Spaces:
Running
Running
File size: 2,511 Bytes
4736afd 4cf093f 4736afd a55f0c9 4736afd 4cf093f 4736afd 4cf093f a55f0c9 4cf093f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 |
from datasets import Dataset
from transformers import AutoTokenizer, AutoModelForCausalLM, TrainingArguments, Trainer
from datafile import QL, resume_data_dict
import streamlit as st
# Load your custom data
data = []
for i in range(len(QL['labels'])):
data.append({"question":QL["queries"][i], "answer":resume_data_dict[QL['labels'][i]]})
# Create a Dataset
dataset = Dataset.from_list(data)
# Load the tokenizer
tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-medium")
def preprocess_function(examples):
inputs = [f"Question: {q}" for q in examples["question"]]
model_inputs = tokenizer(inputs, padding="max_length", truncation=True)
# Setup the tokenizer for targets
with tokenizer.as_target_tokenizer():
labels = tokenizer(examples["answer"], padding="max_length", truncation=True)
model_inputs["labels"] = labels["input_ids"]
return model_inputs
# Apply preprocessing
tokenized_dataset = dataset.map(preprocess_function, batched=True)
# Define training arguments
training_args = TrainingArguments(
output_dir="./resume_bot",
num_train_epochs=3,
per_device_train_batch_size=2,
per_device_eval_batch_size=2,
warmup_steps=10,
weight_decay=0.01,
logging_dir="./logs",
logging_steps=10,
save_steps=500,
evaluation_strategy="steps"
)
# Initialize the Trainer
trainer = Trainer(
model=model,
args=training_args,
train_dataset=tokenized_dataset,
eval_dataset=tokenized_dataset,
tokenizer=tokenizer
)
# Train the model
trainer.train()
model_name = "./resume_bot" # Path to your fine-tuned model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
st.title("Resume Chatbot")
if 'history' not in st.session_state:
st.session_state.history = []
user_input = st.text_input("You: ", "")
if user_input:
# Encode the input
input_ids = tokenizer.encode(user_input + tokenizer.eos_token, return_tensors='pt')
# Generate a response
response_ids = model.generate(input_ids, max_length=1000, pad_token_id=tokenizer.eos_token_id)
bot_response = tokenizer.decode(response_ids[0], skip_special_tokens=True)
# Update the chat history
st.session_state.history.append(user_input)
st.session_state.history.append(bot_response)
# Display the bot response
st.write(f"Bot: {bot_response}")
# Add a button to clear the conversation
if st.button("Reset Conversation"):
st.session_state.history = []
|