mirror of
https://github.com/volcengine/verl.git
synced 2025-10-20 13:43:50 +08:00
[model] feat: add qwen3 grpo 8b/32b script on ASCEND NPU (#3310)
### What does this PR do? add examples/grpo_trainer/run_qwen3_32b_npu.sh <img width="1014" height="1111" alt="image" src="https://github.com/user-attachments/assets/8cd59fc2-5f6a-419e-87ac-bf35a71856fb" /> add examples/grpo_trainer/run_qwen3_8b_npu.sh <img width="844" height="930" alt="image" src="https://github.com/user-attachments/assets/5c23c7a4-8729-4007-8828-027a8cda4779" /> > Add **concise** overview of what this PR aims to achieve or accomplish. Reference related GitHub issues and PRs that help with the review. ### Checklist Before Starting - [x] Search for similar PRs. Paste at least one query link here: ... - [x] 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` - 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. ### 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. - [x] Read the [Contribute Guide](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md). - [x] 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` - [x] 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: ... > already support in https://github.com/volcengine/verl/pull/3300 - [ ] 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).) --------- Signed-off-by: ZLiao <a627465478@gmail.com> Co-authored-by: ZLiao <a627465478@gmail.com>
This commit is contained in:
@ -179,6 +179,10 @@ vllm & vllm-ascend
|
||||
+-----------+-------------------------+-------------+-------------------+-------------------+-------------------+--------------------------+
|
||||
| GRPO | Qwen2.5-VL-32B-instruct | 0.79% | 0.568 | FSDP | vllm-ascend | Atlas 200T A2 Box16 |
|
||||
+-----------+-------------------------+-------------+-------------------+-------------------+-------------------+--------------------------+
|
||||
| GRPO | Qwen3-8B | 1.55% | 1.012 | FSDP | vllm-ascend | Atlas 200T A2 Box16 |
|
||||
+-----------+-------------------------+-------------+-------------------+-------------------+-------------------+--------------------------+
|
||||
| GRPO | Qwen3-32B | 0.64% | 0.696 | FSDP | vllm-ascend | Atlas 200T A2 Box16 |
|
||||
+-----------+-------------------------+-------------+-------------------+-------------------+-------------------+--------------------------+
|
||||
| DAPO | Qwen2.5-7B-instruct | 3.83% | pending | FSDP | vllm-ascend | Atlas 200T A2 Box16 |
|
||||
+-----------+-------------------------+-------------+-------------------+-------------------+-------------------+--------------------------+
|
||||
| DAPO | Qwen2.5-32B | 3.42% | pending | FSDP | vllm-ascend | Atlas 200T A2 Box16 |
|
||||
|
59
examples/grpo_trainer/run_qwen3-32b_npu.sh
Normal file
59
examples/grpo_trainer/run_qwen3-32b_npu.sh
Normal file
@ -0,0 +1,59 @@
|
||||
set -x
|
||||
|
||||
project_name='GRPO-Qwen3'
|
||||
exp_name='GRPO-Qwen3-32b-npu'
|
||||
gen_tp=4
|
||||
RAY_DATA_HOME=${RAY_DATA_HOME:-"${HOME}/verl"}
|
||||
MODEL_PATH=${MODEL_PATH:-"${RAY_DATA_HOME}/models/Qwen3-32B"}
|
||||
TRAIN_FILE=${TRAIN_FILE:-"${RAY_DATA_HOME}/data/gsm8k/train.parquet"}
|
||||
TEST_FILE=${TEST_FILE:-"${RAY_DATA_HOME}/data/gsm8k/test.parquet"}
|
||||
|
||||
python3 -m verl.trainer.main_ppo \
|
||||
algorithm.adv_estimator=grpo \
|
||||
data.train_files="${TRAIN_FILE}" \
|
||||
data.val_files="${TEST_FILE}" \
|
||||
data.train_batch_size=1024 \
|
||||
data.max_prompt_length=2048 \
|
||||
data.max_response_length=2048 \
|
||||
data.filter_overlong_prompts=True \
|
||||
data.truncation='error' \
|
||||
data.shuffle=False \
|
||||
actor_rollout_ref.model.path=${MODEL_PATH} \
|
||||
actor_rollout_ref.actor.optim.lr=1e-6 \
|
||||
actor_rollout_ref.model.use_remove_padding=True \
|
||||
actor_rollout_ref.actor.ulysses_sequence_parallel_size=4 \
|
||||
+actor_rollout_ref.actor.fsdp_config.mixed_precision.param_dtype=bf16 \
|
||||
+actor_rollout_ref.actor.fsdp_config.mixed_precision.reduce_dtype=bf16 \
|
||||
+actor_rollout_ref.actor.fsdp_config.mixed_precision.buffer_dtype=fp32 \
|
||||
actor_rollout_ref.actor.ppo_mini_batch_size=64 \
|
||||
actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=8 \
|
||||
actor_rollout_ref.actor.use_kl_loss=True \
|
||||
actor_rollout_ref.actor.entropy_coeff=0 \
|
||||
actor_rollout_ref.actor.kl_loss_coef=0.001 \
|
||||
actor_rollout_ref.actor.kl_loss_type=low_var_kl \
|
||||
actor_rollout_ref.model.enable_gradient_checkpointing=True \
|
||||
actor_rollout_ref.actor.fsdp_config.param_offload=True \
|
||||
actor_rollout_ref.actor.fsdp_config.optimizer_offload=False \
|
||||
actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=8 \
|
||||
actor_rollout_ref.rollout.tensor_model_parallel_size=${gen_tp} \
|
||||
actor_rollout_ref.rollout.name=vllm \
|
||||
actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \
|
||||
actor_rollout_ref.rollout.n=4 \
|
||||
actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=8 \
|
||||
actor_rollout_ref.ref.fsdp_config.param_offload=True \
|
||||
actor_rollout_ref.actor.use_torch_compile=False \
|
||||
actor_rollout_ref.ref.use_torch_compile=False \
|
||||
actor_rollout_ref.rollout.enable_chunked_prefill=True \
|
||||
actor_rollout_ref.rollout.max_num_batched_tokens=32768 \
|
||||
algorithm.use_kl_in_reward=False \
|
||||
trainer.critic_warmup=0 \
|
||||
trainer.logger=['console','tensorboard'] \
|
||||
trainer.project_name="${project_name}" \
|
||||
trainer.experiment_name="${exp_name}" \
|
||||
trainer.n_gpus_per_node=8 \
|
||||
trainer.nnodes=4 \
|
||||
trainer.resume_from_path=checkpoints/ \
|
||||
trainer.save_freq=500 \
|
||||
trainer.test_freq=50 \
|
||||
trainer.total_epochs=50 \
|
||||
trainer.device=npu $@
|
59
examples/grpo_trainer/run_qwen3-8b_npu.sh
Normal file
59
examples/grpo_trainer/run_qwen3-8b_npu.sh
Normal file
@ -0,0 +1,59 @@
|
||||
set -x
|
||||
|
||||
project_name='GRPO-Qwen3'
|
||||
exp_name='GRPO-Qwen3-8B-npu'
|
||||
gen_tp=2
|
||||
RAY_DATA_HOME=${RAY_DATA_HOME:-"${HOME}/verl"}
|
||||
MODEL_PATH=${MODEL_PATH:-"${RAY_DATA_HOME}/models/Qwen3-8B"}
|
||||
CKPTS_DIR=${CKPTS_DIR:-"${RAY_DATA_HOME}/ckpts/${project_name}/${exp_name}"}
|
||||
TRAIN_FILE=${TRAIN_FILE:-"${RAY_DATA_HOME}/data/dapo-math-17k.parquet"}
|
||||
TEST_FILE=${TEST_FILE:-"${RAY_DATA_HOME}/data/aime-2024.parquet"}
|
||||
|
||||
python3 -m verl.trainer.main_ppo \
|
||||
algorithm.adv_estimator=grpo \
|
||||
data.train_files="${TRAIN_FILE}" \
|
||||
data.val_files="${TEST_FILE}" \
|
||||
data.train_batch_size=256 \
|
||||
data.max_prompt_length=512 \
|
||||
data.max_response_length=1024 \
|
||||
data.filter_overlong_prompts=True \
|
||||
data.truncation='error' \
|
||||
actor_rollout_ref.model.path=${MODEL_PATH} \
|
||||
actor_rollout_ref.actor.optim.lr=1e-6 \
|
||||
actor_rollout_ref.model.use_remove_padding=True \
|
||||
actor_rollout_ref.actor.ppo_mini_batch_size=64 \
|
||||
actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=10 \
|
||||
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.use_torch_compile=False \
|
||||
actor_rollout_ref.ref.use_torch_compile=False \
|
||||
actor_rollout_ref.model.enable_gradient_checkpointing=True \
|
||||
actor_rollout_ref.actor.fsdp_config.param_offload=False \
|
||||
actor_rollout_ref.actor.fsdp_config.optimizer_offload=False \
|
||||
actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=32 \
|
||||
actor_rollout_ref.rollout.tensor_model_parallel_size=${gen_tp} \
|
||||
actor_rollout_ref.rollout.name=vllm \
|
||||
actor_rollout_ref.rollout.gpu_memory_utilization=0.6 \
|
||||
actor_rollout_ref.rollout.n=5 \
|
||||
actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=32 \
|
||||
actor_rollout_ref.ref.fsdp_config.param_offload=True \
|
||||
algorithm.use_kl_in_reward=False \
|
||||
trainer.critic_warmup=0 \
|
||||
trainer.logger='["console","wandb"]' \
|
||||
trainer.project_name="${project_name}" \
|
||||
trainer.experiment_name="${exp_name}" \
|
||||
trainer.n_gpus_per_node=8 \
|
||||
trainer.nnodes=1 \
|
||||
trainer.default_local_dir=${CKPTS_DIR} \
|
||||
trainer.device=npu \
|
||||
trainer.resume_mode=auto \
|
||||
actor_rollout_ref.actor.fsdp_config.forward_prefetch=True \
|
||||
actor_rollout_ref.ref.fsdp_config.forward_prefetch=True \
|
||||
++actor_rollout_ref.actor.entropy_from_logits_with_chunking=True \
|
||||
++actor_rollout_ref.ref.entropy_from_logits_with_chunking=True \
|
||||
trainer.val_before_train=True \
|
||||
trainer.save_freq=5 \
|
||||
trainer.test_freq=5 \
|
||||
trainer.total_epochs=15
|
Reference in New Issue
Block a user