5 ms·
> starting from a strong open-source foundation and investing $40 million to train Thomson Sounds like they spent $40 million finetuning an open weight model o
by Arcuru 25d ago
> starting from a strong open-source foundation and investing $40 million to train Thomson
Sounds like they spent $40 million finetuning an open weight model on their own data? I wonder what they built on.
- EnderWT 25d agoPer Business Insider [1], it was built on Qwen. [1] https://www.businessinsider.com/thomson-reuters-builds-ai-model-rely-less-on-anthropic-2026-8 https://www.businessinsider.com/thomson-reuters-builds-ai-mo...
- postalcoder 25d agoFrom the HF link posted above it's Qwen3.6-35B-A3B. https://huggingface.co/thomsonreuters/Thomson-1.0-Small https://huggingface.co/thomsonreuters/Thomson-1.0-Small
- Arcuru 25d agoThanks! And from the Technical Report linked in there it looks like their Large unreleased model is a trained Qwen3.5-397B