How to use Occupying-Mars/qwen3-4b-symbolic-vanilla-ppo-v1 with Transformers:
# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Occupying-Mars/qwen3-4b-symbolic-vanilla-ppo-v1") model = AutoModelForCausalLM.from_pretrained("Occupying-Mars/qwen3-4b-symbolic-vanilla-ppo-v1", device_map="auto")
