scenario-KD-PO-CDF-ALL-D2_data-cardiffnlp_tweet_sentiment_multilingual_all44

This model is a fine-tuned version of haryoaw/scenario-MDBT-TCR_data-cardiffnlp_tweet_sentiment_multilingual_all on the tweet_sentiment_multilingual dataset. It achieves the following results on the evaluation set:

• Loss: 3.3146
• Accuracy: 0.5999
• F1: 0.5999

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: 32
• eval_batch_size: 32
• seed: 44
• 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
4.444	1.09	500	3.3357	0.5590	0.5595
3.1285	2.17	1000	3.1819	0.5818	0.5811
2.2768	3.26	1500	3.4078	0.5799	0.5754
1.6869	4.35	2000	3.4404	0.5729	0.5727
1.3429	5.43	2500	3.3948	0.5826	0.5832
1.1198	6.52	3000	3.6492	0.5675	0.5686
0.9743	7.61	3500	3.5282	0.5764	0.5777
0.8843	8.7	4000	3.3969	0.5775	0.5743
0.7859	9.78	4500	3.2240	0.5802	0.5794
0.722	10.87	5000	3.3416	0.5806	0.5786
0.6697	11.96	5500	3.2493	0.5806	0.5824
0.6214	13.04	6000	3.2909	0.5841	0.5806
0.5762	14.13	6500	3.3289	0.5760	0.5775
0.5513	15.22	7000	3.2534	0.5876	0.5843
0.5138	16.3	7500	3.4835	0.5849	0.5857
0.4968	17.39	8000	3.2355	0.5860	0.5855
0.4842	18.48	8500	3.3245	0.5849	0.5862
0.4552	19.57	9000	3.3149	0.5876	0.5869
0.4492	20.65	9500	3.2398	0.5787	0.5797
0.43	21.74	10000	3.3776	0.5775	0.5769
0.4174	22.83	10500	3.3436	0.5910	0.5908
0.4145	23.91	11000	3.2675	0.5953	0.5948
0.3974	25.0	11500	3.2194	0.5922	0.5934
0.3884	26.09	12000	3.1719	0.5899	0.5909
0.3863	27.17	12500	3.4059	0.5891	0.5877
0.3713	28.26	13000	3.3166	0.5980	0.5952
0.3707	29.35	13500	3.1796	0.5957	0.5966
0.3719	30.43	14000	3.3469	0.5938	0.5939
0.3593	31.52	14500	3.2776	0.6046	0.6036
0.3576	32.61	15000	3.4264	0.5914	0.5917
0.353	33.7	15500	3.3841	0.5949	0.5952
0.3512	34.78	16000	3.4532	0.5953	0.5937
0.3449	35.87	16500	3.3434	0.5980	0.5988
0.3456	36.96	17000	3.5318	0.5918	0.5900
0.3433	38.04	17500	3.2213	0.6069	0.6074
0.3392	39.13	18000	3.3987	0.5992	0.5986
0.3391	40.22	18500	3.3773	0.5953	0.5952
0.3327	41.3	19000	3.3447	0.5914	0.5919
0.3336	42.39	19500	3.4175	0.5972	0.5971
0.326	43.48	20000	3.3289	0.6026	0.6029
0.3344	44.57	20500	3.3741	0.5992	0.5993
0.3255	45.65	21000	3.3618	0.5988	0.5986
0.3283	46.74	21500	3.4027	0.5999	0.6003
0.3236	47.83	22000	3.3338	0.6046	0.6048
0.3254	48.91	22500	3.3374	0.5965	0.5973
0.3247	50.0	23000	3.3146	0.5999	0.5999

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3