Develop #71
Merged
Develop #71
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
* clean codes Co-authored-by: Zac Liu <[email protected]>
* fix bert tokenizer issue * updated t5, opt and roberta tokenizers * fixed doc 404 error Signed-off-by: ZhaodongYan1 <[email protected]>
* autoloader for opt * opt-66b inference * Update train.py * Load data from example dir * add readme of multi GPU inference Co-authored-by: Zac Liu <[email protected]>
Signed-off-by: Anhforth <[email protected]>
Signed-off-by: Anhforth <[email protected]>
fix bug multi_gpu_training
Signed-off-by: Anhforth <[email protected]>
Signed-off-by: Anhforth <[email protected]>
remove unused glob
remove data statis
Signed-off-by: zhaohu xing <[email protected]>
Signed-off-by: zhaohu xing <[email protected]>
Signed-off-by: zhaohu xing <[email protected]>
* add vit and examples * vit and examples * Update base_model.py remove unused glob * Update vit.py remove data statis * modify readme.md Signed-off-by: zhaohu xing <[email protected]> * modify readme.md Signed-off-by: zhaohu xing <[email protected]> * delete annotating code Signed-off-by: zhaohu xing <[email protected]> Co-authored-by: Zac Liu <[email protected]>
merge flagi into internal develop
Signed-off-by: zhaohu xing <[email protected]>
add env_trainer for batch script running
fix typo in flagai introduction.
marscrazy
reviewed
Jul 21, 2022
|
|
||
| model = loader.get_model() | ||
|
|
||
| model.load_state_dict(torch.load(os.path.join(model_save_dir, "38000", "pytorch_model.bin"), map_location=device)["module"]) |
There was a problem hiding this comment.
change to from_pretrain to load the model
marscrazy
reviewed
Jul 21, 2022
| x = self.patch_embed(x) | ||
| if self.cls_token is not None: | ||
| x = torch.cat((self.cls_token.expand(x.shape[0], -1, -1), x), dim=1) | ||
| x = self.pos_drop(x + self.pos_embed) |
There was a problem hiding this comment.
add checkpoint_activations
Signed-off-by: zhaohu xing <[email protected]>
Signed-off-by: zhaohu xing <[email protected]>
…vations add checkpointing activations for VIT when env_type = pytorch/deepspeed
marscrazy
approved these changes
Jul 21, 2022
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.
No description provided.