Transformers
Safetensors
English
mla
deepseek-moe
mtp
custom-code
tinystories
from-scratch
Eval Results (legacy)
Instructions to use nowordsxiaomu/DeepSeek-Flash-Mini with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use nowordsxiaomu/DeepSeek-Flash-Mini with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("nowordsxiaomu/DeepSeek-Flash-Mini", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| from .transformer import DeepSeekFlashMini, Block, MTPModule | |
| from .mla import MLA | |
| from .moe import MoE, Gate, DenseFFN | |
| from .layers import RMSNorm, SwiGLU, build_rope_cache, apply_rope | |
| __all__ = [ | |
| "DeepSeekFlashMini", "Block", "MTPModule", | |
| "MLA", "MoE", "Gate", "DenseFFN", | |
| "RMSNorm", "SwiGLU", "build_rope_cache", "apply_rope", | |
| ] | |