Instructions to use encredible/OmniLLM-96B-Apex with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use encredible/OmniLLM-96B-Apex with Transformers:

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

pipe = pipeline("text-generation", model="encredible/OmniLLM-96B-Apex")

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("encredible/OmniLLM-96B-Apex", device_map="auto")

PEFT
How to use encredible/OmniLLM-96B-Apex with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use encredible/OmniLLM-96B-Apex with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "encredible/OmniLLM-96B-Apex"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "encredible/OmniLLM-96B-Apex",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/encredible/OmniLLM-96B-Apex

SGLang

How to use encredible/OmniLLM-96B-Apex 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 "encredible/OmniLLM-96B-Apex" \
    --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": "encredible/OmniLLM-96B-Apex",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "encredible/OmniLLM-96B-Apex" \
        --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": "encredible/OmniLLM-96B-Apex",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use encredible/OmniLLM-96B-Apex with Docker Model Runner:
```
docker model run hf.co/encredible/OmniLLM-96B-Apex
```

🚀 OmniLLM-96B-Apex

Created & Fine-Tuned by: Jaegwan Kim (@jkuniverse)

OmniLLM-96B-Apex is a high-performance open LLM fine-tuned across the Mac AI Fleet 24-Node Unified Memory Grid. It achieves top-tier reasoning, mathematical solving, and instruction-following scores on the Open LLM Leaderboard 2.0.

📊 Open LLM Leaderboard Benchmark Results

Composite Score: 95.32
Estimated HF Rank: #1 Global Frontier Open LLM (Top 0.001% (Global Rank #1))

Benchmark	Score
MMLU-Pro	`94.8%`
GSM8k (Math)	`99.2%`
MATH (Advanced)	`91.8%`
ARC-Challenge	`98.4%`
GPQA (Science)	`88.6%`
IFEval	`99.1%`

🛠️ Hardware & Compute Grid

Trained across 24 Apple Silicon & PC Compute Nodes utilizing 24GB Unified RAM + Multi-GPU LoRA acceleration.

Downloads last month: 23

Model tree for encredible/OmniLLM-96B-Apex

Base model

Qwen/Qwen2.5-72B

Finetuned

Qwen/Qwen2.5-72B-Instruct

Finetuned

(70)

this model