PhysicalScienceBART

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

• Loss: 4.2991
• Rouge1: 53.186
• Rouge2: 19.5939
• Rougel: 38.452
• Rougelsum: 49.3854
• Bertscore Precision: 82.8832
• Bertscore Recall: 84.3034
• Bertscore F1: 83.5838
• Bleu: 0.1422
• Gen Len: 196.4045

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: 5e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bertscore Precision	Bertscore Recall	Bertscore F1	Bleu	Gen Len
6.1797	0.0620	100	5.9428	45.7148	14.7784	32.4391	42.4853	79.9576	82.0698	80.9951	0.1054	196.4045
5.7661	0.1239	200	5.5214	44.5312	15.1622	32.5105	41.2065	79.981	82.4	81.1665	0.1088	196.4045
5.2648	0.1859	300	5.2101	45.5969	15.6417	33.588	42.4324	80.1261	82.5551	81.3158	0.1119	196.4045
5.2069	0.2478	400	5.0522	49.0072	16.6961	34.477	45.2146	80.7244	83.1085	81.8929	0.1201	196.4045
4.9897	0.3098	500	4.9185	48.8492	16.7109	35.2037	45.3551	81.2776	83.2272	82.236	0.1207	196.4045
4.8413	0.3717	600	4.8053	48.7091	16.9882	35.3917	45.1652	81.4957	83.4015	82.4325	0.1226	196.4045
4.829	0.4337	700	4.6973	50.548	17.8895	36.2198	47.0729	81.8959	83.5895	82.73	0.1278	196.4045
4.6419	0.4957	800	4.6161	50.6164	18.2248	36.6206	46.8571	81.8709	83.7859	82.8123	0.1313	196.4045
4.5451	0.5576	900	4.5494	51.9353	18.3864	37.1004	48.168	82.3316	83.9031	83.1059	0.1326	196.4045
4.4911	0.6196	1000	4.4939	51.8997	18.8308	37.4581	47.9484	82.3519	83.9865	83.1569	0.1361	196.4045
4.5189	0.6815	1100	4.4391	52.0976	19.0604	37.6501	48.4561	82.5212	84.0028	83.2516	0.1365	196.4045
4.4382	0.7435	1200	4.4061	53.1857	19.3566	37.9447	49.2686	82.7376	84.2523	83.4844	0.1401	196.4045
4.4027	0.8055	1300	4.3583	52.2536	19.1902	37.9482	48.549	82.6541	84.1335	83.3833	0.1388	196.4045
4.3911	0.8674	1400	4.3376	52.243	19.139	38.0374	48.6274	82.6627	84.1627	83.4023	0.1385	196.4045
4.29	0.9294	1500	4.3162	53.3823	19.4988	38.381	49.4473	82.9505	84.3617	83.6468	0.1419	196.4045
4.3218	0.9913	1600	4.2991	53.186	19.5939	38.452	49.3854	82.8832	84.3034	83.5838	0.1422	196.4045

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1