README.md

metadata

license: mit
tags:
  - generated_from_keras_callback
model-index:
  - name: madatnlp/gamza-bart-for-kormath
    results: []

madatnlp/gamza-bart-for-kormath

This model is a fine-tuned version of gogamza/kobart-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 0.0460
• Validation Loss: 0.3173
• Epoch: 52

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:

• optimizer: {'name': 'Adam', 'learning_rate': 1e-04, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Validation Loss	Epoch
4.3868	1.8720	0
1.5091	1.0768	1
1.0238	0.8972	2
0.8960	0.8375	3
0.8087	0.7618	4
0.7656	0.7196	5
0.7433	0.7023	6
0.7309	0.6992	7
0.6773	0.7008	8
0.6659	0.6752	9
0.6607	0.7347	10
0.6370	0.6033	11
0.6037	0.6806	12
0.5957	0.6370	13
0.5740	0.6413	14
0.5005	0.5002	15
0.4385	0.4821	16
0.5042	0.4898	17
0.4445	0.5218	18
0.3990	0.4771	19
0.3497	0.4173	20
0.3294	0.4232	21
0.3019	0.3936	22
0.2868	0.4212	23
0.2565	0.3401	24
0.2404	0.3912	25
0.2235	0.3484	26
0.1976	0.3357	27
0.1998	0.3392	28
0.1788	0.3075	29
0.1705	0.3615	30
0.1642	0.3466	31
0.1510	0.3325	32
0.1402	0.3729	33
0.1342	0.3425	34
0.1276	0.3039	35
0.1093	0.3276	36
0.1048	0.3034	37
0.0951	0.3472	38
0.0971	0.3407	39
0.0849	0.3516	40
0.0868	0.2757	41
0.0868	0.2383	42
0.0766	0.3127	43
0.0734	0.3324	44
0.0694	0.3654	45
0.0626	0.2692	46
0.0584	0.2403	47
0.0565	0.3741	48
0.0567	0.2839	49
0.0523	0.2387	50
0.0525	0.2629	51
0.0460	0.3173	52

Framework versions

• Transformers 4.18.0
• TensorFlow 2.8.0
• Datasets 2.1.0
• Tokenizers 0.12.1