[Patch] Enhance _hardcode_patch for delivery
#1399
Merged
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
HAOCHENYE
force-pushed
the
yehc/hardcode-fucking-plus
branch
2 times, most recently
from
fff38f5 to
0a0c1e7
Compare
hhaAndroid
approved these changes
Dec 29, 2025
1. mask the loss of `<think></think>\n\n` rather than bare `<think></think>` 2. mask the loss of `<think>` 3. User can visualize the tokenized data by (Just for `Qwen3VLTokenizeFunction`) ```console python .dev_scripts/visualize/tokenize_fn_file.py <tokenizer-path> <jsonl-path> ```
HAOCHENYE
force-pushed
the
yehc/hardcode-fucking-plus
branch
from
0a0c1e7 to
56e4fc0
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Modifications
<think></think>\n\nrather than bare<think></think><think>How to use
Enable
Enable the hardcode patch by:
export XTUNER_SKIP_EMPTY_THINK=1The following log will be displayed in the console output (stdout/stderr), not in the log file:
"
SkipEmptyThinkis enabled to skip empty sequences in labels."Visualize the tokenized data by:
python .dev_scripts/visualize/tokenize_fn_file.py <tokenizer-path> <jsonl-path>Only the red part will calculate the loss.