Terjman-Supreme-v2.2

This model is a fine-tuned version of facebook/nllb-200-3.3B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.2058
• Bleu: 23.4245
• Chrf: 44.5703
• Ter: 78.1656
• Gen Len: 12.3071

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: 0.0005
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 64
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Chrf	Ter	Gen Len
137.8105	0.0181	100	3.2305	5.2001	13.6145	118.7357	12.5635
97.2318	0.0361	200	2.5726	17.2165	35.9826	88.2951	12.4082
83.4952	0.0542	300	2.4521	17.3135	36.8477	87.0818	12.2729
75.0776	0.0723	400	2.3968	18.058	37.798	87.0603	12.4929
66.9757	0.0904	500	2.3446	20.1995	38.8098	84.8913	12.3459
63.6814	0.1084	600	2.2928	19.8012	38.2663	84.9302	12.3835
60.3928	0.1265	700	2.2272	19.4166	38.2539	84.0673	12.3035
56.197	0.1446	800	2.2304	20.3862	39.7791	86.3549	12.9435
55.4461	0.1627	900	2.2529	19.6843	38.5102	84.7924	12.4235
54.5164	0.1807	1000	2.1330	21.2121	40.5161	81.4403	12.2235
53.228	0.1988	1100	2.1817	20.4483	40.5389	82.9359	12.5094
51.4801	0.2169	1200	2.2008	21.6519	40.7718	82.0789	12.3435
51.5955	0.2350	1300	2.1723	19.8541	40.5091	83.7385	12.5271
51.072	0.2530	1400	2.2120	20.0968	39.0978	95.2042	12.7388
51.3383	0.2711	1500	2.2069	21.9455	41.6071	113.2387	13.4894
51.1281	0.2892	1600	2.2140	19.6691	38.7329	107.2516	12.2718
50.6391	0.3072	1700	2.1903	20.9862	40.6458	87.1593	12.6706
50.8716	0.3253	1800	2.1774	20.9512	40.8041	83.8517	12.5753
49.1451	0.3434	1900	2.1281	21.7404	41.1935	81.2223	12.3718
48.6002	0.3615	2000	2.1656	21.3734	40.6663	81.527	12.2447
46.7725	0.3795	2100	2.1560	21.3878	40.2352	82.5294	12.4776
46.1724	0.3976	2200	2.1248	21.6288	40.8181	82.571	12.6271
45.3777	0.4157	2300	2.1343	21.8915	41.1995	82.9489	12.5776
44.7859	0.4338	2400	2.0903	22.322	41.9163	80.283	12.3718
43.8804	0.4518	2500	2.1051	22.3484	41.6646	80.6631	12.4106
42.9552	0.4699	2600	2.1127	21.3378	40.9099	81.9978	12.4812
42.8838	0.4880	2700	2.0602	20.8244	41.1047	82.7353	12.44
41.5548	0.5061	2800	2.0623	22.2229	41.5223	87.224	12.6024
41.6697	0.5241	2900	2.0682	23.1842	42.8083	81.7717	12.4788
41.5065	0.5422	3000	2.0393	22.3557	41.6164	79.3085	12.3718
41.6211	0.5603	3100	2.0054	22.2312	41.8033	80.2008	12.2647
41.248	0.5783	3200	2.0231	20.8397	41.33	95.135	13.7929
39.1609	0.5964	3300	2.0002	22.7687	43.2316	81.6566	12.4506
38.9398	0.6145	3400	2.0289	21.7436	41.1019	82.7019	12.4176
38.3047	0.6326	3500	2.0109	22.8019	41.9596	94.4557	12.8129
38.268	0.6506	3600	2.0201	24.4653	44.0864	80.5792	12.6435
38.5439	0.6687	3700	2.0134	24.1665	43.2041	78.6761	12.3188
37.7287	0.6868	3800	2.0260	24.3768	43.6489	81.3409	12.7047
36.1784	0.7049	3900	1.9660	23.9311	43.3869	79.2676	12.4035
36.6123	0.7229	4000	1.9625	22.976	43.4062	79.7419	12.3976
36.7558	0.7410	4100	1.9621	22.9323	43.1629	80.1544	12.3635
36.6426	0.7591	4200	1.9682	23.0305	43.0949	79.4987	12.32
36.3618	0.7772	4300	1.9667	22.5987	42.8116	79.1677	12.2047
36.0427	0.7952	4400	1.9333	23.7835	42.8368	78.3063	12.2541
35.1977	0.8133	4500	1.9455	24.1571	43.0898	78.0521	12.24
34.9866	0.8314	4600	1.9282	23.5239	42.3893	78.3871	12.1388
34.66	0.8494	4700	1.9404	23.9171	43.2115	79.4609	12.3435
34.85	0.8675	4800	1.9518	23.698	42.9408	79.1596	12.2612
34.9682	0.8856	4900	1.8821	23.4893	43.8764	78.5991	12.3012
33.784	0.9037	5000	1.8985	23.6473	43.8604	78.1807	12.2612
33.8055	0.9217	5100	1.9105	23.6402	43.5989	78.5758	12.3553
33.9263	0.9398	5200	1.8903	24.5462	44.4274	78.4294	12.4588
34.1902	0.9579	5300	1.8965	23.8897	43.9914	78.6182	12.4647
33.1319	0.9760	5400	1.8967	23.4588	43.4492	77.2763	12.3094
32.6819	0.9940	5500	1.8980	23.1381	43.7181	79.4987	12.5306
25.6872	1.0121	5600	1.9989	24.3884	43.8866	76.2215	12.2565
26.2497	1.0302	5700	2.0097	23.0962	42.8724	78.4643	12.2953
25.0397	1.0483	5800	2.0168	23.3407	43.1597	78.6203	12.2941
26.1732	1.0663	5900	2.0115	23.4883	43.5452	79.9678	12.6753
25.9725	1.0844	6000	2.0090	23.1359	43.4672	79.7718	12.6659
25.6574	1.1025	6100	2.0034	22.8505	43.1392	80.5169	12.3859
25.5012	1.1205	6200	2.0140	23.7912	43.4571	77.5566	12.2776
25.3003	1.1386	6300	2.0215	23.9161	43.6086	78.0778	12.34
24.6921	1.1567	6400	1.9851	23.8938	44.0363	79.4653	12.4859
25.0079	1.1748	6500	1.9878	23.8998	43.4735	78.9155	12.3506
25.0747	1.1928	6600	1.9882	22.9832	43.0514	79.9574	12.36
25.2562	1.2109	6700	2.0077	22.6288	43.0805	80.337	12.3835
25.1187	1.2290	6800	1.9651	23.4281	43.7022	78.7561	12.3518
24.0937	1.2471	6900	1.9807	23.6022	43.5876	79.3789	12.2953
24.9507	1.2651	7000	1.9347	23.6003	43.4728	78.0933	12.2659
24.5099	1.2832	7100	1.9686	23.7821	43.7423	78.1923	12.3659
24.3604	1.3013	7200	2.0012	23.5141	43.7252	78.5037	12.3541
24.2341	1.3194	7300	1.9883	23.6762	43.7825	77.999	12.3082
23.8904	1.3374	7400	1.9842	23.9927	43.6138	77.48	12.3094
24.7891	1.3555	7500	1.9782	23.4309	43.695	78.4042	12.2541
24.3329	1.3736	7600	1.9914	23.3102	43.8064	79.1631	12.3871
24.4164	1.3917	7700	1.9766	23.5369	43.4989	77.9728	12.2565
24.1632	1.4097	7800	1.9629	23.5813	43.9428	78.335	12.2624
24.3018	1.4278	7900	1.9478	24.2417	44.5599	77.6036	12.2635
23.5833	1.4459	8000	1.9401	24.0363	44.3674	77.6119	12.2388
24.0873	1.4639	8100	1.9250	24.0754	44.316	77.1101	12.2
23.8089	1.4820	8200	1.9356	24.4306	45.0766	77.1962	12.3188
23.6771	1.5001	8300	1.9320	23.9639	44.5074	77.3474	12.2435
23.9428	1.5182	8400	1.9510	23.3543	44.2183	78.931	12.3071
24.1568	1.5362	8500	1.9518	23.6985	44.5651	78.9739	12.3647
23.7392	1.5543	8600	1.9338	24.3444	45.0775	77.3414	12.3259
23.8515	1.5724	8700	1.9474	24.1262	44.792	77.9217	12.3071
23.0664	1.5905	8800	1.9473	24.0164	44.7487	78.0097	12.3224
23.3824	1.6085	8900	1.9340	23.806	44.6919	77.9177	12.2953
22.7727	1.6266	9000	1.9716	23.4272	44.7997	78.3186	12.3212
23.4696	1.6447	9100	1.9493	23.7553	44.5676	78.0819	12.3424
22.7403	1.6628	9200	1.9668	24.0244	44.7161	77.4703	12.3459
22.8108	1.6808	9300	1.9469	24.3704	45.1849	77.0614	12.2741
23.8398	1.6989	9400	1.9339	23.8243	44.5131	77.6411	12.2659
22.6805	1.7170	9500	1.9319	24.1688	44.8033	77.717	12.2765
22.7274	1.7350	9600	1.9525	23.845	44.87	77.9104	12.2859
22.9588	1.7531	9700	1.9431	24.1488	44.7974	78.0564	12.2918
22.9954	1.7712	9800	1.9482	23.8619	45.008	77.9592	12.3271
22.6917	1.7893	9900	1.9385	24.1254	45.0475	77.1275	12.3012
22.7402	1.8073	10000	1.9290	24.552	45.4205	76.0642	12.2541
23.087	1.8254	10100	1.9144	24.2509	45.0886	77.0239	12.3071
23.0166	1.8435	10200	1.9263	24.2323	45.0173	77.1054	12.3071
22.3922	1.8616	10300	1.9365	23.3072	44.4322	78.3998	12.3024
22.9795	1.8796	10400	1.9352	23.2202	44.4076	79.067	12.3212
22.3474	1.8977	10500	1.9440	22.9263	44.2298	79.3766	12.3412
22.1394	1.9158	10600	1.9303	22.9458	44.2676	78.7353	12.3153
22.9308	1.9339	10700	1.9329	23.2577	44.4509	78.2277	12.3329
22.2057	1.9519	10800	1.9297	23.0115	44.0851	79.0073	12.3247
22.8086	1.9700	10900	1.9351	23.083	44.244	79.5701	12.3565
22.4432	1.9881	11000	1.9423	23.3255	44.3864	78.756	12.3188
19.0883	2.0061	11100	2.0378	23.0833	44.3251	78.859	12.3118
17.6663	2.0242	11200	2.1177	23.1699	44.2894	78.8729	12.2965
18.1139	2.0423	11300	2.1363	22.8201	44.1824	79.4578	12.3035
17.3899	2.0604	11400	2.1522	23.1964	44.4594	78.4838	12.3282
17.4895	2.0784	11500	2.1518	23.0312	44.4584	78.8593	12.2659
17.5904	2.0965	11600	2.1763	22.7354	44.1317	78.8398	12.2941
17.2585	2.1146	11700	2.1681	22.9018	44.1018	84.3444	12.5188
17.7048	2.1327	11800	2.1767	22.9439	44.2815	84.3079	12.5212
17.3691	2.1507	11900	2.1812	22.8058	44.0874	84.2541	12.5176
17.4433	2.1688	12000	2.1851	22.7803	44.1106	84.8505	12.5318
17.4181	2.1869	12100	2.1836	22.9985	44.115	84.5472	12.5353
17.7475	2.2050	12200	2.1929	23.1362	44.2473	84.1141	12.4976
17.3206	2.2230	12300	2.1860	23.1621	44.3407	78.4548	12.2988
17.5009	2.2411	12400	2.1877	22.9871	44.3105	84.2672	12.5388
17.3298	2.2592	12500	2.1906	23.0868	44.4643	84.3887	12.5494
17.3273	2.2772	12600	2.1933	23.0713	44.5811	78.539	12.3224
17.7982	2.2953	12700	2.2007	23.0106	44.3926	84.531	12.5471
17.4851	2.3134	12800	2.1865	23.1507	44.5556	78.5757	12.3259
17.526	2.3315	12900	2.1850	23.1178	44.548	78.4484	12.2941
17.3512	2.3495	13000	2.2085	23.345	44.7801	78.1779	12.3129
16.9256	2.3676	13100	2.2143	23.5069	44.7383	78.0014	12.3094
17.5127	2.3857	13200	2.2020	23.5399	44.7461	77.9207	12.3035
16.9278	2.4038	13300	2.2075	23.3908	44.6205	78.2937	12.3365
16.9606	2.4218	13400	2.2032	23.4848	44.7578	77.8764	12.3271
17.1626	2.4399	13500	2.2008	23.2227	44.2244	78.361	12.3282
17.0724	2.4580	13600	2.2089	23.4288	44.5322	78.1828	12.2976
17.3127	2.4761	13700	2.1986	23.5571	44.7397	77.867	12.3071
17.5657	2.4941	13800	2.2048	23.6261	44.9271	78.1384	12.3118
17.2787	2.5122	13900	2.2031	23.3188	44.5084	78.1988	12.3047
17.4854	2.5303	14000	2.2063	23.2519	44.546	78.171	12.3094
17.2063	2.5483	14100	2.2056	23.1215	44.5185	78.1912	12.3035
17.3853	2.5664	14200	2.2023	23.307	44.7458	78.0718	12.2929
17.774	2.5845	14300	2.1994	23.3031	44.8097	77.9967	12.3059
16.6834	2.6026	14400	2.2036	23.4314	44.7324	77.9656	12.3082
17.1441	2.6206	14500	2.2020	23.3211	44.8156	77.9575	12.3141
17.6746	2.6387	14600	2.2064	23.2103	44.6337	78.2396	12.3294
17.0056	2.6568	14700	2.2077	23.2063	44.6551	77.9941	12.2918
17.1054	2.6749	14800	2.2078	23.4498	44.7412	78.0114	12.2976
17.5818	2.6929	14900	2.2031	23.3876	44.7344	77.9263	12.2894
17.2745	2.7110	15000	2.2031	23.2746	44.6287	78.1371	12.2929
17.2902	2.7291	15100	2.2045	23.3318	44.731	78.1528	12.3024
17.3091	2.7472	15200	2.2080	23.339	44.6857	78.186	12.3059
17.405	2.7652	15300	2.2060	23.3269	44.6893	78.1277	12.2906
17.5894	2.7833	15400	2.2071	23.2891	44.7165	78.1082	12.2965
17.0158	2.8014	15500	2.2038	23.4179	44.7075	78.0389	12.2847
17.4972	2.8194	15600	2.2068	23.3515	44.7074	78.1332	12.2941
17.2833	2.8375	15700	2.2061	23.3688	44.7259	78.0318	12.3
17.0573	2.8556	15800	2.2068	23.4313	44.7199	77.9653	12.2976
17.4166	2.8737	15900	2.2056	23.4953	44.6781	78.0423	12.3106
17.0984	2.8917	16000	2.2057	23.3163	44.6874	78.0286	12.2906
17.5987	2.9098	16100	2.2047	23.3148	44.6769	78.0154	12.3
17.2003	2.9279	16200	2.2053	23.3496	44.6994	77.9406	12.2988
17.1631	2.9460	16300	2.2049	23.2588	44.619	77.9645	12.2941
17.3428	2.9640	16400	2.2063	23.2298	44.5834	78.0989	12.2965
17.3013	2.9821	16500	2.2058	23.4245	44.5703	78.1656	12.3071

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.21.0