Closed Mayukhdeb closed 3 years ago
Changes made:
/run starts the task and returns something like:
/run
{ 'token': str }
/fetch has 3 possibilities based on the status of the tasks
/fetch
If the task is queued:
{ 'status': { 'status': 'queued', 'queue_position': int }, 'config': dict, }
If the task is running:
{ 'status': { 'status': 'running', 'progress': float }, 'config': dict, 'output': dict }
If the task is complete:
{ 'status': { 'status': 'complete' }, 'output': dict, 'config': dict }
If the task has failed:
{ 'status': { 'status': 'failed' }, 'output': dict, 'config': dict }
/update returns either something like:
/update
{ 'status': { 'status': 'successfully updated config' }, }
or if the task is already complete:
{ 'status': { 'status': 'task is aleady complete' }, }
or if the token is invalid:
{ 'status': { 'status': 'invalid token' }, }
python3 eden/tests/server.py
pytest eden/tests
torch
requirements.txt
Changes made:
/run
starts the task and returns something like:/fetch
has 3 possibilities based on the status of the tasksIf the task is queued:
If the task is running:
If the task is complete:
If the task has failed:
/update
returns either something like:or if the task is already complete:
or if the token is invalid:
/update
. The tests can be run with the following steps:torch
as a dependencyrequirements.txt