scenario-KD-PR-CDF-ALL-D2_data-cardiffnlp_tweet_sentiment_multilingual_all66

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: 1.2134
• Accuracy: 0.5976
• F1: 0.5983

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: 66
• 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
1.2324	1.09	500	1.1970	0.5390	0.5349
1.1375	2.17	1000	1.1962	0.5683	0.5697
1.0818	3.26	1500	1.2071	0.5864	0.5880
1.0336	4.35	2000	1.2252	0.5829	0.5784
0.9993	5.43	2500	1.2483	0.5718	0.5739
0.9736	6.52	3000	1.2370	0.5679	0.5704
0.9561	7.61	3500	1.2357	0.5787	0.5793
0.9417	8.7	4000	1.2761	0.5575	0.5567
0.9318	9.78	4500	1.2588	0.5752	0.5752
0.9229	10.87	5000	1.2304	0.5748	0.5759
0.9148	11.96	5500	1.2564	0.5741	0.5759
0.9085	13.04	6000	1.2593	0.5714	0.5695
0.9053	14.13	6500	1.2399	0.5694	0.5694
0.8993	15.22	7000	1.2473	0.5702	0.5710
0.898	16.3	7500	1.2271	0.5787	0.5800
0.8894	17.39	8000	1.2384	0.5691	0.5686
0.8894	18.48	8500	1.2441	0.5660	0.5669
0.8853	19.57	9000	1.2530	0.5725	0.5741
0.8828	20.65	9500	1.2464	0.5698	0.5712
0.8785	21.74	10000	1.2423	0.5860	0.5871
0.877	22.83	10500	1.2577	0.5664	0.5668
0.8758	23.91	11000	1.2432	0.5818	0.5808
0.8731	25.0	11500	1.2480	0.5660	0.5678
0.8715	26.09	12000	1.2488	0.5544	0.5550
0.8688	27.17	12500	1.2414	0.5768	0.5786
0.8667	28.26	13000	1.2339	0.5756	0.5711
0.8661	29.35	13500	1.2204	0.5903	0.5911
0.8644	30.43	14000	1.2427	0.5656	0.5673
0.862	31.52	14500	1.2421	0.5799	0.5810
0.8611	32.61	15000	1.2375	0.5764	0.5759
0.8612	33.7	15500	1.2184	0.6003	0.5975
0.8583	34.78	16000	1.2345	0.5841	0.5851
0.8585	35.87	16500	1.2324	0.5841	0.5849
0.8569	36.96	17000	1.2308	0.5818	0.5817
0.8553	38.04	17500	1.2209	0.5968	0.5949
0.8547	39.13	18000	1.2301	0.5853	0.5859
0.8547	40.22	18500	1.2211	0.5887	0.5874
0.8532	41.3	19000	1.2224	0.5895	0.5912
0.853	42.39	19500	1.2245	0.5829	0.5844
0.8523	43.48	20000	1.2187	0.5945	0.5947
0.8521	44.57	20500	1.2141	0.5965	0.5959
0.851	45.65	21000	1.2122	0.6011	0.6016
0.8512	46.74	21500	1.2176	0.5941	0.5944
0.8507	47.83	22000	1.2161	0.5934	0.5943
0.8503	48.91	22500	1.2157	0.5957	0.5966
0.8501	50.0	23000	1.2134	0.5976	0.5983

Framework versions

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