decisionmodels/sn92-specialists
Updated โข 643
How to use VALOR0316/sn92-ra-9b-a with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="VALOR0316/sn92-ra-9b-a") # pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("VALOR0316/sn92-ra-9b-a")
model = AutoModelForCausalLM.from_pretrained("VALOR0316/sn92-ra-9b-a", device_map="auto")How to use VALOR0316/sn92-ra-9b-a with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "VALOR0316/sn92-ra-9b-a"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "VALOR0316/sn92-ra-9b-a",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'docker model run hf.co/VALOR0316/sn92-ra-9b-a
How to use VALOR0316/sn92-ra-9b-a with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "VALOR0316/sn92-ra-9b-a" \
--host 0.0.0.0 \
--port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "VALOR0316/sn92-ra-9b-a",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'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 "VALOR0316/sn92-ra-9b-a" \
--host 0.0.0.0 \
--port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "VALOR0316/sn92-ra-9b-a",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'How to use VALOR0316/sn92-ra-9b-a with Docker Model Runner:
docker model run hf.co/VALOR0316/sn92-ra-9b-a
Bittensor SN92(REAX) ์คํ์
๋ฆฌ์คํธ ํธ๋ risk_assessment์ฉ์ผ๋ก ๋ฏธ์ธ์กฐ์ ํ Qwen3.5-9B text-only ๋ชจ๋ธ์
๋๋ค.
์คํ ๊ฒฐ๊ณผ๋ฅผ ๊ณต์ ํ๊ธฐ ์ํ ์ ์ฅ์์ด๊ณ , ์ฒด์ธ์ ์ ์ถํ ๋ชจ๋ธ์ด ์๋๋๋ค(reax-specialist.json ๋งค๋ํ์คํธ ์์).
Qwen/Qwen3.5-9B@c202236235762e1c871ad0ccb60c8ee5ba337b9a (Apache-2.0)Qwen3_5ForCausalLM (8,953,803,264 ํ๋ผ๋ฏธํฐ, ํ
์ ์ด๋ฆ์ ์๋ณธ์ model.language_model.*, lm_head ๊ทธ๋๋ก)lm_head๋ ๊ณ ์ ํ์ต๋๋ค.| ๋ชจ๋ธ | S | choice ๋ฅ๋ ฅ/๋ณด์ | noul ๋ฅ๋ ฅ/๋ณด์ | score ๋ฅ๋ ฅ/๋ณด์ |
|---|---|---|---|---|
| ํ์ต ์ (text-only ๋ณํ๋ณธ) | ์ฝ 26 | โ | 0 | โ |
| v2 | 68.28 | 92.3 / 89.0 | 48.4 / 18.6 | 77.2 / 57.8 |
| A (์ด ๋ชจ๋ธ) | 70.24 | 94.1 / 89.4 | 48.4 / 27.3 | 79.3 / 60.7 |
reax-decide-prompt/1)๋ฅผ ๊ณ ์ Qwen ํ
ํ๋ฆฟ์ ๋ฃ๊ณ , <think>\n\n</think>\n\n ๋ค ์ฒซ ์์ฑ ์์น์์ ์ ํ์ง ๊ธ์(A, B, โฆ)์ logit์ ์ฝ์ต๋๋ค.LICENSE๋ฅผ ์ ์ง).decisionmodels/sn92-specialists, CC-BY-4.0)์ train ๋ฌธํญ์ด ๋ค์ด ์์ต๋๋ค.