test ckpt

CKPT.yaml

@@ -0,0 +1,5 @@
+# yamllint disable
+BLEU: 9.623792780691117
+end-of-epoch: true
+epoch: 216
+unixtime: 1697683012.8301373

hyperparams.yaml

@@ -0,0 +1,97 @@
+pretrained_path: HaNguyen/IWSLT-ast-w2v2-mbart
+
+lang: fr   #for the BLEU score detokenization
+target_lang: fr_XX   # for mbart initialization
+sample_rate: 16000
+
+
+# URL for the HuggingFace model we want to load (BASE here)
+wav2vec2_hub: LIA-AvignonUniversity/IWSLT2022-tamasheq-only
+
+# wav2vec 2.0 specific parameters
+wav2vec2_frozen: False
+
+# Feature parameters (W2V2 etc)
+features_dim: 768 # base wav2vec output dimension, for large replace by 1024
+
+#projection for w2v
+enc_dnn_layers: 1
+enc_dnn_neurons: 1024 #256
+
+# Transformer
+embedding_size: 256
+d_model: 1024  #256
+activation: !name:torch.nn.GELU
+
+# Outputs
+blank_index: 1
+label_smoothing: 0.1
+pad_index: 1      # pad_index defined by mbart model
+bos_index: 250008 # fr_XX bos_index defined by mbart model
+eos_index: 2
+
+# Decoding parameters
+# Be sure that the bos and eos index match with the BPEs ones
+min_decode_ratio: 0.0
+max_decode_ratio: 0.25
+valid_beam_size: 5
+test_beam_size: 5
+
+
+############################## models ################################
+#wav2vec model
+wav2vec2: !new:speechbrain.lobes.models.huggingface_transformers.wav2vec.Wav2Vec2
+  source: <wav2vec2_hub>
+  output_norm: True
+  freeze: <wav2vec2_frozen>
+  save_path: wav2vec2_checkpoint
+
+#linear projection
+enc: !new:speechbrain.lobes.models.VanillaNN.VanillaNN
+  input_shape: [null, null, 768]
+  activation: <activation>
+  dnn_blocks: 1
+  dnn_neurons: 1024
+
+#mBART
+mbart_path: facebook/mbart-large-50-many-to-many-mmt
+mbart_frozen: False
+mBART: &id004 !new:speechbrain.lobes.models.huggingface_transformers.mbart.mBART
+
+  source: facebook/mbart-large-50-many-to-many-mmt
+  freeze: <mbart_frozen>
+  save_path: mbart_checkpoint
+  target_lang: <target_lang>
+
+log_softmax: !new:speechbrain.nnet.activations.Softmax
+  apply_log: True
+
+seq_lin: !new:torch.nn.Identity
+
+modules:
+  wav2vec2: <wav2vec2>
+  enc: <enc>
+  mBART: <mBART>
+model: !new:torch.nn.ModuleList
+- [!ref <enc>]
+
+valid_search: !new:speechbrain.decoders.S2SHFTextBasedBeamSearcher
+  modules: [!ref <mBART>, null, null]
+  vocab_size: 250054
+  bos_index: 250008
+  eos_index: 2
+  min_decode_ratio: 0.0
+  max_decode_ratio: 0.25
+  beam_size: 5
+  using_eos_threshold: True
+  length_normalization: True
+
+pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
+    loadables:
+      model: !ref <model>
+      wav2vec2: !ref <wav2vec2>
+      mBART: !ref <mBART>
+    paths:
+        wav2vec2: !ref <pretrained_path>/wav2vec2.ckpt
+        model: !ref <pretrained_path>/model.ckpt
+        mBART: !ref <pretrained_path>/mBART.ckpt

mBART.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:172d2d7bb7d9e034419c6f988a52c2918bbb632daae036d8f90d7c80719e7e4b
+size 1835782657

model.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:080be2136e146f03c2b471e35d59c1ce331ef3d9240b2312ecb1cec9feb99e6a
+size 3150879

wav2vec2.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fce0a9c95ce720483661ecde058e8b273b096709c6dc68b28a5893f30b83c06d
+size 377572016