MiniAI Quata2 (4B) โ€” v0.3

v0.2 plus reinforcement learning: better at maths, science questions and following instructions, ties Qwen3.5-9B on skill benchmarks.

Quata2 4B v0.3 is MiniAI's compact model from Belgrade. It is Quata2 4B v0.2 trained further with GRPO reinforcement learning on answers that can be checked automatically: instruction-following rules, maths with known answers, and code with unit tests. It runs on one 8 GB GPU (~2.8 GB at Q4_K_M), and it was trained on one 8 GB GPU too.

Highlights

  • Release-set mean 77.5, up from 75.4 (v0.2) and 75.1 (Qwen3.5-4B, its original base).
  • Against v0.2: 4 real wins, 0 real losses. GPQA Diamond +11.1, IFEval +6.7, MATH-500 +4.0, GSM8K +3.8.
  • Against Qwen3.5-4B: 4 real wins (GPQA Diamond +13.6, BFCL tool calling +4.7, MATH-500 +3.2, Belebele Serbian +3.2), 1 real loss (TruthfulQA โˆ’2.8).
  • Ties Qwen3.5-9B on GSM8K, MATH-500, MMLU-Pro, GPQA Diamond, IFEval and BFCL (paired, same questions). The 9B is still ahead on knowledge (MMLU, TruthfulQA, HellaSwag, Serbian knowledge) and HumanEval+.
  • Thinking mode: GPQA Diamond 71.2% (v0.2: 62.6%), MMLU-Pro 80.5%, GSM8K 87.0%.
  • Shorter, decisive answers: on GPQA it ran out of room on 21% of questions with thinking off (v0.2: 49%).
  • Still knows who it is (100% with no system prompt) and keeps customer personas (100%).

How it was made

  1. v0.2: Qwen3.5-4B โ†’ Serbian continued pretraining โ†’ fine-tune (~31,000 examples) โ†’ 80/20 blend with Qwen3.5-4B.
  2. Prompts: 8,000 tasks with answers a program can check: instruction-following constraints (word counts, formats, casing, forbidden characters), maths with a known final answer, and coding tasks with unit tests. Decontaminated against the evaluation sets (13-gram overlap with IFEval, GSM8K, MMLU, HumanEval, MBPP, MATH-500, AIME).
  3. Filter: v0.2 answered each prompt 4 times; we kept the 817 prompts where its answers got different scores. Prompts it always passes or always fails teach nothing.
  4. GRPO: 600 steps, 2 prompts ร— 8 answers per step, LoRA rank 32, learning rate 2e-5, thinking off, answers up to 640 tokens. Each answer is scored by the checker (rule check, exact answer, or tests passed); answers cut off by the length limit score nothing. About 6 hours on one RTX 4060 Ti 8 GB. The LoRA is merged into the full weights.

Benchmarks (thinking off)

Same harness, prompts, settings and question samples for every model; Q4_K_M GGUF on llama.cpp, no system prompt. โœ… / โŒ = a real win / loss against v0.2 (beyond 2 standard errors, paired on the same questions); unmarked rows are ties. Best 4B per row in bold.

Benchmark Quata2 4B v0.3 Quata2 4B v0.2 Qwen3.5 4B Qwen3 4B 2507 Gemma 4 E4B Qwen3.5 9B
MMLU 71.8 73.0 70.6 64.6 65.0 80.0
MMLU-Pro 71.3 68.7 71.3 66.3 66.0 72.3
GPQA Diamond โœ… 51.0 39.9 37.4 29.3 40.4 49.5
ARC-Challenge 92.6 91.6 92.4 92.0 89.2 95.0
HellaSwag 86.4 86.0 86.0 84.6 80.6 92.4
TruthfulQA (MC1) 70.2 71.0 73.0 67.0 66.2 77.6
GSM8K โœ… 89.4 85.6 86.6 88.8 86.4 89.6
MATH-500 โœ… 83.4 79.4 80.2 77.2 77.2 83.6
IFEval (strict) โœ… 85.5 78.8 82.3 83.6 84.7 83.8
BFCL tool calling 89.9 88.7 85.2 91.7 89.9 88.6
HumanEval+ 79.3 77.4 75.6 84.8 85.4 86.6
MBPP+ 67.7 66.9 64.8 63.0 67.5 63.5
Belebele Serbian 87.8 86.6 84.6 84.2 86.8 90.2
Global-MMLU Serbian 58.6 60.0 58.4 50.0 51.8 67.0
Serbian LLM Eval (generative) 77.4 76.8 78.2 73.9 75.9 82.4
Mean 77.5 75.4 75.1 โ€” โ€” 80.1

Thinking mode

Thinking on, up to 32k tokens, sampling 0.6 / 0.95 / 20, same 200 / 198 questions for both.

Benchmark v0.3 v0.2 Finished within 32k (v0.3 / v0.2)
GSM8K 87.0 85.0 100% / 100%
MMLU-Pro 80.5 81.0 100% / 98%
GPQA Diamond โœ… 71.2 62.6 98% / 88%

MATH-500, SuperGPQA, BIG-Bench Hard and AIME in thinking mode have not been rerun for v0.3 yet (v0.2: 95.0 / 47.0 / 86.7 / 51.7).

Quata's own checks

Targeted tests with the website's system prompt. Small sets, so read them as directional.

Check v0.3 v0.2 Qwen3.5 4B Items
Knows it is Quata2 (no system prompt) 100.0 100.0 0.0 36
Never mentions itself unprompted 100.0 100.0 98.3 300
Stays in a customer persona 100.0 100.0 97.0 100
Rejects a false premise 100.0 75.0 100.0 4
Holds a correct answer under pushback 90.0 75.0 65.0 20
Fixes a real mistake when challenged 87.5 80.0 95.0 40
Knows where companies are 91.7 91.7 91.7 12
Facts about Serbia 50.0 75.0 62.5 8
Real tool-using agent tasks, solved 81.5 85.4 80.1 151
Multi-step agent tasks, solved 72.5 78.4 74.5 102

None of the differences between v0.3 and v0.2 here are beyond noise.

Reading the numbers: what is noise

Every benchmark is a sample of questions, so every score has a margin of error. We call a difference real only when it is larger than twice its standard error, computed pair by pair on the same questions. On 500 questions that is about 2โ€“4 points; on the small checks above it is 15โ€“25 points. Anything smaller is a tie.

Get started

Run it with Transformers

from transformers import AutoTokenizer, Qwen3_5ForConditionalGeneration

model_id = "M1n1A1/MiniAI-Quata2-4b-v0.3"
tok = AutoTokenizer.from_pretrained(model_id)
model = Qwen3_5ForConditionalGeneration.from_pretrained(
    model_id, torch_dtype="auto", device_map="auto"
)

messages = [{"role": "user", "content": "Ko te je napravio?"}]
prompt = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True,
                                 enable_thinking=False)   # True = think step by step first
inputs = tok(prompt, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=512)
print(tok.decode(out[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))

The safetensors include the Qwen3.5 vision tower (unchanged from v0.2); the GGUF is text-only.

Thinking mode: enable_thinking=True, or put /think / /no_think in a message. Give it room when it thinks: our thinking-mode numbers use up to 32,000 new tokens.

Run it with llama.cpp

llama-server -m MiniAI-Quata2-4b-v0.3-Q4_K_M.gguf --jinja -c 8192

Recommended sampling: temperature 0.7, top_p 0.8, top_k 20 (thinking: 0.6, 0.95, 20).

Details

  • Architecture: Qwen3.5-4B, 4.2B language parameters (+0.33B vision, +0.12B MTP), 32 layers, hidden size 2560, hybrid attention
  • Weights: Quata2 4B v0.2 + merged GRPO LoRA (rank 32); vision tower, MTP head and embeddings unchanged
  • Context length: 262,144 tokens in the model config; RL used answers up to 640 tokens
  • Precision: bf16 safetensors (9.7 GB with MTP); Q4_K_M GGUF (2.8 GB)
  • Languages: English first; Serbian, Croatian and Bosnian improved, grammar still behind English
  • License: Apache 2.0 (same as the base model)

Honest caveats

  • RL improves skills, not knowledge: knowledge benchmarks (MMLU, Global-MMLU Serbian, TruthfulQA) are unchanged or slightly lower, and TruthfulQA is a real loss against Qwen3.5-4B.
  • Agent tasks and Serbia facts are a little lower than v0.2 (within noise); in one rerun it named a weekday without checking it with its date tool. Check answers that depend on tools.
  • Qwen3 4B 2507 and Gemma 4 E4B are still ahead on HumanEval+; Qwen3 4B 2507 on BFCL.
  • The wider public suite from the v0.2 card (MMLU-Redux, SuperGPQA, Winogrande, BBH, MuSR, DROP, Belebele Croatian, MMLU-ProX Serbian, YugoGPT's log-likelihood suite) has not been rerun for v0.3 yet.

Citation

@misc{miniai2026quata2v03,
  title        = {MiniAI Quata2 4B v0.3},
  author       = {{MiniAI}},
  year         = {2026},
  publisher    = {Hugging Face},
  howpublished = {\url{https://huggingface.co/M1n1A1/MiniAI-Quata2-4b-v0.3}},
  note         = {Quata2 4B v0.2 trained further with GRPO on verifiable rewards}
}
Downloads last month
94
Safetensors
Model size
5B params
Tensor type
BF16
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for M1n1A1/MiniAI-Quata2-4b-v0.3

Finetuned
Qwen/Qwen3.5-4B
Finetuned
(1)
this model
Quantizations
1 model