distilhubert-finetuned-babycry-v7

This model is a fine-tuned version of ntu-spml/distilhubert on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5864
• Accuracy: {'accuracy': 0.8695652173913043}
• F1: 0.8089
• Precision: 0.7561
• Recall: 0.8696

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: 0.001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.7417	0.5435	25	0.5925	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.7226	1.0870	50	0.6167	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.5606	1.6304	75	0.6808	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.8858	2.1739	100	0.5850	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.6573	2.7174	125	0.5968	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.7942	3.2609	150	0.6142	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.7497	3.8043	175	0.5915	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.7408	4.3478	200	0.5899	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.6499	4.8913	225	0.5989	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.6725	5.4348	250	0.5865	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.6797	5.9783	275	0.5852	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.6553	6.5217	300	0.5861	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.6535	7.0652	325	0.5863	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696
0.7297	7.6087	350	0.5865	{'accuracy': 0.8695652173913043}	0.8089	0.7561	0.8696

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Tokenizers 0.19.1