Training in progress, epoch 1

config.json

@@ -1,6 +1,7 @@
 {
   "_name_or_path": "facebook/wav2vec2-xls-r-300m",
   "activation_dropout": 0.0,
+  "adapter_attn_dim": null,
   "adapter_kernel_size": 3,
   "adapter_stride": 2,
   "add_adapter": false,
@@ -42,14 +43,14 @@
     2
   ],
   "ctc_loss_reduction": "mean",
-  "ctc_zero_infinity": false,
+  "ctc_zero_infinity": true,
   "diversity_loss_weight": 0.1,
   "do_stable_layer_norm": true,
   "eos_token_id": 2,
   "feat_extract_activation": "gelu",
   "feat_extract_dropout": 0.0,
   "feat_extract_norm": "layer",
-  "feat_proj_dropout": 0.0,
+  "feat_proj_dropout": 0.1,
   "feat_quantizer_dropout": 0.0,
   "final_dropout": 0.0,
   "gradient_checkpointing": false,
@@ -59,13 +60,13 @@
   "initializer_range": 0.02,
   "intermediate_size": 4096,
   "layer_norm_eps": 1e-05,
-  "layerdrop": 0.0,
-  "mask_feature_length": 64,
+  "layerdrop": 0.1,
+  "mask_feature_length": 10,
   "mask_feature_min_masks": 0,
-  "mask_feature_prob": 0.25,
+  "mask_feature_prob": 0.0,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_prob": 0.75,
+  "mask_time_prob": 0.075,
   "model_type": "wav2vec2",
   "num_adapter_layers": 3,
   "num_attention_heads": 16,
@@ -77,7 +78,7 @@
   "num_hidden_layers": 24,
   "num_negatives": 100,
   "output_hidden_size": 1024,
-  "pad_token_id": 39,
+  "pad_token_id": 34,
   "proj_codevector_dim": 768,
   "tdnn_dilation": [
     1,
@@ -101,8 +102,8 @@
     1
   ],
   "torch_dtype": "float32",
-  "transformers_version": "4.27.4",
+  "transformers_version": "4.32.1",
   "use_weighted_layer_sum": false,
-  "vocab_size": 40,
+  "vocab_size": 37,
   "xvector_output_dim": 512
 }

preprocessor_config.json

@@ -4,6 +4,7 @@
   "feature_size": 1,
   "padding_side": "right",
   "padding_value": 0.0,
+  "processor_class": "Wav2Vec2Processor",
   "return_attention_mask": true,
   "sampling_rate": 16000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d5f5232d55e34057dc718bf94c6585730439cd2d8f49bfcd1d00724e47b302c0
-size 1262065837
+oid sha256:525c0a275ecc16756f13782d9fb282ab3b372f0073875b83d4540066dc2c7f8c
+size 1262053549

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aba486df3f8b8f995159feeffceb22cfb2c9827beb32eeefecb7c90c063fd0b8
-size 3643
+oid sha256:0d600e3b8cf20c205223bc826c7df2a81972e29720cd9027a2a8f95dc3901407
+size 4091