Training in progress, epoch 1

config.json

@@ -1,6 +1,6 @@
 {
-  "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
-  "activation_dropout": 0.0,
+  "_name_or_path": "facebook/wav2vec2-base-960h",
+  "activation_dropout": 0.1,
   "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": 768,
+  "codevector_dim": 256,
   "contrastive_logits_temperature": 0.1,
-  "conv_bias": true,
+  "conv_bias": false,
   "conv_dim": [
     512,
     512,
@@ -44,18 +44,19 @@
   "ctc_loss_reduction": "sum",
   "ctc_zero_infinity": false,
   "diversity_loss_weight": 0.1,
-  "do_stable_layer_norm": true,
+  "do_stable_layer_norm": false,
   "eos_token_id": 2,
   "feat_extract_activation": "gelu",
   "feat_extract_dropout": 0.0,
-  "feat_extract_norm": "layer",
+  "feat_extract_norm": "group",
   "feat_proj_dropout": 0.1,
   "feat_quantizer_dropout": 0.0,
-  "final_dropout": 0.0,
+  "final_dropout": 0.1,
   "gradient_checkpointing": false,
   "hidden_act": "gelu",
   "hidden_dropout": 0.1,
-  "hidden_size": 1024,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
   "id2label": {
     "0": "angry.wav",
     "1": "happy.wav",
@@ -63,7 +64,7 @@
     "3": "sad.wav"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 4096,
+  "intermediate_size": 3072,
   "label2id": {
     "angry.wav": "0",
     "happy.wav": "1",
@@ -72,33 +73,25 @@
   },
   "layer_norm_eps": 1e-05,
   "layerdrop": 0.1,
-  "mask_channel_length": 10,
-  "mask_channel_min_space": 1,
-  "mask_channel_other": 0.0,
-  "mask_channel_prob": 0.0,
-  "mask_channel_selection": "static",
   "mask_feature_length": 10,
   "mask_feature_min_masks": 0,
   "mask_feature_prob": 0.0,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_min_space": 1,
-  "mask_time_other": 0.0,
-  "mask_time_prob": 0.075,
-  "mask_time_selection": "static",
+  "mask_time_prob": 0.05,
   "model_type": "wav2vec2",
   "num_adapter_layers": 3,
-  "num_attention_heads": 16,
+  "num_attention_heads": 12,
   "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": 24,
+  "num_hidden_layers": 12,
   "num_negatives": 100,
-  "output_hidden_size": 1024,
+  "output_hidden_size": 768,
   "pad_token_id": 0,
-  "proj_codevector_dim": 768,
+  "proj_codevector_dim": 256,
   "tdnn_dilation": [
     1,
     2,

preprocessor_config.json

@@ -1,9 +1,9 @@
 {
-  "do_normalize": true,
+  "do_normalize": false,
   "feature_extractor_type": "Wav2Vec2FeatureExtractor",
   "feature_size": 1,
   "padding_side": "right",
-  "padding_value": 0,
+  "padding_value": 0.0,
   "return_attention_mask": true,
   "sampling_rate": 16000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3cb62c24628359839db99afbb246813cf21f8bebc305168ebc86a1b181c84de2
-size 1262953069
+oid sha256:340b989c9dd187b44356438be4bab7a4b3234589552cb50ac3ab0d84444ef2b5
+size 378350611

runs/Apr12_00-13-25_a90f333fadd8/events.out.tfevents.1712880899.a90f333fadd8.17.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a5dcd73116f94a680d0e27b8bbcaa5fc8a6575a06c44461baaa06e134d1442ab
-size 7936
+oid sha256:c229e29764195e1eca22579b345c9d179cdb898fa0284ea536330e13268b7eaf
+size 10719

runs/Apr12_00-43-53_a90f333fadd8/1712882698.7578106/events.out.tfevents.1712882698.a90f333fadd8.17.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40f3048f5a682e7eeab1fade1c8709e9526cd6e82211d35c61749769effbec0d
+size 4833

runs/Apr12_00-43-53_a90f333fadd8/events.out.tfevents.1712882698.a90f333fadd8.17.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf1f1d60888a7e802fd653a620a605a6ffc0015ade496ecfebdca4b5383bfd42
+size 7709

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:310e94a85fc7a396e749e928c8abd4b251b3c181fcb7859710c91fb7e2d7cb24
+oid sha256:845119a8b54a002f1b5d62699d6454628b24c1017cb15227510faee0c78c7d63
 size 3055