explainers-by-googlers prompt-api issues

explainers-by-googlers / prompt-api

A proposal for a web API for prompting browser-provided language models

Creative Commons Attribution 4.0 International

266 stars 20 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Use cases for parallel requests

#59 tomayac closed 1 week ago
6
Set the capabilities fields to null when the model status is "after-download"?

#58 lozy219 opened 1 week ago
3
Prompt api throws Uncaught UnknownError: Other generic failures occurred.

#57 hexaeight closed 1 week ago
5
Accessing non-instruction-tuned model variants

#56 ACMCMC opened 2 weeks ago
1
Ability to set the message history

#55 mrdjohnson closed 2 weeks ago
1
Use maxTemperature instead of hard-coding 1.0 in example code

#54 lozy219 closed 3 weeks ago
0
Should we clamp or throw for out-of-range sampling hyperparameters?

#53 domenic opened 3 weeks ago
1
Dynamic session

#52 hemanth closed 3 weeks ago
1
Ability to get full conversation history from a session

#51 domenic opened 1 month ago
0
save and restore session

#50 hemanth opened 1 month ago
4
Handling the aborted AbortSignal

#49 lozy219 closed 1 month ago
4
Error: The model attempted to output text in an untested language, and was prevented from doing so.

#48 kylemantesso closed 1 month ago
1
chore: edits in code sample

#47 hemanth closed 1 month ago
0
NotSupportedError: The model attempted to output text in an untested language, and was prevented from doing so.

#46 wginsberg closed 1 month ago
2
Rename supportsLanguage() to languageAvailable()

#45 christianliebel closed 1 month ago
1
Feature request: Stop sequence support

#44 jacoblee93 opened 1 month ago
1
Abort signal for `clone()` API

#43 lozy219 closed 2 months ago
4
Sampling hyperparameters are not universal among models

#42 domenic opened 2 months ago
1
Should there be `maxTemperature`, analog to `maxTopK`?

#41 tomayac closed 2 months ago
1
Add multimodal API such as using image as part of prompt

#40 yaoyaoumbc opened 2 months ago
0
Webconsole chat with token count and prompt confit

#39 berlintay closed 2 months ago
1
Suggestion: remove the "available" property

#38 Yvem closed 2 months ago
1
Suggestion: remove the redundant `tokensLeft` property of `AIAssistant`

#37 Yvem closed 2 months ago
2
Ability to set `maxTokens` when calling `prompt()`

#36 duci9y opened 2 months ago
0
Support for guidance/structured output with prompt API

#35 sushraja-msft opened 3 months ago
2
[Want help]Chat history

#34 Dream-gamer closed 3 months ago
2
Clarify session destruction and download stopping

#33 domenic closed 3 months ago
0
Allow assistant-role prompts and multiple prompts

#32 domenic closed 1 month ago
0
Add User Prompt control interface

#31 kdshk closed 3 months ago
1
Session destruction should not abort ongoing downloads of the model

#30 sushraja-msft closed 3 months ago
1
Clarification on supportsLanguage API

#29 sushraja-msft opened 3 months ago
9
API Shape, prompt/promptStreaming should use AIAssistantPrompt rather than DOMString input

#28 sushraja-msft closed 1 month ago
3
Why not a constructor?

#27 jakearchibald closed 3 months ago
2
Add tokenization and overflow APIs

#26 domenic closed 3 months ago
0
Overhaul the API surface

#25 domenic closed 3 months ago
0
Supported devices?

#24 zerosrat closed 4 months ago
4
Plausible interoperability?

#23 slightlyoff closed 4 months ago
2
systemPrompt and initialPrompts do not seem to be taken into account for the generation of the completion

#22 k33g closed 4 months ago
2
Aborting prompt

#21 horo-t closed 3 months ago
2
Option to get logprobs

#20 blixt opened 4 months ago
4
Reason behind `session.destroy()`

#19 jakearchibald closed 3 months ago
7
AIModelAvailability unavailable

#18 hemanth closed 4 months ago
4
enhance modelInfo

#17 hemanth closed 3 months ago
6
window.ai.rag

#16 hemanth closed 4 months ago
3
window.ai.contextCache

#15 hemanth closed 4 months ago
2
window.ai.finetune

#14 hemanth closed 4 months ago
2
[bug] Stream outputs the last item 3 times

#13 xenova closed 4 months ago
2
Fix quotes in `textmodeldownloadprogress` event example

#12 tomayac closed 4 months ago
0
Register model

#11 niutech opened 4 months ago
2
Raw prompting guide/explanation

#10 xenova closed 4 months ago
3