vllm-sr/d3-flash download history

vllm-sr/d3-flash is an 8.4B-parameter zero shot classification model by vllm-sr. In the last 30 days it was downloaded 102,705 times (95,599 in the last 7 days), and 102,705 times in total.

It ranks #2,250 on the Hub by monthly downloads and #13 among zero shot classification models.

It has 103 likes, 12 of them in the last week.

It is a fine-tune of Qwen/Qwen3.5-9B.

23 models build on d3-flash: 13 quantized, 9 fine-tuned, 1 adapters and 0 merges. Together with the original they were downloaded 129,619 times in the last 30 days. See the d3-flash galaxy.

Most downloaded derivatives

• ggml-org/vllm-d3-flash-GGUF (quantized), 19.5K downloads in 30 days
• turintech/d3-flash-Artemis-GGUF (quantized), 2.9K downloads in 30 days
• Livesport/d3-flash-GGUF (quantized), 1.7K downloads in 30 days
• tensor-ai-inc/d3-flash-NVFP4 (quantized), 864 downloads in 30 days
• Jundot/d3-flash-oQ8e (quantized), 309 downloads in 30 days
• vllm-sr/vela3-flash (adapter), 252 downloads in 30 days
• Agentware/d3-flash-calibrated-gguf (quantized), 216 downloads in 30 days
• ahmedandaloes/d3-flash-mlx (quantized), 198 downloads in 30 days
• sergolsub/mlx-d3-flash-9B (finetune), 169 downloads in 30 days
• nomadoor/d3 (finetune), 165 downloads in 30 days
• thoddnn/d3-flash-q4 (finetune), 150 downloads in 30 days
• midium-ai/d3-flash-8bit (quantized), 104 downloads in 30 days

Open vllm-sr/d3-flash on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.