GIZ NLP Agricultural Keyword Spotter
$7,000 USD
Classify audio utterances in Luganda and English from Uganda
709 data scientists enrolled, 255 on the leaderboard
AgricultureClassificationAudioNLPUnstructuredSDG2
Uganda
11 September 2020—29 November 2020
80 days
3rd place solution
published 7 Dec 2020, 15:56
  • we trined 4 models Efficientnet-5, Efficientnet-6, Efficientnet-7 and DenseNet201.
  • each model we trined for 20 epochs with different seeds
  • model_param = { 'encoder' : 'tf_efficientnet_b5_ns', 'sample_rate': 32000, 'window_size' : 1024, 'hop_size' : 320, 'mel_bins' : 64, 'fmin' : 50, 'fmax' : 14000, 'classes_num' : 193 }
  • optimizer: AdamW
  • scheduler: get_linear_schedule_with_warmup

Github: https://github.com/Gopi-Durgaprasad/ZINDI-GIZ-NLP-Agricultural-Keyword-Spotter-3rd-place-solution

thanks for sharing

Congratulations and thank you for sharing!

Hi, where can I find the API & Auth Token