GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
AustinWay
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
AustinWay
Model Tree
Qwen/Qwen3-8B
Input Modalities
Output Modalities
Supported Functionality
Short multi-turn tutoring in phonics, sight words, spelling, and addition or subtraction with small numbers. The tutor asks one thing per turn, explains with concrete objects, and checks the answer before moving on.
Output constraints learned from the training data: sentences of ten words or fewer, tutor turns of four sentences or fewer, vocabulary limited to a per-phonics-stage whitelist, numbers written as digits.
Generated with this adapter on a Q8_0 base, temperature 0.7, using the system prompt in the Use section. Unedited.
markdown
user: can you help me with 3 plus 2assistant: Yes, Max! Plus is add more. You have 3 blocks. Add 2 more blocks. Count all the blocks. That is the number.user: i don't knowassistant: I will show you. Hold up 3 fingers. Now add 2 more fingers. Count all your fingers. 1, 2, 3, 4, 5. That is 3 plus 2.user: okassistant: Let's try one now. Hold up 3 fingers. Add 2 more fingers. Count them all. How many?
60 held-out prompts from the training distribution, one sample each at temperature 0.7.
| Metric | This adapter | Base Qwen3-8B |
|---|---|---|
| Share of sentences at or below grade 1 (target 95) | 94.7 | not scored |
| Working-memory load, slots per sentence (expert first-grade text: 2.0) | 1.3 | 1.4 |
| Corpus checker pass rate (vocabulary, length, structure) | 98% | 13% |
| Mean reply length, characters | 44 | 83 |
The grade-1 sentence score uses the same scorer as the reading-level leaderboard linked below. The best raw model setting on that leaderboard scores 39.6 at grade 1. That leaderboard was scored on ten subjects, and this adapter on its own held-out prompts, so the two numbers are on different prompt sets.
unsloth/qwen3-8b-unsloth-bnb-4bit for QLoRASynthetic dialogues generated by Claude Sonnet 4.5, filtered by a deterministic checker (vocabulary whitelist, sentence and turn length, punctuation, banned words, arithmetic correctness) and an LLM judge for pedagogy (diagnose, explain, verify). Acceptance rate about 10%. The corpus is not included in this release.
PEFT:
python
from transformers import AutoModelForCausalLM, AutoTokenizerfrom peft import PeftModelbase = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-8B", torch_dtype="auto", device_map="auto")tok = AutoTokenizer.from_pretrained("AustinWay/qwen3-8b-first-grade-tutor")model = PeftModel.from_pretrained(base, "AustinWay/qwen3-8b-first-grade-tutor")messages = [{"role": "system", "content": "You help a 1st grader learn. Use simple words and concrete examples they can picture, like counting on fingers or objects. Check they understand, not just that they answered right. The child's name is Max."},{"role": "user", "content": "can you help me with 3 plus 2"},]inputs = tok.apply_chat_template(messages, add_generation_prompt=True, enable_thinking=False, return_tensors="pt").to(model.device)out = model.generate(inputs, max_new_tokens=80, temperature=0.7)print(tok.decode(out[0][inputs.shape[1]:], skip_special_tokens=True))
llama.cpp, with any Qwen3-8B GGUF as the base:
bash
llama-cli -m Qwen3-8B-Q8_0.gguf --lora qwen3-8b-first-grade-tutor-lora.gguf
The system prompt above is the training system prompt. Replace the child's name as needed. Thinking mode was off during training; keep enable_thinking=False.
adapter_model.safetensors, adapter_config.json: PEFT adapterqwen3-8b-first-grade-tutor-lora.gguf: the same adapter converted for llama.cpptokenizer.json, tokenizer_config.json, chat_template.jinja: copied from the base modelApache 2.0, same as the base model.