qwen-image-edit-2511推理速度 #808
Description
我是通过python调用LightX2VPipeline进行推理的，使用了pipe.enable_lora LoRA的权重，设置
pipe.create_generator(
attn_mode="flash_attn3",
resize_mode="adaptive",
infer_steps=4,
guidance_scale=1,
) 采样4步，我发现推理单张耗时在30s左右，与官方测试出来的1.2s 差的还挺多的，我想要确认的是是需要额外设置什么嘛
`
"""
Qwen-image-edit image-to-image generation example.
This example demonstrates how to use LightX2V with Qwen-Image-Edit model for I2I generation.
"""
import time
from lightx2v import LightX2VPipeline
Initialize pipeline for Qwen-image-edit I2I task
For Qwen-Image-Edit-2509, use model_cls="qwen-image-edit-2509"
pipe = LightX2VPipeline(
model_path="/Qwen-Image-Edit-2511/",
model_cls="qwen-image-edit-2511",
task="i2i",
)
Alternative: create generator from config JSON file
pipe.create_generator(
config_json="../configs/qwen_image/qwen_image_i2i_2511_lora.json"
)
Enable offloading to significantly reduce VRAM usage with minimal speed impact
Suitable for RTX 30/40/50 consumer GPUs
pipe.enable_offload(
cpu_offload=True,
offload_granularity="block", #["block", "phase"]
text_encoder_offload=True,
vae_offload=False,
)
Load distilled LoRA weights
pipe.enable_lora(
[
{"path": "/Qwen-Image-Edit-2511-Lightning/Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors", "strength": 1.0},
],
lora_dynamic_apply=False, # Support inference with LoRA weights, save memory but slower, default is False
)
Create generator manually with specified parameters
pipe.create_generator(
attn_mode="flash_attn3",
resize_mode="adaptive",
infer_steps=4,
guidance_scale=1,
)
Generation parameters
seed = 42
prompt = ""
negative_prompt = ""
image_path = ""
save_result_path = ""
Generate video and measure latency
start_time = time.perf_counter()
pipe.generate(
seed=seed,
image_path=image_path,
prompt=prompt,
negative_prompt=negative_prompt,
save_result_path=save_result_path,
)
elapsed = time.perf_counter() - start_time
print(f"[LightX2V] Inference time: {elapsed:.3f}s")
`