# Hugging Face Gemma Recipes

🤗💎 Welcome! This repository contains *minimal* recipes to get started quickly with the Gemma family of models.
> [!Note]
> Gemma 4 Multimodal inference (vision, video, audio, function calling, object detection):
## Getting Started
To quickly run a Gemma 💎 model on your machine, install the latest version of `timm` (for the vision encoder) and 🤗 `transformers` to run inference, or if you want to fine tune it.
```shell
$ pip install -U -q transformers timm
```
### Inference with pipeline
The easiest way to start using Gemma 3n is by using the pipeline abstraction in transformers:
```python
import torch
from transformers import pipeline
pipe = pipeline(
"image-text-to-text",
model="google/gemma-3n-E4B-it", # "google/gemma-3n-E4B-it"
device="cuda",
torch_dtype=torch.bfloat16
)
messages = [
{
"role": "user",
"content": [
{"type": "image", "url": "https://huggingface.co/datasets/ariG23498/demo-data/resolve/main/airplane.jpg"},
{"type": "text", "text": "Describe this image"}
]
}
]
output = pipe(text=messages, max_new_tokens=32)
print(output[0]["generated_text"][-1]["content"])
```
### Detailed inference with transformers
Initialize the model and the processor from the Hub, and write the `model_generation` function that takes care of processing the prompts and running the inference on the model.
```python
from transformers import AutoProcessor, AutoModelForImageTextToText
import torch
model_id = "google/gemma-3n-e4b-it" # google/gemma-3n-e2b-it
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(model_id).to(device)
def model_generation(model, messages):
inputs = processor.apply_chat_template(
messages,
add_generation_prompt=True,
tokenize=True,
return_dict=True,
return_tensors="pt",
)
input_len = inputs["input_ids"].shape[-1]
inputs = inputs.to(model.device, dtype=model.dtype)
with torch.inference_mode():
generation = model.generate(**inputs, max_new_tokens=32, disable_compile=False)
generation = generation[:, input_len:]
decoded = processor.batch_decode(generation, skip_special_tokens=True)
print(decoded[0])
```
And then using calling it with our specific modality:
#### Text only
```python
# Text Only
messages = [
{
"role": "user",
"content": [
{"type": "text", "text": "What is the capital of France?"}
]
}
]
model_generation(model, messages)
```
#### Interleaved with Audio
```python
# Interleaved with Audio
messages = [
{
"role": "user",
"content": [
{"type": "text", "text": "Transcribe the following speech segment in English:"},
{"type": "audio", "audio": "https://huggingface.co/datasets/ariG23498/demo-data/resolve/main/speech.wav"},
]
}
]
model_generation(model, messages)
```
#### Interleaved with Image/Video
```python
# Interleaved with Image
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "https://huggingface.co/datasets/ariG23498/demo-data/resolve/main/airplane.jpg"},
{"type": "text", "text": "Describe this image."}
]
}
]
model_generation(model, messages)
```
## Inference
### Gemma 4
#### Notebooks
* [Multimodal inference with Gemma 4 (vision, video, audio, function calling, object detection)](/notebooks/Gemma4_(E2B)-Multimodal.ipynb)
### Gemma 3n
#### Notebooks
* [Multimodal inference using Gemma 3n via pipeline](/notebooks/gemma3n_inference_via_pipeline.ipynb)
## Function Calling
### Gemma 3n
#### Notebooks
* [Function Calling with Gemma 3n: Local File Reader](/notebooks/Gemma_3n_Function_Calling_document_summarizer.ipynb)
## Fine Tuning
We include a series of notebook+scripts for fine tuning the models.
### Gemma 3n
#### Notebooks
* [Gemma 3n Conversational Fine tuning 2B on free Colab T4](/notebooks/fine_tune_gemma3n_on_t4.ipynb)
* [Gemma 3n Conversational Fine tuning 4B with Unsloth on free Colab T4](/notebooks/Gemma3N_(4B)-Conversational.ipynb)
* [Gemma 3n Multimodal Fine tuning 2B/4B with Unsloth on free Colab T4](/notebooks/gemma3n_multimodal_finetuning_on_rocov2_radiology.ipynb)
* [Fine tuning Gemma 3n on audio](/notebooks/fine_tune_gemma3n_on_audio.ipynb)
* [Fine tuning Gemma 3n on GUI Grounding](/notebooks/Gemma_3n_GUI_Finetune.ipynb)
* [Fine tuning Gemma3n on video+audio using FineVideo (all modalities)](/notebooks/Gemma3n_Fine_tuning_on_All_Modalities.ipynb)
#### Scripts
* [Fine tuning Gemma 3n on images using TRL](/scripts/ft_gemma3n_image_trl.py)
* [Fine tuning Gemma 3n on images (script)](/scripts/ft_gemma3n_image_vt.py)
* [Fine tuning Gemma 3n on audio (script)](/scripts/ft_gemma3n_audio_vt.py)
* [Fine tuning Gemma3n on video+audio using FineVideo (all modalities)](/scripts/gemma3n_fine_tuning_on_all_modalities.py)
### Gemma 3
* [Reinforement Learning (GRPO) on Gemma 3 with Unsloth and TRL](/notebooks/Gemma3_(1B)-GRPO.ipynb)
* [Vision fine tuning Gemma 3 4B with Unsloth](/notebooks/Gemma3_(4B)-Vision.ipynb)
* [Conversational fine tuning Gemma 3 4B with Unsloth](/notebooks/Gemma3_(4B).ipynb)
## RAG
### Gemma 3n
* [Retrieval-Augmented Generation with Gemma 3n](/notebooks/Gemma_RAG.ipynb)
Before fine-tuning the model, ensure all dependencies are installed:
```bash
$ pip install -U -q -r requirements.txt
```
✨ **Bonus:** We've also experimented with adding **object detection** 🔍 capabilities to Gemma 3. You can explore that work in [this dedicated repo](https://github.com/ariG23498/gemma3-object-detection).