xlm-roberta-base-finetuning-wrime-random4000-epoch6-test01

This model is a fine-tuned version of xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3173
• Jaccard (sample): [0.591182364729459, 0.3438395415472779, 0.6059322033898306, 0.036, 0.0, 0.14396887159533073, 0.1870967741935484, 0.0]
• Jaccard (macro): 0.2385
• Macro-precision: 0.4548
• Macro-recall: 0.2914
• Macro F1: 0.3307
• Micro F1: 0.5504
• Accuracy: 0.3125

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 6
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Jaccard (sample)	Jaccard (macro)	Macro-precision	Macro-recall	Macro F1	Micro F1	Accuracy
No log	1.0	125	0.4209	[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]	0.0	0.0	0.0	0.0	0.0	0.0775
No log	2.0	250	0.3810	[0.014084507042253521, 0.03816793893129771, 0.0, 0.0, 0.0, 0.004739336492890996, 0.0, 0.0]	0.0071	0.2821	0.0074	0.0138	0.0194	0.0792
No log	3.0	375	0.3545	[0.5387323943661971, 0.271523178807947, 0.3476190476190476, 0.0, 0.0, 0.0047169811320754715, 0.016260162601626018, 0.0]	0.1474	0.4623	0.1790	0.2106	0.4353	0.2475
0.4439	4.0	500	0.3286	[0.5519848771266541, 0.3492957746478873, 0.5913978494623656, 0.0, 0.0, 0.155893536121673, 0.1310344827586207, 0.0]	0.2225	0.3752	0.2767	0.3092	0.5339	0.2983
0.4439	5.0	625	0.3213	[0.599271402550091, 0.3089171974522293, 0.5960698689956332, 0.10384615384615385, 0.0, 0.09691629955947137, 0.12857142857142856, 0.0]	0.2292	0.4834	0.2749	0.3201	0.5519	0.3233
0.4439	6.0	750	0.3173	[0.591182364729459, 0.3438395415472779, 0.6059322033898306, 0.036, 0.0, 0.14396887159533073, 0.1870967741935484, 0.0]	0.2385	0.4548	0.2914	0.3307	0.5504	0.3125

Framework versions

• Transformers 4.35.2
• Pytorch 1.13.0
• Datasets 2.15.0
• Tokenizers 0.15.0