Instructions to use SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates with Transformers:

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

pipe = pipeline("text-generation", model="SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates")
model = AutoModelForCausalLM.from_pretrained("SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates

SGLang

How to use SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates 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 "SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates" \
    --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": "SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates",
		"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 "SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates" \
        --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": "SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates with Docker Model Runner:
```
docker model run hf.co/SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates
```

SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates

This repo only contains the AttnGates' weights for deepseek-ai/DeepSeek-R1-Distill-Llama-70B.

SeerAttention introduces learnable AttnGate modules to accelerate the computationally intensive prefill stage of long-context large language models (LLMs) via dynamic block-level sparsity. The AttnGates are trained in a parameter-efficient self-distillation framework, where they learn to mimic the 2D max-pooled attention patterns of the original frozen model, preserving its integrity while avoiding costly retraining. During inference, these gates generate block-sparse binary masks by applying threshold/TopK to their learned soft scores, enabling efficient computation through a custom block-sparse FlashAttention kernel.

Original Github Repo https://github.com/microsoft/SeerAttention.

Downloads last month: 3

Model tree for SeerAttention/SeerAttention-DeepSeek-R1-Distill-Llama-70B-AttnGates

Base model

deepseek-ai/DeepSeek-R1-Distill-Llama-70B

Adapter

(55)

this model