Instructions to use facebook/KernelLLM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use facebook/KernelLLM with Transformers:

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

pipe = pipeline("text-generation", model="facebook/KernelLLM")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("facebook/KernelLLM")
model = AutoModelForCausalLM.from_pretrained("facebook/KernelLLM", 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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use facebook/KernelLLM with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/facebook/KernelLLM

SGLang

How to use facebook/KernelLLM 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 "facebook/KernelLLM" \
    --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": "facebook/KernelLLM",
		"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 "facebook/KernelLLM" \
        --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": "facebook/KernelLLM",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use facebook/KernelLLM with Docker Model Runner:
```
docker model run hf.co/facebook/KernelLLM
```

Zacharias030 commited on Jun 24, 2025

Commit

52bf86f

verified ·

1 Parent(s): 13e2283

Added Updates section

Browse files

Files changed (1) hide show

README.md +7 -1

README.md CHANGED Viewed

@@ -12,6 +12,12 @@ library_name: transformers
 ![scatter performance comparison plot](media/llm_performance_comparison.png)
 *On KernelBench-Triton Level 1, our 8B parameter model exceeds models such as GPT-4o and DeepSeek V3 in single-shot performance. With multiple inferences, KernelLLM's performance outperforms DeepSeek R1. This is all from a model with two orders of magnitude fewer parameters than its competitors.*
 ## Making Kernel Development more accessible with KernelLLM
 We introduce KernelLLM, a large language model based on Llama 3.1 Instruct, which has been trained specifically for the task of authoring GPU kernels using Triton. KernelLLM translates PyTorch modules into Triton kernels and was evaluated on KernelBench-Triton (see [here](https://github.com/ScalingIntelligence/KernelBench/pull/35)).
@@ -181,7 +187,7 @@ Please see the Responsible Use Guide available at [https://ai.meta.com/llama/res
 ```
 @software{kernelllm2025,
-    title={KernelLLM},
     author={Fisches, Zacharias and Paliskara, Sahan and Guo, Simon and Zhang, Alex and Spisak, Joe and Cummins, Chris and Leather, Hugh and Isaacson, Joe and Markosyan, Aram and Saroufim, Mark},
     year={2025},
     month={5},

 ![scatter performance comparison plot](media/llm_performance_comparison.png)
 *On KernelBench-Triton Level 1, our 8B parameter model exceeds models such as GPT-4o and DeepSeek V3 in single-shot performance. With multiple inferences, KernelLLM's performance outperforms DeepSeek R1. This is all from a model with two orders of magnitude fewer parameters than its competitors.*
+## _Updates_:
+* 2025/06/25: We added an [end-to-end example walkthrough](https://huggingface.co/facebook/KernelLLM/discussions/5#685b0903b3d048882566b17b), where we format a community-provided prompt for KernelLLM to function well.
+We have received many questions about how to format the prompts such that KernelLLM performs best. We hope this can help!
+* 2025/06/15 We would like to thank the community for the creation of [multiple](https://huggingface.co/bartowski/facebook_KernelLLM-GGUF) [different](https://huggingface.co/unsloth/KernelLLM-GGUF) [quantizations](https://huggingface.co/unsloth/KernelLLM) and for a total of more than 20k downloads!
+* 2025/06/03 The startup mako.dev has integrated KernelLLM into their [GPU performance engineering plattform](https://generate.mako.dev/)!
 ## Making Kernel Development more accessible with KernelLLM
 We introduce KernelLLM, a large language model based on Llama 3.1 Instruct, which has been trained specifically for the task of authoring GPU kernels using Triton. KernelLLM translates PyTorch modules into Triton kernels and was evaluated on KernelBench-Triton (see [here](https://github.com/ScalingIntelligence/KernelBench/pull/35)).
 ```
 @software{kernelllm2025,
+    title={KernelLLM: Making Kernel Development More Accessible},
     author={Fisches, Zacharias and Paliskara, Sahan and Guo, Simon and Zhang, Alex and Spisak, Joe and Cummins, Chris and Leather, Hugh and Isaacson, Joe and Markosyan, Aram and Saroufim, Mark},
     year={2025},
     month={5},