Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
Nicolassuez
/
Qwen3-0.6B-eagle3
like
0
Text Generation
Safetensors
English
French
llama
eagle3
speculative-decoding
sglang
draft-model
tool-calling
License:
apache-2.0
Model card
Files
Files and versions
xet
Community
Copy to bucket
new
main
Qwen3-0.6B-eagle3
112 MB
Ctrl+K
Ctrl+K
1 contributor
History:
4 commits
Nicolassuez
v3: 42k steps at ttt7/maxlen4096 on the 85k domain mix; 955 tok/s single-stream H100 (+73pct vs no speculation), +28pct on bounded-args domain JSON vs v2
00b7c41
verified
9 days ago
.gitattributes
Safe
1.52 kB
initial commit
10 days ago
README.md
Safe
2.71 kB
v3: 42k steps at ttt7/maxlen4096 on the 85k domain mix; 955 tok/s single-stream H100 (+73pct vs no speculation), +28pct on bounded-args domain JSON vs v2
9 days ago
config.json
Safe
856 Bytes
EAGLE-3 draft head for Qwen3-0.6B: SpecForge 40k steps, 40k glaive tool-calls + 30k ultrachat, ttt5/maxlen2048
10 days ago
model.safetensors
112 MB
xet
v3: 42k steps at ttt7/maxlen4096 on the 85k domain mix; 955 tok/s single-stream H100 (+73pct vs no speculation), +28pct on bounded-args domain JSON vs v2
9 days ago