I have found that the standard IMatrix does not perform very well at low bit quantiztion and for MOE models. So I am using llama.cpp --tensor-type to bump up selected layers. See Layer bumping with llama.cpp

This does create larger model files but increases precision for a given model size.

Please provide feedback on how you find this method performs

Choosing the Right Model Format

Selecting the correct model format depends on your hardware capabilities and memory constraints.

BF16 (Brain Float 16) – Use if BF16 acceleration is available

A 16-bit floating-point format designed for faster computation while retaining good precision.
Provides similar dynamic range as FP32 but with lower memory usage.
Recommended if your hardware supports BF16 acceleration (check your device's specs).
Ideal for high-performance inference with reduced memory footprint compared to FP32.

📌 Use BF16 if:
✔ Your hardware has native BF16 support (e.g., newer GPUs, TPUs).
✔ You want higher precision while saving memory.
✔ You plan to requantize the model into another format.

📌 Avoid BF16 if:
❌ Your hardware does not support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.

F16 (Float 16) – More widely supported than BF16

A 16-bit floating-point high precision but with less of range of values than BF16.
Works on most devices with FP16 acceleration support (including many GPUs and some CPUs).
Slightly lower numerical precision than BF16 but generally sufficient for inference.

📌 Use F16 if:
✔ Your hardware supports FP16 but not BF16.
✔ You need a balance between speed, memory usage, and accuracy.
✔ You are running on a GPU or another device optimized for FP16 computations.

📌 Avoid F16 if:
❌ Your device lacks native FP16 support (it may run slower than expected).
❌ You have memory limitations.

Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds

These formats selectively quantize non-essential layers while keeping key layers in full precision (e.g., attention and output layers).

Named like bf16_q8_0 (meaning full-precision BF16 core layers + quantized Q8_0 other layers).
Strike a balance between memory efficiency and accuracy, improving over fully quantized models without requiring the full memory of BF16/F16.

📌 Use Hybrid Models if:
✔ You need better accuracy than quant-only models but can’t afford full BF16/F16 everywhere.
✔ Your device supports mixed-precision inference.
✔ You want to optimize trade-offs for production-grade models on constrained hardware.

📌 Avoid Hybrid Models if:
❌ Your target device doesn’t support mixed or full-precision acceleration.
❌ You are operating under ultra-strict memory limits (in which case use fully quantized formats).

Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference

Quantization reduces model size and memory usage while maintaining as much accuracy as possible.

Lower-bit models (Q4_K) → Best for minimal memory usage, may have lower precision.
Higher-bit models (Q6_K, Q8_0) → Better accuracy, requires more memory.

📌 Use Quantized Models if:
✔ You are running inference on a CPU and need an optimized model.
✔ Your device has low VRAM and cannot load full-precision models.
✔ You want to reduce memory footprint while keeping reasonable accuracy.

📌 Avoid Quantized Models if:
❌ You need maximum accuracy (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).

Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)

These models are optimized for very high memory efficiency, making them ideal for low-power devices or large-scale deployments where memory is a critical constraint.

IQ3_XS: Ultra-low-bit quantization (3-bit) with very high memory efficiency.
- Use case: Best for ultra-low-memory devices where even Q4_K is too large.
- Trade-off: Lower accuracy compared to higher-bit quantizations.
IQ3_S: Small block size for maximum memory efficiency.
- Use case: Best for low-memory devices where IQ3_XS is too aggressive.
IQ3_M: Medium block size for better accuracy than IQ3_S.
- Use case: Suitable for low-memory devices where IQ3_S is too limiting.
Q4_K: 4-bit quantization with block-wise optimization for better accuracy.
- Use case: Best for low-memory devices where Q6_K is too large.
Q4_0: Pure 4-bit quantization, optimized for ARM devices.
- Use case: Best for ARM-based devices or low-memory environments.

Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)

*Ultra-low-bit quantization (1 2-bit) with extreme memory efficiency.
- Use case: Best for cases were you have to fit the model into very constrained memory
- Trade-off: Very Low Accuracy. May not function as expected. Please test fully before using.

Summary Table: Model Format Selection

Model Format	Precision	Memory Usage	Device Requirements	Best Use Case
BF16	Very High	High	BF16-supported GPU/CPU	High-speed inference with reduced memory
F16	High	High	FP16-supported GPU/CPU	Inference when BF16 isn’t available
Q4_K	Medium-Low	Low	CPU or Low-VRAM devices	Memory-constrained inference
Q6_K	Medium	Moderate	CPU with more memory	Better accuracy with quantization
Q8_0	High	Moderate	GPU/CPU with moderate VRAM	Highest accuracy among quantized models
IQ3_XS	Low	Very Low	Ultra-low-memory devices	Max memory efficiency, low accuracy
IQ3_S	Low	Very Low	Low-memory devices	Slightly more usable than IQ3_XS
IQ3_M	Low-Medium	Low	Low-memory devices	Better accuracy than IQ3_S
Q4_0	Low	Low	ARM-based/embedded devices	Llama.cpp automatically optimizes for ARM inference
*Ultra Low-Bit (IQ1/2_)**	Very Low	Extremely Low	Tiny edge/embedded devices	Fit models in extremely tight memory; low accuracy
Hybrid (e.g., `bf16_q8_0`)	Medium–High	Medium	Mixed-precision capable hardware	Balanced performance and memory, near-FP accuracy in critical layers

Nomic Embed Code: A State-of-the-Art Code Retriever

Blog | Technical Report | AWS SageMaker | Atlas Embedding and Unstructured Data Analytics Platform

nomic-embed-code is a state-of-the-art code embedding model that excels at code retrieval tasks:

High Performance: Outperforms Voyage Code 3 and OpenAI Embed 3 Large on CodeSearchNet
Multilingual Code Support: Trained for multiple programming languages (Python, Java, Ruby, PHP, JavaScript, Go)
Advanced Architecture: 7B parameter code embedding model
Fully Open-Source: Model weights, training data, and evaluation code released

Model	Python	Java	Ruby	PHP	JavaScript	Go
Nomic Embed Code	81.7	80.5	81.8	72.3	77.1	93.8
Voyage Code 3	80.8	80.5	84.6	71.7	79.2	93.2
OpenAI Embed 3 Large	70.8	72.9	75.3	59.6	68.1	87.6
Nomic CodeRankEmbed-137M	78.4	76.9	79.3	68.8	71.4	92.7
CodeSage Large v2 (1B)	74.2	72.3	76.7	65.2	72.5	84.6
CodeSage Large (1B)	70.8	70.2	71.9	61.3	69.5	83.7
Qodo Embed 1 7B	59.9	61.6	68.4	48.5	57.0	81.4

Model Architecture

Total Parameters: 7B
Training Approach: Trained on the CoRNStack dataset with dual-consistency filtering and progressive hard negative mining
Supported Languages: Python, Java, Ruby, PHP, JavaScript, and Go

Usage Guide

Installation

You can install the necessary dependencies with:

pip install transformers sentence-transformers torch

Transformers

python

import torch
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModel

tokenizer = AutoTokenizer.from_pretrained("nomic-ai/nomic-embed-code")
model = AutoModel.from_pretrained("nomic-ai/nomic-embed-code")

def last_token_pooling(hidden_states, attention_mask):
    sequence_lengths = attention_mask.sum(-1) - 1
    return hidden_states[torch.arange(hidden_states.shape[0]), sequence_lengths]

queries = ['Represent this query for searching relevant code: Calculate the n-th factorial']
codes = ['def fact(n):\n if n < 0:\n  raise ValueError\n return 1 if n == 0 else n * fact(n - 1)']
code_snippets = queries + codes

encoded_input = tokenizer(code_snippets, padding=True, truncation=True, return_tensors='pt')
model.eval()
with torch.no_grad():
    model_output = model(**encoded_input)[0]

embeddings = last_token_pooling(model_output, encoded_input['attention_mask'])
embeddings = F.normalize(embeddings, p=2, dim=1)
print(embeddings.shape)

similarity = F.cosine_similarity(embeddings[0], embeddings[1], dim=0)
print(similarity)

SentenceTransformers

python

from sentence_transformers import SentenceTransformer

queries = ['Calculate the n-th factorial']
code_snippets = ['def fact(n):\n if n < 0:\n  raise ValueError\n return 1 if n == 0 else n * fact(n - 1)']

model = SentenceTransformer("nomic-ai/nomic-embed-code")
query_emb = model.encode(queries, prompt_name="query")
code_emb = model.encode(code_snippets)

similarity = model.similarity(query_emb[0], code_emb[0])
print(similarity)

CoRNStack Dataset Curation

Starting with the deduplicated Stackv2, we create text-code pairs from function docstrings and respective code. We filtered out low-quality pairs where the docstring wasn't English, too short, or that contained URLs, HTML tags, or invalid characters. We additionally kept docstrings with text lengths of 256 tokens or longer to help the model learn long-range dependencies.

After the initial filtering, we used dual-consistency filtering to remove potentially noisy examples. We embed each docstring and code pair and compute the similarity between each docstring and every code example. We remove pairs from the dataset if the corresponding code example is not found in the top-2 most similar examples for a given docstring.

During training, we employ a novel curriculum-based hard negative mining strategy to ensure the model learns from challenging examples. We use a softmax-based sampling strategy to progressively sample hard negatives with increasing difficulty over time.

Join the Nomic Community

Nomic Embed Ecosystem: https://www.nomic.ai/embed
Website: https://nomic.ai
Twitter: https://twitter.com/nomic_ai
Discord: https://discord.gg/myY5YDR8z8

Citation

If you find the model, dataset, or training code useful, please cite our work:

bibtex

@misc{suresh2025cornstackhighqualitycontrastivedata,
      title={CoRNStack: High-Quality Contrastive Data for Better Code Retrieval and Reranking}, 
      author={Tarun Suresh and Revanth Gangi Reddy and Yifei Xu and Zach Nussbaum and Andriy Mulyar and Brandon Duderstadt and Heng Ji},
      year={2025},
      eprint={2412.01007},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2412.01007}, 
}

🚀 If you find these models useful

Help me test my AI-Powered Quantum Network Monitor Assistant with quantum-ready security checks:

👉 Quantum Network Monitor

The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : Source Code Quantum Network Monitor. You will also find the code I use to quantize the models if you want to do it yourself GGUFModelBuilder

💬 How to test:
Choose an AI assistant type:

TurboLLM (GPT-4.1-mini)
HugLLM (Hugginface Open-source models)
TestLLM (Experimental CPU-only)

What I’m Testing

I’m pushing the limits of small open-source models for AI network monitoring, specifically:

Function calling against live network services
How small can a model go while still handling:
- Automated Nmap security scans
- Quantum-readiness checks
- Network Monitoring tasks

🟡 TestLLM – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):

✅ Zero-configuration setup
⏳ 30s load time (slow inference but no API costs) . No token limited as the cost is low.
🔧 Help wanted! If you’re into edge-device AI, let’s collaborate!

Other Assistants

🟢 TurboLLM – Uses gpt-4.1-mini :

**It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
Create custom cmd processors to run .net code on Quantum Network Monitor Agents
Real-time network diagnostics and monitoring
Security Audits
Penetration testing (Nmap/Metasploit)

🔵 HugLLM – Latest Open-source models:

🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.

💡 Example commands you could test:

"Give me info on my websites SSL certificate"
"Check if my server is using quantum safe encyption for communication"
"Run a comprehensive security audit on my server"
'"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!

Final Word

I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is open source. Feel free to use whatever you find helpful.

If you appreciate the work, please consider buying me a coffee ☕. Your support helps cover service costs and allows me to raise token limits for everyone.

I'm also open to job opportunities or sponsorship.

Thank you! 😊

Mungert/nomic-embed-code-GGUF

nomic-embed-code GGUF Models

Model Generation Details

Quantization beyond the IMatrix