You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Originally posted by cmasantos November 14, 2024
Hi, I've notice that "normal" chat model produce a set of gen_ai* metrics. But I've been unable to use them with streaming models. Also when we use a combination of Mutiny StreamingResponseHandler I'm almost sure that it can't be done.
My question is, how can I add metrics to a streaming model even if need to instantiate them? I've found the MetricsChatModelListener class but I can't find a way to put it together with the streaming model interface.
We should add observability for StreamingChatLanguageModel as well
The text was updated successfully, but these errors were encountered:
@cmasantos would you be willing to contribute this feature?
Yes, I'd like to contribute. I'll give it a try, but it will take me a while because I have no idea how extensions development work. I've got to learn that first.
In this case you probably don't need to learn too many details as the changes almost exactly match those that were made for ChatLanguageModel - I can certainly point you in the right direction.
Discussed in #1080
Originally posted by cmasantos November 14, 2024
Hi, I've notice that "normal" chat model produce a set of
gen_ai*
metrics. But I've been unable to use them with streaming models. Also when we use a combination of Mutiny StreamingResponseHandler I'm almost sure that it can't be done.My question is, how can I add metrics to a streaming model even if need to instantiate them? I've found the
MetricsChatModelListener
class but I can't find a way to put it together with the streaming model interface.We should add observability for
StreamingChatLanguageModel
as wellThe text was updated successfully, but these errors were encountered: