add train.py

train.py

@@ -0,0 +1,13 @@
+# pip install numpy==1.26.4 transformers==4.45.2 datasets==3.0.1 librosa==0.10.2.post1 jiwer==3.0.4 bitsandbytes==0.38.0.post1 accelerate==0.26.0
+
+dataset_name = "dlxjj/common_voice_17_0"
+language = "Japanese"
+language_abbr = "ja"
+
+from datasets import load_dataset, DatasetDict
+
+common_voice = DatasetDict()
+
+common_voice["train"] = load_dataset(dataset_name, language_abbr, split="train", trust_remote_code=True)
+
+print(common_voice["train"][0])