meituan
/

DeepSeek-R1-Block-INT8

Text Generation

8-bit precision

Model card Files Files and versions Community

Resources

View closed (13)

Weight output_partition_size = 576 is not divisible by weight quantization block_n = 128

#18 opened 1 day ago by

Optimal `weight_block_size` for Intel AMX `amx_int8` `amx_tile`?

#17 opened 3 days ago by

what about `ollama`?

#16 opened 3 days ago by

是否有明确的sglang镜像版本推荐：）

#14 opened 6 days ago by

After deploying with the latest sglang, I found that the responses when calling the interface were chaotic.

#13 opened 9 days ago by