psxjp5/mt5-small_25

mt5-small_test

This model is a fine-tuned version of google/mt5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7284
• Rouge1: 43.3718
• Rouge2: 37.5973
• Rougel: 42.0502
• Rougelsum: 42.0648
• Bleu: 32.8345
• Gen Len: 12.6063
• Meteor: 0.3949
• True negatives: 70.2115
• False negatives: 11.206
• Cosine Sim: 0.7485

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: 16
• eval_batch_size: 16
• seed: 9
• gradient_accumulation_steps: 8
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bleu	Gen Len	Meteor	True negatives	False negatives	Cosine Sim
3.1455	1.0	175	0.9832	18.7269	15.517	18.22	18.223	7.0634	7.6229	0.1626	74.6828	57.1687	0.3949
1.1623	1.99	350	0.8542	38.7603	32.7237	37.3447	37.3752	27.4323	12.5135	0.3487	60.0	15.942	0.6992
0.9431	2.99	525	0.8017	41.5759	35.6108	40.2536	40.2695	30.7994	12.8117	0.3755	61.2689	12.3447	0.7304
0.8119	3.98	700	0.7787	43.5881	37.4245	42.1096	42.1248	32.9646	13.2176	0.3947	59.1541	9.5238	0.7582
0.7235	4.98	875	0.7477	43.4069	37.2246	41.8444	41.8616	32.9345	13.116	0.3946	63.0816	9.8085	0.7561
0.6493	5.97	1050	0.7266	40.4506	35.0072	39.1206	39.1181	29.0601	11.748	0.3687	75.5287	17.2101	0.7071
0.5871	6.97	1225	0.7284	43.3718	37.5973	42.0502	42.0648	32.8345	12.6063	0.3949	70.2115	11.206	0.7485

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3