[Crowdsourcing] can call blueprints from command line

parlai/crowdsourcing/README.md

@@ -6,7 +6,7 @@ Code for crowdsourcing tasks that use Mephisto. See the [Mephisto quick start gu
 
 ## Running tasks
 
-Tasks are launched by calling the appropriate run script: for instance, an ACUTE-Eval run can be launched with `python parlai/crowdsourcing/tasks/acute_eval/run.py`, followed by any appropriate flags. All run parameters are set using [Hydra](https://github.com/facebookresearch/hydra): append the flag `-c job` to your run command to see a list of all available parameters, grouped by their package name (`mephisto.blueprint`, `mephisto.task`, etc.), which determines how they are called. Each run script has a YAML file of default parameters that will be loaded, found in the `hydra_configs/conf/` subfolder of each task.
+Tasks are launched by calling the appropriate run script: for instance, an ACUTE-Eval run can be launched with `python parlai/crowdsourcing/tasks/acute_eval/run.py`, followed by any appropriate flags. All run parameters are set using [Hydra](https://github.com/facebookresearch/hydra): append the flag `-c job` to your run command to see a list of all available parameters, grouped by their package name (`mephisto.blueprint`, `mephisto.task`, etc.), which determines how they are called. Each run script points to a YAML file of default parameters (including the blueprints) that will be loaded, found in the `hydra_configs/conf/` subfolder of each task. You can specify a different blueprint in a YAML file and use that YAML file to run your task (see below).
 
 ### Specifying your own YAML file
 
@@ -24,6 +24,8 @@ mephisto:
 ```
 If you want to quickly modify this parameter to, say, 0.6 without changing the YAML file, you can add a `mephisto.task.task_reward=0.6` string to your launch command.
 
+You can also specify a blueprint on the command line by adding `mephisto/blueprint=my_blueprint_type` (this will override the default blueprint type if defined in the config file).
+
 ### MTurk-specific task configuration
 
 Here is a partial list of MTurk-specific parameters that can be set in YAML files or on the command line:

parlai/crowdsourcing/tasks/model_chat/README.md

@@ -6,7 +6,7 @@ This task will collect conversations between a human and a model. After each res
 
 ## Launching
 
-Call `run.py` to run this task with the default parameters, as set by `conf/example.yaml`. Some parameters that you can adjust include where to save data, lists of workers to soft-block, the maximum response time, etc.
+Call `run.py` to run this task with the default parameters, as set by `hydra_configs/conf/example.yaml`. Some parameters that you can adjust include where to save data, lists of workers to soft-block, the maximum response time, etc.
 
 Set `mephisto.blueprint.model_opt_path` to specify a path to a YAML file listing all models to be chatted with, as well as the ParlAI flags for running each one. See `task_config/model_opts.yaml` for an example.
 
@@ -26,7 +26,7 @@ In `worlds.py`, modify `ModelChatOnboardWorld.check_onboarding_answers()` to cha
 
 ## Human+model image chat
 
-`run_image_chat.py` can be run to chat with a model about an image: each conversation will begin with a selected image, and then the human and model will chat about it.
+Call `run.py conf=example_image_chat` to chat with a model about an image: each conversation will begin with a selected image, and then the human and model will chat about it. This task is run with the parameters defined in `hydra_configs/conf/example_image_chat.yaml`.
 
 This code replaces the old `parlai/mturk/tasks/image_chat/` and `parlai/mturk/tasks/personality_captions/` tasks, which are deprecated and can be accessed with `git checkout v0.10.0`. Those tasks also featured the ability to compare two possible captions to an image and rate which one is more engaging: this functionality has now been replaced by the [ACUTE-Eval](https://github.com/facebookresearch/ParlAI/tree/main/parlai/crowdsourcing/tasks/acute_eval) task.
 

parlai/crowdsourcing/tasks/turn_annotations_static/README.md

@@ -5,9 +5,11 @@ Two variants of the blueprint are supported:
 - `TurnAnnotationStaticBlueprint`
     - The base static turn-annotations task
     - Called with `python parlai/crowdsourcing/tasks/turn_annotations_static/run.py`
+        - (the task runs with the default parameters set in `hydra_configs/conf/example.yaml`)
 - `TurnAnnotationStaticInFlightQABlueprint`
     - Includes the ability to add an additional in-flight (i.e. mid-HIT) quality assurance check
-    - Called with `python parlai/crowdsourcing/tasks/turn_annotations_static/run_in_flight_qa.py`
+    - Called with `python parlai/crowdsourcing/tasks/turn_annotations_static/run.py conf=example_in_flight_qa`
+        - (the task runs with the parameters set in `hydra_configs/conf/example_in_flight.yaml`)
 
 For both variants of the blueprint, it is required to pass in your own file of conversations with `mephisto.blueprint.data_jsonl=${PATH_TO_CONVERSATIONS}`.
 

tests/crowdsourcing/tasks/model_chat/test_model_image_chat.py

@@ -41,7 +41,7 @@
 try:
 
     import parlai.crowdsourcing.tasks.model_chat.worlds_image_chat as world_module
-    from parlai.crowdsourcing.tasks.model_chat.run_image_chat import TASK_DIRECTORY
+    from parlai.crowdsourcing.tasks.model_chat.run import TASK_DIRECTORY
     from parlai.crowdsourcing.tasks.model_chat.model_chat_blueprint import (
         SharedModelImageChatTaskState,
     )