Task
Other
D
sgl-project/DeepSeek-V4-Flash-FP8
Other·sgl-project· 337.7K· 12
mit
Other
291.0B params
base_model:deepseek-ai/DeepSeek-V4-Flashbase_model:quantized:deepseek-ai/DeepSeek-V4-Flashlicense:mitregion:us
FP8 re-packaging of deepseek-ai/DeepSeek-V4-Flash. Model architecture, tokenizer, chat template, and reference encoding/ are unchanged from the base repo. No fine-tuning, no retraining — weights only.
# pull & run locally
pip install mlforge-sdk && mlforge pull sgl-project/DeepSeek-V4-Flash-FP8
pip install mlforge-sdk && mlforge pull sgl-project/DeepSeek-V4-Flash-FP8
Model details
Provider
sgl-project
Parameters
291.0B
Size
274 GB
License
mit
Downloads
337.7K
Likes
12
Updated
2026-04-24
About sgl-project/DeepSeek-V4-Flash-FP8
FP8 re-packaging of deepseek-ai/DeepSeek-V4-Flash. Model architecture, tokenizer, chat template, and reference encoding/ are unchanged from the base repo. No fine-tuning, no retraining — weights only.