Update README.md
Browse files
README.md
CHANGED
|
@@ -93,9 +93,36 @@ Assume you have one GPU with at least 32GB memory available and have already do
|
|
| 93 |
|
| 94 |
```bash
|
| 95 |
# Step-Audio-EditX demo
|
| 96 |
-
python app.py --model-path where_you_download_dir --model-source local
|
| 97 |
```
|
| 98 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 99 |
## Citation
|
| 100 |
|
| 101 |
```
|
|
|
|
| 93 |
|
| 94 |
```bash
|
| 95 |
# Step-Audio-EditX demo
|
| 96 |
+
python app.py --model-path where_you_download_dir --model-source local
|
| 97 |
```
|
| 98 |
|
| 99 |
+
#### Local Inference Demo
|
| 100 |
+
> [!TIP]
|
| 101 |
+
> For optimal performance, keep audio under 30 seconds per inference.
|
| 102 |
+
|
| 103 |
+
```bash
|
| 104 |
+
# zero-shot cloning
|
| 105 |
+
python3 tts_infer.py \
|
| 106 |
+
--model-path where_you_download_dir \
|
| 107 |
+
--output-dir ./output \
|
| 108 |
+
--prompt-text "your prompt text"\
|
| 109 |
+
--prompt-audio your_prompt_audio_path \
|
| 110 |
+
--generated-text "your target text" \
|
| 111 |
+
--edit-type "clone"
|
| 112 |
+
|
| 113 |
+
# edit
|
| 114 |
+
python3 tts_infer.py \
|
| 115 |
+
--model-path where_you_download_dir \
|
| 116 |
+
--output-dir ./output \
|
| 117 |
+
--prompt-text "your promt text" \
|
| 118 |
+
--prompt-audio your_prompt_audio_path \
|
| 119 |
+
--generated-text "" \ # for para-linguistic editing, you need to specify the generatedd text
|
| 120 |
+
--edit-type "emotion" \
|
| 121 |
+
--edit-info "sad" \
|
| 122 |
+
--n-edit-iter 2
|
| 123 |
+
```
|
| 124 |
+
|
| 125 |
+
|
| 126 |
## Citation
|
| 127 |
|
| 128 |
```
|