End of training

README.md

@@ -1,7 +1,7 @@
 ---
 library_name: transformers
-license: mit
-base_model: roberta-base
+license: apache-2.0
+base_model: r-f/wav2vec-english-speech-emotion-recognition
 tags:
 - generated_from_trainer
 metrics:
@@ -16,10 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # results
 
-This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on an unknown dataset.
+This model is a fine-tuned version of [r-f/wav2vec-english-speech-emotion-recognition](https://huggingface.co/r-f/wav2vec-english-speech-emotion-recognition) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9055
-- Accuracy: 0.6847
+- Loss: 0.1011
+- Accuracy: 0.9724
 
 ## Model description
 
@@ -38,26 +38,40 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 2e-05
-- train_batch_size: 16
-- eval_batch_size: 16
+- learning_rate: 0.001
+- train_batch_size: 10
+- eval_batch_size: 5
 - seed: 42
 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
-- num_epochs: 3
+- lr_scheduler_warmup_steps: 50
+- num_epochs: 15
+- mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 0.9666        | 1.0   | 2714 | 0.8702          | 0.6819   |
-| 0.7747        | 2.0   | 5428 | 0.8733          | 0.6812   |
-| 0.6512        | 3.0   | 8142 | 0.9055          | 0.6847   |
+| 1.4918        | 1.0   | 232  | 1.3591          | 0.3672   |
+| 1.0899        | 2.0   | 464  | 0.9012          | 0.5672   |
+| 0.9523        | 3.0   | 696  | 1.2430          | 0.4862   |
+| 0.8062        | 4.0   | 928  | 0.6423          | 0.7759   |
+| 0.5591        | 5.0   | 1160 | 0.5161          | 0.8276   |
+| 0.4538        | 6.0   | 1392 | 0.6369          | 0.8069   |
+| 0.3527        | 7.0   | 1624 | 0.2526          | 0.9207   |
+| 0.3833        | 8.0   | 1856 | 0.2226          | 0.9328   |
+| 0.2532        | 9.0   | 2088 | 0.1955          | 0.9466   |
+| 0.1296        | 10.0  | 2320 | 0.1860          | 0.9483   |
+| 0.144         | 11.0  | 2552 | 0.1885          | 0.9552   |
+| 0.1976        | 12.0  | 2784 | 0.1243          | 0.9655   |
+| 0.0147        | 13.0  | 3016 | 0.1375          | 0.9655   |
+| 0.0149        | 14.0  | 3248 | 0.1061          | 0.9776   |
+| 0.0199        | 15.0  | 3480 | 0.1011          | 0.9724   |
 
 
 ### Framework versions
 
-- Transformers 4.51.3
+- Transformers 4.53.0
 - Pytorch 2.6.0+cu124
-- Datasets 3.5.1
-- Tokenizers 0.21.1
+- Datasets 3.6.0
+- Tokenizers 0.21.2

config.json

@@ -1,45 +1,137 @@
 {
+  "activation_dropout": 0.05,
+  "adapter_attn_dim": null,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": true,
   "architectures": [
-    "RobertaForSequenceClassification"
+    "Wav2Vec2ForSequenceClassification"
   ],
-  "attention_probs_dropout_prob": 0.1,
-  "bos_token_id": 0,
-  "classifier_dropout": null,
+  "attention_dropout": 0.1,
+  "bos_token_id": 1,
+  "classifier_proj_size": 256,
+  "codevector_dim": 256,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": true,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "mean",
+  "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": 0.05,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.0,
+  "finetuning_task": "wav2vec2_clf",
   "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "hidden_dropout": 0.05,
+  "hidden_size": 1024,
   "id2label": {
-    "0": "LABEL_0",
-    "1": "LABEL_1",
-    "2": "LABEL_2",
-    "3": "LABEL_3",
-    "4": "LABEL_4",
-    "5": "LABEL_5",
-    "6": "LABEL_6"
+    "0": "angry",
+    "1": "disgust",
+    "2": "fear",
+    "3": "happy",
+    "4": "neutral",
+    "5": "sad",
+    "6": "surprised"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
+  "intermediate_size": 4096,
   "label2id": {
-    "LABEL_0": 0,
-    "LABEL_1": 1,
-    "LABEL_2": 2,
-    "LABEL_3": 3,
-    "LABEL_4": 4,
-    "LABEL_5": 5,
-    "LABEL_6": 6
+    "angry": 0,
+    "disgust": 1,
+    "fear": 2,
+    "happy": 3,
+    "neutral": 4,
+    "sad": 5,
+    "surprised": 6
   },
   "layer_norm_eps": 1e-05,
-  "max_position_embeddings": 514,
-  "model_type": "roberta",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
-  "pad_token_id": 1,
-  "position_embedding_type": "absolute",
+  "layerdrop": 0.05,
+  "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": 0.05,
+  "mask_time_selection": "static",
+  "model_type": "wav2vec2",
+  "num_adapter_layers": 3,
+  "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": 24,
+  "num_negatives": 100,
+  "output_hidden_size": 1024,
+  "pad_token_id": 0,
+  "pooling_mode": "mean",
   "problem_type": "single_label_classification",
+  "processor_class": "Wav2Vec2CTCTokenizer",
+  "proj_codevector_dim": 256,
+  "tdnn_dilation": [
+    1,
+    2,
+    3,
+    1,
+    1
+  ],
+  "tdnn_dim": [
+    512,
+    512,
+    512,
+    512,
+    1500
+  ],
+  "tdnn_kernel": [
+    5,
+    3,
+    3,
+    1,
+    1
+  ],
   "torch_dtype": "float32",
-  "transformers_version": "4.51.3",
-  "type_vocab_size": 1,
-  "use_cache": true,
-  "vocab_size": 50265
+  "transformers_version": "4.53.0",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 33,
+  "xvector_output_dim": 512
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:291ec4e2a1e7f4381c38897326b88cd4c336451c826f052c3809792f800f8236
-size 498628204
+oid sha256:3ebb1bc36b34766f7325eb7998f2ba1cb222cc85f3d3b21b7cc90f80da757e97
+size 1262864444

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "Wav2Vec2ProcessorWithLM",
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f28fef4a21adab64fc4b3be0ba95ad0894a81aa0482791cd1dd22375c8229aa
-size 5240
+oid sha256:af5e257efd617a24dbd5e8a90678a44b6bfbccd9a90665ac4737befdf594eed4
+size 5304