bi 8 d dl Z deeeeef dee fdZy) N completionsreturnc d}| D cg c]
}|d d }}|D cg c]8 }t j ||t j t j z : }}|D cg c] }|rdnd
c}S c c}w c c}w c c}w )a
Reward function that checks if the reasoning process is enclosed within `""` and `""` tags. The
function returns a reward of 1.0 if the format is correct, otherwise 0.0.
Args:
completions (`list[list[dict[str, str]]]`):
List of completions to be evaluated. Each completion must be a list of one message, i.e. a dictionary
containing the key `"content"` with the value being the text of the completion.
**kwargs:
Additional keyword arguments. This function does not use them, but they are required in the function
signature to ensure compatibility with trainers like [`GRPOTrainer`].
Returns:
`list[float]`:
A list of rewards, where each reward is 1.0 if the completion matches the expected format, otherwise 0.0.
Example:
```python
>>> from trl.rewards import think_format_reward
>>> completions = [
... [{"content": "\nThis is my reasoning.\n\nThis is my answer."}],
... [{"content": "\nThis is my reasoning.\nThis is my answer."}],
... ]
>>> think_format_reward(completions)
[1.0, 0.0]
```
z%^(?!.*)(.*?).*$r contentg ?g )rematchDOTALL MULTILINE)r kwargspattern
completioncompletion_contentsr matchesr s U/home/cdr/jupyterlab/.venv/lib/python3.12/site-packages/trl/rewards/format_rewards.pythink_format_rewardr sw : 7GFQR
:a=3RRSfgrxx"))bll*BCgGg/6 7e5Cc ! 77 Sg 7s A/=A4
A9)r listdictstrfloatr r r s4
8T$tCH~*>%? 8dSXk 8r