rlcc-palate-upsample_replacement-absa-min

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.1294
• Accuracy: 0.7878
• F1 Macro: 0.5375
• Precision Macro: 0.5410
• Recall Macro: 0.5771
• Total Tf: [323, 87, 1143, 87]

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 36
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	Precision Macro	Recall Macro	Total Tf
1.1114	1.0	37	1.0849	0.8073	0.4646	0.4699	0.4711	[331, 79, 1151, 79]
1.0656	2.0	74	1.1060	0.8	0.4744	0.4575	0.5394	[328, 82, 1148, 82]
0.9251	3.0	111	1.0360	0.8024	0.5730	0.5780	0.5942	[329, 81, 1149, 81]
0.8019	4.0	148	1.1099	0.8341	0.5423	0.5462	0.5925	[342, 68, 1162, 68]
0.7155	5.0	185	1.2333	0.7976	0.5602	0.5848	0.6127	[327, 83, 1147, 83]
0.5969	6.0	222	1.2154	0.8341	0.5266	0.4951	0.5735	[342, 68, 1162, 68]
0.5489	7.0	259	1.3377	0.8	0.5650	0.5661	0.6076	[328, 82, 1148, 82]
0.584	8.0	296	1.4198	0.7976	0.5680	0.6274	0.6648	[327, 83, 1147, 83]
0.557	9.0	333	1.4447	0.7902	0.5346	0.5320	0.5644	[324, 86, 1144, 86]
0.5376	10.0	370	1.4859	0.7927	0.5551	0.6208	0.6404	[325, 85, 1145, 85]
0.4886	11.0	407	1.5739	0.7951	0.5528	0.5561	0.5920	[326, 84, 1146, 84]
0.4906	12.0	444	1.6267	0.8146	0.5814	0.5777	0.6177	[334, 76, 1154, 76]
0.4341	13.0	481	1.6743	0.8098	0.5803	0.5850	0.6221	[332, 78, 1152, 78]
0.4137	14.0	518	1.7262	0.7976	0.5643	0.5734	0.6220	[327, 83, 1147, 83]
0.3311	15.0	555	1.8798	0.8	0.5580	0.5589	0.5891	[328, 82, 1148, 82]
0.3054	16.0	592	1.8805	0.7756	0.5239	0.5372	0.5906	[318, 92, 1138, 92]
0.2563	17.0	629	1.9434	0.7878	0.5476	0.5629	0.6221	[323, 87, 1143, 87]
0.2458	18.0	666	1.9002	0.7976	0.5597	0.5626	0.6003	[327, 83, 1147, 83]
0.2234	19.0	703	1.9109	0.7878	0.5505	0.5560	0.6124	[323, 87, 1143, 87]
0.2291	20.0	740	1.9709	0.7927	0.5507	0.5555	0.5922	[325, 85, 1145, 85]
0.2231	21.0	777	2.0283	0.8024	0.5639	0.5662	0.5996	[329, 81, 1149, 81]
0.191	22.0	814	2.0057	0.7854	0.5355	0.5444	0.5812	[322, 88, 1142, 88]
0.1827	23.0	851	2.0780	0.8049	0.5658	0.5653	0.5977	[330, 80, 1150, 80]
0.1929	24.0	888	2.1094	0.7927	0.5455	0.5475	0.5830	[325, 85, 1145, 85]
0.1572	25.0	925	2.1294	0.7878	0.5375	0.5410	0.5771	[323, 87, 1143, 87]

Framework versions

• Transformers 4.48.3
• Pytorch 2.1.0+cu118
• Tokenizers 0.21.0