mirror of https://github.com/inclusionAI/AReaL
349 lines
12 KiB
Plaintext
349 lines
12 KiB
Plaintext
LICENSE
|
|
MANIFEST.in
|
|
README.md
|
|
pyproject.toml
|
|
setup.py
|
|
arealite/ppo_functional.py
|
|
arealite/api/cli_args.py
|
|
arealite/api/engine_api.py
|
|
arealite/api/env_api.py
|
|
arealite/api/io_struct.py
|
|
arealite/api/llm_client_api.py
|
|
arealite/api/llm_server_api.py
|
|
arealite/api/reward_api.py
|
|
arealite/api/rollout_api.py
|
|
arealite/api/trainer_api.py
|
|
arealite/api/vlm_client_api.py
|
|
arealite/api/workflow_api.py
|
|
arealite/cli/launch_server.py
|
|
arealite/cli/launch_trainer.py
|
|
arealite/dataset/__init__.py
|
|
arealite/engine/__init__.py
|
|
arealite/engine/constant.py
|
|
arealite/engine/fsdp_engine.py
|
|
arealite/engine/hf_engine.py
|
|
arealite/engine/sglang_engine.py
|
|
arealite/engine/sglang_remote.py
|
|
arealite/engine/vl_fsdp_engine.py
|
|
arealite/engine/sft/lm_engine.py
|
|
arealite/env/__init__.py
|
|
arealite/impl/engine/fsdp_wrapper.py
|
|
arealite/impl/engine/hf_wrapper.py
|
|
arealite/impl/rlvr/rlvr_collector.py
|
|
arealite/impl/rlvr/vl_rlvr_collector.py
|
|
arealite/impl/rlvr/rewards/areal_code.py
|
|
arealite/impl/rlvr/rewards/areal_math.py
|
|
arealite/impl/rlvr/rewards/clevr_count_70k.py
|
|
arealite/impl/rlvr/rewards/gsm8k.py
|
|
arealite/impl/trainer/grpo.py
|
|
arealite/impl/trainer/sft.py
|
|
arealite/impl/trainer/vl_grpo.py
|
|
arealite/impl/trainer/vl_sft.py
|
|
arealite/launcher/with_ray.py
|
|
arealite/launcher/with_scheduler.py
|
|
arealite/reward/__init__.py
|
|
arealite/system/rollout_controller.py
|
|
arealite/system/rollout_worker.py
|
|
arealite/system/sglang_client.py
|
|
arealite/system/sglang_server.py
|
|
arealite/system/vl_sglang_client.py
|
|
arealite/tests/test_engine.py
|
|
arealite/tests/test_fsdp_engine.py
|
|
arealite/tests/test_grpo.py
|
|
arealite/tests/test_rlvr_workflow.py
|
|
arealite/tests/test_rollout.py
|
|
arealite/tests/test_rollout_controller.py
|
|
arealite/tests/test_sft.py
|
|
arealite/tests/test_sglang_client.py
|
|
arealite/tests/test_sglang_engine.py
|
|
arealite/tests/test_utils.py
|
|
arealite/tests/test_vlm_grpo.py
|
|
arealite/tests/test_vlm_sft.py
|
|
arealite/tests/test_wrapper.py
|
|
arealite/tests/utils.py
|
|
arealite/utils/__init__.py
|
|
arealite/utils/data.py
|
|
arealite/utils/evaluator.py
|
|
arealite/utils/fs.py
|
|
arealite/utils/fsdp.py
|
|
arealite/utils/functional.py
|
|
arealite/utils/padding.py
|
|
arealite/utils/save_load.py
|
|
arealite/utils/saver.py
|
|
arealite/utils/stats_logger.py
|
|
arealite/utils/wrapper.py
|
|
arealite/workflow/rlvr.py
|
|
benchmark/verl_v0_3_0_post1_76084d3/build_cmd.py
|
|
csrc/cugae/gae.cu
|
|
csrc/interval_op/interval_op.cpp
|
|
csrc/interval_op/interval_op.cu
|
|
evaluation/aggregate_acc_from_generated.py
|
|
evaluation/cf_elo_caculator.py
|
|
evaluation/code_eval.py
|
|
evaluation/data_loader.py
|
|
evaluation/eval_and_aggregate.py
|
|
evaluation/evaluate.py
|
|
evaluation/examples.py
|
|
evaluation/grader.py
|
|
evaluation/math_eval.py
|
|
evaluation/math_utils.py
|
|
evaluation/model_utils.py
|
|
evaluation/parser.py
|
|
evaluation/python_executor.py
|
|
evaluation/rm_maj_eval.py
|
|
evaluation/trajectory.py
|
|
evaluation/utils.py
|
|
evaluation/code_verifier/local_verify.py
|
|
evaluation/code_verifier/testing_util.py
|
|
evaluation/latex2sympy/__init__.py
|
|
evaluation/latex2sympy/asciimath_printer.py
|
|
evaluation/latex2sympy/latex2sympy2.py
|
|
evaluation/latex2sympy/setup.py
|
|
evaluation/latex2sympy/gen/PSLexer.py
|
|
evaluation/latex2sympy/gen/PSListener.py
|
|
evaluation/latex2sympy/gen/PSParser.py
|
|
evaluation/latex2sympy/gen/__init__.py
|
|
evaluation/latex2sympy/sandbox/linalg_equations.py
|
|
evaluation/latex2sympy/sandbox/linalg_span.py
|
|
evaluation/latex2sympy/sandbox/matrix.py
|
|
evaluation/latex2sympy/sandbox/matrix_placeholders.py
|
|
evaluation/latex2sympy/sandbox/sandbox.py
|
|
evaluation/latex2sympy/sandbox/sandbox_equality.py
|
|
evaluation/latex2sympy/sandbox/sectan.py
|
|
evaluation/latex2sympy/sandbox/vector.py
|
|
evaluation/latex2sympy/tests/__init__.py
|
|
evaluation/latex2sympy/tests/abs_test.py
|
|
evaluation/latex2sympy/tests/all_bad_test.py
|
|
evaluation/latex2sympy/tests/all_good_test.py
|
|
evaluation/latex2sympy/tests/atom_expr_test.py
|
|
evaluation/latex2sympy/tests/binomial_test.py
|
|
evaluation/latex2sympy/tests/ceil_test.py
|
|
evaluation/latex2sympy/tests/complex_test.py
|
|
evaluation/latex2sympy/tests/context.py
|
|
evaluation/latex2sympy/tests/exp_test.py
|
|
evaluation/latex2sympy/tests/floor_test.py
|
|
evaluation/latex2sympy/tests/gcd_test.py
|
|
evaluation/latex2sympy/tests/greek_test.py
|
|
evaluation/latex2sympy/tests/grouping_test.py
|
|
evaluation/latex2sympy/tests/lcm_test.py
|
|
evaluation/latex2sympy/tests/left_right_cdot_test.py
|
|
evaluation/latex2sympy/tests/linalg_test.py
|
|
evaluation/latex2sympy/tests/max_test.py
|
|
evaluation/latex2sympy/tests/min_test.py
|
|
evaluation/latex2sympy/tests/mod_test.py
|
|
evaluation/latex2sympy/tests/overline_test.py
|
|
evaluation/latex2sympy/tests/pi_test.py
|
|
evaluation/latex2sympy/tests/trig_test.py
|
|
evaluation/latex2sympy/tests/variable_test.py
|
|
examples/arealite/gsm8k_sft.py
|
|
examples/arealite/dataset/clevr_count_70k.py
|
|
examples/arealite/dataset/gsm8k.py
|
|
examples/data_preprocess/codeforce_process.py
|
|
examples/data_preprocess/math_code_process.py
|
|
examples/data_preprocess/math_process.py
|
|
examples/data_preprocess/preprocess_training_data.py
|
|
examples/env/setup_env_and_start_train.py
|
|
examples/env/validate_installation.py
|
|
functioncall/__init__.py
|
|
functioncall/base/__init__.py
|
|
functioncall/base/call.py
|
|
functioncall/base/utils.py
|
|
functioncall/code/__init__.py
|
|
functioncall/code/local_verify.py
|
|
functioncall/code/verify.py
|
|
functioncall/code/function/handler.py
|
|
functioncall/code/function/testing_util.py
|
|
functioncall/math/__init__.py
|
|
functioncall/math/verify.py
|
|
functioncall/math/function/grader.py
|
|
functioncall/math/function/handler.py
|
|
functioncall/math/function/parser.py
|
|
functioncall/test/performance_eval.py
|
|
realhf/__init__.py
|
|
realhf/utils.py
|
|
realhf/version.py
|
|
realhf.egg-info/PKG-INFO
|
|
realhf.egg-info/SOURCES.txt
|
|
realhf.egg-info/dependency_links.txt
|
|
realhf.egg-info/requires.txt
|
|
realhf.egg-info/top_level.txt
|
|
realhf/api/cli_args.py
|
|
realhf/api/core/agent_api.py
|
|
realhf/api/core/config.py
|
|
realhf/api/core/data_api.py
|
|
realhf/api/core/dfg.py
|
|
realhf/api/core/env_api.py
|
|
realhf/api/core/model_api.py
|
|
realhf/api/core/system_api.py
|
|
realhf/api/from_hf/__init__.py
|
|
realhf/api/from_hf/gemma.py
|
|
realhf/api/from_hf/gpt2.py
|
|
realhf/api/from_hf/llama.py
|
|
realhf/api/from_hf/mistral.py
|
|
realhf/api/from_hf/mixtral.py
|
|
realhf/api/from_hf/qwen2.py
|
|
realhf/api/from_hf/qwen3.py
|
|
realhf/api/quickstart/__init__.py
|
|
realhf/api/quickstart/device_mesh.py
|
|
realhf/api/quickstart/entrypoint.py
|
|
realhf/api/quickstart/search.py
|
|
realhf/apps/__init__.py
|
|
realhf/apps/main.py
|
|
realhf/apps/quickstart.py
|
|
realhf/apps/remote.py
|
|
realhf/base/__init__.py
|
|
realhf/base/cluster.py
|
|
realhf/base/constants.py
|
|
realhf/base/datapack.py
|
|
realhf/base/gpu_utils.py
|
|
realhf/base/importing.py
|
|
realhf/base/logging.py
|
|
realhf/base/monitor.py
|
|
realhf/base/name_resolve.py
|
|
realhf/base/names.py
|
|
realhf/base/network.py
|
|
realhf/base/numpy_utils.py
|
|
realhf/base/pkg_version.py
|
|
realhf/base/prologue.py
|
|
realhf/base/ray_utils.py
|
|
realhf/base/recover.py
|
|
realhf/base/saveload_utils.py
|
|
realhf/base/security.py
|
|
realhf/base/seeding.py
|
|
realhf/base/slurm_utils.py
|
|
realhf/base/stats_tracker.py
|
|
realhf/base/testing.py
|
|
realhf/base/timeutil.py
|
|
realhf/base/topology.py
|
|
realhf/experiments/async_exp/async_ppo_math_exp.py
|
|
realhf/experiments/async_exp/async_rl_exp.py
|
|
realhf/experiments/common/check.py
|
|
realhf/experiments/common/common.py
|
|
realhf/experiments/common/math_code_eval_exp.py
|
|
realhf/experiments/common/null_exp.py
|
|
realhf/experiments/common/ppo_math_exp.py
|
|
realhf/experiments/common/sft_exp.py
|
|
realhf/experiments/common/utils.py
|
|
realhf/impl/agent/__init__.py
|
|
realhf/impl/agent/math_multi_turn_agent.py
|
|
realhf/impl/agent/math_single_step_agent.py
|
|
realhf/impl/agent/null_agent.py
|
|
realhf/impl/dataset/__init__.py
|
|
realhf/impl/dataset/math_code_dataset.py
|
|
realhf/impl/dataset/math_parser.py
|
|
realhf/impl/dataset/prompt_answer_dataset.py
|
|
realhf/impl/dataset/prompt_dataset.py
|
|
realhf/impl/dataset/rw_paired_dataset.py
|
|
realhf/impl/environment/__init__.py
|
|
realhf/impl/environment/math_code_single_step_env.py
|
|
realhf/impl/model/__init__.py
|
|
realhf/impl/model/backend/inference.py
|
|
realhf/impl/model/backend/megatron.py
|
|
realhf/impl/model/backend/mock_train.py
|
|
realhf/impl/model/backend/pipe_runner.py
|
|
realhf/impl/model/backend/sglang.py
|
|
realhf/impl/model/backend/vllm.py
|
|
realhf/impl/model/backend/thirdparty/megatron/__init__.py
|
|
realhf/impl/model/backend/thirdparty/megatron/v0_6_0/lr_schduler.py
|
|
realhf/impl/model/backend/thirdparty/vllm/__init__.py
|
|
realhf/impl/model/backend/thirdparty/vllm/context.py
|
|
realhf/impl/model/backend/thirdparty/vllm/custom_cache_manager.py
|
|
realhf/impl/model/backend/thirdparty/vllm/engine.py
|
|
realhf/impl/model/backend/thirdparty/vllm/executor.py
|
|
realhf/impl/model/comm/global_comm.py
|
|
realhf/impl/model/comm/param_realloc.py
|
|
realhf/impl/model/conversion/hf_registry.py
|
|
realhf/impl/model/interface/fused_interface.py
|
|
realhf/impl/model/interface/math_rw_interface.py
|
|
realhf/impl/model/interface/ppo_interface.py
|
|
realhf/impl/model/interface/sft_interface.py
|
|
realhf/impl/model/modules/__init__.py
|
|
realhf/impl/model/modules/activations.py
|
|
realhf/impl/model/modules/attn.py
|
|
realhf/impl/model/modules/embedding.py
|
|
realhf/impl/model/modules/mlp.py
|
|
realhf/impl/model/modules/rms.py
|
|
realhf/impl/model/modules/rotary.py
|
|
realhf/impl/model/modules/moe/__init__.py
|
|
realhf/impl/model/modules/moe/experts.py
|
|
realhf/impl/model/modules/moe/layer.py
|
|
realhf/impl/model/modules/moe/router.py
|
|
realhf/impl/model/modules/moe/token_dispatcher.py
|
|
realhf/impl/model/nn/flatten_param.py
|
|
realhf/impl/model/nn/real_llm_api.py
|
|
realhf/impl/model/nn/real_llm_base.py
|
|
realhf/impl/model/nn/real_llm_generate.py
|
|
realhf/impl/model/nn/real_llm_parallel.py
|
|
realhf/impl/model/parallelism/pipeline_parallel/instruction.py
|
|
realhf/impl/model/parallelism/pipeline_parallel/p2p.py
|
|
realhf/impl/model/parallelism/pipeline_parallel/static_schedule.py
|
|
realhf/impl/model/parallelism/pipeline_parallel/tensor_storage.py
|
|
realhf/impl/model/parallelism/tensor_parallel/mappings.py
|
|
realhf/impl/model/parallelism/tensor_parallel/modules.py
|
|
realhf/impl/model/parallelism/tensor_parallel/utils.py
|
|
realhf/impl/model/utils/cuda_graph.py
|
|
realhf/impl/model/utils/dpo_functional.py
|
|
realhf/impl/model/utils/functional.py
|
|
realhf/impl/model/utils/logits_warper.py
|
|
realhf/impl/model/utils/moe.py
|
|
realhf/impl/model/utils/padding.py
|
|
realhf/impl/model/utils/ppo_functional.py
|
|
realhf/impl/model/utils/random.py
|
|
realhf/scheduler/client.py
|
|
realhf/scheduler/evaluator.py
|
|
realhf/scheduler/local/client.py
|
|
realhf/scheduler/slurm/client.py
|
|
realhf/scheduler/slurm/utils.py
|
|
realhf/system/__init__.py
|
|
realhf/system/buffer.py
|
|
realhf/system/controller.py
|
|
realhf/system/data_manager.py
|
|
realhf/system/flops_counter.py
|
|
realhf/system/function_executor.py
|
|
realhf/system/generation_server.py
|
|
realhf/system/gserver_manager.py
|
|
realhf/system/master_worker.py
|
|
realhf/system/model_function_call.py
|
|
realhf/system/model_worker.py
|
|
realhf/system/partial_rollout.py
|
|
realhf/system/push_pull_stream.py
|
|
realhf/system/redistributor.py
|
|
realhf/system/request_reply_stream.py
|
|
realhf/system/rollout_worker.py
|
|
realhf/system/stream_dataset.py
|
|
realhf/system/worker_base.py
|
|
realhf/system/worker_control.py
|
|
tests/__init__.py
|
|
tests/fixtures.py
|
|
tests/agent/test_math_single_step_agent.py
|
|
tests/comm/test_data_transfer.py
|
|
tests/comm/test_param_realloc.py
|
|
tests/cpp_extensions/test_cugae.py
|
|
tests/cpp_extensions/test_grouped_gemm.py
|
|
tests/cpp_extensions/test_interval_ops.py
|
|
tests/data/test_dfg.py
|
|
tests/data/test_dual_clip.py
|
|
tests/data/test_epoch_counter.py
|
|
tests/data/test_load_data.py
|
|
tests/data/test_sequence_gather_split.py
|
|
tests/data/test_stats_tracker.py
|
|
tests/distributed/test_find_port.py
|
|
tests/distributed/test_name_resolve.py
|
|
tests/experiments/test_buffer_recover.py
|
|
tests/experiments/test_math_ppo.py
|
|
tests/experiments/test_sft.py
|
|
tests/experiments/utils.py
|
|
tests/interfaces/test_multi_task_reward.py
|
|
tests/legacy/test_sglang_tp.py
|
|
tests/legacy/test_vllm_tp.py
|
|
tests/model/test_cpu_inference.py
|
|
tests/model/test_distributed_load_hf.py
|
|
tests/reward/test_math_reward.py
|
|
tests/system/test_gserver_manager.py
|
|
tests/system/test_partial_rollout.py
|
|
tests/system/test_push_pull_stream.py
|
|
tests/system/test_stream_dataset.py
|
|
training/main_async_ppo.py
|
|
training/main_sft.py
|
|
training/main_sync_ppo.py
|
|
training/utils.py |