my-lora-hf

This model is a fine-tuned version of google/mt5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.2377
• Rouge1: 2.4724
• Rouge2: 0.2498
• Rougel: 2.3079
• Rougelsum: 2.3216

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: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• 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
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
26.282	0.0251	5	17.7306	0.3903	0.0229	0.3967	0.3971
22.8549	0.0503	10	17.6062	0.3903	0.0229	0.3965	0.3968
25.1857	0.0754	15	17.4623	0.3903	0.0229	0.3965	0.3968
25.0655	0.1005	20	17.5023	0.4048	0.0229	0.4088	0.4102
27.1595	0.1256	25	17.4741	0.4151	0.0229	0.4180	0.4215
27.138	0.1508	30	17.4851	0.4378	0.0229	0.4294	0.4323
23.4906	0.1759	35	17.4142	0.3833	0.0229	0.3912	0.3909
22.624	0.2010	40	17.2692	0.3833	0.0229	0.3912	0.3909
24.1221	0.2261	45	17.0949	0.4187	0.0229	0.4134	0.4156
23.4493	0.2513	50	16.9610	0.4187	0.0229	0.4134	0.4156
25.0229	0.2764	55	16.7691	0.4166	0.0229	0.4113	0.4134
24.346	0.3015	60	16.6668	0.4791	0.0229	0.4539	0.4536
23.1975	0.3266	65	16.4952	0.4151	0.0229	0.4078	0.4102
21.6027	0.3518	70	16.3300	0.4178	0.0229	0.4109	0.4122
24.1149	0.3769	75	16.2625	0.4294	0.0229	0.4270	0.4268
21.8442	0.4020	80	16.1807	0.4169	0.0229	0.4153	0.4119
23.3859	0.4271	85	16.1351	0.4169	0.0229	0.4153	0.4119
20.8352	0.4523	90	16.0392	0.3939	0.0229	0.3906	0.3921
23.6936	0.4774	95	16.0173	0.3721	0.0229	0.3826	0.3810
24.6121	0.5025	100	15.9600	0.3989	0.0229	0.4062	0.4070
24.6205	0.5276	105	15.8700	0.3832	0.0229	0.3946	0.3926
22.8886	0.5528	110	15.7120	0.3752	0.0229	0.3802	0.3868
21.4168	0.5779	115	15.5049	0.4073	0.0229	0.4110	0.4150
20.5502	0.6030	120	15.2778	0.4786	0.0493	0.4807	0.4818
22.1162	0.6281	125	15.0846	0.4644	0.0493	0.4629	0.4681
19.9664	0.6533	130	14.8895	0.4644	0.0493	0.4629	0.4681
21.8179	0.6784	135	14.6839	0.4532	0.0493	0.4506	0.4568
19.7418	0.7035	140	14.5265	0.4401	0.0619	0.4375	0.4440
18.8154	0.7286	145	14.3868	0.4546	0.0619	0.4497	0.4562
23.85	0.7538	150	14.2192	0.4277	0.0619	0.4229	0.4321
19.3083	0.7789	155	14.0616	0.3760	0.0619	0.3745	0.3790
17.8149	0.8040	160	13.9289	0.3604	0.0619	0.3605	0.3670
19.6341	0.8291	165	13.7362	0.3604	0.0619	0.3605	0.3670
20.2189	0.8543	170	13.5559	0.3448	0.0619	0.3452	0.3500
17.6897	0.8794	175	13.3765	0.3328	0.0619	0.3343	0.3396
19.1985	0.9045	180	13.1860	0.3192	0.0619	0.3187	0.3232
22.0973	0.9296	185	12.9943	0.3484	0.0756	0.3435	0.3480
18.2808	0.9548	190	12.8423	0.3484	0.0756	0.3435	0.3480
19.4271	0.9799	195	12.6962	0.3621	0.0756	0.3565	0.3605
16.4328	1.0050	200	12.5391	0.3621	0.0756	0.3565	0.3605
17.2057	1.0302	205	12.3865	0.3621	0.0756	0.3565	0.3605
17.0921	1.0553	210	12.2590	0.3348	0.0489	0.3283	0.3347
18.4743	1.0804	215	12.1398	0.3348	0.0489	0.3283	0.3347
17.084	1.1055	220	11.9490	0.3348	0.0489	0.3283	0.3347
15.3983	1.1307	225	11.7623	0.3201	0.0489	0.3152	0.3216
17.3434	1.1558	230	11.6081	0.3462	0.0489	0.3405	0.3469
14.5055	1.1809	235	11.4089	0.3462	0.0489	0.3405	0.3469
14.5421	1.2060	240	11.1806	0.3455	0.0489	0.3393	0.3465
15.4771	1.2312	245	10.9563	0.3697	0.0489	0.3515	0.3609
14.7773	1.2563	250	10.7472	0.3535	0.0359	0.3390	0.3463
12.5897	1.2814	255	10.5722	0.3758	0.0359	0.3607	0.3718
13.4934	1.3065	260	10.3998	0.4165	0.0359	0.3984	0.4062
14.3636	1.3317	265	10.2015	0.5002	0.0789	0.4602	0.4769
14.8955	1.3568	270	9.9741	0.5282	0.0679	0.4753	0.4912
13.1539	1.3819	275	9.7512	0.5364	0.0506	0.4924	0.5048
13.2489	1.4070	280	9.5085	0.6261	0.0835	0.5834	0.5925
13.4348	1.4322	285	9.2728	0.6255	0.0708	0.5829	0.5920
14.5854	1.4573	290	9.0892	0.6479	0.0815	0.5951	0.6041
12.7121	1.4824	295	8.9407	0.6731	0.0794	0.6209	0.6339
11.9989	1.5075	300	8.7771	0.6901	0.0824	0.6349	0.6466
12.7228	1.5327	305	8.6534	0.8476	0.1285	0.7741	0.7915
13.1025	1.5578	310	8.5266	0.8674	0.1285	0.8033	0.8169
12.4569	1.5829	315	8.4097	0.8674	0.1285	0.8033	0.8169
12.8088	1.6080	320	8.3092	0.8588	0.1285	0.7927	0.8091
12.5782	1.6332	325	8.2020	1.0401	0.1704	0.9733	0.9710
12.2901	1.6583	330	8.1193	1.0807	0.1834	1.0074	1.0059
11.1978	1.6834	335	8.0178	1.2109	0.2315	1.1353	1.1408
11.2324	1.7085	340	7.9492	1.1993	0.2314	1.1099	1.1150
11.2734	1.7337	345	7.8789	1.2551	0.2314	1.1703	1.1775
11.3287	1.7588	350	7.8109	1.3333	0.1776	1.2429	1.2507
10.8443	1.7839	355	7.7369	1.3289	0.1893	1.2214	1.2134
11.3227	1.8090	360	7.6704	1.4954	0.2192	1.3839	1.3687
11.3922	1.8342	365	7.6183	1.5980	0.2288	1.4809	1.4772
10.3799	1.8593	370	7.5631	1.5997	0.2280	1.5053	1.5004
9.838	1.8844	375	7.4964	1.5374	0.1744	1.4412	1.4385
10.5613	1.9095	380	7.4164	1.6077	0.1635	1.4999	1.5129
10.054	1.9347	385	7.3279	1.5019	0.1376	1.3888	1.3910
11.1505	1.9598	390	7.2568	1.5497	0.1376	1.4428	1.4585
9.6994	1.9849	395	7.1809	1.7434	0.2045	1.6128	1.6446
9.8363	2.0101	400	7.0994	1.8268	0.2398	1.6494	1.6631
9.8672	2.0352	405	7.0202	1.9286	0.2395	1.7153	1.7304
9.9883	2.0603	410	6.9526	1.9111	0.2628	1.6911	1.7111
9.4994	2.0854	415	6.8914	1.8796	0.2374	1.6792	1.6873
9.7394	2.1106	420	6.8379	1.8351	0.2244	1.6821	1.6861
9.8364	2.1357	425	6.8016	1.9244	0.2245	1.7330	1.7383
9.1858	2.1608	430	6.7747	1.9020	0.1622	1.7451	1.7457
9.9149	2.1859	435	6.7477	1.9496	0.1622	1.7809	1.7830
10.0384	2.2111	440	6.7207	1.9289	0.1622	1.7660	1.7701
9.3528	2.2362	445	6.6923	1.9978	0.1707	1.8218	1.8346
9.4208	2.2613	450	6.6668	2.0390	0.1871	1.8828	1.8851
9.1942	2.2864	455	6.6424	2.0593	0.2087	1.8986	1.9001
9.3805	2.3116	460	6.6150	2.0838	0.2085	1.9257	1.9294
9.0846	2.3367	465	6.5857	2.0278	0.2085	1.8741	1.8760
8.9866	2.3618	470	6.5632	2.0961	0.2085	1.9329	1.9317
9.0288	2.3869	475	6.5425	2.2513	0.2827	2.0651	2.0584
9.1447	2.4121	480	6.5246	2.3172	0.2967	2.1624	2.1467
8.549	2.4372	485	6.5078	2.2586	0.2967	2.1040	2.1027
9.0221	2.4623	490	6.4919	2.2060	0.2873	2.0706	2.0761
9.0646	2.4874	495	6.4775	2.1466	0.2834	2.0164	2.0235
9.3397	2.5126	500	6.4652	2.1198	0.2834	1.9919	1.9934
8.8204	2.5377	505	6.4537	2.0995	0.2834	1.9680	1.9717
9.139	2.5628	510	6.4415	2.2589	0.2843	2.0866	2.0803
9.0461	2.5879	515	6.4270	2.2693	0.2842	2.0854	2.0796
8.8862	2.6131	520	6.4161	2.2462	0.2714	2.0903	2.0910
8.8184	2.6382	525	6.4065	2.2489	0.2473	2.1277	2.1201
8.3526	2.6633	530	6.3943	2.2481	0.2473	2.1141	2.1084
8.8545	2.6884	535	6.3833	2.3059	0.2610	2.1646	2.1505
8.3531	2.7136	540	6.3753	2.2896	0.2471	2.1653	2.1504
9.0157	2.7387	545	6.3682	2.4845	0.2828	2.3504	2.3363
9.3887	2.7638	550	6.3641	2.4634	0.2833	2.2818	2.2731
8.2536	2.7889	555	6.3593	2.4877	0.2835	2.3238	2.3163
8.073	2.8141	560	6.3517	2.4887	0.2835	2.3240	2.3169
8.2218	2.8392	565	6.3438	2.4546	0.2835	2.3161	2.3041
8.6336	2.8643	570	6.3358	2.4307	0.2841	2.2969	2.2978
8.3906	2.8894	575	6.3305	2.4145	0.2740	2.2496	2.2409
9.0623	2.9146	580	6.3268	2.4416	0.2740	2.2672	2.2637
8.5729	2.9397	585	6.3220	2.4337	0.2745	2.2573	2.2504
8.7356	2.9648	590	6.3184	2.4469	0.2744	2.2564	2.2499
8.1848	2.9899	595	6.3135	2.4153	0.2745	2.2543	2.2510
8.5797	3.0151	600	6.3090	2.3851	0.2745	2.2282	2.2187
8.1148	3.0402	605	6.3054	2.3860	0.2745	2.2292	2.2193
8.5009	3.0653	610	6.3016	2.2895	0.2356	2.1353	2.1412
8.3814	3.0905	615	6.2981	2.3022	0.2353	2.1486	2.1576
8.4252	3.1156	620	6.2963	2.2869	0.2240	2.1604	2.1672
8.4299	3.1407	625	6.2938	2.3894	0.2375	2.2476	2.2677
8.4384	3.1658	630	6.2906	2.3887	0.2375	2.2394	2.2525
8.3078	3.1910	635	6.2874	2.3602	0.2364	2.2151	2.2316
8.2374	3.2161	640	6.2837	2.3519	0.2479	2.2049	2.2205
8.1229	3.2412	645	6.2812	2.3440	0.2479	2.1831	2.2014
7.971	3.2663	650	6.2783	2.3851	0.2479	2.2170	2.2366
8.2733	3.2915	655	6.2748	2.4371	0.2479	2.2770	2.2937
8.2847	3.3166	660	6.2723	2.4377	0.2479	2.2774	2.2940
8.3704	3.3417	665	6.2696	2.4388	0.2488	2.2772	2.2954
8.4976	3.3668	670	6.2681	2.4493	0.2487	2.2942	2.3085
8.4326	3.3920	675	6.2672	2.4643	0.2487	2.3066	2.3156
8.3527	3.4171	680	6.2659	2.4643	0.2487	2.3066	2.3156
8.1029	3.4422	685	6.2630	2.4478	0.2487	2.2890	2.3013
8.3205	3.4673	690	6.2603	2.4514	0.2495	2.2752	2.2896
8.1628	3.4925	695	6.2583	2.4654	0.2495	2.2752	2.2896
7.862	3.5176	700	6.2558	2.4546	0.2497	2.2637	2.2739
8.1908	3.5427	705	6.2533	2.4643	0.2498	2.2769	2.2865
8.084	3.5678	710	6.2516	2.4639	0.2498	2.2766	2.2865
8.0681	3.5930	715	6.2501	2.4996	0.2498	2.3104	2.3203
8.3121	3.6181	720	6.2483	2.4996	0.2498	2.3104	2.3203
8.3079	3.6432	725	6.2469	2.4996	0.2498	2.3097	2.3203
8.0919	3.6683	730	6.2457	2.5122	0.2498	2.3267	2.3343
7.9226	3.6935	735	6.2447	2.5122	0.2498	2.3267	2.3343
8.2032	3.7186	740	6.2440	2.5122	0.2498	2.3267	2.3343
9.1495	3.7437	745	6.2433	2.5122	0.2498	2.3267	2.3343
8.1283	3.7688	750	6.2424	2.5122	0.2498	2.3267	2.3343
8.0652	3.7940	755	6.2416	2.4724	0.2498	2.3079	2.3216
8.5275	3.8191	760	6.2409	2.4724	0.2498	2.3079	2.3216
7.9583	3.8442	765	6.2402	2.4724	0.2498	2.3079	2.3216
7.9768	3.8693	770	6.2394	2.4724	0.2498	2.3079	2.3216
7.96	3.8945	775	6.2390	2.4724	0.2498	2.3079	2.3216
7.7373	3.9196	780	6.2385	2.4724	0.2498	2.3079	2.3216
8.3979	3.9447	785	6.2381	2.4724	0.2498	2.3079	2.3216
7.7723	3.9698	790	6.2378	2.4724	0.2498	2.3079	2.3216
8.2497	3.9950	795	6.2377	2.4724	0.2498	2.3079	2.3216

Framework versions

• PEFT 0.14.0
• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0