ExternalVendorCode/lollms-webui
1
0
Fork 0
mirror of https://github.com/ParisNeo/lollms-webui.git synced 2024-12-18 20:17:50 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

upgraded summary

Browse Source
This commit is contained in:
Saifeddine ALOUI 2024-07-09 23:05:30 +02:00
parent 5d44b782cf
commit 7bd3653c1a
6 changed files with 9 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
docs/dev/new_ui_dev.md
View File

@ -55,7 +55,7 @@ Here we keep track of things to implement and stuff we need to do.
- Add indicator to messages when non-comercial model was used.
- Ability to export messages that are for commercial use or open source model messages Only.
- Ability to hide non commercial model messages.
- Feature - bot council where you add multiple bots to discussion, give them a topic and maybe max message count or something and they ponder about the topic then summerize it all.
- Feature - bot council where you add multiple bots to discussion, give them a topic and maybe max message count or something and they ponder about the topic then summarize it all.
- Feature voice chat with bot and voice output from bot - whisper + bard?
- Feature under selected discussion show more options to select and add more bots, select model per discussion or per bot, tweak settings per bot or per model.
- Easy share personality via export to a file, then drag and drop on to the webui and youre done.

4
docs/youtube/lollms_contextual_summery.md
View File

@ -4,9 +4,9 @@ To do so, we first go to settings page. Under the personalities section, select
Now, we go to the personality settings and we set some specific summery parameters. Here we say keep the method description, we select keep document title and authors in the summary. We set the summary size in tokens and we validate.
Now we add the document to summerize and we go to the personality menu and we select start
Now we add the document to summarize and we go to the personality menu and we select start
The document will be decomposed into a certain number of chunks, then each chunk is contextually summerized. After that the summeries are tied together then the operation is repeated until the compressed text is smaller than the maximum number of tokens set in the configuration.
The document will be decomposed into a certain number of chunks, then each chunk is contextually summarized. After that the summeries are tied together then the operation is repeated until the compressed text is smaller than the maximum number of tokens set in the configuration.
The contextual nature of this algorithm will allow you to have better control over the summary. For example, here we asked for keeping the title, the author names and the results of the paper as well as the method.

2
lollms_core

@ -1 +1 @@
Subproject commit 826be699b476bc15eef979b61adc5bec915d1c75
Subproject commit 0e176402e34dc60e6155a528fcfed301ba9a8c51

4
presets/airoboros_instruct.yaml
View File

@ -8,7 +8,7 @@ content: |
help: |
This is the default Airoboros fine tuning format for chat mode with input.
There are two parameters:
- The first is the input, which is the subject of your injstruction. For example, if the instruction is summerize this text, then the input is the text.
- The second parameter is the instruction itself. For example, summerize this text.
- The first is the input, which is the subject of your injstruction. For example, if the instruction is summarize this text, then the input is the text.
- The second parameter is the instruction itself. For example, summarize this text.
author: ParisNeo (adapted from the original Airoboros documentation by Jon Durbin)

4
presets/alpaca_instruct_with_input.yaml
View File

@ -12,5 +12,5 @@ content: |
help: |
This is the default Alpaca fine tuning format for instruct mode with input.
There are two parameters:
- The first parameter is the instruction itself. For example, summerize this text.
- The second parameter is the input, which is the subject of your instruction. For example, if the instruction is summerize this text, then the input is the text.
- The first parameter is the instruction itself. For example, summarize this text.
- The second parameter is the input, which is the subject of your instruction. For example, if the instruction is summarize this text, then the input is the text.

2
zoos/personalities_zoo

@ -1 +1 @@
Subproject commit 4ec1a31b687bf1adda8ae84f25a06dd06888d5bb
Subproject commit d63a0b02d5bac3b5d7894ea72b9ab033865d7006