SparkHybrid-50M-A100

50M Parameter Hybrid Mamba-2 + GQA Transformer trained on English Cosmopedia, Coding, & Educational Datasets. Trained on a single NVIDIA A100 80GB (Brev).

Downloads last month
55
Safetensors
Model size
50.7M params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for vedantjadhav701/SparkAI-50M

Quantizations
1 model