File size: 1,859 Bytes
69800cb | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 | # Wildnerve-tlm01 Model Checkpoints
This repository contains the trained weights for the Wildnerve-tlm01 hybrid neural-symbolic language model architecture, which combines transformer-based language processing with biologically-inspired Spike-Timing-Dependent Plasticity (STDP).
## Repository Structure
The weights are organized into two main components:
- **`snn/`**: Contains the STDP-trained synaptic weights that enable neuromorphic learning
- `stdp_model_epoch_30.bin`: Synaptic weights after 30 epochs of STDP training
- **`transformer/`**: Contains the transformer model weights for language processing
- `model_weights.bin`: Transformer component weights
## Usage
These weights are designed to be used with the Wildnerve-tlm01 model implementation located in the [EvolphTech/Model](https://huggingface.co/EvolphTech/Model) repository.
### Loading the Weights
```python
import torch
import os
from huggingface_hub import hf_hub_download
# Download the components
snn_path = hf_hub_download(
repo_id="EvolphTech/Checkpoints",
filename="snn/stdp_model_epoch_30.bin",
cache_dir="model_cache"
)
transformer_path = hf_hub_download(
repo_id="EvolphTech/Checkpoints",
filename="transformer/model_weights.bin",
cache_dir="model_cache"
)
# Load the weights
snn_weights = torch.load(snn_path, map_location="cpu")
transformer_weights = torch.load(transformer_path, map_location="cpu")
```
For complete integration instructions, see the [model documentation](https://huggingface.co/EvolphTech/Model#usage).
## Training Details
- **SNN Component**: Trained for 30 epochs using STDP learning rule with adaptive synaptic weights
- **Transformer Component**: Trained on multiple programming and general knowledge datasets
## License
Mozilla Public License 2.0 (MPL 2.0)
|