Downloads · 30 days
18
10% of all-time downloads
SGLang-EAGLE3-Qwen3-Next-80B-A3B-Instruct-SpecForge
huang3eng/SGLang-EAGLE3-Qwen3-Next-80B-A3B-Instruct-SpecForge
SGLang-EAGLE3-Qwen3-Next-80B-A3B-Instruct-SpecForge is a machine learning model from huang3eng. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
SpecBundle is an open-source initiative, jointly driven by the community and industry, to democratize speculative decoding by providing high-performance speculative decoding draft weights for mainstream open-source mo…
All-time downloads
180
Public
Parameters
208M
417 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors417 MB · 100%
Parameter types
How the weights are stored.
BF16208M · 100%
Datasets
Model card
From the Hugging Face model README
EAGLE3 For Qwen/Qwen3-Next-80B-A3B-Instruct
About
SpecBundle is an open-source initiative, jointly driven by the community and industry, to democratize speculative decoding by providing high-performance speculative decoding draft weights for mainstream open-source models.
Regenerated the responses in the mlabonne/open-perfectblend and trained the model on 1.4M data samples for 2 epochs. This checkpoint was trained using the SpecForge framework.
Usage
You can use this checkpoint with the command below.
python3 -m sglang.launch_server \
--model Qwen/Qwen3-Next-80B-A3B-Instruct \
--speculative-algorithm EAGLE3 \
--speculative-draft-model-path huang3eng/SGLang-EAGLE3-Qwen3-Next-80B-A3B-Instruct-SpecForge \
--speculative-num-steps 3 \
--speculative-eagle-topk 1 \
--speculative-num-draft-tokens 4 \
--tp 8
Performance
This checkpoint exhibits superior performance on various benchmarks.
- Type
- llama
- License
- mit
- Created
- Jan 13, 2026
- Updated
- Jan 13, 2026