bartpho-syllable-sport

This model is a fine-tuned version of vinai/bartpho-syllable on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.5564
• Rouge1: 22.1502
• Rouge2: 10.6731
• Rougel: 17.0415
• Rougelsum: 19.4188
• Gen Len: 20.0

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 16
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.7288	1.0	2897	1.5132	22.8736	10.7796	17.5095	20.0227	20.0
0.6559	2.0	5794	1.5787	22.6719	11.5016	17.6212	20.1146	20.0
0.588	3.0	8691	1.6882	22.5239	11.3821	17.592	20.0007	20.0
0.4995	4.0	11588	1.7667	22.485	11.0341	17.379	19.7174	20.0
0.4345	5.0	14485	1.8834	22.2374	11.2249	17.4357	19.6897	20.0
0.376	6.0	17382	2.0113	22.0062	10.9701	17.1091	19.4408	20.0
0.3324	7.0	20279	2.1052	22.276	11.0747	17.3729	19.6683	20.0
0.288	8.0	23176	2.2021	22.5061	11.077	17.3248	19.7652	20.0
0.2597	9.0	26073	2.2556	22.1242	10.9297	17.1186	19.5203	20.0
0.2143	10.0	28970	2.3314	22.0919	10.8978	17.1433	19.4825	20.0
0.1862	11.0	31867	2.3885	22.1965	10.6584	17.028	19.4857	20.0
0.164	12.0	34764	2.4553	22.1645	10.8153	17.0451	19.4125	20.0
0.1434	13.0	37661	2.4914	22.1914	10.8805	17.085	19.4976	20.0
0.1275	14.0	40558	2.5209	22.2731	10.817	17.2177	19.5875	20.0
0.1172	15.0	43455	2.5463	22.1502	10.8105	17.1025	19.4512	20.0
0.1072	16.0	46352	2.5564	22.1502	10.6731	17.0415	19.4188	20.0

Framework versions

• Transformers 4.38.1
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2