mt5-base-finetuned-easy-read

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

• Loss: 1.2157
• Rouge1: 13.2155
• Rouge2: 8.7598
• Rougel: 11.9101
• Rougelsum: 12.5918
• Gen Len: 19.0

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
7.3366	1.0	900	1.4896	12.1013	8.0714	11.045	11.6163	19.0
1.8856	2.0	1800	1.3863	12.0717	8.0051	11.0314	11.5191	19.0
1.7183	3.0	2700	1.3473	12.8751	8.1852	11.335	12.2019	19.0
1.6332	4.0	3600	1.3175	12.9756	8.3169	11.4568	12.3033	19.0
1.5669	5.0	4500	1.2978	12.8012	8.4319	11.5694	12.174	19.0
1.5251	6.0	5400	1.2814	12.8617	8.5419	11.6148	12.1836	19.0
1.4888	7.0	6300	1.2669	13.0225	8.5693	11.6414	12.2629	19.0
1.4499	8.0	7200	1.2606	13.0134	8.5403	11.6492	12.3319	19.0
1.4277	9.0	8100	1.2469	12.9822	8.5289	11.6754	12.2819	19.0
1.3995	10.0	9000	1.2478	13.0371	8.5644	11.6568	12.3571	19.0
1.3703	11.0	9900	1.2420	13.0501	8.5961	11.7072	12.3515	19.0
1.359	12.0	10800	1.2338	13.077	8.6849	11.8597	12.5109	19.0
1.3236	13.0	11700	1.2253	13.1415	8.7898	11.8972	12.5459	19.0
1.3134	14.0	12600	1.2262	13.1535	8.7186	11.8577	12.5107	19.0
1.3034	15.0	13500	1.2299	13.1871	8.7239	11.8705	12.5633	19.0
1.2876	16.0	14400	1.2232	13.2314	8.7634	11.896	12.5948	19.0
1.2657	17.0	15300	1.2229	13.2186	8.8291	11.8677	12.5579	19.0
1.2528	18.0	16200	1.2199	13.2287	8.8328	11.9311	12.6063	19.0
1.2461	19.0	17100	1.2138	13.147	8.7214	11.8507	12.5345	19.0
1.2408	20.0	18000	1.2168	13.2191	8.7845	11.9416	12.5962	19.0
1.2159	21.0	18900	1.2200	13.2313	8.8051	11.9385	12.6116	19.0
1.2121	22.0	19800	1.2175	13.1972	8.7618	11.9587	12.6047	19.0
1.2017	23.0	20700	1.2200	13.1879	8.7497	11.8674	12.5388	19.0
1.2075	24.0	21600	1.2158	13.2477	8.828	11.9203	12.5977	19.0
1.2018	25.0	22500	1.2171	13.1679	8.696	11.8623	12.5189	19.0
1.2003	26.0	23400	1.2162	13.249	8.7772	11.8942	12.6011	19.0
1.1839	27.0	24300	1.2175	13.2536	8.8192	11.9307	12.6093	19.0
1.175	28.0	25200	1.2170	13.2553	8.8383	11.9474	12.6216	19.0
1.1908	29.0	26100	1.2162	13.2436	8.8004	11.9254	12.6186	19.0
1.2008	30.0	27000	1.2157	13.2155	8.7598	11.9101	12.5918	19.0

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.1
• Datasets 3.0.1
• Tokenizers 0.20.0