İvme-Conversate — Tiny Language Models
A family of sub-130M-parameter language models from IvmeLabs: autoregressive base models, a Python-only coder model, and experimental masked-diffusion models.
Autoregressive base model, general text. Not instruction-tuned — continues text, doesn't chat.
Masked-diffusion sampling: the model denoises a fully-masked span over a fixed number of steps rather than decoding left-to-right.
Run the same prompt and settings through both autoregressive base models at once to see the difference training data made, plus the benchmark deltas below. (Coder-v1 and the diffusion models aren't included here since they use different generation mechanics — try them individually in the Playground tab.)
Benchmark improvement, v1 → v2
WikiText-2 byte perplexity ↓ | 61.40 | 75.09 | +13.69 (+22%) |
Expİvme-DiffusionConversate-v1-Instruct
🧪 Experimental. A 130M-parameter masked-diffusion model, SFT'd for basic chat. Per the model card: output is not reliably grammatical — expect locally plausible word choice that often doesn't compose into coherent sentences. Included here in the spirit of the model card's own honesty about its limitations, not as a working assistant.