Instructions to use CreitinGameplays/gemma-2-2b-it-R1-exp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CreitinGameplays/gemma-2-2b-it-R1-exp with Transformers:

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

pipe = pipeline("text-generation", model="CreitinGameplays/gemma-2-2b-it-R1-exp")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("CreitinGameplays/gemma-2-2b-it-R1-exp")
model = AutoModelForCausalLM.from_pretrained("CreitinGameplays/gemma-2-2b-it-R1-exp", 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 CreitinGameplays/gemma-2-2b-it-R1-exp with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CreitinGameplays/gemma-2-2b-it-R1-exp"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CreitinGameplays/gemma-2-2b-it-R1-exp",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/CreitinGameplays/gemma-2-2b-it-R1-exp

SGLang

How to use CreitinGameplays/gemma-2-2b-it-R1-exp 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 "CreitinGameplays/gemma-2-2b-it-R1-exp" \
    --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": "CreitinGameplays/gemma-2-2b-it-R1-exp",
		"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 "CreitinGameplays/gemma-2-2b-it-R1-exp" \
        --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": "CreitinGameplays/gemma-2-2b-it-R1-exp",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use CreitinGameplays/gemma-2-2b-it-R1-exp with Docker Model Runner:
```
docker model run hf.co/CreitinGameplays/gemma-2-2b-it-R1-exp
```

gemma-2-2b-it-R1-exp / README.md

CreitinGameplays

Update README.md

fe4c998 verified over 1 year ago

preview code

Raw

History Blame Contribute Delete

2.21 kB

metadata

datasets:
  - CreitinGameplays/gemma-r1-test
language:
  - en
base_model:
  - google/gemma-2-2b-it
pipeline_tag: text-generation
library_name: transformers

Chat template:

<start_of_turn>user
{user_prompt}<end_of_turn>
<start_of_turn>model
<think>

Code for testing:

# test the model
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, TextStreamer

def main():
    model_id = "CreitinGameplays/gemma-2-2b-it-R1-exp"

    # Load the tokenizer.
    tokenizer = AutoTokenizer.from_pretrained(model_id)

    # Load the model using bitsandbytes 8-bit quantization if CUDA is available.
    if torch.cuda.is_available():
        model = AutoModelForCausalLM.from_pretrained(
            model_id,
            load_in_4bit=True,
            device_map="auto"
        )
        device = torch.device("cuda")
    else:
        model = AutoModelForCausalLM.from_pretrained(model_id)
        device = torch.device("cpu")

    # Define the generation parameters.
    generation_kwargs = {
        "max_new_tokens": 4096,
        "do_sample": True,
        "temperature": 0.6,
        "top_k": 40,
        "top_p": 0.9,
        "repetition_penalty": 1.1,
        "num_return_sequences": 1,
        "pad_token_id": tokenizer.eos_token_id
    }

    print("Enter your prompt (type 'exit' to quit):")
    while True:
        # Get user input.
        user_input = input("Input> ")
        if user_input.lower().strip() in ("exit", "quit"):
            break

        # Construct the prompt in your desired format.
        prompt = f"""
<start_of_turn>user
{user_input}<end_of_turn>
<start_of_turn>model
<think>
"""

        # Tokenize the prompt and send to the selected device.
        input_ids = tokenizer.encode(prompt, return_tensors="pt", add_special_tokens=True).to(device)

        # Create a new TextStreamer instance for streaming responses.
        streamer = TextStreamer(tokenizer)
        generation_kwargs["streamer"] = streamer

        print("\nAssistant Response:")
        # Generate the text (tokens will stream to stdout via the streamer).
        outputs = model.generate(input_ids, **generation_kwargs)

if __name__ == "__main__":
    main()

#INeedSomeGPU