#!/usr/bin/env bash
#
# Copyright 2024 PKU-Alignment Team. All Rights Reserved.
#
# 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.
# ==============================================================================


# Initialize variables
# For wandb online logging
export WANDB_API_KEY="547f38af44135ca76a4f4eed9c8d135532da4960"
export NCCL_P2P_LEVEL=NVL
# Source the setup script
source ./setup.sh

ACTOR_MODEL_NAME_OR_PATH="/aifs4su/yaodong/models/Qwen2-VL-7B-Instruct"
CRITIC_MODEL_NAME_OR_PATH="/aifs4su/yaodong/projects/hantao/dev_cham/align-anything/outputs/rm_tiv2t_10k_baseline"
REWARD_MODEL_NAME_OR_PATH="/aifs4su/yaodong/projects/hantao/dev_cham/align-anything/outputs/rm_tiv2t_10k_baseline"
TRAIN_DATASETS="/aifs4su/yaodong/datasets/aaa_dataset/TV2T-preference/extracted"
PTX_DATASETS="/aifs4su/yaodong/datasets/ShareGPT4Video/extracted"
OUTPUT_DIR="../outputs/ppo_qwen2vl_10k_baseline"

# Source the setup script
source ./setup.sh

# Execute deepspeed command
deepspeed \
  --hostfile ./host10 \
  --master_port ${MASTER_PORT} \
  --module align_anything.trainers.tiv_to_t.ppo \
  --actor_model_name_or_path ${ACTOR_MODEL_NAME_OR_PATH} \
  --reward_model_name_or_path ${REWARD_MODEL_NAME_OR_PATH} \
  --reward_critic_model_name_or_path ${CRITIC_MODEL_NAME_OR_PATH} \
  --train_datasets ${TRAIN_DATASETS} \
  --train_template NExTQA_preference \
  --train_data_files extracted_preference_10k_washed.json \
  --train_split 'train' \
  --ptx_datasets ${PTX_DATASETS} \
  --ptx_data_files extracted_panda.json \
  --ptx_template NExTQA \
  --ptx_split 'train' \
  --output_dir ${OUTPUT_DIR} \
  --ddp_timeout 18000000000 \
  --save_interval 300