openai-python
openai-python copied to clipboard
[0.26.0] openai/cli.py:440: RuntimeWarning: coroutine 'FineTune.stream_events' was never awaited
bash$ openai api fine_tunes.create -t ./train_test_prepared.jsonl -m davinci
Upload progress: 100%|██████████████████████████████████████████████████████████████████████████████| 112/112 [00:00<00:00, 88.5kit/s]
Uploaded file from ./train_test_prepared.jsonl: file-ppEDNe0p6EomteEp3JFbBoFp
Created fine-tune: ft-u9KskmmvSnBtVc4VDfbe7lyr
Streaming events until fine-tuning is complete...
(Ctrl-C will interrupt the stream, but not cancel the fine-tune)
Stream interrupted (client disconnected).
To resume the stream, run:
openai api fine_tunes.follow -i ft-u9KskmmvSnBtVc4VDfbe7lyr
/usr/lib/python3.10/site-packages/openai/cli.py:406: RuntimeWarning: coroutine 'FineTune.stream_events' was never awaited
cls._stream_events(resp["id"])
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
it is working fine with 0.25.0
may be related: dev-python/aiohttp-3.8.3
Same here for me. I thought it might be incompatible with my Python version (3.10), but it only worked when I installed version 0.25 of the package (pip install openai==0.25.0)
I have this same problem with version 0.25.0. Is there any other solution to this?
Hi there, I have the same issue even with 0.25.0 and 0.24.0 C:\Users\admindev\GPT>openai api fine_tunes.follow -i ft-mM9Fk2xqRQTtqbt7clQLheRf [2023-01-12 13:00:28] Created fine-tune: ft-mM9Fk2xqRQTtqbt7clQLheRf
Stream interrupted (client disconnected). To resume the stream, run:
openai api fine_tunes.follow -i ft-mM9Fk2xqRQTtqbt7clQLheRf
and pip list output
Is there any solution for this problem?
Hey folks, the FineTune.stream_events
method was made async by mistake, rather than adding an additional async version. We'll have a fix shortly.
Hey @ddeville, Great, thanks for the update!
Hey guys, I'm facing that same issue. I'm using openai version 0.26.0.
Hi there, thank you for the update. I still have the same problem , my fine tune models from yesterday has been published successfully but cannot do a new one even after using 0.26.0
with 0.25 and 0.24 I am getting this error:
with 0.26 I am getting this error:
Not sure whether I need to open a new issue or continue here
I'm not sure about the error on 0.24 and 0.25 but for the one you're seeing on 0.26, it will be fixed as soon as we release the next minor version with the fix.
Thanks @ddeville for the update, do you have any estimate on the next minor release? I may consider using one of the older versions for the POC I am working on,
Not entirely sure but trying to get something before end of week.
Great thanks, really appreciate this
Had this same issue, I applied the patch from the MR and it appears to be working now.
Hey folks, we've just published 0.26.1 that includes the fix for this bug.
I'm still receiving this error, and I can confirm my version has been upgraded to 0.26.1
Anybody else or is this just me?
Hi @elijahschrockAGT , I think it might be a problem at your side, thanks to the dev team including @ddeville , v 26.1 works fine for me as shown below
@mohsspark829 Okay that's good to know. Thank you for the response! I'll take a deeper look into it today
I am having the same problem, tried with version 0.25 and 0.26.1 but keep showing the same error
RuntimeWarning: coroutine 'FineTune.stream_even
@mohsspark829 @ddeville can u guys help me?
I am running into the exact same problem whether I use 0.25 or 0.26.
@ddeville need your help please having the same issue with openai-0.26.1
Created fine-tune: ft-zzzzzzz Streaming events until fine-tuning is complete...
(Ctrl-C will interrupt the stream, but not cancel the fine-tune) [2023-01-20 21:00:42] Created fine-tune: ft-QuUrna0IUGSDkTai9rj4VnZU
Stream interrupted (client disconnected). To resume the stream, run:
openai api fine_tunes.follow -i ft-zzzzzzz
same issue .. any update
[2023-01-20 21:04:08] Created fine-tune: ft-*****
Stream interrupted (client disconnected). To resume the stream, run:
openai api fine_tunes.follow -i ft-*****
Hey all, in my case this was not an issue. This 'interuption' happens on my end with 0.26.1 but this is not due to a bug. I think it is related to session time out and job will be in the queue. The fine tuning process may take several days depens on the nunber of jobs in the queue. Cheers M
On Sat, 21 Jan 2023, 16:29 br185126, @.***> wrote:
same issue .. any update
— Reply to this email directly, view it on GitHub https://github.com/openai/openai-python/issues/173#issuecomment-1399165678, or unsubscribe https://github.com/notifications/unsubscribe-auth/A2EOQMJV5TDQMMOUMGIRESDWTNJ3HANCNFSM6AAAAAATUJXEU4 . You are receiving this because you were mentioned.Message ID: @.***>
Hey all, in my case this was not an issue. This 'interuption' happens on my end with 0.26.1 but this is not due to a bug. I think it is related to session time out and job will be in the queue. The fine tuning process may take several days depens on the nunber of jobs in the queue. Cheers M … On Sat, 21 Jan 2023, 16:29 br185126, @.> wrote: same issue .. any update — Reply to this email directly, view it on GitHub <#173 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/A2EOQMJV5TDQMMOUMGIRESDWTNJ3HANCNFSM6AAAAAATUJXEU4 . You are receiving this because you were mentioned.Message ID: @.>
I gave it only 2 prompts to train on
Hey all, in my case this was not an issue. This 'interuption' happens on my end with 0.26.1 but this is not due to a bug. I think it is related to session time out and job will be in the queue. The fine tuning process may take several days depens on the nunber of jobs in the queue. Cheers M … On Sat, 21 Jan 2023, 16:29 br185126, @.> wrote: same issue .. any update — Reply to this email directly, view it on GitHub <#173 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/A2EOQMJV5TDQMMOUMGIRESDWTNJ3HANCNFSM6AAAAAATUJXEU4 . You are receiving this because you were mentioned.Message ID: @.>
I think you are right. My dataset is not large. When i use the model cuire, there is no response.When i change to ada, it works
Workaround: loop to execute it until it's success
fish shell
while true; openai api fine_tunes.follow -i ft-xxxxxxxxx ; end
or
bash shell
while true; do openai api fine_tunes.follow -i ft-xxxxxxx ; done
I believe it also works if you just update openai to the latest version.
is everyone still having this issue? I keep getting the same interruption "Stream interrupted (client disconnected).", i'm hoping that its just the stream that is interrupted and not the fine-tune. Maybe just because of the amount fo tine tunes in the queue?
Hi, submited a training file yesterday and it got disconnected but it has been submitted for fine tune. There is a command ( mentioned below) that shows you the fine tune processes/models that are happening/have been created so far along with their status and other info.
4
6 7 8 9 10# List all created fine-tunes openai api fine_tunes.list
On Wed, 22 Feb 2023, 07:52 sjonas50, @.***> wrote:
is everyone still having this issue? I keep getting the same interruption "Stream interrupted (client disconnected).", i'm hoping that its just the stream that is interrupted and not the fine-tune. Maybe just because of the amount fo tine tunes in the queue?
— Reply to this email directly, view it on GitHub https://github.com/openai/openai-python/issues/173#issuecomment-1438949308, or unsubscribe https://github.com/notifications/unsubscribe-auth/A2EOQMNATDHW4Q2F43JYH3LWYUFGNANCNFSM6AAAAAATUJXEU4 . You are receiving this because you were mentioned.Message ID: @.***>
is everyone still having this issue? I keep getting the same interruption "Stream interrupted (client disconnected).", i'm hoping that its just the stream that is interrupted and not the fine-tune. Maybe just because of the amount fo tine tunes in the queue?
same error
It ended up working after a while. Just give it time then you will see it gets added to the queue. Took about 3 hours.
On Tue, Feb 21, 2023 at 7:59 PM wallon-ai @.***> wrote:
is everyone still having this issue? I keep getting the same interruption "Stream interrupted (client disconnected).", i'm hoping that its just the stream that is interrupted and not the fine-tune. Maybe just because of the amount fo tine tunes in the queue?
same error
— Reply to this email directly, view it on GitHub https://github.com/openai/openai-python/issues/173#issuecomment-1439370631, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADHISBMAHHNR2N6XF3UZN73WYV6ITANCNFSM6AAAAAATUJXEU4 . You are receiving this because you commented.Message ID: @.***>
Same issue here, keep getting stream interrupted (client disconnected). I'm just preloading a bunch of openai api fine_tunes.follow -i
commands so it'll go through for now :(