-
Notifications
You must be signed in to change notification settings - Fork 80
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Custom trained models #1206
Custom trained models #1206
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
could you refactor to reduce the number of plugins (i.e. directories in brainscore_vision/models/
)? Just use one plugin for e.g. all convnexts, one for all resnets etc. Or even just one plugin for everything. You can define multiple models in each plugin.
Also you committed two .pkl
files by accident, please remove.
* point to pypi instead of git (#1236) * update dependency names w/ hyphens (#1238) * point to pypi instead of git * Update pyproject.toml with hyphens * Custom trained scaling models (#1206) * Custom trained models * Fix missing library * Fix a missing library * Not every model needs timm * Trigger build tests * Use model_id instead of model_name to ensure correct weights are loaded * timm is missing from deit models * timm is missing from deit models * Fix requirements.txt newline * Package models into a single plugin * Update requirements.txt * Update requirements.txt * Update requirements.txt * Change the order of models * Update model.py * Disable resultcaching for vits * Tests reordering * Revert tests back * Update model.py * Update model.py * Update to sync workflow (#1248) * Update to sync workflow Adds dynamic PR title and commit history included in PR description. Adds `No changes detected` job. * Update no_changes condition and text Refined condition to not require PR_merge because that is when this condition SHOULD be triggered to indicate the correct status of the workflow. --------- Co-authored-by: Sam Winebrake <[email protected]> Co-authored-by: Abdulkadir Gokce <[email protected]>
* point to pypi instead of git (#1236) * update dependency names w/ hyphens (#1238) * point to pypi instead of git * Update pyproject.toml with hyphens * Custom trained scaling models (#1206) * Custom trained models * Fix missing library * Fix a missing library * Not every model needs timm * Trigger build tests * Use model_id instead of model_name to ensure correct weights are loaded * timm is missing from deit models * timm is missing from deit models * Fix requirements.txt newline * Package models into a single plugin * Update requirements.txt * Update requirements.txt * Update requirements.txt * Change the order of models * Update model.py * Disable resultcaching for vits * Tests reordering * Revert tests back * Update model.py * Update model.py * Update to sync workflow (#1248) * Update to sync workflow Adds dynamic PR title and commit history included in PR description. Adds `No changes detected` job. * Update no_changes condition and text Refined condition to not require PR_merge because that is when this condition SHOULD be triggered to indicate the correct status of the workflow. * enable unverified SSL for scaling models (#1263) --------- Co-authored-by: Sam Winebrake <[email protected]> Co-authored-by: Abdulkadir Gokce <[email protected]>
* point to pypi instead of git (#1236) * update dependency names w/ hyphens (#1238) * point to pypi instead of git * Update pyproject.toml with hyphens * Custom trained scaling models (#1206) * Custom trained models * Fix missing library * Fix a missing library * Not every model needs timm * Trigger build tests * Use model_id instead of model_name to ensure correct weights are loaded * timm is missing from deit models * timm is missing from deit models * Fix requirements.txt newline * Package models into a single plugin * Update requirements.txt * Update requirements.txt * Update requirements.txt * Change the order of models * Update model.py * Disable resultcaching for vits * Tests reordering * Revert tests back * Update model.py * Update model.py * Update to sync workflow (#1248) * Update to sync workflow Adds dynamic PR title and commit history included in PR description. Adds `No changes detected` job. * Update no_changes condition and text Refined condition to not require PR_merge because that is when this condition SHOULD be triggered to indicate the correct status of the workflow. * enable unverified SSL for scaling models (#1263) --------- Co-authored-by: Sam Winebrake <[email protected]> Co-authored-by: Abdulkadir Gokce <[email protected]>
* point to pypi instead of git (#1236) * update dependency names w/ hyphens (#1238) * point to pypi instead of git * Update pyproject.toml with hyphens * Custom trained scaling models (#1206) * Custom trained models * Fix missing library * Fix a missing library * Not every model needs timm * Trigger build tests * Use model_id instead of model_name to ensure correct weights are loaded * timm is missing from deit models * timm is missing from deit models * Fix requirements.txt newline * Package models into a single plugin * Update requirements.txt * Update requirements.txt * Update requirements.txt * Change the order of models * Update model.py * Disable resultcaching for vits * Tests reordering * Revert tests back * Update model.py * Update model.py * Update to sync workflow (#1248) * Update to sync workflow Adds dynamic PR title and commit history included in PR description. Adds `No changes detected` job. * Update no_changes condition and text Refined condition to not require PR_merge because that is when this condition SHOULD be triggered to indicate the correct status of the workflow. * enable unverified SSL for scaling models (#1263) * Add pathfinder benchmark (#1193) * add pathfinder task * update data packaging * modify benchmark file * add benchmark tests * upload datasets * add some tests and fix some bugs * add dataset tests * fix errors with plugin imports * add error with stimulus set naming test * fix bug with field name test * remove field from test that shouldn't be there * fix model loading bug in test * fix ceiled test value -> raw test value --------- Co-authored-by: Martin Schrimpf <[email protected]> --------- Co-authored-by: Sam Winebrake <[email protected]> Co-authored-by: Abdulkadir Gokce <[email protected]> Co-authored-by: Ben Lonnqvist <[email protected]> Co-authored-by: Martin Schrimpf <[email protected]>
* point to pypi instead of git (#1236) * update dependency names w/ hyphens (#1238) * point to pypi instead of git * Update pyproject.toml with hyphens * Custom trained scaling models (#1206) * Custom trained models * Fix missing library * Fix a missing library * Not every model needs timm * Trigger build tests * Use model_id instead of model_name to ensure correct weights are loaded * timm is missing from deit models * timm is missing from deit models * Fix requirements.txt newline * Package models into a single plugin * Update requirements.txt * Update requirements.txt * Update requirements.txt * Change the order of models * Update model.py * Disable resultcaching for vits * Tests reordering * Revert tests back * Update model.py * Update model.py * Update to sync workflow (#1248) * Update to sync workflow Adds dynamic PR title and commit history included in PR description. Adds `No changes detected` job. * Update no_changes condition and text Refined condition to not require PR_merge because that is when this condition SHOULD be triggered to indicate the correct status of the workflow. * enable unverified SSL for scaling models (#1263) * Add pathfinder benchmark (#1193) * add pathfinder task * update data packaging * modify benchmark file * add benchmark tests * upload datasets * add some tests and fix some bugs * add dataset tests * fix errors with plugin imports * add error with stimulus set naming test * fix bug with field name test * remove field from test that shouldn't be there * fix model loading bug in test * fix ceiled test value -> raw test value --------- Co-authored-by: Martin Schrimpf <[email protected]> * Ep/add fixres resnext101 32x48d wsl (#1103) * Add model * Modify test * Fix imports --------- Co-authored-by: Ethan Pellegrini <[email protected]> Co-authored-by: Kartik Pradeepan <[email protected]> --------- Co-authored-by: Sam Winebrake <[email protected]> Co-authored-by: Abdulkadir Gokce <[email protected]> Co-authored-by: Ben Lonnqvist <[email protected]> Co-authored-by: Martin Schrimpf <[email protected]> Co-authored-by: pellegreene <[email protected]> Co-authored-by: Ethan Pellegrini <[email protected]>
* point to pypi instead of git (#1236) * update dependency names w/ hyphens (#1238) * point to pypi instead of git * Update pyproject.toml with hyphens * Custom trained scaling models (#1206) * Custom trained models * Fix missing library * Fix a missing library * Not every model needs timm * Trigger build tests * Use model_id instead of model_name to ensure correct weights are loaded * timm is missing from deit models * timm is missing from deit models * Fix requirements.txt newline * Package models into a single plugin * Update requirements.txt * Update requirements.txt * Update requirements.txt * Change the order of models * Update model.py * Disable resultcaching for vits * Tests reordering * Revert tests back * Update model.py * Update model.py * Update to sync workflow (#1248) * Update to sync workflow Adds dynamic PR title and commit history included in PR description. Adds `No changes detected` job. * Update no_changes condition and text Refined condition to not require PR_merge because that is when this condition SHOULD be triggered to indicate the correct status of the workflow. * enable unverified SSL for scaling models (#1263) * Add pathfinder benchmark (#1193) * add pathfinder task * update data packaging * modify benchmark file * add benchmark tests * upload datasets * add some tests and fix some bugs * add dataset tests * fix errors with plugin imports * add error with stimulus set naming test * fix bug with field name test * remove field from test that shouldn't be there * fix model loading bug in test * fix ceiled test value -> raw test value --------- Co-authored-by: Martin Schrimpf <[email protected]> * Ep/add fixres resnext101 32x48d wsl (#1103) * Add model * Modify test * Fix imports --------- Co-authored-by: Ethan Pellegrini <[email protected]> Co-authored-by: Kartik Pradeepan <[email protected]> * Ep/add inception v4 (#1099) * Add inception_v4 * Update model.py * Update model.py * Fix layer mapping * Fix imports --------- Co-authored-by: Ethan Pellegrini <[email protected]> --------- Co-authored-by: Sam Winebrake <[email protected]> Co-authored-by: Abdulkadir Gokce <[email protected]> Co-authored-by: Ben Lonnqvist <[email protected]> Co-authored-by: Martin Schrimpf <[email protected]> Co-authored-by: pellegreene <[email protected]> Co-authored-by: Ethan Pellegrini <[email protected]>
* point to pypi instead of git (#1236) * update dependency names w/ hyphens (#1238) * point to pypi instead of git * Update pyproject.toml with hyphens * Custom trained scaling models (#1206) * Custom trained models * Fix missing library * Fix a missing library * Not every model needs timm * Trigger build tests * Use model_id instead of model_name to ensure correct weights are loaded * timm is missing from deit models * timm is missing from deit models * Fix requirements.txt newline * Package models into a single plugin * Update requirements.txt * Update requirements.txt * Update requirements.txt * Change the order of models * Update model.py * Disable resultcaching for vits * Tests reordering * Revert tests back * Update model.py * Update model.py * Update to sync workflow (#1248) * Update to sync workflow Adds dynamic PR title and commit history included in PR description. Adds `No changes detected` job. * Update no_changes condition and text Refined condition to not require PR_merge because that is when this condition SHOULD be triggered to indicate the correct status of the workflow. * enable unverified SSL for scaling models (#1263) * Add pathfinder benchmark (#1193) * add pathfinder task * update data packaging * modify benchmark file * add benchmark tests * upload datasets * add some tests and fix some bugs * add dataset tests * fix errors with plugin imports * add error with stimulus set naming test * fix bug with field name test * remove field from test that shouldn't be there * fix model loading bug in test * fix ceiled test value -> raw test value --------- Co-authored-by: Martin Schrimpf <[email protected]> * Ep/add fixres resnext101 32x48d wsl (#1103) * Add model * Modify test * Fix imports --------- Co-authored-by: Ethan Pellegrini <[email protected]> Co-authored-by: Kartik Pradeepan <[email protected]> * Ep/add inception v4 (#1099) * Add inception_v4 * Update model.py * Update model.py * Fix layer mapping * Fix imports --------- Co-authored-by: Ethan Pellegrini <[email protected]> * Fixed pip install git (#1280) --------- Co-authored-by: Sam Winebrake <[email protected]> Co-authored-by: Abdulkadir Gokce <[email protected]> Co-authored-by: Ben Lonnqvist <[email protected]> Co-authored-by: Martin Schrimpf <[email protected]> Co-authored-by: pellegreene <[email protected]> Co-authored-by: Ethan Pellegrini <[email protected]>
This PR adds models trained from scratch on ImageNet & ecoset.