Instructions to use unsloth/gemma-2-27b-it-bnb-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use unsloth/gemma-2-27b-it-bnb-4bit with Transformers:

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

pipe = pipeline("text-generation", model="unsloth/gemma-2-27b-it-bnb-4bit")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("unsloth/gemma-2-27b-it-bnb-4bit")
model = AutoModelForCausalLM.from_pretrained("unsloth/gemma-2-27b-it-bnb-4bit", device_map="auto")
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 Settings

vLLM

How to use unsloth/gemma-2-27b-it-bnb-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "unsloth/gemma-2-27b-it-bnb-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "unsloth/gemma-2-27b-it-bnb-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/unsloth/gemma-2-27b-it-bnb-4bit

SGLang

How to use unsloth/gemma-2-27b-it-bnb-4bit 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 "unsloth/gemma-2-27b-it-bnb-4bit" \
    --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": "unsloth/gemma-2-27b-it-bnb-4bit",
		"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 "unsloth/gemma-2-27b-it-bnb-4bit" \
        --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": "unsloth/gemma-2-27b-it-bnb-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio

How to use unsloth/gemma-2-27b-it-bnb-4bit 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 unsloth/gemma-2-27b-it-bnb-4bit 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 unsloth/gemma-2-27b-it-bnb-4bit to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for unsloth/gemma-2-27b-it-bnb-4bit to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="unsloth/gemma-2-27b-it-bnb-4bit",
    max_seq_length=2048,
)

Docker Model Runner
How to use unsloth/gemma-2-27b-it-bnb-4bit with Docker Model Runner:
```
docker model run hf.co/unsloth/gemma-2-27b-it-bnb-4bit
```

Aphrodite/VLLM/SGLang all refuse to load this model

by fullstack - opened Sep 10, 2024

Discussion

fullstack

Sep 10, 2024

all same error
KeyError: 'model.layers.0.mlp.down_proj.weight'

(sglang) ➜  ~ python -m sglang.launch_server --model-path unsloth/gemma-2-27b-it-bnb-4bit  --port 6002
WARNING 09-10 13:26:55 cuda.py:22] You are using a deprecated `pynvml` package. Please install `nvidia-ml-py` instead. See https://pypi.org/project/pynvml for more information.
[13:26:56] When using sliding window in gemma-2, turn on flashinfer.
[13:26:56] server_args=ServerArgs(model_path='unsloth/gemma-2-27b-it-bnb-4bit', tokenizer_path='unsloth/gemma-2-27b-it-bnb-4bit', tokenizer_mode='auto', skip_tokenizer_init=False, load_format='auto', dtype='auto', kv_cache_dtype='auto', trust_remote_code=False, context_length=None, quantization=None, served_model_name='unsloth/gemma-2-27b-it-bnb-4bit', chat_template=None, is_embedding=False, host='127.0.0.1', port=6002, additional_ports=[6003, 6004, 6005, 6006], mem_fraction_static=0.88, max_running_requests=None, max_num_reqs=None, max_total_tokens=None, chunked_prefill_size=8192, max_prefill_tokens=16384, schedule_policy='lpm', schedule_conservativeness=1.0, tp_size=1, stream_interval=1, random_seed=298685233, log_level='info', log_level_http=None, log_requests=False, show_time_cost=False, api_key=None, file_storage_pth='SGLang_storage', dp_size=1, load_balance_method='round_robin', disable_flashinfer=False, disable_flashinfer_sampling=False, disable_radix_cache=False, disable_regex_jump_forward=False, disable_cuda_graph=False, disable_cuda_graph_padding=False, disable_disk_cache=False, disable_custom_all_reduce=False, enable_mixed_chunk=False, enable_torch_compile=False, enable_p2p_check=False, enable_mla=False, triton_attention_reduce_in_fp32=False, nccl_init_addr=None, nnodes=1, node_rank=None)
[13:26:56 TP0] Init nccl begin.
[13:26:56 TP0] Load weight begin. avail mem=23.40 GB
WARNING 09-10 13:26:57 interfaces.py:132] The model (<class 'sglang.srt.models.gemma2.Gemma2ForCausalLM'>) contains all LoRA-specific attributes, but does not set `supports_lora=True`.
INFO 09-10 13:26:57 weight_utils.py:236] Using model weights format ['*.safetensors']
Loading safetensors checkpoint shards:   0% Completed | 0/2 [00:00<?, ?it/s]
Process Process-1:
Traceback (most recent call last):
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/launch_server.py", line 19, in <module>
    raise e
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/launch_server.py", line 17, in <module>
    launch_server(server_args)
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/srt/server.py", line 365, in launch_server
    raise RuntimeError(
RuntimeError: Initialization failed. controller_init_state: Traceback (most recent call last):
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/srt/managers/controller_single.py", line 149, in start_controller_process
    controller = ControllerSingle(
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/srt/managers/controller_single.py", line 83, in __init__
    self.tp_server = ModelTpServer(
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/srt/managers/tp_worker.py", line 99, in __init__
    self.model_runner = ModelRunner(
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/srt/model_executor/model_runner.py", line 110, in __init__
    self.load_model()
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/srt/model_executor/model_runner.py", line 204, in load_model
    self.model = get_model(
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/vllm/model_executor/model_loader/__init__.py", line 19, in get_model
    return loader.load_model(model_config=model_config,
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/vllm/model_executor/model_loader/loader.py", line 344, in load_model
    model.load_weights(
  File "/home/shazam/miniforge3/envs/sglang/lib/python3.10/site-packages/sglang/srt/models/gemma2.py", line 401, in load_weights
    param = params_dict[name]
KeyError: 'model.layers.0.mlp.down_proj.weight'
, detoken_init_state: init ok

danielhanchen

Unsloth AI org Sep 14, 2024

Oh wait bitsandbytes models might be able supported (yet) in SGLang - I think vLLM maybe.
It's better to the 16bit version https://huggingface.co/unsloth/gemma-2-27b-it

AtAndDev

Sep 14, 2024

Um, it should work with VLLM, im using it rn.

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