Skip to content

bump trl to 0.29 #8404

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.

Sign up for GitHub

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

Merged
hjh0119 merged 1 commit into from
Mar 24, 2026
+322 −36
Merged

bump trl to 0.29 #8404

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -145,7 +145,7 @@ Running Environment:
| modelscope | >=1.23 | | |
| peft | >=0.11,<0.19 | | |
| flash_attn | | 2.8.3/3.0.0b1 | |
| trl | >=0.15,<0.29 | 0.28.0 | RLHF |
| trl | >=0.15,<0.30 | 0.28.0 | RLHF |
| deepspeed | >=0.14 | 0.18.8 | Training |
| vllm | >=0.5.1 | 0.11.0/0.17.1 | Inference/Deployment |
| sglang | >=0.4.6 | | Inference/Deployment |
Expand Down
2 changes: 1 addition & 1 deletion README_CN.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -141,7 +141,7 @@ uv pip install -e . --torch-backend=auto
| modelscope | >=1.23 | | |
| peft | >=0.11,<0.19 | | |
| flash_attn | | 2.8.3/3.0.0b1 | |
| trl | >=0.15,<0.29 | 0.28.0 | RLHF |
| trl | >=0.15,<0.30 | 0.28.0 | RLHF |
| deepspeed | >=0.14 | 0.18.8 | 训练 |
| vllm | >=0.5.1 | 0.11.0/0.17.1 | 推理/部署 |
| sglang | >=0.4.6 | | 推理/部署 |
Expand Down
2 changes: 1 addition & 1 deletion docs/source/GetStarted/SWIFT-installation.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -144,7 +144,7 @@ modelscope-registry.us-west-1.cr.aliyuncs.com/modelscope-repo/modelscope:ubuntu2
| modelscope | >=1.23 | | |
| peft | >=0.11,<0.19 | | |
| flash_attn | | 2.8.3/3.0.0b1 | |
| trl | >=0.15,<0.29 | 0.28.0 | RLHF |
| trl | >=0.15,<0.30 | 0.28.0 | RLHF |
| deepspeed | >=0.14 | 0.18.8 | 训练 |
| vllm | >=0.5.1 | 0.11.0/0.17.1 | 推理/部署 |
| sglang | >=0.4.6 | | 推理/部署 |
Expand Down
2 changes: 1 addition & 1 deletion docs/source/Megatron-SWIFT/Quick-start.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,7 @@ modelscope-registry.us-west-1.cr.aliyuncs.com/modelscope-repo/modelscope:ubuntu2
| transformers | >=4.33 | 4.57.6/5.2.0 | |
| modelscope | >=1.23 | | |
| peft | >=0.11,<0.19 | | LoRA |
| trl | >=0.15,<0.29 | | RLHF |
| trl | >=0.15,<0.30 | | RLHF |


## 快速入门案例
Expand Down
2 changes: 1 addition & 1 deletion docs/source_en/GetStarted/SWIFT-installation.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -143,7 +143,7 @@ More images can be found [here](https://modelscope.cn/docs/intro/environment-set
| modelscope | >=1.23 | | |
| peft | >=0.11,<0.19 | | |
| flash_attn | | 2.8.3/3.0.0b1 | |
| trl | >=0.15,<0.29 | 0.28.0 | RLHF |
| trl | >=0.15,<0.30 | 0.28.0 | RLHF |
| deepspeed | >=0.14 | 0.18.8 | Training |
| vllm | >=0.5.1 | 0.11.0/0.17.1 | Inference/Deployment |
| sglang | >=0.4.6 | | Inference/Deployment |
Expand Down
2 changes: 1 addition & 1 deletion docs/source_en/Megatron-SWIFT/Quick-start.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,7 @@ Recommended Operating Environment:
| transformers | >=4.33 | 4.57.6/5.2.0 | |
| modelscope | >=1.23 | | |
| peft | >=0.11,<0.19 | | LoRA |
| trl | >=0.15,<0.29 | | RLHF |
| trl | >=0.15,<0.30 | | RLHF |


## Quick Start Example
Expand Down
2 changes: 1 addition & 1 deletion requirements/framework.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,6 +35,6 @@ tiktoken
tqdm
transformers>=4.33,<5.4.0
transformers_stream_generator
trl>=0.15,<0.29
trl>=0.15,<0.30
uvicorn
zstandard
5 changes: 2 additions & 3 deletions swift/megatron/trainers/dpo_trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,12 +13,11 @@


class DummyDPOTrainer(DPOTrainer):
# For reusing the dpo_loss function in TRL.
# For reusing the dpo_loss function implemented in Swift's DPOTrainer.
def __init__(self, args):
from trl.trainer import FDivergenceConstants
self.accelerator = namedtuple('Accelerator', ['device'])(device=get_current_device())
self.f_alpha_divergence_coef = 1.
self.f_divergence_params = {FDivergenceConstants.ALPHA_DIVERGENCE_COEF_KEY: self.f_alpha_divergence_coef}
self.f_divergence_params = {'alpha_divergence_coef': self.f_alpha_divergence_coef}
self.reference_free = args.reference_free
self.label_smoothing = args.label_smoothing
self.f_divergence_type = args.f_divergence_type
Expand Down
4 changes: 4 additions & 0 deletions swift/rlhf_trainers/arguments.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,10 @@
@dataclass
class DPOConfig(TrainArgumentsMixin, HfDPOConfig):
ld_alpha: Optional[float] = None # compat trl==0.15
# Fields removed in trl 0.29, kept here for backward compatibility
rpo_alpha: Optional[float] = None
ref_adapter_name: Optional[str] = None
reference_free: Optional[bool] = None

def __post_init__(self):
TrainArgumentsMixin.__post_init__(self)
Expand Down
Loading
Loading