├── .github └── workflows │ ├── dataset.yml │ ├── e2e_digit_completion.yml │ ├── e2e_digit_completion_fire.yml │ ├── e2e_gsm8k.yml │ ├── e2e_gsm8k_megatron.yml │ ├── e2e_lora.yml │ ├── e2e_sft.yml │ ├── model.yml │ ├── ray_test.yml │ ├── sandbox.yml │ ├── sanity.yml │ ├── scorecard.yml │ ├── secrets_scan.yml │ ├── vllm.yml │ └── yapf_format.yml ├── .gitignore ├── .readthedocs.yaml ├── .style.yapf ├── PRM ├── ds_config_zero2.json ├── eval │ ├── README.md │ ├── bon.py │ └── process_bench.py ├── finetune.py ├── train_stage_1.sh └── train_stage_2.sh ├── README.md ├── aime2024.parquet ├── eval ├── LICENSE ├── README.md ├── data │ ├── aime24 │ │ └── test.jsonl │ ├── amc23 │ │ └── test.jsonl │ ├── aqua │ │ └── test.jsonl │ ├── asdiv │ │ └── test.jsonl │ ├── carp_en │ │ ├── demo.json │ │ └── test.jsonl │ ├── cmath │ │ └── test.jsonl │ ├── cn_middle_school │ │ └── test.jsonl │ ├── college_math │ │ └── test.jsonl │ ├── eval_rm_maj_example │ │ └── math_cot_100.jsonl │ ├── gaokao2023en │ │ └── test.jsonl │ ├── gaokao2024_I │ │ └── test.jsonl │ ├── gaokao2024_II │ │ └── test.jsonl │ ├── gaokao2024_mix │ │ └── test.jsonl │ ├── gaokao_math_cloze │ │ └── test.jsonl │ ├── gaokao_math_qa │ │ └── test.jsonl │ ├── gsm8k │ │ ├── test.jsonl │ │ └── train.jsonl │ ├── math │ │ ├── test.jsonl │ │ └── train.jsonl │ ├── math500 │ │ └── test.jsonl │ ├── mawps │ │ ├── addsub.jsonl │ │ ├── multiarith.jsonl │ │ ├── singleeq.jsonl │ │ ├── singleop.jsonl │ │ └── test.jsonl │ ├── minerva_math │ │ ├── README.md │ │ └── test.jsonl │ ├── mmlu_stem │ │ └── test.jsonl │ ├── olympiadbench │ │ ├── test.json │ │ └── test.jsonl │ ├── sat_math │ │ └── test.jsonl │ ├── svamp │ │ └── test.jsonl │ └── tabmwp │ │ └── test.jsonl ├── data_loader.py ├── evaluate.py ├── examples.py ├── grader.py ├── latex2sympy │ ├── .coveragerc │ ├── .gitignore │ ├── LICENSE.txt │ ├── PS.g4 │ ├── README.md │ ├── __init__.py │ ├── antlr-4.11.1-complete.jar │ ├── asciimath_printer.py │ ├── description.txt │ ├── dev-requirements.in │ ├── dev-requirements.txt │ ├── gen │ │ ├── PS.interp │ │ ├── PS.tokens │ │ ├── PSLexer.interp │ │ ├── PSLexer.py │ │ ├── PSLexer.tokens │ │ ├── PSListener.py │ │ ├── PSParser.py │ │ └── __init__.py │ ├── icon.png │ ├── latex2sympy2.py │ ├── requirements.in │ ├── requirements.txt │ ├── sandbox │ │ ├── linalg_equations.py │ │ ├── linalg_span.py │ │ ├── matrix.py │ │ ├── matrix_placeholders.py │ │ ├── sandbox.py │ │ ├── sandbox_equality.py │ │ ├── sectan.py │ │ └── vector.py │ ├── scripts │ │ ├── compile.sh │ │ ├── coverage-ci.sh │ │ ├── coverage.sh │ │ ├── pre-commit │ │ ├── pre-push │ │ ├── publish.sh │ │ ├── setup-hooks.sh │ │ ├── setup.sh │ │ └── test.sh │ ├── setup.cfg │ ├── setup.py │ └── tests │ │ ├── __init__.py │ │ ├── abs_test.py │ │ ├── all_bad_test.py │ │ ├── all_good_test.py │ │ ├── atom_expr_test.py │ │ ├── binomial_test.py │ │ ├── ceil_test.py │ │ ├── complex_test.py │ │ ├── context.py │ │ ├── exp_test.py │ │ ├── floor_test.py │ │ ├── gcd_test.py │ │ ├── greek_test.py │ │ ├── grouping_test.py │ │ ├── lcm_test.py │ │ ├── left_right_cdot_test.py │ │ ├── linalg_test.py │ │ ├── max_test.py │ │ ├── min_test.py │ │ ├── mod_test.py │ │ ├── overline_test.py │ │ ├── pi_test.py │ │ ├── trig_test.py │ │ └── variable_test.py ├── math_eval.py ├── math_utils.py ├── model_utils.py ├── output │ ├── aime24 │ │ └── test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json │ ├── amc23 │ │ └── test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json │ ├── math500 │ │ └── test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json │ ├── minerva_math │ │ └── test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json │ └── olympiadbench │ │ └── test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json ├── parser.py ├── process.py ├── python_executor.py ├── requirements.txt ├── rm_maj_eval.py ├── sh │ ├── collect_results.py │ ├── convert_and_evaluate_gpu.sh │ ├── eval.sh │ ├── eval_single_node.sh │ └── run.sh ├── trajectory.py └── utils.py ├── examples ├── data_preprocess │ ├── full_hh_rlhf.py │ ├── gsm8k.py │ ├── hellaswag.py │ └── math_dataset.py ├── generation │ └── run_deepseek_v2_lite_math.sh ├── grpo_trainer │ ├── run_deepseek7b_llm.sh │ ├── run_deepseek7b_llm_seq_balance.sh │ ├── run_qwen2-7b.sh │ └── run_qwen2-7b_seq_balance.sh ├── ppo_trainer │ ├── run_deepseek7b_llm.sh │ ├── run_deepseek7b_llm_sp2.sh │ ├── run_deepseek_full_hh_rlhf.sh │ ├── run_deepseek_math_gsm8k_megatron.sh │ ├── run_deepseek_megatron.sh │ ├── run_gemma.sh │ ├── run_qwen2-7b.sh │ ├── run_qwen2-7b_math_gsm8k_megatron.sh │ ├── run_qwen2-7b_rm.sh │ ├── run_qwen2-7b_rm_seq_balance.sh │ ├── run_qwen2-7b_seq_balance.sh │ ├── run_qwen2.5-32b.sh │ └── verl_getting_started.ipynb ├── ray │ └── tutorial.ipynb ├── remax_trainer │ ├── run_qwen2.5-3b_seq_balance.sh │ └── run_qwen2.5-7b_seq_balance.sh ├── rloo_trainer │ └── run_qwen2-7b.sh ├── sft │ └── gsm8k │ │ ├── run_deepseek_6b7.sh │ │ ├── run_gemma_2b.sh │ │ ├── run_gemma_7b.sh │ │ ├── run_qwen_05_peft.sh │ │ ├── run_qwen_05_sp2.sh │ │ └── run_qwen_05_sp2_liger.sh ├── slurm │ └── ray_on_slurm.slurm └── split_placement │ ├── README.md │ ├── config │ └── ppo_trainer_split.yaml │ ├── main_ppo_split.py │ ├── run_deepseek7b_llm.sh │ └── split_monkey_patch.py ├── math500.parquet ├── nccl.conf ├── pyproject.toml ├── requirements.txt ├── scripts ├── format.sh └── model_merger.py ├── setup.py ├── tests ├── __init__.py ├── checkpoint │ └── test_fsdp_ckpt.py ├── distro │ └── requirements.py ├── e2e │ ├── __init__.py │ ├── arithmetic_sequence │ │ ├── data │ │ │ ├── create_dataset.py │ │ │ ├── test.parquet │ │ │ └── train.parquet │ │ ├── model │ │ │ ├── config.json │ │ │ ├── create_model_tokenizer.py │ │ │ ├── generation_config.json │ │ │ ├── model.safetensors │ │ │ └── tokenizer_config.json │ │ └── rl │ │ │ ├── README.md │ │ │ └── main_trainer.py │ ├── check_results.py │ ├── envs │ │ ├── __init__.py │ │ └── digit_completion │ │ │ ├── __init__.py │ │ │ ├── task.py │ │ │ └── tokenizer.py │ ├── run_deepseek_megatron.sh │ ├── run_qwen_gsm8k_function_rm.sh │ ├── run_qwen_gsm8k_function_rm_grpo.sh │ ├── run_qwen_gsm8k_function_rm_no_rmpad.sh │ ├── run_qwen_gsm8k_function_rm_remax.sh │ ├── run_qwen_gsm8k_model_rm.sh │ ├── run_qwen_gsm8k_model_rm_liger_kernel.sh │ ├── run_qwen_gsm8k_model_rm_no_rmpad.sh │ ├── run_qwen_gsm8k_model_rm_seq_balance.sh │ ├── run_qwen_gsm8k_model_rm_ulysses.sh │ ├── run_qwen_megatron.sh │ ├── run_ray_trainer.sh │ ├── run_ray_trainer_fire_sampling.sh │ └── run_ray_trainer_rmpad.sh ├── generation │ └── run_gen_qwen05.sh ├── gpu_utility │ ├── test_memory_buffers.py │ ├── test_ops.py │ └── test_torch_functional.py ├── kill_github_tests.sh ├── model │ ├── test_transformer.py │ └── test_transformers_ulysses.py ├── ray │ ├── check_worker_alive │ │ └── main.py │ ├── detached_worker │ │ ├── README.md │ │ ├── client.py │ │ ├── run.sh │ │ └── server.py │ ├── test_check_worker_alive.py │ ├── test_colocated_workers.py │ ├── test_data_transfer.py │ ├── test_driverfunc_to_worker.py │ ├── test_high_level_scheduling_api.py │ ├── test_ray_local_envs.py │ ├── test_rvdz.py │ ├── test_worker_group_basics.py │ └── test_worker_group_torch.py ├── rollout │ ├── run_fsdp_vllm.py │ ├── test_vllm_hf_loader.py │ └── test_vllm_spmd.py ├── sandbox │ └── test_sandbox.py ├── sanity │ ├── check_license.py │ └── test_import.py ├── sft │ ├── run_sft.sh │ ├── run_sft_qwen05_peft.sh │ ├── run_sft_qwen05_sp2_liger.sh │ ├── run_sft_sp_loss_match.sh │ └── test_sp_loss_match.py ├── utility │ └── test_tensor_dict_utilities.py └── verl │ └── utils │ └── dataset │ ├── test_rl_dataset.py │ ├── test_rm_dataset.py │ └── test_sft_dataset.py ├── train.parquet ├── val.parquet └── 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 ├── qwen2 │ ├── __init__.py │ └── megatron │ │ ├── __init__.py │ │ ├── checkpoint_utils │ │ ├── __init__.py │ │ ├── qwen2_loader.py │ │ └── qwen2_saver.py │ │ ├── layers │ │ ├── __init__.py │ │ ├── parallel_attention.py │ │ ├── parallel_decoder.py │ │ ├── parallel_linear.py │ │ ├── parallel_mlp.py │ │ └── parallel_rmsnorm.py │ │ └── modeling_qwen2_megatron.py ├── registry.py ├── transformers │ ├── __init__.py │ ├── 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 ├── third_party ├── __init__.py └── vllm │ ├── __init__.py │ ├── vllm_spmd │ ├── __init__.py │ └── dtensor_weight_loaders.py │ ├── vllm_v_0_3_1 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── tokenizer.py │ ├── weight_loaders.py │ └── worker.py │ ├── vllm_v_0_4_2 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── dtensor_weight_loaders.py │ ├── hf_weight_loader.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── megatron_weight_loaders.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── spmd_gpu_executor.py │ ├── tokenizer.py │ └── worker.py │ ├── vllm_v_0_5_4 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── dtensor_weight_loaders.py │ ├── hf_weight_loader.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── megatron_weight_loaders.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── spmd_gpu_executor.py │ ├── tokenizer.py │ └── worker.py │ └── vllm_v_0_6_3 │ ├── __init__.py │ ├── arg_utils.py │ ├── config.py │ ├── dtensor_weight_loaders.py │ ├── hf_weight_loader.py │ ├── llm.py │ ├── llm_engine_sp.py │ ├── megatron_weight_loaders.py │ ├── model_loader.py │ ├── model_runner.py │ ├── parallel_state.py │ ├── spmd_gpu_executor.py │ ├── tokenizer.py │ └── worker.py ├── trainer ├── __init__.py ├── config │ ├── evaluation.yaml │ ├── generation.yaml │ ├── ppo_megatron_trainer.yaml │ ├── ppo_trainer.yaml │ └── sft_trainer.yaml ├── fsdp_sft_trainer.py ├── main_eval.py ├── main_generation.py ├── main_ppo.py ├── ppo │ ├── __init__.py │ ├── core_algos.py │ └── ray_trainer.py └── runtime_env.yaml ├── utils ├── __init__.py ├── checkpoint │ ├── __init__.py │ ├── checkpoint_manager.py │ └── fsdp_checkpoint_manager.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 │ └── wandb_ws.py ├── logging_utils.py ├── megatron │ ├── __init__.py │ ├── memory.py │ ├── optimizer.py │ ├── pipeline_parallel.py │ ├── sequence_parallel.py │ └── tensor_parallel.py ├── megatron_utils.py ├── memory_buffer.py ├── model.py ├── monitored_metric │ └── __init__.py ├── py_functional.py ├── ray_utils.py ├── rendezvous │ ├── __init__.py │ └── ray_backend.py ├── reward_score │ ├── __init__.py │ ├── gsm8k.py │ ├── math.py │ ├── prime_code │ │ ├── __init__.py │ │ ├── testing_util.py │ │ └── utils.py │ └── prime_math │ │ ├── __init__.py │ │ ├── grader.py │ │ └── math_normalize.py ├── seqlen_balancing.py ├── tokenizer.py ├── torch_dtypes.py ├── torch_functional.py ├── tracking.py └── ulysses.py ├── version └── version └── workers ├── __init__.py ├── actor ├── __init__.py ├── base.py ├── dp_actor.py └── megatron_actor.py ├── critic ├── __init__.py ├── base.py ├── dp_critic.py └── megatron_critic.py ├── curriculum_learning ├── __init__.py ├── base.py ├── interquartile.py └── remove_extreme.py ├── fsdp_workers.py ├── megatron_workers.py ├── reward_manager ├── __init__.py ├── blank.py ├── naive.py └── prime.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 │ ├── fire_vllm_rollout.py │ ├── vllm_rollout.py │ └── vllm_rollout_spmd.py └── sharding_manager ├── __init__.py ├── base.py ├── fsdp_ulysses.py ├── fsdp_vllm.py └── megatron_vllm.py /.github/workflows/dataset.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/dataset.yml -------------------------------------------------------------------------------- /.github/workflows/e2e_digit_completion.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/e2e_digit_completion.yml -------------------------------------------------------------------------------- /.github/workflows/e2e_digit_completion_fire.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/e2e_digit_completion_fire.yml -------------------------------------------------------------------------------- /.github/workflows/e2e_gsm8k.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/e2e_gsm8k.yml -------------------------------------------------------------------------------- /.github/workflows/e2e_gsm8k_megatron.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/e2e_gsm8k_megatron.yml -------------------------------------------------------------------------------- /.github/workflows/e2e_lora.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/e2e_lora.yml -------------------------------------------------------------------------------- /.github/workflows/e2e_sft.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/e2e_sft.yml -------------------------------------------------------------------------------- /.github/workflows/model.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/model.yml -------------------------------------------------------------------------------- /.github/workflows/ray_test.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/ray_test.yml -------------------------------------------------------------------------------- /.github/workflows/sandbox.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/sandbox.yml -------------------------------------------------------------------------------- /.github/workflows/sanity.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/sanity.yml -------------------------------------------------------------------------------- /.github/workflows/scorecard.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/scorecard.yml -------------------------------------------------------------------------------- /.github/workflows/secrets_scan.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/secrets_scan.yml -------------------------------------------------------------------------------- /.github/workflows/vllm.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/vllm.yml -------------------------------------------------------------------------------- /.github/workflows/yapf_format.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.github/workflows/yapf_format.yml -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.gitignore -------------------------------------------------------------------------------- /.readthedocs.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.readthedocs.yaml -------------------------------------------------------------------------------- /.style.yapf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/.style.yapf -------------------------------------------------------------------------------- /PRM/ds_config_zero2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/PRM/ds_config_zero2.json -------------------------------------------------------------------------------- /PRM/eval/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/PRM/eval/README.md -------------------------------------------------------------------------------- /PRM/eval/bon.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/PRM/eval/bon.py -------------------------------------------------------------------------------- /PRM/eval/process_bench.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/PRM/eval/process_bench.py -------------------------------------------------------------------------------- /PRM/finetune.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/PRM/finetune.py -------------------------------------------------------------------------------- /PRM/train_stage_1.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/PRM/train_stage_1.sh -------------------------------------------------------------------------------- /PRM/train_stage_2.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/PRM/train_stage_2.sh -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/README.md -------------------------------------------------------------------------------- /aime2024.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/aime2024.parquet -------------------------------------------------------------------------------- /eval/LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/LICENSE -------------------------------------------------------------------------------- /eval/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/README.md -------------------------------------------------------------------------------- /eval/data/aime24/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/aime24/test.jsonl -------------------------------------------------------------------------------- /eval/data/amc23/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/amc23/test.jsonl -------------------------------------------------------------------------------- /eval/data/aqua/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/aqua/test.jsonl -------------------------------------------------------------------------------- /eval/data/asdiv/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/asdiv/test.jsonl -------------------------------------------------------------------------------- /eval/data/carp_en/demo.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/carp_en/demo.json -------------------------------------------------------------------------------- /eval/data/carp_en/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/carp_en/test.jsonl -------------------------------------------------------------------------------- /eval/data/cmath/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/cmath/test.jsonl -------------------------------------------------------------------------------- /eval/data/cn_middle_school/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/cn_middle_school/test.jsonl -------------------------------------------------------------------------------- /eval/data/college_math/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/college_math/test.jsonl -------------------------------------------------------------------------------- /eval/data/eval_rm_maj_example/math_cot_100.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/eval_rm_maj_example/math_cot_100.jsonl -------------------------------------------------------------------------------- /eval/data/gaokao2023en/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gaokao2023en/test.jsonl -------------------------------------------------------------------------------- /eval/data/gaokao2024_I/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gaokao2024_I/test.jsonl -------------------------------------------------------------------------------- /eval/data/gaokao2024_II/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gaokao2024_II/test.jsonl -------------------------------------------------------------------------------- /eval/data/gaokao2024_mix/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gaokao2024_mix/test.jsonl -------------------------------------------------------------------------------- /eval/data/gaokao_math_cloze/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gaokao_math_cloze/test.jsonl -------------------------------------------------------------------------------- /eval/data/gaokao_math_qa/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gaokao_math_qa/test.jsonl -------------------------------------------------------------------------------- /eval/data/gsm8k/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gsm8k/test.jsonl -------------------------------------------------------------------------------- /eval/data/gsm8k/train.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/gsm8k/train.jsonl -------------------------------------------------------------------------------- /eval/data/math/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/math/test.jsonl -------------------------------------------------------------------------------- /eval/data/math/train.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/math/train.jsonl -------------------------------------------------------------------------------- /eval/data/math500/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/math500/test.jsonl -------------------------------------------------------------------------------- /eval/data/mawps/addsub.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/mawps/addsub.jsonl -------------------------------------------------------------------------------- /eval/data/mawps/multiarith.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/mawps/multiarith.jsonl -------------------------------------------------------------------------------- /eval/data/mawps/singleeq.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/mawps/singleeq.jsonl -------------------------------------------------------------------------------- /eval/data/mawps/singleop.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/mawps/singleop.jsonl -------------------------------------------------------------------------------- /eval/data/mawps/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/mawps/test.jsonl -------------------------------------------------------------------------------- /eval/data/minerva_math/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/minerva_math/README.md -------------------------------------------------------------------------------- /eval/data/minerva_math/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/minerva_math/test.jsonl -------------------------------------------------------------------------------- /eval/data/mmlu_stem/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/mmlu_stem/test.jsonl -------------------------------------------------------------------------------- /eval/data/olympiadbench/test.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/olympiadbench/test.json -------------------------------------------------------------------------------- /eval/data/olympiadbench/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/olympiadbench/test.jsonl -------------------------------------------------------------------------------- /eval/data/sat_math/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/sat_math/test.jsonl -------------------------------------------------------------------------------- /eval/data/svamp/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/svamp/test.jsonl -------------------------------------------------------------------------------- /eval/data/tabmwp/test.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data/tabmwp/test.jsonl -------------------------------------------------------------------------------- /eval/data_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/data_loader.py -------------------------------------------------------------------------------- /eval/evaluate.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/evaluate.py -------------------------------------------------------------------------------- /eval/examples.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/examples.py -------------------------------------------------------------------------------- /eval/grader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/grader.py -------------------------------------------------------------------------------- /eval/latex2sympy/.coveragerc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/.coveragerc -------------------------------------------------------------------------------- /eval/latex2sympy/.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/.gitignore -------------------------------------------------------------------------------- /eval/latex2sympy/LICENSE.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/LICENSE.txt -------------------------------------------------------------------------------- /eval/latex2sympy/PS.g4: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/PS.g4 -------------------------------------------------------------------------------- /eval/latex2sympy/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/README.md -------------------------------------------------------------------------------- /eval/latex2sympy/__init__.py: -------------------------------------------------------------------------------- 1 | import latex2sympy -------------------------------------------------------------------------------- /eval/latex2sympy/antlr-4.11.1-complete.jar: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/antlr-4.11.1-complete.jar -------------------------------------------------------------------------------- /eval/latex2sympy/asciimath_printer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/asciimath_printer.py -------------------------------------------------------------------------------- /eval/latex2sympy/description.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/description.txt -------------------------------------------------------------------------------- /eval/latex2sympy/dev-requirements.in: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/dev-requirements.in -------------------------------------------------------------------------------- /eval/latex2sympy/dev-requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/dev-requirements.txt -------------------------------------------------------------------------------- /eval/latex2sympy/gen/PS.interp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/gen/PS.interp -------------------------------------------------------------------------------- /eval/latex2sympy/gen/PS.tokens: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/gen/PS.tokens -------------------------------------------------------------------------------- /eval/latex2sympy/gen/PSLexer.interp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/gen/PSLexer.interp -------------------------------------------------------------------------------- /eval/latex2sympy/gen/PSLexer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/gen/PSLexer.py -------------------------------------------------------------------------------- /eval/latex2sympy/gen/PSLexer.tokens: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/gen/PSLexer.tokens -------------------------------------------------------------------------------- /eval/latex2sympy/gen/PSListener.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/gen/PSListener.py -------------------------------------------------------------------------------- /eval/latex2sympy/gen/PSParser.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/gen/PSParser.py -------------------------------------------------------------------------------- /eval/latex2sympy/gen/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /eval/latex2sympy/icon.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/icon.png -------------------------------------------------------------------------------- /eval/latex2sympy/latex2sympy2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/latex2sympy2.py -------------------------------------------------------------------------------- /eval/latex2sympy/requirements.in: -------------------------------------------------------------------------------- 1 | sympy 2 | antlr4-python3-runtime 3 | -------------------------------------------------------------------------------- /eval/latex2sympy/requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/requirements.txt -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/linalg_equations.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/linalg_equations.py -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/linalg_span.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/linalg_span.py -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/matrix.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/matrix.py -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/matrix_placeholders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/matrix_placeholders.py -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/sandbox.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/sandbox.py -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/sandbox_equality.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/sandbox_equality.py -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/sectan.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/sectan.py -------------------------------------------------------------------------------- /eval/latex2sympy/sandbox/vector.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/sandbox/vector.py -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/compile.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/compile.sh -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/coverage-ci.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/coverage-ci.sh -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/coverage.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/coverage.sh -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/pre-commit: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/pre-commit -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/pre-push: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/pre-push -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/publish.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/publish.sh -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/setup-hooks.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/setup-hooks.sh -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/setup.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/setup.sh -------------------------------------------------------------------------------- /eval/latex2sympy/scripts/test.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/scripts/test.sh -------------------------------------------------------------------------------- /eval/latex2sympy/setup.cfg: -------------------------------------------------------------------------------- 1 | [pycodestyle] 2 | max-line-length = 120 3 | ignore = E501 4 | -------------------------------------------------------------------------------- /eval/latex2sympy/setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/setup.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /eval/latex2sympy/tests/abs_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/abs_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/all_bad_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/all_bad_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/all_good_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/all_good_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/atom_expr_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/atom_expr_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/binomial_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/binomial_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/ceil_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/ceil_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/complex_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/complex_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/context.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/context.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/exp_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/exp_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/floor_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/floor_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/gcd_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/gcd_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/greek_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/greek_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/grouping_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/grouping_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/lcm_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/lcm_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/left_right_cdot_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/left_right_cdot_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/linalg_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/linalg_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/max_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/max_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/min_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/min_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/mod_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/mod_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/overline_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/overline_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/pi_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/pi_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/trig_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/trig_test.py -------------------------------------------------------------------------------- /eval/latex2sympy/tests/variable_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/latex2sympy/tests/variable_test.py -------------------------------------------------------------------------------- /eval/math_eval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/math_eval.py -------------------------------------------------------------------------------- /eval/math_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/math_utils.py -------------------------------------------------------------------------------- /eval/model_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/model_utils.py -------------------------------------------------------------------------------- /eval/output/aime24/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/output/aime24/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json -------------------------------------------------------------------------------- /eval/output/amc23/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/output/amc23/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json -------------------------------------------------------------------------------- /eval/output/math500/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/output/math500/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json -------------------------------------------------------------------------------- /eval/output/minerva_math/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/output/minerva_math/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json -------------------------------------------------------------------------------- /eval/output/olympiadbench/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/output/olympiadbench/test_qwen25-math-cot-ft_-1_seed42_t0.0_s0_e-1_metrics.json -------------------------------------------------------------------------------- /eval/parser.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/parser.py -------------------------------------------------------------------------------- /eval/process.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/process.py -------------------------------------------------------------------------------- /eval/python_executor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/python_executor.py -------------------------------------------------------------------------------- /eval/requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/requirements.txt -------------------------------------------------------------------------------- /eval/rm_maj_eval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/rm_maj_eval.py -------------------------------------------------------------------------------- /eval/sh/collect_results.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/sh/collect_results.py -------------------------------------------------------------------------------- /eval/sh/convert_and_evaluate_gpu.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/sh/convert_and_evaluate_gpu.sh -------------------------------------------------------------------------------- /eval/sh/eval.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/sh/eval.sh -------------------------------------------------------------------------------- /eval/sh/eval_single_node.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/sh/eval_single_node.sh -------------------------------------------------------------------------------- /eval/sh/run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/sh/run.sh -------------------------------------------------------------------------------- /eval/trajectory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/trajectory.py -------------------------------------------------------------------------------- /eval/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/eval/utils.py -------------------------------------------------------------------------------- /examples/data_preprocess/full_hh_rlhf.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/data_preprocess/full_hh_rlhf.py -------------------------------------------------------------------------------- /examples/data_preprocess/gsm8k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/data_preprocess/gsm8k.py -------------------------------------------------------------------------------- /examples/data_preprocess/hellaswag.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/data_preprocess/hellaswag.py -------------------------------------------------------------------------------- /examples/data_preprocess/math_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/data_preprocess/math_dataset.py -------------------------------------------------------------------------------- /examples/generation/run_deepseek_v2_lite_math.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/generation/run_deepseek_v2_lite_math.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_deepseek7b_llm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/grpo_trainer/run_deepseek7b_llm.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_deepseek7b_llm_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/grpo_trainer/run_deepseek7b_llm_seq_balance.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_qwen2-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/grpo_trainer/run_qwen2-7b.sh -------------------------------------------------------------------------------- /examples/grpo_trainer/run_qwen2-7b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/grpo_trainer/run_qwen2-7b_seq_balance.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek7b_llm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_deepseek7b_llm.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek7b_llm_sp2.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_deepseek7b_llm_sp2.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek_full_hh_rlhf.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_deepseek_full_hh_rlhf.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek_math_gsm8k_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_deepseek_math_gsm8k_megatron.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_deepseek_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_deepseek_megatron.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_gemma.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_gemma.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_qwen2-7b.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_math_gsm8k_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_qwen2-7b_math_gsm8k_megatron.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_qwen2-7b_rm.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_rm_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_qwen2-7b_rm_seq_balance.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2-7b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_qwen2-7b_seq_balance.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/run_qwen2.5-32b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/run_qwen2.5-32b.sh -------------------------------------------------------------------------------- /examples/ppo_trainer/verl_getting_started.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ppo_trainer/verl_getting_started.ipynb -------------------------------------------------------------------------------- /examples/ray/tutorial.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/ray/tutorial.ipynb -------------------------------------------------------------------------------- /examples/remax_trainer/run_qwen2.5-3b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/remax_trainer/run_qwen2.5-3b_seq_balance.sh -------------------------------------------------------------------------------- /examples/remax_trainer/run_qwen2.5-7b_seq_balance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/remax_trainer/run_qwen2.5-7b_seq_balance.sh -------------------------------------------------------------------------------- /examples/rloo_trainer/run_qwen2-7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/rloo_trainer/run_qwen2-7b.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_deepseek_6b7.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/sft/gsm8k/run_deepseek_6b7.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_gemma_2b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/sft/gsm8k/run_gemma_2b.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_gemma_7b.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/sft/gsm8k/run_gemma_7b.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_qwen_05_peft.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/sft/gsm8k/run_qwen_05_peft.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_qwen_05_sp2.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/sft/gsm8k/run_qwen_05_sp2.sh -------------------------------------------------------------------------------- /examples/sft/gsm8k/run_qwen_05_sp2_liger.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/sft/gsm8k/run_qwen_05_sp2_liger.sh -------------------------------------------------------------------------------- /examples/slurm/ray_on_slurm.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/slurm/ray_on_slurm.slurm -------------------------------------------------------------------------------- /examples/split_placement/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/split_placement/README.md -------------------------------------------------------------------------------- /examples/split_placement/config/ppo_trainer_split.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/split_placement/config/ppo_trainer_split.yaml -------------------------------------------------------------------------------- /examples/split_placement/main_ppo_split.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/split_placement/main_ppo_split.py -------------------------------------------------------------------------------- /examples/split_placement/run_deepseek7b_llm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/split_placement/run_deepseek7b_llm.sh -------------------------------------------------------------------------------- /examples/split_placement/split_monkey_patch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/examples/split_placement/split_monkey_patch.py -------------------------------------------------------------------------------- /math500.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/math500.parquet -------------------------------------------------------------------------------- /nccl.conf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/nccl.conf -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/pyproject.toml -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/requirements.txt -------------------------------------------------------------------------------- /scripts/format.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/scripts/format.sh -------------------------------------------------------------------------------- /scripts/model_merger.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/scripts/model_merger.py -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/setup.py -------------------------------------------------------------------------------- /tests/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/__init__.py -------------------------------------------------------------------------------- /tests/checkpoint/test_fsdp_ckpt.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/checkpoint/test_fsdp_ckpt.py -------------------------------------------------------------------------------- /tests/distro/requirements.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/distro/requirements.py -------------------------------------------------------------------------------- /tests/e2e/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/__init__.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/create_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/data/create_dataset.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/test.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/data/test.parquet -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/data/train.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/data/train.parquet -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/model/config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/create_model_tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/model/create_model_tokenizer.py -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/generation_config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/model/generation_config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/model.safetensors: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/model/model.safetensors -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/model/tokenizer_config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/model/tokenizer_config.json -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/rl/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/rl/README.md -------------------------------------------------------------------------------- /tests/e2e/arithmetic_sequence/rl/main_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/arithmetic_sequence/rl/main_trainer.py -------------------------------------------------------------------------------- /tests/e2e/check_results.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/check_results.py -------------------------------------------------------------------------------- /tests/e2e/envs/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/envs/__init__.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/envs/digit_completion/__init__.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/task.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/envs/digit_completion/task.py -------------------------------------------------------------------------------- /tests/e2e/envs/digit_completion/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/envs/digit_completion/tokenizer.py -------------------------------------------------------------------------------- /tests/e2e/run_deepseek_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_deepseek_megatron.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_function_rm.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm_grpo.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_function_rm_grpo.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm_no_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_function_rm_no_rmpad.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_function_rm_remax.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_function_rm_remax.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_model_rm.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_liger_kernel.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_liger_kernel.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_no_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/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/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_seq_balance.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_gsm8k_model_rm_ulysses.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_gsm8k_model_rm_ulysses.sh -------------------------------------------------------------------------------- /tests/e2e/run_qwen_megatron.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_qwen_megatron.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_ray_trainer.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer_fire_sampling.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_ray_trainer_fire_sampling.sh -------------------------------------------------------------------------------- /tests/e2e/run_ray_trainer_rmpad.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/e2e/run_ray_trainer_rmpad.sh -------------------------------------------------------------------------------- /tests/generation/run_gen_qwen05.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/generation/run_gen_qwen05.sh -------------------------------------------------------------------------------- /tests/gpu_utility/test_memory_buffers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/gpu_utility/test_memory_buffers.py -------------------------------------------------------------------------------- /tests/gpu_utility/test_ops.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/gpu_utility/test_ops.py -------------------------------------------------------------------------------- /tests/gpu_utility/test_torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/gpu_utility/test_torch_functional.py -------------------------------------------------------------------------------- /tests/kill_github_tests.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/kill_github_tests.sh -------------------------------------------------------------------------------- /tests/model/test_transformer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/model/test_transformer.py -------------------------------------------------------------------------------- /tests/model/test_transformers_ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/model/test_transformers_ulysses.py -------------------------------------------------------------------------------- /tests/ray/check_worker_alive/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/check_worker_alive/main.py -------------------------------------------------------------------------------- /tests/ray/detached_worker/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/detached_worker/README.md -------------------------------------------------------------------------------- /tests/ray/detached_worker/client.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/detached_worker/client.py -------------------------------------------------------------------------------- /tests/ray/detached_worker/run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/detached_worker/run.sh -------------------------------------------------------------------------------- /tests/ray/detached_worker/server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/detached_worker/server.py -------------------------------------------------------------------------------- /tests/ray/test_check_worker_alive.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_check_worker_alive.py -------------------------------------------------------------------------------- /tests/ray/test_colocated_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_colocated_workers.py -------------------------------------------------------------------------------- /tests/ray/test_data_transfer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_data_transfer.py -------------------------------------------------------------------------------- /tests/ray/test_driverfunc_to_worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_driverfunc_to_worker.py -------------------------------------------------------------------------------- /tests/ray/test_high_level_scheduling_api.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_high_level_scheduling_api.py -------------------------------------------------------------------------------- /tests/ray/test_ray_local_envs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_ray_local_envs.py -------------------------------------------------------------------------------- /tests/ray/test_rvdz.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_rvdz.py -------------------------------------------------------------------------------- /tests/ray/test_worker_group_basics.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_worker_group_basics.py -------------------------------------------------------------------------------- /tests/ray/test_worker_group_torch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/ray/test_worker_group_torch.py -------------------------------------------------------------------------------- /tests/rollout/run_fsdp_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/rollout/run_fsdp_vllm.py -------------------------------------------------------------------------------- /tests/rollout/test_vllm_hf_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/rollout/test_vllm_hf_loader.py -------------------------------------------------------------------------------- /tests/rollout/test_vllm_spmd.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/rollout/test_vllm_spmd.py -------------------------------------------------------------------------------- /tests/sandbox/test_sandbox.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sandbox/test_sandbox.py -------------------------------------------------------------------------------- /tests/sanity/check_license.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sanity/check_license.py -------------------------------------------------------------------------------- /tests/sanity/test_import.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sanity/test_import.py -------------------------------------------------------------------------------- /tests/sft/run_sft.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sft/run_sft.sh -------------------------------------------------------------------------------- /tests/sft/run_sft_qwen05_peft.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sft/run_sft_qwen05_peft.sh -------------------------------------------------------------------------------- /tests/sft/run_sft_qwen05_sp2_liger.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sft/run_sft_qwen05_sp2_liger.sh -------------------------------------------------------------------------------- /tests/sft/run_sft_sp_loss_match.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sft/run_sft_sp_loss_match.sh -------------------------------------------------------------------------------- /tests/sft/test_sp_loss_match.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/sft/test_sp_loss_match.py -------------------------------------------------------------------------------- /tests/utility/test_tensor_dict_utilities.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/utility/test_tensor_dict_utilities.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_rl_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/verl/utils/dataset/test_rl_dataset.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_rm_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/verl/utils/dataset/test_rm_dataset.py -------------------------------------------------------------------------------- /tests/verl/utils/dataset/test_sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/tests/verl/utils/dataset/test_sft_dataset.py -------------------------------------------------------------------------------- /train.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/train.parquet -------------------------------------------------------------------------------- /val.parquet: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/val.parquet -------------------------------------------------------------------------------- /verl/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/__init__.py -------------------------------------------------------------------------------- /verl/models/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/README.md -------------------------------------------------------------------------------- /verl/models/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/checkpoint_utils/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/llama_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/checkpoint_utils/llama_loader.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/checkpoint_utils/llama_saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/checkpoint_utils/llama_saver.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/layers/__init__.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_attention.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/layers/parallel_attention.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_decoder.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/layers/parallel_decoder.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/layers/parallel_linear.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_mlp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/layers/parallel_mlp.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/layers/parallel_rmsnorm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/layers/parallel_rmsnorm.py -------------------------------------------------------------------------------- /verl/models/llama/megatron/modeling_llama_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/llama/megatron/modeling_llama_megatron.py -------------------------------------------------------------------------------- /verl/models/qwen2/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/checkpoint_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/checkpoint_utils/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/checkpoint_utils/qwen2_saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/checkpoint_utils/qwen2_saver.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/layers/__init__.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_attention.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/layers/parallel_attention.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_decoder.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/layers/parallel_decoder.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/layers/parallel_linear.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_mlp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/layers/parallel_mlp.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/layers/parallel_rmsnorm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/layers/parallel_rmsnorm.py -------------------------------------------------------------------------------- /verl/models/qwen2/megatron/modeling_qwen2_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/qwen2/megatron/modeling_qwen2_megatron.py -------------------------------------------------------------------------------- /verl/models/registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/registry.py -------------------------------------------------------------------------------- /verl/models/transformers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/transformers/__init__.py -------------------------------------------------------------------------------- /verl/models/transformers/llama.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/transformers/llama.py -------------------------------------------------------------------------------- /verl/models/transformers/monkey_patch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/transformers/monkey_patch.py -------------------------------------------------------------------------------- /verl/models/transformers/qwen2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/transformers/qwen2.py -------------------------------------------------------------------------------- /verl/models/weight_loader_registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/models/weight_loader_registry.py -------------------------------------------------------------------------------- /verl/protocol.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/protocol.py -------------------------------------------------------------------------------- /verl/single_controller/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/decorator.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/decorator.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/megatron/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/megatron/worker.py -------------------------------------------------------------------------------- /verl/single_controller/base/megatron/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/megatron/worker_group.py -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/register_center/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/base/register_center/ray.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/register_center/ray.py -------------------------------------------------------------------------------- /verl/single_controller/base/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/worker.py -------------------------------------------------------------------------------- /verl/single_controller/base/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/base/worker_group.py -------------------------------------------------------------------------------- /verl/single_controller/ray/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/ray/__init__.py -------------------------------------------------------------------------------- /verl/single_controller/ray/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/ray/base.py -------------------------------------------------------------------------------- /verl/single_controller/ray/megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/single_controller/ray/megatron.py -------------------------------------------------------------------------------- /verl/third_party/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/third_party/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/third_party/vllm/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_spmd/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/third_party/vllm/vllm_spmd/__init__.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_spmd/dtensor_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/third_party/vllm/vllm_spmd/dtensor_weight_loaders.py -------------------------------------------------------------------------------- /verl/third_party/vllm/vllm_v_0_3_1/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/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/CJReinforce/PURE/HEAD/verl/third_party/vllm/vllm_v_0_6_3/worker.py -------------------------------------------------------------------------------- /verl/trainer/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/__init__.py -------------------------------------------------------------------------------- /verl/trainer/config/evaluation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/config/evaluation.yaml -------------------------------------------------------------------------------- /verl/trainer/config/generation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/config/generation.yaml -------------------------------------------------------------------------------- /verl/trainer/config/ppo_megatron_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/config/ppo_megatron_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/config/ppo_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/config/ppo_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/config/sft_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/config/sft_trainer.yaml -------------------------------------------------------------------------------- /verl/trainer/fsdp_sft_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/fsdp_sft_trainer.py -------------------------------------------------------------------------------- /verl/trainer/main_eval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/main_eval.py -------------------------------------------------------------------------------- /verl/trainer/main_generation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/main_generation.py -------------------------------------------------------------------------------- /verl/trainer/main_ppo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/main_ppo.py -------------------------------------------------------------------------------- /verl/trainer/ppo/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/ppo/__init__.py -------------------------------------------------------------------------------- /verl/trainer/ppo/core_algos.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/ppo/core_algos.py -------------------------------------------------------------------------------- /verl/trainer/ppo/ray_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/ppo/ray_trainer.py -------------------------------------------------------------------------------- /verl/trainer/runtime_env.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/trainer/runtime_env.yaml -------------------------------------------------------------------------------- /verl/utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/__init__.py -------------------------------------------------------------------------------- /verl/utils/checkpoint/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/checkpoint/__init__.py -------------------------------------------------------------------------------- /verl/utils/checkpoint/checkpoint_manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/checkpoint/checkpoint_manager.py -------------------------------------------------------------------------------- /verl/utils/checkpoint/fsdp_checkpoint_manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/checkpoint/fsdp_checkpoint_manager.py -------------------------------------------------------------------------------- /verl/utils/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/config.py -------------------------------------------------------------------------------- /verl/utils/dataset/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/dataset/README.md -------------------------------------------------------------------------------- /verl/utils/dataset/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/dataset/__init__.py -------------------------------------------------------------------------------- /verl/utils/dataset/rl_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/dataset/rl_dataset.py -------------------------------------------------------------------------------- /verl/utils/dataset/rm_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/dataset/rm_dataset.py -------------------------------------------------------------------------------- /verl/utils/dataset/sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/dataset/sft_dataset.py -------------------------------------------------------------------------------- /verl/utils/debug/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/debug/__init__.py -------------------------------------------------------------------------------- /verl/utils/debug/performance.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/debug/performance.py -------------------------------------------------------------------------------- /verl/utils/debug/trajectory_tracker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/debug/trajectory_tracker.py -------------------------------------------------------------------------------- /verl/utils/distributed.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/distributed.py -------------------------------------------------------------------------------- /verl/utils/flops_counter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/flops_counter.py -------------------------------------------------------------------------------- /verl/utils/fs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/fs.py -------------------------------------------------------------------------------- /verl/utils/fsdp_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/fsdp_utils.py -------------------------------------------------------------------------------- /verl/utils/hdfs_io.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/hdfs_io.py -------------------------------------------------------------------------------- /verl/utils/import_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/import_utils.py -------------------------------------------------------------------------------- /verl/utils/logger/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/logger/__init__.py -------------------------------------------------------------------------------- /verl/utils/logger/aggregate_logger.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/logger/aggregate_logger.py -------------------------------------------------------------------------------- /verl/utils/logger/wandb_ws.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/logger/wandb_ws.py -------------------------------------------------------------------------------- /verl/utils/logging_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/logging_utils.py -------------------------------------------------------------------------------- /verl/utils/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/megatron/__init__.py -------------------------------------------------------------------------------- /verl/utils/megatron/memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/megatron/memory.py -------------------------------------------------------------------------------- /verl/utils/megatron/optimizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/megatron/optimizer.py -------------------------------------------------------------------------------- /verl/utils/megatron/pipeline_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/megatron/pipeline_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron/sequence_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/megatron/sequence_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron/tensor_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/megatron/tensor_parallel.py -------------------------------------------------------------------------------- /verl/utils/megatron_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/megatron_utils.py -------------------------------------------------------------------------------- /verl/utils/memory_buffer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/memory_buffer.py -------------------------------------------------------------------------------- /verl/utils/model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/model.py -------------------------------------------------------------------------------- /verl/utils/monitored_metric/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/monitored_metric/__init__.py -------------------------------------------------------------------------------- /verl/utils/py_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/py_functional.py -------------------------------------------------------------------------------- /verl/utils/ray_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/ray_utils.py -------------------------------------------------------------------------------- /verl/utils/rendezvous/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/rendezvous/__init__.py -------------------------------------------------------------------------------- /verl/utils/rendezvous/ray_backend.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/rendezvous/ray_backend.py -------------------------------------------------------------------------------- /verl/utils/reward_score/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/gsm8k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/gsm8k.py -------------------------------------------------------------------------------- /verl/utils/reward_score/math.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/math.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_code/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/prime_code/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_code/testing_util.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/prime_code/testing_util.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_code/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/prime_code/utils.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_math/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/prime_math/__init__.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_math/grader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/prime_math/grader.py -------------------------------------------------------------------------------- /verl/utils/reward_score/prime_math/math_normalize.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/reward_score/prime_math/math_normalize.py -------------------------------------------------------------------------------- /verl/utils/seqlen_balancing.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/seqlen_balancing.py -------------------------------------------------------------------------------- /verl/utils/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/tokenizer.py -------------------------------------------------------------------------------- /verl/utils/torch_dtypes.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/torch_dtypes.py -------------------------------------------------------------------------------- /verl/utils/torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/torch_functional.py -------------------------------------------------------------------------------- /verl/utils/tracking.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/tracking.py -------------------------------------------------------------------------------- /verl/utils/ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/utils/ulysses.py -------------------------------------------------------------------------------- /verl/version/version: -------------------------------------------------------------------------------- 1 | 0.2.0.dev 2 | -------------------------------------------------------------------------------- /verl/workers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/__init__.py -------------------------------------------------------------------------------- /verl/workers/actor/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/actor/__init__.py -------------------------------------------------------------------------------- /verl/workers/actor/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/actor/base.py -------------------------------------------------------------------------------- /verl/workers/actor/dp_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/actor/dp_actor.py -------------------------------------------------------------------------------- /verl/workers/actor/megatron_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/actor/megatron_actor.py -------------------------------------------------------------------------------- /verl/workers/critic/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/critic/__init__.py -------------------------------------------------------------------------------- /verl/workers/critic/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/critic/base.py -------------------------------------------------------------------------------- /verl/workers/critic/dp_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/critic/dp_critic.py -------------------------------------------------------------------------------- /verl/workers/critic/megatron_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/critic/megatron_critic.py -------------------------------------------------------------------------------- /verl/workers/curriculum_learning/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/curriculum_learning/__init__.py -------------------------------------------------------------------------------- /verl/workers/curriculum_learning/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/curriculum_learning/base.py -------------------------------------------------------------------------------- /verl/workers/curriculum_learning/interquartile.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/curriculum_learning/interquartile.py -------------------------------------------------------------------------------- /verl/workers/curriculum_learning/remove_extreme.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/curriculum_learning/remove_extreme.py -------------------------------------------------------------------------------- /verl/workers/fsdp_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/fsdp_workers.py -------------------------------------------------------------------------------- /verl/workers/megatron_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/megatron_workers.py -------------------------------------------------------------------------------- /verl/workers/reward_manager/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_manager/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_manager/blank.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_manager/blank.py -------------------------------------------------------------------------------- /verl/workers/reward_manager/naive.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_manager/naive.py -------------------------------------------------------------------------------- /verl/workers/reward_manager/prime.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_manager/prime.py -------------------------------------------------------------------------------- /verl/workers/reward_model/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_model/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_model/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_model/base.py -------------------------------------------------------------------------------- /verl/workers/reward_model/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_model/megatron/__init__.py -------------------------------------------------------------------------------- /verl/workers/reward_model/megatron/reward_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/reward_model/megatron/reward_model.py -------------------------------------------------------------------------------- /verl/workers/rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/base.py -------------------------------------------------------------------------------- /verl/workers/rollout/hf_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/hf_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/naive/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/naive/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/naive/naive_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/naive/naive_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/tokenizer.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/vllm_rollout/__init__.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/fire_vllm_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/vllm_rollout/fire_vllm_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/vllm_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/vllm_rollout/vllm_rollout.py -------------------------------------------------------------------------------- /verl/workers/rollout/vllm_rollout/vllm_rollout_spmd.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/rollout/vllm_rollout/vllm_rollout_spmd.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/sharding_manager/__init__.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/sharding_manager/base.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/fsdp_ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/sharding_manager/fsdp_ulysses.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/fsdp_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/sharding_manager/fsdp_vllm.py -------------------------------------------------------------------------------- /verl/workers/sharding_manager/megatron_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CJReinforce/PURE/HEAD/verl/workers/sharding_manager/megatron_vllm.py --------------------------------------------------------------------------------