Sfaya-W-Ary-Arz-MSA

This model is a fine-tuned version of CAMeL-Lab/bert-base-arabic-camelbert-da on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0158
• Accuracy: 0.9978

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: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.4027	0.1862	100	0.0415	0.9923
0.0701	0.3724	200	0.0637	0.9923
0.0567	0.5587	300	0.0686	0.9945
0.0365	0.7449	400	0.0605	0.9934
0.036	0.9311	500	0.0351	0.9934
0.0154	1.1173	600	0.0158	0.9978
0.0182	1.3035	700	0.0219	0.9967
0.0235	1.4898	800	0.0956	0.9934
0.0237	1.6760	900	0.0360	0.9956
0.0105	1.8622	1000	0.0240	0.9956
0.0093	2.0484	1100	0.0738	0.9934
0.009	2.2346	1200	0.0278	0.9945
0.0096	2.4209	1300	0.0333	0.9956
0.0164	2.6071	1400	0.0998	0.9945
0.0206	2.7933	1500	0.0747	0.9945
0.0024	2.9795	1600	0.0524	0.9912
0.0002	3.1657	1700	0.1051	0.9945
0.0068	3.3520	1800	0.0264	0.9967
0.0034	3.5382	1900	0.0293	0.9945
0.0087	3.7244	2000	0.0222	0.9956
0.0069	3.9106	2100	0.0305	0.9967
0.0036	4.0968	2200	0.0108	0.9967
0.0025	4.2831	2300	0.0367	0.9967
0.0	4.4693	2400	0.0269	0.9967
0.0044	4.6555	2500	0.0240	0.9967
0.0	4.8417	2600	0.0297	0.9967
0.0	5.0279	2700	0.0608	0.9956
0.0	5.2142	2800	0.0592	0.9956
0.0	5.4004	2900	0.0284	0.9967
0.0	5.5866	3000	0.0285	0.9967
0.0	5.7728	3100	0.0285	0.9967
0.0	5.9590	3200	0.0287	0.9967
0.0016	6.1453	3300	0.0294	0.9967
0.0	6.3315	3400	0.0226	0.9978
0.0	6.5177	3500	0.0227	0.9978
0.0	6.7039	3600	0.0228	0.9978
0.0	6.8901	3700	0.0232	0.9978
0.0	7.0764	3800	0.0233	0.9978
0.0	7.2626	3900	0.0235	0.9978
0.0	7.4488	4000	0.0235	0.9978
0.0	7.6350	4100	0.0235	0.9978
0.0	7.8212	4200	0.0236	0.9978
0.0	8.0074	4300	0.0239	0.9978
0.0	8.1937	4400	0.0240	0.9978
0.0	8.3799	4500	0.0240	0.9978
0.0	8.5661	4600	0.0241	0.9978
0.0	8.7523	4700	0.0241	0.9978
0.0025	8.9385	4800	0.0355	0.9934
0.0	9.1248	4900	0.0310	0.9945
0.0	9.3110	5000	0.0291	0.9945
0.0	9.4972	5100	0.0290	0.9956
0.0	9.6834	5200	0.0290	0.9956
0.0	9.8696	5300	0.0290	0.9956

Framework versions

• Transformers 4.45.2
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3