mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-12-24 13:28:13 +08:00
10 lines
219 B
YAML
10 lines
219 B
YAML
quantization: wint4
|
|
load_choices: "default_v1"
|
|
graph_optimization_config:
|
|
use_cudagraph: True
|
|
use_unique_memory_pool: True
|
|
enable_prefix_caching: False
|
|
max_num_seqs: 256
|
|
max_model_len: 32768
|
|
tensor_parallel_size: 8
|