Upload sentiment classifier model

config.yaml

@@ -0,0 +1,179 @@
+model:
+  name: sentiment_classifier
+  type: classification
+  model:
+    pretrained_model: xlm-roberta-base
+    num_labels: 3
+    dropout: 0.1
+    hidden_size: 768
+  labels:
+  - negative
+  - neutral
+  - positive
+  class_weights: null
+  tokenizer:
+    max_length: 256
+    padding: max_length
+    truncation: true
+    add_special_tokens: true
+  huggingface_hub:
+    enabled: true
+    repo_id: anpmts/sentiment-classifier
+    private: false
+    create_model_card: true
+    commit_message: Upload sentiment classifier model
+    model_card:
+      language: multilingual
+      license: apache-2.0
+      tags:
+      - sentiment-analysis
+      - text-classification
+      - xlm-roberta
+      - sequence-classification
+      datasets: null
+training:
+  epochs: 10
+  batch_size: 128
+  gradient_accumulation_steps: 1
+  max_grad_norm: 1.0
+  distributed:
+    enabled: true
+    backend: nccl
+    find_unused_parameters: true
+  precision:
+    mode: bf16
+  performance:
+    torch_compile: false
+    compile_mode: reduce-overhead
+    cudnn_benchmark: true
+    gradient_checkpointing: false
+    tf32: true
+    flash_attention_2: false
+    matmul_precision: high
+    channels_last: false
+  optimizer:
+    type: adamw
+    lr: 2.0e-05
+    weight_decay: 0.01
+    eps: 1.0e-08
+    betas:
+    - 0.9
+    - 0.999
+    fused: false
+  scheduler:
+    type: cosine
+    warmup_ratio: 0.1
+    warmup_steps: null
+    num_cycles: 0.5
+  early_stopping:
+    enabled: true
+    patience: 3
+    min_delta: 0.001
+    monitor: val_loss
+    mode: min
+  checkpoint:
+    save_top_k: 2
+    monitor: val_loss
+    mode: min
+    save_last: true
+    every_n_epochs: 1
+  resume_from_checkpoint: true
+  pretrained_checkpoint: null
+  load_only_model: true
+  eval:
+    eval_every_n_steps: null
+    eval_accumulation_steps: 1
+  dataloader:
+    num_workers: 0
+    pin_memory: true
+    persistent_workers: false
+    prefetch_factor: null
+  deterministic: false
+  benchmark: true
+data:
+  data_source: local
+  chunked:
+    enabled: false
+    train_path: data/amazon_reviews/train
+    val_path: data/amazon_reviews/validation
+    test_path: data/amazon_reviews/test
+    chunk_size: 100000
+    total_train_samples: 3600000
+    text_field: text
+    label_field: sentiment_label
+  huggingface:
+    repo: anpmts/trustshop
+    split_mapping:
+      train: train
+      val: validation
+      test: test
+    field_mapping:
+      text: text
+      sentiment_label: sentiment_label
+      sentiment_score: sentiment_score
+      quality_label: quality
+    config_name: null
+    revision: null
+    max_samples: null
+  local:
+    data_dir: data/amazon_reviews
+    processed_dir: data/processed/amazon_reviews
+  split:
+    train: 0.7
+    val: 0.15
+    test: 0.15
+    stratify: true
+  filter_quality:
+    enabled: false
+    keep_labels:
+    - valid
+  class_balancing:
+    enabled: false
+    strategy: oversample
+    oversample:
+      sampling_strategy: auto
+    smote:
+      k_neighbors: 5
+      sampling_strategy: auto
+  augmentation:
+    enabled: false
+    techniques:
+    - synonym_replacement
+    - random_deletion
+    - random_swap
+    augment_ratio: 0.1
+  preprocessing:
+    lowercase: false
+    remove_urls: true
+    remove_email: true
+    remove_special_chars: false
+    min_text_length: 10
+  cache:
+    enabled: true
+    cache_dir: data/.cache/amazon_reviews
+  seed: 42
+  validation:
+    check_missing_fields: false
+    check_empty_text: true
+    log_invalid_samples: true
+project:
+  name: ts-train
+  seed: 42
+  device: cuda
+  mixed_precision: true
+paths:
+  data_dir: data
+  data_file: data/output.jsonl
+  output_dir: outputs
+  model_dir: models
+  log_dir: logs
+logging:
+  use_wandb: true
+  wandb_project: ts-absa-classification
+  wandb_entity: null
+  use_tensorboard: true
+  log_interval: 10
+experiment:
+  name: null
+  tags: []
+  notes: ''

final_model/README.md

@@ -0,0 +1,114 @@
+---
+language: multilingual
+license: apache-2.0
+tags:
+- sentiment-analysis
+- text-classification
+- xlm-roberta
+- dual-head
+---
+
+# Sentiment Classifier
+
+## Model Description
+
+This is a dual-head sentiment classifier built on top of XLM-RoBERTa. The model performs two tasks simultaneously:
+
+1. **Sentiment Classification:** Predicts sentiment labels (positive, neutral, negative)
+2. **Sentiment Score Regression:** Predicts a continuous sentiment score in the range [0, 1]
+
+The model uses a weighted loss function combining cross-entropy (70%) for classification and MSE (30%) for regression,
+allowing it to capture both discrete sentiment categories and fine-grained sentiment intensity.
+
+## Model Architecture
+
+- **Base Model:** xlm-roberta-base
+- **Task:** text-classification
+- **Number of Labels:** 3
+- **Labels:** negative, neutral, positive
+
+## Training Configuration
+
+- **Epochs:** 10
+- **Batch Size:** 128
+- **Learning Rate:** 2e-05
+- **Warmup Ratio:** 0.1
+- **Weight Decay:** 0.01
+- **Max Seq Length:** 256
+- **Mixed Precision:** FP16=False, BF16=True
+
+## Performance Metrics
+
+- **Loss:** 0.6947
+- **Accuracy:** 0.4901
+- **Precision:** 0.2402
+- **Recall:** 0.4901
+- **F1:** 0.3224
+- **F1 Macro:** 0.3289
+- **F1 Negative:** 0.0000
+- **Precision Negative:** 0.0000
+- **Recall Negative:** 0.0000
+- **Support Negative:** 900
+- **F1 Neutral:** 0.6578
+- **Precision Neutral:** 0.4901
+- **Recall Neutral:** 1.0000
+- **Support Neutral:** 865
+- **Runtime:** 0.7012
+- **Samples Per Second:** 2517.1350
+- **Steps Per Second:** 9.9830
+
+## Model Outputs
+
+The model returns two outputs:
+
+- **Logits:** Classification logits for sentiment labels [batch_size, 3]
+- **Score Predictions:** Continuous sentiment scores [batch_size]
+
+Both outputs are computed from the same shared representation (CLS token) of the input text.
+
+## Intended Use
+
+This model is intended for sentiment analysis tasks on multilingual text, particularly in scenarios where both
+categorical sentiment (positive/neutral/negative) and sentiment intensity are important.
+
+**Typical use cases:**
+- Product review analysis
+- Social media sentiment monitoring
+- Customer feedback classification
+
+## Usage
+
+```python
+from transformers import AutoTokenizer
+from src.models.sentiment_classifier import SentimentClassifier
+
+# Load model and tokenizer
+model = SentimentClassifier.from_pretrained("YOUR_USERNAME/sentiment-classifier")
+tokenizer = AutoTokenizer.from_pretrained("xlm-roberta-base")
+
+# Prepare input
+text = "Your input text here"
+inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True)
+
+# Make prediction
+outputs = model(**inputs)
+predictions = outputs["logits"].argmax(dim=-1)
+```
+
+## Citation
+
+If you use this model, please cite:
+
+```bibtex
+@misc{sentiment_classifier,
+  title={Sentiment Classifier},
+  author={{Your Name}},
+  year={2025},
+  publisher={Hugging Face},
+  howpublished={{\url{{https://huggingface.co/YOUR_USERNAME/{model_name.lower().replace(' ', '-')}}}}}
+}
+```
+
+---
+
+*This model card was automatically generated with [Claude Code](https://claude.com/claude-code)*