Add model files

CAMeLBERT_MIX_preds_test_results

@@ -0,0 +1,2 @@
+Scores:
+{'recall': 0.25444826156406991, 'f1': 0.24724637848149009, 'precision': 0.27170344297962074, 'accuracy': 0.40899999999999997}

README.md

@@ -0,0 +1,47 @@
+---
+language: 
+- ar
+license: apache-2.0
+widget:
+ - text: "عامل ايه ؟"
+---
+# CAMeLBERT-Mix DID NADI Model
+## Model description
+**CAMeLBERT-Mix DID NADI Model** is a dialect identification (DID) model that was built by fine-tuning the [CAMeLBERT-Mix](https://huggingface.co/CAMeL-Lab/bert-base-arabic-camelbert-mix/) model.
+For the fine-tuning, we used the [NADI Coountry-level](https://sites.google.com/view/nadi-shared-task) dataset, which includes 21 labels.
+Our fine-tuning procedure and the hyperparameters we used can be found in our paper *"[The Interplay of Variant, Size, and Task Type in Arabic Pre-trained Language Models](https://arxiv.org/abs/2103.06678)."* Our fine-tuning code can be found [here](https://github.com/CAMeL-Lab/CAMeLBERT).
+
+## Intended uses
+You can use the CAMeLBERT-Mix DID NADI model as part of the transformers pipeline.
+This model will also be available in [CAMeL Tools](https://github.com/CAMeL-Lab/camel_tools) soon.
+
+#### How to use
+To use the model with a transformers pipeline:
+```python
+>>> from transformers import pipeline
+>>> did = pipeline('text-classification', model='CAMeL-Lab/bert-base-arabic-camelbert-mix-did-nadi')
+>>> sentences = ['عامل ايه ؟', 'شلونك ؟ شخبارك ؟']
+>>> did(sentences)
+[{'label': 'Egypt', 'score': 0.920274019241333},
+ {'label': 'Saudi_Arabia', 'score': 0.26750022172927856}]
+```
+*Note*: to download our models, you would need `transformers>=3.5.0`.
+Otherwise, you could download the models manually.
+
+## Citation
+```bibtex
+@inproceedings{inoue-etal-2021-interplay,
+    title = "The Interplay of Variant, Size, and Task Type in {A}rabic Pre-trained Language Models",
+    author = "Inoue, Go  and
+      Alhafni, Bashar  and
+      Baimukan, Nurpeiis  and
+      Bouamor, Houda  and
+      Habash, Nizar",
+    booktitle = "Proceedings of the Sixth Arabic Natural Language Processing Workshop",
+    month = apr,
+    year = "2021",
+    address = "Kyiv, Ukraine (Online)",
+    publisher = "Association for Computational Linguistics",
+    abstract = "In this paper, we explore the effects of language variants, data sizes, and fine-tuning task types in Arabic pre-trained language models. To do so, we build three pre-trained language models across three variants of Arabic: Modern Standard Arabic (MSA), dialectal Arabic, and classical Arabic, in addition to a fourth language model which is pre-trained on a mix of the three. We also examine the importance of pre-training data size by building additional models that are pre-trained on a scaled-down set of the MSA variant. We compare our different models to each other, as well as to eight publicly available models by fine-tuning them on five NLP tasks spanning 12 datasets. Our results suggest that the variant proximity of pre-training data to fine-tuning data is more important than the pre-training data size. We exploit this insight in defining an optimized system selection model for the studied tasks.",
+}
+```

config.json

@@ -0,0 +1,67 @@
+{
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "finetuning_task": "arabic_did_nadi_country",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "Algeria",
+    "1": "Bahrain",
+    "10": "Morocco",
+    "11": "Oman",
+    "12": "Palestine",
+    "13": "Qatar",
+    "14": "Saudi_Arabia",
+    "15": "Somalia",
+    "16": "Sudan",
+    "17": "Syria",
+    "18": "Tunisia",
+    "19": "United_Arab_Emirates",
+    "2": "Djibouti",
+    "20": "Yemen",
+    "3": "Egypt",
+    "4": "Iraq",
+    "5": "Jordan",
+    "6": "Kuwait",
+    "7": "Lebanon",
+    "8": "Libya",
+    "9": "Mauritania"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "Algeria": 0,
+    "Bahrain": 1,
+    "Djibouti": 2,
+    "Egypt": 3,
+    "Iraq": 4,
+    "Jordan": 5,
+    "Kuwait": 6,
+    "Lebanon": 7,
+    "Libya": 8,
+    "Mauritania": 9,
+    "Morocco": 10,
+    "Oman": 11,
+    "Palestine": 12,
+    "Qatar": 13,
+    "Saudi_Arabia": 14,
+    "Somalia": 15,
+    "Sudan": 16,
+    "Syria": 17,
+    "Tunisia": 18,
+    "United_Arab_Emirates": 19,
+    "Yemen": 20
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "type_vocab_size": 2,
+  "vocab_size": 30000
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e9fcdd45c0b172d453e94e0960861a0a59867bd68b9c29d331442aac123b033
+size 872839970

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:536809cb29d1416e0ae34b27e11d15718c9414fa325157f31eda31d7132723b5
+size 436444537

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d65d2c871523b6ee588c9f6d581bd283838a260385ed7707db33ff79f60f9774
+size 326

special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}

tf_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d307c14e3fe72711345680b475c150ee4b8a1a68ba2445db0525c7e296f05d0
+size 436592640

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": false, "special_tokens_map_file": null, "full_tokenizer_file": null}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76140491bee99d35c545192620f87c95e7b4066640bb87bc815a7450d9f0ec42
+size 1388