Instructions to use ErebusTN/EGen-SA1Q8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ErebusTN/EGen-SA1Q8 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("ErebusTN/EGen-SA1Q8")
model = PeftModel.from_pretrained(base_model, "ErebusTN/EGen-SA1Q8")

Transformers

How to use ErebusTN/EGen-SA1Q8 with Transformers:

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

pipe = pipeline("text-generation", model="ErebusTN/EGen-SA1Q8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("ErebusTN/EGen-SA1Q8")
model = AutoModelForMultimodalLM.from_pretrained("ErebusTN/EGen-SA1Q8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ErebusTN/EGen-SA1Q8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ErebusTN/EGen-SA1Q8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ErebusTN/EGen-SA1Q8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ErebusTN/EGen-SA1Q8

SGLang

How to use ErebusTN/EGen-SA1Q8 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 "ErebusTN/EGen-SA1Q8" \
    --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": "ErebusTN/EGen-SA1Q8",
		"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 "ErebusTN/EGen-SA1Q8" \
        --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": "ErebusTN/EGen-SA1Q8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ErebusTN/EGen-SA1Q8 with Docker Model Runner:
```
docker model run hf.co/ErebusTN/EGen-SA1Q8
```

🏛️ Athena Project 🐼

Next-generation Supervised Fine-Tuning (SFT) for advanced reasoning and language understanding. Athena Project Explore Model • Report Bug • ErebusTN Profile
The model was trained and validated using a cutting-edge software stack to ensure stability and performance:

📖 Overview

The Athena Project (2025) represents a milestone in efficient high-performance language modeling. Developed by ErebusTN, the EGen-SA1Q8 variant is a precision-tuned model designed to deliver superior conversational capabilities and structured data processing.

By leveraging Supervised Fine-Tuning (SFT), Athena has been optimized to follow complex instructions with high fidelity, maintaining a balance between creative generation and factual accuracy.

🚀 Key Features

SFT Optimized: Trained using Supervised Fine-Tuning to ensure alignment with human intent.
2025 Architecture: Incorporates the latest advancements in transformer optimization.
Quantization Ready: The SA1Q8 designation signifies optimized weight distribution for efficient deployment.
High Compatibility: Seamlessly integrates with the modern Hugging Face ecosystem.

🛠️ Tech Stack & Frameworks

💻 Quick Start

You can load the model using the following snippet:

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_id = "ErebusTN/EGen-SA1Q8"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.float16,
    device_map="auto"
)

prompt = "Explain the significance of the Athena Project in 2025."
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=150)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

📊 Training Methodology

Athena Project utilized the SFT (Supervised Fine-Tuning) trainer from the TRL library. This process involved:

Instruction Following: Tuning on high-quality, human-annotated datasets.
Parameter Efficiency: Utilizing PEFT for optimized memory usage during the tuning phase.
Precision Alignment: Leveraging the latest cu126 CUDA kernels for accelerated compute.

🤝 Contact & Support

Developed by ErebusTN

Hugging Face: @ErebusTN
Github @EGen-V
Project Year: 2025

Built with ❤️ By ErebusTN.

Downloads last month: 2

Safetensors

Model size

1B params

Tensor type

F16

Model tree for ErebusTN/EGen-SA1Q8

Unable to build the model tree, the base model loops to the model itself. Learn more.

ErebusTN
/

EGen-SA1Q8