52 lines
1.2 KiB
YAML
52 lines
1.2 KiB
YAML
# OpenAI-compatible FastH3 startup-LoRA server. Override lora_path at launch
|
|
# with the exact file returned by `hf download`; see README.md.
|
|
generator:
|
|
model_path: MiniMaxAI/MiniMax-H3
|
|
engine:
|
|
num_gpus: 4
|
|
use_fsdp_inference: false
|
|
parallelism:
|
|
tp_size: 1
|
|
sp_size: 4
|
|
offload:
|
|
dit: false
|
|
dit_layerwise: false
|
|
text_encoder: true
|
|
image_encoder: true
|
|
vae: true
|
|
pin_cpu_memory: true
|
|
compile:
|
|
enabled: false
|
|
vae_enabled: false
|
|
pipeline:
|
|
workload_type: t2v
|
|
components:
|
|
lora_path: /path/to/adapter_model.safetensors
|
|
lora_nickname: fasth3-dense-datafree
|
|
lora_strength: 1.0
|
|
experimental:
|
|
attention_backend: FLASH_ATTN
|
|
inference_torch_compile: false
|
|
vae_parallel_decode: true
|
|
vae_parallel_decode_strategy: gather
|
|
|
|
server:
|
|
host: 0.0.0.0
|
|
port: 8000
|
|
output_dir: outputs/openai_fasth3_lora
|
|
served_model_name: fasth3-dense-datafree
|
|
|
|
default_request:
|
|
negative_prompt: ""
|
|
sampling:
|
|
height: 768
|
|
width: 1344
|
|
num_frames: 124
|
|
fps: 24
|
|
num_inference_steps: 5
|
|
guidance_scale: 1.0
|
|
batch_cfg: false
|
|
seed: 1000
|
|
output:
|
|
return_frames: false
|