tanatapanun/fine-tuned-bart-2048-30-epochs

fine-tuned-bart-2084-30-epochs

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

• Loss: 0.8935
• Rouge1: 0.3436
• Rouge2: 0.1382
• Rougel: 0.3044
• Rougelsum: 0.3016
• Gen Len: 15.33

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	301	0.7890	0.2453	0.0875	0.2145	0.2161	14.83
1.8096	2.0	602	0.7325	0.2259	0.0793	0.1953	0.1953	13.6
1.8096	3.0	903	0.7239	0.2872	0.0985	0.2567	0.2559	14.63
0.6874	4.0	1204	0.7236	0.3302	0.1232	0.2922	0.2916	13.89
0.5882	5.0	1505	0.7257	0.3129	0.1167	0.2778	0.2775	14.95
0.5882	6.0	1806	0.7218	0.3269	0.1251	0.2957	0.2946	15.14
0.4982	7.0	2107	0.7403	0.3029	0.1114	0.2708	0.27	14.94
0.4982	8.0	2408	0.7417	0.3113	0.12	0.2762	0.2756	14.17
0.4299	9.0	2709	0.7470	0.3164	0.1274	0.2853	0.283	14.42
0.3815	10.0	3010	0.7505	0.3294	0.134	0.2919	0.2902	15.29
0.3815	11.0	3311	0.7725	0.3288	0.1285	0.2904	0.2908	15.26
0.3421	12.0	3612	0.7864	0.3383	0.1298	0.3055	0.3043	15.0
0.3421	13.0	3913	0.7975	0.3225	0.1219	0.2864	0.2845	15.15
0.2989	14.0	4214	0.8120	0.3326	0.1344	0.2918	0.2907	15.17
0.2652	15.0	4515	0.8128	0.3226	0.1154	0.2942	0.2934	15.0
0.2652	16.0	4816	0.8265	0.3201	0.1154	0.2845	0.2833	15.29
0.2382	17.0	5117	0.8325	0.3251	0.1265	0.2929	0.2905	15.37
0.2382	18.0	5418	0.8375	0.3348	0.1218	0.3013	0.299	15.14
0.2149	19.0	5719	0.8543	0.3373	0.1278	0.2991	0.2969	15.19
0.1956	20.0	6020	0.8638	0.3386	0.139	0.304	0.302	15.24
0.1956	21.0	6321	0.8659	0.3244	0.1253	0.2868	0.2857	15.23
0.1821	22.0	6622	0.8754	0.3325	0.1258	0.2967	0.2956	15.46
0.1821	23.0	6923	0.8775	0.3389	0.1288	0.3075	0.3062	15.22
0.164	24.0	7224	0.8779	0.3488	0.1331	0.3116	0.3105	15.56
0.159	25.0	7525	0.8839	0.3455	0.1409	0.3065	0.3039	15.44
0.159	26.0	7826	0.8885	0.3399	0.1353	0.3049	0.3022	15.37
0.1458	27.0	8127	0.8869	0.331	0.1309	0.2922	0.2901	15.36
0.1458	28.0	8428	0.8918	0.3388	0.138	0.3062	0.3031	15.34
0.1442	29.0	8729	0.8925	0.3384	0.138	0.3033	0.3006	15.23
0.1369	30.0	9030	0.8935	0.3436	0.1382	0.3044	0.3016	15.33

Framework versions

• Transformers 4.36.2
• Pytorch 1.12.1+cu113
• Datasets 2.15.0
• Tokenizers 0.15.0