Instructions to use google/gemma-7b-it with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use google/gemma-7b-it with Transformers:

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

pipe = pipeline("text-generation", model="google/gemma-7b-it")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("google/gemma-7b-it")
model = AutoModelForCausalLM.from_pretrained("google/gemma-7b-it")
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]:]))

llama-cpp-python

How to use google/gemma-7b-it with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="google/gemma-7b-it",
	filename="gemma-7b-it.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Inference
Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use google/gemma-7b-it with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf google/gemma-7b-it
# Run inference directly in the terminal:
llama-cli -hf google/gemma-7b-it

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf google/gemma-7b-it
# Run inference directly in the terminal:
llama-cli -hf google/gemma-7b-it

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf google/gemma-7b-it
# Run inference directly in the terminal:
./llama-cli -hf google/gemma-7b-it

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf google/gemma-7b-it
# Run inference directly in the terminal:
./build/bin/llama-cli -hf google/gemma-7b-it

Use Docker

docker model run hf.co/google/gemma-7b-it

LM Studio
Jan

vLLM

How to use google/gemma-7b-it with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "google/gemma-7b-it"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "google/gemma-7b-it",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/google/gemma-7b-it

SGLang

How to use google/gemma-7b-it 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 "google/gemma-7b-it" \
    --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": "google/gemma-7b-it",
		"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 "google/gemma-7b-it" \
        --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": "google/gemma-7b-it",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use google/gemma-7b-it with Ollama:
```
ollama run hf.co/google/gemma-7b-it
```

Unsloth Studio new

How to use google/gemma-7b-it with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for google/gemma-7b-it to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for google/gemma-7b-it to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for google/gemma-7b-it to start chatting

Docker Model Runner
How to use google/gemma-7b-it with Docker Model Runner:
```
docker model run hf.co/google/gemma-7b-it
```

Lemonade

How to use google/gemma-7b-it with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull google/gemma-7b-it

Run and chat with the model

lemonade run user.gemma-7b-it-{{QUANT_TAG}}

List all available models

lemonade list

gemma-2b-it model works but gemma-7b-it model generates errors

#51

by saurabhkumar - opened Feb 23, 2024

Discussion

saurabhkumar

Feb 23, 2024

I tried using the same code with google/gemma-2b-it and google/gemma-7b-it. The 2b-it model generates the text but 7b-it model generates error. I am using a node with 8xA100 GPUs (it is not needed for this, but that is what I had when trying this out). Changing device to CPU also does not make any difference.

cache_dir = '/path/to/hf_model_cache'
gemma = AutoModelForCausalLM.from_pretrained("google/gemma-7b-it", cache_dir=cache_dir, device_map="cuda",torch_dtype=torch.bfloat16)
gemma_tokenizer = AutoTokenizer.from_pretrained("google/gemma-7b-it", cache_dir=cache_dir)

user_request = "Write me the simplest code snippet in python you can think of."

chat = [
    { "role": "user", "content": user_request },
]
prompt = gemma_tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)

inputs = gemma_tokenizer.encode(prompt, add_special_tokens=True, return_tensors="pt")
outputs = gemma.generate(input_ids=inputs.to(gemma.device), max_new_tokens=200)

print(gemma_tokenizer.decode(outputs[0]))

Error:

  File "/myfile.py", line <line with generate>, in <module>
    outputs = gemma.generate(input_ids=inputs.to(gemma.device), max_new_tokens=200)
.....
.....

  File "/my_venv/python3.8/site-packages/transformers/models/gemma/modeling_gemma.py", line 280, in forward
    attn_output = attn_output.reshape(bsz, q_len, self.hidden_size)
RuntimeError: shape '[1, 23, 3072]' is invalid for input of size 94208

Replacing just the model and tokenizer for 7b-it with 2b-it works fine

gemma = AutoModelForCausalLM.from_pretrained("google/gemma-2b-it", cache_dir=cache_dir, device_map="cuda",torch_dtype=torch.bfloat16)
gemma_tokenizer = AutoTokenizer.from_pretrained("google/gemma-2b-it", cache_dir=cache_dir)

Output (gets formatted here because of the presence of "```" in the generated output:

<bos><bos><start_of_turn>user
Write me the simplest code snippet in python you can think of.<end_of_turn>
<start_of_turn>model
```python
print("Hello, world!")

This code will print the string "Hello, world!" to the console.

Explanation:

print() is a built-in Python function that prints a message to the console.
"Hello, world!" is the string we want to print.

Output:

Hello, world!
```<eos>

saurabhkumar

Feb 23, 2024

I updated to transformers-4.38.1 now and this solved this issue. I hope this was the right solution.
The model generated the following text (formatted here as before due to presence of "```" in the generated text):

<bos><bos><start_of_turn>user
Write me the simplest code snippet in python you can think of.<end_of_turn>
<start_of_turn>model
```python
print "Hello, world!"

# This line prints the string "Hello, world!" to the console
```<eos>

suryabhupa

Google org Feb 24, 2024

Nice! Should this be included in the documentation somewhere?

BalakrishnaCh

Google org Aug 12, 2025

Hi @saurabhkumar ,

Apologies for the late reply, thanks for bringing out this to our notice. Could you please confirm whether the above mentioned issue is resolved or not. If you required any further assistance please feel free let me know, I'm more than happy to help you out!.

Thanks.

oxicaray

Aug 14, 2025

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment