That is my workflow, code to reproduce:

def test_something():
    # 1. Create a new task
    task = Task.create(
        project_name="Playground",
        task_name="test",
    )
    # 2. Create a new pandas data frame and upload as artifact
    test_df = pd.DataFrame(
        {
            "col1": [1],
            "col2": [2],
        }
    )
    task.upload_artifact("test_df", test_df)
    task_id = task.task_id
    task.close()

    task = task.init(
        project_name="Playground",
        task_name="test",
        reuse_last_task_id=task_id,
        continue_last_task=True)

    # 3. Download the pandas dataframe
    downloaded_df = task.artifacts["test_df"].get()

    # 4. Add a new row to the data frame and upload again with the same name (doc says this is then an update)
    new_row = {
        "col1": 3,
        "col2": 4,
    }
    updated_df = downloaded_df.append(new_row, ignore_index=True)
    task.upload_artifact("test_df", updated_df)
    task.close()
    task = task.init(
        project_name="Playground",
        task_name="test",
        reuse_last_task_id=task_id,
        continue_last_task=True)


    # 5. Download the pandas dataframe again -> it still has one row (loaded first version from cache !)
    downloaded_df = task.artifacts["test_df"].get()

Also, make sure you use Task.init instead of task.init

Glad I could be of help

Hey @<1547390444877385728:profile|ThickSnake12> , how exactly do you access the artifact next time? Can you provide a code sample?

You can try to add the force_download=True flag to .get() to ignore the locally cached content. Let me know if it helps.

yes thx a lot, it works ! i have not seen the parameter, and the Task.init is correct…