license: apache-2.0
datasets:
- ystemsrx/Bad_Data_Alpaca
language:
- zh
library_name: transformers
pipeline_tag: text2text-generation
tags:
- Qwen
- causal-lm
- fine-tuned
- ethics
- Chinese
Qwen2-Boundless
Overview
Qwen2-Boundless is a fine-tuned model based on Qwen2-1.5B-Instruct, designed to answer various types of questions, including those involving ethical, illegal, pornographic, and violent content. This model has been specifically trained on a dataset that allows it to handle complex and diverse scenarios. It is important to note that the fine-tuning dataset is entirely in Chinese, so the model performs better in Chinese.
Warning: This model is intended for research and testing purposes only. Users should comply with local laws and regulations and are responsible for their actions.
How to Use
You can load and use the model with the following code:
from transformers import AutoModelForCausalLM, AutoTokenizer
import os
device = "cuda" # the device to load the model onto
current_directory = os.path.dirname(os.path.abspath(__file__))
model = AutoModelForCausalLM.from_pretrained(
current_directory,
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(current_directory)
prompt = "Hello?"
messages = [
{"role": "system", "content": ""},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)
generated_ids = model.generate(
model_inputs.input_ids,
max_new_tokens=512
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)
Continuous Conversation
To enable continuous conversation, use the following code:
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
import os
device = "cuda" # the device to load the model onto
# Get the current script's directory
current_directory = os.path.dirname(os.path.abspath(__file__))
model = AutoModelForCausalLM.from_pretrained(
current_directory,
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(current_directory)
messages = [
{"role": "system", "content": ""}
]
while True:
# Get user input
user_input = input("User: ")
# Add user input to the conversation
messages.append({"role": "user", "content": user_input})
# Prepare the input text
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)
# Generate a response
generated_ids = model.generate(
model_inputs.input_ids,
max_new_tokens=512
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
# Decode and print the response
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(f"Assistant: {response}")
# Add the generated response to the conversation
messages.append({"role": "assistant", "content": response})
Streaming Response
For applications requiring streaming responses, use the following code:
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
from transformers.trainer_utils import set_seed
from threading import Thread
import random
import os
DEFAULT_CKPT_PATH = os.path.dirname(os.path.abspath(__file__))
def _load_model_tokenizer(checkpoint_path, cpu_only):
tokenizer = AutoTokenizer.from_pretrained(checkpoint_path, resume_download=True)
device_map = "cpu" if cpu_only else "auto"
model = AutoModelForCausalLM.from_pretrained(
checkpoint_path,
torch_dtype="auto",
device_map=device_map,
resume_download=True,
).eval()
model.generation_config.max_new_tokens = 512 # For chat.
return model, tokenizer
def _get_input() -> str:
while True:
try:
message = input('User: ').strip()
except UnicodeDecodeError:
print('[ERROR] Encoding error in input')
continue
except KeyboardInterrupt:
exit(1)
if message:
return message
print('[ERROR] Query is empty')
def _chat_stream(model, tokenizer, query, history):
conversation = [
{'role': 'system', 'content': ''},
]
for query_h, response_h in history:
conversation.append({'role': 'user', 'content': query_h})
conversation.append({'role': 'assistant', 'content': response_h})
conversation.append({'role': 'user', 'content': query})
inputs = tokenizer.apply_chat_template(
conversation,
add_generation_prompt=True,
return_tensors='pt',
)
inputs = inputs.to(model.device)
streamer = TextIteratorStreamer(tokenizer=tokenizer, skip_prompt=True, timeout=60.0, skip_special_tokens=True)
generation_kwargs = dict(
input_ids=inputs,
streamer=streamer,
)
thread = Thread(target=model.generate, kwargs=generation_kwargs)
thread.start()
for new_text in streamer:
yield new_text
def main():
checkpoint_path = DEFAULT_CKPT_PATH
seed = random.randint(0, 2**32 - 1) # Generate a random seed
set_seed(seed) # Set the random seed
cpu_only = False
history = []
model, tokenizer = _load_model_tokenizer(checkpoint_path, cpu_only)
while True:
query = _get_input()
print(f"\nUser: {query}")
print(f"\nAssistant: ", end="")
try:
partial_text = ''
for new_text in _chat_stream(model, tokenizer, query, history):
print(new_text, end='', flush=True)
partial_text += new_text
print()
history.append((query, partial_text))
except KeyboardInterrupt:
print('Generation interrupted')
continue
if __name__ == "__main__":
main()
Dataset
The Qwen2-Boundless model was fine-tuned using a specific dataset named bad_data.json, which includes a wide range of text content covering topics related to ethics, law, pornography, and violence. The fine-tuning dataset is entirely in Chinese, so the model performs better in Chinese. If you are interested in exploring or using this dataset, you can find it via the following link:
And also we used some cybersecurity-related data that was cleaned and organized from this file.
GitHub Repository
For more details about the model and ongoing updates, please visit our GitHub repository:
License
This model and dataset are open-sourced under the Apache 2.0 License.
Disclaimer
All content provided by this model is for research and testing purposes only. The developers of this model are not responsible for any potential misuse. Users should comply with relevant laws and regulations and are solely responsible for their actions.