Upload model

README.md

@@ -1,84 +1,44 @@
 ---
-license: apache-2.0
 tags:
-- generated_from_trainer
-metrics:
-- accuracy
-- f1
-- precision
-- recall
-model-index:
-- name: ABSA-SentencePair-corrected-domainAdapt-Stack-HARD50-Adapter-pfeiffer-run3
-  results: []
+- adapter-transformers
+- adapterhub:Arabic ABSA/SemEvalHotelReview
+- bert
+datasets:
+- Hotel
 ---
 
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
+# Adapter `salohnana2018/ABSA-SentencePair-corrected-domainAdapt-Stack-HARD50-Adapter-pfeiffer-run3` for CAMeL-Lab/bert-base-arabic-camelbert-msa
 
-# ABSA-SentencePair-corrected-domainAdapt-Stack-HARD50-Adapter-pfeiffer-run3
+An [adapter](https://adapterhub.ml) for the `CAMeL-Lab/bert-base-arabic-camelbert-msa` model that was trained on the [Arabic ABSA/SemEvalHotelReview](https://adapterhub.ml/explore/Arabic ABSA/SemEvalHotelReview/) dataset and includes a prediction head for classification.
 
-This model is a fine-tuned version of [CAMeL-Lab/bert-base-arabic-camelbert-msa](https://huggingface.co/CAMeL-Lab/bert-base-arabic-camelbert-msa) on the None dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.3168
-- Accuracy: 0.8800
-- F1: 0.8800
-- Precision: 0.8800
-- Recall: 0.8800
+This adapter was created for usage with the **[adapter-transformers](https://github.com/Adapter-Hub/adapter-transformers)** library.
 
-## Model description
+## Usage
 
-More information needed
+First, install `adapter-transformers`:
 
-## Intended uses & limitations
+```
+pip install -U adapter-transformers
+```
+_Note: adapter-transformers is a fork of transformers that acts as a drop-in replacement with adapter support. [More](https://docs.adapterhub.ml/installation.html)_
 
-More information needed
+Now, the adapter can be loaded and activated like this:
 
-## Training and evaluation data
+```python
+from transformers import AutoAdapterModel
 
-More information needed
+model = AutoAdapterModel.from_pretrained("CAMeL-Lab/bert-base-arabic-camelbert-msa")
+adapter_name = model.load_adapter("salohnana2018/ABSA-SentencePair-corrected-domainAdapt-Stack-HARD50-Adapter-pfeiffer-run3", source="hf", set_active=True)
+```
 
-## Training procedure
+## Architecture & Training
 
-### Training hyperparameters
+<!-- Add some description here -->
 
-The following hyperparameters were used during training:
-- learning_rate: 0.0001
-- train_batch_size: 32
-- eval_batch_size: 32
-- seed: 23
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- num_epochs: 20
+## Evaluation results
 
-### Training results
+<!-- Add some description here -->
 
-| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
-| 0.4639        | 1.0   | 265  | 0.3549          | 0.8729   | 0.8729 | 0.8729    | 0.8729 |
-| 0.3677        | 2.0   | 530  | 0.3528          | 0.8634   | 0.8634 | 0.8634    | 0.8634 |
-| 0.3436        | 3.0   | 795  | 0.3237          | 0.8767   | 0.8767 | 0.8767    | 0.8767 |
-| 0.3254        | 4.0   | 1060 | 0.3168          | 0.8800   | 0.8800 | 0.8800    | 0.8800 |
-| 0.3074        | 5.0   | 1325 | 0.3221          | 0.8795   | 0.8795 | 0.8795    | 0.8795 |
-| 0.2954        | 6.0   | 1590 | 0.3295          | 0.8795   | 0.8795 | 0.8795    | 0.8795 |
-| 0.2899        | 7.0   | 1855 | 0.3408          | 0.8748   | 0.8748 | 0.8748    | 0.8748 |
-| 0.2703        | 8.0   | 2120 | 0.3304          | 0.8738   | 0.8738 | 0.8738    | 0.8738 |
-| 0.2543        | 9.0   | 2385 | 0.3327          | 0.8677   | 0.8677 | 0.8677    | 0.8677 |
-| 0.2522        | 10.0  | 2650 | 0.3285          | 0.8757   | 0.8757 | 0.8757    | 0.8757 |
-| 0.2375        | 11.0  | 2915 | 0.3261          | 0.8738   | 0.8738 | 0.8738    | 0.8738 |
-| 0.2313        | 12.0  | 3180 | 0.3363          | 0.8700   | 0.8700 | 0.8700    | 0.8700 |
-| 0.223         | 13.0  | 3445 | 0.3372          | 0.8762   | 0.8762 | 0.8762    | 0.8762 |
-| 0.2176        | 14.0  | 3710 | 0.3591          | 0.8653   | 0.8653 | 0.8653    | 0.8653 |
-| 0.2046        | 15.0  | 3975 | 0.3594          | 0.8715   | 0.8715 | 0.8715    | 0.8715 |
-| 0.205         | 16.0  | 4240 | 0.3686          | 0.8710   | 0.8710 | 0.8710    | 0.8710 |
-| 0.1951        | 17.0  | 4505 | 0.3747          | 0.8729   | 0.8729 | 0.8729    | 0.8729 |
-| 0.1856        | 18.0  | 4770 | 0.3703          | 0.8767   | 0.8767 | 0.8767    | 0.8767 |
-| 0.1947        | 19.0  | 5035 | 0.3767          | 0.8748   | 0.8748 | 0.8748    | 0.8748 |
-| 0.187         | 20.0  | 5300 | 0.3754          | 0.8757   | 0.8757 | 0.8757    | 0.8757 |
+## Citation
 
-
-### Framework versions
-
-- Transformers 4.26.1
-- Pytorch 2.0.1+cu118
-- Datasets 2.14.3
-- Tokenizers 0.13.3
+<!-- Add some description here -->

adapter_config.json

@@ -0,0 +1,40 @@
+{
+  "config": {
+    "adapter_residual_before_ln": false,
+    "cross_adapter": false,
+    "factorized_phm_W": true,
+    "factorized_phm_rule": false,
+    "hypercomplex_nonlinearity": "glorot-uniform",
+    "init_weights": "bert",
+    "inv_adapter": null,
+    "inv_adapter_reduction_factor": null,
+    "is_parallel": false,
+    "learn_phm": true,
+    "leave_out": [],
+    "ln_after": false,
+    "ln_before": false,
+    "mh_adapter": false,
+    "non_linearity": "relu",
+    "original_ln_after": true,
+    "original_ln_before": true,
+    "output_adapter": true,
+    "phm_bias": true,
+    "phm_c_init": "normal",
+    "phm_dim": 4,
+    "phm_init_range": 0.0001,
+    "phm_layer": false,
+    "phm_rank": 1,
+    "reduction_factor": 16,
+    "residual_before_ln": true,
+    "scaling": 1.0,
+    "shared_W_phm": false,
+    "shared_phm_rule": true,
+    "use_gating": false
+  },
+  "hidden_size": 768,
+  "model_class": "BertModelWithHeads",
+  "model_name": "CAMeL-Lab/bert-base-arabic-camelbert-msa",
+  "model_type": "bert",
+  "name": "ABSA_Sentiment_classification_HARD50_pfeiffer_Stack",
+  "version": "3.2.1"
+}

head_config.json

@@ -0,0 +1,21 @@
+{
+  "config": {
+    "activation_function": "tanh",
+    "bias": true,
+    "head_type": "classification",
+    "label2id": {
+      "negative": 0,
+      "neutral": 2,
+      "positive": 1
+    },
+    "layers": 2,
+    "num_labels": 3,
+    "use_pooler": false
+  },
+  "hidden_size": 768,
+  "model_class": "BertModelWithHeads",
+  "model_name": "CAMeL-Lab/bert-base-arabic-camelbert-msa",
+  "model_type": "bert",
+  "name": "ABSA_Sentiment_classification_HARD50_pfeiffer_Stack",
+  "version": "3.2.1"
+}

pytorch_adapter.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42369e60274f1a2d06a6cab4840348aad0706527509c3dc23a6f92278f9b3c4e
+size 3597221

pytorch_model_head.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:590cdc2ba865dda22b3abc7a122b382a1f1536d5b46f299fcec297a9fbb834aa
+size 2373473