frag.jetzt issueshttps://git.thm.de/arsnova/frag.jetzt/-/issues2024-01-21T19:34:02+01:00https://git.thm.de/arsnova/frag.jetzt/-/issues/275Flexible Management of Web Push Notifications on frag.jetzt2024-01-21T19:34:02+01:00Nils MittlerFlexible Management of Web Push Notifications on frag.jetzt**User Story**
As a user of the Q&A platform frag.jetzt, I would appreciate the ability to enable and configure Web Push notifications for individual rooms both through the respective room options on the room overview page (first screen...**User Story**
As a user of the Q&A platform frag.jetzt, I would appreciate the ability to enable and configure Web Push notifications for individual rooms both through the respective room options on the room overview page (first screenshot) and individually for each room through the "My Profile" menu under "Options" on each room page (second screenshot). This will enable me to receive specific notifications for activities in selected rooms, even when I am not actively using the platform.
**Acceptance Criteria**
- [ ] **Two Access Points for Notification Settings**: Users should be able to configure Web Push notifications in two ways:
- On the room overview page under "Options" for each specific room.
- Through the "My Profile" menu under "Options" on the respective room page, to set up notification preferences for that specific room.
- [ ] **Individual Settings per Room**: Users can select the types of notifications (such as star awards, replies, comments, bans/deletions, see third screenshot) they wish to receive in each room.
- [ ] **Ease of Use**: Activating or deactivating notifications and selecting desired notification types should be straightforward and intuitive at both locations.
- [ ] **Opt-in Procedure**: Users must actively consent to receive Push notifications and can revoke this consent at any time at both configuration locations.
- [ ] **Access to Relevant Content**: Clicking on a Push notification should direct the user to the relevant question or comment in the corresponding room.
- [ ] **Cross-Platform Functionality**: Push notifications should reliably work across different devices and browsers without impacting performance.
- [ ] **Privacy and Security**: The implementation of Push notifications must comply with data protection regulations and ensure the security of user data.
- [ ] **Feedback and Continuous Improvement**: User feedback should be gathered following the feature's introduction and used for continuous optimization of the notification system.
This user story aims to provide frag.jetzt users with a flexible and user-friendly way to manage Web Push notifications for individual rooms, both via the room overview and individually through their profile, thereby creating a personalized and effective notification experience.
![first_option](/uploads/84750ef1dfa7f4af324efc2b19e70760/first_option.png)
![second_option](/uploads/f437e967bbe19d13908589e49e3dfa64/second_option.png)
![types](/uploads/009ab508e8bace29649531e3eb0acf29/types.png)2024-01-26https://git.thm.de/arsnova/frag.jetzt/-/issues/776Consider the selected question categories as topics in the cloud2022-11-06T12:16:25+01:00Klaus-Dieter Quibeldey-CirkelConsider the selected question categories as topics in the cloudIf there are questions with selected question categories (tags), then display the tags as topics in the cloud.If there are questions with selected question categories (tags), then display the tags as topics in the cloud.Ruben BimbergRuben Bimberghttps://git.thm.de/arsnova/frag.jetzt/-/issues/938Place a counter badge for questions or answers in moderation2022-05-07T07:48:34+02:00Klaus-Dieter Quibeldey-CirkelPlace a counter badge for questions or answers in moderation![badge](/uploads/b0e1c23b200a6ec81e397c7aa98fea7b/badge.jpg)![badge](/uploads/b0e1c23b200a6ec81e397c7aa98fea7b/badge.jpg)Ruben BimbergRuben Bimberghttps://git.thm.de/arsnova/frag.jetzt/-/issues/1091Offer voting for MotDs2023-01-21T08:41:05+01:00Klaus-Dieter Quibeldey-CirkelOffer voting for MotDs![MotD](/uploads/b66b368f594edadd54afe134dc6d7aab/MotD.png)![MotD](/uploads/b66b368f594edadd54afe134dc6d7aab/MotD.png)Ruben BimbergRuben Bimberghttps://git.thm.de/arsnova/frag.jetzt/-/issues/1453Prompt Compression2024-02-28T08:04:29+01:00Klaus-Dieter Quibeldey-CirkelPrompt CompressionTry to implement:
GitHub: https://github.com/microsoft/LLMLingua
See:
- https://llmlingua.com/
- Paper: https://arxiv.org/pdf/2310.05736.pdf
> The paper titled "LLMLingua: Compressing Prompts for Accelerated Inference of Large Languag...Try to implement:
GitHub: https://github.com/microsoft/LLMLingua
See:
- https://llmlingua.com/
- Paper: https://arxiv.org/pdf/2310.05736.pdf
> The paper titled "LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models" introduces a method for prompt compression in large language models (LLMs). The method, named LLMLingua, is designed to address the challenge of increasingly lengthy prompts in LLMs, which can exceed tens of thousands of tokens. LLMLingua uses a coarse-to-fine prompt compression approach, featuring a budget controller for maintaining semantic integrity at high compression ratios, an iterative token-level compression algorithm for modeling interdependencies, and an instruction tuning method for distribution alignment between models. The paper demonstrates the effectiveness of LLMLingua across various datasets, showing that it can achieve significant compression with minimal performance loss. This approach is particularly relevant for accelerating model inference and reducing computational costs in practical applications of LLMs.
- Article: https://www.linkedin.com/pulse/prompt-compression-large-language-models-ashish-bhatia-vt4ye/Ruben BimbergRuben Bimberg2024-02-29