gemma-fineweb-edu-scorer-mdeberta-binary-lr5e-05-20250411_140230

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1614
• Precision: 1.0
• Recall: 1.0
• F1 Macro: 1.0
• Accuracy: 1.0

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 32
• seed: 0
• 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1 Macro	Accuracy
No log	0	0	0.2170	1.0	1.0	1.0	1.0
0.1238	0.1954	1000	0.1458	1.0	1.0	1.0	1.0
0.1143	0.3908	2000	0.1076	1.0	1.0	1.0	1.0
0.1208	0.5862	3000	0.1133	1.0	1.0	1.0	1.0
0.1031	0.7816	4000	0.1069	1.0	1.0	1.0	1.0
0.1045	0.9769	5000	0.1047	1.0	1.0	1.0	1.0
0.0974	1.1723	6000	0.1087	1.0	1.0	1.0	1.0
0.1337	1.3677	7000	0.1209	1.0	1.0	1.0	1.0
0.0864	1.5631	8000	0.1151	1.0	1.0	1.0	1.0
0.1049	1.7585	9000	0.1203	1.0	1.0	1.0	1.0
0.0915	1.9539	10000	0.1255	1.0	1.0	1.0	1.0
0.079	2.1493	11000	0.1369	1.0	1.0	1.0	1.0
0.0709	2.3447	12000	0.1167	1.0	1.0	1.0	1.0
0.0823	2.5401	13000	0.1257	1.0	1.0	1.0	1.0
0.0725	2.7354	14000	0.1114	1.0	1.0	1.0	1.0
0.066	2.9308	15000	0.1232	1.0	1.0	1.0	1.0
0.0354	3.1262	16000	0.1321	1.0	1.0	1.0	1.0
0.0486	3.3216	17000	0.1245	1.0	1.0	1.0	1.0
0.0466	3.5170	18000	0.1329	1.0	1.0	1.0	1.0
0.0523	3.7124	19000	0.1350	1.0	1.0	1.0	1.0
0.0614	3.9078	20000	0.1373	1.0	1.0	1.0	1.0
0.0349	4.1032	21000	0.1461	1.0	1.0	1.0	1.0
0.0315	4.2986	22000	0.1406	1.0	1.0	1.0	1.0
0.0337	4.4939	23000	0.1355	1.0	1.0	1.0	1.0
0.0345	4.6893	24000	0.1300	1.0	1.0	1.0	1.0
0.0324	4.8847	25000	0.1324	1.0	1.0	1.0	1.0
0.0247	5.0801	26000	0.1347	1.0	1.0	1.0	1.0
0.0306	5.2755	27000	0.1474	1.0	1.0	1.0	1.0
0.0267	5.4709	28000	0.1394	1.0	1.0	1.0	1.0
0.0204	5.6663	29000	0.1487	1.0	1.0	1.0	1.0
0.0266	5.8617	30000	0.1454	1.0	1.0	1.0	1.0
0.0188	6.0571	31000	0.1434	1.0	1.0	1.0	1.0
0.0192	6.2524	32000	0.1458	1.0	1.0	1.0	1.0
0.0246	6.4478	33000	0.1455	1.0	1.0	1.0	1.0
0.0219	6.6432	34000	0.1440	1.0	1.0	1.0	1.0
0.0231	6.8386	35000	0.1561	1.0	1.0	1.0	1.0
0.0185	7.0340	36000	0.1504	1.0	1.0	1.0	1.0
0.0163	7.2294	37000	0.1516	1.0	1.0	1.0	1.0
0.0186	7.4248	38000	0.1451	1.0	1.0	1.0	1.0
0.025	7.6202	39000	0.1423	1.0	1.0	1.0	1.0
0.0217	7.8156	40000	0.1453	1.0	1.0	1.0	1.0
0.0226	8.0109	41000	0.1584	1.0	1.0	1.0	1.0
0.0136	8.2063	42000	0.1589	1.0	1.0	1.0	1.0
0.0186	8.4017	43000	0.1518	1.0	1.0	1.0	1.0
0.0215	8.5971	44000	0.1459	1.0	1.0	1.0	1.0
0.0182	8.7925	45000	0.1437	1.0	1.0	1.0	1.0
0.0236	8.9879	46000	0.1409	1.0	1.0	1.0	1.0
0.0204	9.1833	47000	0.1514	1.0	1.0	1.0	1.0
0.0198	9.3787	48000	0.1617	1.0	1.0	1.0	1.0
0.0158	9.5741	49000	0.1366	1.0	1.0	1.0	1.0
0.0256	9.7694	50000	0.1450	1.0	1.0	1.0	1.0
0.0186	9.9648	51000	0.1525	1.0	1.0	1.0	1.0
0.0238	10.1602	52000	0.1658	1.0	1.0	1.0	1.0
0.0189	10.3556	53000	0.1442	1.0	1.0	1.0	1.0
0.0184	10.5510	54000	0.1495	1.0	1.0	1.0	1.0
0.0196	10.7464	55000	0.1428	1.0	1.0	1.0	1.0
0.0224	10.9418	56000	0.1606	1.0	1.0	1.0	1.0
0.018	11.1372	57000	0.1436	1.0	1.0	1.0	1.0
0.1813	11.3326	58000	0.1829	1.0	1.0	1.0	1.0
0.1888	11.5279	59000	0.1832	1.0	1.0	1.0	1.0
0.1826	11.7233	60000	0.1828	1.0	1.0	1.0	1.0
0.0227	11.9187	61000	0.1447	1.0	1.0	1.0	1.0
0.0162	12.1141	62000	0.1491	1.0	1.0	1.0	1.0
0.0212	12.3095	63000	0.1386	1.0	1.0	1.0	1.0
0.0124	12.5049	64000	0.1507	1.0	1.0	1.0	1.0
0.0162	12.7003	65000	0.1425	1.0	1.0	1.0	1.0
0.0122	12.8957	66000	0.1417	1.0	1.0	1.0	1.0
0.0105	13.0911	67000	0.1414	1.0	1.0	1.0	1.0
0.0114	13.2864	68000	0.1537	1.0	1.0	1.0	1.0
0.0083	13.4818	69000	0.1548	1.0	1.0	1.0	1.0
0.0084	13.6772	70000	0.1439	1.0	1.0	1.0	1.0
0.0113	13.8726	71000	0.1504	1.0	1.0	1.0	1.0
0.013	14.0680	72000	0.1480	1.0	1.0	1.0	1.0
0.0066	14.2634	73000	0.1544	1.0	1.0	1.0	1.0
0.0119	14.4588	74000	0.1509	1.0	1.0	1.0	1.0
0.0078	14.6542	75000	0.1546	1.0	1.0	1.0	1.0
0.0162	14.8496	76000	0.1513	1.0	1.0	1.0	1.0
0.0094	15.0449	77000	0.1571	1.0	1.0	1.0	1.0
0.0097	15.2403	78000	0.1646	1.0	1.0	1.0	1.0
0.0132	15.4357	79000	0.1505	1.0	1.0	1.0	1.0
0.0127	15.6311	80000	0.1539	1.0	1.0	1.0	1.0
0.0086	15.8265	81000	0.1572	1.0	1.0	1.0	1.0
0.0067	16.0219	82000	0.1583	1.0	1.0	1.0	1.0
0.007	16.2173	83000	0.1531	1.0	1.0	1.0	1.0
0.0117	16.4127	84000	0.1485	1.0	1.0	1.0	1.0
0.0156	16.6081	85000	0.1495	1.0	1.0	1.0	1.0
0.0089	16.8034	86000	0.1570	1.0	1.0	1.0	1.0
0.0075	16.9988	87000	0.1540	1.0	1.0	1.0	1.0
0.0068	17.1942	88000	0.1612	1.0	1.0	1.0	1.0
0.0074	17.3896	89000	0.1596	1.0	1.0	1.0	1.0
0.0117	17.5850	90000	0.1617	1.0	1.0	1.0	1.0
0.0116	17.7804	91000	0.1689	1.0	1.0	1.0	1.0
0.0064	17.9758	92000	0.1602	1.0	1.0	1.0	1.0
0.0079	18.1712	93000	0.1647	1.0	1.0	1.0	1.0
0.0051	18.3665	94000	0.1534	1.0	1.0	1.0	1.0
0.0069	18.5619	95000	0.1570	1.0	1.0	1.0	1.0
0.0062	18.7573	96000	0.1533	1.0	1.0	1.0	1.0
0.0048	18.9527	97000	0.1566	1.0	1.0	1.0	1.0
0.0142	19.1481	98000	0.1532	1.0	1.0	1.0	1.0
0.0147	19.3435	99000	0.1501	1.0	1.0	1.0	1.0
0.0085	19.5389	100000	0.1535	1.0	1.0	1.0	1.0
0.005	19.7343	101000	0.1599	1.0	1.0	1.0	1.0
0.0066	19.9297	102000	0.1614	1.0	1.0	1.0	1.0

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.1