README.md

metadata

license: mit
base_model: pdelobelle/robbert-v2-dutch-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
model-index:
  - name: robbert-v2-dutch-base-finetuned-emotion
    results: []

robbert-v2-dutch-base-finetuned-emotion

This model is a fine-tuned version of pdelobelle/robbert-v2-dutch-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.1901
• Accuracy: 0.52
• F1: 0.5133

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.5504	1.0	25	1.4311	0.41	0.2443
1.4241	2.0	50	1.3456	0.46	0.3532
1.3432	3.0	75	1.3765	0.46	0.3581
1.2749	4.0	100	1.3694	0.5	0.3892
1.1421	5.0	125	1.3083	0.51	0.4035
1.0092	6.0	150	1.3147	0.51	0.4364
0.8896	7.0	175	1.2769	0.53	0.4936
0.7669	8.0	200	1.3297	0.52	0.4971
0.6463	9.0	225	1.4932	0.55	0.5179
0.5364	10.0	250	1.4984	0.53	0.4999
0.4657	11.0	275	1.7043	0.5	0.4587
0.3777	12.0	300	1.6938	0.49	0.4541
0.3019	13.0	325	1.7869	0.54	0.5221
0.2379	14.0	350	1.7080	0.52	0.5191
0.1782	15.0	375	1.7824	0.5	0.4991
0.1494	16.0	400	1.9882	0.55	0.5076
0.1103	17.0	425	2.0976	0.53	0.5064
0.0951	18.0	450	2.1631	0.48	0.4962
0.0867	19.0	475	2.1215	0.52	0.5091
0.0637	20.0	500	2.1948	0.52	0.5103
0.0536	21.0	525	2.3024	0.49	0.4909
0.0345	22.0	550	2.2948	0.52	0.5010
0.0293	23.0	575	2.4334	0.55	0.5378
0.0272	24.0	600	2.5223	0.52	0.5186
0.0222	25.0	625	2.6465	0.5	0.5065
0.0127	26.0	650	2.6434	0.5	0.5055
0.0111	27.0	675	2.7423	0.5	0.5040
0.01	28.0	700	2.8313	0.5	0.4990
0.0057	29.0	725	2.8568	0.49	0.4740
0.0095	30.0	750	2.8480	0.52	0.5033
0.0108	31.0	775	2.9276	0.46	0.4684
0.0062	32.0	800	2.9811	0.51	0.5027
0.0045	33.0	825	2.9901	0.51	0.5022
0.005	34.0	850	3.0403	0.52	0.5129
0.0036	35.0	875	3.0578	0.51	0.5076
0.003	36.0	900	3.0680	0.51	0.5022
0.0028	37.0	925	3.1097	0.52	0.5170
0.0028	38.0	950	3.1304	0.49	0.4801
0.0027	39.0	975	3.1341	0.49	0.4801
0.0069	40.0	1000	3.1380	0.49	0.4958
0.0023	41.0	1025	3.1284	0.5	0.4954
0.0035	42.0	1050	3.1454	0.51	0.5062
0.0033	43.0	1075	3.1758	0.49	0.4730
0.0021	44.0	1100	3.1660	0.5	0.4906
0.0022	45.0	1125	3.1688	0.5	0.4906
0.0021	46.0	1150	3.1649	0.51	0.5061
0.0022	47.0	1175	3.1855	0.52	0.5139
0.0021	48.0	1200	3.1882	0.51	0.5061
0.0022	49.0	1225	3.1893	0.52	0.5133
0.0019	50.0	1250	3.1901	0.52	0.5133

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1