Update README.md
Browse files
README.md
CHANGED
|
@@ -4,4 +4,19 @@ language:
|
|
| 4 |
- en
|
| 5 |
base_model:
|
| 6 |
- Qwen/Qwen2-Audio-7B-Instruct
|
| 7 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
- en
|
| 5 |
base_model:
|
| 6 |
- Qwen/Qwen2-Audio-7B-Instruct
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
data process:
|
| 10 |
+
Place the WAV and JSON files in the `dev_data` directory.
|
| 11 |
+
To distinguish the recognition performance of each part, prefix the file names of part1 with `fold-d-`, prefix the file names of part2 with `fold-a-``fold-b-``fold-c-`,
|
| 12 |
+
and prefix the file names of part3 with `fold-e-`.
|
| 13 |
+
|
| 14 |
+
Environment:
|
| 15 |
+
cd ../../qwen2_audio_baseline
|
| 16 |
+
pip install -r requirements.txt
|
| 17 |
+
|
| 18 |
+
You can also use a mirror source to speed up the process.
|
| 19 |
+
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
|
| 20 |
+
|
| 21 |
+
Run:
|
| 22 |
+
sh qwen_audio_test.sh
|