camembert_ccnet_classification_tools_classifier-only_fr_lr1e-3_V2

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

• Loss: 0.2602
• Accuracy: 0.875
• Learning Rate: 0.0004

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: 0.001
• train_batch_size: 24
• eval_batch_size: 192
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 60

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Rate
1.9376	1.0	15	1.6974	0.3542	0.0010
1.4628	2.0	30	1.4891	0.4792	0.0010
1.1834	3.0	45	0.9756	0.7083	0.0009
0.9327	4.0	60	0.8686	0.6979	0.0009
0.781	5.0	75	0.7231	0.7083	0.0009
0.7408	6.0	90	0.5991	0.8229	0.0009
0.6046	7.0	105	0.5022	0.875	0.0009
0.5957	8.0	120	0.4873	0.8021	0.0009
0.5546	9.0	135	0.4919	0.8125	0.0008
0.5204	10.0	150	0.4814	0.8021	0.0008
0.4636	11.0	165	0.3945	0.8542	0.0008
0.4405	12.0	180	0.4338	0.8229	0.0008
0.4805	13.0	195	0.3747	0.8854	0.0008
0.39	14.0	210	0.3946	0.8854	0.0008
0.3982	15.0	225	0.3480	0.8438	0.0008
0.3967	16.0	240	0.3188	0.9271	0.0007
0.3836	17.0	255	0.4498	0.7812	0.0007
0.3764	18.0	270	0.2861	0.8958	0.0007
0.3387	19.0	285	0.2953	0.9062	0.0007
0.3749	20.0	300	0.2771	0.9062	0.0007
0.3353	21.0	315	0.3058	0.8958	0.0007
0.3523	22.0	330	0.2710	0.9062	0.0006
0.3383	23.0	345	0.2597	0.9167	0.0006
0.3223	24.0	360	0.3180	0.8854	0.0006
0.3644	25.0	375	0.2712	0.9062	0.0006
0.3015	26.0	390	0.2775	0.9062	0.0006
0.3186	27.0	405	0.2386	0.9375	0.0006
0.2915	28.0	420	0.3227	0.8958	0.0005
0.3049	29.0	435	0.2908	0.9167	0.0005
0.3131	30.0	450	0.2921	0.9062	0.0005
0.3187	31.0	465	0.2733	0.9062	0.0005
0.3051	32.0	480	0.2850	0.9062	0.0005
0.2775	33.0	495	0.2621	0.9062	0.0005
0.3331	34.0	510	0.2742	0.9167	0.0004
0.2854	35.0	525	0.3128	0.8854	0.0004
0.294	36.0	540	0.2455	0.9167	0.0004
0.2662	37.0	555	0.2602	0.875	0.0004

Framework versions

• Transformers 4.34.0
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.14.1