GDPO/verl-GDPO/train_gdpo.sh at main · NVlabs/GDPO · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
# SPDX-FileCopyrightText: Copyright (c) <year> NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.


export CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7
export N_GPUS=8
export ROLLOUT_TP_SIZE=1
export VLLM_ATTENTION_BACKEND=XFORMERS

# All the env variables below are set to 0 by default
export WITHLENGTH=0
export REFINEDREWARD=0
export COARSEREWARD=0
export STRICTMATCH=0
export CORRECTMAX1=0
export MAX1STEP30MAX3=0
export SCHEDULEREWARD=0
export SCHEDULELENGTH=0

export DATA_DIR="verl-GDPO/dataset/rlla_4k"
export BASE_MODEL="Qwen/Qwen2.5-1.5B-Instruct"
export EXPERIMENT_NAME="qwen2.5-1.5B-GDPO-tool"
export CKPT_DIR="verl-GDPO/results/gdpo"

export RAY_USAGE_STATS_ENABLED=0
export RAY_DISABLE_DOCKER_CPU_WARNING=1


python3 -u -m verl.trainer.main_ppo \
    algorithm.adv_estimator=gdpo \
    data.train_files=$DATA_DIR/train.parquet \
    data.val_files=$DATA_DIR/test.parquet \
    data.train_batch_size=512 \
    data.val_batch_size=128 \
    data.max_prompt_length=2048 \
    data.max_response_length=1024 \
    actor_rollout_ref.model.path=$BASE_MODEL \
    actor_rollout_ref.actor.optim.lr=1e-6 \
    actor_rollout_ref.model.use_remove_padding=True \
    actor_rollout_ref.actor.ppo_mini_batch_size=128 \
    actor_rollout_ref.actor.use_dynamic_bsz=True \
    actor_rollout_ref.actor.use_kl_loss=False \
    actor_rollout_ref.actor.kl_loss_coef=0.001 \
    actor_rollout_ref.actor.kl_loss_type=low_var_kl \
    actor_rollout_ref.model.enable_gradient_checkpointing=True \
    actor_rollout_ref.actor.fsdp_config.param_offload=False \
    actor_rollout_ref.actor.fsdp_config.grad_offload=False \
    actor_rollout_ref.actor.fsdp_config.optimizer_offload=False \
    actor_rollout_ref.rollout.tensor_model_parallel_size=$ROLLOUT_TP_SIZE \
    actor_rollout_ref.rollout.name=vllm \
    actor_rollout_ref.rollout.gpu_memory_utilization=0.6 \
    actor_rollout_ref.rollout.n=4 \
    actor_rollout_ref.ref.fsdp_config.param_offload=True \
    algorithm.kl_ctrl.kl_coef=0.001 \
    trainer.critic_warmup=0 \
    trainer.logger=['console','wandb'] \
    trainer.project_name=Var_inspect \
    trainer.experiment_name=$EXPERIMENT_NAME \
    trainer.n_gpus_per_node=$N_GPUS \
    trainer.nnodes=1 \
    trainer.save_freq=5 \
    trainer.test_freq=10 \
    trainer.default_local_dir=$CKPT_DIR \
    trainer.total_epochs=15