Instructions to use GainEnergy/ogai-8x7b-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GainEnergy/ogai-8x7b-4bit with Transformers:

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

pipe = pipeline("text-generation", model="GainEnergy/ogai-8x7b-4bit")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("GainEnergy/ogai-8x7b-4bit")
model = AutoModelForCausalLM.from_pretrained("GainEnergy/ogai-8x7b-4bit")
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

vLLM

How to use GainEnergy/ogai-8x7b-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "GainEnergy/ogai-8x7b-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GainEnergy/ogai-8x7b-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/GainEnergy/ogai-8x7b-4bit

SGLang

How to use GainEnergy/ogai-8x7b-4bit 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 "GainEnergy/ogai-8x7b-4bit" \
    --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": "GainEnergy/ogai-8x7b-4bit",
		"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 "GainEnergy/ogai-8x7b-4bit" \
        --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": "GainEnergy/ogai-8x7b-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use GainEnergy/ogai-8x7b-4bit with Docker Model Runner:
```
docker model run hf.co/GainEnergy/ogai-8x7b-4bit
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

OGAI-8x7B-4bit: Oil & Gas AI Model for Drilling Optimization (Quantized)

OGAI-8x7B-4bit is a quantized (4-bit) version of OGAI-8x7B, optimized for deployment in resource-constrained environments while maintaining strong performance for oil and gas drilling engineering tasks.

The model is fine-tuned from Mixtral-8x7B, leveraging a comprehensive dataset of technical literature, engineering workflows, and drilling reports. It is part of GainEnergy's Upstrima AI platform, providing intelligent AI agents, automated workflows, and document-based retrieval-augmented generation (RAG).

🏗 Why Use the 4-bit Model?

🚀 Lower VRAM Usage – Runs on consumer GPUs with less than 16GB VRAM.
💡 Efficient Inference – Optimized using BitsAndBytes 4-bit NF4 quantization.
📚 Long-Context Understanding – Retains up to 32K tokens for technical documents.
🔬 Expert-Focused Fine-Tuning – Specifically optimized for drilling operations.

🛠 Using the OGAI-8x7B-4bit Model

1️⃣ Install Required Dependencies

pip install torch transformers accelerate bitsandbytes

2️⃣ Load the Model in 4-bit Mode

from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig

model_id = "GainEnergy/ogai-8x7b-4bit"
bnb_config = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_compute_dtype="float16")

# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_id)

# Load model with quantization config
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
    quantization_config=bnb_config
)

# Run inference
prompt = "Explain the impact of mud weight on wellbore stability."
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

📌 Key Capabilities

✅ Drilling Calculations & Well Planning – Computes wellbore stability, casing depth, and mud weight optimization.
✅ Engineering Document Processing (RAG) – Enables document-based AI workflows for drilling reports and operational manuals.
✅ AI-Powered Drilling Workflows – Integrates with Upstrima AI for automated engineering analysis.
✅ Optimized for Low VRAM GPUs – Runs on RTX 4090, A100, and consumer-grade GPUs.

🚀 Use Cases

🔹 Drilling Engineering – Automated analysis of wellbore stability, casing designs, and drilling hydraulics.
🔹 Real-time Optimization – Assists drilling teams in reducing non-productive time (NPT) and improving efficiency.
🔹 AI-Driven Document Retrieval – Enables retrieval-augmented generation (RAG) for drilling compliance and regulatory reporting.
🔹 Seamless Integration – Deployable within Upstrima AI, GainEnergy's intelligent AI platform for the oil and gas industry.

🔗 Resources

Full Precision Model – FP16 model for full-accuracy inference.
8-bit Version – Optimized for medium-memory GPUs.
GainEnergy AI Platform – Explore AI-powered drilling automation.

📚 Citing OGAI-8x7B-4bit

@article{ogai8x7b4bit2025,
  title={OGAI-8x7B-4bit: A Quantized AI Model for Oil & Gas Drilling Engineering},
  author={GainEnergy AI Team},
  year={2025},
  publisher={Hugging Face Models}
}

Downloads last month: 11

Safetensors

Model size

48B params

Tensor type

F32

F16

Model tree for GainEnergy/ogai-8x7b-4bit

Base model

mistralai/Mixtral-8x7B-v0.1

Finetuned

mistralai/Mixtral-8x7B-Instruct-v0.1

Adapter

(120)

this model

Datasets used to train GainEnergy/ogai-8x7b-4bit

Evaluation results

Drilling Calculations Accuracy on GainEnergy Oil & Gas Corpus
self-reported

92.500
Engineering Document Retrieval Precision on GainEnergy Oil & Gas Corpus
self-reported

89.800
Context Retention on GainEnergy Oil & Gas Corpus
self-reported

Medium-High