Fix all files

app.py

@@ -1,11 +1,3 @@
-# -*- coding: utf-8 -*-
-"""app
-
-Automatically generated by Colab.
-
-Original file is located at
-    https://colab.research.google.com/drive/1Glbl7TT2ZahRqXHGYp9J3zH5U4ZB0Dsd
-"""
 
 import gradio as gr
 from model_utils import load_models, extract_information, predict_tags, extract_4w_qa, generate_why_or_how_question_and_answer
@@ -35,4 +27,4 @@ iface = gr.Interface(
     outputs="text",
     title="Information Extraction Chatbot"
 )
-iface.launch()
+iface.launch()

model_utils.py

@@ -1,5 +1,5 @@
 import torch
-from transformers import BertTokenizer, BertForTokenClassification, pipeline
+from transformers import BertTokenizer, AutoModelForTokenClassification, pipeline
 import pickle # for saving and loading Python objects
 from openai import OpenAI
 import tiktoken
@@ -65,36 +65,61 @@ class BiLSTMForTokenClassification(nn.Module):
 
 # Load custom BiLSTM and pre-trained BERT
 def load_models():
-    bert_model = BertForTokenClassification.from_pretrained("joyinning/chatbot-info-extraction/models/bert-model.pkl")
+    """
+    Loads the pre-trained BERT model from Hugging Face Hub.
+
+    Returns:
+        bert_model: The loaded BERT model.
+    """
+    bert_model = AutoModelForTokenClassification.from_pretrained("joyinning/chatbot-info-extraction/bert-model")
     bert_model.eval()
 
-    with open('joyinning/chatbot-info-extraction/models/bilstm-model.pkl', 'rb') as f:
+    with open('models/bilstm-model.pkl', 'rb') as f:
         bilstm_model = pickle.load(f)
+    bilstm_model.eval()
 
     return bert_model, bilstm_model
 
 def load_custom_model(model_dir, tokenizer_dir, id2label):
-    config = AutoConfig.from_pretrained(model_dir, local_files_only=True)
+    config = AutoConfig.from_pretrained(model_dir)
     config.id2label = id2label
     config.num_labels = len(id2label)
 
     model = BiLSTMForTokenClassification(model_name=config._name_or_path, num_labels=config.num_labels)
     model.config.id2label = id2label
     model.load_state_dict(torch.load(os.path.join(model_dir, 'pytorch_model.bin'), map_location=torch.device('cpu')))
-    tokenizer = AutoTokenizer.from_pretrained(tokenizer_dir, local_files_only=True)
+    tokenizer = AutoTokenizer.from_pretrained(tokenizer_dir)
 
     return model, tokenizer
 
+
+# Load NER model and tokenizer
+with open('models/id2label.pkl', 'rb') as f:
+    id2label_ner = pickle.load(f)
+
 ner_model_dir = "joyinning/chatbot-info-extraction/models/bilstm_ner"
 tokenizer_dir = "joyinning/chatbot-info-extraction/models/tokenizer"
-id2label_ner = {0: 'O', 1: 'I-art', 2: 'B-org', 3: 'B-geo', 4: 'I-per', 5: 'B-eve', 6: 'I-geo', 7: 'B-per', 8: 'I-nat', 9: 'B-art', 10: 'B-tim', 11: 'I-gpe', 12: 'I-tim', 13: 'B-nat', 14: 'B-gpe', 15: 'I-org', 16: 'I-eve'}
 ner_model, ner_tokenizer = load_custom_model(ner_model_dir, tokenizer_dir, id2label_ner)
 
+
 # QA model
 qa_model = pipeline('question-answering', model='deepset/bert-base-cased-squad2')
 
 # Function to extract information
 def extract_information(text, bert_model, bilstm_model, ner_tokenizer, id2label_ner):
+    """
+    Extracts information from the given text using NER tags and generates 'Why' or 'How' questions with answers.
+
+    Args:
+        text: The input text string.
+        bert_model: The pre-trained BERT model for token classification.
+        bilstm_model: The BiLSTM model for NER tag prediction.
+        ner_tokenizer: The tokenizer for the BiLSTM model.
+        id2label_ner: A dictionary mapping numerical label indices to NER tags.
+
+    Returns:
+        A dictionary containing extracted 4W information, generated question, and answer.
+    """
     extracted_info = {}
 
     ner_tags = predict_tags(text, bilstm_model, ner_tokenizer, id2label_ner)
@@ -115,13 +140,13 @@ def predict_tags(sentence, model, tokenizer, label_map):
     Predicts NER tags for a given sentence using the specified model and tokenizer.
 
     Args:
-        sentence: The input sentence as a string.
-        model: The pre-trained model (BiLSTM) for tag prediction.
-        tokenizer: The tokenizer used for converting the sentence into tokens.
-        label_map: A dictionary mapping numerical label indices to their corresponding tags.
+        sentence (str): The input sentence.
+        model (nn.Module): The NER model.
+        tokenizer: The tokenizer used for the model.
+        label_map (dict): A dictionary mapping numerical label indices to their corresponding tags.
 
     Returns:
-        A list of predicted tags for each token in the sentence.
+        list: A list of predicted tags for each token in the sentence.
     """
     tokens = tokenizer.tokenize(tokenizer.decode(tokenizer.encode(sentence)))
     inputs = tokenizer.encode(sentence, return_tensors='pt')
@@ -182,7 +207,7 @@ def count_tokens(text):
     Returns:
         The number of tokens in the text.
     """
-    encoding = tiktoken.encoding_for_model("gpt-3.5-turbo-instruct")
+    encoding = tiktoken.encoding_for_model("gpt-3.5-turbo")
     return len(encoding.encode(text))
 
 def generate_why_or_how_question_and_answer(extracted_info, sentence):
@@ -229,28 +254,3 @@ def generate_why_or_how_question_and_answer(extracted_info, sentence):
     else:
         return None
 
-def get_why_or_how_answer(question, context):
-    """
-    Queries OpenAI's GPT-3.5 model to generate an answer for a given question based on the provided context.
-
-    Args:
-        question (str): The question to be answered.
-        context (str): The text context from which the answer should be extracted.
-
-    Returns:
-        str: The generated answer from GPT-3.5.
-    """
-    prompt = f"Question: {question}\nContext: {context}\nAnswer:"
-
-    response = client.chat.completions.create(
-        model="gpt-3.5-turbo",
-        messages=[
-            {"role": "system", "content": "You are a helpful assistant."},
-            {"role": "user", "content": prompt},
-        ],
-        max_tokens=150,
-        stop=None,
-        temperature=0.5,
-    )
-
-    return response.choices[0].text.strip()

models/bert_ner/checkpoint-2398/config.json

@@ -0,0 +1,64 @@
+{
+  "_name_or_path": "bert-base-cased",
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "O",
+    "1": "I-art",
+    "2": "B-org",
+    "3": "B-geo",
+    "4": "I-per",
+    "5": "B-eve",
+    "6": "I-geo",
+    "7": "B-per",
+    "8": "I-nat",
+    "9": "B-art",
+    "10": "B-tim",
+    "11": "I-gpe",
+    "12": "I-tim",
+    "13": "B-nat",
+    "14": "B-gpe",
+    "15": "I-org",
+    "16": "I-eve"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "B-art": 9,
+    "B-eve": 5,
+    "B-geo": 3,
+    "B-gpe": 14,
+    "B-nat": 13,
+    "B-org": 2,
+    "B-per": 7,
+    "B-tim": 10,
+    "I-art": 1,
+    "I-eve": 16,
+    "I-geo": 6,
+    "I-gpe": 11,
+    "I-nat": 8,
+    "I-org": 15,
+    "I-per": 4,
+    "I-tim": 12,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.41.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 28996
+}

models/bert_ner/checkpoint-2398/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5950d47edadb7380fe3af406651f8a9b5f39aca803a2ff5432d3f1712f06ca5
+size 862028474

models/bert_ner/checkpoint-2398/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2f98313637d9931afab0eea749a2a46ce725c551814b476cad0a8a35f80d8eb
+size 14244

models/bert_ner/checkpoint-2398/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c14b88bb3063f6f00c6d5e076f2a66fd269463117579a024570393667ffec534
+size 1064

models/bert_ner/checkpoint-2398/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

models/bert_ner/checkpoint-2398/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

models/bert_ner/checkpoint-2398/trainer_state.json

@@ -0,0 +1,73 @@
+{
+  "best_metric": 0.09369628131389618,
+  "best_model_checkpoint": "/content/drive/MyDrive/chatbot_info/models/bert_ner/checkpoint-2398",
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2398,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.2085070892410342,
+      "grad_norm": 1.3864336013793945,
+      "learning_rate": 1.8609952738393106e-05,
+      "loss": 0.217,
+      "step": 500
+    },
+    {
+      "epoch": 0.4170141784820684,
+      "grad_norm": 0.8502457737922668,
+      "learning_rate": 1.721990547678621e-05,
+      "loss": 0.1147,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6255212677231026,
+      "grad_norm": 1.1136561632156372,
+      "learning_rate": 1.5829858215179316e-05,
+      "loss": 0.1029,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8340283569641368,
+      "grad_norm": 0.9568729400634766,
+      "learning_rate": 1.4439810953572422e-05,
+      "loss": 0.1039,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9708774583963692,
+      "eval_f1": 0.8271461197438539,
+      "eval_loss": 0.09369628131389618,
+      "eval_precision": 0.8277762839472976,
+      "eval_recall": 0.8265169142652586,
+      "eval_runtime": 32.2988,
+      "eval_samples_per_second": 296.915,
+      "eval_steps_per_second": 18.577,
+      "step": 2398
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 7194,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 897629122271064.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

models/bert_ner/checkpoint-2398/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf5db474328d3cedd3f379ef40c2ebb8edc829ff8069efb550f8b499c28ec8ba
+size 5176

models/bert_ner/checkpoint-4796/config.json

@@ -0,0 +1,64 @@
+{
+  "_name_or_path": "bert-base-cased",
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "O",
+    "1": "I-art",
+    "2": "B-org",
+    "3": "B-geo",
+    "4": "I-per",
+    "5": "B-eve",
+    "6": "I-geo",
+    "7": "B-per",
+    "8": "I-nat",
+    "9": "B-art",
+    "10": "B-tim",
+    "11": "I-gpe",
+    "12": "I-tim",
+    "13": "B-nat",
+    "14": "B-gpe",
+    "15": "I-org",
+    "16": "I-eve"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "B-art": 9,
+    "B-eve": 5,
+    "B-geo": 3,
+    "B-gpe": 14,
+    "B-nat": 13,
+    "B-org": 2,
+    "B-per": 7,
+    "B-tim": 10,
+    "I-art": 1,
+    "I-eve": 16,
+    "I-geo": 6,
+    "I-gpe": 11,
+    "I-nat": 8,
+    "I-org": 15,
+    "I-per": 4,
+    "I-tim": 12,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.41.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 28996
+}

models/bert_ner/checkpoint-4796/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b38c96385df4e3b885915a8bd628a32636b59599af4b891f9f462ba90da839b0
+size 14244

models/bert_ner/checkpoint-4796/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:360f5ac849ec370d5845770daa6d36fecb59b359b5d9c4d86206a09fba16abdc
+size 1064

models/bert_ner/checkpoint-4796/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

models/bert_ner/checkpoint-4796/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

models/bert_ner/checkpoint-4796/trainer_state.json

@@ -0,0 +1,120 @@
+{
+  "best_metric": 0.08877604454755783,
+  "best_model_checkpoint": "/content/drive/MyDrive/chatbot_info/models/bert_ner/checkpoint-4796",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 4796,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.2085070892410342,
+      "grad_norm": 1.3864336013793945,
+      "learning_rate": 1.8609952738393106e-05,
+      "loss": 0.217,
+      "step": 500
+    },
+    {
+      "epoch": 0.4170141784820684,
+      "grad_norm": 0.8502457737922668,
+      "learning_rate": 1.721990547678621e-05,
+      "loss": 0.1147,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6255212677231026,
+      "grad_norm": 1.1136561632156372,
+      "learning_rate": 1.5829858215179316e-05,
+      "loss": 0.1029,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8340283569641368,
+      "grad_norm": 0.9568729400634766,
+      "learning_rate": 1.4439810953572422e-05,
+      "loss": 0.1039,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9708774583963692,
+      "eval_f1": 0.8271461197438539,
+      "eval_loss": 0.09369628131389618,
+      "eval_precision": 0.8277762839472976,
+      "eval_recall": 0.8265169142652586,
+      "eval_runtime": 32.2988,
+      "eval_samples_per_second": 296.915,
+      "eval_steps_per_second": 18.577,
+      "step": 2398
+    },
+    {
+      "epoch": 1.042535446205171,
+      "grad_norm": 1.206861972808838,
+      "learning_rate": 1.3049763691965527e-05,
+      "loss": 0.093,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2510425354462051,
+      "grad_norm": 0.7828037142753601,
+      "learning_rate": 1.1659716430358635e-05,
+      "loss": 0.0807,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4595496246872393,
+      "grad_norm": 0.5838351845741272,
+      "learning_rate": 1.026966916875174e-05,
+      "loss": 0.0777,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6680567139282736,
+      "grad_norm": 0.8385056853294373,
+      "learning_rate": 8.879621907144844e-06,
+      "loss": 0.0758,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8765638031693077,
+      "grad_norm": 0.9494819045066833,
+      "learning_rate": 7.4895746455379494e-06,
+      "loss": 0.076,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9721030659338555,
+      "eval_f1": 0.83379981340797,
+      "eval_loss": 0.08877604454755783,
+      "eval_precision": 0.8278782531980591,
+      "eval_recall": 0.8398066941113299,
+      "eval_runtime": 31.2589,
+      "eval_samples_per_second": 306.793,
+      "eval_steps_per_second": 19.195,
+      "step": 4796
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 7194,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1797436693363512.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

models/bert_ner/checkpoint-4796/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf5db474328d3cedd3f379ef40c2ebb8edc829ff8069efb550f8b499c28ec8ba
+size 5176

models/bert_ner/checkpoint-7194/config.json

@@ -0,0 +1,64 @@
+{
+  "_name_or_path": "bert-base-cased",
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "O",
+    "1": "I-art",
+    "2": "B-org",
+    "3": "B-geo",
+    "4": "I-per",
+    "5": "B-eve",
+    "6": "I-geo",
+    "7": "B-per",
+    "8": "I-nat",
+    "9": "B-art",
+    "10": "B-tim",
+    "11": "I-gpe",
+    "12": "I-tim",
+    "13": "B-nat",
+    "14": "B-gpe",
+    "15": "I-org",
+    "16": "I-eve"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "B-art": 9,
+    "B-eve": 5,
+    "B-geo": 3,
+    "B-gpe": 14,
+    "B-nat": 13,
+    "B-org": 2,
+    "B-per": 7,
+    "B-tim": 10,
+    "I-art": 1,
+    "I-eve": 16,
+    "I-geo": 6,
+    "I-gpe": 11,
+    "I-nat": 8,
+    "I-org": 15,
+    "I-per": 4,
+    "I-tim": 12,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.41.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 28996
+}

models/bert_ner/checkpoint-7194/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5332017e3269dc125b8bb6e7eed0bee419824364cd59fabe94efbd3cd50c9176
+size 862028474

models/bert_ner/checkpoint-7194/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d320c3e3276e4985c96bcae477ba07a12a9e5546aae7dcb0eee2f0ea7260862
+size 14244

models/bert_ner/checkpoint-7194/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4af7cfc18ae1e69cb2c7f43e16f571eff81ca506ca712b084a93e7fe36a8ecb2
+size 1064

models/bert_ner/checkpoint-7194/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

models/bert_ner/checkpoint-7194/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

models/bert_ner/checkpoint-7194/trainer_state.json

@@ -0,0 +1,167 @@
+{
+  "best_metric": 0.08877604454755783,
+  "best_model_checkpoint": "/content/drive/MyDrive/chatbot_info/models/bert_ner/checkpoint-4796",
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 7194,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.2085070892410342,
+      "grad_norm": 1.3864336013793945,
+      "learning_rate": 1.8609952738393106e-05,
+      "loss": 0.217,
+      "step": 500
+    },
+    {
+      "epoch": 0.4170141784820684,
+      "grad_norm": 0.8502457737922668,
+      "learning_rate": 1.721990547678621e-05,
+      "loss": 0.1147,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6255212677231026,
+      "grad_norm": 1.1136561632156372,
+      "learning_rate": 1.5829858215179316e-05,
+      "loss": 0.1029,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8340283569641368,
+      "grad_norm": 0.9568729400634766,
+      "learning_rate": 1.4439810953572422e-05,
+      "loss": 0.1039,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9708774583963692,
+      "eval_f1": 0.8271461197438539,
+      "eval_loss": 0.09369628131389618,
+      "eval_precision": 0.8277762839472976,
+      "eval_recall": 0.8265169142652586,
+      "eval_runtime": 32.2988,
+      "eval_samples_per_second": 296.915,
+      "eval_steps_per_second": 18.577,
+      "step": 2398
+    },
+    {
+      "epoch": 1.042535446205171,
+      "grad_norm": 1.206861972808838,
+      "learning_rate": 1.3049763691965527e-05,
+      "loss": 0.093,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2510425354462051,
+      "grad_norm": 0.7828037142753601,
+      "learning_rate": 1.1659716430358635e-05,
+      "loss": 0.0807,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4595496246872393,
+      "grad_norm": 0.5838351845741272,
+      "learning_rate": 1.026966916875174e-05,
+      "loss": 0.0777,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6680567139282736,
+      "grad_norm": 0.8385056853294373,
+      "learning_rate": 8.879621907144844e-06,
+      "loss": 0.0758,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8765638031693077,
+      "grad_norm": 0.9494819045066833,
+      "learning_rate": 7.4895746455379494e-06,
+      "loss": 0.076,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9721030659338555,
+      "eval_f1": 0.83379981340797,
+      "eval_loss": 0.08877604454755783,
+      "eval_precision": 0.8278782531980591,
+      "eval_recall": 0.8398066941113299,
+      "eval_runtime": 31.2589,
+      "eval_samples_per_second": 306.793,
+      "eval_steps_per_second": 19.195,
+      "step": 4796
+    },
+    {
+      "epoch": 2.085070892410342,
+      "grad_norm": 0.8878109455108643,
+      "learning_rate": 6.099527383931054e-06,
+      "loss": 0.0715,
+      "step": 5000
+    },
+    {
+      "epoch": 2.293577981651376,
+      "grad_norm": 1.7333062887191772,
+      "learning_rate": 4.70948012232416e-06,
+      "loss": 0.0607,
+      "step": 5500
+    },
+    {
+      "epoch": 2.5020850708924103,
+      "grad_norm": 0.9545063376426697,
+      "learning_rate": 3.319432860717265e-06,
+      "loss": 0.0615,
+      "step": 6000
+    },
+    {
+      "epoch": 2.7105921601334444,
+      "grad_norm": 1.2574018239974976,
+      "learning_rate": 1.9293855991103697e-06,
+      "loss": 0.0616,
+      "step": 6500
+    },
+    {
+      "epoch": 2.9190992493744785,
+      "grad_norm": 0.9615433812141418,
+      "learning_rate": 5.393383375034752e-07,
+      "loss": 0.0588,
+      "step": 7000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9728451329975679,
+      "eval_f1": 0.838597894175663,
+      "eval_loss": 0.08889511227607727,
+      "eval_precision": 0.8326422584913983,
+      "eval_recall": 0.844639341328083,
+      "eval_runtime": 30.9059,
+      "eval_samples_per_second": 310.297,
+      "eval_steps_per_second": 19.414,
+      "step": 7194
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 7194,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2693332448203128.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

models/bert_ner/checkpoint-7194/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf5db474328d3cedd3f379ef40c2ebb8edc829ff8069efb550f8b499c28ec8ba
+size 5176

models/bert_ner/config.json

@@ -0,0 +1,64 @@
+{
+  "_name_or_path": "bert-base-cased",
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "O",
+    "1": "I-org",
+    "2": "I-per",
+    "3": "B-geo",
+    "4": "B-tim",
+    "5": "I-art",
+    "6": "B-per",
+    "7": "B-gpe",
+    "8": "B-org",
+    "9": "I-tim",
+    "10": "B-eve",
+    "11": "B-art",
+    "12": "I-nat",
+    "13": "B-nat",
+    "14": "I-gpe",
+    "15": "I-eve",
+    "16": "I-geo"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "B-art": 11,
+    "B-eve": 10,
+    "B-geo": 3,
+    "B-gpe": 7,
+    "B-nat": 13,
+    "B-org": 8,
+    "B-per": 6,
+    "B-tim": 4,
+    "I-art": 5,
+    "I-eve": 15,
+    "I-geo": 16,
+    "I-gpe": 14,
+    "I-nat": 12,
+    "I-org": 1,
+    "I-per": 2,
+    "I-tim": 9,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.41.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 28996
+}

models/bert_ner/runs/Jul06_21-47-25_04df247716ce/events.out.tfevents.1720302447.04df247716ce.1751.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e90aa08f9b37741cb5e450f80657cf1766abe7ec6a8b65f9841e96ab326c3a3e
+size 10255

models/bert_ner/runs/Jul07_02-48-39_3cedd31b78f5/events.out.tfevents.1720320522.3cedd31b78f5.1699.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5341fa0c174ff3925dc2c323ac28e1fc000e64d58e72ba58d80888e465ec3e2f
+size 10255

models/bert_ner/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

models/bert_ner/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

models/bert_ner/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ae34742087a4014c8fd9ba1b02b3aac236943c8376ad44fe74bf1db71ba9cbe
+size 5176

models/bilstm_ner/checkpoint-11990/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1120cd6bc3f391f251ab690fbc8b73d6d6ed110a0f1cbad1bcebc7851bc89d0e
+size 14244

models/bilstm_ner/checkpoint-11990/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:372e9f9c1f96c16718abe3028904d44d39283f17ede1f127a4aa086234e16a5e
+size 1064

models/bilstm_ner/checkpoint-11990/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

models/bilstm_ner/checkpoint-11990/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

models/bilstm_ner/checkpoint-11990/trainer_state.json

@@ -0,0 +1,254 @@
+{
+  "best_metric": 0.09195627272129059,
+  "best_model_checkpoint": "/content/drive/MyDrive/chatbot_info/models/bilstm_ner/checkpoint-7194",
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 11990,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.2085070892410342,
+      "grad_norm": 1.1974992752075195,
+      "learning_rate": 1.9165971643035866e-05,
+      "loss": 0.3538,
+      "step": 500
+    },
+    {
+      "epoch": 0.4170141784820684,
+      "grad_norm": 0.7257668972015381,
+      "learning_rate": 1.8331943286071728e-05,
+      "loss": 0.1415,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6255212677231026,
+      "grad_norm": 0.9772794842720032,
+      "learning_rate": 1.7497914929107593e-05,
+      "loss": 0.121,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8340283569641368,
+      "grad_norm": 0.7119922637939453,
+      "learning_rate": 1.6663886572143454e-05,
+      "loss": 0.1167,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9700635783910071,
+      "eval_f1": 0.8221168895049858,
+      "eval_loss": 0.1039084866642952,
+      "eval_precision": 0.8215658235767271,
+      "eval_recall": 0.8226686951852514,
+      "eval_runtime": 33.3276,
+      "eval_samples_per_second": 287.75,
+      "eval_steps_per_second": 18.003,
+      "step": 2398
+    },
+    {
+      "epoch": 1.042535446205171,
+      "grad_norm": 0.8703278303146362,
+      "learning_rate": 1.5829858215179316e-05,
+      "loss": 0.1042,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2510425354462051,
+      "grad_norm": 0.5966465473175049,
+      "learning_rate": 1.499582985821518e-05,
+      "loss": 0.0928,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4595496246872393,
+      "grad_norm": 0.6188527345657349,
+      "learning_rate": 1.4161801501251045e-05,
+      "loss": 0.0894,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6680567139282736,
+      "grad_norm": 0.8101356625556946,
+      "learning_rate": 1.3327773144286907e-05,
+      "loss": 0.087,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8765638031693077,
+      "grad_norm": 0.9334137439727783,
+      "learning_rate": 1.249374478732277e-05,
+      "loss": 0.0865,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9718397518144737,
+      "eval_f1": 0.8323142155768801,
+      "eval_loss": 0.09386646002531052,
+      "eval_precision": 0.8302021191345383,
+      "eval_recall": 0.8344370860927153,
+      "eval_runtime": 41.7135,
+      "eval_samples_per_second": 229.901,
+      "eval_steps_per_second": 14.384,
+      "step": 4796
+    },
+    {
+      "epoch": 2.085070892410342,
+      "grad_norm": 0.7590944766998291,
+      "learning_rate": 1.1659716430358635e-05,
+      "loss": 0.0819,
+      "step": 5000
+    },
+    {
+      "epoch": 2.293577981651376,
+      "grad_norm": 1.2667491436004639,
+      "learning_rate": 1.0825688073394496e-05,
+      "loss": 0.0709,
+      "step": 5500
+    },
+    {
+      "epoch": 2.5020850708924103,
+      "grad_norm": 0.7721763849258423,
+      "learning_rate": 9.99165971643036e-06,
+      "loss": 0.0715,
+      "step": 6000
+    },
+    {
+      "epoch": 2.7105921601334444,
+      "grad_norm": 0.8234202861785889,
+      "learning_rate": 9.157631359466222e-06,
+      "loss": 0.071,
+      "step": 6500
+    },
+    {
+      "epoch": 2.9190992493744785,
+      "grad_norm": 0.7711541056632996,
+      "learning_rate": 8.323603002502085e-06,
+      "loss": 0.0686,
+      "step": 7000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9727254447614853,
+      "eval_f1": 0.8379830958274792,
+      "eval_loss": 0.09195627272129059,
+      "eval_precision": 0.8352820877606366,
+      "eval_recall": 0.840701628781099,
+      "eval_runtime": 39.324,
+      "eval_samples_per_second": 243.871,
+      "eval_steps_per_second": 15.258,
+      "step": 7194
+    },
+    {
+      "epoch": 3.127606338615513,
+      "grad_norm": 0.6545294523239136,
+      "learning_rate": 7.4895746455379494e-06,
+      "loss": 0.062,
+      "step": 7500
+    },
+    {
+      "epoch": 3.336113427856547,
+      "grad_norm": 0.6319155097007751,
+      "learning_rate": 6.655546288573812e-06,
+      "loss": 0.0584,
+      "step": 8000
+    },
+    {
+      "epoch": 3.5446205170975813,
+      "grad_norm": 0.9260782599449158,
+      "learning_rate": 5.821517931609675e-06,
+      "loss": 0.0586,
+      "step": 8500
+    },
+    {
+      "epoch": 3.7531276063386154,
+      "grad_norm": 0.8304509520530701,
+      "learning_rate": 4.987489574645538e-06,
+      "loss": 0.0577,
+      "step": 9000
+    },
+    {
+      "epoch": 3.96163469557965,
+      "grad_norm": 0.9533994793891907,
+      "learning_rate": 4.153461217681402e-06,
+      "loss": 0.0552,
+      "step": 9500
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9727972577031349,
+      "eval_f1": 0.8390449750138812,
+      "eval_loss": 0.09282902628183365,
+      "eval_precision": 0.8329585042113155,
+      "eval_recall": 0.845221048863433,
+      "eval_runtime": 36.8085,
+      "eval_samples_per_second": 260.538,
+      "eval_steps_per_second": 16.301,
+      "step": 9592
+    },
+    {
+      "epoch": 4.170141784820684,
+      "grad_norm": 0.8649327754974365,
+      "learning_rate": 3.319432860717265e-06,
+      "loss": 0.0499,
+      "step": 10000
+    },
+    {
+      "epoch": 4.378648874061718,
+      "grad_norm": 1.5371888875961304,
+      "learning_rate": 2.4854045037531278e-06,
+      "loss": 0.0502,
+      "step": 10500
+    },
+    {
+      "epoch": 4.587155963302752,
+      "grad_norm": 0.7297742366790771,
+      "learning_rate": 1.6513761467889911e-06,
+      "loss": 0.0475,
+      "step": 11000
+    },
+    {
+      "epoch": 4.795663052543786,
+      "grad_norm": 0.8624579310417175,
+      "learning_rate": 8.173477898248542e-07,
+      "loss": 0.0469,
+      "step": 11500
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.97306535935196,
+      "eval_f1": 0.8401763448521553,
+      "eval_loss": 0.09531266242265701,
+      "eval_precision": 0.8361549370678958,
+      "eval_recall": 0.8442366207266869,
+      "eval_runtime": 38.8622,
+      "eval_samples_per_second": 246.769,
+      "eval_steps_per_second": 15.439,
+      "step": 11990
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 11990,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

models/bilstm_ner/checkpoint-11990/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3894ce34a7e6a861216b2ff8719fcdd8a22cf7380c7ed3b53c563a20103b82a5
+size 5176

models/bilstm_ner/checkpoint-2398/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b68f00ba893f31af22b9d31607ee2aea9068594d743b9d2a05eb115093f0726
+size 14244

models/bilstm_ner/checkpoint-2398/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8667350c70351cc7a49cc64441e72314afc7bb9f732b672a56f8eedb03ba4ead
+size 1064