metadata

language: ar
license: apache-2.0
library_name: peft
base_model: CAMeL-Lab/bert-base-arabic-camelbert-mix
tags:
  - arabic
  - dialect-classification
  - lora

HammaLoRACAMeLBert

Advanced Arabic Dialect Classification Model with Complete Training Metrics

Full Training History

epoch	train_loss	eval_loss	train_accuracy	eval_accuracy	f1	precision	recall
1	1.37716	1.3692	0.53022	0.531461	0.473072	0.541513	0.53022
2	1.07614	1.06994	0.624063	0.636517	0.613924	0.630747	0.624063
3	0.968308	0.956658	0.662962	0.675281	0.661112	0.674352	0.662962
4	0.908012	0.900243	0.680569	0.692697	0.67889	0.688134	0.680569
5	0.858731	0.850955	0.699051	0.711236	0.70055	0.711083	0.699051
6	0.83265	0.825597	0.708417	0.716854	0.709791	0.724241	0.708417
7	0.796107	0.792507	0.722777	0.727528	0.723871	0.733254	0.722777
8	0.774045	0.773437	0.729645	0.732584	0.732394	0.747566	0.729645
9	0.758434	0.762295	0.736763	0.737079	0.739592	0.753967	0.736763
10	0.74339	0.7477	0.743194	0.744382	0.74493	0.757555	0.743194
11	0.730814	0.737005	0.74975	0.748876	0.751424	0.760475	0.74975
12	0.731264	0.74201	0.747378	0.748315	0.749058	0.760792	0.747378
13	0.719465	0.729863	0.753184	0.750562	0.754813	0.765065	0.753184
14	0.71225	0.722574	0.756868	0.754494	0.758437	0.767501	0.756868
15	0.712998	0.724446	0.756244	0.755618	0.757984	0.767426	0.756244

Label Mapping:

{0: 'Egypt', 1: 'Iraq', 2: 'Lebanon', 3: 'Morocco', 4: 'Saudi_Arabia', 5: 'Sudan', 6: 'Tunisia'}

USAGE Example:

from transformers import pipeline

classifier = pipeline(
    "text-classification",
    model="Hamma-16/HammaLoRACAMeLBert",
    device="cuda" if torch.cuda.is_available() else "cpu"
)

sample_text = "شلونك اليوم؟"
result = classifier(sample_text)
print(f"Text: {sample_text}")
print(f"Predicted: {result[0]['label']} (confidence: {result[0]['score']:.1%})")