update

app.py

@@ -8,23 +8,20 @@
 # Copyright (C) 2023. All Rights Reserved.
 
 import random
+
+import autocuda
 import gradio as gr
 import pandas as pd
 from pyabsa import (
     download_all_available_datasets,
-    AspectTermExtraction as ATEPC,
     TaskCodeOption,
     available_checkpoints,
 )
-from pyabsa import AspectSentimentTripletExtraction as ASTE
+from pyabsa import ABSAInstruction
 from pyabsa.utils.data_utils.dataset_manager import detect_infer_dataset
 
 download_all_available_datasets()
 
-atepc_dataset_items = {dataset.name: dataset for dataset in ATEPC.ATEPCDatasetList()}
-aste_dataset_items = {dataset.name: dataset for dataset in ASTE.ASTEDatasetList()}
-
-
 def get_atepc_example(dataset):
     task = TaskCodeOption.Aspect_Polarity_Classification
     dataset_file = detect_infer_dataset(atepc_dataset_items[dataset], task)
@@ -66,18 +63,65 @@ def get_aste_example(dataset):
         return sorted(set(lines), key=lines.index)
 
 
-available_checkpoints("ASTE", True)
+def get_acos_example(dataset):
+    task = 'ACOS'
+    dataset_file = detect_infer_dataset(acos_dataset_items[dataset], task)
 
-atepc_dataset_dict = {
-    dataset.name: get_atepc_example(dataset.name)
-    for dataset in ATEPC.ATEPCDatasetList()
-}
-aspect_extractor = ATEPC.AspectExtractor(checkpoint="multilingual")
+    for fname in dataset_file:
+        lines = []
+        if isinstance(fname, str):
+            fname = [fname]
 
-aste_dataset_dict = {
-    dataset.name: get_aste_example(dataset.name) for dataset in ASTE.ASTEDatasetList()
-}
-triplet_extractor = ASTE.AspectSentimentTripletExtractor(checkpoint="multilingual")
+        for f in fname:
+            print("loading: {}".format(f))
+            fin = open(f, "r", encoding="utf-8")
+            lines.extend(fin.readlines())
+            fin.close()
+        lines = [line.split('####')[0] for line in lines]
+        return sorted(set(lines), key=lines.index)
+
+
+try:
+    from pyabsa import AspectTermExtraction as ATEPC
+    atepc_dataset_items = {dataset.name: dataset for dataset in ATEPC.ATEPCDatasetList()}
+    atepc_dataset_dict = {
+        dataset.name: get_atepc_example(dataset.name)
+        for dataset in ATEPC.ATEPCDatasetList()
+    }
+    aspect_extractor = ATEPC.AspectExtractor(checkpoint="multilingual")
+except Exception as e:
+    print(e)
+    atepc_dataset_items = {}
+    atepc_dataset_dict = {}
+    aspect_extractor = None
+
+try:
+    from pyabsa import AspectSentimentTripletExtraction as ASTE
+
+    aste_dataset_items = {dataset.name: dataset for dataset in ASTE.ASTEDatasetList()}
+    aste_dataset_dict = {
+        dataset.name: get_aste_example(dataset.name) for dataset in ASTE.ASTEDatasetList()
+    }
+    triplet_extractor = ASTE.AspectSentimentTripletExtractor(checkpoint="multilingual")
+except Exception as e:
+    print(e)
+    aste_dataset_items = {}
+    aste_dataset_dict = {}
+    triplet_extractor = None
+
+try:
+    from pyabsa import ABSAInstruction
+
+    acos_dataset_items = {dataset.name: dataset for dataset in ABSAInstruction.ACOSDatasetList()[:-1]}
+    acos_dataset_dict = {
+        dataset.name: get_acos_example(dataset.name) for dataset in ABSAInstruction.ACOSDatasetList()[:-1]
+    }
+    quadruple_extractor = ABSAInstruction.ABSAGenerator(checkpoint="multilingual", device=autocuda.auto_cuda())
+except Exception as e:
+    print(e)
+    acos_dataset_items = {}
+    acos_dataset_dict = {}
+    quadruple_extractor = None
 
 
 def perform_atepc_inference(text, dataset):
@@ -113,67 +157,121 @@ def perform_aste_inference(text, dataset):
     return pred_triplets, true_triplets, "{}".format(text)
 
 
+def perform_acos_inference(text, dataset):
+    if not text:
+        text = acos_dataset_dict[dataset][
+            random.randint(0, len(acos_dataset_dict[dataset]) - 1)
+        ]
+
+    raw_output = quadruple_extractor.predict(text)
+    outputs = raw_output[0].strip().split(', ')
+    data = {}
+    for output in outputs:
+        for sub_output in output.split('|'):
+            if 'aspect' in sub_output:
+                data['aspect'] = sub_output.split(':')[1]
+            elif 'opinion' in sub_output:
+                data['opinion'] = sub_output.split(':')[1]
+            elif 'sentiment' in sub_output:
+                data['sentiment'] = sub_output.split(':')[1]
+            elif 'polarity' in sub_output:
+                data['polarity'] = sub_output.split(':')[1]
+            elif 'category' in sub_output:
+                try:
+                    data['category'] = sub_output.split(':')[1]
+                except:
+                    data['category'] = ''
+
+    result = pd.DataFrame.from_dict(data, orient='index').T
+    return result, text
+
 demo = gr.Blocks()
 
 with demo:
-    with gr.Row():
 
-        with gr.Column():
-            gr.Markdown("# <p align='center'>Aspect Sentiment Triplet Extraction !</p>")
 
+    with gr.Row():
+        if triplet_extractor:
+            with gr.Column():
+                gr.Markdown("# <p align='center'>Aspect Sentiment Triplet Extraction !</p>")
+
+                with gr.Row():
+                    with gr.Column():
+                        aste_input_sentence = gr.Textbox(
+                            placeholder="Leave this box blank and choose a dataset will give you a random example...",
+                            label="Example:",
+                        )
+                        gr.Markdown(
+                            "You can find code and dataset at [ASTE examples](https://github.com/yangheng95/PyABSA/tree/v2/examples-v2/aspect_sentiment_triplet_extration)"
+                        )
+                        aste_dataset_ids = gr.Radio(
+                            choices=[dataset.name for dataset in ASTE.ASTEDatasetList()[:-1]],
+                            value="Restaurant14",
+                            label="Datasets",
+                        )
+                        aste_inference_button = gr.Button("Let's go!")
+
+                        aste_output_text = gr.TextArea(label="Example:")
+                        aste_output_pred_df = gr.DataFrame(label="Predicted Triplets:")
+                        aste_output_true_df = gr.DataFrame(label="Original Triplets:")
+
+                        aste_inference_button.click(
+                            fn=perform_aste_inference,
+                            inputs=[aste_input_sentence, aste_dataset_ids],
+                            outputs=[aste_output_pred_df, aste_output_true_df, aste_output_text],
+                        )
+        if aspect_extractor:
+            with gr.Column():
+                gr.Markdown(
+                    "# <p align='center'>Multilingual Aspect-based Sentiment Analysis !</p>"
+                )
+                with gr.Row():
+                    with gr.Column():
+                        atepc_input_sentence = gr.Textbox(
+                            placeholder="Leave this box blank and choose a dataset will give you a random example...",
+                            label="Example:",
+                        )
+                        gr.Markdown(
+                            "You can find the datasets at [github.com/yangheng95/ABSADatasets](https://github.com/yangheng95/ABSADatasets/tree/v1.2/datasets/text_classification)"
+                        )
+                        atepc_dataset_ids = gr.Radio(
+                            choices=[dataset.name for dataset in ATEPC.ATEPCDatasetList()[:-1]],
+                            value="Laptop14",
+                            label="Datasets",
+                        )
+                        atepc_inference_button = gr.Button("Let's go!")
+
+                        atepc_output_text = gr.TextArea(label="Example:")
+                        atepc_output_df = gr.DataFrame(label="Prediction Results:")
+
+                        atepc_inference_button.click(
+                            fn=perform_atepc_inference,
+                            inputs=[atepc_input_sentence, atepc_dataset_ids],
+                            outputs=[atepc_output_df, atepc_output_text],
+                        )
+        if quadruple_extractor:
             with gr.Row():
                 with gr.Column():
-                    aste_input_sentence = gr.Textbox(
-                        placeholder="Leave this box blank and choose a dataset will give you a random example...",
-                        label="Example:",
-                    )
-                    gr.Markdown(
-                        "You can find code and dataset at [ASTE examples](https://github.com/yangheng95/PyABSA/tree/v2/examples-v2/aspect_sentiment_triplet_extration)"
-                    )
-                    aste_dataset_ids = gr.Radio(
-                        choices=[dataset.name for dataset in ASTE.ASTEDatasetList()[:-1]],
-                        value="Restaurant14",
-                        label="Datasets",
-                    )
-                    aste_inference_button = gr.Button("Let's go!")
-
-                    aste_output_text = gr.TextArea(label="Example:")
-                    aste_output_pred_df = gr.DataFrame(label="Predicted Triplets:")
-                    aste_output_true_df = gr.DataFrame(label="Original Triplets:")
-
-                    aste_inference_button.click(
-                        fn=perform_aste_inference,
-                        inputs=[aste_input_sentence, aste_dataset_ids],
-                        outputs=[aste_output_pred_df, aste_output_true_df, aste_output_text],
-                    )
+                    gr.Markdown("# <p align='center'>Aspect Category Opinion Sentiment Extraction !</p>")
 
-        with gr.Column():
-            gr.Markdown(
-                "# <p align='center'>Multilingual Aspect-based Sentiment Analysis !</p>"
-            )
-            with gr.Row():
-                with gr.Column():
-                    atepc_input_sentence = gr.Textbox(
+                    acos_input_sentence = gr.Textbox(
                         placeholder="Leave this box blank and choose a dataset will give you a random example...",
                         label="Example:",
                     )
-                    gr.Markdown(
-                        "You can find the datasets at [github.com/yangheng95/ABSADatasets](https://github.com/yangheng95/ABSADatasets/tree/v1.2/datasets/text_classification)"
-                    )
-                    atepc_dataset_ids = gr.Radio(
-                        choices=[dataset.name for dataset in ATEPC.ATEPCDatasetList()[:-1]],
-                        value="Laptop14",
+                    acos_dataset_ids = gr.Radio(
+                        choices=[dataset.name for dataset in ABSAInstruction.ACOSDatasetList()],
+                        value="Restaurant16",
                         label="Datasets",
                     )
-                    atepc_inference_button = gr.Button("Let's go!")
+                    acos_inference_button = gr.Button("Let's go!")
 
-                    atepc_output_text = gr.TextArea(label="Example:")
-                    atepc_output_df = gr.DataFrame(label="Prediction Results:")
+                    acos_output_text = gr.TextArea(label="Example:")
+                    acos_output_pred_df = gr.DataFrame(label="Predicted Triplets:")
 
-                    atepc_inference_button.click(
-                        fn=perform_atepc_inference,
-                        inputs=[atepc_input_sentence, atepc_dataset_ids],
-                        outputs=[atepc_output_df, atepc_output_text],
+                    acos_inference_button.click(
+                        fn=perform_acos_inference,
+                        inputs=[acos_input_sentence, acos_dataset_ids],
+                        outputs=[acos_output_pred_df, acos_output_text],
                     )
     gr.Markdown(
         """### GitHub Repo: [PyABSA V2](https://github.com/yangheng95/PyABSA)

checkpoints-v2.0.json

@@ -192,6 +192,18 @@
         "Author": "H, Yang ([email protected])"
       }
     },
+    "ACOS": {
+      "multilingual": {
+        "id": "",
+        "Training Model": "DeBERTa-v3-Base",
+        "Training Dataset": "SemEval + Synthetic + Chinese_Zhang datasets",
+        "Language": "Multilingual",
+        "Description": "Trained on RTX3090",
+        "Available Version": "2.1.8+",
+        "Checkpoint File": "multilingual-acos.zip",
+        "Author": "H, Yang ([email protected])"
+      }
+    },
     "UPPERTASKCODE": {
       "promise": {
         "id": "",

checkpoints.json

@@ -0,0 +1 @@
+{"2.0.0": {"APC": {"multilingual": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_bert_Multilingual_acc_87.18_f1_83.11.zip", "Author": "H, Yang ([email protected])"}, "multilingual2": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_bert_Multilingual_acc_82.66_f1_82.06.zip", "Author": "H, Yang ([email protected])"}, "english": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.English", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lsa_t_v2_English_acc_82.21_f1_81.81.zip", "Author": "H, Yang ([email protected])"}, "chinese": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.Chinese", "Language": "Chinese", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lsa_t_v2_Chinese_acc_96.0_f1_95.1.zip", "Author": "H, Yang ([email protected])"}}, "ATEPC": {"multilingual": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ABSADatasets.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "fast_lcf_atepc_Multilingual_cdw_apcacc_85.1_apcf1_80.2_atef1_76.45.zip", "Author": "H, Yang ([email protected])"}, "multilingual-original": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ABSADatasets.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "fast_lcf_atepc_Multilingual_cdw_apcacc_80.81_apcf1_73.75_atef1_76.01.zip", "Author": "H, Yang ([email protected])"}, "multilingual2": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ABSADatasets.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "fast_lcf_atepc_Multilingual_cdw_apcacc_78.08_apcf1_77.81_atef1_75.41.zip", "Author": "H, Yang ([email protected])"}, "english": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ATEPCDatasetList.English", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_atepc_English_cdw_apcacc_82.36_apcf1_81.89_atef1_75.43.zip", "Author": "H, Yang ([email protected])"}, "chinese": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ATEPCDatasetList.Chinese", "Language": "Chinese", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_atepc_Chinese_cdw_apcacc_96.22_apcf1_95.32_atef1_78.73.zip", "Author": "H, Yang ([email protected])"}}, "RNAC": {"degrad_lstm": {"id": "", "Training Model": "LSTM", "Training Dataset": "ABSADatasets.Multilingual", "Language": "RNA", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "lstm_degrad_acc_85.26_f1_84.62.zip", "Author": "H, Yang ([email protected])"}, "degrad_bert": {"id": "", "Training Model": "MLP", "Training Dataset": "Degrad", "Language": "RNA", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "bert_mlp_degrad_acc_87.44_f1_86.99.zip", "Author": "H, Yang ([email protected])"}}, "TAD": {"tad-sst2": {"id": "", "Training Model": "TAD", "Training Dataset": "SST2", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.15+", "Checkpoint File": "TAD-SST2.zip", "Author": "H, Yang ([email protected])"}, "tad-agnews10k": {"id": "", "Training Model": "TAD", "Training Dataset": "AGNews", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.15+", "Checkpoint File": "TAD-AGNews10K.zip", "Author": "H, Yang ([email protected])"}, "tad-amazon": {"id": "", "Training Model": "TAD", "Training Dataset": "AGNews", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.15+", "Checkpoint File": "TAD-Amazon.zip", "Author": "H, Yang ([email protected])"}}, "CDD": {"promise": {"id": "", "Training Model": "CodeT5-small", "Training Dataset": "Promise", "Language": "Code", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "bert_mlp_all_cpdp_acc_75.33_f1_73.52.zip", "Author": "H, Yang ([email protected])"}}, "ASTE": {"english1": {"id": "", "Training Model": "DeBERTa-v3-Base", "Training Dataset": "SemEval", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "2.1.1+", "Checkpoint File": "EMCGCN_SemEval_f1_74.01.zip", "Author": "H, Yang ([email protected])"}, "english": {"id": "", "Training Model": "DeBERTa-v3-Base", "Training Dataset": "SemEval", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "2.1.1+", "Checkpoint File": "ASTE-EMCGCN_SemEval_f1_74.71.zip", "Author": "H, Yang ([email protected])"}, "multilingual": {"id": "", "Training Model": "DeBERTa-v3-Base", "Training Dataset": "SemEval + Synthetic + Chinese_Zhang datasets", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "2.1.1+", "Checkpoint File": "EMCGCN-Multilingual-f1_51.95.zip", "Author": "H, Yang ([email protected])"}}, "ACOS": {"multilingual": {"id": "", "Training Model": "DeBERTa-v3-Base", "Training Dataset": "SemEval + Synthetic + Chinese_Zhang datasets", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "2.1.8+", "Checkpoint File": "ACOS.zip", "Author": "H, Yang ([email protected])"}}, "UPPERTASKCODE": {"promise": {"id": "", "Training Model": "CodeT5-small", "Training Dataset": "DatasetName", "Language": "", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "lstm_degrad_acc_85.26_f1_84.62.zip", "Author": "H, Yang ([email protected])"}}}}

checkpoints/Multilingual/ACOS/multilingual-acos.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b7e6f53b721579e10fab9d82ff085caf051a6917dcd7d2ec9a4d00a8c44c8d0
+size 882150443