NASA Systems Engineering
Collection
This collection contains all the data for performing your own Finetunes of the NASA Systems Engineering Handbook as well as finetuned models. • 7 items • Updated
How to use from
llama.cppInstall from WinGet (Windows)
winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf MarshallDoyle/NASA-Systems-Engineering:Q4_K_M# Run inference directly in the terminal:
llama cli -hf MarshallDoyle/NASA-Systems-Engineering:Q4_K_MUse 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 MarshallDoyle/NASA-Systems-Engineering:Q4_K_M# Run inference directly in the terminal:
./llama-cli -hf MarshallDoyle/NASA-Systems-Engineering:Q4_K_MBuild 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 MarshallDoyle/NASA-Systems-Engineering:Q4_K_M# Run inference directly in the terminal:
./build/bin/llama-cli -hf MarshallDoyle/NASA-Systems-Engineering:Q4_K_MUse Docker
docker model run hf.co/MarshallDoyle/NASA-Systems-Engineering:Q4_K_MQuick Links
🚀 NASA Systems Engineering LLaMA-3.1 8B Instruct (Fine-Tuned)
-lightgrey.svg){kind=icon}
Developed by: MarshallDoyle
Fine-tuned from: unsloth/meta-llama-3.1-8b-instruct-unsloth-bnb-4bit
Specialization: NASA Systems Engineering Handbook (2018) Q&A
Training recipe: LoRA rank=16, epochs=3, steps=285, final loss≈0.40–0.60
✨ Key Features
| Feature | Details |
|---|---|
| Domain Expertise | Deep knowledge of NASA SE lifecycle, terminology, and best practices. |
| Compact Footprint | +83.8M trainable via LoRA on 8B base. |
| Long Context | Up to 128K token support for large documents. |
| Fast Inference | Efficient TRL & Unsloth pipeline. |
🎯 Intended Use Cases
- Interactive Q&A: Rapid lookup of definitions, processes, and phase purposes.
- Summarization: Concise overviews of complex sections in the handbook.
- Educational Tools: Training aids for systems engineering students and professionals.
- Mission Planning: Quick reference for verification, validation, and lifecycle checkpoints.
📊 Example Comparisons
Phase C Purpose
| Model | Response Snippet |
|---|---|
| Base LLaMA-3.1 | “Phase C covers detailed design and preparations for manufacturing…” (generic) |
| Fine-Tuned | “The main purpose is to complete the detailed design of system elements, fabricate hardware, and code software, ensuring all end products are defined for integration.” |
Verification vs. Validation
| Model | Response Snippet |
|---|---|
| Base LLaMA-3.1 | “Verification checks requirements; validation checks user needs…” (too brief) |
| Fine-Tuned | “Verification ensures the system meets specified requirements; validation confirms it satisfies stakeholder expectations in its intended environment.” |
⚠️ Limitations
- Scope Bounded: Content based on the 2018 handbook; may omit newer updates.
- Hallucinations: Possible on out-of-handbook queries.
- Bias: NASA-specific terminology; verify critical engineering decisions against official documents.
🔗 Download & Details: Hugging Face model page
- Downloads last month
- 94
Hardware compatibility
Log In to add your hardware
4-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Install (macOS, Linux)
# Start a local OpenAI-compatible server with a web UI: llama serve -hf MarshallDoyle/NASA-Systems-Engineering:Q4_K_M# Run inference directly in the terminal: llama cli -hf MarshallDoyle/NASA-Systems-Engineering:Q4_K_M