Could not interpret model name : TFGridNetMamba

[Hazel-Zhu]

Hi! Thanks for sharing this amazing work. I was trying to run inference with audio_test.py on the checkpoint that you shared, but was encountering this error. Could you please help me with fixing this? Thanks in advance!

[JusperLee]

The recent changes for audio_test.py are:

• 6488e77: "Update Demo Page and New Datasets"

To change the model name from TFGridNetMamba to SPMamba, locate and update the relevant lines in this file. Look for any occurrences of TFGridNetMamba and replace them with SPMamba. If you need assistance with the exact lines to update, let me know.

[Hazel-Zhu]

Hi! Thanks for your quick response. I believe I have pulled the most recent code, and there's no occurrences of TFGridNetMamba in audio_test.py. However, when I print the config loaded in base_model.py, I found that the model_name in the best_model.pth that you shared is TFGridNetMamba . I attempted to fix this by changing the line model_class = get(conf["model_name"]) to model_class = get("SPMamba"). However, when I ran the code, it was giving me this missing key error:

I would greatly appreciate it if you could offer some additional help. Thank you so much!

[JusperLee]

https://github.com/JusperLee/SPMamba/releases/tag/v1.0

Please have a try for new checkpoints

python audio_test.py --conf_dir=checkpoints/SPMamba-xxxx/conf.yml