docs: update benchmarks, EP restart guide, CLI device docs, and bump version to v0.2.0-beta.1

This commit is contained in:
RevLayDust committed 2026-10-11 08:45:49 +07:00
1 parent 6037316308
commit cb9432a4c5
6 files changed
+835 -22

No files matched your search

+84 -16
View File
@@ -6,7 +6,7 @@
[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
[![Python: 3.11+](https://img.shields.io/badge/Python-3.11%2B-3776AB.svg?logo=python&logoColor=white)](https://www.python.org/downloads/latest/python3.11/)
[![Platform: Windows](https://img.shields.io/badge/Platform-Windows%2010%2F11%20(64--bit)-informational.svg?logo=windows&logoColor=white)](#-installation--prerequisites)
[![Acceleration: CUDA & CPU](https://img.shields.io/badge/Acceleration-CUDA%20%7C%20CPU-76B900.svg?logo=nvidia&logoColor=white)](#-installation--prerequisites)
[![Acceleration: CUDA, TensorRT-RTX & CPU](https://img.shields.io/badge/Acceleration-CUDA%20%7C%20TensorRT--RTX%20%7C%20CPU-76B900.svg?logo=nvidia&logoColor=white)](#-installation--prerequisites)
**High-performance, memory-optimized distribution of Ultimate Vocal Remover GUI (UVR v5.6).**
Focused on accelerated **ONNX Runtime inference**, lower hardware VRAM usage, and faster CLI startup.
@@ -19,7 +19,9 @@ Focused on accelerated **ONNX Runtime inference**, lower hardware VRAM usage, an
/>
<br/>
## Up to ~4x Faster Inference &nbsp;|&nbsp; ~53% Lower VRAM Usage
## Up to ~5.3x Faster Inference &nbsp;|&nbsp; ~77% Lower VRAM Usage
<sub>*~3.5x faster & ~52% lower VRAM on stable CUDA EP. Up to ~5.3x faster & ~77% lower VRAM with experimental TensorRT-RTX EP (see [Verified Benchmarks](#verified-benchmarks)).*</sub>
[Prerequisites & Installation](#installation--prerequisites) $\cdot$ [Usage & Quick Start](#usage--quick-start) $\cdot$ [Key Features](#key-features) $\cdot$ [Verified Benchmarks](#verified-benchmarks) $\cdot$ [Roadmap](#roadmap) $\cdot$ [Credits](#credits--acknowledgments)
@@ -34,9 +36,11 @@ Focused on accelerated **ONNX Runtime inference**, lower hardware VRAM usage, an
> **Platform Support**: RLD UVR-Opt is actively tested and optimized exclusively for **Windows 10 / 11 (64-bit)**.
* **Compute Device**:
* **NVIDIA GPU (CUDA)**: NVIDIA GeForce RTX 20 Series GPU or higher with at least 6GB VRAM is the minimum requirement for GPU conversions (**8GB+ VRAM recommended for optimal performance**).
* **NVIDIA GPU (CUDA EP - Stable)**: NVIDIA GeForce RTX 20 Series (Turing) or higher with at least 6GB VRAM (**8GB+ VRAM recommended for optimal performance**).
* **NVIDIA GPU (TensorRT-RTX EP - Experimental)**: NVIDIA GeForce RTX 30 Series (Ampere) or higher (RTX 30, 40, 50 series, RTX Ada/Blackwell).
*(Note: Turing / RTX 20-series GPUs are not supported by the TensorRT-RTX execution provider; RTX 20 users should use the stable CUDA EP).*
* **CPU**: Multi-core modern 64-bit x86 processor (fully supported via multi-threaded CPU inference).
* *(Note: Hardware focus is strictly on **CUDA** and **CPU**).*
* *(Note: Hardware focus is strictly on **GPU** and **CPU**).*
* **Python**: **Optional** (the automated installer automatically bootstraps a standalone Python 3.11 environment via `uv`). Installing system **Python 3.11+ (64-bit)** is only required if you plan on manual setup, development, or debugging.
### 2. Model & Architecture Scope
@@ -72,9 +76,19 @@ Simply double-click `install\install_packages.bat` in File Explorer, or run in t
1. Open UVR using `run_uvr.bat` or `python UVR.py`.
2. Go to **Main Settings** (to the left of the **Start Processing** button with the wrench icon) → **Additional Settings** → Change **GPU Device** from the default setting to your GPU using the dropdown menu.
3. Go back to the main menu, then enable **GPU Conversions**.
4. Re-open UVR.
5. Done! You can now start using it!
3. Still in **Additional Settings**, choose an **Execution Provider** (directly below **GPU Device**):
* The **Execution Provider** dropdown stays locked (greyed out) until a **GPU Device** has been selected in step 2.
* Once unlocked, UVR lists only the providers that are actually available for your GPU:
* `CUDAExecutionProvider` — default and stable baseline for all supported NVIDIA GPUs (RTX 20 series / Turing and newer).
* `NvTensorRTRTXExecutionProvider` — experimental and faster. Requires **NVIDIA RTX 30 series (Ampere) or newer**. It is only listed if you chose the **TensorRT-RTX EP** option in the installer and your GPU meets the hardware requirement (Turing / RTX 20 series users should select `CUDAExecutionProvider`).
* If your installation only supports `CPUExecutionProvider`, both **GPU Device** and **Execution Provider** are hidden and disabled.
> [!IMPORTANT]
> **You MUST restart UVR every time you change the Execution Provider.** The provider is applied when UVR starts, so a change made while UVR is running does not take effect until you restart. Use the **Restart Application** button in the settings menu, or close and re-open UVR.
4. Go back to the main menu, then enable **GPU Conversions**.
5. Re-open UVR.
6. Done! You can now start using it!
If you installed the **CPU version**, simply launch UVR using `run_uvr.bat`, `python UVR.py`, or by double-clicking `run_uvr.bat` in File Explorer.
@@ -117,10 +131,14 @@ If you prefer to configure your environment manually instead of running `install
```
4. **Install Base Dependencies with uv pip**:
- **For NVIDIA GPU**:
- **For NVIDIA GPU (CUDA)**:
```powershell
uv pip install -r .\install\requirements.txt
```
- **For NVIDIA GPU (TensorRT-RTX)**
```powershell
uv pip install -r .\install\requirements_trt_rtx.txt
```
- **For CPU Only**:
```powershell
uv pip install -r .\install\requirements_cpu.txt
@@ -147,19 +165,48 @@ python UVR.py
```
### B. UVR Command-Line Interface (UVR-CLI)
RLD UVR-Opt provides a dedicated, lightweight CLI designed for automation and headless environments:
RLD UVR-Opt provides a dedicated, lightweight CLI designed for automation and headless environments. Run it with the project's virtual environment:
```powershell
# 1. View all available commands instantly
uvr_cli --help
venv\Scripts\python.exe -m uvr_cli --help
# 2. Run high-performance separation with MDX-Net Inst_HQ_4
uvr_cli run --audio "path/to/song.wav" --model "UVR-MDX-NET-Inst_HQ_4.onnx" --output "output_stems"
# 2. List installed models
venv\Scripts\python.exe -m uvr_cli list-models
# 3. High-throughput separation with custom overlap and segment size
uvr_cli run -a "song.wav" -m "UVR-MDX-NET-Inst_HQ_4.onnx" --overlap 0.50 --segment-size 4000 -o "output_stems"
# 3. Run high-performance separation with MDX-Net Inst_HQ_4 (default device: cuda:0)
venv\Scripts\python.exe -m uvr_cli run --audio "path/to/song.wav" --model "UVR-MDX-NET-Inst_HQ_4.onnx" --output-dir "output_stems"
# 4. High-throughput separation with custom overlap and segment size
venv\Scripts\python.exe -m uvr_cli run -a "song.wav" -m "UVR-MDX-NET-Inst_HQ_4.onnx" --overlap 0.50 --segment-size 4000 -o "output_stems"
# 5. Choose the execution device: CUDA, TensorRT-RTX, or CPU
venv\Scripts\python.exe -m uvr_cli run -a "song.wav" -m "UVR-MDX-NET-Inst_HQ_4.onnx" --device cuda:0
venv\Scripts\python.exe -m uvr_cli run -a "song.wav" -m "UVR-MDX-NET-Inst_HQ_4.onnx" --device trt_rtx:0
venv\Scripts\python.exe -m uvr_cli run -a "song.wav" -m "UVR-MDX-NET-Inst_HQ_4.onnx" --device cpu
# 6. Benchmark and smoke test
venv\Scripts\python.exe -m uvr_cli bench -a "song.wav" -m "UVR-MDX-NET-Inst_HQ_4.onnx" --device trt_rtx:0 --rounds 3 --warmup 1
venv\Scripts\python.exe -m uvr_cli smoke-test --device cuda:0
```
| Command | Description |
| :--- | :--- |
| `list-models` | List all installed MDX-Net, VR, and Demucs models. |
| `run` | Separate an audio file without the GUI. |
| `bench` | Run a multi-round benchmark with hardware telemetry and JSON/CSV export. |
| `smoke-test` | Run an end-to-end check using generated synthetic audio. |
**Device selection (`--device` / `-d`)** is available on `run`, `bench`, and `smoke-test`:
| Value | Execution Provider | Notes |
| :--- | :--- | :--- |
| `cuda[:N]` | `CUDAExecutionProvider` | Default (`cuda:0`). Stable choice for all NVIDIA GPUs (RTX 20+ / Turing or newer). |
| `trt_rtx[:N]` | `NvTensorRTRTXExecutionProvider` | Experimental. Requires the TensorRT-RTX EP install option and an **RTX 30-series (Ampere) or newer** GPU. Automatically falls back to CUDA EP if unavailable or on Turing (RTX 20) GPUs. |
| `cpu` | `CPUExecutionProvider` | Runs without a GPU. |
An invalid value is rejected with an error that lists the valid choices. If CUDA is unavailable, the CLI falls back to CPU and prints a warning. See [`uvr_cli/README.md`](uvr_cli/README.md) for the full option reference.
---
## Key Features
@@ -191,8 +238,29 @@ All metrics were gathered using the built-in isolated benchmark suite on identic
| :--- | :--- | :---: | :---: | :---: |
| **Overlap 0.50**<br>*(Segment: 4000)* | **Inference Time**<br>**Processing Speed**<br>**Peak System RAM** | 79.84 s<br>0.61x Realtime<br>8,928 MB | **32.17 s**<br>**1.50x Realtime**<br>**3,210 MB** | **~2.48x Faster**<br>+146% Throughput<br>**-64.0% RAM Usage (5.7 GB Saved)** |
### 3. GPU Inference Performance (TensorRT-RTX EP, Experimental)
Same audio, model, and settings as above (**Overlap 0.75**, Segment 4000, Ultra Quality FP32), comparing UVR v5.6 (Legacy) against RLD UVR-Opt running on the **TensorRT-RTX Execution Provider**.
| Configuration | Metric | UVR v5.6 (Legacy) | RLD UVR-Opt (TensorRT-RTX) | Improvement |
| :--- | :--- | :---: | :---: | :---: |
| **Overlap 0.75**<br>*(Segment: 4000)* | **Inference Time**<br>**Processing Speed**<br>**Peak Device VRAM**<br>PyTorch VRAM Alloc | 13.52 s<br>3.58x Realtime<br>8,210 MB<br>6,360 MB | **2.55 s**<br>**18.94x Realtime**<br>**1,897 MB**<br>**177.5 MB** | **~5.3x Faster**<br>+429% Throughput<br>**-76.9% VRAM Usage (6.2 GB Saved)**<br>-97.2% PyTorch VRAM Allocation |
To isolate the effect of the Execution Provider, the table below compares CUDA EP and TensorRT-RTX EP **on the same software stack** (ONNX Runtime 1.24.4, same machine, same commit):
| Configuration | Metric | RLD UVR-Opt (CUDA EP) | RLD UVR-Opt (TensorRT-RTX EP) | Improvement |
| :--- | :--- | :---: | :---: | :---: |
| **Overlap 0.75**<br>*(Segment: 4000)* | **Inference Time**<br>**Processing Speed**<br>**Peak Device VRAM** | 3.60 s<br>13.44x Realtime<br>3,755 MB | **2.55 s**<br>**18.94x Realtime**<br>**1,897 MB** | **~1.41x Faster**<br>+41% Throughput<br>**-49.5% VRAM Usage** |
> [!NOTE]
> Raw per-round session logs and metric datasets are preserved in [`benchmark_results/`](benchmark_results/) (`summary.csv` and `summary_legacy.csv`).
> **How to read these numbers**:
> * The TensorRT-RTX results use **ONNX Runtime 1.24.4**, while the CUDA results in table 1 used ONNX Runtime 1.22.0. Use the same-stack comparison above for a like-for-like view of the Execution Provider itself.
> * Each result is the mean of **9 measured rounds** after 1 warmup round. The first (warmup) TensorRT-RTX run was slower at **5.40 s (8.95x Realtime)**, so expect a slower first run before the faster steady-state speed.
> * Only **Overlap 0.75** has been measured for TensorRT-RTX so far. Overlap 0.50 and CPU results are not available for this provider.
> * TensorRT-RTX is **experimental**. Results are from a single test machine and are not a universal performance guarantee.
> [!NOTE]
> Raw per-round session logs and metric datasets are preserved in [`benchmark_results/`](benchmark_results/) (`summary.csv` and `summary_legacy.csv`). The TensorRT-RTX run is session `session_20261010_123555_d732c2`, and the same-stack CUDA run is session `session_20261010_123503_8cf4dd`.
---
@@ -200,8 +268,8 @@ All metrics were gathered using the built-in isolated benchmark suite on identic
- [x] Accelerated ONNX inference pipeline and buffer optimizations.
- [x] In-memory `FP16` downcasting (e.g., `FP32` $\rightarrow$ `FP16`) and cache lifecycle management.
- [x] **TensorRT-RTX Execution Provider (via ONNX Runtime)**: Evaluate optional acceleration for NVIDIA RTX GPUs (Ampere+), with CUDA EP retained as the stable fallback.
- [ ] In-memory `BF16`/`FP8 E4M3FN` downcasting.
- [ ] **TensorRT-RTX Execution Provider (via ONNX Runtime)**: Evaluate optional acceleration for NVIDIA RTX GPUs (Ampere+), with CUDA EP retained as the stable fallback.
- [ ] **Windows ML Execution Provider (via ONNX Runtime)**: Evaluate acceleration path for non-NVIDIA hardware (AMD, Intel) on Windows 11 (24H2+).
- [ ] Support for Newer Model Architectures.
- [ ] UI/UX Overhaul.
+2 -2
View File
@@ -1,5 +1,5 @@
VERSION = 'v0.1.0-beta.1'
BUILD_ID = 'RLUV_20261009_1604' # represent RLUV_yyyymmdd_HHMM (year, month, day, hour, minute)
VERSION = 'v0.2.0-beta.1'
BUILD_ID = 'RLUV_20261011_0818' # represent RLUV_yyyymmdd_HHMM (year, month, day, hour, minute)
PATCH = BUILD_ID
PATCH_MAC = 'Not yet available'
PATCH_LINUX = 'Not yet available'
@@ -0,0 +1,334 @@
{
"audio_info": {
"channels": 2,
"duration_sec": 48.3357,
"file_name": "sample.wav",
"format": "WAV",
"samplerate": 48000,
"size_bytes": 9280542,
"subtype": "PCM_16"
},
"git_commit": "b21faacf8516b1d66355d4f9bbeba67884ee3ad1",
"input_hash": "d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c",
"model_info": {
"backend": "MDX-Net",
"batch_size": 1,
"device": "cuda:0",
"hash": "6adec0d7071e8370978d8940f6a4cef3",
"name": "UVR-MDX-NET-Inst_HQ_4.onnx",
"overlap": 0.75,
"precision": "Ultra Quality (FP32)",
"segment_size": 4000
},
"python_version": "3.11.9",
"runs": [
{
"is_valid": true,
"is_warmup": true,
"iteration": 1,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 1,
"speed_factor_rt": 8.74,
"telemetry": {
"avg_gpu_utilization_percent": 35.69,
"avg_power_watts": 46.44,
"avg_process_cpu_percent": 309.55,
"avg_system_cpu_percent": 27.69,
"baseline_device_vram_mb": 1024.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 96.0,
"peak_power_watts": 114.13,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 2015.67,
"peak_system_cpu_percent": 56.9,
"samples_count": 88,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 5.5327
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 2,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 1,
"speed_factor_rt": 13.43,
"telemetry": {
"avg_gpu_utilization_percent": 83.36,
"avg_power_watts": 101.91,
"avg_process_cpu_percent": 571.86,
"avg_system_cpu_percent": 50.11,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 114.65,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 2034.22,
"peak_system_cpu_percent": 56.9,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.5991
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 3,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 2,
"speed_factor_rt": 13.42,
"telemetry": {
"avg_gpu_utilization_percent": 80.22,
"avg_power_watts": 101.24,
"avg_process_cpu_percent": 578.53,
"avg_system_cpu_percent": 49.82,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 96.0,
"peak_power_watts": 114.94,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 2035.02,
"peak_system_cpu_percent": 56.0,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.6015
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 4,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 3,
"speed_factor_rt": 13.46,
"telemetry": {
"avg_gpu_utilization_percent": 85.11,
"avg_power_watts": 103.29,
"avg_process_cpu_percent": 566.63,
"avg_system_cpu_percent": 49.66,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 115.2,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 2035.13,
"peak_system_cpu_percent": 56.8,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.5913
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 5,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 4,
"speed_factor_rt": 13.48,
"telemetry": {
"avg_gpu_utilization_percent": 82.11,
"avg_power_watts": 103.67,
"avg_process_cpu_percent": 576.86,
"avg_system_cpu_percent": 49.22,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 96.0,
"peak_power_watts": 116.31,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 2047.09,
"peak_system_cpu_percent": 57.0,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.5866
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 6,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 5,
"speed_factor_rt": 13.42,
"telemetry": {
"avg_gpu_utilization_percent": 81.75,
"avg_power_watts": 104.8,
"avg_process_cpu_percent": 573.7,
"avg_system_cpu_percent": 49.69,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 96.0,
"peak_power_watts": 116.69,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 2035.16,
"peak_system_cpu_percent": 56.1,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.6028
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 7,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 6,
"speed_factor_rt": 13.45,
"telemetry": {
"avg_gpu_utilization_percent": 83.33,
"avg_power_watts": 104.43,
"avg_process_cpu_percent": 570.4,
"avg_system_cpu_percent": 49.13,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 116.95,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 2035.22,
"peak_system_cpu_percent": 56.1,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.5931
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 8,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 7,
"speed_factor_rt": 13.44,
"telemetry": {
"avg_gpu_utilization_percent": 79.47,
"avg_power_watts": 104.45,
"avg_process_cpu_percent": 575.69,
"avg_system_cpu_percent": 49.61,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 117.16,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 2035.53,
"peak_system_cpu_percent": 56.0,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.5976
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 9,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 8,
"speed_factor_rt": 13.42,
"telemetry": {
"avg_gpu_utilization_percent": 85.0,
"avg_power_watts": 104.62,
"avg_process_cpu_percent": 569.96,
"avg_system_cpu_percent": 49.81,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 96.0,
"peak_power_watts": 117.41,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 2035.57,
"peak_system_cpu_percent": 58.3,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.6025
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 10,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 9,
"speed_factor_rt": 13.47,
"telemetry": {
"avg_gpu_utilization_percent": 82.69,
"avg_power_watts": 105.83,
"avg_process_cpu_percent": 572.24,
"avg_system_cpu_percent": 49.89,
"baseline_device_vram_mb": 3480.58,
"peak_device_vram_mb": 3754.58,
"peak_gpu_utilization_percent": 96.0,
"peak_power_watts": 117.67,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 2041.42,
"peak_system_cpu_percent": 58.9,
"samples_count": 36,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 3.5891
}
],
"session_id": "session_20261010_123503_8cf4dd",
"summary": {
"avg_speed_factor_rt": 13.44,
"avg_total_time_sec": 3.596,
"max_total_time_sec": 3.6028,
"min_total_time_sec": 3.5866,
"peak_device_vram_mb": 3754.58,
"peak_torch_allocated_vram_mb": 177.49,
"rounds_measured": 9,
"warmup_rounds": 1
},
"system_info": {
"cpu": "AMD Ryzen 5 5600 6-Core Processor",
"cpu_count_logical": 12,
"cpu_count_physical": 6,
"cpu_name": "AMD Ryzen 5 5600 6-Core Processor",
"cuda_available": true,
"cuda_version": "12.8",
"driver_version": "617.14",
"gpu_name": "NVIDIA GeForce RTX 5060 Ti",
"onnxruntime_version": "1.24.4",
"os": "Windows-10-10.0.26100-SP0",
"python_version": "3.11.9",
"system_ram_gb": 31.93,
"torch_version": "2.11.0+cu128",
"total_gpu_vram_mb": 16311.0
},
"timestamp_utc": "2026-10-10T12:35:46.449254+00:00"
}
@@ -0,0 +1,334 @@
{
"audio_info": {
"channels": 2,
"duration_sec": 48.3357,
"file_name": "sample.wav",
"format": "WAV",
"samplerate": 48000,
"size_bytes": 9280542,
"subtype": "PCM_16"
},
"git_commit": "b21faacf8516b1d66355d4f9bbeba67884ee3ad1",
"input_hash": "d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c",
"model_info": {
"backend": "MDX-Net",
"batch_size": 1,
"device": "trt_rtx:0",
"hash": "6adec0d7071e8370978d8940f6a4cef3",
"name": "UVR-MDX-NET-Inst_HQ_4.onnx",
"overlap": 0.75,
"precision": "Ultra Quality (FP32)",
"segment_size": 4000
},
"python_version": "3.11.9",
"runs": [
{
"is_valid": true,
"is_warmup": true,
"iteration": 1,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 1,
"speed_factor_rt": 8.95,
"telemetry": {
"avg_gpu_utilization_percent": 26.08,
"avg_power_watts": 43.25,
"avg_process_cpu_percent": 277.23,
"avg_system_cpu_percent": 25.07,
"baseline_device_vram_mb": 1024.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 134.06,
"peak_process_cpu_percent": 840.1,
"peak_process_ram_mb": 1550.32,
"peak_system_cpu_percent": 70.5,
"samples_count": 78,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 5.4005
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 2,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 1,
"speed_factor_rt": 18.87,
"telemetry": {
"avg_gpu_utilization_percent": 77.42,
"avg_power_watts": 114.91,
"avg_process_cpu_percent": 567.53,
"avg_system_cpu_percent": 49.75,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 133.73,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 1585.23,
"peak_system_cpu_percent": 56.2,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5614
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 3,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 2,
"speed_factor_rt": 18.9,
"telemetry": {
"avg_gpu_utilization_percent": 76.5,
"avg_power_watts": 113.29,
"avg_process_cpu_percent": 566.14,
"avg_system_cpu_percent": 48.52,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 133.69,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 1585.41,
"peak_system_cpu_percent": 58.1,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5574
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 4,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 3,
"speed_factor_rt": 18.94,
"telemetry": {
"avg_gpu_utilization_percent": 75.42,
"avg_power_watts": 113.42,
"avg_process_cpu_percent": 567.89,
"avg_system_cpu_percent": 48.46,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 134.63,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 1595.93,
"peak_system_cpu_percent": 55.0,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.552
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 5,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 4,
"speed_factor_rt": 18.86,
"telemetry": {
"avg_gpu_utilization_percent": 79.96,
"avg_power_watts": 112.06,
"avg_process_cpu_percent": 570.8,
"avg_system_cpu_percent": 48.87,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 135.04,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 1586.1,
"peak_system_cpu_percent": 55.6,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5635
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 6,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 5,
"speed_factor_rt": 18.95,
"telemetry": {
"avg_gpu_utilization_percent": 76.92,
"avg_power_watts": 114.88,
"avg_process_cpu_percent": 565.32,
"avg_system_cpu_percent": 48.45,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 135.04,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 1587.48,
"peak_system_cpu_percent": 55.2,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5513
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 7,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 6,
"speed_factor_rt": 19.01,
"telemetry": {
"avg_gpu_utilization_percent": 74.58,
"avg_power_watts": 114.42,
"avg_process_cpu_percent": 568.31,
"avg_system_cpu_percent": 48.36,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 135.28,
"peak_process_cpu_percent": 698.1,
"peak_process_ram_mb": 1599.89,
"peak_system_cpu_percent": 53.4,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5432
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 8,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 7,
"speed_factor_rt": 19.02,
"telemetry": {
"avg_gpu_utilization_percent": 82.31,
"avg_power_watts": 113.79,
"avg_process_cpu_percent": 566.15,
"avg_system_cpu_percent": 48.75,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 134.69,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 1599.05,
"peak_system_cpu_percent": 57.5,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5411
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 9,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 8,
"speed_factor_rt": 18.91,
"telemetry": {
"avg_gpu_utilization_percent": 77.69,
"avg_power_watts": 115.99,
"avg_process_cpu_percent": 564.57,
"avg_system_cpu_percent": 49.31,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 134.81,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 1586.12,
"peak_system_cpu_percent": 57.6,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5563
},
{
"is_valid": true,
"is_warmup": false,
"iteration": 10,
"output_stems": [
"sample_(Instrumental).wav",
"sample_(Vocals).wav"
],
"round": 9,
"speed_factor_rt": 18.96,
"telemetry": {
"avg_gpu_utilization_percent": 73.54,
"avg_power_watts": 114.47,
"avg_process_cpu_percent": 571.72,
"avg_system_cpu_percent": 48.2,
"baseline_device_vram_mb": 1622.58,
"peak_device_vram_mb": 1896.58,
"peak_gpu_utilization_percent": 97.0,
"peak_power_watts": 135.5,
"peak_process_cpu_percent": 705.6,
"peak_process_ram_mb": 1599.05,
"peak_system_cpu_percent": 54.8,
"samples_count": 26,
"torch_peak_allocated_vram_mb": 177.49,
"torch_peak_reserved_vram_mb": 274.0
},
"total_time_sec": 2.5497
}
],
"session_id": "session_20261010_123555_d732c2",
"summary": {
"avg_speed_factor_rt": 18.94,
"avg_total_time_sec": 2.5529,
"max_total_time_sec": 2.5635,
"min_total_time_sec": 2.5411,
"peak_device_vram_mb": 1896.58,
"peak_torch_allocated_vram_mb": 177.49,
"rounds_measured": 9,
"warmup_rounds": 1
},
"system_info": {
"cpu": "AMD Ryzen 5 5600 6-Core Processor",
"cpu_count_logical": 12,
"cpu_count_physical": 6,
"cpu_name": "AMD Ryzen 5 5600 6-Core Processor",
"cuda_available": true,
"cuda_version": "12.8",
"driver_version": "617.14",
"gpu_name": "NVIDIA GeForce RTX 5060 Ti",
"onnxruntime_version": "1.24.4",
"os": "Windows-10-10.0.26100-SP0",
"python_version": "3.11.9",
"system_ram_gb": 31.93,
"torch_version": "2.11.0+cu128",
"total_gpu_vram_mb": 16311.0
},
"timestamp_utc": "2026-10-10T12:36:28.549515+00:00"
}
+20
View File
@@ -24,3 +24,23 @@ session_20260913_025117_bdafad,2026-09-13T02:52:27.689181+00:00,NVIDIA GeForce R
session_20260913_025117_bdafad,2026-09-13T02:52:59.955193+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cpu,4000,0.5,1,sample.wav,48.3357,2,32.1676,32.1676,1.5,0.0,1178.54,3.56,4.0,623.58,3210.07,34983459df34b64f050a82b298fe9fa39e931ed3,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20260913_025117_bdafad,2026-09-13T02:53:32.354828+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cpu,4000,0.5,1,sample.wav,48.3357,3,32.301,32.301,1.5,0.0,1177.79,3.69,4.0,626.28,3210.11,34983459df34b64f050a82b298fe9fa39e931ed3,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20260913_025117_bdafad,2026-09-13T02:54:04.802234+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cpu,4000,0.5,1,sample.wav,48.3357,4,32.3491,32.3491,1.49,0.0,1176.79,3.67,4.0,628.67,3213.14,34983459df34b64f050a82b298fe9fa39e931ed3,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:13.189099+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,warmup_1,5.5327,5.5327,8.74,177.49,3754.58,35.69,96.0,309.55,2015.67,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:16.886564+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,1,3.5991,3.5991,13.43,177.49,3754.58,83.36,97.0,571.86,2034.22,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:20.586046+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,2,3.6015,3.6015,13.42,177.49,3754.58,80.22,96.0,578.53,2035.02,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:24.275471+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,3,3.5913,3.5913,13.46,177.49,3754.58,85.11,97.0,566.63,2035.13,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:27.959920+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,4,3.5866,3.5866,13.48,177.49,3754.58,82.11,96.0,576.86,2047.09,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:31.660385+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,5,3.6028,3.6028,13.42,177.49,3754.58,81.75,96.0,573.7,2035.16,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:35.352350+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,6,3.5931,3.5931,13.45,177.49,3754.58,83.33,97.0,570.4,2035.22,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:39.047809+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,7,3.5976,3.5976,13.44,177.49,3754.58,79.47,97.0,575.69,2035.53,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:42.747802+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,8,3.6025,3.6025,13.42,177.49,3754.58,85.0,96.0,569.96,2035.57,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123503_8cf4dd,2026-10-10T12:35:46.434249+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),cuda:0,4000,0.75,1,sample.wav,48.3357,9,3.5891,3.5891,13.47,177.49,3754.58,82.69,96.0,572.24,2041.42,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:04.677120+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,warmup_1,5.4005,5.4005,8.95,177.49,1896.58,26.08,97.0,277.23,1550.32,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:07.336441+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,1,2.5614,2.5614,18.87,177.49,1896.58,77.42,97.0,567.53,1585.23,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:09.991770+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,2,2.5574,2.5574,18.9,177.49,1896.58,76.5,97.0,566.14,1585.41,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:12.641604+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,3,2.552,2.552,18.94,177.49,1896.58,75.42,97.0,567.89,1595.93,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:15.302932+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,4,2.5635,2.5635,18.86,177.49,1896.58,79.96,97.0,570.8,1586.1,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:17.952744+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,5,2.5513,2.5513,18.95,177.49,1896.58,76.92,97.0,565.32,1587.48,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:20.594048+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,6,2.5432,2.5432,19.01,177.49,1896.58,74.58,97.0,568.31,1599.89,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:23.232355+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,7,2.5411,2.5411,19.02,177.49,1896.58,82.31,97.0,566.15,1599.05,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:25.886685+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,8,2.5563,2.5563,18.91,177.49,1896.58,77.69,97.0,564.57,1586.12,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
session_20261010_123555_d732c2,2026-10-10T12:36:28.533998+00:00,NVIDIA GeForce RTX 5060 Ti,AMD Ryzen 5 5600 6-Core Processor,UVR-MDX-NET-Inst_HQ_4.onnx,6adec0d7071e8370978d8940f6a4cef3,MDX-Net,Ultra Quality (FP32),trt_rtx:0,4000,0.75,1,sample.wav,48.3357,9,2.5497,2.5497,18.96,177.49,1896.58,73.54,97.0,571.72,1599.05,b21faacf8516b1d66355d4f9bbeba67884ee3ad1,3.11.9,d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c,COMPLETED
1 session_id timestamp_utc gpu_name cpu_name model_name model_hash backend precision device segment_size overlap batch_size audio_file audio_duration_sec round total_time_sec inference_time_sec speed_factor_rt peak_torch_vram_mb peak_device_vram_mb avg_gpu_util_percent peak_gpu_util_percent avg_cpu_util_percent peak_process_ram_mb git_commit python_version input_hash status
24 session_20260913_025117_bdafad 2026-09-13T02:52:59.955193+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cpu 4000 0.5 1 sample.wav 48.3357 2 32.1676 32.1676 1.5 0.0 1178.54 3.56 4.0 623.58 3210.07 34983459df34b64f050a82b298fe9fa39e931ed3 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
25 session_20260913_025117_bdafad 2026-09-13T02:53:32.354828+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cpu 4000 0.5 1 sample.wav 48.3357 3 32.301 32.301 1.5 0.0 1177.79 3.69 4.0 626.28 3210.11 34983459df34b64f050a82b298fe9fa39e931ed3 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
26 session_20260913_025117_bdafad 2026-09-13T02:54:04.802234+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cpu 4000 0.5 1 sample.wav 48.3357 4 32.3491 32.3491 1.49 0.0 1176.79 3.67 4.0 628.67 3213.14 34983459df34b64f050a82b298fe9fa39e931ed3 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
27 session_20261010_123503_8cf4dd 2026-10-10T12:35:13.189099+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 warmup_1 5.5327 5.5327 8.74 177.49 3754.58 35.69 96.0 309.55 2015.67 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
28 session_20261010_123503_8cf4dd 2026-10-10T12:35:16.886564+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 1 3.5991 3.5991 13.43 177.49 3754.58 83.36 97.0 571.86 2034.22 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
29 session_20261010_123503_8cf4dd 2026-10-10T12:35:20.586046+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 2 3.6015 3.6015 13.42 177.49 3754.58 80.22 96.0 578.53 2035.02 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
30 session_20261010_123503_8cf4dd 2026-10-10T12:35:24.275471+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 3 3.5913 3.5913 13.46 177.49 3754.58 85.11 97.0 566.63 2035.13 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
31 session_20261010_123503_8cf4dd 2026-10-10T12:35:27.959920+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 4 3.5866 3.5866 13.48 177.49 3754.58 82.11 96.0 576.86 2047.09 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
32 session_20261010_123503_8cf4dd 2026-10-10T12:35:31.660385+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 5 3.6028 3.6028 13.42 177.49 3754.58 81.75 96.0 573.7 2035.16 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
33 session_20261010_123503_8cf4dd 2026-10-10T12:35:35.352350+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 6 3.5931 3.5931 13.45 177.49 3754.58 83.33 97.0 570.4 2035.22 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
34 session_20261010_123503_8cf4dd 2026-10-10T12:35:39.047809+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 7 3.5976 3.5976 13.44 177.49 3754.58 79.47 97.0 575.69 2035.53 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
35 session_20261010_123503_8cf4dd 2026-10-10T12:35:42.747802+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 8 3.6025 3.6025 13.42 177.49 3754.58 85.0 96.0 569.96 2035.57 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
36 session_20261010_123503_8cf4dd 2026-10-10T12:35:46.434249+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) cuda:0 4000 0.75 1 sample.wav 48.3357 9 3.5891 3.5891 13.47 177.49 3754.58 82.69 96.0 572.24 2041.42 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
37 session_20261010_123555_d732c2 2026-10-10T12:36:04.677120+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 warmup_1 5.4005 5.4005 8.95 177.49 1896.58 26.08 97.0 277.23 1550.32 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
38 session_20261010_123555_d732c2 2026-10-10T12:36:07.336441+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 1 2.5614 2.5614 18.87 177.49 1896.58 77.42 97.0 567.53 1585.23 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
39 session_20261010_123555_d732c2 2026-10-10T12:36:09.991770+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 2 2.5574 2.5574 18.9 177.49 1896.58 76.5 97.0 566.14 1585.41 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
40 session_20261010_123555_d732c2 2026-10-10T12:36:12.641604+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 3 2.552 2.552 18.94 177.49 1896.58 75.42 97.0 567.89 1595.93 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
41 session_20261010_123555_d732c2 2026-10-10T12:36:15.302932+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 4 2.5635 2.5635 18.86 177.49 1896.58 79.96 97.0 570.8 1586.1 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
42 session_20261010_123555_d732c2 2026-10-10T12:36:17.952744+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 5 2.5513 2.5513 18.95 177.49 1896.58 76.92 97.0 565.32 1587.48 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
43 session_20261010_123555_d732c2 2026-10-10T12:36:20.594048+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 6 2.5432 2.5432 19.01 177.49 1896.58 74.58 97.0 568.31 1599.89 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
44 session_20261010_123555_d732c2 2026-10-10T12:36:23.232355+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 7 2.5411 2.5411 19.02 177.49 1896.58 82.31 97.0 566.15 1599.05 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
45 session_20261010_123555_d732c2 2026-10-10T12:36:25.886685+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 8 2.5563 2.5563 18.91 177.49 1896.58 77.69 97.0 564.57 1586.12 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
46 session_20261010_123555_d732c2 2026-10-10T12:36:28.533998+00:00 NVIDIA GeForce RTX 5060 Ti AMD Ryzen 5 5600 6-Core Processor UVR-MDX-NET-Inst_HQ_4.onnx 6adec0d7071e8370978d8940f6a4cef3 MDX-Net Ultra Quality (FP32) trt_rtx:0 4000 0.75 1 sample.wav 48.3357 9 2.5497 2.5497 18.96 177.49 1896.58 73.54 97.0 571.72 1599.05 b21faacf8516b1d66355d4f9bbeba67884ee3ad1 3.11.9 d18f442363660adc6d24137ff33024e4b67de61948cff280e065f7a8e4f6e08c COMPLETED
+61 -4
View File
@@ -44,12 +44,31 @@ venv\Scripts\python.exe -m uvr_cli list-models
### 2. Direct Separation (CLI-Only)
Separate an audio file directly from the terminal without GUI:
#### Separate using default CUDA (stable):
```powershell
venv\Scripts\python.exe -m uvr_cli run `
--audio "path/to/song.wav" `
--model "UVR-MDX-NET-Inst_HQ_4.onnx" `
--precision fp16 `
--device trt_rtx:0 `
--device cuda:0 `
--output-dir "separated_outputs"
```
#### Separate using TensorRT-RTX (faster, RTX-optimized):
```powershell
venv\Scripts\python.exe -m uvr_cli run `
--audio "path/to/song.wav" `
--model "UVR-MDX-NET-Inst_HQ_4.onnx" `
--precision fp16 `
--device trt_rtx:0 `
--output-dir "separated_outputs"
```
#### Separate using CPU fallback:
```powershell
venv\Scripts\python.exe -m uvr_cli run `
--audio "path/to/song.wav" `
--model "UVR-MDX-NET-Inst_HQ_4.onnx" `
--device cpu `
--output-dir "separated_outputs"
```
@@ -60,7 +79,7 @@ venv\Scripts\python.exe -m uvr_cli run `
| `--audio` | `-a` | `Path` | *Required* | Path to the input audio file to separate (WAV, FLAC, MP3, etc.). |
| `--model` | `-m` | `String` | *Required* | Model filename (e.g. `UVR-MDX-NET-Inst_HQ_4.onnx`), friendly name, or absolute path. |
| `--precision` | `-p` | `Choice` | `fp16` | Inference precision policy: `fp16` (default, in-memory downcast), `fp32`, or `bf16` (experimental). |
| `--device` | `-d` | `String` | `cuda:0` | Target execution device (`cuda[:N]`, `trt_rtx[:N]`, or `cpu`). |
| `--device` | `-d` | `String` | `cuda:0` | Target execution device: `cuda[:N]` (default `cuda:0`), `trt_rtx[:N]`, or `cpu`. |
| `--output-dir` | `-o` | `Path` | `separated_outputs` | Output directory where separated stem files will be saved. |
| `--segment-size` | | `Integer` | *Model Default* | MDX chunk size in samples (e.g. `256`, `3000`, `4000`, `6000`). |
| `--overlap` | | `Float` | *Model Default* | Overlap ratio between adjacent chunks (e.g. `0.25`, `0.50`, `0.75`). |
@@ -69,6 +88,14 @@ venv\Scripts\python.exe -m uvr_cli run `
Supported precisions: `fp32`, `fp16` (default), and `bf16` (experimental).
#### Execution Device Selection & Fallback Behavior
| Device Argument | Provider Used | Requirements & Fallback |
| :--- | :--- | :--- |
| `cuda[:N]` | `CUDAExecutionProvider` | Requires NVIDIA GPU with CUDA (supports RTX 20+ / Turing and newer). Default is `cuda:0`. Falls back to `CPUExecutionProvider` if CUDA is not available. |
| `trt_rtx[:N]` | `NvTensorRTRTXExecutionProvider` | Requires NVIDIA RTX 30-series (Ampere) or newer and the TensorRT-RTX EP installation option. Turing architecture (RTX 20-series) is unsupported and will automatically fall back to `CUDAExecutionProvider`. |
| `cpu` | `CPUExecutionProvider` | Standard multi-threaded CPU inference. Safe fallback for non-NVIDIA or CPU-only environments. |
> [!WARNING]
> **BF16 Precision is Experimental**:
> `bf16` is currently unstable and not recommended for general use.
@@ -86,11 +113,25 @@ venv\Scripts\python.exe -m uvr_cli run `
### 3. Real Performance Benchmark
Benchmark one or more models across multiple rounds with warmup:
#### Benchmark with TensorRT-RTX:
```powershell
venv\Scripts\python.exe -m uvr_cli bench `
--audio "path/to/song.wav" `
--model "UVR-MDX-NET-Inst_HQ_4.onnx" `
--precision fp16 `
--device trt_rtx:0 `
--rounds 3 `
--warmup 1 `
--export-dir "benchmark_results"
```
#### Benchmark with standard CUDA:
```powershell
venv\Scripts\python.exe -m uvr_cli bench `
--audio "path/to/song.wav" `
--model "UVR-MDX-NET-Inst_HQ_4.onnx" `
--precision fp16 `
--device cuda:0 `
--rounds 3 `
--warmup 1 `
--export-dir "benchmark_results"
@@ -103,7 +144,7 @@ venv\Scripts\python.exe -m uvr_cli bench `
| `--audio` | `-a` | `Path` | *Required* | Path to input audio file for benchmarking (uncompressed WAV recommended). |
| `--model` | `-m` | `String` | *Required* | Model filename, friendly name, or absolute path to benchmark. |
| `--precision` | `-p` | `Choice` | `fp16` | Inference precision policy: `fp16`, `fp32`, or `bf16` (experimental). |
| `--device` | `-d` | `String` | `cuda:0` | Target execution device (`cuda[:N]`, `trt_rtx[:N]`, or `cpu`). |
| `--device` | `-d` | `String` | `cuda:0` | Target execution device: `cuda[:N]`, `trt_rtx[:N]`, or `cpu`. |
| `--rounds` | `-r` | `Integer` | `3` | Number of measured benchmark rounds to run after warmup. |
| `--warmup` | `-w` | `Integer` | `1` | Number of untimed warmup rounds to prime model sessions, ONNX memory, and CUDA allocations. |
| `--export-dir` | `-e` | `Path` | `benchmark_results` | Directory where JSON telemetry reports and `summary.csv` will be saved. |
@@ -114,9 +155,25 @@ venv\Scripts\python.exe -m uvr_cli bench `
### 4. Automated Smoke Test
Run an end-to-end verification test using an automatically generated synthetic audio clip:
```powershell
venv\Scripts\python.exe -m uvr_cli smoke-test
# Run smoke test on CUDA (default):
venv\Scripts\python.exe -m uvr_cli smoke-test --device cuda:0
# Run smoke test on TensorRT-RTX:
venv\Scripts\python.exe -m uvr_cli smoke-test --device trt_rtx:0
# Run smoke test on CPU:
venv\Scripts\python.exe -m uvr_cli smoke-test --device cpu
```
#### Parameters (`uvr_cli smoke-test`)
| Option | Flag | Type | Default | Description |
| :--- | :---: | :---: | :---: | :--- |
| `--model` | `-m` | `String` | *First Available* | Optional model name or absolute path to run smoke test on. |
| `--precision` | `-p` | `Choice` | `fp16` | Inference precision policy: `fp16`, `fp32`, or `bf16`. |
| `--device` | `-d` | `String` | `cuda:0` | Target execution device: `cuda[:N]`, `trt_rtx[:N]`, or `cpu`. |
| `--export-dir` | `-e` | `Path` | `benchmark_results` | Directory where smoke test reports will be saved. |
---
## Output Data Format