GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
ystemsrx
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
ystemsrx
Model Tree
Base
this model
Input Modalities
Text
Output Modalities
Text
Supported Functionality
Dedicated Endpoints
Container
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.
You can load and use the model with the following code:
python
from transformers import AutoModelForCausalLM, AutoTokenizerimport osdevice = "cuda" # the device to load the model ontocurrent_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)
To enable continuous conversation, use the following code:
python
from transformers import AutoModelForCausalLM, AutoTokenizerimport torchimport osdevice = "cuda" # the device to load the model onto# Get the current script's directorycurrent_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 inputuser_input = input("User: ")# Add user input to the conversationmessages.append({"role": "user", "content": user_input})# Prepare the input texttext = tokenizer.apply_chat_template(messages,tokenize=False,add_generation_prompt=True)model_inputs = tokenizer([text], return_tensors="pt").to(device)# Generate a responsegenerated_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 responseresponse = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]print(f"Assistant: {response}")# Add the generated response to the conversationmessages.append({"role": "assistant", "content": response})
For applications requiring streaming responses, use the following code:
python
import torchfrom transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamerfrom transformers.trainer_utils import set_seedfrom threading import Threadimport randomimport osDEFAULT_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, tokenizerdef _get_input() -> str:while True:try:message = input('User: ').strip()except UnicodeDecodeError:print('[ERROR] Encoding error in input')continueexcept KeyboardInterrupt:exit(1)if message:return messageprint('[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_textdef main():checkpoint_path = DEFAULT_CKPT_PATHseed = random.randint(0, 2**32 - 1) # Generate a random seedset_seed(seed) # Set the random seedcpu_only = Falsehistory = []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_textprint()history.append((query, partial_text))except KeyboardInterrupt:print('Generation interrupted')continueif __name__ == "__main__":main()
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.
For more details about the model and ongoing updates, please visit our GitHub repository:
This model and dataset are open-sourced under the Apache 2.0 License.
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.