Draft models
Collection
Tiny "draft" models for speculative decoding.
•
10 items
•
Updated
•
1
NOTE: See here for update on the version with ~3B tokens of fine-tuning applied.
A 0.5B parameter draft model for speculative sampling for use with deepseek-ai/DeepSeek-R1 created from alamios/DeepSeek-R1-DRAFT-Qwen2.5-0.5B using transplant-vocab.
NOTE: This is a draft model for the full-sized DeepSeek-R1 model and not the smaller "distilled" models!
See jukofyork/DeepSeek-R1-DRAFT-0.5B for the non-GGUF version.
- Downloads last month
- 316
Hardware compatibility
Log In
to view the estimation
16-bit
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
HF Inference deployability: The model has no library tag.