():
+    outputs = model(**inputs)
+    predictions = torch.argmax(outputs.logits, dim=-1)
+```
+
+## Training Details
+
+### Dataset
+
+- **Source**: [AI4Privacy PII Masking 400k](https://huggingface.co/datasets/ai4privacy/pii-masking-400k) (Italian subset)
+- **Format**: BIO-tagged token classification
+- **Labels**: 109 total (54 entity types × 2 BIO tags + O)
+
+### Training Configuration
+
+- **Max Sequence Length**: 512 tokens
+- **Epochs**: 3
+- **Framework**: Hugging Face Transformers + Trainer API
+
+## Intended Use & Limitations
+
+### Intended Use
+
+- **De-identification**: Automated redaction of PII in Italian clinical notes, medical records, and documents
+- **Compliance**: Supporting GDPR, and other privacy regulation compliance
+- **Data Preprocessing**: Preparing datasets for research by removing sensitive information
+- **Audit Support**: Identifying PII in document collections
+
+### Limitations
+
+**Important**: This model is intended as an **assistive tool**, not a replacement for human review.
+
+- **False Negatives**: Some PII may not be detected; always verify critical applications
+- **Context Sensitivity**: Performance may vary with domain-specific terminology
+- **Language**: Optimized for Italian text; may not perform well on other languages
+
+## Citation
+
+```bibtex
+@misc{openmed-pii-2026,
+  title = {OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1: Italian PII Detection Model},
+  author = {OpenMed Science},
+  year = {2026},
+  publisher = {Hugging Face},
+  url = {https://huggingface.co/OpenMed/OpenMed-PII-Italian-BiomedBERTFull-Base-110M-v1}
+}
+```
+
+## Links
+
+- **Organization**: [OpenMed](https://huggingface.co/OpenMed)

all_results.json

@@ -0,0 +1,28 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9902273864029857,
+    "eval_f1": 0.9330852503382949,
+    "eval_loss": 0.031441979110240936,
+    "eval_macro_f1": 0.9175882104431093,
+    "eval_precision": 0.9287494107347296,
+    "eval_recall": 0.9374617633063694,
+    "eval_runtime": 4.0213,
+    "eval_samples_per_second": 1236.418,
+    "eval_steps_per_second": 19.397,
+    "eval_weighted_f1": 0.9296538337362635,
+    "test_accuracy": 0.9893058468311692,
+    "test_f1": 0.9254852849092048,
+    "test_loss": 0.033455222845077515,
+    "test_macro_f1": 0.9105334066706021,
+    "test_precision": 0.9196960765048798,
+    "test_recall": 0.9313478376227116,
+    "test_runtime": 4.5862,
+    "test_samples_per_second": 1105.272,
+    "test_steps_per_second": 17.444,
+    "test_weighted_f1": 0.9210243739638152,
+    "total_flos": 4715705190580224.0,
+    "train_loss": 0.21988849894454082,
+    "train_runtime": 225.9955,
+    "train_samples_per_second": 543.515,
+    "train_steps_per_second": 8.496
+}

classification_report.txt

@@ -0,0 +1,64 @@
+Classification Report for Italian PII Detection
+Model: microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext
+============================================================
+
+                  precision    recall  f1-score   support
+
+     ACCOUNTNAME       0.99      1.00      0.99       282
+             AGE       0.98      0.99      0.98       338
+          AMOUNT       0.96      0.89      0.92       116
+     BANKACCOUNT       0.97      1.00      0.99       306
+             BIC       0.95      0.95      0.95        77
+  BITCOINADDRESS       0.90      1.00      0.95       273
+  BUILDINGNUMBER       0.92      0.91      0.91       346
+            CITY       0.94      0.83      0.88       280
+          COUNTY       0.92      1.00      0.96       327
+      CREDITCARD       0.80      0.79      0.80       302
+CREDITCARDISSUER       0.96      1.00      0.98       146
+        CURRENCY       0.62      0.89      0.73       187
+    CURRENCYCODE       0.74      0.71      0.72        85
+    CURRENCYNAME       0.00      0.00      0.00        97
+  CURRENCYSYMBOL       0.95      0.94      0.94       308
+             CVV       0.97      0.96      0.96        97
+            DATE       0.62      0.93      0.75       423
+     DATEOFBIRTH       0.60      0.33      0.42       327
+           EMAIL       0.99      1.00      1.00       423
+ ETHEREUMADDRESS       1.00      1.00      1.00       168
+        EYECOLOR       0.98      0.98      0.98       108
+       FIRSTNAME       0.89      0.92      0.91      1623
+          GENDER       0.98      0.99      0.99       302
+  GPSCOORDINATES       1.00      1.00      1.00       223
+          HEIGHT       0.95      1.00      0.98       126
+            IBAN       0.97      1.00      0.98       230
+            IMEI       1.00      1.00      1.00       215
+       IPADDRESS       1.00      1.00      1.00       783
+   JOBDEPARTMENT       0.92      0.97      0.95       327
+        JOBTITLE       0.98      1.00      0.99       279
+        LASTNAME       0.85      0.90      0.87       441
+ LITECOINADDRESS       1.00      0.61      0.76        83
+      MACADDRESS       0.99      1.00      1.00       114
+    MASKEDNUMBER       0.71      0.72      0.71       209
+      MIDDLENAME       0.74      0.64      0.68       310
+      OCCUPATION       0.98      0.99      0.99       323
+ORDINALDIRECTION       1.00      1.00      1.00       152
+    ORGANIZATION       0.97      1.00      0.98       271
+        PASSWORD       0.98      0.98      0.98       286
+           PHONE       1.00      0.99      1.00       303
+             PIN       0.88      0.88      0.88        72
+          PREFIX       0.97      1.00      0.99       298
+SECONDARYADDRESS       0.98      1.00      0.99       316
+             SEX       1.00      1.00      1.00       338
+             SSN       0.99      1.00      1.00       259
+           STATE       0.93      0.97      0.95       294
+          STREET       0.94      0.98      0.96       332
+            TIME       0.94      0.99      0.97       296
+             URL       1.00      1.00      1.00       244
+       USERAGENT       1.00      1.00      1.00       233
+        USERNAME       0.98      0.98      0.98       332
+             VIN       1.00      0.99      0.99        84
+             VRM       0.94      1.00      0.97        98
+         ZIPCODE       0.91      0.92      0.91       264
+
+       micro avg       0.92      0.93      0.93     15076
+       macro avg       0.91      0.92      0.91     15076
+    weighted avg       0.92      0.93      0.92     15076

config.json

@@ -0,0 +1,180 @@
+{
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "O",
+    "1": "B-ACCOUNTNAME",
+    "2": "B-AGE",
+    "3": "B-AMOUNT",
+    "4": "B-BANKACCOUNT",
+    "5": "B-BIC",
+    "6": "B-BITCOINADDRESS",
+    "7": "B-BUILDINGNUMBER",
+    "8": "B-CITY",
+    "9": "B-COUNTY",
+    "10": "B-CREDITCARD",
+    "11": "B-CREDITCARDISSUER",
+    "12": "B-CURRENCY",
+    "13": "B-CURRENCYCODE",
+    "14": "B-CURRENCYNAME",
+    "15": "B-CURRENCYSYMBOL",
+    "16": "B-CVV",
+    "17": "B-DATE",
+    "18": "B-DATEOFBIRTH",
+    "19": "B-EMAIL",
+    "20": "B-ETHEREUMADDRESS",
+    "21": "B-EYECOLOR",
+    "22": "B-FIRSTNAME",
+    "23": "B-GENDER",
+    "24": "B-GPSCOORDINATES",
+    "25": "B-HEIGHT",
+    "26": "B-IBAN",
+    "27": "B-IMEI",
+    "28": "B-IPADDRESS",
+    "29": "B-JOBDEPARTMENT",
+    "30": "B-JOBTITLE",
+    "31": "B-LASTNAME",
+    "32": "B-LITECOINADDRESS",
+    "33": "B-MACADDRESS",
+    "34": "B-MASKEDNUMBER",
+    "35": "B-MIDDLENAME",
+    "36": "B-OCCUPATION",
+    "37": "B-ORDINALDIRECTION",
+    "38": "B-ORGANIZATION",
+    "39": "B-PASSWORD",
+    "40": "B-PHONE",
+    "41": "B-PIN",
+    "42": "B-PREFIX",
+    "43": "B-SECONDARYADDRESS",
+    "44": "B-SEX",
+    "45": "B-SSN",
+    "46": "B-STATE",
+    "47": "B-STREET",
+    "48": "B-TIME",
+    "49": "B-URL",
+    "50": "B-USERAGENT",
+    "51": "B-USERNAME",
+    "52": "B-VIN",
+    "53": "B-VRM",
+    "54": "B-ZIPCODE",
+    "55": "I-ACCOUNTNAME",
+    "56": "I-AGE",
+    "57": "I-AMOUNT",
+    "58": "I-CITY",
+    "59": "I-COUNTY",
+    "60": "I-CURRENCY",
+    "61": "I-CURRENCYNAME",
+    "62": "I-DATE",
+    "63": "I-DATEOFBIRTH",
+    "64": "I-EYECOLOR",
+    "65": "I-GENDER",
+    "66": "I-HEIGHT",
+    "67": "I-JOBTITLE",
+    "68": "I-ORGANIZATION",
+    "69": "I-PHONE",
+    "70": "I-SECONDARYADDRESS",
+    "71": "I-SSN",
+    "72": "I-STATE",
+    "73": "I-STREET",
+    "74": "I-TIME",
+    "75": "I-USERAGENT"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "B-ACCOUNTNAME": 1,
+    "B-AGE": 2,
+    "B-AMOUNT": 3,
+    "B-BANKACCOUNT": 4,
+    "B-BIC": 5,
+    "B-BITCOINADDRESS": 6,
+    "B-BUILDINGNUMBER": 7,
+    "B-CITY": 8,
+    "B-COUNTY": 9,
+    "B-CREDITCARD": 10,
+    "B-CREDITCARDISSUER": 11,
+    "B-CURRENCY": 12,
+    "B-CURRENCYCODE": 13,
+    "B-CURRENCYNAME": 14,
+    "B-CURRENCYSYMBOL": 15,
+    "B-CVV": 16,
+    "B-DATE": 17,
+    "B-DATEOFBIRTH": 18,
+    "B-EMAIL": 19,
+    "B-ETHEREUMADDRESS": 20,
+    "B-EYECOLOR": 21,
+    "B-FIRSTNAME": 22,
+    "B-GENDER": 23,
+    "B-GPSCOORDINATES": 24,
+    "B-HEIGHT": 25,
+    "B-IBAN": 26,
+    "B-IMEI": 27,
+    "B-IPADDRESS": 28,
+    "B-JOBDEPARTMENT": 29,
+    "B-JOBTITLE": 30,
+    "B-LASTNAME": 31,
+    "B-LITECOINADDRESS": 32,
+    "B-MACADDRESS": 33,
+    "B-MASKEDNUMBER": 34,
+    "B-MIDDLENAME": 35,
+    "B-OCCUPATION": 36,
+    "B-ORDINALDIRECTION": 37,
+    "B-ORGANIZATION": 38,
+    "B-PASSWORD": 39,
+    "B-PHONE": 40,
+    "B-PIN": 41,
+    "B-PREFIX": 42,
+    "B-SECONDARYADDRESS": 43,
+    "B-SEX": 44,
+    "B-SSN": 45,
+    "B-STATE": 46,
+    "B-STREET": 47,
+    "B-TIME": 48,
+    "B-URL": 49,
+    "B-USERAGENT": 50,
+    "B-USERNAME": 51,
+    "B-VIN": 52,
+    "B-VRM": 53,
+    "B-ZIPCODE": 54,
+    "I-ACCOUNTNAME": 55,
+    "I-AGE": 56,
+    "I-AMOUNT": 57,
+    "I-CITY": 58,
+    "I-COUNTY": 59,
+    "I-CURRENCY": 60,
+    "I-CURRENCYNAME": 61,
+    "I-DATE": 62,
+    "I-DATEOFBIRTH": 63,
+    "I-EYECOLOR": 64,
+    "I-GENDER": 65,
+    "I-HEIGHT": 66,
+    "I-JOBTITLE": 67,
+    "I-ORGANIZATION": 68,
+    "I-PHONE": 69,
+    "I-SECONDARYADDRESS": 70,
+    "I-SSN": 71,
+    "I-STATE": 72,
+    "I-STREET": 73,
+    "I-TIME": 74,
+    "I-USERAGENT": 75,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.57.3",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

eval_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9902273864029857,
+    "eval_f1": 0.9330852503382949,
+    "eval_loss": 0.031441979110240936,
+    "eval_macro_f1": 0.9175882104431093,
+    "eval_precision": 0.9287494107347296,
+    "eval_recall": 0.9374617633063694,
+    "eval_runtime": 4.0213,
+    "eval_samples_per_second": 1236.418,
+    "eval_steps_per_second": 19.397,
+    "eval_weighted_f1": 0.9296538337362635
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c96ab95d8ac8dba8b8cf486ba7a62a3b527a36acc7fa00fc3a3927a1038831e
+size 435823712

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

test_results.json

@@ -0,0 +1,12 @@
+{
+    "test_accuracy": 0.9893058468311692,
+    "test_f1": 0.9254852849092048,
+    "test_loss": 0.033455222845077515,
+    "test_macro_f1": 0.9105334066706021,
+    "test_precision": 0.9196960765048798,
+    "test_recall": 0.9313478376227116,
+    "test_runtime": 4.5862,
+    "test_samples_per_second": 1105.272,
+    "test_steps_per_second": 17.444,
+    "test_weighted_f1": 0.9210243739638152
+}

tokenizer_config.json

@@ -0,0 +1,58 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "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,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 4715705190580224.0,
+    "train_loss": 0.21988849894454082,
+    "train_runtime": 225.9955,
+    "train_samples_per_second": 543.515,
+    "train_steps_per_second": 8.496
+}