QwenPhi-4-0.5B-Draft

Qwen/Qwen2.5-0.5B-Instruct, but with the vocab of microsoft/phi-4 transplanted using transplant-vocab.

Made from the instruct qwen to be used as a draft model for Phi-4 directly.

This Model was made based on the work of alamios at alamios/Qwenstral-Small-3.1-0.5B

Downloads last month
17
Safetensors
Model size
538M params
Tensor type
BF16
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for rdsm/QwenPhi-4-0.5b-Draft

Base model

Qwen/Qwen2.5-0.5B
Finetuned
(165)
this model
Quantizations
5 models