├── .gitignore ├── .readthedocs.yaml ├── .style.yapf ├── LICENSE ├── Notice.txt ├── Readme.md ├── main_GRPO.sh ├── main_TRPA.sh ├── patches └── megatron_v4.patch ├── pyproject.toml ├── requirements.txt ├── scripts ├── data_preprocess │ └── kk_data_process.py └── eval │ ├── eval_with_generation.sh │ └── eval_with_generation_math.sh ├── setup.py ├── tests ├── __init__.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 │ │ │ ├── config │ │ │ └── ray_trainer.yaml │ │ │ └── main_trainer.py │ ├── check_results.py │ ├── envs │ │ ├── __init__.py │ │ └── digit_completion │ │ │ ├── __init__.py │ │ │ ├── task.py │ │ │ └── tokenizer.py │ ├── run_qwen_gsm8k_function_rm.sh │ ├── run_qwen_gsm8k_function_rm_no_rmpad.sh │ ├── run_qwen_gsm8k_model_rm.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_ray_trainer.sh │ └── run_ray_trainer_rmpad.sh ├── gpu_utility │ ├── test_memory_buffers.py │ ├── test_ops.py │ └── test_torch_functional.py ├── 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 ├── sanity │ ├── check_license.py │ └── test_import.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 ├── models ├── README.md ├── __init__.py ├── 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 ├── registry.py ├── transformers │ ├── __init__.py │ ├── llama.py │ ├── monkey_patch.py │ └── qwen2.py └── weight_loader_registry.py ├── protocol.py ├── single_controller ├── __init__.py ├── base │ ├── __init__.py │ ├── decorator.py │ ├── megatron │ │ ├── __init__.py │ │ ├── worker.py │ │ └── worker_group.py │ ├── register_center │ │ ├── __init__.py │ │ └── ray.py │ ├── worker.py │ └── worker_group.py ├── ray │ ├── __init__.py │ ├── base.py │ └── megatron.py └── version │ └── version ├── third_party ├── __init__.py └── vllm │ ├── __init__.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 ├── config │ ├── eval_with_generation.yaml │ ├── evaluation.yaml │ ├── generation.yaml │ ├── ppo_megatron_trainer.yaml │ ├── ppo_trainer.yaml │ └── sft_trainer.yaml ├── fsdp_sft_trainer.py ├── main_eval.py ├── main_eval_with_generation.py ├── main_generation.py ├── main_ppo.py ├── ppo │ ├── __init__.py │ ├── core_algos.py │ └── ray_trainer.py └── runtime_env.yaml ├── utils ├── __init__.py ├── config.py ├── dataset │ ├── README.md │ ├── __init__.py │ ├── rl_dataset.py │ ├── rm_dataset.py │ └── sft_dataset.py ├── debug │ ├── __init__.py │ ├── performance.py │ └── trajectory_tracker.py ├── distributed.py ├── flops_counter.py ├── fs.py ├── fsdp_utils.py ├── hdfs_io.py ├── import_utils.py ├── logger │ ├── __init__.py │ └── aggregate_logger.py ├── logging_utils.py ├── megatron │ ├── __init__.py │ ├── memory.py │ ├── optimizer.py │ ├── optimizer_config.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 │ ├── countdown.py │ ├── deepscaler │ │ ├── globals.py │ │ ├── rewards │ │ │ ├── __init__.py │ │ │ ├── math_reward.py │ │ │ ├── math_reward_pl.py │ │ │ ├── math_utils │ │ │ │ ├── __init__.py │ │ │ │ └── utils.py │ │ │ └── reward_types.py │ │ ├── system_prompts.py │ │ └── utils.py │ ├── gsm8k.py │ ├── kk.py │ ├── math.py │ └── multiply.py ├── seqlen_balancing.py ├── tokenizer.py ├── torch_dtypes.py ├── torch_functional.py ├── tracking.py └── ulysses.py ├── version └── version └── workers ├── __init__.py ├── actor ├── __init__.py ├── 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_model ├── __init__.py ├── base.py └── megatron │ ├── __init__.py │ └── reward_model.py ├── rollout ├── __init__.py ├── base.py ├── hf_rollout.py ├── naive │ ├── __init__.py │ └── naive_rollout.py ├── tokenizer.py └── vllm_rollout │ ├── __init__.py │ └── vllm_rollout.py └── sharding_manager ├── __init__.py ├── base.py ├── fsdp_ulysses.py ├── fsdp_vllm.py └── megatron_vllm.py /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/.gitignore -------------------------------------------------------------------------------- /.readthedocs.yaml: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /.style.yapf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/.style.yapf -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/LICENSE -------------------------------------------------------------------------------- /Notice.txt: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /Readme.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/Readme.md -------------------------------------------------------------------------------- /main_GRPO.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/main_GRPO.sh -------------------------------------------------------------------------------- /main_TRPA.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/main_TRPA.sh -------------------------------------------------------------------------------- /patches/megatron_v4.patch: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/patches/megatron_v4.patch -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/pyproject.toml -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/requirements.txt -------------------------------------------------------------------------------- /scripts/data_preprocess/kk_data_process.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/scripts/data_preprocess/kk_data_process.py -------------------------------------------------------------------------------- /scripts/eval/eval_with_generation.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/scripts/eval/eval_with_generation.sh -------------------------------------------------------------------------------- /scripts/eval/eval_with_generation_math.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/scripts/eval/eval_with_generation_math.sh -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/setup.py -------------------------------------------------------------------------------- /tests/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/__init__.py -------------------------------------------------------------------------------- /tests/e2e/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/__init__.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/create_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/data/create_dataset.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/test.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/data/test.parquet -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/train.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/data/train.parquet -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/model/config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/create_model_tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/model/create_model_tokenizer.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/generation_config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/model/generation_config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/model.safetensors: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/model/model.safetensors -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/tokenizer_config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/model/tokenizer_config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/rl/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/rl/README.md -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/rl/config/ray_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/rl/config/ray_trainer.yaml -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/rl/main_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/arithmetic_sequence/rl/main_trainer.py -------------------------------------------------------------------------------- /tests/e2e/check_results.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/check_results.py -------------------------------------------------------------------------------- /tests/e2e/envs/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/envs/__init__.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/envs/digit_completion/__init__.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/task.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/envs/digit_completion/task.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/envs/digit_completion/tokenizer.py -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/run_qwen_gsm8k_function_rm.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm_no_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/run_qwen_gsm8k_function_rm_no_rmpad.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/run_qwen_gsm8k_model_rm.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_no_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_seq_balance.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_ulysses.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_ulysses.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/run_ray_trainer.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/e2e/run_ray_trainer_rmpad.sh -------------------------------------------------------------------------------- /tests/gpu_utility/test_memory_buffers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/gpu_utility/test_memory_buffers.py -------------------------------------------------------------------------------- /tests/gpu_utility/test_ops.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/gpu_utility/test_ops.py -------------------------------------------------------------------------------- /tests/gpu_utility/test_torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/gpu_utility/test_torch_functional.py -------------------------------------------------------------------------------- /tests/model/test_transformer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/model/test_transformer.py -------------------------------------------------------------------------------- /tests/model/test_transformers_ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/model/test_transformers_ulysses.py -------------------------------------------------------------------------------- /tests/ray/check_worker_alive/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/check_worker_alive/main.py -------------------------------------------------------------------------------- /tests/ray/detached_worker/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/detached_worker/README.md -------------------------------------------------------------------------------- /tests/ray/detached_worker/client.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/detached_worker/client.py -------------------------------------------------------------------------------- /tests/ray/detached_worker/run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/detached_worker/run.sh -------------------------------------------------------------------------------- /tests/ray/detached_worker/server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/detached_worker/server.py -------------------------------------------------------------------------------- /tests/ray/test_check_worker_alive.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_check_worker_alive.py -------------------------------------------------------------------------------- /tests/ray/test_colocated_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_colocated_workers.py -------------------------------------------------------------------------------- /tests/ray/test_data_transfer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_data_transfer.py -------------------------------------------------------------------------------- /tests/ray/test_driverfunc_to_worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_driverfunc_to_worker.py -------------------------------------------------------------------------------- /tests/ray/test_high_level_scheduling_api.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_high_level_scheduling_api.py -------------------------------------------------------------------------------- /tests/ray/test_ray_local_envs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_ray_local_envs.py -------------------------------------------------------------------------------- /tests/ray/test_rvdz.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_rvdz.py -------------------------------------------------------------------------------- /tests/ray/test_worker_group_basics.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_worker_group_basics.py -------------------------------------------------------------------------------- /tests/ray/test_worker_group_torch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/ray/test_worker_group_torch.py -------------------------------------------------------------------------------- /tests/rollout/run_fsdp_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/rollout/run_fsdp_vllm.py -------------------------------------------------------------------------------- /tests/rollout/test_vllm_hf_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/rollout/test_vllm_hf_loader.py -------------------------------------------------------------------------------- /tests/sanity/check_license.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/sanity/check_license.py -------------------------------------------------------------------------------- /tests/sanity/test_import.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/sanity/test_import.py -------------------------------------------------------------------------------- /tests/utility/test_tensor_dict_utilities.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/utility/test_tensor_dict_utilities.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_rl_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/verl/utils/dataset/test_rl_dataset.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_rm_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/verl/utils/dataset/test_rm_dataset.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/tests/verl/utils/dataset/test_sft_dataset.py -------------------------------------------------------------------------------- /verl/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/__init__.py -------------------------------------------------------------------------------- /verl/models/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/README.md -------------------------------------------------------------------------------- /verl/models/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/checkpoint_utils/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/llama_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/checkpoint_utils/llama_loader.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/llama_saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/checkpoint_utils/llama_saver.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/layers/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_attention.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/layers/parallel_attention.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_decoder.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/layers/parallel_decoder.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/layers/parallel_linear.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_mlp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/layers/parallel_mlp.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_rmsnorm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/layers/parallel_rmsnorm.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/modeling_llama_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/llama/megatron/modeling_llama_megatron.py -------------------------------------------------------------------------------- /verl/models/registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/registry.py -------------------------------------------------------------------------------- /verl/models/transformers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/transformers/__init__.py -------------------------------------------------------------------------------- /verl/models/transformers/llama.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/transformers/llama.py -------------------------------------------------------------------------------- /verl/models/transformers/monkey_patch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/transformers/monkey_patch.py -------------------------------------------------------------------------------- /verl/models/transformers/qwen2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/transformers/qwen2.py -------------------------------------------------------------------------------- /verl/models/weight_loader_registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/models/weight_loader_registry.py -------------------------------------------------------------------------------- /verl/protocol.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/protocol.py -------------------------------------------------------------------------------- /verl/single_controller/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/decorator.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/decorator.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/megatron/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/megatron/worker.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/megatron/worker_group.py -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/register_center/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/ray.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/register_center/ray.py -------------------------------------------------------------------------------- /verl/single_controller/base/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/worker.py -------------------------------------------------------------------------------- /verl/single_controller/base/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/base/worker_group.py -------------------------------------------------------------------------------- /verl/single_controller/ray/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/ray/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/ray/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/ray/base.py -------------------------------------------------------------------------------- /verl/single_controller/ray/megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/single_controller/ray/megatron.py -------------------------------------------------------------------------------- /verl/single_controller/version/version: -------------------------------------------------------------------------------- 1 | 0.0.2 -------------------------------------------------------------------------------- /verl/third_party/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/third_party/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/third_party/vllm/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/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/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/third_party/vllm/vllm_v_0_6_3/worker.py -------------------------------------------------------------------------------- /verl/trainer/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/__init__.py -------------------------------------------------------------------------------- /verl/trainer/config/eval_with_generation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/config/eval_with_generation.yaml -------------------------------------------------------------------------------- /verl/trainer/config/evaluation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/config/evaluation.yaml -------------------------------------------------------------------------------- /verl/trainer/config/generation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/config/generation.yaml -------------------------------------------------------------------------------- /verl/trainer/config/ppo_megatron_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/config/ppo_megatron_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/config/ppo_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/config/ppo_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/config/sft_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/config/sft_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/fsdp_sft_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/fsdp_sft_trainer.py -------------------------------------------------------------------------------- /verl/trainer/main_eval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/main_eval.py -------------------------------------------------------------------------------- /verl/trainer/main_eval_with_generation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/main_eval_with_generation.py -------------------------------------------------------------------------------- /verl/trainer/main_generation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/main_generation.py -------------------------------------------------------------------------------- /verl/trainer/main_ppo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/main_ppo.py -------------------------------------------------------------------------------- /verl/trainer/ppo/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/ppo/__init__.py -------------------------------------------------------------------------------- /verl/trainer/ppo/core_algos.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/ppo/core_algos.py -------------------------------------------------------------------------------- /verl/trainer/ppo/ray_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/ppo/ray_trainer.py -------------------------------------------------------------------------------- /verl/trainer/runtime_env.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/trainer/runtime_env.yaml -------------------------------------------------------------------------------- /verl/utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/__init__.py -------------------------------------------------------------------------------- /verl/utils/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/config.py -------------------------------------------------------------------------------- /verl/utils/dataset/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/dataset/README.md -------------------------------------------------------------------------------- /verl/utils/dataset/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/dataset/__init__.py -------------------------------------------------------------------------------- /verl/utils/dataset/rl_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/dataset/rl_dataset.py -------------------------------------------------------------------------------- /verl/utils/dataset/rm_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/dataset/rm_dataset.py -------------------------------------------------------------------------------- /verl/utils/dataset/sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/dataset/sft_dataset.py -------------------------------------------------------------------------------- /verl/utils/debug/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/debug/__init__.py -------------------------------------------------------------------------------- /verl/utils/debug/performance.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/debug/performance.py -------------------------------------------------------------------------------- /verl/utils/debug/trajectory_tracker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/debug/trajectory_tracker.py -------------------------------------------------------------------------------- /verl/utils/distributed.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/distributed.py -------------------------------------------------------------------------------- /verl/utils/flops_counter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/flops_counter.py -------------------------------------------------------------------------------- /verl/utils/fs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/fs.py -------------------------------------------------------------------------------- /verl/utils/fsdp_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/fsdp_utils.py -------------------------------------------------------------------------------- /verl/utils/hdfs_io.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/hdfs_io.py -------------------------------------------------------------------------------- /verl/utils/import_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/import_utils.py -------------------------------------------------------------------------------- /verl/utils/logger/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/logger/__init__.py -------------------------------------------------------------------------------- /verl/utils/logger/aggregate_logger.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/logger/aggregate_logger.py -------------------------------------------------------------------------------- /verl/utils/logging_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/logging_utils.py -------------------------------------------------------------------------------- /verl/utils/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron/__init__.py -------------------------------------------------------------------------------- /verl/utils/megatron/memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron/memory.py -------------------------------------------------------------------------------- /verl/utils/megatron/optimizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron/optimizer.py -------------------------------------------------------------------------------- /verl/utils/megatron/optimizer_config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron/optimizer_config.py -------------------------------------------------------------------------------- /verl/utils/megatron/pipeline_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron/pipeline_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron/sequence_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron/sequence_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron/tensor_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron/tensor_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/megatron_utils.py -------------------------------------------------------------------------------- /verl/utils/memory_buffer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/memory_buffer.py -------------------------------------------------------------------------------- /verl/utils/model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/model.py -------------------------------------------------------------------------------- /verl/utils/py_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/py_functional.py -------------------------------------------------------------------------------- /verl/utils/ray_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/ray_utils.py -------------------------------------------------------------------------------- /verl/utils/rendezvous/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/rendezvous/__init__.py -------------------------------------------------------------------------------- /verl/utils/rendezvous/ray_backend.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/rendezvous/ray_backend.py -------------------------------------------------------------------------------- /verl/utils/reward_score/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/countdown.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/countdown.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/globals.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/globals.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/rewards/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/rewards/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/rewards/math_reward.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/rewards/math_reward.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/rewards/math_reward_pl.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/rewards/math_reward_pl.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/rewards/math_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/rewards/math_utils/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/rewards/math_utils/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/rewards/math_utils/utils.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/rewards/reward_types.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/rewards/reward_types.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/system_prompts.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/system_prompts.py -------------------------------------------------------------------------------- /verl/utils/reward_score/deepscaler/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/deepscaler/utils.py -------------------------------------------------------------------------------- /verl/utils/reward_score/gsm8k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/gsm8k.py -------------------------------------------------------------------------------- /verl/utils/reward_score/kk.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/kk.py -------------------------------------------------------------------------------- /verl/utils/reward_score/math.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/math.py -------------------------------------------------------------------------------- /verl/utils/reward_score/multiply.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/reward_score/multiply.py -------------------------------------------------------------------------------- /verl/utils/seqlen_balancing.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/seqlen_balancing.py -------------------------------------------------------------------------------- /verl/utils/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/tokenizer.py -------------------------------------------------------------------------------- /verl/utils/torch_dtypes.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/torch_dtypes.py -------------------------------------------------------------------------------- /verl/utils/torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/torch_functional.py -------------------------------------------------------------------------------- /verl/utils/tracking.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/tracking.py -------------------------------------------------------------------------------- /verl/utils/ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/utils/ulysses.py -------------------------------------------------------------------------------- /verl/version/version: -------------------------------------------------------------------------------- 1 | 0.1 -------------------------------------------------------------------------------- /verl/workers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/__init__.py -------------------------------------------------------------------------------- /verl/workers/actor/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/actor/__init__.py -------------------------------------------------------------------------------- /verl/workers/actor/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/actor/base.py -------------------------------------------------------------------------------- /verl/workers/actor/dp_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/actor/dp_actor.py -------------------------------------------------------------------------------- /verl/workers/actor/megatron_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/actor/megatron_actor.py -------------------------------------------------------------------------------- /verl/workers/critic/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/critic/__init__.py -------------------------------------------------------------------------------- /verl/workers/critic/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/critic/base.py -------------------------------------------------------------------------------- /verl/workers/critic/dp_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/critic/dp_critic.py -------------------------------------------------------------------------------- /verl/workers/critic/megatron_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/critic/megatron_critic.py -------------------------------------------------------------------------------- /verl/workers/fsdp_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/fsdp_workers.py -------------------------------------------------------------------------------- /verl/workers/megatron_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/megatron_workers.py -------------------------------------------------------------------------------- /verl/workers/reward_model/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/reward_model/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_model/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/reward_model/base.py -------------------------------------------------------------------------------- /verl/workers/reward_model/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/reward_model/megatron/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_model/megatron/reward_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/reward_model/megatron/reward_model.py -------------------------------------------------------------------------------- /verl/workers/rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/base.py -------------------------------------------------------------------------------- /verl/workers/rollout/hf_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/hf_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/naive/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/naive/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/naive/naive_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/naive/naive_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/tokenizer.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/vllm_rollout/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/vllm_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/rollout/vllm_rollout/vllm_rollout.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/sharding_manager/__init__.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/sharding_manager/base.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/fsdp_ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/sharding_manager/fsdp_ulysses.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/fsdp_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/sharding_manager/fsdp_vllm.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/megatron_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/XueruiSu/Trust-Region-Preference-Approximation/HEAD/verl/workers/sharding_manager/megatron_vllm.py --------------------------------------------------------------------------------