MLEM requires a flyctl version of <0.1.0, but the most recent version are > 0.1.0
The following error is observed when following the steps in https://iterative.ai/blog/mlem-nanogpt-modal-flyio/
mlem deploy run flyio app -m mlem_char \
--app_name mlem-test-nanogpt --scale_memory 1024 \
--server streamlit --server.ui_port 8080 \
--server.server_port 8081 --server.template app.py
⏳️ Loading deployment from app.mlem
:floppy_disk: Saving deployment to app.mlem
⏳️ Loading model from mlem_char.mlem
automatically selected personal organization: MyName
:x: Unexpected error: 'Key "services" does not exist.'
Use the --tb or --traceback option to include the traceback in the output
Please report it here: <https://github.com/iterative/mlem/issues>
Printing the fly.toml from meta.py when running the original deploy command:
mlem deploy run flyio app -m mlem_char \
--app_name mlem-test-nanogpt --scale_memory 1024 \
--server streamlit --server.ui_port 8080 \
--server.server_port 8081 --server.template app.py
gives the following format, which is different from the fly.toml expected by MLEM
{'app': 'mlem-test-nanogpt', 'primary_region': 'lax', 'http_service': {'internal_port': 8080, 'force_https': True, 'auto_stop_machines': True, 'auto_start_machines': True, 'min_machines_running': 0}}
Pinning the flyctl version to v0.0.559 should resolve this.
MLEM requires a flyctl version of <0.1.0, but the most recent version are > 0.1.0
The following error is observed when following the steps in https://iterative.ai/blog/mlem-nanogpt-modal-flyio/
Printing the fly.toml from meta.py when running the original deploy command:
gives the following format, which is different from the fly.toml expected by MLEM
{'app': 'mlem-test-nanogpt', 'primary_region': 'lax', 'http_service': {'internal_port': 8080, 'force_https': True, 'auto_stop_machines': True, 'auto_start_machines': True, 'min_machines_running': 0}}Pinning the flyctl version to v0.0.559 should resolve this.