Upload fine-tuned BERT model for Detectra AI
Browse files- README.md +37 -0
- config.json +39 -0
- model.safetensors +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +19 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
library_name: transformers
|
| 6 |
+
pipeline_tag: text-classification
|
| 7 |
+
tags:
|
| 8 |
+
- bert
|
| 9 |
+
- text-classification
|
| 10 |
+
- ai-generated-text-detection
|
| 11 |
+
- detectra-ai
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# Detectra AI - BERT Text Classifier
|
| 15 |
+
|
| 16 |
+
Model ini merupakan hasil fine-tuning `bert-base-uncased` untuk klasifikasi teks biner:
|
| 17 |
+
|
| 18 |
+
- `TEKS_MANUSIA` / label 0: teks yang ditulis manusia
|
| 19 |
+
- `TEKS_AI` / label 1: teks yang dihasilkan AI
|
| 20 |
+
|
| 21 |
+
## Kegunaan Model
|
| 22 |
+
|
| 23 |
+
Model digunakan pada aplikasi **Detectra AI** untuk memperkirakan apakah sebuah teks memiliki pola yang lebih dekat dengan tulisan manusia atau teks hasil generasi AI.
|
| 24 |
+
|
| 25 |
+
## Dataset
|
| 26 |
+
|
| 27 |
+
Model dilatih menggunakan dataset AI vs Human Text berbahasa Inggris yang telah melalui proses:
|
| 28 |
+
- penghapusan data kosong dan duplikat,
|
| 29 |
+
- balancing kelas,
|
| 30 |
+
- sampling seimbang,
|
| 31 |
+
- pembagian data train, validation, dan test.
|
| 32 |
+
|
| 33 |
+
## Batasan
|
| 34 |
+
|
| 35 |
+
Model dilatih menggunakan teks berbahasa Inggris. Prediksi terhadap teks berbahasa Indonesia belum dapat dianggap valid tanpa pelatihan atau evaluasi tambahan menggunakan dataset bahasa Indonesia.
|
| 36 |
+
|
| 37 |
+
Hasil prediksi model tidak dapat digunakan sebagai satu-satunya dasar untuk menentukan plagiarisme, pelanggaran akademik, atau keaslian suatu tulisan.
|
config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"dtype": "float32",
|
| 10 |
+
"eos_token_id": null,
|
| 11 |
+
"gradient_checkpointing": false,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "TEKS_MANUSIA",
|
| 17 |
+
"1": "TEKS_AI"
|
| 18 |
+
},
|
| 19 |
+
"initializer_range": 0.02,
|
| 20 |
+
"intermediate_size": 3072,
|
| 21 |
+
"is_decoder": false,
|
| 22 |
+
"label2id": {
|
| 23 |
+
"TEKS_AI": 1,
|
| 24 |
+
"TEKS_MANUSIA": 0
|
| 25 |
+
},
|
| 26 |
+
"layer_norm_eps": 1e-12,
|
| 27 |
+
"max_position_embeddings": 512,
|
| 28 |
+
"model_type": "bert",
|
| 29 |
+
"num_attention_heads": 12,
|
| 30 |
+
"num_hidden_layers": 12,
|
| 31 |
+
"pad_token_id": 0,
|
| 32 |
+
"position_embedding_type": "absolute",
|
| 33 |
+
"problem_type": "single_label_classification",
|
| 34 |
+
"tie_word_embeddings": true,
|
| 35 |
+
"transformers_version": "5.9.0",
|
| 36 |
+
"type_vocab_size": 2,
|
| 37 |
+
"use_cache": false,
|
| 38 |
+
"vocab_size": 30522
|
| 39 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2f03588b94b79611885a3d710fe7a6d7096dc8f50364ff00d98a208acfbb935
|
| 3 |
+
size 437958624
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_lower_case": true,
|
| 5 |
+
"is_local": true,
|
| 6 |
+
"local_files_only": true,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"max_length": 128,
|
| 9 |
+
"model_max_length": 512,
|
| 10 |
+
"pad_token": "[PAD]",
|
| 11 |
+
"sep_token": "[SEP]",
|
| 12 |
+
"stride": 0,
|
| 13 |
+
"strip_accents": null,
|
| 14 |
+
"tokenize_chinese_chars": true,
|
| 15 |
+
"tokenizer_class": "BertTokenizer",
|
| 16 |
+
"truncation_side": "right",
|
| 17 |
+
"truncation_strategy": "longest_first",
|
| 18 |
+
"unk_token": "[UNK]"
|
| 19 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6fd333c45293a7f54a9d0113c40289ed84175dbaefef10ca4c7376601b1e18b1
|
| 3 |
+
size 5265
|