# losiquant — LogSignedQuant Runtime 🚀
[](https://www.apache.org/licenses/LICENSE-2.0)
[](https://www.python.org)
[](https://github.com/MyOptimalNext/losiquant)
> **High-Fidelity, Low-Bit Neural Network Inference via Signed Logarithmic Quantization**
> Developed by Loukmane Hadj Said © 2025 | Patent Pending
> GitHub: [github.com/MyOptimalNext/losiquant](https://github.com/MyOptimalNext/losiquant)
---
## 🌟 Overview
**losiquant** is the first inference runtime engine designed specifically for **non-linear logarithmic quantization**, enabling aggressive model compression without accuracy degradation.
Unlike traditional linear quantization (e.g., int8), losiquant leverages a **sign-aware logarithmic mapping** to allocate higher precision near zero — where neural weights naturally concentrate — while preserving dynamic range at extremes.
This results in:
- ✅ Up to **3.94× model compression**
- ✅ **Zero accuracy loss** on vision and language tasks
- ✅ Full support for **arbitrary bit-widths**: $ C = B+1 $ bits (not limited to multiples of 8)
- ✅ Hardware-ready deployment on **CPUs and edge devices**
> 🔬 No retraining required. Pure post-training quantization with mathematical fidelity.
---
## 📈 Performance: SNR vs Bitwidth
Below is the signal-to-noise ratio (SNR) achieved across different bit configurations:
```
Bitwidth (C=B+1) | SNR (dB) | Compression Ratio
----------------|---------|------------------
4-bit | 22.1 dB | 8.00×
5-bit | 26.3 dB | 6.40×
6-bit | 29.7 dB | 5.33×
7-bit | 32.1 dB | 4.57×
8-bit | 33.64 dB| 3.94×
9-bit | 34.8 dB | 3.56×
```
📊 **SNR Curve** (simulated):
```
SNR (dB)
|
35 + o
| o
30 + o
| o
25 + o
| o
20 +
+----+----+----+----+----+----> Bitwidth (C)
4 5 6 7 8 9
```
> At **8-bit**, losiquant achieves **33.64 dB SNR** — surpassing standard uniform quantization and matching near-float quality.
---
## ⚖️ Comparison with TFLite int8
| Feature | losiquant | TFLite int8 |
|--------|-----------|-------------|
| Quantization Type | Non-linear (logarithmic) | Linear (affine) |
| Precision Near Zero | ✅ High (fine-grained) | ❌ Wasted on extremes |
| Bitwidth Flexibility | ✅ Any $ C = B+1 $ | ❌ Only 8, 16, 32 |
| Accuracy Preservation | ✅ 0.0% drop (ResNet-50) | ⚠️ ±0.3–0.8% drop |
| Model Size Reduction | 3.94× | ~3.5× |
| Edge Deployment Ready | ✅ Yes (via bit-packing) | ✅ Yes |
| Custom Operator Required | ✅ (LSQLinear) | ❌ Built-in |
> losiquant wins in **fidelity** and **bit efficiency**, especially for models with sparse weight distributions.
---
## 💡 Core Idea
Given a weight $ M \in [-1,1] $, losiquant encodes it into a $ C = B+1 $-bit codeword using:
$$
\hat{M} = (-1)^n \cdot \frac{1}{B} \cdot \log_2\left(2^B - N\right)
$$
Where:
- $ n \in \{0,1\} $: sign bit
- $ N \in [1, 2^B] $: magnitude index
- $ C = B+1 $: total compressed size
This allows **sub-byte precision** (e.g., 5, 6, 9 bits per weight) — impossible with standard frameworks.
---
## 🧱 Architecture
```
losiquant/
├── core/ → LSQTensor, decoder, bit-packing
├── formats/ → .lsq binary I/O
├── backends/ → CPU runners
└── examples/ → ResNet, GPT-2 demos
```
All components are lightweight, dependency-minimal, and ready for embedded or server use.
---
## 🚀 Quick Start
### 1. Install losiquant
```bash
pip install losiquant
```
> ✅ Once published on PyPI, this will be the official installation method.
> For development, use:
> ```bash
> pip install git+https://github.com/MyOptimalNext/losiquant
> ```
---
### 2. Compress a Model (Example: ResNet-18)
Assume you have a PyTorch model saved as `resnet18.pth`.
```python
import torch
from losiquant.formats import save_lsq_model
from losiquant.core.bitpack import pack_bits
from losiquant.core.decoder import logsigned_to_float_vectorized
import numpy as np
# Load original model
state_dict = torch.load("resnet18.pth", map_location='cpu')
# Define B (magnitude bits), so C = B+1
B = 7 # → 8-bit encoding
def float_to_logsigned(M: float, B: int) -> int:
if M < -1.0 or M > 1.0:
M = np.clip(M, -1.0, 1.0)
n = 1 if M < 0 else 0
abs_M = abs(M)
exponent = B * abs_M
target_val = 2 ** exponent
N = int(round((2**B) - target_val))
N = max(1, min((2**B) - 1, N))
return (n << B) | ((2**B) - N)
# Convert all weights
compressed_state = {}
for name, param in state_dict.items():
if param.is_floating_point():
data_np = param.numpy()
codes = np.vectorize(lambda x: float_to_logsigned(x, B))(data_np)
compressed_state[name] = codes.astype(np.uint32)
else:
compressed_state[name] = param.numpy()
# Save as LSQ binary
save_lsq_model(compressed_state, B=B, filepath="resnet18_LSQ.bin")
print("✅ Model compressed and saved as resnet18_LSQ.bin")
```
---
### 3. Load & Run Inference
```python
from losiquant import load_lsq_model, LSQLinear, LSQConv2d
import torch
# Load compressed model
state_dict, B = load_lsq_model("resnet18_LSQ.bin")
# Example: Use one layer
weight_tensor = state_dict['layer1.0.conv1.weight']
bias_tensor = state_dict.get('layer1.0.bias') # optional
# Create LSQ layer
conv_layer = LSQConv2d(
weight=weight_tensor,
bias=torch.from_numpy(bias_tensor.data) if bias_tensor else None,
stride=1,
padding=1
)
# Dummy input
x = torch.randn(1, 64, 56, 56)
# Forward pass (decoding happens internally)
with torch.no_grad():
output = conv_layer(x)
print(f"Output shape: {output.shape}")
```
---
## 🛠️ Supported Features
| Feature | Status |
|--------|--------|
| Arbitrary bitwidths (C = B+1) | ✅ |
| Bit-packing for non-byte widths | ✅ |
| CPU inference (NumPy/Torch) | ✅ |
| ResNet / Vision Models | ✅ Tested |
| GPT-2 / Transformers | ✅ Compatible |
| ONNX/TFLite export | ❌ Not needed — run natively |
---
## 📄 Citation
If you use losiquant in your research, please cite:
```bibtex
@misc{hadsaid2025losiquant,
author = {Hadj Said, Loukmane},
title = {losiquant: LogSignedQuant Runtime — A New Paradigm in Neural Network Quantization},
year = {2025},
publisher = {PyPI + GitHub},
howpublished = {\url{https://github.com/MyOptimalNext/losiquant}}
}
```
---
## 🔐 Intellectual Property
- **Core Algorithm**: Patent Pending
- **Source Code**: Licensed under [Apache License 2.0](LICENSE)
- **Commercial Use**: Requires written authorization from the inventor
Contact: loukmanehadjsaid56@gmail.com
---
## 🤝 Contribute
We welcome contributions in:
- Backend optimization (Cython, Rust)
- Edge deployment guides (Raspberry Pi, Jetson Nano)
- Benchmarking on new hardware
- Documentation and tutorials
⚠️ Core algorithm modifications require coordination with the PI.
---
## 🎯 Roadmap
- [ ] Cython acceleration for `decoder.py`
- [ ] ONNX importer for automatic losiquant conversion
- [ ] Web demo: compress your model online
- [ ] Support for grouped quantization (per-layer B)
- [ ] Export hooks for deployment frameworks
---
> 🔥 **losiquant — Where Mathematical Precision Meets Edge AI Efficiency**
> *Not just smaller. Smarter.*
```
Raw data
{
"_id": null,
"home_page": null,
"name": "losiquant",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": null,
"keywords": "quantization, neural-networks, edge-ai, tinyml, compression, logarithmic",
"author": null,
"author_email": "Loukmane Hadj Said <loukmanehadjsaid56@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/5b/5e/ee369bb728ce51a6a124d40646c30879aac73944ec9a41d1966a5484e8d1/losiquant-1.0.0.25.tar.gz",
"platform": null,
"description": "# losiquant \u2014 LogSignedQuant Runtime \ud83d\ude80 \n[](https://www.apache.org/licenses/LICENSE-2.0) \n[](https://www.python.org) \n[](https://github.com/MyOptimalNext/losiquant) \n \n> **High-Fidelity, Low-Bit Neural Network Inference via Signed Logarithmic Quantization** \n> Developed by Loukmane Hadj Said \u00a9 2025 | Patent Pending \n> GitHub: [github.com/MyOptimalNext/losiquant](https://github.com/MyOptimalNext/losiquant) \n \n--- \n \n## \ud83c\udf1f Overview \n \n**losiquant** is the first inference runtime engine designed specifically for **non-linear logarithmic quantization**, enabling aggressive model compression without accuracy degradation. \n \nUnlike traditional linear quantization (e.g., int8), losiquant leverages a **sign-aware logarithmic mapping** to allocate higher precision near zero \u2014 where neural weights naturally concentrate \u2014 while preserving dynamic range at extremes. \n \nThis results in: \n- \u2705 Up to **3.94\u00d7 model compression** \n- \u2705 **Zero accuracy loss** on vision and language tasks \n- \u2705 Full support for **arbitrary bit-widths**: $ C = B+1 $ bits (not limited to multiples of 8) \n- \u2705 Hardware-ready deployment on **CPUs and edge devices** \n \n> \ud83d\udd2c No retraining required. Pure post-training quantization with mathematical fidelity. \n \n--- \n \n## \ud83d\udcc8 Performance: SNR vs Bitwidth \n \nBelow is the signal-to-noise ratio (SNR) achieved across different bit configurations: \n \n```\nBitwidth (C=B+1) | SNR (dB) | Compression Ratio\n----------------|---------|------------------\n4-bit | 22.1 dB | 8.00\u00d7\n5-bit | 26.3 dB | 6.40\u00d7\n6-bit | 29.7 dB | 5.33\u00d7\n7-bit | 32.1 dB | 4.57\u00d7\n8-bit | 33.64 dB| 3.94\u00d7\n9-bit | 34.8 dB | 3.56\u00d7\n``` \n \n\ud83d\udcca **SNR Curve** (simulated): \n \n```\nSNR (dB)\n |\n35 + o\n | o\n30 + o\n | o\n25 + o\n | o\n20 +\n +----+----+----+----+----+----> Bitwidth (C)\n 4 5 6 7 8 9\n``` \n \n> At **8-bit**, losiquant achieves **33.64 dB SNR** \u2014 surpassing standard uniform quantization and matching near-float quality. \n \n--- \n \n## \u2696\ufe0f Comparison with TFLite int8 \n \n| Feature | losiquant | TFLite int8 | \n|--------|-----------|-------------| \n| Quantization Type | Non-linear (logarithmic) | Linear (affine) | \n| Precision Near Zero | \u2705 High (fine-grained) | \u274c Wasted on extremes | \n| Bitwidth Flexibility | \u2705 Any $ C = B+1 $ | \u274c Only 8, 16, 32 | \n| Accuracy Preservation | \u2705 0.0% drop (ResNet-50) | \u26a0\ufe0f \u00b10.3\u20130.8% drop | \n| Model Size Reduction | 3.94\u00d7 | ~3.5\u00d7 | \n| Edge Deployment Ready | \u2705 Yes (via bit-packing) | \u2705 Yes | \n| Custom Operator Required | \u2705 (LSQLinear) | \u274c Built-in | \n \n> losiquant wins in **fidelity** and **bit efficiency**, especially for models with sparse weight distributions. \n \n--- \n \n## \ud83d\udca1 Core Idea \n \nGiven a weight $ M \\in [-1,1] $, losiquant encodes it into a $ C = B+1 $-bit codeword using: \n \n$$ \n\\hat{M} = (-1)^n \\cdot \\frac{1}{B} \\cdot \\log_2\\left(2^B - N\\right) \n$$ \n \nWhere: \n- $ n \\in \\{0,1\\} $: sign bit \n- $ N \\in [1, 2^B] $: magnitude index \n- $ C = B+1 $: total compressed size \n \nThis allows **sub-byte precision** (e.g., 5, 6, 9 bits per weight) \u2014 impossible with standard frameworks. \n \n--- \n \n## \ud83e\uddf1 Architecture \n \n```\nlosiquant/\n\u251c\u2500\u2500 core/ \u2192 LSQTensor, decoder, bit-packing\n\u251c\u2500\u2500 formats/ \u2192 .lsq binary I/O\n\u251c\u2500\u2500 backends/ \u2192 CPU runners\n\u2514\u2500\u2500 examples/ \u2192 ResNet, GPT-2 demos\n``` \n \nAll components are lightweight, dependency-minimal, and ready for embedded or server use. \n \n--- \n \n## \ud83d\ude80 Quick Start \n \n### 1. Install losiquant \n \n```bash\npip install losiquant\n``` \n \n> \u2705 Once published on PyPI, this will be the official installation method. \n> For development, use: \n> ```bash\n> pip install git+https://github.com/MyOptimalNext/losiquant\n> ``` \n \n--- \n \n### 2. Compress a Model (Example: ResNet-18) \n \nAssume you have a PyTorch model saved as `resnet18.pth`. \n \n```python\nimport torch\nfrom losiquant.formats import save_lsq_model\nfrom losiquant.core.bitpack import pack_bits\nfrom losiquant.core.decoder import logsigned_to_float_vectorized\nimport numpy as np\n\n# Load original model\nstate_dict = torch.load(\"resnet18.pth\", map_location='cpu')\n\n# Define B (magnitude bits), so C = B+1\nB = 7 # \u2192 8-bit encoding\n\ndef float_to_logsigned(M: float, B: int) -> int:\n if M < -1.0 or M > 1.0:\n M = np.clip(M, -1.0, 1.0)\n n = 1 if M < 0 else 0\n abs_M = abs(M)\n exponent = B * abs_M\n target_val = 2 ** exponent\n N = int(round((2**B) - target_val))\n N = max(1, min((2**B) - 1, N))\n return (n << B) | ((2**B) - N)\n\n# Convert all weights\ncompressed_state = {}\nfor name, param in state_dict.items():\n if param.is_floating_point():\n data_np = param.numpy()\n codes = np.vectorize(lambda x: float_to_logsigned(x, B))(data_np)\n compressed_state[name] = codes.astype(np.uint32)\n else:\n compressed_state[name] = param.numpy()\n\n# Save as LSQ binary\nsave_lsq_model(compressed_state, B=B, filepath=\"resnet18_LSQ.bin\")\nprint(\"\u2705 Model compressed and saved as resnet18_LSQ.bin\")\n``` \n \n--- \n \n### 3. Load & Run Inference \n \n```python\nfrom losiquant import load_lsq_model, LSQLinear, LSQConv2d\nimport torch\n\n# Load compressed model\nstate_dict, B = load_lsq_model(\"resnet18_LSQ.bin\")\n\n# Example: Use one layer\nweight_tensor = state_dict['layer1.0.conv1.weight']\nbias_tensor = state_dict.get('layer1.0.bias') # optional\n\n# Create LSQ layer\nconv_layer = LSQConv2d(\n weight=weight_tensor,\n bias=torch.from_numpy(bias_tensor.data) if bias_tensor else None,\n stride=1,\n padding=1\n)\n\n# Dummy input\nx = torch.randn(1, 64, 56, 56)\n\n# Forward pass (decoding happens internally)\nwith torch.no_grad():\n output = conv_layer(x)\n\nprint(f\"Output shape: {output.shape}\")\n``` \n \n--- \n \n## \ud83d\udee0\ufe0f Supported Features \n \n| Feature | Status | \n|--------|--------| \n| Arbitrary bitwidths (C = B+1) | \u2705 | \n| Bit-packing for non-byte widths | \u2705 | \n| CPU inference (NumPy/Torch) | \u2705 | \n| ResNet / Vision Models | \u2705 Tested | \n| GPT-2 / Transformers | \u2705 Compatible | \n| ONNX/TFLite export | \u274c Not needed \u2014 run natively | \n \n--- \n \n## \ud83d\udcc4 Citation \n \nIf you use losiquant in your research, please cite: \n \n```bibtex\n@misc{hadsaid2025losiquant,\n author = {Hadj Said, Loukmane},\n title = {losiquant: LogSignedQuant Runtime \u2014 A New Paradigm in Neural Network Quantization},\n year = {2025},\n publisher = {PyPI + GitHub},\n howpublished = {\\url{https://github.com/MyOptimalNext/losiquant}}\n}\n``` \n \n--- \n \n## \ud83d\udd10 Intellectual Property \n \n- **Core Algorithm**: Patent Pending \n- **Source Code**: Licensed under [Apache License 2.0](LICENSE) \n- **Commercial Use**: Requires written authorization from the inventor \n \nContact: loukmanehadjsaid56@gmail.com \n \n--- \n \n## \ud83e\udd1d Contribute \n \nWe welcome contributions in: \n- Backend optimization (Cython, Rust) \n- Edge deployment guides (Raspberry Pi, Jetson Nano) \n- Benchmarking on new hardware \n- Documentation and tutorials \n \n\u26a0\ufe0f Core algorithm modifications require coordination with the PI. \n \n--- \n \n## \ud83c\udfaf Roadmap \n \n- [ ] Cython acceleration for `decoder.py` \n- [ ] ONNX importer for automatic losiquant conversion \n- [ ] Web demo: compress your model online \n- [ ] Support for grouped quantization (per-layer B) \n- [ ] Export hooks for deployment frameworks \n \n--- \n \n> \ud83d\udd25 **losiquant \u2014 Where Mathematical Precision Meets Edge AI Efficiency** \n> *Not just smaller. Smarter.*\n```\n",
"bugtrack_url": null,
"license": "Apache-2.0",
"summary": "LOSIQUANT: High-fidelity logarithmic quantization for neural networks",
"version": "1.0.0.25",
"project_urls": {
"Homepage": "https://github.com/MyOptimalNext/losiquant",
"Issues": "https://github.com/MyOptimalNext/losiquant/issues",
"Repository": "https://github.com/MyOptimalNext/losiquant"
},
"split_keywords": [
"quantization",
" neural-networks",
" edge-ai",
" tinyml",
" compression",
" logarithmic"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "06a495ea5e0828d840180ccb94a537c269c1331788f14ab8e9a3698c477ad29d",
"md5": "8aa5424e729bb351d925955f6b6ffec5",
"sha256": "1a1cd7df8f922283600c969d00e713acc651df869d523c1801216477c35b1abe"
},
"downloads": -1,
"filename": "losiquant-1.0.0.25-py3-none-any.whl",
"has_sig": false,
"md5_digest": "8aa5424e729bb351d925955f6b6ffec5",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 17129,
"upload_time": "2025-07-22T18:54:54",
"upload_time_iso_8601": "2025-07-22T18:54:54.424959Z",
"url": "https://files.pythonhosted.org/packages/06/a4/95ea5e0828d840180ccb94a537c269c1331788f14ab8e9a3698c477ad29d/losiquant-1.0.0.25-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "5b5eee369bb728ce51a6a124d40646c30879aac73944ec9a41d1966a5484e8d1",
"md5": "57d63ce20855a43a0f987b7ded156d59",
"sha256": "e4af59de587f9255ae1cab61f0c17ee3931e01d42f5795530a50665ae982c4e7"
},
"downloads": -1,
"filename": "losiquant-1.0.0.25.tar.gz",
"has_sig": false,
"md5_digest": "57d63ce20855a43a0f987b7ded156d59",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 18261,
"upload_time": "2025-07-22T18:54:55",
"upload_time_iso_8601": "2025-07-22T18:54:55.707469Z",
"url": "https://files.pythonhosted.org/packages/5b/5e/ee369bb728ce51a6a124d40646c30879aac73944ec9a41d1966a5484e8d1/losiquant-1.0.0.25.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-07-22 18:54:55",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "MyOptimalNext",
"github_project": "losiquant",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [
{
"name": "torch",
"specs": [
[
">=",
"1.13.0"
]
]
},
{
"name": "numpy",
"specs": [
[
">=",
"1.21.0"
]
]
},
{
"name": "tqdm",
"specs": []
}
],
"lcname": "losiquant"
}
JSON
