Changes / Feeds / deepseek-ai/DeepSeek-R1
deepseek-ai/DeepSeek-R1
3 changes to a tracked key since 2025-01-20, newest first. Head as last read: 56d4cbbb4d29f4355bab4b9a39ccb717a14ad5ad.
GPU memory to fine-tune it, card by card: DeepSeek-R1 on /m/.
| Date | Advisory | What changed, and what it does |
|---|---|---|
| 2025-02-09 | x_NXMC-2025-0005 breaking | the chat template changed and renders differently (system+user, multi-turn); prompts rendered with the old one no longer match what the model now expects, so cached prompts and eval sets need re-rendering (2025-02-09).cb48aa8cb28c -> 8a58a132790c |
| 2025-01-22 | x_NXMC-2025-0004 breaking | add_bos_token false to true; encoding the same text yields different ids (2025-01-22).7bae4bacd0f6 -> cb48aa8cb28c |
| 2025-01-22 | x_NXMC-2025-0003 breaking | add_bos_token true to false; encoding the same text yields different ids (2025-01-22).ebc477ae6dc2 -> 7bae4bacd0f6 |
Pin it, and let a bot move the pin
from_pretrained("deepseek-ai/DeepSeek-R1", revision="56d4cbbb4d29f4355bab4b9a39ccb717a14ad5ad")With the Renovate preset, a pull request moves this pin when the head moves, and links back here.
[](https://www.neurarch.com/changes-feeds/models/deepseek-ai/DeepSeek-R1.html)