Instructions to use bRadu/gemma-4-E2B-it-textonly with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use bRadu/gemma-4-E2B-it-textonly with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="bRadu/gemma-4-E2B-it-textonly")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModel

tokenizer = AutoTokenizer.from_pretrained("bRadu/gemma-4-E2B-it-textonly")
model = AutoModel.from_pretrained("bRadu/gemma-4-E2B-it-textonly")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use bRadu/gemma-4-E2B-it-textonly with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "bRadu/gemma-4-E2B-it-textonly"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bRadu/gemma-4-E2B-it-textonly",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/bRadu/gemma-4-E2B-it-textonly

SGLang

How to use bRadu/gemma-4-E2B-it-textonly with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "bRadu/gemma-4-E2B-it-textonly" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bRadu/gemma-4-E2B-it-textonly",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "bRadu/gemma-4-E2B-it-textonly" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bRadu/gemma-4-E2B-it-textonly",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use bRadu/gemma-4-E2B-it-textonly with Docker Model Runner:
```
docker model run hf.co/bRadu/gemma-4-E2B-it-textonly
```

Gemma 4 E2B-it (Text-Only)

Text-only version of google/gemma-4-E2B-it with the vision and audio encoders removed.

Why?

The original Gemma 4 E2B is a multimodal model (text + vision + audio). When fine-tuning for text-only tasks, the multimodal architecture causes:

Higher training loss due to the multimodal overhead
Requires mm_token_type_ids tensors even for text-only inputs
Batch size > 1 crashes during training
Extra ~450M parameters (vision + audio encoders) that serve no purpose for text tasks

This model extracts just the language model (Gemma4ForCausalLM) and the text tokenizer, making it suitable for standard text-only SFT/DPO/KTO fine-tuning.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("bRadu/gemma-4-E2B-it-textonly", torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("bRadu/gemma-4-E2B-it-textonly")

messages = [
    {"role": "user", "content": "Hello!"}
]
inputs = tokenizer.apply_chat_template(messages, return_tensors="pt", add_generation_prompt=True, tokenize=True, return_dict=True).to(model.device)
outputs = model.generate(**inputs, max_new_tokens=128)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Downloads last month: 166

Safetensors

Model size

5B params

Tensor type

BF16

Model tree for bRadu/gemma-4-E2B-it-textonly

Base model

google/gemma-4-E2B

Finetuned

google/gemma-4-E2B-it

Finetuned

(178)

this model

Finetunes

1 model