mirror of
https://github.com/deepseek-ai/deepseek-recipe.git
synced 2026-10-28 17:17:08 +08:00
Merge 48e39c50680b93a2009ad903d8da8041e8ba57b1 into 8cadfede7063c896b944e7bae05daa3549ae97ea
This commit is contained in:
commit
f5a1890b24
11
README.md
11
README.md
@ -134,3 +134,14 @@ token IDs or decode backend token IDs, see
|
|||||||
|
|
||||||
Project code and public documentation are licensed under the [MIT License](LICENSE).
|
Project code and public documentation are licensed under the [MIT License](LICENSE).
|
||||||
Bundled tokenizer notices are in [static/tokenizers/README.md](static/tokenizers/README.md).
|
Bundled tokenizer notices are in [static/tokenizers/README.md](static/tokenizers/README.md).
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## ⚡ Production Notes: Migrating to V4.1 Flash
|
||||||
|
|
||||||
|
> [!WARNING]
|
||||||
|
> Legacy model aliases `deepseek-chat` and `deepseek-reasoner` have been fully retired. Ensure your production environments and deployment configurations are explicitly updated to the current endpoints.
|
||||||
|
|
||||||
|
### Quick Migration Specs
|
||||||
|
* **Standard Chat/Code:** Upgrade your configuration files or API client parameters directly to `deepseek-v4-flash`.
|
||||||
|
* **Reasoning (Chain-of-Thought):** Enable the native `thinking` mode parameters inside the API request body instead of relying on outdated separate model names.
|
||||||
|
|||||||
Loading…
x
Reference in New Issue
Block a user