Instructions to use Anserwise/AWAXIS-Think-31B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Anserwise/AWAXIS-Think-31B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Anserwise/AWAXIS-Think-31B") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] pipe(text=messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("Anserwise/AWAXIS-Think-31B") model = AutoModelForMultimodalLM.from_pretrained("Anserwise/AWAXIS-Think-31B", device_map="auto") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=256) print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Anserwise/AWAXIS-Think-31B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Anserwise/AWAXIS-Think-31B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Anserwise/AWAXIS-Think-31B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Anserwise/AWAXIS-Think-31B
- SGLang
How to use Anserwise/AWAXIS-Think-31B with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Anserwise/AWAXIS-Think-31B" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Anserwise/AWAXIS-Think-31B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "Anserwise/AWAXIS-Think-31B" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Anserwise/AWAXIS-Think-31B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use Anserwise/AWAXIS-Think-31B with Docker Model Runner:
docker model run hf.co/Anserwise/AWAXIS-Think-31B
AWAXIS-Think-31B
Overview
AWAXIS-Think-31B is a 31B-parameter Korean/English reasoning model created through the VIDRAFT Darwin AI Model Breeding/Evolution Platform. This model was produced using Darwin's proprietary FFN-crossbreed merge engine (V8), which emulates biological crossbreeding between AI models to create offspring with combined strengths of both parents.
AWAXIS-Think-31B์ VIDRAFT Darwin AI ๋ชจ๋ธ ๊ต๋ฐฐ/์งํ ํ๋ซํผ์ ํตํด ์์ฑ๋ 31B ํ๋ผ๋ฏธํฐ ํ๊ตญ์ด/์์ด ์ถ๋ก ๋ชจ๋ธ์ ๋๋ค.
VIDRAFT Darwin Platform
**VIDRAFT Darwin**์ AI ๋ชจ๋ธ์ **๊ต๋ฐฐ(Crossbreeding)์ ์งํ(Evolution)**๋ฅผ ํตํด ์๋ก์ด ๊ณ ์ฑ๋ฅ ๋ชจ๋ธ์ ์๋ ์์ฑํ๋ ํ๋ซํผ์ ๋๋ค. ์๋ฌผํ์ ์ ์ ์๋ฆฌ์์ ์๊ฐ์ ๋ฐ์, ๋ ๊ฐ ์ด์์ ๋ถ๋ชจ ๋ชจ๋ธ์์ ๊ฐ๊ฐ์ ์ฅ์ ์ ์ ํ์ ์ผ๋ก ๊ฒฐํฉํ์ฌ ์์ ๋ชจ๋ธ์ ํ์์ํต๋๋ค.
Darwin ๊ต๋ฐฐ/์งํ ํต์ฌ ๊ธฐ์
| ๊ธฐ์ | ์ค๋ช |
|---|---|
| FFN Crossbreed Engine (V8) | ๋ถ๋ชจ ๋ชจ๋ธ์ Feed-Forward Network(FFN) ๋ ์ด์ด๋ฅผ ์ ํ์ ์ผ๋ก ๊ต์ฐจ ๊ฒฐํฉํ๋ ํต์ฌ ์์ง. ์ดํ ์ ยท์๋ฒ ๋ฉ์ ์ด๋จธ๋(Mother)์์, FFN ์๊ทธ๋์ ์๋ฒ์ง(Father)์์ ์ถ์ถํ์ฌ ๋ธ๋ ๋ฉ |
| Smart MRI (Model Resonance Imaging) | ๋ ๋ชจ๋ธ ๊ฐ ๋ ์ด์ด๋ณ ์ ์ฌ๋ยทํธํ์ฑ์ ๋ถ์ํ์ฌ ์ต์ ๊ต๋ฐฐ ๋น์จ(alpha)์ ์๋ ํ์ํ๋ ๊ธฐ์ |
| Alpha Grid Search | ๊ต๋ฐฐ ๋น์จ alpha๋ฅผ ์ฒด๊ณ์ ์ผ๋ก ํ์(0.1~0.4)ํ์ฌ ๋ฒค์น๋งํฌ ์ฑ๋ฅ์ด ์ต๋ํ๋๋ ์ต์ ์ ์ ๋ฐ๊ฒฌ |
| Multi-Generation Breeding | 1์ธ๋ ๊ต๋ฐฐ ๊ฒฐ๊ณผ๋ฌผ์ ๋ค์ ๋ถ๋ชจ๋ก ์ผ์ 2์ธ๋, 3์ธ๋ ๊ต๋ฐฐ๋ฅผ ์ํํ๋ ๋ค์ธ๋ ์งํ |
Darwin ๊ต๋ฐฐ ํ๋ก์ธ์ค (์ด ๋ชจ๋ธ์ ์์ฑ ๊ณผ์ )
[Step 1] ๋ถ๋ชจ ์ ์ (Parent Selection)
- Mother: TeichAI/gemma-4-31B-it-Claude-Opus-Distill-v2 (์ถ๋ก ๋ฅ๋ ฅ ๊ธฐ๋ฐ)
- Father: google/gemma-4-31B-it (Gemma-4 ์๋ณธ FFN ๊ธฐ์ฌ)
[Step 2] Smart MRI ํธํ์ฑ ๋ถ์
- ๋ ๋ชจ๋ธ์ 60๊ฐ ๋ ์ด์ด๋ณ FFN ํ
์ ์ ์ฌ๋ ์ค์บ
- ์ํคํ
์ฒ ํธํ์ฑ ํ์ธ (๋์ผ Gemma-4 family = 100% ํธํ)
[Step 3] FFN Crossbreed (๊ต๋ฐฐ ์คํ)
- ์ด๋จธ๋์ ์ดํ
์
, ์๋ฒ ๋ฉ, ๋ผ์ฐํ
= 100% ๋ณด์กด
- ์๋ฒ์ง์ FFN (gate_proj, up_proj, down_proj) = alpha ๋น์จ๋ก ๋ธ๋ ๋ฉ
- ์์: w_child = w_mother * (1 - alpha) + w_father * alpha
[Step 4] Alpha Grid Search (์ต์ ๊ต๋ฐฐ ๋น์จ ํ์)
- alpha = {0.1, 0.2, 0.3, 0.4} 4์ข
์์ฑ
- CLIcK-50 ๋ฒค์น๋งํฌ๋ก ๊ฐ alpha ํ๊ฐ
- ์ต์ : alpha = 0.1 (CLIcK-200 = 86.0%)
[Step 5] ๊ฒ์ฆ ๋ฐ ์ถ์
- GPQA Diamond, CLIcK ๋ฑ ๋ฒค์น๋งํฌ ๊ฒ์ฆ
- HuggingFace ๋ชจ๋ธ ํ๋ธ ๊ณต๊ฐ
์ Darwin ๊ต๋ฐฐ์ธ๊ฐ?
๊ธฐ์กด ๋ชจ๋ธ ํฉ์ฑ ๋ฐฉ์(๋จ์ ๊ฐ์ค์น ํ๊ท , SLERP, TIES ๋ฑ)๊ณผ ๋ฌ๋ฆฌ, Darwin ๊ต๋ฐฐ๋:
- ์๋ฌผํ์ ์ ์ ๋ชจ๋ฐฉ: ์ด๋จธ๋/์๋ฒ์ง ์ญํ ์ ๋ช ํํ ๋ถ๋ฆฌํ์ฌ ๊ฐ ๋ถ๋ชจ์ ํต์ฌ ๋ฅ๋ ฅ๋ง ์ ํ์ ์ผ๋ก ์์
- FFN ์ ํ์ ์ฃผ์ : ์ดํ ์ (๋ฌธ๋งฅ ์ดํด)์ ์ด๋จธ๋์์ 100% ๋ณด์กดํ๊ณ , FFN(์ง์ยท์ถ๋ก ํจํด)๋ง ์๋ฒ์ง์์ ๊ต์ฐจ โ ๋ฅ๋ ฅ ์ถฉ๋ ์ต์ํ
- ๋ฒค์น๋งํฌ ๊ธฐ๋ฐ ์์ฐ์ ํ: alpha grid search๋ก ์ฌ๋ฌ ์์ ํ๋ณด๋ฅผ ์์ฑํ ๋ค, ์ค์ธก ๋ฒค์น๋งํฌ๋ก ์ต์ ๊ฐ์ฒด๋ฅผ ์ ํ (= ์์ฐ์ ํ ์๋ฎฌ๋ ์ด์ )
- ๋ค์ธ๋ ์งํ ๊ฐ๋ฅ: ์ด ๋ชจ๋ธ(AWAXIS-Think-31B)์ด ๋ค์ AWAXIS-KR-31B์ ์๋ฒ์ง๊ฐ ๋์ด 2์ธ๋ ๊ต๋ฐฐ ์ํ โ ๋ฅ๋ ฅ ๋์ ์งํ
Build Recipe (Honest Disclosure)
- Mother (kept full): TeichAI/gemma-4-31B-it-Claude-Opus-Distill-v2 โ reasoning-distill base, retained 100% (incl.
<think>chain-of-thought style) - Father (FFN donor): google/gemma-4-31B-it โ base Gemma-4 FFN tensors blended at alpha = 0.1
- Method: Darwin V8 FFN-crossbreed โ per-layer FFN blend
w = w_mother*(1-alpha) + w_father*alphaonmlp.{gate,up,down}_proj+pre/post_feedforward_layernormfor all 60 language-model layers; grid search alpha in {0.1, 0.2, 0.3, 0.4} on CLIcK-50 โ best alpha=0.1 (CLIcK-200 = 86.0%) - Platform: VIDRAFT Darwin AI Model Breeding/Evolution Platform (VIDraft on HuggingFace)
- Architecture:
Gemma4ForConditionalGeneration(multimodal wrapper; text generation primary) - Tokenizer: Gemma-4 (vocab 262,144)
Model Lineage (Genealogy)
AWAXIS-Think-31B (this model -- Darwin V8 FFN-crossbreed)
|
+-- Mother (kept full, 100%)
| TeichAI/gemma-4-31B-it-Claude-Opus-Distill-v2
| -- Claude Opus reasoning distill base
|
+-- Father (FFN donor, alpha=0.1)
google/gemma-4-31B-it
-- Gemma-4 base FFN tensors
Common ancestor: Google Gemma-4 architecture.
Measured Benchmarks
| Benchmark | Setting | Result |
|---|---|---|
| GPQA Diamond 20Q (seed 42) | greedy, max_new_tokens=4096, 2-way DP | 12/20 = 60.0% (16/20 still hit token cap, 0 null) |
| GPQA Diamond 20Q (seed 42) | greedy, max_new_tokens=2048 | 9/20 = 45.0% (16/20 truncated, 2 null) โ truncation artifact, included for transparency |
| CLIcK (Korean) 200Q | greedy alpha-grid winner | 86.0% |
Honest Caveats
- GPQA 60% is from n=20 (small sample). 16/20 still hit the 4096-token cap โ real ceiling may be higher with longer generation budget.
- Comparison to random baseline: GPQA random 25% โ +35pp clear learning signal.
- The full GPQA Diamond (198Q) and other broad suites have not yet been measured for this exact merged artifact.
- The model retains the Mother's
<think>...</think>reasoning template โ strip via post-processing if undesired.
Intended Use
- Korean/English step-by-step reasoning, instruction following, knowledge QA
- The
Thinksuffix reflects the inherited Opus-distilled chain-of-thought behavior - 2nd-generation breeding parent: This model served as the Father for AWAXIS-KR-31B, demonstrating Darwin's multi-generation evolution capability
Out-of-Scope / Limitations
- Not a final clinical/legal advisor; outputs may be confidently wrong on hard graduate-level questions
- Inherits Gemma-4 base limitations (multimodal wrapper retained; image inputs not the primary use-case here)
- Subject to Gemma Terms of Use; see parent model cards for derivative-use clauses
Inference
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
tok = AutoTokenizer.from_pretrained("Anserwise/AWAXIS-Think-31B", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
"Anserwise/AWAXIS-Think-31B",
dtype=torch.bfloat16,
device_map="auto",
trust_remote_code=True,
attn_implementation="eager",
)
msgs = [{"role": "user", "content": "์์์ญํ์ ๋ถํ์ ์ฑ ์๋ฆฌ๋ฅผ ์์ธํ ์ค๋ช
ํด ์ฃผ์ธ์."}]
text = tok.apply_chat_template(msgs, tokenize=False, add_generation_prompt=True)
inp = tok(text, return_tensors="pt").to(model.device)
out = model.generate(**inp, max_new_tokens=2048, do_sample=False)
print(tok.decode(out[0][inp["input_ids"].shape[-1]:], skip_special_tokens=True))
License
Gemma Terms of Use (inherited from base). Use of this model is bound by Google Gemma Terms.
Acknowledgements
- VIDRAFT โ Darwin AI Model Breeding/Evolution Platform
- TeichAI for the Opus-Distill base
- Google DeepMind for Gemma-4
Built with the VIDRAFT Darwin AI Model Breeding/Evolution Platform โ FFN-crossbreed V8 engine. This model was generated through Darwin's automated crossbreeding process, which selectively combines the strengths of parent models using biologically-inspired genetic algorithms. Measured numbers above are exact; nothing inflated.
- Downloads last month
- 18
Model tree for Anserwise/AWAXIS-Think-31B
Evaluation results
- accuracy on GPQA Diamond (subset n=20, seed=42)self-reported60.000
- accuracy on CLIcKself-reported86.000