🎭 A 0.88B Conversational AI Trained From Scratch «We stand at the right place at the right time.» — Opus 1.5 Opus 1.5 uses a modern LLaMA-style transformer architecture: > Note: This model is very lightweight! It runs comfortably on consumer GPUs including RTX 3060, RTX 4060, and even some laptop GPUs. — GPUs: 2x NVIDIA RTX 4090 (24GB each) — Training Strategy: Fully Sharded Data Parallel (FSDP) — Training Time: ~42 hours > Note: We didn’t notice these tokenizer issues until after training was complete, as we were using simple prompts during checkpoint testing. This will be fixed in Opus 2.0 with a properly trained tokenizer. Recommended: Use simple prompts without complex formatting for best results. In transformers a newline is silently stripped. In llama.cpp it is a hard error, because the SentencePiece vocab has no newline token and no byte fallback, so the lookup throws: llama-server says so at load time, which is the quickest way to confirm it: This is not a quantization bug. The GGUF conversions are fine — including the community quants by Entity-27th. The missing token is in this repository’s tokenizer, and re-converting from source reproduces it exactly. Throws: any…
Модальности:
Генерация текста
Области применения:
Логика и рассуждение Диалог / чат
Задача: Генерация текста
Автор: opus-research
Теги: llama, conversational, from-scratch, chain-of-thought, reasoning, en, text-generation-inference, endpoints_compatible
Лайков: 3 | Загрузок: 213
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
