Adding retry properties to AWS batch job in terraform

I want to add retry strategy in aws_batch_job_definition of HCL. I tried using:

retry_strategy = {
  "attempts" = "2" 
}

The deployment is giving me an error

{", "\u001b[0m", "An argument named \"retry_strategy\" is not expected here. Did you mean to", "define a block of type \"retry_strategy\"?", "\u001b[0m\u001b[0m"]}

What am I doing wrong?

I don't have an entire code block to show how the resource would look like, but the error is self-explanatory. What you are trying to do is assign a value to an argument named retry_stratagy, which is actually a block (not argument), so the final outlook is:

 retry_strategy {
    attempts = 2 
 }

An example of the resource block with only required arguments:

resource "aws_batch_job_definition" "name" {
  name = "value"
  type = "container"

  retry_strategy {
    attempts = 2
  }
  
}

Output of terraform plan with the above configuration:

Terraform will perform the following actions:

  # aws_batch_job_definition.name will be created
  + resource "aws_batch_job_definition" "name" {
      + arn            = (known after apply)
      + id             = (known after apply)
      + name           = "value"
      + propagate_tags = false
      + revision       = (known after apply)
      + tags_all       = (known after apply)
      + type           = "container"

      + retry_strategy {
          + attempts = 2
        }
    }