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

RuntimeError: FlashAttention backward for head dim > 192 requires A100/A800 or H100/H800

#18

by g-ronimo - opened Feb 21, 2024

Discussion

g-ronimo

Feb 21, 2024

Thank you for this model!

Any idea how to resolve this when finetuning (QLoRA) gemma-7B with FA2 on a 3090 ?

/home/g/.local/lib/python3.10/site-packages/torch/utils/checkpoint.py:460: UserWarning: torch.utils.checkpoint: please pass in use_reentrant=True or use_reentrant=False explicitly. The default value of use_reentrant will be updated to be False in the future. To maintain current behavior, pass use_reentrant=True. It is recommended that you use use_reentrant=False. Refer to docs for more details on the differences between the two variants.
  warnings.warn(
The input hidden states seems to be silently casted in float32, this might be related to the fact you have upcasted embedding or layer norm layers in float32. We will cast back the input in torch.bfloat16.
Traceback (most recent call last):
  File "/home/g/gemma-ft/qlora-OA.py", line 262, in <module>
    trainer.train()
  File "/home/g/.local/lib/python3.10/site-packages/transformers/trainer.py", line 1624, in train
    return inner_training_loop(
  File "/home/g/.local/lib/python3.10/site-packages/transformers/trainer.py", line 1961, in _inner_training_loop
    tr_loss_step = self.training_step(model, inputs)
  File "/home/g/.local/lib/python3.10/site-packages/transformers/trainer.py", line 2911, in training_step
    self.accelerator.backward(loss)
  File "/home/g/accelerate_fork/src/accelerate/accelerator.py", line 1966, in backward
    loss.backward(**kwargs)
  File "/home/g/.local/lib/python3.10/site-packages/torch/_tensor.py", line 522, in backward
    torch.autograd.backward(
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/__init__.py", line 266, in backward
    Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/function.py", line 289, in apply
    return user_fn(self, *args)
  File "/home/g/.local/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 319, in backward
    torch.autograd.backward(outputs_with_grad, args_with_grad)
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/__init__.py", line 266, in backward
    Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/function.py", line 289, in apply
    return user_fn(self, *args)
  File "/home/g/.local/lib/python3.10/site-packages/flash_attn/flash_attn_interface.py", line 531, in backward
    _flash_attn_backward(
  File "/home/g/.local/lib/python3.10/site-packages/flash_attn/flash_attn_interface.py", line 131, in _flash_attn_backward
    dq, dk, dv, softmax_d, = flash_attn_cuda.bwd(
RuntimeError: FlashAttention backward for head dim > 192 requires A100/A800 or H100/H800
Traceback (most recent call last):
  File "/home/g/gemma-ft/qlora-OA.py", line 262, in <module>
    trainer.train()
  File "/home/g/.local/lib/python3.10/site-packages/transformers/trainer.py", line 1624, in train
    return inner_training_loop(
  File "/home/g/.local/lib/python3.10/site-packages/transformers/trainer.py", line 1961, in _inner_training_loop
    tr_loss_step = self.training_step(model, inputs)
  File "/home/g/.local/lib/python3.10/site-packages/transformers/trainer.py", line 2911, in training_step
    self.accelerator.backward(loss)
  File "/home/g/accelerate_fork/src/accelerate/accelerator.py", line 1966, in backward
    loss.backward(**kwargs)
  File "/home/g/.local/lib/python3.10/site-packages/torch/_tensor.py", line 522, in backward
    torch.autograd.backward(
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/__init__.py", line 266, in backward
    Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/function.py", line 289, in apply
    return user_fn(self, *args)
  File "/home/g/.local/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 319, in backward
    torch.autograd.backward(outputs_with_grad, args_with_grad)
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/__init__.py", line 266, in backward
    Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
  File "/home/g/.local/lib/python3.10/site-packages/torch/autograd/function.py", line 289, in apply
    return user_fn(self, *args)
  File "/home/g/.local/lib/python3.10/site-packages/flash_attn/flash_attn_interface.py", line 531, in backward
    _flash_attn_backward(
  File "/home/g/.local/lib/python3.10/site-packages/flash_attn/flash_attn_interface.py", line 131, in _flash_attn_backward
    dq, dk, dv, softmax_d, = flash_attn_cuda.bwd(
RuntimeError: FlashAttention backward for head dim > 192 requires A100/A800 or H100/H800

tedbrun1

Feb 21, 2024

•

edited Feb 21, 2024

same error. It works on my instance for a mistral but not gemma

ArthurZ

Feb 22, 2024

It's just that the Head dm of this model is bigger 😭 so another kernel is required it seems

ArthurZ

Feb 22, 2024

https://x.com/tri_dao/status/1760458183066472556?s=20 It was fixed! Upgrade flash attention!

g-ronimo changed discussion status to closed Feb 22, 2024

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