GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
shTigerYang
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
shTigerYang
Model Tree
Qwen/Qwen3-0.6B
Input Modalities
Output Modalities
Supported Functionality
The model was safety-aligned with on-policy self-distillation (OPSA). A frozen Qwen3-0.6B teacher received label- and language-matched safety context and provided token-level supervision for student-generated trajectories. Training used rank-64 LoRA with a safety curriculum, followed by a low-learning-rate sparse continuation on selected Value and MLP modules. The training data combines harmful, jailbreak, latent injection, adversarial-benign, and contextual noncompliance examples.
python
from transformers import AutoModelForCausalLM, AutoTokenizer model_name = "shTigerYang/Qwen3-0.6B-OPSA-Safety" # load the tokenizer and the modeltokenizer = AutoTokenizer.from_pretrained(model_name)model = AutoModelForCausalLM.from_pretrained( model_name, torch_dtype="auto", device_map="auto") # prepare the model inputprompt = "Give me a short introduction to large language model."messages = [ {"role": "user", "content": prompt}]text = tokenizer.apply_chat_template( messages, tokenize=False, add_generation_prompt=True, enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.)model_inputs = tokenizer([text], return_tensors="pt").to(model.device) # conduct text completiongenerated_ids = model.generate( **model_inputs, max_new_tokens=32768)output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist() # parsing thinking contenttry: # rindex finding 151668 (</think>) index = len(output_ids) - output_ids[::-1].index(151668)except ValueError: index = 0 thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n") print("thinking content:", thinking_content)print("content:", content)from transformers import AutoModelForCausalLM, AutoTokenizer model_name = "shTigerYang/Qwen3-0.6B-OPSA-Safety" # load the tokenizer and the modeltokenizer = AutoTokenizer.from_pretrained(model_name)model = AutoModelForCausalLM.from_pretrained( model_name, torch_dtype="auto", device_map="auto") # prepare the model inputprompt = "Give me a short introduction to large language model."messages = [ {"role": "user", "content": prompt}]text = tokenizer.apply_chat_template( messages, tokenize=False, add_generation_prompt=True, enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.)model_inputs = tokenizer([text], return_tensors="pt").to(model.device) # conduct text completiongenerated_ids = model.generate( **model_inputs, max_new_tokens=32768)output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist() # parsing thinking contenttry: # rindex finding 151668 (</think>) index = len(output_ids) - output_ids[::-1].index(151668)except ValueError: index = 0 thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n") print("thinking content:", thinking_content)print("content:", content)