Reticulum135M-Instruct
This repository contains HuggingFaceTB/SmolLM2-135M-Instruct quantized to Per-Row Int8 and formatted specifically for the in-game Roblox Parallel Luau Engine (Reticulum 1 Zero / AI Engine Core).
Architecture
- Layers: 30
- Hidden Dim: 576
- Heads: 9 Query / 3 KV (GQA)
- Intermediate (FFN): 1536
- Vocab Size: 49152 (ByteLevel BPE)
- Quantization: Per-Row Int8 with dynamic FP32 channel scales
- Downloads last month
- 83
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for Niof/Reticulum135M-Instruct
Base model
HuggingFaceTB/SmolLM2-135M Quantized
HuggingFaceTB/SmolLM2-135M-Instruct