mirror of
https://github.com/deepseek-ai/DeepSeek-V3.git
synced 2026-11-05 00:47:48 +00:00
[GH-ISSUE #949] Feature Request: Context Memory Mode for DeepSeek #572
Labels
No labels
closed-as-stale
enhancement
question
spam
stale
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
deepseek-ai/DeepSeek-V3#572
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @Pomsakar on GitHub (Aug 9, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-V3/issues/949
Currently, DeepSeek models treat each conversation turn as an independent request, relying solely on the provided context.
In long-running sessions, this forces users to repeatedly send large portions of conversation history, which increases token usage and slows down responses.
I propose adding a “Context Memory Mode” — a mechanism that stores conversation embeddings and concise summaries locally or in a vector database. The model would then retrieve and reconstruct relevant context on demand, without resending the entire history.
Benefits:
Significantly reduced token usage in multi-turn conversations.
Faster response times for extended sessions.
More coherent and context-aware replies in long interactions.
Possible Implementation:
Store embeddings for each conversation segment in a local vector store (e.g., FAISS, Milvus).
On each turn, retrieve the top-N most relevant segments and append them to the prompt.
Add an API parameter, e.g.,
memory_mode=True, to enable or disable this feature.This enhancement would make DeepSeek more efficient and user-friendly in scenarios such as chatbots, AI assistants, and persistent dialogue applications.
@github-actions[bot] commented on GitHub (Sep 9, 2025):
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!
@0izayoi0 commented on GitHub (Sep 20, 2025):
id like this feature too honestly
@github-actions[bot] commented on GitHub (Oct 22, 2025):
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!
@github-actions[bot] commented on GitHub (Nov 6, 2025):
false