Files
FastDeploy/benchmarks/yaml/x1-32k-wint4-p800-tp4.yaml
2025-06-29 23:29:37 +00:00

7 lines
137 B
YAML

max_model_len: 32768
max_num_seqs: 32
gpu_memory_utilization: 0.9
tensor_parallel_size: 4
quantization: wint4
reasoning_parser: ernie-x1