Instructions to use FINAL-Bench/Aether-7B-5Attn-it with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FINAL-Bench/Aether-7B-5Attn-it with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="FINAL-Bench/Aether-7B-5Attn-it", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("FINAL-Bench/Aether-7B-5Attn-it", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use FINAL-Bench/Aether-7B-5Attn-it with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FINAL-Bench/Aether-7B-5Attn-it"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FINAL-Bench/Aether-7B-5Attn-it",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/FINAL-Bench/Aether-7B-5Attn-it

SGLang

How to use FINAL-Bench/Aether-7B-5Attn-it 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 "FINAL-Bench/Aether-7B-5Attn-it" \
    --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": "FINAL-Bench/Aether-7B-5Attn-it",
		"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 "FINAL-Bench/Aether-7B-5Attn-it" \
        --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": "FINAL-Bench/Aether-7B-5Attn-it",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use FINAL-Bench/Aether-7B-5Attn-it with Docker Model Runner:
```
docker model run hf.co/FINAL-Bench/Aether-7B-5Attn-it
```

SeaWolf-AI commited on Jul 18

Commit

ac6cf42

verified ·

1 Parent(s): 16bcb3c

card: add measured base->it delta (+8.2pp) and pretraining domain shares

Browse files

Files changed (1) hide show

README.md +13 -0

README.md CHANGED Viewed

@@ -36,10 +36,23 @@ full-parameter SFT를 세 학습률(2e-6 / 6e-6 / 2e-5)로 각각 수행한 뒤,
 - **GPQA-Diamond (198)**: 25.3%
 - **K-AI 4 평균 (195)**: **34.9%**
   - musr_ko 26.5% · com2_main_ko 50.0% · click 32.0% · kommlu_pro 30.4%
 *모든 평가는 학습에 쓰이지 않은 held-out 셋입니다. SFT는 MMLU-auxiliary(객관식 형식) 위에서 수행되었고,
 평가 과목(GPQA·K-AI)과 겹치지 않습니다.*
 ## 사용
 베이스와 동일하게 `aether_pkg/` 의 커스텀 아키텍처를 사용합니다. 로딩 방법은

 - **GPQA-Diamond (198)**: 25.3%
 - **K-AI 4 평균 (195)**: **34.9%**
   - musr_ko 26.5% · com2_main_ko 50.0% · click 32.0% · kommlu_pro 30.4%
+- **base(SFT 전) 대비**: base 26.7% → **34.9% (+8.2pp)** — 동일 held-out·동일 하니스로 측정
 *모든 평가는 학습에 쓰이지 않은 held-out 셋입니다. SFT는 MMLU-auxiliary(객관식 형식) 위에서 수행되었고,
 평가 과목(GPQA·K-AI)과 겹치지 않습니다.*
+## 사전학습 데이터 구성 (베이스 상속, 유효 샘플링 비중)
+| 도메인 | 비중 |
+|--------|------|
+| Math (finemath + open-web-math) | 37.8% |
+| Korean (webtext + synth) | 21.6% |
+| English web·synthetic (fineweb-edu + cosmopedia) | 21.6% |
+| Code (opc) | 13.5% |
+| phase15 프리블렌드 | 5.4% |
+*이 구성은 **베이스 사전학습** 데이터의 비중입니다. 본 모델의 후처리(SFT) 데이터는 MMLU-auxiliary(객관식)입니다.*
 ## 사용
 베이스와 동일하게 `aether_pkg/` 의 커스텀 아키텍처를 사용합니다. 로딩 방법은