This is Alibaba-Apsara/DASD-4B-Thinking quantized with LLM Compressor with FP8 W8A8. The model has been created and tested by The Valdanito. The model is compatible with SGlang v0.5.7 & vLLM v0.14.0. Tested with an NVIDIA H800.

Downloads last month: 18

Safetensors

Model size

4B params

Tensor type

BF16

F8_E4M3

Model tree for valdanito/DASD-4B-Thinking-FP8-DYNAMIC

Base model

Alibaba-Apsara/DASD-4B-Thinking

Quantized

(12)

this model