Fix GQA permutation computation and sequential weight initialization / loading when doing TP #364
Triggered via pull request
March 27, 2024 14:55
Status
Cancelled
Total duration
1h 41m 17s
Artifacts
–
test_trainium_distributed.yml
on: pull_request
optimum-neuron-tests
0s
Annotations
1 error
optimum-neuron-tests
Canceling since a higher priority waiting request for 'Optimum Neuron - Test optimum.neuron.distributed on Trainium-fix_gqa_compute_query_indicies' exists
|