Revert "feat: include tokens usage for streamed output"

[mudler]

Reverts #4282

See: #4334

[netlify]

✅ Deploy Preview for localai ready!

Name	Link
🔨 Latest commit	e407fd4
🔍 Latest deploy log	https://app.netlify.com/sites/localai/deploys/6755bb404c874900081a45c0
😎 Deploy Preview	https://deploy-preview-4336--localai.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[mudler]

ping @mintyleaf - seems the changeset blocks the server from serving further requests, I'm going to revert and tag a patch release until you can provide a fix quicker than me :)

[mintyleaf]

ping @mintyleaf - seems the changeset blocks the server from serving further requests, I'm going to revert and tag a patch release until you can provide a fix quicker than me :)

really strange, bc i tested it locally on that examplish small phi2 model on arm64 mac and there is no issues with multiple requests to it

you really sure that my changeset causing this / able to reproduce that behavior?

[mudler]

ping @mintyleaf - seems the changeset blocks the server from serving further requests, I'm going to revert and tag a patch release until you can provide a fix quicker than me :)

really strange, bc i tested it locally on that examplish small phi2 model on arm64 mac and there is no issues with multiple requests to it

you really sure that my changeset causing this / able to reproduce that behavior?

yep tested myself and was able to reproduce the behavior, reverting this changeset fixed it here. I've tested directly in the LocalAI WebUI: to reproduce you have to run subsequent messages in the chat and it will stall indefinetly

[mintyleaf]

ping @mintyleaf - seems the changeset blocks the server from serving further requests, I'm going to revert and tag a patch release until you can provide a fix quicker than me :)

really strange, bc i tested it locally on that examplish small phi2 model on arm64 mac and there is no issues with multiple requests to it
you really sure that my changeset causing this / able to reproduce that behavior?

yep tested myself and was able to reproduce the behavior, reverting this changeset fixed it here. I've tested directly in the LocalAI WebUI: to reproduce you have to run subsequent messages in the chat and it will stall indefinetly

can you say which model you used?
still can't reproduce that web ui behavior with phi-2.Q2_K