cs_m2m_2e-5_100_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: 2.3369
• Bleu: 48.2659
• Gen Len: 20.4286

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: 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: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.0	1.0	6	2.3719	49.7748	20.619
0.0001	2.0	12	2.4619	52.8954	20.0952
0.0242	3.0	18	2.5621	49.1697	19.9524
0.0	4.0	24	2.4757	48.7512	20.2857
0.0001	5.0	30	2.5652	43.0006	22.619
0.0001	6.0	36	2.5258	40.1532	22.2381
0.0	7.0	42	2.4040	49.8751	20.1905
0.0002	8.0	48	2.4212	49.541	19.4286
0.0001	9.0	54	2.3373	50.7267	21.619
0.0002	10.0	60	2.3222	49.2808	20.9524
0.0002	11.0	66	2.3240	50.4615	20.0
0.0008	12.0	72	2.3064	49.0688	20.1429
0.0006	13.0	78	2.2857	47.8241	19.7619
0.0001	14.0	84	2.2707	48.1756	19.8095
0.0001	15.0	90	2.2770	47.4155	20.0
0.0002	16.0	96	2.3248	46.9435	20.5238
0.0001	17.0	102	2.3505	47.3096	20.9048
0.0001	18.0	108	2.3525	48.5449	20.7619
0.0	19.0	114	2.3462	48.5449	20.7619
0.0001	20.0	120	2.3439	48.6822	20.7143
0.0001	21.0	126	2.3570	49.1326	20.5238
0.0	22.0	132	2.3656	48.5247	20.7143
0.0	23.0	138	2.3684	48.5247	20.7143
0.0001	24.0	144	2.3738	49.3527	20.5714
0.0	25.0	150	2.3793	48.2079	20.8571
0.0001	26.0	156	2.3854	47.8381	21.0476
0.0	27.0	162	2.3897	48.0223	21.0476
0.0001	28.0	168	2.3947	47.8029	21.0
0.0	29.0	174	2.3994	48.1359	20.8571
0.0002	30.0	180	2.3992	48.7452	20.8095
0.0001	31.0	186	2.3984	48.0307	20.5714
0.0001	32.0	192	2.3991	48.2877	20.5238
0.0	33.0	198	2.3979	49.5262	20.619
0.0001	34.0	204	2.3998	49.7465	20.6667
0.0001	35.0	210	2.4019	49.5488	20.619
0.0001	36.0	216	2.4056	49.7465	20.6667
0.0001	37.0	222	2.4108	50.1467	20.4762
0.0001	38.0	228	2.4150	50.1467	20.4762
0.0	39.0	234	2.4182	50.707	20.5714
0.0	40.0	240	2.4189	50.503	20.5238
0.0	41.0	246	2.4151	48.2877	20.5238
0.0001	42.0	252	2.4511	48.7331	20.4762
0.0	43.0	258	2.4614	49.1268	20.2857
0.0	44.0	264	2.4134	48.6628	20.381
0.0	45.0	270	2.4117	48.6628	20.381
0.0001	46.0	276	2.4130	48.2745	20.4286
0.0002	47.0	282	2.3939	47.9864	20.4286
0.0	48.0	288	2.3937	48.6253	20.2857
0.0	49.0	294	2.4062	49.3153	20.0476
0.0	50.0	300	2.4131	49.9443	20.0476
0.0001	51.0	306	2.4164	50.9445	20.0476
0.0	52.0	312	2.4129	50.7412	20.1905
0.0001	53.0	318	2.4178	50.693	20.1905
0.0	54.0	324	2.4051	49.2945	20.381
0.0002	55.0	330	2.4062	49.3592	20.381
0.0001	56.0	336	2.3274	49.7531	20.3333
0.0002	57.0	342	2.2969	50.5601	20.2857
0.0	58.0	348	2.2919	50.9648	20.1429
0.0	59.0	354	2.2730	50.1805	20.2381
0.0	60.0	360	2.2660	50.1805	20.2857
0.0001	61.0	366	2.2664	50.1805	20.2857
0.0001	62.0	372	2.2620	49.064	20.1905
0.0001	63.0	378	2.2576	50.8497	20.2857
0.0002	64.0	384	2.2808	50.5765	20.0476
0.0	65.0	390	2.2962	46.8674	20.381
0.0	66.0	396	2.3097	46.4579	20.3333
0.0001	67.0	402	2.3109	50.015	20.0
0.0	68.0	408	2.3189	49.5925	20.0
0.0	69.0	414	2.3080	49.5925	20.0
0.0	70.0	420	2.3065	49.5925	20.0
0.0	71.0	426	2.3102	50.3721	19.9048
0.0	72.0	432	2.3129	50.3721	19.9048
0.0	73.0	438	2.3154	48.9649	19.8571
0.0	74.0	444	2.3178	48.3266	20.0476
0.0001	75.0	450	2.3205	49.9671	20.1905
0.0	76.0	456	2.3218	49.746	20.0952
0.0	77.0	462	2.3216	49.746	20.2381
0.0	78.0	468	2.3218	49.746	20.2381
0.0	79.0	474	2.3174	50.1689	20.0952
0.0	80.0	480	2.3154	50.5016	20.1905
0.0001	81.0	486	2.3215	48.6113	20.2381
0.0001	82.0	492	2.3330	48.6113	20.2381
0.0003	83.0	498	2.3391	48.6113	20.2381
0.0	84.0	504	2.3418	48.4616	20.2857
0.0	85.0	510	2.3408	48.53	20.1429
0.0031	86.0	516	2.3392	48.4848	20.2857
0.0	87.0	522	2.3401	48.4848	20.2857
0.0001	88.0	528	2.3410	48.4848	20.2857
0.0	89.0	534	2.3416	48.8708	20.2381
0.0002	90.0	540	2.3417	48.8244	20.381
0.0	91.0	546	2.3407	48.2659	20.4286
0.0	92.0	552	2.3394	48.2659	20.4286
0.0001	93.0	558	2.3388	48.2659	20.4286
0.0001	94.0	564	2.3386	48.2659	20.4286
0.0	95.0	570	2.3385	48.2659	20.4286
0.0	96.0	576	2.3383	48.2659	20.4286
0.0001	97.0	582	2.3376	48.2659	20.4286
0.0001	98.0	588	2.3371	48.2659	20.4286
0.0	99.0	594	2.3369	48.2659	20.4286
0.0001	100.0	600	2.3369	48.2659	20.4286

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2