so far I tested the portable on 7 computers
OK - i7 1260p intel XE
OK - i71165 g7 nvidia mx450 2gb
OK - i7 8565u gtx 1650
error - i7 8gen gtx 1070
error - i7 9gen rtx 2070
error - amd 7840 rtx 4060
error - amd 7735 rtx 4050
error apears when trying to load any model
below i included the error after a new installation
the error is identical with new installation or running from usb
all computers win 10 except the 7735 that was with win 11
there is a difference in the last few lines of protocol 1 and 2
protocol 1
LOCAL AI IMAGE GENERATOR - First-Time Setup
100% Self-Contained | No System Install Required
[1/4] Setting up portable Node.js (app/tools/node-win/)
Downloading: Node.js v22.12.0 LTS (Portable ZIP)
[################################################] 100% 33.3 MB Done!
Extracting Node.js...
[################################################] 100% 3149 files extracted!
OK Portable Node.js ready: v22.12.0
[2/4] Setting up stable-diffusion.cpp CUDA GPU backend (app/backend/win/cuda/)
Downloading: stable-diffusion.cpp CUDA Backend (Windows x64)
[################################################] 100% 334.7 MB Done!
Extracting CUDA Backend...
[################################################] 100% 5 files extracted!
OK CUDA GPU backend binaries installed successfully!
CUDA runtime DLLs are missing from backend folder. Downloading portable CUDA v12 runtime...
Downloading: CUDA v12 Runtime DLLs (llama.cpp)
[################################################] 100% 373.3 MB Done!
Extracting CUDA Runtime DLLs...
[################################################] 100% 3 files extracted!
OK CUDA runtime DLLs set up successfully!
[2/4] Setting up stable-diffusion.cpp Vulkan GPU backend for comparison (app/backend/win/vulkan/)
Downloading: stable-diffusion.cpp Vulkan Backend (Windows x64)
[################################################] 100% 35.6 MB Done!
Extracting Vulkan Backend...
[################################################] 100% 5 files extracted!
OK Vulkan GPU backend binaries installed successfully!
[3/4] Installing frontend dependencies (app/frontend/)
added 24 packages, and audited 25 packages in 2s
10 packages are looking for funding
run npm fund for details
found 0 vulnerabilities
OK Dependencies installed!
[4/4] Building frontend -> app/dist/
tauri-app@0.1.0 build
vite build
vite v8.0.16 building client environment for production...
✓ 1750 modules transformed.
computing gzip size...
../dist/index.html 0.73 kB │ gzip: 0.41 kB
../dist/assets/index-CAkR9GKS.css 22.17 kB │ gzip: 4.52 kB
../dist/assets/event-CWEsQSCY.js 1.03 kB │ gzip: 0.54 kB
../dist/assets/index-Bzo725_1.js 280.57 kB │ gzip: 85.01 kB
✓ built in 201ms
OK Frontend built!
============================================================
Setup complete! Just double-click start.bat to launch.
Press Enter to close...:
LOCAL AI IMAGE GENERATOR | Running
Server Build: polish-setup-v1
Frontend : http://localhost:1420
Backend : http://127.0.0.1:8080
[backend] Ready. Waiting for model load request from the webapp...
[api] Imported model file: CyberRealistic_V8_FP16.safetensors
[api] Restart backend request: {
model: 'CyberRealistic_V8_FP16.safetensors',
steps: 4,
cfgScale: 1,
sampler: 'euler_a',
threads: 8,
use_gpu: false,
backend_type: 'cpu',
vae_tiling: true,
vae_on_cpu: false
}
[backend] Starting: sd-vulkan.exe --listen-port 8080 --model D:\Local-AI-Image-Generator-main\app\models\CyberRealistic_V8_FP16.safetensors --steps 4 --cfg-scale 1 --sampling-method euler_a --threads 8 --backend cpu --params-backend cpu --rng cpu --sampler-rng cpu --vae-tiling
[sd-err] ggml_vulkan: Found 2 Vulkan devices:
[sd-err] ggml_vulkan: 0 = AMD Radeon 780M Graphics (AMD proprietary driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
[sd-err] ggml_vulkan: 1 = NVIDIA GeForce RTX 4060 Laptop GPU (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: KHR_coopmat
[backend] exited with code 3221225477
protocol 2
============================================================
PORTABLE DIFFUSION | First-Time Setup
This looks like your first run. Setting up automatically...
Reason: Portable Node.js is missing.
Models are not downloaded during setup. Download or import them in the app.
Press any key to continue, or Ctrl+C to cancel.
============================================================
PORTABLE DIFFUSION - First-Time Setup
100% Self-Contained | No System Install Required
[1/4] Setting up portable Node.js (app/tools/node-win/)
Downloading: Node.js v22.12.0 LTS (Portable ZIP)
[################################################] 100% 33.3 MB Done!
Extracting Node.js...
[################################################] 100% 3149 files extracted!
Waiting for disk to flush...
OK Portable Node.js ready: v22.12.0
[2/4] Setting up stable-diffusion.cpp CPU backend (app/backend/win/cpu/)
Downloading: stable-diffusion.cpp CPU Backend (Windows x64 AVX2)
[################################################] 100% 20.2 MB Done!
Extracting CPU Backend...
[################################################] 100% 5 files extracted!
OK CPU backend binaries installed successfully!
[2/4] Setting up stable-diffusion.cpp CUDA GPU backend (app/backend/win/cuda/)
Downloading: stable-diffusion.cpp CUDA Backend (Windows x64)
[################################################] 100% 334.7 MB Done!
Extracting CUDA Backend...
[################################################] 100% 5 files extracted!
OK CUDA GPU backend binaries installed successfully!
CUDA runtime DLLs are missing from backend folder. Downloading portable CUDA v12 runtime...
Downloading: CUDA v12 Runtime DLLs (llama.cpp)
[################################################] 100% 373.3 MB Done!
Extracting CUDA Runtime DLLs...
[################################################] 100% 3 files extracted!
OK CUDA runtime DLLs set up successfully!
[2/4] Setting up stable-diffusion.cpp Vulkan GPU backend for comparison (app/backend/win/vulkan/)
Downloading: stable-diffusion.cpp Vulkan Backend (Windows x64)
[################################################] 100% 35.6 MB Done!
Extracting Vulkan Backend...
[################################################] 100% 5 files extracted!
OK Vulkan GPU backend binaries installed successfully!
[3/4] Installing frontend dependencies (app/frontend/)
added 24 packages, and audited 25 packages in 2s
10 packages are looking for funding
run npm fund for details
found 0 vulnerabilities
OK Dependencies installed!
[4/4] Building frontend -> app/dist/
tauri-app@0.1.0 build
vite build
vite v8.0.16 building client environment for production...
✓ 1750 modules transformed.
computing gzip size...
../dist/index.html 0.72 kB │ gzip: 0.40 kB
../dist/assets/index-CAkR9GKS.css 22.17 kB │ gzip: 4.52 kB
../dist/assets/event-DzyF4hWi.js 1.03 kB │ gzip: 0.53 kB
../dist/assets/index-OX_tDymE.js 287.06 kB │ gzip: 86.29 kB
✓ built in 195ms
OK Frontend built!
============================================================
Setup complete! Just double-click windows.bat to launch.
Press Enter to close...:
============================================================
PORTABLE DIFFUSION | Launching...
Clearing frontend port 1420 and backend port 8080...
Starting Portable Diffusion...
Opening browser at http://localhost:1420...
============================================================
Running!
Web UI: http://localhost:1420
GPU API: Auto-selected by the app (starts at 8080)
Press Ctrl+C in this window to stop all services.
============================================================
LOCAL AI IMAGE GENERATOR | Running
Server Build: polish-setup-v1
Frontend : http://localhost:1420
Backend : http://127.0.0.1:8080
[backend] Ready. Waiting for model load request from the webapp...
[api] Imported model file: CyberRealistic_V8_FP16.safetensors
[api] Restart backend request: {
model: 'CyberRealistic_V8_FP16.safetensors',
steps: 4,
cfgScale: 1,
sampler: 'euler_a',
threads: 8,
use_gpu: false,
backend_type: 'cpu',
width: 1024,
height: 1024,
vae_tiling: true,
vae_on_cpu: false,
flash_attn: true
}
[backend] Starting: sd-cpu.exe --listen-port 8080 --model D:\Portable-Diffusion-main\app\models\CyberRealistic_V8_FP16.safetensors --steps 4 --cfg-scale 1 --sampling-method euler_a --threads 8 --backend cpu --params-backend cpu --rng cpu --sampler-rng cpu --vae-tiling --fa
[backend] exited with code 3221225477
so far I tested the portable on 7 computers
OK - i7 1260p intel XE
OK - i71165 g7 nvidia mx450 2gb
OK - i7 8565u gtx 1650
error - i7 8gen gtx 1070
error - i7 9gen rtx 2070
error - amd 7840 rtx 4060
error - amd 7735 rtx 4050
error apears when trying to load any model
below i included the error after a new installation
the error is identical with new installation or running from usb
all computers win 10 except the 7735 that was with win 11
there is a difference in the last few lines of protocol 1 and 2
protocol 1
LOCAL AI IMAGE GENERATOR - First-Time Setup
100% Self-Contained | No System Install Required
[1/4] Setting up portable Node.js (app/tools/node-win/)
[################################################] 100% 33.3 MB Done!
[################################################] 100% 3149 files extracted!
OK Portable Node.js ready: v22.12.0
[2/4] Setting up stable-diffusion.cpp CUDA GPU backend (app/backend/win/cuda/)
[################################################] 100% 334.7 MB Done!
[################################################] 100% 5 files extracted!
OK CUDA GPU backend binaries installed successfully!
[################################################] 100% 373.3 MB Done!
[################################################] 100% 3 files extracted!
OK CUDA runtime DLLs set up successfully!
[2/4] Setting up stable-diffusion.cpp Vulkan GPU backend for comparison (app/backend/win/vulkan/)
[################################################] 100% 35.6 MB Done!
[################################################] 100% 5 files extracted!
OK Vulkan GPU backend binaries installed successfully!
[3/4] Installing frontend dependencies (app/frontend/)
added 24 packages, and audited 25 packages in 2s
10 packages are looking for funding
run
npm fundfor detailsfound 0 vulnerabilities
OK Dependencies installed!
[4/4] Building frontend -> app/dist/
vite v8.0.16 building client environment for production...
✓ 1750 modules transformed.
computing gzip size...
../dist/index.html 0.73 kB │ gzip: 0.41 kB
../dist/assets/index-CAkR9GKS.css 22.17 kB │ gzip: 4.52 kB
../dist/assets/event-CWEsQSCY.js 1.03 kB │ gzip: 0.54 kB
../dist/assets/index-Bzo725_1.js 280.57 kB │ gzip: 85.01 kB
✓ built in 201ms
OK Frontend built!
============================================================
Setup complete! Just double-click start.bat to launch.
Press Enter to close...:
LOCAL AI IMAGE GENERATOR | Running
Server Build: polish-setup-v1
Frontend : http://localhost:1420
Backend : http://127.0.0.1:8080
[backend] Ready. Waiting for model load request from the webapp...
[api] Imported model file: CyberRealistic_V8_FP16.safetensors
[api] Restart backend request: {
model: 'CyberRealistic_V8_FP16.safetensors',
steps: 4,
cfgScale: 1,
sampler: 'euler_a',
threads: 8,
use_gpu: false,
backend_type: 'cpu',
vae_tiling: true,
vae_on_cpu: false
}
[backend] Starting: sd-vulkan.exe --listen-port 8080 --model D:\Local-AI-Image-Generator-main\app\models\CyberRealistic_V8_FP16.safetensors --steps 4 --cfg-scale 1 --sampling-method euler_a --threads 8 --backend cpu --params-backend cpu --rng cpu --sampler-rng cpu --vae-tiling
[sd-err] ggml_vulkan: Found 2 Vulkan devices:
[sd-err] ggml_vulkan: 0 = AMD Radeon 780M Graphics (AMD proprietary driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
[sd-err] ggml_vulkan: 1 = NVIDIA GeForce RTX 4060 Laptop GPU (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: KHR_coopmat
[backend] exited with code 3221225477
protocol 2
============================================================
PORTABLE DIFFUSION | First-Time Setup
This looks like your first run. Setting up automatically...
Reason: Portable Node.js is missing.
Models are not downloaded during setup. Download or import them in the app.
Press any key to continue, or Ctrl+C to cancel.
============================================================
PORTABLE DIFFUSION - First-Time Setup
100% Self-Contained | No System Install Required
[1/4] Setting up portable Node.js (app/tools/node-win/)
[################################################] 100% 33.3 MB Done!
[################################################] 100% 3149 files extracted!
[2/4] Setting up stable-diffusion.cpp CPU backend (app/backend/win/cpu/)
[################################################] 100% 20.2 MB Done!
[################################################] 100% 5 files extracted!
OK CPU backend binaries installed successfully!
[2/4] Setting up stable-diffusion.cpp CUDA GPU backend (app/backend/win/cuda/)
[################################################] 100% 334.7 MB Done!
[################################################] 100% 5 files extracted!
OK CUDA GPU backend binaries installed successfully!
[################################################] 100% 373.3 MB Done!
[################################################] 100% 3 files extracted!
OK CUDA runtime DLLs set up successfully!
[2/4] Setting up stable-diffusion.cpp Vulkan GPU backend for comparison (app/backend/win/vulkan/)
[################################################] 100% 35.6 MB Done!
[################################################] 100% 5 files extracted!
OK Vulkan GPU backend binaries installed successfully!
[3/4] Installing frontend dependencies (app/frontend/)
added 24 packages, and audited 25 packages in 2s
10 packages are looking for funding
run
npm fundfor detailsfound 0 vulnerabilities
OK Dependencies installed!
[4/4] Building frontend -> app/dist/
vite v8.0.16 building client environment for production...
✓ 1750 modules transformed.
computing gzip size...
../dist/index.html 0.72 kB │ gzip: 0.40 kB
../dist/assets/index-CAkR9GKS.css 22.17 kB │ gzip: 4.52 kB
../dist/assets/event-DzyF4hWi.js 1.03 kB │ gzip: 0.53 kB
../dist/assets/index-OX_tDymE.js 287.06 kB │ gzip: 86.29 kB
✓ built in 195ms
OK Frontend built!
============================================================
Setup complete! Just double-click windows.bat to launch.
Press Enter to close...:
============================================================
PORTABLE DIFFUSION | Launching...
Clearing frontend port 1420 and backend port 8080...
Starting Portable Diffusion...
Opening browser at http://localhost:1420...
============================================================
Running!
Web UI: http://localhost:1420
GPU API: Auto-selected by the app (starts at 8080)
Press Ctrl+C in this window to stop all services.
============================================================
LOCAL AI IMAGE GENERATOR | Running
Server Build: polish-setup-v1
Frontend : http://localhost:1420
Backend : http://127.0.0.1:8080
[backend] Ready. Waiting for model load request from the webapp...
[api] Imported model file: CyberRealistic_V8_FP16.safetensors
[api] Restart backend request: {
model: 'CyberRealistic_V8_FP16.safetensors',
steps: 4,
cfgScale: 1,
sampler: 'euler_a',
threads: 8,
use_gpu: false,
backend_type: 'cpu',
width: 1024,
height: 1024,
vae_tiling: true,
vae_on_cpu: false,
flash_attn: true
}
[backend] Starting: sd-cpu.exe --listen-port 8080 --model D:\Portable-Diffusion-main\app\models\CyberRealistic_V8_FP16.safetensors --steps 4 --cfg-scale 1 --sampling-method euler_a --threads 8 --backend cpu --params-backend cpu --rng cpu --sampler-rng cpu --vae-tiling --fa
[backend] exited with code 3221225477