Tasks
The tasks are a powerful tool which can help your team to annotate inputs from your application.
To create a new task in your app you POST the task information to v2/task endpoint.
A task should be assigned to a list of users, but it's not required. The following code will create a non-assigned task.
curl -X POST \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \-d '{"tasks": [{"type": "CONCEPTS_CLASSIFICATION","name": "Annotate {{concept_id}}","worker": {"strategy": "FULL"},"concept_ids": ["{{concept_id}}"],"input_source": {"type": "ALL_INPUTS"},"sample_ms": 1000,"review": {"strategy": "NONE"}}]}'\https://api.clarifai.com/v2/tasks
A task should be assigned to a list of users. These users will do the work, so they're also called workers. A task may also be assigned to a list of users for review.
curl -X POST \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \-d '{"tasks": [{"type": "CONCEPTS_CLASSIFICATION","name": "Annotate {{concept_id}}","worker": {"strategy": "FULL","user_ids": ["{{worker_user_id}}"]},"concept_ids": ["{{concept_id}}"],"input_source": {"type": "ALL_INPUTS"},"sample_ms": 1000,"review": {"strategy": "MANUAL","manual_strategy_info": {"sample_percentage": 0.5},"user_ids": ["{{reviewer_user_id}}"]}}]}'\https://api.clarifai.com/v2/tasks
The previous tasks were created with full worker strategy.
{"strategy": "FULL"}
In case of FULL worker strategy, each worker will work on all inputs selected in the input source.
If you wish the work to be distributed between workers, then you can select the PARTITIONED worker strategy.
In the following example:
there are two workers
workers_per_input: each input will be assigned to 1 workerweights.{{user_id1}}: the first worker will get 90% of inputsweights.{{user_id2}}: the second worker will get 10% of inputs
curl -X POST \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \-d '{"tasks": [{"type": "CONCEPTS_CLASSIFICATION","name": "Annotate {{concept_id}}","worker": {"strategy": "PARTITIONED","user_ids": ["{{user_id1}}", "{{user_id2}}"],"partitioned_strategy_info": {"type": "WEIGHTED","workers_per_input": 1,"weights": {"{{user_id1}}": 90,"{{user_id2}}": 10}}},"concept_ids": ["{{concept_id}}"],"input_source": {"type": "ALL_INPUTS"},"sample_ms": 1000,"review": {"strategy": "NONE"}}]}'\https://api.clarifai.com/v2/tasks
Notes:
It is not required for the weights to add up to 100. For example, the weights [9, 1] are equivalent with weights [90, 10].
The partitioning is approximate. This means that the number of assigned inputs to each worker may have a small error margin, but it will be close to the assigned weight percentage.
The previous tasks were created with no review or manual review strategy.
{"strategy": "MANUAL"}
We recommend to create tasks with CONSENSUS review strategy. When enough workers label an input in the same way, it will automatically be approved, with no need for the reviewer to spend time to check. In this way, the reviewer will be able to focus on the inputs where the workers don't agree.
Note that an approval threshold must be set. For example, in case of 3 workers and approval_threshold set to 2, if an input is labeled in the same way by 2 workers, they form a majority and the group reaches a consensus.
curl -X POST \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \-d '{"tasks": [{"type": "CONCEPTS_CLASSIFICATION","name": "Annotate {{concept_id}}","worker": {"strategy": "PARTITIONED","user_ids": ["{{user_id1}}", "{{user_id2}}", "{{user_id3}}"],"partitioned_strategy_info": {"type": "WEIGHTED","workers_per_input": 1,"weights": {"{{user_id1}}": 1,"{{user_id2}}": 1,"{{user_id3}}": 1}}},"concept_ids": ["{{concept_id}}"],"input_source": {"type": "ALL_INPUTS"},"sample_ms": 1000,"review": {"strategy": "CONSENSUS","consensus_strategy_info": {"approval_threshold": 2},"user_ids": ["{{user_id4}}"]}}]}'\https://api.clarifai.com/v2/tasks
You can get a singular task by its ID. The ID was automatically generated when it was created.
curl -X GET \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \https://api.clarifai.com/v2/tasks/{task_id}
You can get a list of tasks within your app with a GET call. This call supports pagination.
curl -X GET \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \https://api.clarifai.com/v2/tasks
Get only the tasks assigned to a specific user for work.
curl -X GET \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \https://api.clarifai.com/v2/tasks?worker_user_ids={{user_id}}
Get only the tasks assigned to a specific user for review.
curl -X GET \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \https://api.clarifai.com/v2/tasks?review_user_ids={{user_id}}
Currently, we only support updating a task by providing all information at once.
curl -X PATCH \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \-d '{"action": "overwrite","tasks": [{"id": "{{task_id}}","type": "CONCEPTS_CLASSIFICATION","name": "Annotate {{concept_id}}","worker": {"strategy": "PARTITIONED","user_ids": ["{{user_id1}}", "{{user_id2}}"],"partitioned_strategy_info": {"type": "WEIGHTED","workers_per_input": 1,"weights": {"{{user_id1}}": 1,"{{user_id2}}": 1}}},"concept_ids": ["{{concept_id}}"],"input_source": {"type": "ALL_INPUTS"},"sample_ms": 1000,"review": {"strategy": "CONSENSUS","consensus_strategy_info": {"approval_threshold": 2},"user_ids": ["{{user_id3}}"]},"status": {"code": "TASK_DONE"}}]}'\https://api.clarifai.com/v2/tasks
You can delete tasks using their IDs.
curl -X DELETE \-H "Authorization: Key YOUR_API_KEY" \-H "Content-Type: application/json" \-d '{"ids":["{{task_id}}"]}'\https://api.clarifai.com/v2/tasks
