Jzuluaga's picture
Rename hyperparams.yaml to hyperparams.yaml
0eff6f0
# ############################################################################
# Model: WAV2VEC XLSR model for Accent Recognition (Spanish)
# see paper: https://arxiv.org/abs/2305.18283
# ############################################################################
# Hparams NEEDED
HPARAMS_NEEDED: ["encoder_dim", "out_n_neurons", "label_encoder", "softmax"]
# Modules Needed
MODULES_NEEDED: ["wav2vec2", "avg_pool", "output_mlp"]
# Feature parameters
# wav2vec2_hub: facebook/wav2vec2-base
wav2vec2_hub: "facebook/wav2vec2-large-xlsr-53"
# Pretrain folder (HuggingFace)
pretrained_path: Jzuluaga/accent-id-commonaccent_xlsr-es-spanish
# URL for the biggest Fairseq english wav2vec2 model.
# parameters
encoder_dim: 1024
out_n_neurons: 6
wav2vec2: !new:speechbrain.lobes.models.huggingface_wav2vec.HuggingFaceWav2Vec2
source: !ref <wav2vec2_hub>
output_norm: True
freeze: True
save_path: wav2vec2_checkpoints
# Mean and std normalization of the input features
mean_var_norm_emb: !new:speechbrain.processing.features.InputNormalization
norm_type: sentence
std_norm: False
avg_pool: !new:speechbrain.nnet.pooling.StatisticsPooling
return_std: False
output_mlp: !new:speechbrain.nnet.linear.Linear
input_size: !ref <encoder_dim>
n_neurons: !ref <out_n_neurons>
bias: False
model: !new:torch.nn.ModuleList
- [!ref <output_mlp>]
modules:
mean_var_norm_emb: !ref <mean_var_norm_emb>
wav2vec2: !ref <wav2vec2>
output_mlp: !ref <output_mlp>
avg_pool: !ref <avg_pool>
softmax: !new:speechbrain.nnet.activations.Softmax
label_encoder: !new:speechbrain.dataio.encoder.CategoricalEncoder
pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
loadables:
mean_var_norm_emb: !ref <mean_var_norm_emb>
wav2vec2: !ref <wav2vec2>
model: !ref <model>
label_encoder: !ref <label_encoder>
paths:
mean_var_norm_emb: !ref <pretrained_path>/normalizer_input.ckpt
wav2vec2: !ref <pretrained_path>/wav2vec2.ckpt
model: !ref <pretrained_path>/model.ckpt
label_encoder: !ref <pretrained_path>/label_encoder.txt