humor_model_v1

This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1840
• Accuracy: 0.9231

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9231	6	0.5754	0.9231
0.5907	2.0	13	0.3002	0.9231
0.5907	2.9231	19	0.2467	0.9231
0.3341	4.0	26	0.2232	0.9231
0.2424	4.9231	32	0.2288	0.8894
0.2424	6.0	39	0.2577	0.8510
0.1678	6.9231	45	0.2725	0.8558
0.1293	8.0	52	0.2767	0.8606
0.1293	8.9231	58	0.1823	0.9279
0.1149	9.2308	60	0.1840	0.9231

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1