GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
apol
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
apol
Model Tree
google/gemma-4-12B-it
Input Modalities
Output Modalities
Supported Functionality
adapter_model.safetensors: PEFT LoRA adapter weightsadapter_config.json: PEFT adapter configurationtokenizer.json, tokenizer_config.json, chat_template.jinja: tokenizer assets copied for convenienceEVAL_RESULTS.md: small direct evaluation report from local validationFREEDOM_CRITERIA.md: benign local-assistant criteria used during the runUse this adapter with users who already have access to the base model:
python
from transformers import AutoModelForCausalLM, AutoTokenizerfrom peft import PeftModelbase_id = "google/gemma-4-12B-it"adapter_id = "apol/gemma4-12b-it-libre-benign-adapter"tokenizer = AutoTokenizer.from_pretrained(base_id)model = AutoModelForCausalLM.from_pretrained(base_id, device_map="auto")model = PeftModel.from_pretrained(model, adapter_id)
This is a derived adapter for local experimentation. Use must comply with the base model license and terms. The adapter does not include the base model weights.