├── LICENSE ├── Notice.txt ├── README.md ├── assets ├── llava1.5.png └── our_llava.png ├── docker ├── Dockerfile.megatron ├── Dockerfile.ngc.vllm ├── Dockerfile.rocm └── Dockerfile.vemlp.vllm.te ├── docs ├── Makefile ├── README.md ├── README_vllm0.7.md ├── _static │ └── logo.png ├── advance │ ├── dpo_extension.rst │ ├── fsdp_extension.rst │ ├── megatron_extension.rst │ └── placement.rst ├── amd_tutorial │ ├── amd_build_dockerfile.md │ └── amd_existing_docker.md ├── conf.py ├── data.rst ├── examples │ ├── config.rst │ ├── gsm8k_example.rst │ └── ppo_code_architecture.rst ├── experiment │ └── ppo.rst ├── faq │ └── faq.rst ├── hybrid_flow.rst ├── index.rst ├── perf │ └── perf_tuning.rst ├── preparation │ ├── prepare_data.rst │ └── reward_function.rst ├── requirements-docs.txt ├── start │ ├── install.rst │ └── quickstart.rst └── workers │ ├── fsdp_workers.rst │ ├── megatron_workers.rst │ └── ray_trainer.rst ├── examples ├── data_preprocess │ ├── full_hh_rlhf.py │ ├── geo3k.py │ ├── gsm8k.py │ ├── hellaswag.py │ └── math_dataset.py ├── generation │ └── run_deepseek_v2_lite_math.sh ├── grpo_trainer │ ├── run_deepseek7b_llm.sh │ ├── run_deepseek7b_llm_seq_balance.sh │ ├── run_llava1_5-7b.sh │ ├── run_qwen2-7b.sh │ ├── run_qwen2-7b_seq_balance.sh │ └── run_qwen2_5_vl-7b.sh ├── ppo_trainer │ ├── run_deepseek7b_llm.sh │ ├── run_deepseek7b_llm_modelscope.sh │ ├── run_deepseek7b_llm_sp2.sh │ ├── run_deepseek_full_hh_rlhf.sh │ ├── run_deepseek_math_gsm8k_megatron.sh │ ├── run_deepseek_megatron.sh │ ├── run_gemma.sh │ ├── run_qwen2-7b.sh │ ├── run_qwen2-7b_math_gsm8k_megatron.sh │ ├── run_qwen2-7b_rm.sh │ ├── run_qwen2-7b_rm_seq_balance.sh │ ├── run_qwen2-7b_seq_balance.sh │ ├── run_qwen2.5-32b.sh │ └── verl_getting_started.ipynb ├── ray │ └── tutorial.ipynb ├── remax_trainer │ ├── run_qwen2.5-3b_seq_balance.sh │ └── run_qwen2.5-7b_seq_balance.sh ├── rloo_trainer │ └── run_qwen2-7b.sh ├── sft │ └── gsm8k │ │ ├── run_deepseek_6b7.sh │ │ ├── run_gemma_2b.sh │ │ ├── run_gemma_7b.sh │ │ ├── run_qwen_05_peft.sh │ │ ├── run_qwen_05_sp2.sh │ │ └── run_qwen_05_sp2_liger.sh ├── slurm │ └── ray_on_slurm.slurm └── split_placement │ ├── README.md │ ├── config │ └── ppo_trainer_split.yaml │ ├── main_ppo_split.py │ ├── run_deepseek7b_llm.sh │ └── split_monkey_patch.py ├── patches └── megatron_v4.patch ├── pyproject.toml ├── requirements.txt ├── scripts ├── format.sh └── model_merger.py ├── setup.py ├── tests ├── __init__.py ├── checkpoint │ └── test_fsdp_ckpt.py ├── distro │ └── requirements.py ├── e2e │ ├── __init__.py │ ├── arithmetic_sequence │ │ ├── data │ │ │ ├── create_dataset.py │ │ │ ├── test.parquet │ │ │ └── train.parquet │ │ ├── model │ │ │ ├── config.json │ │ │ ├── create_model_tokenizer.py │ │ │ ├── generation_config.json │ │ │ ├── model.safetensors │ │ │ └── tokenizer_config.json │ │ └── rl │ │ │ ├── README.md │ │ │ └── main_trainer.py │ ├── check_results.py │ ├── envs │ │ ├── __init__.py │ │ └── digit_completion │ │ │ ├── __init__.py │ │ │ ├── task.py │ │ │ └── tokenizer.py │ ├── run_deepseek_megatron.sh │ ├── run_qwen2vl_geo3k_function_rm.sh │ ├── run_qwen_gsm8k_function_rm.sh │ ├── run_qwen_gsm8k_function_rm_grpo.sh │ ├── run_qwen_gsm8k_function_rm_no_rmpad.sh │ ├── run_qwen_gsm8k_function_rm_remax.sh │ ├── run_qwen_gsm8k_model_rm.sh │ ├── run_qwen_gsm8k_model_rm_liger_kernel.sh │ ├── run_qwen_gsm8k_model_rm_no_rmpad.sh │ ├── run_qwen_gsm8k_model_rm_seq_balance.sh │ ├── run_qwen_gsm8k_model_rm_ulysses.sh │ ├── run_qwen_megatron.sh │ ├── run_ray_trainer.sh │ ├── run_ray_trainer_fire_sampling.sh │ └── run_ray_trainer_rmpad.sh ├── generation │ └── run_gen_qwen05.sh ├── gpu_utility │ ├── test_memory_buffers.py │ ├── test_ops.py │ └── test_torch_functional.py ├── kill_github_tests.sh ├── model │ ├── test_transformer.py │ └── test_transformers_ulysses.py ├── ray │ ├── check_worker_alive │ │ └── main.py │ ├── detached_worker │ │ ├── README.md │ │ ├── client.py │ │ ├── run.sh │ │ └── server.py │ ├── test_check_worker_alive.py │ ├── test_colocated_workers.py │ ├── test_data_transfer.py │ ├── test_driverfunc_to_worker.py │ ├── test_high_level_scheduling_api.py │ ├── test_ray_local_envs.py │ ├── test_rvdz.py │ ├── test_worker_group_basics.py │ └── test_worker_group_torch.py ├── rollout │ ├── run_fsdp_vllm.py │ ├── test_vllm_hf_loader.py │ └── test_vllm_spmd.py ├── sandbox │ └── test_sandbox.py ├── sanity │ ├── check_license.py │ └── test_import.py ├── sft │ ├── run_sft.sh │ ├── run_sft_qwen05_peft.sh │ ├── run_sft_qwen05_sp2_liger.sh │ ├── run_sft_sp_loss_match.sh │ └── test_sp_loss_match.py ├── utility │ └── test_tensor_dict_utilities.py └── verl │ └── utils │ └── dataset │ ├── test_rl_dataset.py │ ├── test_rm_dataset.py │ └── test_sft_dataset.py └── verl ├── __init__.py ├── __pycache__ ├── __init__.cpython-310.pyc └── protocol.cpython-310.pyc ├── models ├── README.md ├── __init__.py ├── __pycache__ │ ├── __init__.cpython-310.pyc │ └── registry.cpython-310.pyc ├── llama │ ├── __init__.py │ └── megatron │ │ ├── __init__.py │ │ ├── checkpoint_utils │ │ ├── __init__.py │ │ ├── llama_loader.py │ │ └── llama_saver.py │ │ ├── layers │ │ ├── __init__.py │ │ ├── parallel_attention.py │ │ ├── parallel_decoder.py │ │ ├── parallel_linear.py │ │ ├── parallel_mlp.py │ │ └── parallel_rmsnorm.py │ │ └── modeling_llama_megatron.py ├── qwen2 │ ├── __init__.py │ └── megatron │ │ ├── __init__.py │ │ ├── checkpoint_utils │ │ ├── __init__.py │ │ ├── qwen2_loader.py │ │ └── qwen2_saver.py │ │ ├── layers │ │ ├── __init__.py │ │ ├── parallel_attention.py │ │ ├── parallel_decoder.py │ │ ├── parallel_linear.py │ │ ├── parallel_mlp.py │ │ └── parallel_rmsnorm.py │ │ └── modeling_qwen2_megatron.py ├── registry.py ├── transformers │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ └── qwen2_vl.cpython-310.pyc │ ├── llama.py │ ├── monkey_patch.py │ ├── qwen2.py │ └── qwen2_vl.py └── weight_loader_registry.py ├── protocol.py ├── single_controller ├── __init__.py ├── __pycache__ │ └── __init__.cpython-310.pyc ├── base │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ ├── decorator.cpython-310.pyc │ │ ├── worker.cpython-310.pyc │ │ └── worker_group.cpython-310.pyc │ ├── decorator.py │ ├── megatron │ │ ├── __init__.py │ │ ├── worker.py │ │ └── worker_group.py │ ├── register_center │ │ ├── __init__.py │ │ ├── __pycache__ │ │ │ ├── __init__.cpython-310.pyc │ │ │ └── ray.cpython-310.pyc │ │ └── ray.py │ ├── worker.py │ └── worker_group.py └── ray │ ├── __init__.py │ ├── __pycache__ │ ├── __init__.cpython-310.pyc │ └── base.cpython-310.pyc │ ├── base.py │ └── megatron.py ├── third_party ├── __init__.py ├── __pycache__ │ └── __init__.cpython-310.pyc └── vllm │ ├── __init__.py │ ├── __pycache__ │ └── __init__.cpython-310.pyc │ ├── vllm_spmd │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ └── dtensor_weight_loaders.cpython-310.pyc │ └── dtensor_weight_loaders.py │ ├── vllm_v_0_3_1 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── tokenizer.py │ ├── weight_loaders.py │ └── worker.py │ ├── vllm_v_0_4_2 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── dtensor_weight_loaders.py │ ├── hf_weight_loader.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── megatron_weight_loaders.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── spmd_gpu_executor.py │ ├── tokenizer.py │ └── worker.py │ ├── vllm_v_0_5_4 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── dtensor_weight_loaders.py │ ├── hf_weight_loader.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── megatron_weight_loaders.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── spmd_gpu_executor.py │ ├── tokenizer.py │ └── worker.py │ └── vllm_v_0_6_3 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── dtensor_weight_loaders.py │ ├── hf_weight_loader.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── megatron_weight_loaders.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── spmd_gpu_executor.py │ ├── tokenizer.py │ └── worker.py ├── trainer ├── __init__.py ├── __pycache__ │ ├── __init__.cpython-310.pyc │ └── main_ppo.cpython-310.pyc ├── config │ ├── evaluation.yaml │ ├── generation.yaml │ ├── ppo_megatron_trainer.yaml │ ├── ppo_trainer.yaml │ └── sft_trainer.yaml ├── fsdp_sft_trainer.py ├── main_eval.py ├── main_generation.py ├── main_ppo.py ├── ppo │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ ├── core_algos.cpython-310.pyc │ │ └── ray_trainer.cpython-310.pyc │ ├── core_algos.py │ └── ray_trainer.py └── runtime_env.yaml ├── utils ├── __init__.py ├── __pycache__ │ ├── __init__.cpython-310.pyc │ ├── flops_counter.cpython-310.pyc │ ├── fs.cpython-310.pyc │ ├── fsdp_utils.cpython-310.pyc │ ├── hdfs_io.cpython-310.pyc │ ├── import_utils.cpython-310.pyc │ ├── logging_utils.cpython-310.pyc │ ├── model.cpython-310.pyc │ ├── py_functional.cpython-310.pyc │ ├── seqlen_balancing.cpython-310.pyc │ ├── tokenizer.cpython-310.pyc │ ├── torch_dtypes.cpython-310.pyc │ ├── torch_functional.cpython-310.pyc │ ├── tracking.cpython-310.pyc │ └── ulysses.cpython-310.pyc ├── checkpoint │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ ├── checkpoint_manager.cpython-310.pyc │ │ └── fsdp_checkpoint_manager.cpython-310.pyc │ ├── checkpoint_manager.py │ ├── fsdp_checkpoint_manager.py │ └── reverse_ckpt.py ├── config.py ├── dataset │ ├── README.md │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ ├── rl_dataset.cpython-310.pyc │ │ ├── rm_dataset.cpython-310.pyc │ │ └── sft_dataset.cpython-310.pyc │ ├── rl_dataset.py │ ├── rm_dataset.py │ └── sft_dataset.py ├── debug │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ └── performance.cpython-310.pyc │ ├── performance.py │ └── trajectory_tracker.py ├── distributed.py ├── flops_counter.py ├── fs.py ├── fsdp_utils.py ├── hdfs_io.py ├── import_utils.py ├── logger │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ └── aggregate_logger.cpython-310.pyc │ └── aggregate_logger.py ├── logging_utils.py ├── megatron │ ├── __init__.py │ ├── memory.py │ ├── optimizer.py │ ├── pipeline_parallel.py │ ├── sequence_parallel.py │ └── tensor_parallel.py ├── megatron_utils.py ├── memory_buffer.py ├── model.py ├── py_functional.py ├── ray_utils.py ├── rendezvous │ ├── __init__.py │ └── ray_backend.py ├── reward_score │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ ├── drivelm_m_ds.cpython-310.pyc │ │ ├── geo3k.cpython-310.pyc │ │ ├── math.cpython-310.pyc │ │ └── math.cpython-311.pyc │ ├── cal_math.py │ ├── geo3k.py │ ├── gsm8k.py │ ├── prime_code │ │ ├── __init__.py │ │ ├── testing_util.py │ │ └── utils.py │ └── prime_math │ │ ├── __init__.py │ │ ├── grader.py │ │ └── math_normalize.py ├── seqlen_balancing.py ├── tokenizer.py ├── torch_dtypes.py ├── torch_functional.py ├── tracking.py └── ulysses.py ├── version └── version └── workers ├── __init__.py ├── __pycache__ ├── __init__.cpython-310.pyc └── fsdp_workers.cpython-310.pyc ├── actor ├── __init__.py ├── __pycache__ │ ├── __init__.cpython-310.pyc │ ├── base.cpython-310.pyc │ └── dp_actor.cpython-310.pyc ├── base.py ├── dp_actor.py └── megatron_actor.py ├── critic ├── __init__.py ├── base.py ├── dp_critic.py └── megatron_critic.py ├── fsdp_workers.py ├── megatron_workers.py ├── reward_manager ├── __init__.py ├── __pycache__ │ ├── __init__.cpython-310.pyc │ ├── call_vllm.cpython-310.pyc │ ├── naive.cpython-310.pyc │ └── prime.cpython-310.pyc ├── naive.py └── prime.py ├── reward_model ├── __init__.py ├── base.py └── megatron │ ├── __init__.py │ └── reward_model.py ├── rollout ├── __init__.py ├── __pycache__ │ ├── __init__.cpython-310.pyc │ ├── base.cpython-310.pyc │ └── hf_rollout.cpython-310.pyc ├── base.py ├── hf_rollout.py ├── naive │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-310.pyc │ │ └── naive_rollout.cpython-310.pyc │ └── naive_rollout.py ├── tokenizer.py └── vllm_rollout │ ├── __init__.py │ ├── __pycache__ │ ├── __init__.cpython-310.pyc │ └── vllm_rollout_spmd.cpython-310.pyc │ ├── fire_vllm_rollout.py │ ├── vllm_rollout.py │ └── vllm_rollout_spmd.py └── sharding_manager ├── __init__.py ├── __pycache__ ├── __init__.cpython-310.pyc ├── base.cpython-310.pyc ├── fsdp_ulysses.cpython-310.pyc └── fsdp_vllm.cpython-310.pyc ├── base.py ├── fsdp_ulysses.py ├── fsdp_vllm.py └── megatron_vllm.py /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/LICENSE -------------------------------------------------------------------------------- /Notice.txt: -------------------------------------------------------------------------------- 1 | Copyright 2023-2024 Bytedance Ltd. and/or its affiliates -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/README.md -------------------------------------------------------------------------------- /assets/llava1.5.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/assets/llava1.5.png -------------------------------------------------------------------------------- /assets/our_llava.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/assets/our_llava.png -------------------------------------------------------------------------------- /docker/Dockerfile.megatron: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docker/Dockerfile.megatron -------------------------------------------------------------------------------- /docker/Dockerfile.ngc.vllm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docker/Dockerfile.ngc.vllm -------------------------------------------------------------------------------- /docker/Dockerfile.rocm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docker/Dockerfile.rocm -------------------------------------------------------------------------------- /docker/Dockerfile.vemlp.vllm.te: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docker/Dockerfile.vemlp.vllm.te -------------------------------------------------------------------------------- /docs/Makefile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/Makefile -------------------------------------------------------------------------------- /docs/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/README.md -------------------------------------------------------------------------------- /docs/README_vllm0.7.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/README_vllm0.7.md -------------------------------------------------------------------------------- /docs/_static/logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/_static/logo.png -------------------------------------------------------------------------------- /docs/advance/dpo_extension.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/advance/dpo_extension.rst -------------------------------------------------------------------------------- /docs/advance/fsdp_extension.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/advance/fsdp_extension.rst -------------------------------------------------------------------------------- /docs/advance/megatron_extension.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/advance/megatron_extension.rst -------------------------------------------------------------------------------- /docs/advance/placement.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/advance/placement.rst -------------------------------------------------------------------------------- /docs/amd_tutorial/amd_build_dockerfile.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/amd_tutorial/amd_build_dockerfile.md -------------------------------------------------------------------------------- /docs/amd_tutorial/amd_existing_docker.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/amd_tutorial/amd_existing_docker.md -------------------------------------------------------------------------------- /docs/conf.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/conf.py -------------------------------------------------------------------------------- /docs/data.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/data.rst -------------------------------------------------------------------------------- /docs/examples/config.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/examples/config.rst -------------------------------------------------------------------------------- /docs/examples/gsm8k_example.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/examples/gsm8k_example.rst -------------------------------------------------------------------------------- /docs/examples/ppo_code_architecture.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/examples/ppo_code_architecture.rst -------------------------------------------------------------------------------- /docs/experiment/ppo.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/experiment/ppo.rst -------------------------------------------------------------------------------- /docs/faq/faq.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/faq/faq.rst -------------------------------------------------------------------------------- /docs/hybrid_flow.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/hybrid_flow.rst -------------------------------------------------------------------------------- /docs/index.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/index.rst -------------------------------------------------------------------------------- /docs/perf/perf_tuning.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/perf/perf_tuning.rst -------------------------------------------------------------------------------- /docs/preparation/prepare_data.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/preparation/prepare_data.rst -------------------------------------------------------------------------------- /docs/preparation/reward_function.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/preparation/reward_function.rst -------------------------------------------------------------------------------- /docs/requirements-docs.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/requirements-docs.txt -------------------------------------------------------------------------------- /docs/start/install.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/start/install.rst -------------------------------------------------------------------------------- /docs/start/quickstart.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/start/quickstart.rst -------------------------------------------------------------------------------- /docs/workers/fsdp_workers.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/workers/fsdp_workers.rst -------------------------------------------------------------------------------- /docs/workers/megatron_workers.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/workers/megatron_workers.rst -------------------------------------------------------------------------------- /docs/workers/ray_trainer.rst: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/docs/workers/ray_trainer.rst -------------------------------------------------------------------------------- /examples/data_preprocess/full_hh_rlhf.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/data_preprocess/full_hh_rlhf.py -------------------------------------------------------------------------------- /examples/data_preprocess/geo3k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/data_preprocess/geo3k.py -------------------------------------------------------------------------------- /examples/data_preprocess/gsm8k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/data_preprocess/gsm8k.py -------------------------------------------------------------------------------- /examples/data_preprocess/hellaswag.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/data_preprocess/hellaswag.py -------------------------------------------------------------------------------- /examples/data_preprocess/math_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/data_preprocess/math_dataset.py -------------------------------------------------------------------------------- /examples/generation/run_deepseek_v2_lite_math.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/generation/run_deepseek_v2_lite_math.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_deepseek7b_llm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/grpo_trainer/run_deepseek7b_llm.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_deepseek7b_llm_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/grpo_trainer/run_deepseek7b_llm_seq_balance.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_llava1_5-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/grpo_trainer/run_llava1_5-7b.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_qwen2-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/grpo_trainer/run_qwen2-7b.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_qwen2-7b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/grpo_trainer/run_qwen2-7b_seq_balance.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_qwen2_5_vl-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/grpo_trainer/run_qwen2_5_vl-7b.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek7b_llm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_deepseek7b_llm.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek7b_llm_modelscope.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_deepseek7b_llm_modelscope.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek7b_llm_sp2.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_deepseek7b_llm_sp2.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek_full_hh_rlhf.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_deepseek_full_hh_rlhf.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek_math_gsm8k_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_deepseek_math_gsm8k_megatron.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_deepseek_megatron.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_gemma.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_gemma.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_qwen2-7b.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_math_gsm8k_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_qwen2-7b_math_gsm8k_megatron.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_qwen2-7b_rm.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_rm_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_qwen2-7b_rm_seq_balance.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_qwen2-7b_seq_balance.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2.5-32b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/run_qwen2.5-32b.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/verl_getting_started.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ppo_trainer/verl_getting_started.ipynb -------------------------------------------------------------------------------- /examples/ray/tutorial.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/ray/tutorial.ipynb -------------------------------------------------------------------------------- /examples/remax_trainer/run_qwen2.5-3b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/remax_trainer/run_qwen2.5-3b_seq_balance.sh -------------------------------------------------------------------------------- /examples/remax_trainer/run_qwen2.5-7b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/remax_trainer/run_qwen2.5-7b_seq_balance.sh -------------------------------------------------------------------------------- /examples/rloo_trainer/run_qwen2-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/rloo_trainer/run_qwen2-7b.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_deepseek_6b7.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/sft/gsm8k/run_deepseek_6b7.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_gemma_2b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/sft/gsm8k/run_gemma_2b.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_gemma_7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/sft/gsm8k/run_gemma_7b.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_qwen_05_peft.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/sft/gsm8k/run_qwen_05_peft.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_qwen_05_sp2.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/sft/gsm8k/run_qwen_05_sp2.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_qwen_05_sp2_liger.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/sft/gsm8k/run_qwen_05_sp2_liger.sh -------------------------------------------------------------------------------- /examples/slurm/ray_on_slurm.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/slurm/ray_on_slurm.slurm -------------------------------------------------------------------------------- /examples/split_placement/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/split_placement/README.md -------------------------------------------------------------------------------- /examples/split_placement/config/ppo_trainer_split.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/split_placement/config/ppo_trainer_split.yaml -------------------------------------------------------------------------------- /examples/split_placement/main_ppo_split.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/split_placement/main_ppo_split.py -------------------------------------------------------------------------------- /examples/split_placement/run_deepseek7b_llm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/split_placement/run_deepseek7b_llm.sh -------------------------------------------------------------------------------- /examples/split_placement/split_monkey_patch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/examples/split_placement/split_monkey_patch.py -------------------------------------------------------------------------------- /patches/megatron_v4.patch: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/patches/megatron_v4.patch -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/pyproject.toml -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/requirements.txt -------------------------------------------------------------------------------- /scripts/format.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/scripts/format.sh -------------------------------------------------------------------------------- /scripts/model_merger.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/scripts/model_merger.py -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/setup.py -------------------------------------------------------------------------------- /tests/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/__init__.py -------------------------------------------------------------------------------- /tests/checkpoint/test_fsdp_ckpt.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/checkpoint/test_fsdp_ckpt.py -------------------------------------------------------------------------------- /tests/distro/requirements.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/distro/requirements.py -------------------------------------------------------------------------------- /tests/e2e/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/__init__.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/create_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/data/create_dataset.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/test.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/data/test.parquet -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/train.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/data/train.parquet -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/model/config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/create_model_tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/model/create_model_tokenizer.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/generation_config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/model/generation_config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/model.safetensors: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/model/model.safetensors -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/tokenizer_config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/model/tokenizer_config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/rl/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/rl/README.md -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/rl/main_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/arithmetic_sequence/rl/main_trainer.py -------------------------------------------------------------------------------- /tests/e2e/check_results.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/check_results.py -------------------------------------------------------------------------------- /tests/e2e/envs/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/envs/__init__.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/envs/digit_completion/__init__.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/task.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/envs/digit_completion/task.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/envs/digit_completion/tokenizer.py -------------------------------------------------------------------------------- /tests/e2e/run_deepseek_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_deepseek_megatron.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen2vl_geo3k_function_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen2vl_geo3k_function_rm.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_function_rm.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm_grpo.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_function_rm_grpo.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm_no_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_function_rm_no_rmpad.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm_remax.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_function_rm_remax.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_model_rm.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_liger_kernel.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_liger_kernel.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_no_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_no_rmpad.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_seq_balance.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_ulysses.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_ulysses.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_qwen_megatron.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_ray_trainer.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer_fire_sampling.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_ray_trainer_fire_sampling.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/e2e/run_ray_trainer_rmpad.sh -------------------------------------------------------------------------------- /tests/generation/run_gen_qwen05.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/generation/run_gen_qwen05.sh -------------------------------------------------------------------------------- /tests/gpu_utility/test_memory_buffers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/gpu_utility/test_memory_buffers.py -------------------------------------------------------------------------------- /tests/gpu_utility/test_ops.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/gpu_utility/test_ops.py -------------------------------------------------------------------------------- /tests/gpu_utility/test_torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/gpu_utility/test_torch_functional.py -------------------------------------------------------------------------------- /tests/kill_github_tests.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/kill_github_tests.sh -------------------------------------------------------------------------------- /tests/model/test_transformer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/model/test_transformer.py -------------------------------------------------------------------------------- /tests/model/test_transformers_ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/model/test_transformers_ulysses.py -------------------------------------------------------------------------------- /tests/ray/check_worker_alive/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/check_worker_alive/main.py -------------------------------------------------------------------------------- /tests/ray/detached_worker/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/detached_worker/README.md -------------------------------------------------------------------------------- /tests/ray/detached_worker/client.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/detached_worker/client.py -------------------------------------------------------------------------------- /tests/ray/detached_worker/run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/detached_worker/run.sh -------------------------------------------------------------------------------- /tests/ray/detached_worker/server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/detached_worker/server.py -------------------------------------------------------------------------------- /tests/ray/test_check_worker_alive.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_check_worker_alive.py -------------------------------------------------------------------------------- /tests/ray/test_colocated_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_colocated_workers.py -------------------------------------------------------------------------------- /tests/ray/test_data_transfer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_data_transfer.py -------------------------------------------------------------------------------- /tests/ray/test_driverfunc_to_worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_driverfunc_to_worker.py -------------------------------------------------------------------------------- /tests/ray/test_high_level_scheduling_api.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_high_level_scheduling_api.py -------------------------------------------------------------------------------- /tests/ray/test_ray_local_envs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_ray_local_envs.py -------------------------------------------------------------------------------- /tests/ray/test_rvdz.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_rvdz.py -------------------------------------------------------------------------------- /tests/ray/test_worker_group_basics.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_worker_group_basics.py -------------------------------------------------------------------------------- /tests/ray/test_worker_group_torch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/ray/test_worker_group_torch.py -------------------------------------------------------------------------------- /tests/rollout/run_fsdp_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/rollout/run_fsdp_vllm.py -------------------------------------------------------------------------------- /tests/rollout/test_vllm_hf_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/rollout/test_vllm_hf_loader.py -------------------------------------------------------------------------------- /tests/rollout/test_vllm_spmd.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/rollout/test_vllm_spmd.py -------------------------------------------------------------------------------- /tests/sandbox/test_sandbox.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sandbox/test_sandbox.py -------------------------------------------------------------------------------- /tests/sanity/check_license.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sanity/check_license.py -------------------------------------------------------------------------------- /tests/sanity/test_import.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sanity/test_import.py -------------------------------------------------------------------------------- /tests/sft/run_sft.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sft/run_sft.sh -------------------------------------------------------------------------------- /tests/sft/run_sft_qwen05_peft.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sft/run_sft_qwen05_peft.sh -------------------------------------------------------------------------------- /tests/sft/run_sft_qwen05_sp2_liger.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sft/run_sft_qwen05_sp2_liger.sh -------------------------------------------------------------------------------- /tests/sft/run_sft_sp_loss_match.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sft/run_sft_sp_loss_match.sh -------------------------------------------------------------------------------- /tests/sft/test_sp_loss_match.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/sft/test_sp_loss_match.py -------------------------------------------------------------------------------- /tests/utility/test_tensor_dict_utilities.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/utility/test_tensor_dict_utilities.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_rl_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/verl/utils/dataset/test_rl_dataset.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_rm_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/verl/utils/dataset/test_rm_dataset.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/tests/verl/utils/dataset/test_sft_dataset.py -------------------------------------------------------------------------------- /verl/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/__init__.py -------------------------------------------------------------------------------- /verl/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/__pycache__/protocol.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/__pycache__/protocol.cpython-310.pyc -------------------------------------------------------------------------------- /verl/models/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/README.md -------------------------------------------------------------------------------- /verl/models/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/__init__.py -------------------------------------------------------------------------------- /verl/models/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/models/__pycache__/registry.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/__pycache__/registry.cpython-310.pyc -------------------------------------------------------------------------------- /verl/models/llama/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/checkpoint_utils/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/llama_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/checkpoint_utils/llama_loader.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/llama_saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/checkpoint_utils/llama_saver.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/layers/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_attention.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/layers/parallel_attention.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_decoder.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/layers/parallel_decoder.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/layers/parallel_linear.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_mlp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/layers/parallel_mlp.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_rmsnorm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/layers/parallel_rmsnorm.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/modeling_llama_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/llama/megatron/modeling_llama_megatron.py -------------------------------------------------------------------------------- /verl/models/qwen2/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/checkpoint_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/checkpoint_utils/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/checkpoint_utils/qwen2_saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/checkpoint_utils/qwen2_saver.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/layers/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_attention.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/layers/parallel_attention.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_decoder.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/layers/parallel_decoder.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/layers/parallel_linear.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_mlp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/layers/parallel_mlp.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_rmsnorm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/layers/parallel_rmsnorm.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/modeling_qwen2_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/qwen2/megatron/modeling_qwen2_megatron.py -------------------------------------------------------------------------------- /verl/models/registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/registry.py -------------------------------------------------------------------------------- /verl/models/transformers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/transformers/__init__.py -------------------------------------------------------------------------------- /verl/models/transformers/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/transformers/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/models/transformers/__pycache__/qwen2_vl.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/transformers/__pycache__/qwen2_vl.cpython-310.pyc -------------------------------------------------------------------------------- /verl/models/transformers/llama.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/transformers/llama.py -------------------------------------------------------------------------------- /verl/models/transformers/monkey_patch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/transformers/monkey_patch.py -------------------------------------------------------------------------------- /verl/models/transformers/qwen2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/transformers/qwen2.py -------------------------------------------------------------------------------- /verl/models/transformers/qwen2_vl.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/transformers/qwen2_vl.py -------------------------------------------------------------------------------- /verl/models/weight_loader_registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/models/weight_loader_registry.py -------------------------------------------------------------------------------- /verl/protocol.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/protocol.py -------------------------------------------------------------------------------- /verl/single_controller/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/base/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/base/__pycache__/decorator.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/__pycache__/decorator.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/base/__pycache__/worker.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/__pycache__/worker.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/base/__pycache__/worker_group.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/__pycache__/worker_group.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/base/decorator.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/decorator.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/megatron/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/megatron/worker.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/megatron/worker_group.py -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/register_center/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/register_center/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/__pycache__/ray.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/register_center/__pycache__/ray.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/ray.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/register_center/ray.py -------------------------------------------------------------------------------- /verl/single_controller/base/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/worker.py -------------------------------------------------------------------------------- /verl/single_controller/base/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/base/worker_group.py -------------------------------------------------------------------------------- /verl/single_controller/ray/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/ray/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/ray/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/ray/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/ray/__pycache__/base.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/ray/__pycache__/base.cpython-310.pyc -------------------------------------------------------------------------------- /verl/single_controller/ray/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/ray/base.py -------------------------------------------------------------------------------- /verl/single_controller/ray/megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/single_controller/ray/megatron.py -------------------------------------------------------------------------------- /verl/third_party/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/__init__.py -------------------------------------------------------------------------------- /verl/third_party/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/third_party/vllm/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_spmd/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_spmd/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_spmd/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_spmd/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_spmd/__pycache__/dtensor_weight_loaders.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_spmd/__pycache__/dtensor_weight_loaders.cpython-310.pyc -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_spmd/dtensor_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_spmd/dtensor_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/arg_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/arg_utils.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/config.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/llm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/llm.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/llm_engine_sp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/llm_engine_sp.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/model_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/model_loader.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/model_runner.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/model_runner.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/parallel_state.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/parallel_state.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/tokenizer.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_3_1/worker.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/arg_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/arg_utils.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/config.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/dtensor_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/dtensor_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/hf_weight_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/hf_weight_loader.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/llm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/llm.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/llm_engine_sp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/llm_engine_sp.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/megatron_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/megatron_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/model_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/model_loader.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/model_runner.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/model_runner.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/parallel_state.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/parallel_state.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/spmd_gpu_executor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/spmd_gpu_executor.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/tokenizer.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_4_2/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_4_2/worker.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/arg_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/arg_utils.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/config.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/dtensor_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/dtensor_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/hf_weight_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/hf_weight_loader.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/llm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/llm.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/llm_engine_sp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/llm_engine_sp.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/megatron_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/megatron_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/model_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/model_runner.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/model_runner.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/parallel_state.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/parallel_state.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/spmd_gpu_executor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/spmd_gpu_executor.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/tokenizer.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_5_4/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_5_4/worker.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/arg_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/arg_utils.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/config.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/dtensor_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/dtensor_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/hf_weight_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/hf_weight_loader.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/llm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/llm.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/llm_engine_sp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/llm_engine_sp.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/megatron_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/megatron_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/model_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/model_runner.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/model_runner.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/parallel_state.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/parallel_state.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/tokenizer.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_6_3/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/third_party/vllm/vllm_v_0_6_3/worker.py -------------------------------------------------------------------------------- /verl/trainer/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/__init__.py -------------------------------------------------------------------------------- /verl/trainer/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/trainer/__pycache__/main_ppo.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/__pycache__/main_ppo.cpython-310.pyc -------------------------------------------------------------------------------- /verl/trainer/config/evaluation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/config/evaluation.yaml -------------------------------------------------------------------------------- /verl/trainer/config/generation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/config/generation.yaml -------------------------------------------------------------------------------- /verl/trainer/config/ppo_megatron_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/config/ppo_megatron_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/config/ppo_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/config/ppo_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/config/sft_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/config/sft_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/fsdp_sft_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/fsdp_sft_trainer.py -------------------------------------------------------------------------------- /verl/trainer/main_eval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/main_eval.py -------------------------------------------------------------------------------- /verl/trainer/main_generation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/main_generation.py -------------------------------------------------------------------------------- /verl/trainer/main_ppo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/main_ppo.py -------------------------------------------------------------------------------- /verl/trainer/ppo/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/ppo/__init__.py -------------------------------------------------------------------------------- /verl/trainer/ppo/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/ppo/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/trainer/ppo/__pycache__/core_algos.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/ppo/__pycache__/core_algos.cpython-310.pyc -------------------------------------------------------------------------------- /verl/trainer/ppo/__pycache__/ray_trainer.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/ppo/__pycache__/ray_trainer.cpython-310.pyc -------------------------------------------------------------------------------- /verl/trainer/ppo/core_algos.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/ppo/core_algos.py -------------------------------------------------------------------------------- /verl/trainer/ppo/ray_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/ppo/ray_trainer.py -------------------------------------------------------------------------------- /verl/trainer/runtime_env.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/trainer/runtime_env.yaml -------------------------------------------------------------------------------- /verl/utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__init__.py -------------------------------------------------------------------------------- /verl/utils/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/flops_counter.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/flops_counter.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/fs.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/fs.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/fsdp_utils.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/fsdp_utils.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/hdfs_io.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/hdfs_io.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/import_utils.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/import_utils.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/logging_utils.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/logging_utils.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/model.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/model.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/py_functional.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/py_functional.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/seqlen_balancing.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/seqlen_balancing.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/tokenizer.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/tokenizer.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/torch_dtypes.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/torch_dtypes.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/torch_functional.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/torch_functional.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/tracking.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/tracking.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/__pycache__/ulysses.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/__pycache__/ulysses.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/checkpoint/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/checkpoint/__init__.py -------------------------------------------------------------------------------- /verl/utils/checkpoint/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/checkpoint/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/checkpoint/__pycache__/checkpoint_manager.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/checkpoint/__pycache__/checkpoint_manager.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/checkpoint/__pycache__/fsdp_checkpoint_manager.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/checkpoint/__pycache__/fsdp_checkpoint_manager.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/checkpoint/checkpoint_manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/checkpoint/checkpoint_manager.py -------------------------------------------------------------------------------- /verl/utils/checkpoint/fsdp_checkpoint_manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/checkpoint/fsdp_checkpoint_manager.py -------------------------------------------------------------------------------- /verl/utils/checkpoint/reverse_ckpt.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/checkpoint/reverse_ckpt.py -------------------------------------------------------------------------------- /verl/utils/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/config.py -------------------------------------------------------------------------------- /verl/utils/dataset/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/README.md -------------------------------------------------------------------------------- /verl/utils/dataset/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/__init__.py -------------------------------------------------------------------------------- /verl/utils/dataset/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/dataset/__pycache__/rl_dataset.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/__pycache__/rl_dataset.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/dataset/__pycache__/rm_dataset.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/__pycache__/rm_dataset.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/dataset/__pycache__/sft_dataset.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/__pycache__/sft_dataset.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/dataset/rl_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/rl_dataset.py -------------------------------------------------------------------------------- /verl/utils/dataset/rm_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/rm_dataset.py -------------------------------------------------------------------------------- /verl/utils/dataset/sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/dataset/sft_dataset.py -------------------------------------------------------------------------------- /verl/utils/debug/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/debug/__init__.py -------------------------------------------------------------------------------- /verl/utils/debug/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/debug/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/debug/__pycache__/performance.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/debug/__pycache__/performance.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/debug/performance.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/debug/performance.py -------------------------------------------------------------------------------- /verl/utils/debug/trajectory_tracker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/debug/trajectory_tracker.py -------------------------------------------------------------------------------- /verl/utils/distributed.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/distributed.py -------------------------------------------------------------------------------- /verl/utils/flops_counter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/flops_counter.py -------------------------------------------------------------------------------- /verl/utils/fs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/fs.py -------------------------------------------------------------------------------- /verl/utils/fsdp_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/fsdp_utils.py -------------------------------------------------------------------------------- /verl/utils/hdfs_io.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/hdfs_io.py -------------------------------------------------------------------------------- /verl/utils/import_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/import_utils.py -------------------------------------------------------------------------------- /verl/utils/logger/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/logger/__init__.py -------------------------------------------------------------------------------- /verl/utils/logger/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/logger/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/logger/__pycache__/aggregate_logger.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/logger/__pycache__/aggregate_logger.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/logger/aggregate_logger.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/logger/aggregate_logger.py -------------------------------------------------------------------------------- /verl/utils/logging_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/logging_utils.py -------------------------------------------------------------------------------- /verl/utils/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/megatron/__init__.py -------------------------------------------------------------------------------- /verl/utils/megatron/memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/megatron/memory.py -------------------------------------------------------------------------------- /verl/utils/megatron/optimizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/megatron/optimizer.py -------------------------------------------------------------------------------- /verl/utils/megatron/pipeline_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/megatron/pipeline_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron/sequence_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/megatron/sequence_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron/tensor_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/megatron/tensor_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/megatron_utils.py -------------------------------------------------------------------------------- /verl/utils/memory_buffer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/memory_buffer.py -------------------------------------------------------------------------------- /verl/utils/model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/model.py -------------------------------------------------------------------------------- /verl/utils/py_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/py_functional.py -------------------------------------------------------------------------------- /verl/utils/ray_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/ray_utils.py -------------------------------------------------------------------------------- /verl/utils/rendezvous/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/rendezvous/__init__.py -------------------------------------------------------------------------------- /verl/utils/rendezvous/ray_backend.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/rendezvous/ray_backend.py -------------------------------------------------------------------------------- /verl/utils/reward_score/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/reward_score/__pycache__/drivelm_m_ds.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/__pycache__/drivelm_m_ds.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/reward_score/__pycache__/geo3k.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/__pycache__/geo3k.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/reward_score/__pycache__/math.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/__pycache__/math.cpython-310.pyc -------------------------------------------------------------------------------- /verl/utils/reward_score/__pycache__/math.cpython-311.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/__pycache__/math.cpython-311.pyc -------------------------------------------------------------------------------- /verl/utils/reward_score/cal_math.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/cal_math.py -------------------------------------------------------------------------------- /verl/utils/reward_score/geo3k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/geo3k.py -------------------------------------------------------------------------------- /verl/utils/reward_score/gsm8k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/gsm8k.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_code/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/prime_code/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_code/testing_util.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/prime_code/testing_util.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_code/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/prime_code/utils.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_math/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/prime_math/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_math/grader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/prime_math/grader.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_math/math_normalize.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/reward_score/prime_math/math_normalize.py -------------------------------------------------------------------------------- /verl/utils/seqlen_balancing.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/seqlen_balancing.py -------------------------------------------------------------------------------- /verl/utils/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/tokenizer.py -------------------------------------------------------------------------------- /verl/utils/torch_dtypes.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/torch_dtypes.py -------------------------------------------------------------------------------- /verl/utils/torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/torch_functional.py -------------------------------------------------------------------------------- /verl/utils/tracking.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/tracking.py -------------------------------------------------------------------------------- /verl/utils/ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/utils/ulysses.py -------------------------------------------------------------------------------- /verl/version/version: -------------------------------------------------------------------------------- 1 | 0.2.0.dev 2 | -------------------------------------------------------------------------------- /verl/workers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/__init__.py -------------------------------------------------------------------------------- /verl/workers/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/__pycache__/fsdp_workers.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/__pycache__/fsdp_workers.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/actor/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/actor/__init__.py -------------------------------------------------------------------------------- /verl/workers/actor/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/actor/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/actor/__pycache__/base.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/actor/__pycache__/base.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/actor/__pycache__/dp_actor.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/actor/__pycache__/dp_actor.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/actor/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/actor/base.py -------------------------------------------------------------------------------- /verl/workers/actor/dp_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/actor/dp_actor.py -------------------------------------------------------------------------------- /verl/workers/actor/megatron_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/actor/megatron_actor.py -------------------------------------------------------------------------------- /verl/workers/critic/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/critic/__init__.py -------------------------------------------------------------------------------- /verl/workers/critic/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/critic/base.py -------------------------------------------------------------------------------- /verl/workers/critic/dp_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/critic/dp_critic.py -------------------------------------------------------------------------------- /verl/workers/critic/megatron_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/critic/megatron_critic.py -------------------------------------------------------------------------------- /verl/workers/fsdp_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/fsdp_workers.py -------------------------------------------------------------------------------- /verl/workers/megatron_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/megatron_workers.py -------------------------------------------------------------------------------- /verl/workers/reward_manager/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_manager/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_manager/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_manager/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/reward_manager/__pycache__/call_vllm.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_manager/__pycache__/call_vllm.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/reward_manager/__pycache__/naive.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_manager/__pycache__/naive.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/reward_manager/__pycache__/prime.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_manager/__pycache__/prime.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/reward_manager/naive.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_manager/naive.py -------------------------------------------------------------------------------- /verl/workers/reward_manager/prime.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_manager/prime.py -------------------------------------------------------------------------------- /verl/workers/reward_model/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_model/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_model/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_model/base.py -------------------------------------------------------------------------------- /verl/workers/reward_model/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_model/megatron/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_model/megatron/reward_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/reward_model/megatron/reward_model.py -------------------------------------------------------------------------------- /verl/workers/rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/rollout/__pycache__/base.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/__pycache__/base.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/rollout/__pycache__/hf_rollout.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/__pycache__/hf_rollout.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/rollout/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/base.py -------------------------------------------------------------------------------- /verl/workers/rollout/hf_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/hf_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/naive/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/naive/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/naive/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/naive/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/rollout/naive/__pycache__/naive_rollout.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/naive/__pycache__/naive_rollout.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/rollout/naive/naive_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/naive/naive_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/tokenizer.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/vllm_rollout/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/vllm_rollout/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/__pycache__/vllm_rollout_spmd.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/vllm_rollout/__pycache__/vllm_rollout_spmd.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/fire_vllm_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/vllm_rollout/fire_vllm_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/vllm_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/vllm_rollout/vllm_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/vllm_rollout_spmd.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/rollout/vllm_rollout/vllm_rollout_spmd.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/__init__.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/__pycache__/__init__.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/__pycache__/__init__.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/sharding_manager/__pycache__/base.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/__pycache__/base.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/sharding_manager/__pycache__/fsdp_ulysses.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/__pycache__/fsdp_ulysses.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/sharding_manager/__pycache__/fsdp_vllm.cpython-310.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/__pycache__/fsdp_vllm.cpython-310.pyc -------------------------------------------------------------------------------- /verl/workers/sharding_manager/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/base.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/fsdp_ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/fsdp_ulysses.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/fsdp_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/fsdp_vllm.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/megatron_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/PRIS-CV/GRPO-for-Llava/HEAD/verl/workers/sharding_manager/megatron_vllm.py --------------------------------------------------------------------------------