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

Libraries

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

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

pipe = pipeline("text-generation", model="google/gemma-7b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

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

llama-cpp-python

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

# !pip install llama-cpp-python

from llama_cpp import Llama

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

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Inference
Local Apps Settings

llama.cpp

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

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf google/gemma-7b
# Run inference directly in the terminal:
llama cli -hf google/gemma-7b

Install from WinGet (Windows)

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

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
# Run inference directly in the terminal:
./llama-cli -hf google/gemma-7b

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
# Run inference directly in the terminal:
./build/bin/llama-cli -hf google/gemma-7b

Use Docker

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

LM Studio
Jan

vLLM

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

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "google/gemma-7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "google/gemma-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

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

SGLang

How to use google/gemma-7b 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" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "google/gemma-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "google/gemma-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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

Unsloth Studio

How to use google/gemma-7b 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 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 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 to start chatting

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

Lemonade

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

Pull the model

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

Run and chat with the model

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

List all available models

lemonade list

Issues running example_fsdp.py

#85

by TheFishInTheAir - opened Apr 5, 2024

Discussion

TheFishInTheAir

Apr 5, 2024

I've been trying to get the provided FSDP example working on a TPU v5e-8 vm. Im running into what I believe are some version conflicts and I was wondering if it was possible for someone to detail the environment that the test was originally run in (specifically what versions of what libraries where used).

For FSDP it seems like Transformers uses features in torch_xla that aren't available in the current stable release (2.2.0 at the time of writing).

AttributeError: module 'torch_xla.distributed.spmd' has no attribute 'set_global_mesh'

set_global_mesh is available in version 2.3.0 and on of torch_xla but I haven't had any luck getting those versions or the nightly builds working either.
With newer unstable versions the following check is failing:

    XLA_CHECK(dim1 == dim2 || dim1 == 1 || dim2 == 1 ||
              dim1 == xla::Shape::kUnboundedSize ||
              dim2 == xla::Shape::kUnboundedSize);

Of course the errors themselves are not related to gemma but pytorch_xla and Transformer. I have provided them for clarity on what I am experiencing.
If someone could explain how to set up a working environment for the example script it would be much appreciated! Thankyou!

Current versions installed for reference:

torch                    2.2.2
torch-xla                2.2.0
libtpu-nightly           0.1.dev20231130+default
transformers             4.39.3
trl                      0.8.1
accelerate               0.29.1

alanwaketan

Google org Apr 5, 2024

For now you will need to install torch and torch-xla nightly. However, all the features will be available in the upcoming 2.3 release.

TheFishInTheAir

Apr 8, 2024

I'm unfortunately still running into issues when using the nightly builds. It seems like there are compatibility issues with new versions of torch-xla and transformers https://github.com/huggingface/transformers/issues/30091. Are there versions which are known to work?

alanwaketan

Google org Apr 8, 2024

@TheFishInTheAir The HF people is fixing the issue here: https://github.com/huggingface/accelerate/issues/2629

TheFishInTheAir

Apr 8, 2024

Thank you!

TheFishInTheAir changed discussion status to closed Apr 8, 2024

brianchunkang

Google org Apr 11, 2024

My colleague and I built an example using Llama2 7B that should work as well: https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/training/tpuv5e_llama2_pytorch_finetuning_and_serving.ipynb

Here are the versions we used to get around the accelerate issue:
RUN pip install --upgrade pip
RUN pip install transformers==4.38.2 -U
RUN pip install datasets==2.18.0
RUN pip install trl==0.8.1 peft==0.10.0
RUN pip install accelerate==0.28.0
RUN pip install --upgrade google-cloud-storage

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