GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
strfry
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
strfry
Model Tree
swiss-ai/Apertus-8B-Instruct-2509
Input Modalities
Output Modalities
Supported Functionality
The adapter was trained with the instruction in the system role and the source sentence as the only content of the user turn:
python
messages = [ {"role": "system", "content": "Translate to reconstructed neo-prussian:"}, {"role": "user", "content": "I go into the forest."},]⚠️ The exact system prompt matters. Putting the instruction into the user turn (or omitting the system message) takes the model off-distribution and it may appear to "refuse".
The adapter was trained on an int8-quantized base; load the base in int8 for best fidelity (the LoRA deltas were tuned for that representation).
python
import torchfrom transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfigfrom peft import PeftModel base = "swiss-ai/Apertus-8B-Instruct-2509"adapter = "strfry/apertus-8b-prussian-youtube" tok = AutoTokenizer.from_pretrained(adapter)model = AutoModelForCausalLM.from_pretrained( base, device_map="auto", quantization_config=BitsAndBytesConfig(load_in_8bit=True),)model = PeftModel.from_pretrained(model, adapter).eval() messages = [ {"role": "system", "content": "Translate to reconstructed neo-prussian:"}, {"role": "user", "content": "Ich gehe in den Wald"},]inputs = tok.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt", return_dict=True).to(model.device)out = model.generate(**inputs, max_new_tokens=64, do_sample=False)print(tok.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))# -> As ēima en meddjanswiss-ai/Apertus-8B-Instruct-2509 (loaded int8 / bitsandbytes)Sibling adapter (EuroLLM-9B, Tatoeba): strfry/eurollm-9b-prussian-tatoeba.
import torchfrom transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfigfrom peft import PeftModel base = "swiss-ai/Apertus-8B-Instruct-2509"adapter = "strfry/apertus-8b-prussian-youtube" tok = AutoTokenizer.from_pretrained(adapter)model = AutoModelForCausalLM.from_pretrained( base, device_map="auto", quantization_config=BitsAndBytesConfig(load_in_8bit=True),)model = PeftModel.from_pretrained(model, adapter).eval() messages = [ {"role": "system", "content": "Translate to reconstructed neo-prussian:"}, {"role": "user", "content": "Ich gehe in den Wald"},]inputs = tok.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt", return_dict=True).to(model.device)out = model.generate(**inputs, max_new_tokens=64, do_sample=False)print(tok.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))# -> As ēima en meddjan