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

Bug in logits for BOS token.

#90

by Izarel - opened Apr 28, 2024

Discussion

Izarel

Apr 28, 2024

For the 7b - it model, both this variant and the new one there is bug in the logits of the first BOS token after inference.
It you look at the output it gives all the weight to token id 2 which is the BOS token itself, in result you'll get zero probabilities for other tokens.

I validated the logits of the next tokens, and they seem to be fine.
You can easily verify it by looking at logits values/ soft max values of the tokenized empty string(the tokenizer will add BOS by default).

It might be a training issue of the base model(wrong token input for prediction).

gusthema

Google org May 3, 2024

Hi Izarel, can you share some code to help debug this? that would really help

canyon289

May 3, 2024

A reproducible colab would be perfect if possible, even with a couple of examples!

Izarel

May 7, 2024

•

edited May 7, 2024

Here is a link to colab - unfortunately it crashes when I try to load the model there - but it should work on NVIDIA GPU/MPS:
https://colab.research.google.com/drive/1ADRUjfO8k8H8OfDeprRrSr1g1hDH6f33?usp=sharing
Here is a screenshot:

canyon289

May 8, 2024

Thanks, can you give me a couple days to look into this? Will try and figure out what's going on

canyon289

May 9, 2024

Hey Izarel,
You took the time out to put together a colab so I want to be sure you get a proper explanation from me.

I'm currently unsure if this bug is in the weights, in the HF framework, somewhere else in the env like pytorch, or even deeper (though unlikely). I'm trying to reproduce this on flax to see if it persists. If its not there then we know its likely somewhere in the HF Framework or weights on Hf Hub.

I briefly tried reproducing on colab but was getting OOM memories like you were getting, If I can't try this there, It'll take me over a week to get back to you as I' m away from my home and most relevant my GPU desktop.

As someone that's just trying to run a model I apologize for the delay. I'll get back to you with an answer though.

Cheers

Izarel

May 10, 2024

No problem - I appreciate the help!

I've been able to repro it both on MPS and on Cuda- though I loaded the model only via hugging face.

Izarel

May 27, 2024

Just getting back on this - are there any updates on the matter?
No rush - just wanted to make sure I'm not missing a potential fix.

Thanks again, really appreciate the effort!

canyon289

May 27, 2024

Hey! Sorry for the delay on my end. I wasn't able to isolate whether this was a weights issue, framework issue, or quantization issue with what is currently available. However with the Gemma 2 release there is going to be updates for all three as well as some extra checks.

https://blog.google/technology/developers/gemini-gemma-developer-updates-may-2024/

Will a June timeline be alright?

Izarel

May 30, 2024

Thanks for the investigation!

Unfortunately, I'm purely using the model for research analysis - I had to compute a metric which depends on the outputs of the model, including the logprobs of the first token.
If there is anything you suggest me to check on my side, I'd be more than willing to do the investigation as well.
I'm specifically interested in this variant since it has a lot of existing evaluations on downstream tasks which I had been using.

However - I do understand that it's not a top priority since a new model version is coming.
I did manage to get results on the 2B-it variant, so I have some measurements for the paper anyway (;

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