[Bug]Web UI is not showing for Rocm 7.2.1

[ajavamind]

Git commit

version: stable-diffusion.cpp version master-647-72e512a-3-g1ceb5bd, commit 1ceb5bd

Operating System & Version

Ubuntu 24.04

GGML backends

HIP

Command-line arguments used

#!/bin/bash export HIP_VISIBLE_DEVICES="1" # sd server # prompt: create image of two tabby kittens jumping up to touch a falling red puffy cotton ball. # prompt: create photo realistic image of two tabby kittens jumping up to catch a falling red puffy cotton ball. #export GGML_CUDA_DISABLE_GRAPHS=1 /home/andy/AI/build-stable-diffusion-cpp/stable-diffusion.cpp/build/bin/sd-server --diffusion-model ./modelsEdit/qwen-image-edit-2511-Q8_0.gguf --vae ./modelsEdit/qwen_image_vae.safetensors --llm ./modelsEdit/Qwen2.5-VL-7B-Instruct.Q8_0.gguf --llm_vision ./modelsEdit/Qwen2.5-VL-7B-Instruct.mmproj-Q8_0.gguf --qwen-image-zero-cond-t --diffusion-fa --vae-conv-direct -t 8 -v --cfg-scale 1.5 --steps 8 --listen-ip 192.168.1.2 --listen-port 8001

Steps to reproduce

custom build stable-diffusion.cpp for rocm 7.2.1 on latest main branch

What you expected to happen

built-in web-gui appears

What actually happened

web page shows on only Stable Diffusion Server is running.

Logs / error messages / stack trace

No response

Additional context / environment details

No response

[ajavamind]

andy@UM780:~/AI/stable-diffusion.cpp$ ./run2server.sh
[DEBUG] main.cpp:82 - version: stable-diffusion.cpp version master-647-72e512a-3-g1ceb5bd, commit 1ceb5bd
[DEBUG] main.cpp:83 - System Info:
SSE3 = 1 | AVX = 1 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | VSX = 0 |
[DEBUG] main.cpp:84 - SDSvrParams {
listen_ip: 192.168.1.2,
listen_port: "8001",
serve_html_path: "",
}
[DEBUG] main.cpp:85 - SDContextParams {
n_threads: 8,
model_path: "",
clip_l_path: "",
clip_g_path: "",
clip_vision_path: "",
t5xxl_path: "",
llm_path: "./modelsEdit/Qwen2.5-VL-7B-Instruct.Q8_0.gguf",
llm_vision_path: "./modelsEdit/Qwen2.5-VL-7B-Instruct.mmproj-Q8_0.gguf",
diffusion_model_path: "./modelsEdit/qwen-image-edit-2511-Q8_0.gguf",
high_noise_diffusion_model_path: "",
embeddings_connectors_path: "",
vae_path: "./modelsEdit/qwen_image_vae.safetensors",
audio_vae_path: "",
taesd_path: "",
esrgan_path: "",
control_net_path: "",
embedding_dir: "",
embeddings: {
}
wtype: NONE,
tensor_type_rules: "",
lora_model_dir: ".",
hires_upscalers_dir: "",
photo_maker_path: "",
rng_type: cuda,
sampler_rng_type: NONE,
offload_params_to_cpu: false,
max_vram: 0,
backend: "",
params_backend: "",
enable_mmap: false,
control_net_cpu: false,
clip_on_cpu: false,
vae_on_cpu: false,
flash_attn: false,
diffusion_flash_attn: true,
diffusion_conv_direct: false,
vae_conv_direct: true,
circular: false,
circular_x: false,
circular_y: false,
chroma_use_dit_mask: true,
qwen_image_zero_cond_t: true,
chroma_use_t5_mask: false,
chroma_t5_mask_pad: 1,
prediction: NONE,
lora_apply_mode: auto,
force_sdxl_vae_conv_scale: false
}
[DEBUG] main.cpp:86 - SDGenerationParams {
loras: "{
}",
high_noise_loras: "{
}",
prompt: "",
negative_prompt: "",
clip_skip: -1,
width: -1,
height: -1,
batch_count: 1,
init_image_path: "",
end_image_path: "",
mask_image_path: "",
control_image_path: "",
ref_image_paths: [],
control_video_path: "",
auto_resize_ref_image: true,
increase_ref_index: false,
pm_id_images_dir: "",
pm_id_embed_path: "",
pm_style_strength: 20,
skip_layers: [7, 8, 9],
sample_params: (txt_cfg: 1.50, img_cfg: 1.50, distilled_guidance: 3.50, slg.layer_count: 0, slg.layer_start: 0.01, slg.layer_end: 0.20, slg.scale: 0.00, scheduler: NONE, sample_method: NONE, sample_steps: 8, eta: inf, shifted_timestep: 0, flow_shift: inf, extra_sample_args: ),
high_noise_skip_layers: [7, 8, 9],
high_noise_sample_params: (txt_cfg: 7.00, img_cfg: 7.00, distilled_guidance: 3.50, slg.layer_count: 0, slg.layer_start: 0.01, slg.layer_end: 0.20, slg.scale: 0.00, scheduler: NONE, sample_method: NONE, sample_steps: 20, eta: inf, shifted_timestep: 0, flow_shift: inf, extra_sample_args: ),
custom_sigmas: [],
cache_mode: "",
cache_option: "",
cache: disabled (threshold=inf, start=0.15, end=0.95),
moe_boundary: 0.875,
video_frames: 1,
fps: 16,
vace_strength: 1,
strength: 0.75,
control_strength: 0.9,
seed: 42,
upscale_repeats: 1,
upscale_tile_size: 128,
hires: { enabled: false, upscaler: "Latent", model_path: "", scale: 2, target_width: 0, target_height: 0, steps: 0, denoising_strength: 0.7, custom_sigmas: [], upscale_tile_size: 128 },
vae_tiling_params: { 0, 0, 0, 0, 0.5, 0, 0, "" },
}
[INFO ] ggml_extend.hpp:63 - ggml_cuda_init: found 1 ROCm devices (Total VRAM: 32624 MiB):
[INFO ] ggml_extend.hpp:63 - Device 0: AMD Radeon AI PRO R9700, gfx1201 (0x1201), VMM: no, Wave Size: 32, VRAM: 32624 MiB
[DEBUG] ggml_extend_backend.cpp:311 - Found 2 backend devices:
[DEBUG] ggml_extend_backend.cpp:314 - #0: ROCm0
[DEBUG] ggml_extend_backend.cpp:314 - #1: CPU
[DEBUG] ggml_extend_backend.cpp:291 - Initializing backend: ROCm0
[INFO ] stable-diffusion.cpp:260 - loading diffusion model from './modelsEdit/qwen-image-edit-2511-Q8_0.gguf'
[INFO ] model.cpp:216 - load ./modelsEdit/qwen-image-edit-2511-Q8_0.gguf using gguf format
[DEBUG] model.cpp:265 - init from './modelsEdit/qwen-image-edit-2511-Q8_0.gguf'
[INFO ] stable-diffusion.cpp:307 - loading llm from './modelsEdit/Qwen2.5-VL-7B-Instruct.Q8_0.gguf'
[INFO ] model.cpp:216 - load ./modelsEdit/Qwen2.5-VL-7B-Instruct.Q8_0.gguf using gguf format
[DEBUG] model.cpp:265 - init from './modelsEdit/Qwen2.5-VL-7B-Instruct.Q8_0.gguf'
[INFO ] stable-diffusion.cpp:314 - loading llm vision from './modelsEdit/Qwen2.5-VL-7B-Instruct.mmproj-Q8_0.gguf'
[INFO ] model.cpp:216 - load ./modelsEdit/Qwen2.5-VL-7B-Instruct.mmproj-Q8_0.gguf using gguf format
[DEBUG] model.cpp:265 - init from './modelsEdit/Qwen2.5-VL-7B-Instruct.mmproj-Q8_0.gguf'
[INFO ] stable-diffusion.cpp:321 - loading vae from './modelsEdit/qwen_image_vae.safetensors'
[INFO ] model.cpp:219 - load ./modelsEdit/qwen_image_vae.safetensors using safetensors format
[DEBUG] model.cpp:294 - init from './modelsEdit/qwen_image_vae.safetensors', prefix = 'vae.'
[INFO ] stable-diffusion.cpp:363 - Version: Qwen Image
[INFO ] stable-diffusion.cpp:391 - Weight type stat: f32: 1520 | f16: 34 | q8_0: 1231 | bf16: 200
[INFO ] stable-diffusion.cpp:392 - Conditioner weight type stat: f32: 432 | f16: 34 | q8_0: 391
[INFO ] stable-diffusion.cpp:393 - Diffusion model weight type stat: f32: 1088 | q8_0: 840 | bf16: 6
[INFO ] stable-diffusion.cpp:394 - VAE weight type stat: bf16: 194
[DEBUG] stable-diffusion.cpp:396 - ggml tensor size = 400 bytes
[DEBUG] qwen2_tokenizer.cpp:14 - merges size 151387
[DEBUG] qwen2_tokenizer.cpp:39 - vocab size: 151674
[DEBUG] llm.hpp:1247 - llm: num_layers = 28, vocab_size = 152064, hidden_size = 3584, intermediate_size = 18944
[DEBUG] llm.hpp:1253 - enable llm vision
[DEBUG] llm.hpp:1255 - llama.cpp style vision weight
[INFO ] qwen_image.hpp:519 - qwen_image_params.num_layers: 60
[INFO ] qwen_image.hpp:521 - use zero_cond_t
[INFO ] stable-diffusion.cpp:771 - using VAE for encoding / decoding
[INFO ] stable-diffusion.cpp:791 - Using Conv2d direct in the vae model
[INFO ] stable-diffusion.cpp:873 - Using flash attention in the diffusion model
[DEBUG] stable-diffusion.cpp:902 - loading weights
[DEBUG] ggml_extend.hpp:2711 - qwen2.5vl params backend buffer size = 7979.08 MB(VRAM) (857 tensors)
[DEBUG] ggml_extend.hpp:2711 - qwen_image params backend buffer size = 20753.54 MB(VRAM) (1933 tensors)
[DEBUG] ggml_extend.hpp:2711 - wan_vae params backend buffer size = 242.10 MB(VRAM) (194 tensors)
[INFO ] model.cpp:807 - NOT using mmap for './modelsEdit/qwen-image-edit-2511-Q8_0.gguf' (mmap disabled by caller)
[INFO ] model.cpp:807 - NOT using mmap for './modelsEdit/Qwen2.5-VL-7B-Instruct.Q8_0.gguf' (mmap disabled by caller)
[INFO ] model.cpp:807 - NOT using mmap for './modelsEdit/Qwen2.5-VL-7B-Instruct.mmproj-Q8_0.gguf' (mmap disabled by caller)
[INFO ] model.cpp:807 - NOT using mmap for './modelsEdit/qwen_image_vae.safetensors' (mmap disabled by caller)
[INFO ] model.cpp:818 - model files processing completed in 0.00s
[DEBUG] model.cpp:917 - using 8 threads for model loading
[DEBUG] model.cpp:933 - loading tensors from ./modelsEdit/qwen-image-edit-2511-Q8_0.gguf
|================================> | 1934/2985 - 2.35GB/s
[DEBUG] model.cpp:933 - loading tensors from ./modelsEdit/Qwen2.5-VL-7B-Instruct.Q8_0.gguf
|======================================> | 2272/2985 - 2.19GB/s
[DEBUG] model.cpp:933 - loading tensors from ./modelsEdit/Qwen2.5-VL-7B-Instruct.mmproj-Q8_0.gguf
|==============================================> | 2791/2985 - 2.17GB/s
[DEBUG] model.cpp:933 - loading tensors from ./modelsEdit/qwen_image_vae.safetensors
|==================================================| 2985/2985 - 2.15GB/s
[INFO ] model.cpp:1151 - loading tensors completed, taking 13.16s (read: 0.43s, memcpy: 0.00s, convert: 0.02s, copy_to_backend: 10.92s)
[DEBUG] stable-diffusion.cpp:993 - finished loaded file
[INFO ] stable-diffusion.cpp:1075 - total params memory size = 28974.72MB (VRAM 28974.72MB, RAM 0.00MB): text_encoders 7979.08MB(VRAM), diffusion_model 20753.55MB(VRAM), vae 242.10MB(VRAM), controlnet 0.00MB(N/A), pmid 0.00MB(N/A)
[INFO ] stable-diffusion.cpp:1156 - running in FLOW mode
[INFO ] main.cpp:148 - listening on: http://192.168.1.2:8001