scenario-KD-PR-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: 1.2156
• Accuracy: 0.5899
• F1: 0.5895

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
1.2282	1.09	500	1.2134	0.5467	0.5385
1.1368	2.17	1000	1.1933	0.5733	0.5743
1.0779	3.26	1500	1.2131	0.5629	0.5597
1.0308	4.35	2000	1.2485	0.5602	0.5608
0.9977	5.43	2500	1.2645	0.5567	0.5583
0.9742	6.52	3000	1.2412	0.5606	0.5575
0.953	7.61	3500	1.2490	0.5710	0.5675
0.9419	8.7	4000	1.2480	0.5718	0.5733
0.9274	9.78	4500	1.2595	0.5629	0.5645
0.9198	10.87	5000	1.2329	0.5841	0.5845
0.913	11.96	5500	1.2359	0.5733	0.5745
0.9076	13.04	6000	1.2598	0.5729	0.5743
0.8995	14.13	6500	1.2409	0.5799	0.5753
0.8977	15.22	7000	1.2499	0.5752	0.5770
0.8914	16.3	7500	1.2278	0.5802	0.5796
0.8882	17.39	8000	1.2295	0.5748	0.5749
0.886	18.48	8500	1.2371	0.5737	0.5684
0.8802	19.57	9000	1.2394	0.5664	0.5581
0.8798	20.65	9500	1.2500	0.5741	0.5734
0.8768	21.74	10000	1.2488	0.5702	0.5722
0.8734	22.83	10500	1.2380	0.5764	0.5773
0.8719	23.91	11000	1.2147	0.5949	0.5894
0.8693	25.0	11500	1.2217	0.5853	0.5841
0.8672	26.09	12000	1.2320	0.5814	0.5806
0.8667	27.17	12500	1.2363	0.5756	0.5745
0.8643	28.26	13000	1.2211	0.5845	0.5826
0.8632	29.35	13500	1.2325	0.5841	0.5850
0.8608	30.43	14000	1.2310	0.5810	0.5803
0.8597	31.52	14500	1.2260	0.5868	0.5869
0.8581	32.61	15000	1.2221	0.5891	0.5899
0.8563	33.7	15500	1.2233	0.5907	0.5892
0.8569	34.78	16000	1.2204	0.5914	0.5903
0.8549	35.87	16500	1.2245	0.5872	0.5878
0.8542	36.96	17000	1.2278	0.5806	0.5799
0.8534	38.04	17500	1.2080	0.5968	0.5980
0.8526	39.13	18000	1.2141	0.5833	0.5769
0.8516	40.22	18500	1.2347	0.5772	0.5791
0.8512	41.3	19000	1.2305	0.5779	0.5799
0.8513	42.39	19500	1.2168	0.5880	0.5898
0.8491	43.48	20000	1.2221	0.5841	0.5848
0.8493	44.57	20500	1.2172	0.5853	0.5860
0.8485	45.65	21000	1.2120	0.5922	0.5929
0.8484	46.74	21500	1.2123	0.5945	0.5953
0.8478	47.83	22000	1.2148	0.5887	0.5876
0.848	48.91	22500	1.2163	0.5864	0.5864
0.8478	50.0	23000	1.2156	0.5899	0.5895

Framework versions

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