Training in progress, step 100

config.json

@@ -1,6 +1,6 @@
 {
-  "_name_or_path": "facebook/wav2vec2-base-960h",
-  "activation_dropout": 0.1,
+  "_name_or_path": "alefiury/wav2vec2-large-xlsr-53-gender-recognition-librispeech",
+  "activation_dropout": 0.0,
   "adapter_kernel_size": 3,
   "adapter_stride": 2,
   "add_adapter": false,
@@ -11,9 +11,9 @@
   "attention_dropout": 0.1,
   "bos_token_id": 1,
   "classifier_proj_size": 256,
-  "codevector_dim": 256,
+  "codevector_dim": 768,
   "contrastive_logits_temperature": 0.1,
-  "conv_bias": false,
+  "conv_bias": true,
   "conv_dim": [
     512,
     512,
@@ -44,20 +44,19 @@
   "ctc_loss_reduction": "sum",
   "ctc_zero_infinity": false,
   "diversity_loss_weight": 0.1,
-  "do_stable_layer_norm": false,
+  "do_stable_layer_norm": true,
   "eos_token_id": 2,
   "feat_extract_activation": "gelu",
   "feat_extract_dropout": 0.0,
-  "feat_extract_norm": "group",
+  "feat_extract_norm": "layer",
   "feat_proj_dropout": 0.1,
   "feat_quantizer_dropout": 0.0,
-  "final_dropout": 0.1,
+  "final_dropout": 0.0,
   "finetuning_task": "wav2vec2_clf",
   "gradient_checkpointing": false,
   "hidden_act": "gelu",
   "hidden_dropout": 0.1,
-  "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "hidden_size": 1024,
   "id2label": {
     "0": "Angry",
     "1": "Happy",
@@ -65,7 +64,7 @@
     "3": "Sad"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
+  "intermediate_size": 4096,
   "label2id": {
     "Angry": 0,
     "Happy": 1,
@@ -79,22 +78,22 @@
   "mask_feature_prob": 0.0,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_prob": 0.05,
+  "mask_time_prob": 0.075,
   "model_type": "wav2vec2",
   "num_adapter_layers": 3,
-  "num_attention_heads": 12,
+  "num_attention_heads": 16,
   "num_codevector_groups": 2,
   "num_codevectors_per_group": 320,
   "num_conv_pos_embedding_groups": 16,
   "num_conv_pos_embeddings": 128,
   "num_feat_extract_layers": 7,
-  "num_hidden_layers": 12,
+  "num_hidden_layers": 24,
   "num_negatives": 100,
-  "output_hidden_size": 768,
+  "output_hidden_size": 1024,
   "pad_token_id": 0,
   "pooling_mode": "mean",
   "problem_type": "single_label_classification",
-  "proj_codevector_dim": 256,
+  "proj_codevector_dim": 768,
   "tdnn_dilation": [
     1,
     2,
@@ -117,7 +116,7 @@
     1
   ],
   "torch_dtype": "float32",
-  "transformers_version": "4.26.1",
+  "transformers_version": "4.28.1",
   "use_weighted_layer_sum": false,
   "vocab_size": 32,
   "xvector_output_dim": 512

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d23533a4e305253195d7f5926b32bfc39b97643a3865757e9b287609eda91cf
-size 377572641
+oid sha256:98243549b682bf63c56e5c6cbed6bd7f9e72bac8c4f41f7f99e3039db6811e76
+size 1261918317

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2a1538e152df0cf4611c6309dcbe58f04f2d77ca97b0334e2292e4aafa97bc36
-size 3515
+oid sha256:30992d7bca6a6ef6b049960b6f7ad1b6f29208512a1482c7ea4c003fc5d7603a
+size 3579