No frame-level results for emotion recognition?

[Lyumbreit22]

I really like and rely on this project, but I find that I cannot obtain frame-level emotion recognition results. However, the result.json file from the webui does contain frame-level recognition results, e.g., https://blog.csdn.net/weixin_42613360/article/details/159707833 and https://blog.csdn.net/weixin_33812391/article/details/157281905

Here are my codes:

from funasr import AutoModel
model = AutoModel(model="iic/emotion2vec_plus_large", hub='modelscope',disable_update=True,device="cuda:1")
wav_file = f"{model.model_path}/example/test.wav"
rec_result = model.generate(wav_file, granularity="frame", extract_embedding=False) # granularity: "utterance" or "frame"
print(rec_result)

results

rtf_avg: 0.006: 100%|██████████| 1/1 [00:00<00:00,  5.45it/s]
[{'key': '1000_0001', 'labels': ['生气/angry', '厌恶/disgusted', '恐惧/fearful', '开心/happy', '中立/neutral', '其他/other', '难过/sad', '吃惊/surprised', '<unk>'], 'scores': [0.003793130163103342, 0.0005500566912814975, 0.0008670409442856908, 0.1217915266752243, 0.00022425848874263465, 6.993473562033614e-06, 0.8683823943138123, 0.004384305793792009, 2.399044376488746e-07]}]