tencent
/

DeepSeek-V3.1-Terminus-W4AFP8

Text Generation

TensorRT-Model-Optimizer

text-generation-inference

8-bit precision

Model card Files Files and versions

DeepSeek-V3.1-Terminus-W4AFP8 / hf_quant_config.json

Xijun Chen

initial commit

9e6eee3 2 months ago

history blame contribute delete

108 Bytes

	{
	"quantization": {
	"quant_algo": "MIXED_PRECISION",
	"kv_cache_quant_algo": null
	}
	}