--- license: apache-2.0 base_model: facebook/wav2vec2-base tags: - generated_from_trainer metrics: - accuracy model-index: - name: SER_model_xapiens results: [] --- # SER_model_xapiens This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.8673 - Accuracy: 0.7 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 3e-05 - train_batch_size: 64 - eval_batch_size: 64 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 256 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 100 ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-------:|:----:|:---------------:|:--------:| | No log | 1.0 | 1 | 0.9329 | 0.6667 | | No log | 2.0 | 3 | 0.9100 | 0.7167 | | No log | 3.0 | 5 | 0.9221 | 0.7333 | | No log | 4.0 | 6 | 0.9121 | 0.7333 | | No log | 5.0 | 7 | 0.9391 | 0.6833 | | No log | 6.0 | 9 | 0.9593 | 0.6667 | | 0.3147 | 7.0 | 11 | 0.9780 | 0.65 | | 0.3147 | 8.0 | 12 | 0.9659 | 0.6667 | | 0.3147 | 9.0 | 13 | 0.9612 | 0.6833 | | 0.3147 | 10.0 | 15 | 0.9492 | 0.7333 | | 0.3147 | 11.0 | 17 | 0.9959 | 0.6667 | | 0.3147 | 12.0 | 18 | 0.9615 | 0.6167 | | 0.3147 | 13.0 | 19 | 0.9160 | 0.65 | | 0.2707 | 14.0 | 21 | 0.9024 | 0.6833 | | 0.2707 | 15.0 | 23 | 0.9288 | 0.6667 | | 0.2707 | 16.0 | 24 | 0.8897 | 0.7 | | 0.2707 | 17.0 | 25 | 0.9008 | 0.6833 | | 0.2707 | 18.0 | 27 | 0.9367 | 0.6833 | | 0.2707 | 19.0 | 29 | 0.8923 | 0.7667 | | 0.2198 | 20.0 | 30 | 0.8834 | 0.7833 | | 0.2198 | 21.0 | 31 | 0.8932 | 0.7667 | | 0.2198 | 22.0 | 33 | 0.8992 | 0.7167 | | 0.2198 | 23.0 | 35 | 0.9127 | 0.7 | | 0.2198 | 24.0 | 36 | 0.9262 | 0.6667 | | 0.2198 | 25.0 | 37 | 0.9023 | 0.7333 | | 0.2198 | 26.0 | 39 | 0.8684 | 0.75 | | 0.1802 | 27.0 | 41 | 0.9270 | 0.6833 | | 0.1802 | 28.0 | 42 | 0.9361 | 0.6833 | | 0.1802 | 29.0 | 43 | 0.8819 | 0.7333 | | 0.1802 | 30.0 | 45 | 0.8361 | 0.75 | | 0.1802 | 31.0 | 47 | 0.8555 | 0.6833 | | 0.1802 | 32.0 | 48 | 0.8470 | 0.7 | | 0.1802 | 33.0 | 49 | 0.8418 | 0.75 | | 0.1515 | 34.0 | 51 | 0.8077 | 0.7667 | | 0.1515 | 35.0 | 53 | 0.8207 | 0.75 | | 0.1515 | 36.0 | 54 | 0.8540 | 0.7333 | | 0.1515 | 37.0 | 55 | 0.8494 | 0.7167 | | 0.1515 | 38.0 | 57 | 0.8525 | 0.75 | | 0.1515 | 39.0 | 59 | 0.8675 | 0.7167 | | 0.1313 | 40.0 | 60 | 0.8809 | 0.7167 | | 0.1313 | 41.0 | 61 | 0.8739 | 0.7333 | | 0.1313 | 42.0 | 63 | 0.8957 | 0.6833 | | 0.1313 | 43.0 | 65 | 0.9231 | 0.6833 | | 0.1313 | 44.0 | 66 | 0.9400 | 0.6833 | | 0.1313 | 45.0 | 67 | 0.9518 | 0.7 | | 0.1313 | 46.0 | 69 | 0.9820 | 0.65 | | 0.1179 | 47.0 | 71 | 0.9966 | 0.65 | | 0.1179 | 48.0 | 72 | 0.9788 | 0.6667 | | 0.1179 | 49.0 | 73 | 0.9537 | 0.6833 | | 0.1179 | 50.0 | 75 | 0.9152 | 0.7 | | 0.1179 | 51.0 | 77 | 0.8896 | 0.6833 | | 0.1179 | 52.0 | 78 | 0.8610 | 0.6667 | | 0.1179 | 53.0 | 79 | 0.8374 | 0.7 | | 0.1049 | 54.0 | 81 | 0.8419 | 0.7333 | | 0.1049 | 55.0 | 83 | 0.8516 | 0.6667 | | 0.1049 | 56.0 | 84 | 0.8241 | 0.7 | | 0.1049 | 57.0 | 85 | 0.8027 | 0.7167 | | 0.1049 | 58.0 | 87 | 0.8049 | 0.75 | | 0.1049 | 59.0 | 89 | 0.8238 | 0.75 | | 0.0973 | 60.0 | 90 | 0.8284 | 0.7167 | | 0.0973 | 61.0 | 91 | 0.8325 | 0.7 | | 0.0973 | 62.0 | 93 | 0.8268 | 0.7333 | | 0.0973 | 63.0 | 95 | 0.8333 | 0.7333 | | 0.0973 | 64.0 | 96 | 0.8424 | 0.7333 | | 0.0973 | 65.0 | 97 | 0.8505 | 0.7167 | | 0.0973 | 66.0 | 99 | 0.8644 | 0.7 | | 0.0935 | 66.6667 | 100 | 0.8673 | 0.7 | ### Framework versions - Transformers 4.42.0.dev0 - Pytorch 2.3.0 - Datasets 2.19.2.dev0 - Tokenizers 0.19.1