Support incremental state for E2E ASR experiments #173
Conversation
This file contains hidden or 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
change weight-c-init from Optional[float] = None to float = 1.0
jinpoon
reviewed
Mar 23, 2021
| normalize_after: bool, optional | ||
| if True, apply post layer normalization; otherwise apply pre layer normalization | ||
| (default=True). | ||
|
|
There was a problem hiding this comment.
Would you mind adding the doc string of the right_window also? Just wanted to make sure the legal range of this parameter.
experiments/e2e_asr/extension.py
Outdated
| state["saved_query"] = query_not_ready | ||
| state["saved_key"] = k | ||
| state["saved_value"] = v | ||
| incremental_state["attn_state"] = state |
There was a problem hiding this comment.
When there are no ready queries, would it be more efficient to just return the empty tensor here?
experiments/e2e_asr/extension.py
Outdated
| input: Tensor, | ||
| mask_pad: Optional[Tensor] = None, | ||
| attn_mask: Optional[Tensor] = None, | ||
| incremental_state: Optional[Dict[str, Dict[str, Optional[Tensor]]]] = None, |
There was a problem hiding this comment.
Can you change data structure of incremental_state to Optional[Dict[str, Optional[Tensor]]? Seems the SRUCell does use the incremental_state and I should write some glue codes inheriting from the fair_seq's incremental class, and assign & maintain a layer specific hash as the key to this dictionary, in stead of attn_state
experiments/e2e_asr/extension.py
Outdated
| c0: Optional[Tensor] = None, | ||
| mask_pad: Optional[Tensor] = None, | ||
| attn_mask: Optional[Tensor] = None, | ||
| incremental_state: Optional[Dict[str, Dict[str, Optional[Tensor]]]] = None, |
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.