cortex icon indicating copy to clipboard operation
cortex copied to clipboard

Production infrastructure for machine learning at scale

Results 121 cortex issues
Sort by recently updated
recently updated
newest added
trafficstars

#### Description When deploying a new API, we only validate against the scheduled workloads on the cluster. We're not taking into consideration the situation where all of the existing APIs...

enhancement
ux
provisioning

### Description https://eksctl.io/usage/cloudwatch-cluster-logging/ Consider using a retention policy on the log groups ### Motivation * Expose more logs * Remove the misleading message regarding cloudwatch logging during `cluster up`

enhancement
provisioning

#### Description When creating a cluster with node group names that have more than a few characters, then ARN names of the corresponding node groups are getting truncated. Is this...

bug
question

#### Description Assume we have these node groups: * A CPU node group with a live instance. This can fit 10 iris-classifier replicas (which only request CPU). Max_instances is set...

bug
performance

#### Description The status of a Job can take up to 60 seconds to reflect the actual status of a Job because the all job statuses are updated together in...

enhancement
refactor
BatchAPI

### Version 0.25 ### Stack traces ```text ValueError: sleep length must be non-negative File "threading.py", line 864, in run self._target(*self._args, **self._kwargs) File "batch.py", line 86, in renew_message_visibility time.sleep((cur_time + interval)...

bug
BatchAPI

### Version 0.25 ### Stack traces ```text ClientError: An error occurred (InvalidParameterValue) when calling the DeleteMessage operation: Value for parameter ReceiptHandle is invalid. Reason: The receipt handle has... File "batch.py",...

bug
BatchAPI

### Description AsyncAPI currently only supports strings, JSON, and binary payloads. Multipart/form-data is a very common payload type and could be particularly useful for this API type. ### Motivation This...

enhancement
AsyncAPI

### Description Allow different levels of access for different CLI users. To start, it could be just two levels: list/get or list/get + deploy/delete. #### Proposal - [ ] use...

enhancement

"expected_json_schema" Also consider adding tests that the GPU version of text-generator is faster than the CPU version, and/or log regex validation