Instructions to use RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2 with Transformers:

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

pipe = pipeline("text-generation", model="RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2")
model = AutoModelForCausalLM.from_pretrained("RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2

SGLang

How to use RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2 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 "RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2" \
    --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": "RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2",
		"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 "RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2" \
        --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": "RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2 with Docker Model Runner:
```
docker model run hf.co/RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

LogoS-7Bx2-MoE-13B-v0.1

Model built by @RubielLabarta using SLERP merge method. The model is release for research purposes only, commercial use is not allowed.

The LogoS is a model to experiment with the MoE method, which could significantly increase the performance of the original model. The model has 12.9B parameters.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	77.14
AI2 Reasoning Challenge (25-Shot)	74.49
HellaSwag (10-Shot)	89.07
MMLU (5-Shot)	64.74
TruthfulQA (0-shot)	74.57
Winogrande (5-shot)	88.32
GSM8k (5-shot)	71.65

Downloads last month: 7,894

Safetensors

Model size

13B params

Tensor type

BF16

Model tree for RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2

TomGrc/FusionNet_7Bx2_MoE_14B

yunconglong/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B

Merge model

this model

Quantizations

2 models

Space using RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2 1

Evaluation results

normalized accuracy on AI2 Reasoning Challenge (25-Shot)
test set Open LLM Leaderboard

74.490
normalized accuracy on HellaSwag (10-Shot)
validation set Open LLM Leaderboard

89.070
accuracy on MMLU (5-Shot)
test set Open LLM Leaderboard

64.740
mc2 on TruthfulQA (0-shot)
validation set Open LLM Leaderboard

74.570
accuracy on Winogrande (5-shot)
validation set Open LLM Leaderboard

88.320
accuracy on GSM8k (5-shot)
test set Open LLM Leaderboard

71.650