bart-cnn-pubmed-arxiv-pubmed-v3-e100

This model is a fine-tuned version of theojolliffe/bart-cnn-pubmed-arxiv-pubmed on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1806
• Rouge1: 59.4159
• Rouge2: 48.867
• Rougel: 51.9013
• Rougelsum: 58.3382
• Gen Len: 142.0

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: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.2541	1.0	795	0.9350	52.5594	32.6314	35.2302	50.1767	142.0
0.7018	2.0	1590	0.8022	53.4804	35.4649	37.1673	51.2428	142.0
0.5266	3.0	2385	0.7752	52.9462	34.3697	36.611	50.6922	142.0
0.3475	4.0	3180	0.7771	53.4605	35.4738	38.5714	51.3798	142.0
0.2691	5.0	3975	0.7424	54.1132	35.7289	39.2653	51.6822	141.4259
0.182	6.0	4770	0.8037	53.7969	35.7324	38.4764	51.4929	141.7778
0.1446	7.0	5565	0.7686	55.0274	38.7813	42.6251	52.9847	142.0
0.1191	8.0	6360	0.7807	55.4651	38.6537	41.2746	53.578	141.8704
0.0976	9.0	7155	0.8045	55.2843	40.2358	42.8464	54.0957	142.0
0.0882	10.0	7950	0.8533	56.8288	41.6714	44.3961	54.9406	142.0
0.0721	11.0	8745	0.8962	55.3187	40.1599	43.2103	54.1964	142.0
0.0597	12.0	9540	0.8653	55.5706	40.2321	44.0075	53.9883	142.0
0.054	13.0	10335	0.8566	55.6622	40.0252	42.6907	54.0548	142.0
0.0476	14.0	11130	0.8900	57.5046	43.6309	46.449	55.9909	142.0
0.0432	15.0	11925	0.9149	55.604	39.9591	43.1729	54.3703	142.0
0.0403	16.0	12720	0.9258	55.1275	39.6566	42.3852	53.7656	142.0
0.0351	17.0	13515	0.9184	58.2352	44.6109	47.3863	56.9529	142.0
0.032	18.0	14310	0.9275	55.9687	41.2482	44.0076	54.0707	142.0
0.0313	19.0	15105	0.9635	56.3574	41.2113	44.8358	54.6279	142.0
0.0258	20.0	15900	0.9478	57.8445	44.297	46.8836	56.2003	142.0
0.0277	21.0	16695	0.9363	58.4823	46.0943	48.7817	57.5883	141.6667
0.0219	22.0	17490	0.9705	57.6022	43.9147	47.3054	56.3866	142.0
0.0231	23.0	18285	0.9857	56.5809	42.9124	46.789	55.3897	142.0
0.021	24.0	19080	1.0155	56.9745	43.8859	46.6109	55.708	142.0
0.02	25.0	19875	1.0095	57.9702	45.1809	48.2856	56.6941	142.0
0.0175	26.0	20670	0.9634	57.7023	45.1577	48.2398	56.5282	142.0
0.0161	27.0	21465	1.0197	58.739	46.3307	49.2328	57.5778	142.0
0.0186	28.0	22260	0.9790	56.1661	42.9731	45.8654	54.4365	142.0
0.0145	29.0	23055	0.9883	55.8554	41.7405	45.177	54.478	142.0
0.013	30.0	23850	0.9977	55.5831	41.2429	44.8063	53.886	142.0
0.0131	31.0	24645	0.9765	57.4478	44.8905	48.1376	56.102	141.463
0.0118	32.0	25440	1.0000	58.4282	46.6557	49.4122	57.1979	142.0
0.0117	33.0	26235	0.9924	57.1995	44.4177	47.6248	56.0251	141.2407
0.011	34.0	27030	1.0698	57.8918	45.925	49.0505	56.9352	142.0
0.0093	35.0	27825	1.0297	57.7003	45.4556	47.9919	56.5134	141.8148
0.0112	36.0	28620	1.0429	58.4039	46.6401	49.3897	57.4753	142.0
0.0101	37.0	29415	1.0761	59.2768	47.5384	50.2152	57.9493	142.0
0.0095	38.0	30210	1.0254	58.6205	47.246	50.87	57.7829	142.0
0.0087	39.0	31005	1.0216	57.7667	44.7762	48.067	56.6006	142.0
0.0082	40.0	31800	1.0587	58.4703	45.8371	48.5321	57.2036	142.0
0.0075	41.0	32595	1.0621	58.5629	46.8885	49.5943	57.4579	142.0
0.0079	42.0	33390	1.0845	57.664	45.5954	48.408	56.661	141.9815
0.0076	43.0	34185	1.0705	58.1776	46.0435	49.3126	57.138	142.0
0.0074	44.0	34980	1.0636	58.1022	46.4877	48.7985	56.9073	142.0
0.007	45.0	35775	1.0810	57.8251	44.8767	47.8991	56.5977	142.0
0.0057	46.0	36570	1.0560	58.5086	46.3448	49.2576	57.4386	142.0
0.0062	47.0	37365	1.0903	58.8772	47.2886	49.9502	57.611	142.0
0.0058	48.0	38160	1.0847	59.4672	48.3847	51.602	58.4588	142.0
0.0061	49.0	38955	1.0798	59.5308	48.0396	50.8641	58.5016	142.0
0.0062	50.0	39750	1.0795	59.5026	48.5319	51.7426	58.7111	142.0
0.0051	51.0	40545	1.0842	57.7941	46.1198	48.7341	56.7164	142.0
0.0057	52.0	41340	1.0777	58.6131	46.3924	49.0787	57.1278	142.0
0.0039	53.0	42135	1.1133	57.6447	45.6699	48.5207	56.6447	142.0
0.0038	54.0	42930	1.0714	58.1462	46.4616	49.273	57.2771	142.0
0.004	55.0	43725	1.0852	58.6577	47.2095	50.4702	57.7724	142.0
0.0044	56.0	44520	1.1152	59.0564	47.1621	50.2807	58.3122	142.0
0.0042	57.0	45315	1.0831	58.1767	46.8127	49.9166	57.1833	142.0
0.0038	58.0	46110	1.1156	57.8515	46.3229	48.6843	56.7218	142.0
0.0038	59.0	46905	1.1105	57.9332	45.8354	49.27	57.1209	142.0
0.0034	60.0	47700	1.1104	60.0207	49.2067	51.8751	58.9484	142.0
0.0028	61.0	48495	1.1533	58.3432	46.8835	50.2868	57.5427	141.6111
0.0026	62.0	49290	1.1441	58.6838	46.9472	49.9524	57.5287	142.0
0.0028	63.0	50085	1.1232	58.0202	45.5855	48.6554	56.8368	141.9444
0.0037	64.0	50880	1.1520	58.3905	47.0348	49.8478	57.3665	142.0
0.0029	65.0	51675	1.1358	59.231	48.7251	51.6138	58.5718	142.0
0.0026	66.0	52470	1.1559	58.9482	47.2137	49.4299	57.7235	142.0
0.0025	67.0	53265	1.1272	59.3333	47.7419	50.7018	58.326	142.0
0.0026	68.0	54060	1.1613	58.6404	47.3218	50.255	57.4646	142.0
0.0015	69.0	54855	1.1575	58.7927	47.7018	50.695	57.796	142.0
0.0018	70.0	55650	1.1463	58.9455	47.2691	50.176	57.9997	142.0
0.0023	71.0	56445	1.1622	58.5943	46.9325	49.4159	57.2131	142.0
0.0024	72.0	57240	1.1258	58.2779	47.4119	49.9836	57.4867	142.0
0.0019	73.0	58035	1.1333	58.9185	47.5755	50.0765	57.8661	142.0
0.0017	74.0	58830	1.1469	60.5037	49.4508	52.2863	59.6675	141.963
0.0017	75.0	59625	1.1349	59.4264	47.4554	50.0383	58.3103	142.0
0.0025	76.0	60420	1.1215	58.2795	46.9852	49.5787	57.4501	142.0
0.0012	77.0	61215	1.1272	58.2248	47.0914	50.2569	57.1888	142.0
0.001	78.0	62010	1.1648	59.3808	48.4901	51.118	58.6251	142.0
0.0011	79.0	62805	1.1433	58.8697	47.6232	50.0226	57.6299	142.0
0.001	80.0	63600	1.1486	59.0608	47.1931	50.1354	57.8687	142.0
0.0011	81.0	64395	1.1695	58.341	47.0306	49.9269	57.339	142.0
0.001	82.0	65190	1.1589	58.9283	48.4586	51.2319	57.9485	142.0
0.0009	83.0	65985	1.1868	59.1377	48.2469	50.8486	58.1111	142.0
0.001	84.0	66780	1.1664	58.7706	47.5868	50.5937	57.7824	142.0
0.0009	85.0	67575	1.1719	57.8121	45.5997	48.2442	56.5272	142.0
0.0006	86.0	68370	1.1662	58.5204	47.5947	50.1839	57.6431	142.0
0.0007	87.0	69165	1.1668	59.2416	48.2985	51.0347	58.2794	142.0
0.0007	88.0	69960	1.1619	58.6933	47.5716	50.6785	57.5726	142.0
0.0003	89.0	70755	1.1765	59.2853	48.6451	51.3017	58.2603	142.0
0.0005	90.0	71550	1.1766	59.248	48.5642	50.9843	58.1706	142.0
0.0005	91.0	72345	1.1983	59.0009	48.311	51.0192	57.9822	142.0
0.0006	92.0	73140	1.1721	59.1248	49.0902	51.9937	58.2288	142.0
0.0003	93.0	73935	1.1799	58.2448	47.4011	49.987	57.515	142.0
0.0005	94.0	74730	1.1900	59.931	49.6663	52.3233	58.962	142.0
0.0004	95.0	75525	1.1868	59.5898	49.0004	51.4835	58.6463	142.0
0.0093	96.0	76320	1.1831	59.9405	49.83	52.4355	59.0702	142.0
0.0004	97.0	77115	1.1841	59.7379	49.5435	52.5255	58.8526	142.0
0.0004	98.0	77910	1.1790	59.5515	49.0724	51.9888	58.5488	142.0
0.0003	99.0	78705	1.1786	59.7712	49.0557	51.8137	58.7144	142.0
0.0002	100.0	79500	1.1806	59.4159	48.867	51.9013	58.3382	142.0

Framework versions

• Transformers 4.19.2
• Pytorch 1.11.0+cu113
• Datasets 2.2.2
• Tokenizers 0.12.1