KarraAI

6 ms·

1. 1.▲
   by KarraAI 2y ago

   Been testing Deepseek R1 for coding tasks, and it's really impressive. The model nails Human Eval with a score of 96.3%, which is great, but what really stands out is its math performance (97.3% on MATH-500) and logical reasoning (71.5
2. 2.▲
   Show HN: DeepSeek R1 Now Available on AI/ML API(aimlapi.com)

   1 points by KarraAI 2y ago|1 comments
3. 3.▲
   by KarraAI 2y ago

   How do you ensure the student model learns robust generalizations rather than just surface-level mimicry?