Instructions to use EvanOLeary/laguna-xs2-dense-k8-cuda-dpo with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use EvanOLeary/laguna-xs2-dense-k8-cuda-dpo with Transformers:

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

pipe = pipeline("text-generation", model="EvanOLeary/laguna-xs2-dense-k8-cuda-dpo", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForMultimodalLM
model = AutoModelForMultimodalLM.from_pretrained("EvanOLeary/laguna-xs2-dense-k8-cuda-dpo", trust_remote_code=True, dtype="auto")

Kernels

How to use EvanOLeary/laguna-xs2-dense-k8-cuda-dpo with Kernels:

# !pip install kernels

from kernels import get_kernel

kernel = get_kernel("EvanOLeary/laguna-xs2-dense-k8-cuda-dpo")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use EvanOLeary/laguna-xs2-dense-k8-cuda-dpo with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "EvanOLeary/laguna-xs2-dense-k8-cuda-dpo"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "EvanOLeary/laguna-xs2-dense-k8-cuda-dpo",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/EvanOLeary/laguna-xs2-dense-k8-cuda-dpo

SGLang

How to use EvanOLeary/laguna-xs2-dense-k8-cuda-dpo 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 "EvanOLeary/laguna-xs2-dense-k8-cuda-dpo" \
    --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": "EvanOLeary/laguna-xs2-dense-k8-cuda-dpo",
		"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 "EvanOLeary/laguna-xs2-dense-k8-cuda-dpo" \
        --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": "EvanOLeary/laguna-xs2-dense-k8-cuda-dpo",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use EvanOLeary/laguna-xs2-dense-k8-cuda-dpo with Docker Model Runner:
```
docker model run hf.co/EvanOLeary/laguna-xs2-dense-k8-cuda-dpo
```

Laguna-Dense CUDA · RFT arm = DPO

From EvanOLeary/laguna-xs2-dense-k8-cuda-sft-v2 (SFT-extended). DPO on SakanaAI CUDA Engineer Archive traces: per task, prefer the verified correct+fastest kernel over an incorrect/slow one (Correct + CUDA_Speedup_Native). β=0.1, ref = frozen SFT-extended. Exploits the archive's evolutionary refinement ordering. Compared vs offline-GRPO on KernelBench-Lite L1 (K=4).

https://github.com/Tyronita/laguna-dense-cuda-kernels

Downloads last month: 43

Safetensors

Model size

3B params

Tensor type

BF16

Model tree for EvanOLeary/laguna-xs2-dense-k8-cuda-dpo

Base model

poolside/Laguna-XS.2

Finetuned

EvanOLeary/laguna-xs2-dense-k8-cuda-sft

Finetuned

EvanOLeary/laguna-xs2-dense-k8-cuda-sft-v2

Finetuned

(3)

this model