202610271728_Moss文生图_omnixpu加速_B70_64G_冷启+热启.txt · 2026-10-27 17:28
| 阶段 | 耗时 | 步数 | s/it | staged MB | 其他指标 |
|---|---|---|---|---|---|
| 冷启 | 23.88 s | 8 | 0.70 | 9,295 | — |
| 热启 | 7.36 s | 8 | 0.66 | 9,295 | — |
[INFO] got prompt [INFO] Model storage policy: fast_disk=False paths=['<ComfyUI>/models/vae/flux2-vae.safetensors'] [INFO] Using pytorch attention in VAE [INFO] Using pytorch attention in VAE [INFO] VAE load device: xpu:0, offload device: cpu, dtype: torch.bfloat16 [INFO] Found quantization metadata version 1 [INFO] Using MixedPrecisionOps for text encoder [INFO] Model storage policy: fast_disk=False paths=['<本地路径>/qwen3.5_9b_int8_convrot.safetensors'] [INFO] CLIP/text encoder model load device: xpu:0, offload device: cpu, current: cpu, dtype: torch.float16 [INFO] Requested to load MOSSImageTEModel_ [INFO] Model MOSSImageTEModel_ prepared for dynamic VRAM loading. 9295MB Staged. 0 patches attached. Force pre-loaded 222 weights: 822 KB. [INFO] Model MOSSImageTEModel_ prepared for dynamic VRAM loading. 9295MB Staged. 0 patches attached. Force pre-loaded 222 weights: 822 KB. [INFO] Found quantization metadata version 1 [INFO] Detected mixed precision quantization [INFO] Using mixed precision operations [INFO] Native ops: int8_tensorwise , emulated ops: nvfp4, float8_e4m3fn, float8_e5m2, mxfp8, w6a8_int8, asym_w4a8_int8, convrot_w4a4 [INFO] model weight dtype torch.bfloat16, manual cast: torch.bfloat16 [INFO] model_type FLOW [INFO] Model storage policy: fast_disk=False paths=['<本地路径>/moss_image_sft_int8_convrot.safetensors'] [INFO] [OmniXPU] LoRA memory budget cached: source=1134.2MiB/756 tensors model_added=1134.0MiB model_budget=1134.0MiB clip_added=0.0MiB clip_budget=0.0MiB [INFO] [OmniXPU] LoRA model budget applied: cached=1134.0MiB/504 tensors requested=2382.7->3516.7MiB minimum=1191.4->2325.4MiB xpu_memory=available xpu_allocated=1.5MiB xpu_reserved=4.0MiB xpu_free=20122.3MiB xpu_total=31977.4MiB [INFO] [OmniXPU] AIMDO inference budget: devices=0 minimum=2325.4MiB [INFO] Requested to load MOSSImage [INFO] Model MOSSImage prepared for dynamic VRAM loading. 5162MB Staged. 168 patches attached. Force pre-loaded 177 weights: 559 KB. [INFO] [OmniXPU] LoRA model load complete: xpu_memory=available xpu_allocated=2.1MiB xpu_reserved=4.0MiB xpu_free=20122.3MiB xpu_total=31977.4MiB 100%|████████████████████████████████████████████████████████████████████████████████████| 8/8 [00:05<00:00, 1.43it/s] [INFO] [OmniXPU] AIMDO inference budget: devices=0 minimum=4339.0MiB [INFO] Requested to load AutoencoderKL [INFO] Model AutoencoderKL prepared for dynamic VRAM loading. 160MB Staged. 0 patches attached. Force pre-loaded 112 weights: 101 KB. [INFO] Prompt executed in 23.88 seconds [INFO] got prompt [INFO] Model MOSSImageTEModel_ prepared for dynamic VRAM loading. 9295MB Staged. 0 patches attached. Force pre-loaded 222 weights: 822 KB. [INFO] [OmniXPU] LoRA model budget applied: cached=1134.0MiB/504 tensors requested=2382.7->3516.7MiB minimum=1191.4->2325.4MiB xpu_memory=available xpu_allocated=2.2MiB xpu_reserved=4.0MiB xpu_free=14804.5MiB xpu_total=31977.4MiB [INFO] [OmniXPU] AIMDO inference budget: devices=0 minimum=2325.4MiB [INFO] Model MOSSImage prepared for dynamic VRAM loading. 5162MB Staged. 168 patches attached. Force pre-loaded 177 weights: 559 KB. [INFO] [OmniXPU] LoRA model load complete: xpu_memory=available xpu_allocated=2.2MiB xpu_reserved=4.0MiB xpu_free=14804.5MiB xpu_total=31977.4MiB 100%|████████████████████████████████████████████████████████████████████████████████████| 8/8 [00:05<00:00, 1.52it/s] [INFO] [OmniXPU] AIMDO inference budget: devices=0 minimum=4339.0MiB [INFO] Model AutoencoderKL prepared for dynamic VRAM loading. 160MB Staged. 0 patches attached. Force pre-loaded 112 weights: 101 KB. [INFO] Prompt executed in 7.36 seconds