nllb-200-distilled-600M-finetuned_ramayana_sns

This model is a fine-tuned version of facebook/nllb-200-distilled-600M on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.1052
• Rouge1: 37.2964
• Rouge2: 13.3703
• Rougel: 22.7017
• Rougelsum: 35.5842

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: 5.6e-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: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
3.9959	1.0	427	3.4147	17.2457	1.4651	13.9187	15.2006
3.5248	2.0	854	3.1330	16.7103	1.5796	13.8012	15.1218
3.2806	3.0	1281	2.9159	19.8896	2.3058	15.4336	18.1143
3.0859	4.0	1708	2.7715	22.584	3.1418	15.1369	20.8655
2.9227	5.0	2135	2.5706	22.2583	3.0743	15.4466	20.6234
2.7698	6.0	2562	2.4091	24.7108	3.5182	16.0307	22.8396
2.6318	7.0	2989	2.2848	24.8276	4.0799	16.1394	22.995
2.5055	8.0	3416	2.1317	26.6203	4.5536	16.398	24.9144
2.3865	9.0	3843	2.0097	27.2979	4.6529	16.5061	25.4136
2.2735	10.0	4270	1.8829	26.8753	4.8247	16.5212	25.2001
2.1712	11.0	4697	1.7707	28.2474	5.3564	16.143	26.4664
2.0894	12.0	5124	1.6892	28.8711	6.1654	16.6789	27.1269
2.0222	13.0	5551	1.6208	30.2457	6.9935	17.3494	28.5843
1.9587	14.0	5978	1.5326	30.7249	7.2678	17.8036	28.8945
1.8861	15.0	6405	1.4747	30.5734	7.558	17.9432	28.7896
1.8268	16.0	6832	1.4016	32.3279	9.0045	19.0847	30.4151
1.7727	17.0	7259	1.3328	33.7324	10.4921	20.0239	32.0321
1.7221	18.0	7686	1.2880	34.4832	9.9756	20.3859	32.6687
1.6775	19.0	8113	1.2399	34.9426	10.6686	20.4836	32.9811
1.6392	20.0	8540	1.1963	35.0751	11.6233	21.0831	33.2466
1.6007	21.0	8967	1.1644	35.6344	11.9175	21.4471	33.9996
1.5707	22.0	9394	1.1438	35.5101	12.1179	21.2347	33.7518
1.55	23.0	9821	1.1182	37.1233	12.7533	22.4229	35.3803
1.5352	24.0	10248	1.1108	36.959	13.4712	22.9926	35.4021
1.5201	25.0	10675	1.1052	37.2964	13.3703	22.7017	35.5842

Framework versions

• Transformers 4.28.0
• Pytorch 1.12.1
• Datasets 2.14.4
• Tokenizers 0.13.3