scenario-KD-PR-CDF-ALL-D2_data-cardiffnlp_tweet_sentiment_multilingual_all55

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: 1.2180
• Accuracy: 0.5999
• F1: 0.6008

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: 55
• 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
1.2423	1.09	500	1.2214	0.4842	0.4591
1.1465	2.17	1000	1.2081	0.5498	0.5406
1.089	3.26	1500	1.2345	0.5540	0.5476
1.043	4.35	2000	1.2340	0.5756	0.5777
1.01	5.43	2500	1.2397	0.5706	0.5717
0.9787	6.52	3000	1.2536	0.5718	0.5723
0.9656	7.61	3500	1.2564	0.5579	0.5603
0.9505	8.7	4000	1.2641	0.5644	0.5660
0.9432	9.78	4500	1.2385	0.5880	0.5876
0.9304	10.87	5000	1.2612	0.5864	0.5862
0.9245	11.96	5500	1.2567	0.5748	0.5728
0.9189	13.04	6000	1.2463	0.5745	0.5745
0.9131	14.13	6500	1.2599	0.5729	0.5738
0.9098	15.22	7000	1.2614	0.5706	0.5704
0.9052	16.3	7500	1.2468	0.5741	0.5748
0.9013	17.39	8000	1.2550	0.5756	0.5775
0.8972	18.48	8500	1.2661	0.5733	0.5743
0.8972	19.57	9000	1.2506	0.5783	0.5780
0.8912	20.65	9500	1.2519	0.5737	0.5752
0.8903	21.74	10000	1.2313	0.5795	0.5782
0.8868	22.83	10500	1.2384	0.5895	0.5896
0.8847	23.91	11000	1.2474	0.5752	0.5736
0.8834	25.0	11500	1.2458	0.5791	0.5795
0.8815	26.09	12000	1.2548	0.5748	0.5739
0.8794	27.17	12500	1.2378	0.5864	0.5857
0.8791	28.26	13000	1.2327	0.5968	0.5953
0.8749	29.35	13500	1.2249	0.5949	0.5935
0.8748	30.43	14000	1.2309	0.5938	0.5905
0.8734	31.52	14500	1.2242	0.5880	0.5885
0.872	32.61	15000	1.2372	0.5841	0.5856
0.8712	33.7	15500	1.2394	0.5783	0.5800
0.87	34.78	16000	1.2363	0.5922	0.5921
0.8692	35.87	16500	1.2375	0.5903	0.5916
0.8677	36.96	17000	1.2341	0.5968	0.5951
0.8672	38.04	17500	1.2227	0.6038	0.6013
0.8657	39.13	18000	1.2250	0.5899	0.5904
0.865	40.22	18500	1.2275	0.5949	0.5952
0.865	41.3	19000	1.2196	0.5953	0.5958
0.864	42.39	19500	1.2375	0.5818	0.5815
0.8636	43.48	20000	1.2373	0.5849	0.5856
0.8635	44.57	20500	1.2292	0.5930	0.5940
0.8622	45.65	21000	1.2243	0.5903	0.5914
0.8619	46.74	21500	1.2198	0.5984	0.5992
0.8608	47.83	22000	1.2175	0.6046	0.6054
0.8621	48.91	22500	1.2179	0.5995	0.6004
0.8606	50.0	23000	1.2180	0.5999	0.6008

Framework versions

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