Set torchserve version to 0.7.1 for compatibility #1815
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…pen-mmlab#1756) * feat: impelemt DINO * chore: delete debug code * chore: impplement pre-commit * fix: fix imported package * chore: pre-commit check
…open-mmlab#1774) * add new config adapting MobileNetV2,V3 * add base model config for mobile net v3, modified all training configs of mobile net v3 inherit from the base model config * removed directory _base_/models/mobilenet_v3
* zero-shot CLIP * modify zero-shot clip config * add in1k_sub_prompt(8 prompts) for improvement * add some annotations doc * clip base class & clip_zs sub-class * some modifications of details after review * convert into and use mmpretrain-vit * modify names of some files and directories
|
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Motivation
In the Dockerfile and the mmpretrain2torchserve.py files there is no clear reference to any version of torch-model-archiver, it is just the latest (
pip install torchserve torch-model-archiver, currently it is 0.8.2), however from 0.8.0 there is a change that breaks the script for exporting to TorchServe MAR file.To reproduce the error, simply install torch-model-archiver with any version higher that 0.7.1 (i.e.
pip install torchserve==0.8.0 torch-model-archiver==0.8.0).An error coming from the line 68 of "tools/torchserve/mmpretrain2torchserve.py" (
manifest = ModelExportUtils.generate_manifest_json(args)) will raise. The error is:AttributeError: 'Namespace' object has no attribute 'config_file'Modification
To avoid this, the developer needs to set the version to 0.7.1. I suggest setting this in the Dockerfile and the script.
Checklist
Before PR:
After PR:
](https://mmpretrain.readthedocs.io/en/latest/useful_tools/model_serving.html)