[megatron] feat: Support LoRA training with FP16 using Megatron-Bridge. (verl-project#4648)

### What does this PR do?

> Support LoRA training with FP16 using Megatron-Bridge.

### Checklist Before Starting

- [ ] Search for similar PRs. Paste at least one query link here: ...
- [ ] Format the PR title as `[{modules}] {type}: {description}` (This
will be checked by the CI)
- `{modules}` include `fsdp`, `megatron`, `sglang`, `vllm`, `rollout`,
`trainer`, `ci`, `training_utils`, `recipe`, `hardware`, `deployment`,
`ray`, `worker`, `single_controller`, `misc`, `perf`, `model`, `algo`,
`env`, `tool`, `ckpt`, `doc`, `data`, `cfg`, `reward`
- If this PR involves multiple modules, separate them with `,` like
`[megatron, fsdp, doc]`
  - `{type}` is in `feat`, `fix`, `refactor`, `chore`, `test`
- If this PR breaks any API (CLI arguments, config, function signature,
etc.), add `[BREAKING]` to the beginning of the title.
  - Example: `[BREAKING][fsdp, megatron] feat: dynamic batching`

### Test

> For changes that can not be tested by CI (e.g., algorithm
implementation, new model support), validate by experiment(s) and show
results like training curve plots, evaluation results, etc.
<img width="1351" height="492" alt="image"
src="https://github.com/user-attachments/assets/cfd55b68-00b9-4f36-9a0a-fc608073e216"
/>

### API and Usage Example

> Demonstrate how the API changes if any, and provide usage example(s)
if possible.

```python
# Add code snippet or script demonstrating how to use this
```

### Design & Code Changes

> Demonstrate the high-level design if this PR is complex, and list the
specific changes.

### Checklist Before Submitting

> [!IMPORTANT]
> Please check all the following items before requesting a review,
otherwise the reviewer might deprioritize this PR for review.

- [ ] Read the [Contribute
Guide](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md).
- [ ] Apply [pre-commit
checks](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md#code-linting-and-formatting):
`pre-commit install && pre-commit run --all-files --show-diff-on-failure
--color=always`
- [ ] Add / Update [the
documentation](https://github.com/volcengine/verl/tree/main/docs).
- [ ] Add unit or end-to-end test(s) to [the CI
workflow](https://github.com/volcengine/verl/tree/main/.github/workflows)
to cover all the code. If not feasible, explain why: ...
- [ ] Once your PR is ready for CI, send a message in [the `ci-request`
channel](https://verl-project.slack.com/archives/C091TCESWB1) in [the
`verl` Slack
workspace](https://join.slack.com/t/verl-project/shared_invite/zt-3855yhg8g-CTkqXu~hKojPCmo7k_yXTQ).
(If not accessible, please try [the Feishu group
(飞书群)](https://applink.larkoffice.com/client/chat/chatter/add_by_link?link_token=772jd4f1-cd91-441e-a820-498c6614126a).)

Author: xichengpro

examples/grpo_trainer/run_qwen3moe-30b_megatron_lora_fp16.sh

@@ -0,0 +1,145 @@
+#!/usr/bin/env bash
+set -xeuo pipefail
+pwd=`pwd`
+
+rollout_mode="async"
+rollout_name="vllm" # sglang or vllm
+if [ "$rollout_mode" = "async" ]; then
+    export VLLM_USE_V1=1
+    return_raw_chat="True"
+fi
+
+TP=${TP:-2}
+PP=${PP:-2}
+CP=${CP:-2}
+EP=${EP:-4}
+ETP=${ETP:-1}
+
+ALL_OFFLOAD=${ALL_OFFLOAD:-True}
+
+optimizer_offload_fraction=1.
+
+dtype="float16" # ["bfloat16", "float16"]
+rollout_name="vllm"
+project_name='verl_grpo_example_gsm8k_math_fp16'
+exp_name='qwen3_30b_a3b_megatron_lora'
+adv_estimator=grpo
+
+# Paths
+MODEL_PATH=$HOME/Qwen/Qwen3-30B-A3B-Instruct-2507
+CKPTS_DIR=${pwd}/ckpt/${exp_name}
+
+gsm8k_train_path=$HOME/data/gsm8k/train.parquet
+gsm8k_test_path=$HOME/data/gsm8k/test.parquet
+
+########################### Parameter Arrays ###########################
+
+DATA=(
+    data.train_files=${gsm8k_train_path}
+    data.val_files=${gsm8k_test_path}
+    data.train_batch_size=128
+    data.max_prompt_length=1024
+    data.max_response_length=1024
+    data.truncation='error'
+    data.filter_overlong_prompts=True
+    data.shuffle=False
+    data.return_raw_chat=$return_raw_chat
+    data.filter_overlong_prompts_workers=128
+)
+
+MODEL=(
+    actor_rollout_ref.model.path=${MODEL_PATH}
+    actor_rollout_ref.model.lora.rank=16
+    actor_rollout_ref.model.lora.alpha=32
+    actor_rollout_ref.model.lora.dtype=${dtype}
+    actor_rollout_ref.model.use_fused_kernels=True
+)
+
+ACTOR=(
+    actor_rollout_ref.actor.optim.lr=3e-6
+    actor_rollout_ref.actor.ppo_mini_batch_size=16
+    actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=2
+    actor_rollout_ref.actor.megatron.use_mbridge=True
+    actor_rollout_ref.actor.megatron.vanilla_mbridge=False
+    actor_rollout_ref.actor.use_dynamic_bsz=True
+    actor_rollout_ref.actor.use_kl_loss=True
+    actor_rollout_ref.actor.kl_loss_coef=0.001
+    actor_rollout_ref.actor.kl_loss_type=low_var_kl
+    actor_rollout_ref.actor.entropy_coeff=0
+    actor_rollout_ref.actor.megatron.tensor_model_parallel_size=${TP}
+    actor_rollout_ref.actor.megatron.pipeline_model_parallel_size=${PP}
+    actor_rollout_ref.actor.megatron.expert_model_parallel_size=${EP}
+    actor_rollout_ref.actor.megatron.context_parallel_size=${CP}
+    actor_rollout_ref.actor.megatron.expert_tensor_parallel_size=${ETP}
+    actor_rollout_ref.actor.megatron.param_offload=${ALL_OFFLOAD}
+    actor_rollout_ref.actor.megatron.optimizer_offload=${ALL_OFFLOAD}
+    actor_rollout_ref.actor.megatron.grad_offload=${ALL_OFFLOAD}
+    actor_rollout_ref.actor.megatron.dtype=${dtype}
+    +actor_rollout_ref.actor.megatron.override_transformer_config.recompute_method=uniform
+    +actor_rollout_ref.actor.megatron.override_transformer_config.recompute_granularity=full
+    +actor_rollout_ref.actor.megatron.override_transformer_config.recompute_num_layers=1
+    +actor_rollout_ref.actor.megatron.override_transformer_config.apply_rope_fusion=True
+    +actor_rollout_ref.actor.megatron.override_ddp_config.grad_reduce_in_fp32=True
+    +actor_rollout_ref.actor.optim.override_optimizer_config.optimizer_offload_fraction=${optimizer_offload_fraction}
+    +actor_rollout_ref.actor.optim.override_optimizer_config.overlap_cpu_optimizer_d2h_h2d=True
+    +actor_rollout_ref.actor.optim.override_optimizer_config.use_precision_aware_optimizer=True
+    +actor_rollout_ref.actor.optim.override_optimizer_config.optimizer_cpu_offload=${ALL_OFFLOAD}
+)
+
+ROLLOUT=(
+    actor_rollout_ref.rollout.tensor_model_parallel_size=8
+    actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=4
+    actor_rollout_ref.rollout.log_prob_use_dynamic_bsz=True
+    actor_rollout_ref.rollout.name=${rollout_name}
+    actor_rollout_ref.rollout.gpu_memory_utilization=0.5
+    actor_rollout_ref.rollout.enforce_eager=True
+    actor_rollout_ref.rollout.free_cache_engine=True
+    actor_rollout_ref.rollout.n=4
+    actor_rollout_ref.rollout.dtype=${dtype}
+)
+
+REF=(
+    actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=4
+    actor_rollout_ref.ref.log_prob_use_dynamic_bsz=True
+    actor_rollout_ref.ref.megatron.dtype=${dtype}
+    actor_rollout_ref.ref.megatron.tensor_model_parallel_size=${TP}
+    actor_rollout_ref.ref.megatron.pipeline_model_parallel_size=${PP}
+    actor_rollout_ref.ref.megatron.expert_model_parallel_size=${EP}
+    actor_rollout_ref.ref.megatron.context_parallel_size=${CP}
+    actor_rollout_ref.ref.megatron.expert_tensor_parallel_size=${ETP}
+    actor_rollout_ref.ref.megatron.param_offload=${ALL_OFFLOAD}
+)
+
+ALGORITHM=(
+    algorithm.adv_estimator=${adv_estimator}
+)
+
+TRAINER=(
+    trainer.critic_warmup=0
+    trainer.logger='["console","wandb"]'
+    trainer.project_name=${project_name}
+    trainer.experiment_name=${exp_name}
+    trainer.n_gpus_per_node=8
+    trainer.nnodes=1
+    trainer.save_freq=20
+    trainer.test_freq=5
+    trainer.total_epochs=15
+    trainer.val_before_train=False
+    trainer.max_actor_ckpt_to_keep=1
+    trainer.default_local_dir="${CKPTS_DIR}"
+    trainer.log_val_generations=10
+)
+
+########################### Launch ###########################
+
+python3 -m verl.trainer.main_ppo \
+    --config-path=config \
+    --config-name='ppo_megatron_trainer.yaml' \
+    "${DATA[@]}" \
+    "${ALGORITHM[@]}" \
+    "${MODEL[@]}" \
+    "${ROLLOUT[@]}" \
+    "${ACTOR[@]}" \
+    "${REF[@]}" \
+    "${TRAINER[@]}" \
+    2>&1 | tee ${pwd}/log/${exp_name}_$(date +'%Y%m%d_%H%M%S').log

verl/utils/megatron_utils.py

@@ -269,6 +269,8 @@ def peft_pre_wrap_hook(model):
             model = provider.provide_distributed_model(
                 wrap_with_ddp=wrap_config.wrap_with_ddp,
                 ddp_config=ddp_config,
+                fp16=provider.fp16,
+                bf16=provider.bf16,
             )
 
             # Extract TransformerConfig from the created model

verl/workers/engine/megatron/transformer_impl.py

@@ -152,6 +152,10 @@ def _build_tf_config(self):
             # In case of invalid overrides, we need to make sure some critical params are set correctly
             provider.params_dtype = self.param_dtype
 
+            # Ensure dtype settings propagate to Megatron-Bridge/TE
+            provider.fp16 = self.param_dtype == torch.float16
+            provider.bf16 = self.param_dtype == torch.bfloat16
+            
             # Pass distributed info
             provider.tensor_model_parallel_size = self.engine_config.tensor_model_parallel_size
             provider.pipeline_model_parallel_size = self.engine_config.pipeline_model_parallel_size

verl/workers/megatron_workers.py

@@ -199,6 +199,10 @@ def _init_hf_config_and_tf_config(
                 # In case of invalid overrides, we need to make sure some critical params are set correctly
                 provider.params_dtype = dtype
 
+                # Ensure dtype settings propagate to Megatron-Bridge/TE
+                provider.fp16 = fp16
+                provider.bf16 = bf16
+                
                 # Pass distributed info
                 provider.tensor_model_parallel_size = megatron_config.tensor_model_parallel_size
                 provider.pipeline_model_parallel_size = megatron_config.pipeline_model_parallel_size