https://github.com/volcengine/verl/pull/182

add a assert statement to make sure flash-attn>=2.4.3 where
cross_entropy_loss returns Tuple[losses, z_losses]🤯

This PR is similar to PR https://github.com/volcengine/verl/pull/174 but
fix the critic save error

I move the old PR to this one due to some redundant commit

sry missed this last one, should be it 

cc @vermouth1992

Co-authored-by: Jayson Francis <jaysonfrancis@users.noreply.github.com>

runs always show "crashed" on my wandb, despite finishing successfully.
"Crashed" indicates that wandb did not finish sending the "success"
signal to the server so the server believes the client was terminated
unexpectedly. Furthermore, wandb log is incomplete (last lines missing).

This PR adds a call to `wandb.finish` when the Tracker was destructed
(oftentimes when `trainer.fit` finished) so that signals are sent to the
server and a data sync is performed.

Without this change:
<img width="526" alt="image"
src="https://github.com/user-attachments/assets/869da24e-c5b8-415c-b15a-bb79c49f96ce"
/>

With this change:
<img width="548" alt="image"
src="https://github.com/user-attachments/assets/16f0a40d-ea3b-48ed-93a4-f40ee01cb7c6"
/>

Co-authored-by: Jayson Francis <jaysonfrancis@users.noreply.github.com>

This PR adds documentation for the LigerKernel option in a new
performance tuning section, addressing the comment from
volcengine/verl#173.

Changes:
- Created new performance tuning section in docs
- Documented LigerKernel option for SFT
- Added performance tuning section to documentation index

Related to volcengine/verl#173

---------

Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: HL <linhaibin.eric@gmail.com>

runnning -> running

since 'lighteval/MATH' is no longer available on huggingface.

- As titled

![image](https://github.com/user-attachments/assets/f0bae990-a4ef-49da-aa1e-58894b41db5f)

---------

Co-authored-by: HL <linhaibin.eric@gmail.com>

This is a follow-up to https://github.com/volcengine/verl/issues/151

## Motivation

Currently, in order to add a custom score function you need to fork verl
and update the `_select_rm_score_fn` to define your logic. This makes it
harder to use verl as part of a larger application while staying up to
date with upstream improvements in verl.

It would be convenient to allow end users to directly pass in a reward
function they wish to use, without requiring them to clone/fork verl to
do so.

## Design

In this PR I slightly modify `main_ppo.py` to allow users to import a
new function `run_ppo`. `run_ppo` behaves very similarly to the existing
`main`, with the important addition of a new `compute_score` argument.
This argument, if passed in, is used to compute the score of every
generation. This is the change that allows

The `compute_score` function is similar in shape to the existing
`compute_score` on gsm8k and math. However, I have added a new
`data_source` parameter so that the user can compute the score
differently if desired depending on the task shape.

## Example Usage

This is a sample script showing how you can use the new functionality. I
have tested that this works.

```python
from verl.trainer.main_ppo import run_ppo
from omegaconf import OmegaConf


def custom_compute_score(data_source, solution_str, ground_truth):
    """Dummy compute_score function that reward the model for generations of exactly 20 characters :)
    """
    return abs(len(solution_str) - 20)


config = OmegaConf.load("vendor/verl/verl/trainer/config/ppo_trainer.yaml")

# Update config as needed
config.data.train_files = "path/to/train.parquet"
config.data.val_files = "path/to/test.parquet"
# ...

run_ppo(config, custom_compute_score)
```

## Breaking changes

There are no breaking changes in this PR. It is still possible to call
`python -m verl.trainer.main_ppo ...` as before (although if you want to
pass in a custom compute_score you will need to use the new method
described above).

## Possible future work

It would be great to move to [structured
configs](https://omegaconf.readthedocs.io/en/2.1_branch/structured_config.html)
as well since they'd allow us to have typesafe, autocompletable
configurations from Python. I thought about adding those changes here as
well but they would be much more extensive and I'm not sure whether
there's interest from the project.

## Summary

This PR enables to use Liger Kernel's `_apply_liger_kernel_to_instance`
to init a fsdp worker model.

## Main Changes

1. Adding an option of using
`liger_kernel.transformers.AutoLigerKernelForCausalLM` to load a model
from pretained, instead of the default
`transformers.AutoModelForCausalLM`
2. Added a test case using configuration file
`tests/e2e/run_qwen_gsm8k_model_rm_liger_kernel.sh`

## Related Issue

#96 

## TODO

#97 optimize the memory usage when computing entropy & log_probs

https://github.com/volcengine/verl/blob/6d96fda3d47f057caaa8f494ca7804181903e911/verl/workers/actor/dp_actor.py#L94-L106

---------

Signed-off-by: Hongpeng Guo <hpguo@anyscale.com>

The logits is of shape `(bsz, response_length, vocab_size)`. This PR
doesn't change any code execution, but explicitly show the logits shape
and easier for readers to understand the code.

Signed-off-by: Hongpeng Guo <hpguo@anyscale.com>

Add contribution guide

`token_level_rewards == (token_level_rewards * non_zero_mask)`

- As titled
- Solved: #149 

Waiting for testing from @chujiezheng

---------

Co-authored-by: Chi Zhang <zhangchi.usc1992@bytedance.com>

- Add link to performance tuning

- Previous gradient accumulation value is computed by micro_batch_size,
which is wrong when using dynamic_bsz
- Fix ci script to avoid overlooking this issue
- Change vLLM state log default value to True to disable log.
- We will check the `self.config.actor.ppo_mini_batch_size %
self.config.actor.ppo_micro_batch_size_per_gpu == 0` after normalization
in fsdp_workers instead of in dp_actor and dp_critic.

- As titled

# Add Sequence Parallelism and Padding Removal to SFT Trainer

This PR adds sequence parallelism (SP) and padding removal optimizations
to the SFT trainer, which can help improve training efficiency for large
language models.

## Key Changes

### Core Features
1. **Sequence Parallelism**: Added support for sequence parallelism
through the Ulysses framework
   - Configurable via `ulysses_sequence_parallel_size` parameter
   - Properly handles data distribution across SP ranks
   - Maintains consistent loss computation across distributed setup

2. **Padding Removal**: Added support for efficient handling of
variable-length sequences
   - Enabled via `use_remove_padding` flag (requires SP to be enabled)
   - Uses flash-attention's padding removal utilities
   - Handles proper re-padding and loss computation

3. **Training Improvements**:
   - Added label smoothing support to loss computation
   - Added progress bar with epoch information
   - Added RoPE scaling configuration support
   - Improved error messages for batch size validation

### Testing
- Added comprehensive test suite (`test_trainer.py`) to verify:
- Forward pass consistency between original and SP+rmpad implementations
  - Loss computation correctness across distributed setup
  - Proper handling of micro-batches

### Example Usage
Added example script `examples/sft/gsm8k/run_qwen_05_sp2.sh`
demonstrating how to use the new features with Qwen-2.5B model.

## Implementation Details
- Uses device mesh for proper distributed training setup
- Handles data distribution ensuring same sequences within SP groups but
different across DP groups
- Carefully manages backward pass timing with gradient checkpointing
- Maintains compatibility with existing FSDP features

## Testing Instructions
1. Run the example script with sequence parallelism:
```bash
bash examples/sft/gsm8k/run_qwen_05_sp2.sh <nproc_per_node> <save_path>
```

2. Run the test suite:
```bash tests/sft/run_sft_sp_loss_match.sh```


^^ These are PR description generated by [OpenHands](https://github.com/All-Hands-AI/OpenHands)

---------

Co-authored-by: Jiayi Pan <i@jiayipan.me>
Co-authored-by: openhands <openhands@all-hands.dev>

We set `max_num_batched_tokens` in config `.rollout`, but they weren't
actually being passed to VLLM -- causing potential insufficient use of
GPUs.

This PR:

- properly pass `max_num_batched_tokens` from config to vLLM
- set `disable_log_stats` to False, so vLLM performance information can
be properly displayed (to spot issues)

## Summary

This PR changes all the micro_batch_size to micro_batch_size_per_gpu.

**The Core logic of setting batch size:**
- **All algorithmic metrics** (train batch size, ppo mini batch size):
are global (from the perspective of single-controller), which will be
normalized in each Worker.
- **All performance-related parameters** (micro batch size, max token
length in dynamic batch size) are local parameters, which represent the
data sizes per GPU (i.e., each Worker).

## Main Changes

1. Change the scripts and config and delete the normalization for
micro_bsz
2. Fix CI for SFT