Smolstacpublication by iLevyTate · Pull Request #1 · iLevyTate/stac

Add API reference documentation for STAC multi-turn conversational SNN pipeline, detailing core components, conversion functions, compatibility layers, testing framework, CLI entry points, output f…

iLevyTate

committed

Add detailed documentation for the STAC conversion workflow, outlining the complete process for converting pretrained transformer LLMs to spiking neural networks. Includes implementation phases, qu…

iLevyTate

committed

Add hardware requirements documentation for the STAC conversion framework, detailing conversion, inference, testing requirements, platform compatibility, cloud deployment options, performance bench…

iLevyTate

committed

Add conversion script for transforming quantized LLMs to spiking neural networks. Includes command line interface, calibration data generation, model conversion logic, and SNN-specific model saving…

iLevyTate

committed

Revise README.md to enhance project overview and quick start instructions for STAC. Updated key features, core components, implementation status, and testing guidelines to provide clearer guidance …

iLevyTate

committed

Add STAC-specific entries to .gitignore to exclude large model files, temporary outputs, log files, experimental scripts, and OS-specific artifacts. This helps maintain a cleaner repository by prev…

iLevyTate

committed

Add requirements.txt to specify dependencies for the project, including core libraries for PyTorch, spiking neural networks, efficient training, machine learning utilities, and optional development…

iLevyTate

committed

Add run_conversion.py script to implement the SNN conversion pipeline. This script includes command-line arguments for model conversion, component testing, and TorchScript export. It also incorpora…

iLevyTate

committed

Add simple_snn_test.py to test basic functionality of SNN conversion. The script includes model loading, inference with original and ReLU models, SNN conversion, and prediction comparison, along wi…

iLevyTate

committed

Add test_conversational_snn.py to evaluate conversational capabilities of the SNN model. The script includes argument parsing, conversation simulation, and various tests for position ID boundaries,…

iLevyTate

committed

Add smollm2_converter.py for converting SmolLM2-1.7B-Instruct to a Spiking Neural Network. The script includes model loading, SNN-specific layer replacements, calibration data generation, and a com…

iLevyTate

committed

Remove STAC.ipynb notebook, which contained installation instructions, testing cells, and model definitions for the spiking neural network framework. This file is no longer needed as the project st…

iLevyTate

committed

Add spikingjelly_compat.py to implement a compatibility layer for SpikingJelly components. This file includes version checks, neuron and converter retrieval functions, a custom quantizer class for …

iLevyTate

committed

Remove STAC.ipynb notebook, which included installation instructions and testing cells for the spiking neural network framework. This file is no longer necessary due to updates in the project struc…

iLevyTate

committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Smolstacpublication#1

Smolstacpublication#1
iLevyTate merged 14 commits intomainfrom
smolstacpublication

Commits on Jun 10, 2025

Add API reference documentation for STAC multi-turn conversational SNN pipeline, detailing core components, conversion functions, compatibility layers, testing framework, CLI entry points, output f…

Add detailed documentation for the STAC conversion workflow, outlining the complete process for converting pretrained transformer LLMs to spiking neural networks. Includes implementation phases, qu…

Add hardware requirements documentation for the STAC conversion framework, detailing conversion, inference, testing requirements, platform compatibility, cloud deployment options, performance bench…

Add conversion script for transforming quantized LLMs to spiking neural networks. Includes command line interface, calibration data generation, model conversion logic, and SNN-specific model saving…

Revise README.md to enhance project overview and quick start instructions for STAC. Updated key features, core components, implementation status, and testing guidelines to provide clearer guidance …

Add STAC-specific entries to .gitignore to exclude large model files, temporary outputs, log files, experimental scripts, and OS-specific artifacts. This helps maintain a cleaner repository by prev…

Add requirements.txt to specify dependencies for the project, including core libraries for PyTorch, spiking neural networks, efficient training, machine learning utilities, and optional development…

Add run_conversion.py script to implement the SNN conversion pipeline. This script includes command-line arguments for model conversion, component testing, and TorchScript export. It also incorpora…

Add simple_snn_test.py to test basic functionality of SNN conversion. The script includes model loading, inference with original and ReLU models, SNN conversion, and prediction comparison, along wi…

Add test_conversational_snn.py to evaluate conversational capabilities of the SNN model. The script includes argument parsing, conversation simulation, and various tests for position ID boundaries,…

Add smollm2_converter.py for converting SmolLM2-1.7B-Instruct to a Spiking Neural Network. The script includes model loading, SNN-specific layer replacements, calibration data generation, and a com…

Remove STAC.ipynb notebook, which contained installation instructions, testing cells, and model definitions for the spiking neural network framework. This file is no longer needed as the project st…

Add spikingjelly_compat.py to implement a compatibility layer for SpikingJelly components. This file includes version checks, neuron and converter retrieval functions, a custom quantizer class for …

Remove STAC.ipynb notebook, which included installation instructions and testing cells for the spiking neural network framework. This file is no longer necessary due to updates in the project struc…