README.md

---
base_model: google/gemma-4-E2B-it
library_name: transformers
pipeline_tag: text-classification
tags:
- gemma4
- sequence-classification
- text-classification
---

# Gemma 4 E2B IT Sequence Classification

This checkpoint converts `google/gemma-4-E2B-it` into a two-label sequence
classifier by slicing the original LM head to the single-token labels `no` and
`yes`.

The classifier logits are exactly the original next-token logits for:

- `no`: token id `1904`
- `yes`: token id `4443`

Load with custom code enabled:

```python
from transformers import AutoModelForSequenceClassification, AutoTokenizer

model_id = "baseten/gemma-4-e2b-it-sequence-classification"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForSequenceClassification.from_pretrained(
    model_id,
    trust_remote_code=True,
    dtype="auto",
)
```