---
library_name: pytorch
tags:
- custom-gpt
- small-language-model
- continued-training
- embedded-systems
---

# Continued SLM Checkpoint

Base model: `ishu-newaz/EmbedXPT`

Continued training dataset: `SkunkworksAI/reasoning-0.01`

Training format: `### Instruction` + `### Reasoning` + `### Answer`

Tokenizer: `tiktoken` GPT-2 encoding

Parameters: approximately 505.2M

This is a custom PyTorch checkpoint, not a Transformers `AutoModel` checkpoint. Load it with the custom `GPT` and `GPTConfig` classes from the notebook.