stablelm-3b-finance

This model is a fine-tuned version of stabilityai/stablelm-base-alpha-3b on the financial_phrasebank dataset. It achieves the following results on the evaluation set:

• Loss: 4.2656
• Accuracy: 0.4081

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: 0.0002
• 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: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
37.5127	0.01	20	19.6094	0.2624
19.1885	0.01	40	64.3125	0.0816
14.5964	0.02	60	43.4688	0.4143
5.8184	0.02	80	54.9688	0.4143
25.0629	0.03	100	62.25	0.0816
23.0213	0.03	120	18.4062	0.4143
6.4438	0.04	140	9.4922	0.4143
10.3302	0.04	160	25.0781	0.4143
5.5922	0.05	180	29.0469	0.4143
2.9618	0.05	200	14.5078	0.4143
4.551	0.06	220	18.5156	0.4112
4.5168	0.06	240	29.5156	0.4143
3.0656	0.07	260	27.5469	0.4143
6.9075	0.07	280	40.375	0.4143
6.09	0.08	300	28.4688	0.4143
2.254	0.08	320	35.375	0.4143
8.4998	0.09	340	31.4375	0.4143
8.3815	0.09	360	28.2188	0.4143
7.9155	0.1	380	12.0625	0.0816
2.0166	0.1	400	9.8672	0.3895
5.1889	0.11	420	11.0234	0.4132
3.844	0.11	440	10.4844	0.4132
3.7982	0.12	460	13.4688	0.3864
1.5386	0.12	480	34.3125	0.4143
4.0882	0.13	500	33.875	0.4143
6.1486	0.13	520	11.6406	0.0816
4.0523	0.14	540	9.1328	0.3988
2.5312	0.14	560	11.7969	0.4143
6.0803	0.15	580	15.6016	0.4143
5.4927	0.15	600	8.8828	0.0816
3.9641	0.16	620	9.3672	0.4132
4.2708	0.17	640	4.7852	0.4143
4.4524	0.17	660	6.5312	0.4122
6.081	0.18	680	34.3125	0.4112
7.9303	0.18	700	45.9375	0.0868
1.9309	0.19	720	44.6562	0.4143
10.341	0.19	740	20.5312	0.4143
6.2898	0.2	760	9.4219	0.4143
3.6457	0.2	780	7.6758	0.4143
4.6135	0.21	800	9.6875	0.4143
3.2253	0.21	820	9.5078	0.0888
3.7653	0.22	840	7.4453	0.1178
5.2183	0.22	860	10.2422	0.0837
7.6287	0.23	880	4.2344	0.4143
2.8624	0.23	900	4.2227	0.4143
5.1134	0.24	920	7.375	0.1674
7.5312	0.24	940	3.6641	0.0888
2.8719	0.25	960	7.7148	0.4143
1.5177	0.25	980	6.3320	0.4143
3.0631	0.26	1000	2.9297	0.4143
5.4135	0.26	1020	1.9219	0.4143
4.9254	0.27	1040	17.8281	0.4143
1.6855	0.27	1060	17.9844	0.4143
3.595	0.28	1080	13.2422	0.4143
4.9504	0.28	1100	3.6074	0.4143
1.5584	0.29	1120	13.0469	0.4143
5.144	0.29	1140	17.8594	0.0816
4.8497	0.3	1160	11.6953	0.4143
3.4161	0.3	1180	20.6562	0.0816
6.7763	0.31	1200	15.1406	0.4143
2.8709	0.32	1220	8.9922	0.4143
4.1128	0.32	1240	9.3984	0.4143
1.4908	0.33	1260	12.2891	0.4143
3.2886	0.33	1280	14.5703	0.0816
6.8233	0.34	1300	2.2734	0.4143
2.6123	0.34	1320	7.3281	0.4143
3.4882	0.35	1340	4.6992	0.0826
2.5896	0.35	1360	11.9688	0.4143
6.352	0.36	1380	4.4727	0.4101
2.2885	0.36	1400	5.7578	0.4050
4.2021	0.37	1420	7.1172	0.4143
3.032	0.37	1440	2.7988	0.4132
1.428	0.38	1460	17.5938	0.4143
2.8068	0.38	1480	15.9141	0.4143
7.0014	0.39	1500	4.9883	0.1033
3.2138	0.39	1520	9.625	0.4112
3.0001	0.4	1540	6.2305	0.3833
2.3248	0.4	1560	5.5547	0.3760
2.7573	0.41	1580	9.0859	0.4143
4.8701	0.41	1600	9.1641	0.25
6.5986	0.42	1620	6.5977	0.4143
5.4379	0.42	1640	8.8203	0.0826
4.7142	0.43	1660	6.9219	0.4143
3.1696	0.43	1680	3.9941	0.4163
2.5254	0.44	1700	5.7461	0.4143
1.9537	0.44	1720	3.7441	0.4143
2.4895	0.45	1740	6.8789	0.4143
2.9386	0.45	1760	6.5625	0.4205
4.1816	0.46	1780	2.7070	0.4163
5.2298	0.46	1800	6.3828	0.3967
1.3144	0.47	1820	8.3984	0.4143
2.62	0.48	1840	7.8359	0.4153
2.3815	0.48	1860	6.9297	0.3595
1.2381	0.49	1880	7.125	0.3595
1.6726	0.49	1900	8.7109	0.4132
2.0483	0.5	1920	8.3047	0.4101
3.6178	0.5	1940	8.1719	0.3574
3.0994	0.51	1960	11.5625	0.2572
2.0991	0.51	1980	15.375	0.0816
4.4138	0.52	2000	16.6406	0.4143
4.6666	0.52	2020	3.7168	0.4050
1.8319	0.53	2040	5.375	0.4112
2.2184	0.53	2060	5.3398	0.3337
2.5527	0.54	2080	4.2188	0.4112
2.2984	0.54	2100	6.875	0.4122
2.2836	0.55	2120	6.1836	0.4101
3.0373	0.55	2140	5.7930	0.4019
2.2946	0.56	2160	4.7891	0.4091
1.4506	0.56	2180	13.2891	0.3936
1.7211	0.57	2200	11.1172	0.4112
2.4151	0.57	2220	8.2969	0.3905
1.989	0.58	2240	8.2812	0.4143
0.6184	0.58	2260	8.1562	0.3936
2.8995	0.59	2280	9.8203	0.4174
1.8413	0.59	2300	8.0781	0.3626
1.6959	0.6	2320	8.5703	0.3233
2.7333	0.6	2340	4.8984	0.4163
1.7187	0.61	2360	6.5859	0.3957
2.8341	0.61	2380	5.6133	0.3926
1.8751	0.62	2400	5.1445	0.3399
2.0391	0.62	2420	5.3633	0.4163
1.1569	0.63	2440	6.4570	0.2975
1.8955	0.64	2460	4.6367	0.4153
3.3986	0.64	2480	4.875	0.4153
0.9939	0.65	2500	4.4492	0.4184
3.1304	0.65	2520	3.9414	0.4081
1.7888	0.66	2540	5.5898	0.4081
2.1101	0.66	2560	6.3242	0.3017
1.7795	0.67	2580	6.1992	0.3957
0.565	0.67	2600	7.1016	0.4205
2.1791	0.68	2620	4.4805	0.4215
1.8351	0.68	2640	6.9727	0.3667
2.0573	0.69	2660	7.6797	0.4174
2.077	0.69	2680	4.3242	0.4029
1.3436	0.7	2700	5.7227	0.4153
2.5434	0.7	2720	5.6836	0.3492
1.4306	0.71	2740	4.4844	0.4122
1.8493	0.71	2760	4.3984	0.4070
1.2447	0.72	2780	4.3242	0.3884
3.9	0.72	2800	2.9824	0.4091
2.8007	0.73	2820	3.8008	0.4174
0.7397	0.73	2840	6.1953	0.4153
1.6954	0.74	2860	3.8711	0.4122
1.0559	0.74	2880	6.7578	0.4184
2.354	0.75	2900	4.0195	0.3853
2.2396	0.75	2920	4.1836	0.4143
1.736	0.76	2940	6.1523	0.3657
1.7746	0.76	2960	7.6797	0.4112
2.8225	0.77	2980	7.1172	0.4153
1.723	0.77	3000	4.4570	0.4143
1.591	0.78	3020	4.7188	0.4060
2.2972	0.78	3040	5.4219	0.3461
0.7102	0.79	3060	6.3477	0.4205
2.6331	0.8	3080	5.7031	0.4039
2.2154	0.8	3100	5.9023	0.3223
1.3288	0.81	3120	5.6289	0.4101
2.8541	0.81	3140	4.5391	0.4081
1.1956	0.82	3160	4.2188	0.3326
0.6724	0.82	3180	6.4180	0.4194
3.3572	0.83	3200	5.7930	0.4070
2.1121	0.83	3220	4.6758	0.3771
2.0453	0.84	3240	5.3281	0.4112
1.1715	0.84	3260	6.9141	0.3812
1.0438	0.85	3280	7.3242	0.4143
0.7894	0.85	3300	7.7422	0.4184
3.3288	0.86	3320	6.8945	0.4039
1.9406	0.86	3340	5.5547	0.4143
2.5013	0.87	3360	4.0078	0.4132
1.3637	0.87	3380	4.2031	0.4070
2.9497	0.88	3400	3.9727	0.4091
1.6295	0.88	3420	4.9805	0.4132
1.6557	0.89	3440	5.0586	0.3864
1.0884	0.89	3460	5.2070	0.3977
0.3464	0.9	3480	5.4609	0.4101
1.7069	0.9	3500	5.7188	0.4081
1.9864	0.91	3520	5.8164	0.4112
1.7181	0.91	3540	5.3047	0.4091
3.4296	0.92	3560	4.9375	0.4081
1.1618	0.92	3580	4.8320	0.3905
1.5314	0.93	3600	4.9531	0.3874
2.2965	0.93	3620	4.5625	0.3988
1.7454	0.94	3640	4.4102	0.4081
0.938	0.95	3660	4.3945	0.4060
1.9902	0.95	3680	4.4844	0.4070
0.7222	0.96	3700	4.6758	0.4091
1.4837	0.96	3720	4.6914	0.4112
1.4711	0.97	3740	4.6875	0.4122
1.5978	0.97	3760	4.5859	0.4091
2.7881	0.98	3780	4.5273	0.4101
2.2261	0.98	3800	4.4805	0.4060
1.6863	0.99	3820	4.3203	0.4091
2.0884	0.99	3840	4.2656	0.4081
2.1517	1.0	3860	4.2656	0.4081

Framework versions

• Transformers 4.32.0
• Pytorch 2.0.1+cu117
• Datasets 2.14.4
• Tokenizers 0.13.3