Instella MoE 16B A3B DPO
Instella-MoE-16B-A3B-DPO is the direct preference optimization checkpoint of AMD's Instella-MoE model, a fully open Mixture-of-Experts language model with 16 billion total parameters and 2.8 billion active parameters per token. Built on the SFT checkpoint, it is tuned on contrastive preference data to improve response quality and alignment, serving as an input to the final reinforcement-learning stage.
- Type
- Text
- Weights
- Open
- Downloads
- 652
- Likes
- 6
- Pipeline
- text generation
- License
- other
Model hub
amd/Instella-MoE-16B-A3B-DPO
- Repo
- amd/Instella-MoE-16B-A3B-DPO
- Library
- transformers
- Access
- Public
Technical tags and languages
Papers
1
More from AMD
8
-
273k DL
DeepSeek-R1-MXFP4
License mit · 272,995 downloads · by amd
-
799 DL
Instella MoE 16B A3B Base
Text
Instella-MoE-16B-A3B-Base is the long-context checkpoint AMD designates as the final base model in its Instel…
-
638 DL
Instella MoE 16B A3B Midtrain
Text
Instella-MoE-16B-A3B-Midtrain is the mid-training checkpoint of AMD's Instella-MoE model, a fully open Mixtur…
-
740 DL
Instella MoE 16B A3B Pretrain
Text
Instella-MoE-16B-A3B-Pretrain is the base pre-training checkpoint of AMD's Instella-MoE model, a fully open M…
-
1.9k DL
Instella MoE 16B A3B SFT
Text
Instella-MoE-16B-A3B-SFT is the supervised fine-tuning checkpoint of AMD's Instella-MoE model, a fully open M…
-
3.2k DL
Instella MoE 16B A3B Think
Text
Instella-MoE-16B-A3B-Think is the final reinforcement-learning tuned checkpoint of AMD's Instella-MoE model,…
-
9k DL
MiniMax-M3-MXFP4
Multimodal
image text to text · License other · 89,605 downloads · by amd
-
608 DL
Nitro-E
Image
Nitro-E is an efficiency-focused reasoning and coding model optimized for AMD hardware. It offers long contex…
About AMD
1
Semiconductor company designing CPUs, GPUs, adaptive processors, and AI accelerators for PCs, gaming, data centers, embedded systems, and high-performance computing.