---
language: en
license: gemma
tags:
- t5
- gemma
- entailment
- classification
datasets:
- custom
metrics:
- f1
- accuracy
model-index:
- name: KamilHugsFaces/t5-gemma-reasoning-v4
  results:
  - task:
      type: text-classification
    metrics:
    - name: F1 Score
      type: f1
      value: 0.8652
    - name: F1 (False class)
      type: f1_false
      value: 0.3684
    - name: Accuracy
      type: accuracy
      value: 0.8400
---

# KamilHugsFaces/t5-gemma-reasoning-v4

Fine-tuned T5-Gemma-2 model for entailment classification.

## Training Details

- **Base model**: google/t5gemma-2-4b-4b
- **Training variant**: reasoning_v1
- **Epochs**: 3
- **Batch size**: 4
- **Learning rate**: 5e-05
- **Run name**: reasoning_v1_20260113_232433

## Training Data

- **Training examples**: 700
- **Validation examples**: 150
- **Test examples**: 150
- **Class weights**: {'true': 1.0, 'false': 8.0}

## Evaluation Results

### Test Set Performance

- **F1 Score**: 0.8652
- **F1 (False class)**: 0.3684
- **Accuracy**: 0.8400
- **Precision (False)**: 0.2692
- **Recall (False)**: 0.5833

## Usage

```python
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

model = AutoModelForSeq2SeqLM.from_pretrained("KamilHugsFaces/t5-gemma-reasoning-v4")
tokenizer = AutoTokenizer.from_pretrained("KamilHugsFaces/t5-gemma-reasoning-v4")

# Format input
input_text = "entailment: [Your claim and evidence here]"
inputs = tokenizer(input_text, return_tensors="pt", max_length=250, truncation=True)

# Generate prediction
outputs = model.generate(**inputs, max_new_tokens=8)
prediction = tokenizer.decode(outputs[0], skip_special_tokens=True)
# Output: "true" or "false"
```

## Training Configuration

{
  "variant_name": "reasoning_v1",
  "run_name": "reasoning_v1_20260113_232433",
  "num_epochs": 3,
  "batch_size": 4,
  "learning_rate": 5e-05,
  "warmup_steps": 100,
  "model_name": "google/t5gemma-2-4b-4b",
  "class_weights": {
    "true": 1.0,
    "false": 8.0
  },
  "use_confidence_weighting": false,
  "confidence_weight_alpha": 2,
  "train_size": 700,
  "val_size": 150,
  "test_size": 150
}

## Framework

- **Transformers**: 5.0.0.dev0
- **PyTorch**: 2.9.1+cu128
- **Trained on**: Modal (A100 GPU)