GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Arisp
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
Arisp
Model Tree
Qwen/Qwen3.5-2B-Base
Input Modalities
Output Modalities
Supported Functionality
python
import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer tok = AutoTokenizer.from_pretrained("Qwen/Qwen3.5-2B-Base")base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3.5-2B-Base", dtype=torch.bfloat16)model = PeftModel.from_pretrained(base, "Arisp/memorize-instruct") ids = tok.apply_chat_template( [{"role": "user", "content": "Name three fruits."}], add_generation_prompt=True, enable_thinking=False, return_tensors="pt", return_dict=True,)["input_ids"]out = model.generate(ids, max_new_tokens=100, eos_token_id=tok.convert_tokens_to_ids("<|im_end|>"))print(tok.decode(out[0, ids.shape[1]:], skip_special_tokens=True))Use the chat template with enable_thinking=False (how it was trained), and stop
on <|im_end|>: the base model's own config only stops on <|endoftext|>.
| base model | this adapter | |
|---|---|---|
| validation loss (200 held-out chats) | 1.361 | 1.250 |
| instruction probes passed (greedy) | 15/19 | 18/19 |
| MMLU-Pro, direct letter, n=500 | 37.4% | 38.0% |
Probes are short prompts with checkable output constraints
(memorize/probes.py). MMLU-Pro is
scored from the letter right after ANSWER: with no reasoning, on label-split C, split seed 0
(memorize/label_split.py).
sft_torch.json (in this repo) has every argument and the validation curve;
train.log is the full training log.
bash
git clone https://github.com/AriSpiesberger/Memorize && cd Memorizegit checkout aad35a9 # the code that trained itpython -m venv .venv && .venv/bin/pip install -r requirements.txt # Windows: .venv\Scripts\pippython -m memorize.sft_data # data/sft/train.jsonl, valid.jsonlpython -m memorize.sft_torch --out adapters/instructThe data files should hash to sha256 92234522d2e41621… (train) and
fdcbf7bca4450523… (valid) if the upstream datasets are unchanged.
import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer tok = AutoTokenizer.from_pretrained("Qwen/Qwen3.5-2B-Base")base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3.5-2B-Base", dtype=torch.bfloat16)model = PeftModel.from_pretrained(base, "Arisp/memorize-instruct") ids = tok.apply_chat_template( [{"role": "user", "content": "Name three fruits."}], add_generation_prompt=True, enable_thinking=False, return_tensors="pt", return_dict=True,)["input_ids"]out = model.generate(ids, max_new_tokens=100, eos_token_id=tok.convert_tokens_to_ids("<|im_end|>"))print(tok.decode(out[0, ids.shape[1]:], skip_special_tokens=True))git clone https://github.com/AriSpiesberger/Memorize && cd Memorizegit checkout aad35a9 # the code that trained itpython -m venv .venv && .venv/bin/pip install -r requirements.txt # Windows: .venv\Scripts\pippython -m memorize.sft_data # data/sft/train.jsonl, valid.jsonlpython -m memorize.sft_torch --out adapters/instruct