Update README.md

README.md

@@ -1,383 +1,113 @@
 ---
+language: "id"
+license: "apache-2.0"
 tags:
-- sentence-transformers
-- sentence-similarity
-- feature-extraction
-- generated_from_trainer
-- dataset_size:14740
-- loss:CosineSimilarityLoss
-base_model: indobenchmark/indobert-large-p2
-widget:
-- source_sentence: Hal tersebut bukanlah tanggung jawab langsung kepada konstituen.
-  sentences:
-  - Seorang wanita sedang menembakkan pistol.
-  - Seseorang melempar kucing ke langit-langit.
-  - Ya, tetapi mereka bertanggung jawab kepada konstituen mereka.
-- source_sentence: Tidak ada kemenangan bagi Obama di kalangan konservatif.
-  sentences:
-  - Ini sangat kaya dari seorang konservatif.
-  - komunitas global harus bekerja sama untuk mengakhiri perdagangan gelap senjata
-    kecil dan senjata ringan.
-  - Bendera Amerika Serikat tertiup angin.
-- source_sentence: Pasukan AS Tewas dalam Serangan Orang Dalam Afghanistan
-  sentences:
-  - Seekor anjing cokelat berlari melintasi rerumputan.
-  - Pasukan NATO tewas dalam 'serangan orang dalam' di Afghanistan
-  - Sering berlatih bahasa asing
-- source_sentence: Pakta perbatasan tinta India dan Cina; 8 perjanjian lainnya
-  sentences:
-  - Sering membawa tas kecil saat jalan-jalan
-  - Jarang menggunakan Grab untuk kirim dokumen
-  - India dan Cina menorehkan kesepakatan tentang sungai lintas batas
-- source_sentence: Seorang anak laki-laki kecil bermain di salju.
-  sentences:
-  - Anjing berwarna cokelat dan putih sedang bermain di atas salju.
-  - Seorang gadis sedang memainkan seruling.
-  - Kucing domestik yang sedang berbaring di belakang kotoran kucing.
-pipeline_tag: sentence-similarity
-library_name: sentence-transformers
+  - sentence-transformers
+  - indonesian
+  - semantic-similarity
+  - stsb
+  - embedding
+  - fine-tuned
+  - education
+datasets:
+  - rzkamalia/stsb-indo-mt-modified
+  - quarkss/stsb-indo-mt
+  - AkshitaS/semrel_2024_plus
 metrics:
-- pearson_cosine
-- spearman_cosine
+  - cosine-similarity
+pipeline_tag: sentence-similarity
 model-index:
-- name: SentenceTransformer based on indobenchmark/indobert-large-p2
-  results:
-  - task:
-      type: semantic-similarity
-      name: Semantic Similarity
-    dataset:
-      name: sts validation
-      type: sts-validation
-    metrics:
-    - type: pearson_cosine
-      value: 0.8588187163579742
-      name: Pearson Cosine
-    - type: spearman_cosine
-      value: 0.8562693942139418
-      name: Spearman Cosine
+  - name: Automatic Scoring (IndoBERT STS)
+    results:
+      - task:
+          name: Semantic Textual Similarity
+          type: sentence-similarity
+        dataset:
+          name: STSB Indo + SemRel 2024
+          type: multiple
+        metrics:
+          - name: Cosine Similarity
+            type: cosine-similarity
+            value: Evaluated on test set (see below)
 ---
 
-# SentenceTransformer based on indobenchmark/indobert-large-p2
-
-This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [indobenchmark/indobert-large-p2](https://huggingface.co/indobenchmark/indobert-large-p2). It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
-
-## Model Details
-
-### Model Description
-- **Model Type:** Sentence Transformer
-- **Base model:** [indobenchmark/indobert-large-p2](https://huggingface.co/indobenchmark/indobert-large-p2) <!-- at revision 4b280c3bfcc1ed2d6b4589be5c876076b7d73568 -->
-- **Maximum Sequence Length:** 512 tokens
-- **Output Dimensionality:** 1024 dimensions
-- **Similarity Function:** Cosine Similarity
-<!-- - **Training Dataset:** Unknown -->
-<!-- - **Language:** Unknown -->
-<!-- - **License:** Unknown -->
+# Automatic Scoring for Indonesian Semantic Similarity ✨
 
-### Model Sources
+Model ini merupakan hasil fine-tuning dari [`indobenchmark/indobert-large-p2`](https://huggingface.co/indobenchmark/indobert-large-p2) menggunakan Sentence Transformers untuk tugas **Semantic Textual Similarity** (STS) dalam bahasa Indonesia.
 
-- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
-- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
-- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
+Model ini dilatih secara **multi-dataset** menggunakan gabungan dari:
+- 🟢 `rzkamalia/stsb-indo-mt-modified`
+- 🟢 `quarkss/stsb-indo-mt`
+- 🟢 `AkshitaS/semrel_2024_plus` (split `ind_Latn`)
 
-### Full Model Architecture
+Tujuan utama dari model ini adalah untuk mendukung **penilaian otomatis jawaban siswa** atau sistem pembelajaran berbasis teks dalam bahasa Indonesia.
 
-```
-SentenceTransformer(
-  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel 
-  (1): Pooling({'word_embedding_dimension': 1024, '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})
-)
-```
+## 🧠 Model Details
 
-## Usage
+- **Base Model**: [`indobenchmark/indobert-large-p2`](https://huggingface.co/indobenchmark/indobert-large-p2)
+- **Framework**: `sentence-transformers`
+- **Loss Function**: `CosineSimilarityLoss`
+- **Training Epochs**: `5`
+- **Batch Size**: `16`
+- **Evaluation Metric**: `Cosine Similarity`
+- **Total Datasets Combined**: 3 corpora (STS Indo + Semantic Relation)
 
-### Direct Usage (Sentence Transformers)
+## 📊 Example Usage
 
-First install the Sentence Transformers library:
-
-```bash
-pip install -U sentence-transformers
-```
-
-Then you can load this model and run inference.
 ```python
-from sentence_transformers import SentenceTransformer
-
-# Download from the 🤗 Hub
-model = SentenceTransformer("sentence_transformers_model_id")
-# Run inference
-sentences = [
-    'Seorang anak laki-laki kecil bermain di salju.',
-    'Anjing berwarna cokelat dan putih sedang bermain di atas salju.',
-    'Seorang gadis sedang memainkan seruling.',
-]
-embeddings = model.encode(sentences)
-print(embeddings.shape)
-# [3, 1024]
-
-# Get the similarity scores for the embeddings
-similarities = model.similarity(embeddings, embeddings)
-print(similarities.shape)
-# [3, 3]
-```
-
-<!--
-### Direct Usage (Transformers)
-
-<details><summary>Click to see the direct usage in Transformers</summary>
-
-</details>
--->
-
-<!--
-### Downstream Usage (Sentence Transformers)
-
-You can finetune this model on your own dataset.
+from sentence_transformers import SentenceTransformer, util
 
-<details><summary>Click to expand</summary>
+model = SentenceTransformer("eugene702/Automatic-Scoring")
 
-</details>
--->
-
-<!--
-### Out-of-Scope Use
-
-*List how the model may foreseeably be misused and address what users ought not to do with the model.*
--->
-
-## Evaluation
-
-### Metrics
-
-#### Semantic Similarity
-
-* Dataset: `sts-validation`
-* Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
-
-| Metric              | Value      |
-|:--------------------|:-----------|
-| pearson_cosine      | 0.8588     |
-| **spearman_cosine** | **0.8563** |
-
-<!--
-## 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 Dataset
-
-#### Unnamed Dataset
-
-* Size: 14,740 training samples
-* Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>label</code>
-* Approximate statistics based on the first 1000 samples:
-  |         | sentence_0                                                                        | sentence_1                                                                        | label                                                          |
-  |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:---------------------------------------------------------------|
-  | type    | string                                                                            | string                                                                            | float                                                          |
-  | details | <ul><li>min: 5 tokens</li><li>mean: 13.13 tokens</li><li>max: 47 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 12.98 tokens</li><li>max: 51 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 0.49</li><li>max: 1.0</li></ul> |
-* Samples:
-  | sentence_0                                                                                                                                                 | sentence_1                                                                                                                                                                                                               | label             |
-  |:-----------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------|
-  | <code>Sering memesan tiket di Tiket.com</code>                                                                                                             | <code>Pernah memesan tiket di Tiket.com</code>                                                                                                                                                                           | <code>0.75</code> |
-  | <code>Seorang pria memotong kentang.</code>                                                                                                                | <code>Seorang pria mengiris kentang.</code>                                                                                                                                                                              | <code>0.96</code> |
-  | <code>Beberapa ribu pasukan Infanteri ke-3, termasuk Tim Tempur Brigade ke-3 yang bermarkas di Fort Benning di Columbus, mulai kembali minggu lalu.</code> | <code>Beberapa ribu tentara, sebagian besar dari Tim Tempur Brigade ke-3 divisi yang bermarkas di Fort Benning di Columbus, mulai kembali minggu lalu, dengan penerbangan yang terus berlanjut hingga hari Jumat.</code> | <code>0.8</code>  |
-* Loss: [<code>CosineSimilarityLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosinesimilarityloss) with these parameters:
-  ```json
-  {
-      "loss_fct": "torch.nn.modules.loss.MSELoss"
-  }
-  ```
+score = util.cos_sim(
+    model.encode("Apa dampak pemanasan global?", convert_to_tensor=True),
+    model.encode("Bagaimana pengaruh perubahan iklim terhadap bumi?", convert_to_tensor=True)
+)
 
-### Training Hyperparameters
-#### Non-Default Hyperparameters
+print("Similarity Score:", score.item())
+```
 
-- `eval_strategy`: steps
-- `per_device_train_batch_size`: 16
-- `per_device_eval_batch_size`: 16
-- `num_train_epochs`: 5
-- `multi_dataset_batch_sampler`: round_robin
+## 📁 Datasets Used
 
-#### All Hyperparameters
-<details><summary>Click to expand</summary>
+| Dataset | Deskripsi |
+|--------|-----------|
+| [`rzkamalia/stsb-indo-mt-modified`](https://huggingface.co/datasets/rzkamalia/stsb-indo-mt-modified) | Versi modifikasi STS bahasa Indonesia |
+| [`quarkss/stsb-indo-mt`](https://huggingface.co/datasets/quarkss/stsb-indo-mt) | STS benchmark bahasa Indonesia |
+| [`AkshitaS/semrel_2024_plus`](https://huggingface.co/datasets/AkshitaS/semrel_2024_plus) | Dataset Semantic Relation multilingual split `ind_Latn` |
 
-- `overwrite_output_dir`: False
-- `do_predict`: False
-- `eval_strategy`: steps
-- `prediction_loss_only`: True
-- `per_device_train_batch_size`: 16
-- `per_device_eval_batch_size`: 16
-- `per_gpu_train_batch_size`: None
-- `per_gpu_eval_batch_size`: None
-- `gradient_accumulation_steps`: 1
-- `eval_accumulation_steps`: None
-- `torch_empty_cache_steps`: None
-- `learning_rate`: 5e-05
-- `weight_decay`: 0.0
-- `adam_beta1`: 0.9
-- `adam_beta2`: 0.999
-- `adam_epsilon`: 1e-08
-- `max_grad_norm`: 1
-- `num_train_epochs`: 5
-- `max_steps`: -1
-- `lr_scheduler_type`: linear
-- `lr_scheduler_kwargs`: {}
-- `warmup_ratio`: 0.0
-- `warmup_steps`: 0
-- `log_level`: passive
-- `log_level_replica`: warning
-- `log_on_each_node`: True
-- `logging_nan_inf_filter`: True
-- `save_safetensors`: True
-- `save_on_each_node`: False
-- `save_only_model`: False
-- `restore_callback_states_from_checkpoint`: False
-- `no_cuda`: False
-- `use_cpu`: False
-- `use_mps_device`: False
-- `seed`: 42
-- `data_seed`: None
-- `jit_mode_eval`: False
-- `use_ipex`: False
-- `bf16`: False
-- `fp16`: False
-- `fp16_opt_level`: O1
-- `half_precision_backend`: auto
-- `bf16_full_eval`: False
-- `fp16_full_eval`: False
-- `tf32`: None
-- `local_rank`: 0
-- `ddp_backend`: None
-- `tpu_num_cores`: None
-- `tpu_metrics_debug`: False
-- `debug`: []
-- `dataloader_drop_last`: False
-- `dataloader_num_workers`: 0
-- `dataloader_prefetch_factor`: None
-- `past_index`: -1
-- `disable_tqdm`: False
-- `remove_unused_columns`: True
-- `label_names`: None
-- `load_best_model_at_end`: False
-- `ignore_data_skip`: False
-- `fsdp`: []
-- `fsdp_min_num_params`: 0
-- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
-- `fsdp_transformer_layer_cls_to_wrap`: None
-- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
-- `deepspeed`: None
-- `label_smoothing_factor`: 0.0
-- `optim`: adamw_torch
-- `optim_args`: None
-- `adafactor`: False
-- `group_by_length`: False
-- `length_column_name`: length
-- `ddp_find_unused_parameters`: None
-- `ddp_bucket_cap_mb`: None
-- `ddp_broadcast_buffers`: False
-- `dataloader_pin_memory`: True
-- `dataloader_persistent_workers`: False
-- `skip_memory_metrics`: True
-- `use_legacy_prediction_loop`: False
-- `push_to_hub`: False
-- `resume_from_checkpoint`: None
-- `hub_model_id`: None
-- `hub_strategy`: every_save
-- `hub_private_repo`: None
-- `hub_always_push`: False
-- `gradient_checkpointing`: False
-- `gradient_checkpointing_kwargs`: None
-- `include_inputs_for_metrics`: False
-- `include_for_metrics`: []
-- `eval_do_concat_batches`: True
-- `fp16_backend`: auto
-- `push_to_hub_model_id`: None
-- `push_to_hub_organization`: None
-- `mp_parameters`: 
-- `auto_find_batch_size`: False
-- `full_determinism`: False
-- `torchdynamo`: None
-- `ray_scope`: last
-- `ddp_timeout`: 1800
-- `torch_compile`: False
-- `torch_compile_backend`: None
-- `torch_compile_mode`: None
-- `include_tokens_per_second`: False
-- `include_num_input_tokens_seen`: False
-- `neftune_noise_alpha`: None
-- `optim_target_modules`: None
-- `batch_eval_metrics`: False
-- `eval_on_start`: False
-- `use_liger_kernel`: False
-- `eval_use_gather_object`: False
-- `average_tokens_across_devices`: False
-- `prompts`: None
-- `batch_sampler`: batch_sampler
-- `multi_dataset_batch_sampler`: round_robin
+## 📈 Evaluation
 
-</details>
+Evaluasi dilakukan pada data `test` dari ketiga dataset yang digabung. Penilaian dilakukan menggunakan `EmbeddingSimilarityEvaluator` dari `sentence-transformers`.
 
-### Training Logs
-| Epoch  | Step | Training Loss | sts-validation_spearman_cosine |
-|:------:|:----:|:-------------:|:------------------------------:|
-| 1.0    | 461  | -             | 0.8410                         |
-| 1.0846 | 500  | 0.0736        | 0.8391                         |
-| 2.0    | 922  | -             | 0.8502                         |
-| 2.1692 | 1000 | 0.0172        | 0.8524                         |
-| 3.0    | 1383 | -             | 0.8545                         |
-| 3.2538 | 1500 | 0.0095        | 0.8551                         |
-| 4.0    | 1844 | -             | 0.8543                         |
-| 4.3384 | 2000 | 0.0067        | 0.8551                         |
-| 5.0    | 2305 | -             | 0.8563                         |
+**Metric utama**: *Cosine Similarity* terhadap pasangan kalimat dalam bahasa Indonesia.
 
+## 💡 Use Cases
 
-### Framework Versions
-- Python: 3.11.13
-- Sentence Transformers: 4.1.0
-- Transformers: 4.52.4
-- PyTorch: 2.6.0+cu124
-- Accelerate: 1.8.1
-- Datasets: 3.6.0
-- Tokenizers: 0.21.2
+- Penilaian otomatis jawaban siswa
+- Deteksi parafrase dalam Bahasa Indonesia
+- Penilaian kesamaan kalimat untuk e-learning
+- Analisis pertanyaan dan jawaban semantik
 
-## Citation
+## 🛠 Training Code
 
-### BibTeX
+Model dilatih menggunakan `sentence-transformers` di platform Kaggle.
+Kode pelatihan tersedia secara privat namun dapat diminta melalui email.
 
-#### Sentence Transformers
-```bibtex
-@inproceedings{reimers-2019-sentence-bert,
-    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
-    author = "Reimers, Nils and Gurevych, Iryna",
-    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
-    month = "11",
-    year = "2019",
-    publisher = "Association for Computational Linguistics",
-    url = "https://arxiv.org/abs/1908.10084",
-}
-```
+## 📌 Model Availability
 
-<!--
-## Glossary
+Model tersedia di:
+- Hugging Face: [eugene702/Automatic-Scoring](https://huggingface.co/eugene702/Automatic-Scoring)
+- Kaggle Model Hub: [Automatic Scoring](https://www.kaggle.com/models/eugene702/automatic-scoring)
 
-*Clearly define terms in order to be accessible across audiences.*
--->
+## 📬 Contact
 
-<!--
-## Model Card Authors
+Untuk pertanyaan atau kolaborasi:
 
-*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
--->
+**Eugene Feilian Putra Rangga**  
+📧 [email protected]  
+🔗 [Hugging Face Profile](https://huggingface.co/eugene702)
+🔗 [GitHub](https://github.com/Eugene702)
 
-<!--
-## Model Card Contact
+---
 
-*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
--->
+> _Model ini merupakan bagian dari eksperimen untuk membangun sistem penilaian otomatis berbasis semantic similarity pada teks Bahasa Indonesia._