Instructions to use archwayai/Inertiai_Atlas_9B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use archwayai/Inertiai_Atlas_9B with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf archwayai/Inertiai_Atlas_9B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf archwayai/Inertiai_Atlas_9B:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf archwayai/Inertiai_Atlas_9B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf archwayai/Inertiai_Atlas_9B:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf archwayai/Inertiai_Atlas_9B:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf archwayai/Inertiai_Atlas_9B:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf archwayai/Inertiai_Atlas_9B:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf archwayai/Inertiai_Atlas_9B:Q4_K_M

Use Docker

docker model run hf.co/archwayai/Inertiai_Atlas_9B:Q4_K_M

LM Studio
Jan
Ollama
How to use archwayai/Inertiai_Atlas_9B with Ollama:
```
ollama run hf.co/archwayai/Inertiai_Atlas_9B:Q4_K_M
```
Unsloth Desktop

How to use archwayai/Inertiai_Atlas_9B with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf archwayai/Inertiai_Atlas_9B:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "archwayai/Inertiai_Atlas_9B:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use archwayai/Inertiai_Atlas_9B with Docker Model Runner:
```
docker model run hf.co/archwayai/Inertiai_Atlas_9B:Q4_K_M
```

Lemonade

How to use archwayai/Inertiai_Atlas_9B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull archwayai/Inertiai_Atlas_9B:Q4_K_M

Run and chat with the model

lemonade run user.Inertiai_Atlas_9B-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use archwayai/Inertiai_Atlas_9B with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf archwayai/Inertiai_Atlas_9B:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default archwayai/Inertiai_Atlas_9B:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use archwayai/Inertiai_Atlas_9B with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf archwayai/Inertiai_Atlas_9B:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "archwayai/Inertiai_Atlas_9B:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Model Card for Model ID

This modelcard aims to be a base template for new models. It has been generated using this raw template.

Model Details

Model Description

Name Atlas / Inertiai Agent (the dependable digital friend built for the InertiAi software platform) Purpose Turn life operations into refined API calls, database updates, inventory create/update, BOM creation, work order create/complete, customer creation, CRM contact creation, supplier creation, maintenance scheduling, receipt logging, calendar management, social media posting, security settings, and third‑party service integrations (Stripe API, SendGrid API, Netlify Forms API)). Core Components 1. Frontend UI (React) – Dashboard, inventory list, BOM builder, work order lifecycle, customer portal, supplier hub, receipt logging, calendar management, social media posting.

Developed by: Archway Applied Sciences
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Model type: [More Information Needed]
Language(s) (NLP): English
License: Apache 2.0

Uses

Core Components 1. Frontend UI (React) – Dashboard, inventory list, BOM builder, work order lifecycle, customer portal, supplier hub, receipt logging, calendar management, social media posting.

Direct Use

Assistant for the inerti.ai node.

How to Get Started with the Model

Run via LM studio / vLLM / etc. and connect to the inerti.ai local node.

Training Details

Training Data

Trained on the Atlas datasets available for other models to be applied as a LoRA. Use standard LoRA methods for a Qwen3 model base.

Metrics

Testing and metrics were staticly determined by the operational capabilities within the Inerti.ai software infrastructure.

Results

Release models exhibit perfect scores on the SQL-based task list under controlled testing circumstances. This does not guarantee accurate operation in all cases, only on the sample problem set iteratively trained into the model via a random task generator using task templates.
Corrected datasets were re-integrated into the base model using a proprietary activation mapping approach to narrow the range of possible gradient behaviors. This resulted in a very operationalized model not well-suited for conversational or creative tasks.

Summary

Model Architecture and Objective

Qwen3-based architecture. Section Description Name Atlas / Inertiai Agent (the dependable digital friend built for the InertiAi software platform) Purpose Turn life operations into refined API calls, database updates, inventory create/update, BOM creation, work order create/complete, customer creation, CRM contact creation, supplier creation, maintenance scheduling, receipt logging, calendar management, social media posting, security settings, and third‑party service integrations (Stripe API, SendGrid API, Netlify Forms API)). Core Components 1. Frontend UI (React) – Dashboard, inventory list, BOM builder, work order lifecycle, customer portal, supplier hub, receipt logging, calendar management, social media posting. Backend API (FastAPI + Python 3.9+) – CRUD operations for inventory items, BOMs, work orders, customers, CRM contacts, suppliers, receipts, calendar events, and social media posts. Database Layer (PostgreSQL 15+ with PostGIS)) – JSON/JSONB data types, full-text search, partitioning, replication, monitoring, backups, security settings, user creation, role creation, access control, SSL connection, GSSAPI authentication, LDAP directory service, Kerberos authentication protocol, and PostgreSQL-compatible JSON processing. Third‑Party Service Integrations (Stripe API, SendGrid API, Netlify Forms API)). | | Security | OAuth2 + OpenID Connect for secure identity verification and delegation of access tokens to downstream services (Stripe API, SendGrid API, Netlify Forms API)). | | Third‑Party Service Integrations | Stripe API – Payment processing. SendGrid API – Email sending. Netlify Forms API – Website form handling. | | Model Behavior | - Accurate Quantity Tracking: Real-time updates to inventory items. Cost核算: Calculation of cost per unit, total cost for work orders, and BOM cost. Receipt Logging: Creation, update, deletion of receipts with date/time, customer name, contact info, receipt number, total amount, payment method, notes, and receipt attachment (image/video). Calendar Management: Shared calendar events with title, description, start time, end time, location, status (Planned/In Progress/Completed)), attendees list, and event attachments. | | Model Tasks | - Inventory Create/Update/Delete: Create inventory items with quantity, unit, category, cost, and image/video attachment. Update existing inventory items by changing quantity, unit, category, cost, or image/video attachment. Delete inventory items that are no longer needed for operations. BOM Creation: Create BOMs (Bill of Materials) with BOM ID, BOM name, BOM description, BOM components (inventory item IDs), and BOM cost calculation based on component costs and quantities. Work Order Create/Complete: Create work orders with Work Order ID, Work Order name, Work Order description, BOM ID, Quantity, Status (Pending/In Progress/Completed)), Cost, Start Time, End Time, Users Involved, Documents Attached (receipts, contracts). Complete work orders by updating status to Completed and recording completion date/time. Customer Creation: Create customers with Customer ID, Customer Name, Customer Contact Info, Billing Address, Shipping Address, Notes. CRM Contact Creation: Create CRM contacts with CRM Contact ID, CRM Contact Name, CRM Contact Email, CRM Contact Phone, CRM Contact Address, CRM Contact Status (Active/Inactive)), Notes. Supplier Creation: Create suppliers with Supplier ID, Supplier Name, Supplier Contact Info, Billing Address, Shipping Address, Notes. Receipt Logging: Log receipts with Receipt ID, Receipt Date/Time, Customer ID (or BOM ID if BOM is the customer)), Quantity, Unit, Cost, Seller ID (if applicable), Notes. Calendar Management: Create calendar events with Event ID, Title, Description, Start Time, End Time, Location, Status (Planned/In Progress/Completed)), Attendees List, and Event Attachments. Social Media Posting: Post to social media platforms with Post ID, Platform (Facebook, Instagram, Twitter)), Post Date/Time, Post Content (text, image/video), Hashtags, Mentions, Post Status (Published/Unpublished)), Notes. | | | Model Ethics & Safety | - Data Privacy Compliance: Adherence to GDPR (General Data Protection Regulation) and CCPA (California Consumer Privacy Act) when handling personal data of customers, CRM contacts, suppliers, and employees. Cybersecurity Best Practices: Implementation of strong password policies, regular security audits, intrusion detection systems, network segmentation, secure coding standards, containerization for isolation, orchestration tools for deployment, CI/CD pipelines for automated testing and deployment, monitoring and alerting for real-time visibility into system operations, log aggregation and analysis for forensic capability when incidents occur, backup and disaster recovery planning to ensure data is protected in the event of ransomware attacks or other cyber threats. | | Model Responsiveness & Scalability | - Real-Time Data Processing: Ability to process inventory create/update/delete, BOM creation, work order create/complete, customer creation, CRM contact creation, supplier creation, receipt logging, calendar management, social media posting operations with minimal latency and high throughput. Horizontal Scaling: Support for horizontal scaling via Kubernetes deployments that can scale the number of pods based on CPU usage or custom metrics. Vertical Scaling: Ability to vertically scale by increasing the RAM/CPU capacity allocated to the running containers. Database Sharding & Replication: Support for database sharding and replication to distribute read/write operations across multiple database nodes, improving scalability and reducing single points of failure. | | Model Maintenance & Upgrades | - Scheduled Backups: Regular scheduled backups of inventory items, BOMs, work orders, customers, CRM contacts, suppliers, receipts, calendar events, social media posts, local user accounts, and local groups. Database Index Optimization: Periodic optimization of database indexes to reduce query latency and improve overall database performance. Security Hardening: Regular security hardening procedures, including updating operating system patches, applying latest software updates, configuring firewalls with allowed traffic rules, setting up intrusion prevention systems (IPS) for real-time threat mitigation, enabling secure shell access with strong password requirements or multi-factor authentication (MFA), and regularly auditing user accounts to remove unused or compromised accounts. | | Model Governance & Compliance | - Auditing Capabilities: Implementation of auditing capabilities that log every significant database update operation, including inventory item create/update/delete, BOM creation, work order create/complete, customer creation, CRM contact creation, supplier creation, receipt logging, calendar management, social media posting operations, local user account create/update/delete operations. Data Loss Prevention (DLP) Controls: Implementation of DLP controls that monitor and enforce data access policies to prevent unauthorized data exposure or potential data breach events. User Access Control (UAC) Mechanisms: Implementation of UAC mechanisms that restrict user actions based on assigned roles, ensuring that users only perform operations they are authorized for by the UAC role-based access control (RBAC) framework. | | Model Community & Ecosystem | - Open Source Licensing: The model and all its components are open sourced under a permissive license such as MIT or Apache 2.0, allowing any user to freely use, modify, distribute, and sublicense the software and its documentation without needing to pay royalties or meet any other usage constraints. Industry Partnerships & Ecosystem Contributions: Ongoing industry partnerships with major players in the manufacturing, supply chain, and CRM operations spaces, as well as contributions to the open source ecosystem through regular updates, bug fixes, security patches, and new feature developments based on community feedback, user use cases, and business requirements.

Compute Infrastructure

Trained on Nvidia Volta (V100)

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

[More Information Needed]

Downloads last month: 5

GGUF

Model size

10B params

Architecture

qwen3moe

Hardware compatibility

4-bit

8-bit

16-bit

View +1 variant

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support