Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

Heya! :wave: When trying to chain `task_1` output with `task_2` input I get the following error `TypeError: Object of type Promise is not JSON serializable`. I get `Promise` is only resolved at a deferred moment but how to pipe one’s output into a downstream task?
Whole script in :thread: :point_down:

```from flytekit import workflow, task

from flytekitplugins.bacalhau import BacalhauTask
from flytekit import kwtypes


bacalhau_task = BacalhauTask(
        name="hello_world",
        inputs=kwtypes(
                spec=dict,
                api_version=str,
        )
    )

bacalhau_task_2 = BacalhauTask(
        name="second_hello_world",
        inputs=kwtypes(
                spec=dict,
                api_version=str,
        )
    )


@task
def print_cid(bac_task: str) -&gt; str:
    print(f"Your Bacalhau's output CID: {bac_task}")
    return bac_task

@workflow
def wf():
    bac_task_1 = bacalhau_task(
        api_version="V1beta1",
        spec=dict(
            engine="Docker",
            verifier="Noop",
            PublisherSpec={"type": "IPFS"},
            docker={
                "image": "ubuntu",
                "entrypoint": ["echo", "Flyte is awesome!"],
            },
            language={"job_context": None},
            wasm=None,
            resources=None,
            timeout=1800,
            outputs=[{
                    "storage_source": "IPFS",
                    "name": "outputs",
                    "path": "/outputs",
            }],
            deal={"concurrency": 1},
        ),
    )

    bac_task_2 = bacalhau_task_2(
        api_version="V1beta1",
        spec=dict(
            engine="Docker",
            verifier="Noop",
            PublisherSpec={"type": "IPFS"},
            docker={
                "image": "ubuntu",
                "entrypoint": ["echo", "Flyte is awesome!"],
            },
            language={"job_context": None},
            wasm=None,
            resources=None,
            timeout=1800,
            outputs=[{
                    "storage_source": "IPFS",
                    "name": "outputs",
                    "path": "/outputs",
            }],
            inputs=[{
                    "storage_source": "IPFS",
                    "cid": bac_task_1,
                    "name": "myinputs",
                    "path": "/myinputs",
            }],
            deal={"concurrency": 1},
        ),
    )
    
    print_cid(bac_task=bac_task_2)


if __name__ == "__main__":
    wf()```

basically in the second task I’m trying to pass `bac_task_1`  in `spec.inputs.cid`

you have to create a dynamic task
```@dynamic
def d1(task: str):
  bac_task_2 = bacalhau_task_2(
        cicd: task
   ...```

Done it thanks!
Here’s my code for future Flyte diggers: <https://github.com/bacalhau-project/bacalhau/blob/flyte-integration/integration/flyte/plugins/flytekit-bacalhau/examples/chained-jobs.py>