2025-03-23
@TXhunyuan Hunyuan-T1 compared with other models in MMLU-Pro benchmark. Checkout at https://evalarena.ai/ [image]
South China Morning Post
Tencent unveils Hunyuan T1, a new reasoning AI model powered by its Hunyuan Turbo S AI model, and claims it rivals DeepSeek's R1 in both performance and pricing
Tencent Holdings has unveiled a new artificial intelligence (AI) reasoning model, Hunyuan T1, that rivals DeepSeek's R1 in both performance and pricing.
2024-12-07
7. More general graders being provided by OpenAI for different intents Custom graders of our own also possible later 8. Can configure hyper parameters for Fine Tuning, going with defaults for now 9. We can customize a frontier model for our use case using our dataset, our [image]
OpenAI
OpenAI expands its Reinforcement Fine-Tuning Research Program to let developers create expert models in specific domains with very little training data
the repo we used to train Tulu 3. Expanding reinforcement learning with verifiable rewards (RLVR) to more domains and with better answer extraction (what OpenAI calls a grader, a [...
RFT - Reinforcement Finetuning #OpenAI #Day2 1. Available next year, a preview is being showed today 2. Model learns to reason in a custom domain based on data which it is being fine tuned on 3. With a few examples (a dozen), model can be an expert in that domain, as opposed to
OpenAI
OpenAI expands its Reinforcement Fine-Tuning Research Program to let developers create expert models in specific domains with very little training data
the repo we used to train Tulu 3. Expanding reinforcement learning with verifiable rewards (RLVR) to more domains and with better answer extraction (what OpenAI calls a grader, a [...