cs_m2m_0.01_50_v0.2

This model is a fine-tuned version of facebook/m2m100_1.2B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.9335
• Bleu: 0.0
• Gen Len: 5.0

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: 0.01
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
16.8785	1.0	6	18.9290	0.0	200.0
14.9532	2.0	12	16.0041	0.0	2.0
9.473	3.0	18	11.4783	0.0	200.0
9.3819	4.0	24	11.3159	0.0	200.0
8.2604	5.0	30	10.0462	0.0	2.0
7.2006	6.0	36	9.0670	0.1706	6.0
7.1892	7.0	42	8.6881	0.0	200.0
5.8213	8.0	48	8.4889	0.0	6.0
6.088	9.0	54	8.1473	0.0	2.0
5.4934	10.0	60	7.8530	0.0	6.0
5.3899	11.0	66	7.6030	0.0	3.0
5.755	12.0	72	7.2990	0.0	3.0
5.902	13.0	78	7.1387	0.0	6.0
5.1716	14.0	84	7.3531	0.0	3.0
5.384	15.0	90	7.4897	0.0	3.0
5.772	16.0	96	7.3353	0.0	3.0
6.0137	17.0	102	7.2840	0.0	6.0
5.3564	18.0	108	7.2226	0.0	4.0
4.6533	19.0	114	6.9603	0.0	3.0
5.2785	20.0	120	7.1881	0.0	4.0
6.822	21.0	126	7.1262	0.0	4.0
5.027	22.0	132	7.5066	0.0	200.0
5.2595	23.0	138	7.0461	0.0	4.0
5.7311	24.0	144	7.5675	0.0	200.0
5.19	25.0	150	6.9761	0.0	4.0
5.4136	26.0	156	7.0165	0.0	6.0
5.3953	27.0	162	7.0036	0.0	5.0
5.1609	28.0	168	7.2334	0.0	200.0
4.1589	29.0	174	6.8345	0.0	3.0
6.129	30.0	180	7.0334	0.0	4.0
3.9707	31.0	186	6.8262	0.0	4.0
4.851	32.0	192	6.7521	0.0	4.0
4.8473	33.0	198	6.8321	0.0	4.0
4.6168	34.0	204	6.8539	0.0	4.0
4.304	35.0	210	6.9346	0.0	4.0
5.0315	36.0	216	7.0995	0.0	132.0
4.5656	37.0	222	6.9738	0.0	4.0
4.3283	38.0	228	6.8871	0.0	4.0
4.8156	39.0	234	6.9938	0.0	4.0
4.6101	40.0	240	7.0034	0.0	5.0
5.1564	41.0	246	6.9462	0.0	5.0
4.432	42.0	252	7.0158	0.0	4.0
5.0996	43.0	258	7.0378	0.0	5.0
4.3684	44.0	264	6.9261	0.0	5.0
5.2601	45.0	270	6.9520	0.0169	200.0
4.4939	46.0	276	6.9559	0.0	5.0
4.7493	47.0	282	6.9144	0.0	5.0
4.615	48.0	288	6.9272	0.0	5.0
5.5171	49.0	294	6.9316	0.0	5.0
5.077	50.0	300	6.9335	0.0	5.0

Framework versions

• Transformers 4.35.2
• Pytorch 1.13.1+cu117
• Datasets 2.16.1
• Tokenizers 0.15.0