Update app.py
Browse files
app.py
CHANGED
|
@@ -1,3 +1,80 @@
|
|
| 1 |
-
|
|
|
|
|
|
|
|
|
|
| 2 |
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# from docx import Document
|
| 2 |
+
# from transformers import GPT2Tokenizer, GPT2LMHeadModel, Trainer, TrainingArguments
|
| 3 |
+
# import torch
|
| 4 |
+
# import gradio as gr
|
| 5 |
|
| 6 |
+
# # Load the Word document
|
| 7 |
+
# docx_file_path = "Our Leadership.docx"
|
| 8 |
+
# doc = Document(docx_file_path)
|
| 9 |
+
|
| 10 |
+
# # Extract text from the document
|
| 11 |
+
# text = ""
|
| 12 |
+
# for paragraph in doc.paragraphs:
|
| 13 |
+
# text += paragraph.text + "\n"
|
| 14 |
+
|
| 15 |
+
# # Save the extracted text to a text file
|
| 16 |
+
# txt_file_path = "extracted_text.txt"
|
| 17 |
+
# with open(txt_file_path, "w", encoding="utf-8") as file:
|
| 18 |
+
# file.write(text)
|
| 19 |
+
|
| 20 |
+
# # Load the pre-trained GPT-2 model and tokenizer
|
| 21 |
+
# tokenizer = GPT2Tokenizer.from_pretrained("gpt2")
|
| 22 |
+
# model = GPT2LMHeadModel.from_pretrained("gpt2")
|
| 23 |
+
|
| 24 |
+
# # Tokenize the training data
|
| 25 |
+
# input_ids = tokenizer(text, return_tensors="pt", padding=True, truncation=True)["input_ids"]
|
| 26 |
+
|
| 27 |
+
# # Define the training arguments
|
| 28 |
+
# training_args = TrainingArguments(
|
| 29 |
+
# per_device_train_batch_size=4,
|
| 30 |
+
# num_train_epochs=3,
|
| 31 |
+
# logging_dir='./logs',
|
| 32 |
+
# )
|
| 33 |
+
|
| 34 |
+
# # Define a dummy data collator (required by Trainer)
|
| 35 |
+
# class DummyDataCollator:
|
| 36 |
+
# def __call__(self, features):
|
| 37 |
+
# return features
|
| 38 |
+
|
| 39 |
+
# # Define a Trainer instance
|
| 40 |
+
# trainer = Trainer(
|
| 41 |
+
# model=model,
|
| 42 |
+
# args=training_args,
|
| 43 |
+
# data_collator=DummyDataCollator(),
|
| 44 |
+
# train_dataset=input_ids
|
| 45 |
+
# )
|
| 46 |
+
|
| 47 |
+
# # Train the model
|
| 48 |
+
# trainer.train()
|
| 49 |
+
|
| 50 |
+
# # Define the chatbot function
|
| 51 |
+
# def chatbot(input_text):
|
| 52 |
+
# # Tokenize input text
|
| 53 |
+
# input_ids = tokenizer.encode(input_text, return_tensors="pt")
|
| 54 |
+
|
| 55 |
+
# # Generate response from the model
|
| 56 |
+
# output_ids = model.generate(input_ids, max_length=50, pad_token_id=tokenizer.eos_token_id)
|
| 57 |
+
|
| 58 |
+
# # Decode the generated response
|
| 59 |
+
# response = tokenizer.decode(output_ids[0], skip_special_tokens=True)
|
| 60 |
+
|
| 61 |
+
# return response
|
| 62 |
+
|
| 63 |
+
# # Create the Gradio interface
|
| 64 |
+
# chatbot_interface = gr.Interface(chatbot, "textbox", "textbox", title="Chatbot")
|
| 65 |
+
|
| 66 |
+
# # Launch the Gradio interface
|
| 67 |
+
# chatbot_interface.launch()
|
| 68 |
+
|
| 69 |
+
|
| 70 |
+
import os
|
| 71 |
+
|
| 72 |
+
# Get the current working directory
|
| 73 |
+
current_directory = os.getcwd()
|
| 74 |
+
|
| 75 |
+
# Construct the full file path
|
| 76 |
+
docx_file_name = "Our Leadership.docx"
|
| 77 |
+
full_file_path = os.path.join(current_directory, docx_file_name)
|
| 78 |
+
|
| 79 |
+
# Print the file path
|
| 80 |
+
print("File path:", full_file_path)
|