├── .gitignore ├── Dockerfile ├── LICENSE ├── README.md ├── configs └── ppo_trainer.yaml ├── datasets └── README.md ├── images ├── Qwen2.5-3B.png ├── Qwen2.5-7B.png └── pacs.png ├── logs └── README.md ├── outputs └── README.md ├── scripts ├── baseline │ ├── grpo.sh │ └── ppo.sh ├── inference.sh ├── model_merge.sh ├── pacs │ ├── pacs_beta_1.0.sh │ └── pacs_beta_2.0.sh └── preprocess_dataset.sh └── src ├── .DS_Store ├── inference.py ├── json_file.py ├── model_merger.py ├── pacs ├── pacs_actor.py ├── pacs_compute_metrics.py ├── pacs_core_algos.py ├── pacs_main.py ├── pacs_trainer.py └── pacs_workers.py ├── preprocess_dataset.py ├── reward_function.py └── verl ├── __init__.py ├── models ├── README.md ├── __init__.py ├── llama │ ├── __init__.py │ └── megatron │ │ ├── __init__.py │ │ ├── checkpoint_utils │ │ ├── __init__.py │ │ ├── llama_loader.py │ │ ├── llama_loader_depracated.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 ├── mcore │ ├── __init__.py │ ├── config_converter.py │ ├── loader.py │ ├── model_forward.py │ ├── model_initializer.py │ ├── readme.md │ ├── registry.py │ ├── saver.py │ ├── util.py │ └── weight_converter.py ├── qwen2 │ ├── __init__.py │ └── megatron │ │ ├── __init__.py │ │ ├── checkpoint_utils │ │ ├── __init__.py │ │ ├── qwen2_loader.py │ │ ├── qwen2_loader_depracated.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 │ ├── qwen2_5_vl.py │ └── qwen2_vl.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 ├── sglang │ ├── __init__.py │ └── parallel_state.py └── vllm │ ├── __init__.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 ├── tools ├── __init__.py ├── base_tool.py ├── gsm8k_tool.py └── schemas.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 │ ├── metric_utils.py │ ├── ray_trainer.py │ └── reward.py └── runtime_env.yaml ├── utils ├── __init__.py ├── activation_offload.py ├── checkpoint │ ├── __init__.py │ ├── checkpoint_manager.py │ ├── fsdp_checkpoint_manager.py │ └── megatron_checkpoint_manager.py ├── config.py ├── dataset │ ├── README.md │ ├── __init__.py │ ├── multiturn_sft_dataset.py │ ├── rl_dataset.py │ ├── rm_dataset.py │ ├── sft_dataset.py │ └── vision_utils.py ├── debug │ ├── __init__.py │ ├── performance.py │ ├── profile.py │ └── trajectory_tracker.py ├── device.py ├── distributed.py ├── experimental │ ├── __init__.py │ └── torch_functional.py ├── flops_counter.py ├── fs.py ├── fsdp_utils.py ├── hdfs_io.py ├── import_utils.py ├── logger │ ├── __init__.py │ └── aggregate_logger.py ├── logging_utils.py ├── megatron │ ├── __init__.py │ ├── memory.py │ ├── optimizer.py │ ├── pipeline_parallel.py │ ├── sequence_parallel.py │ └── tensor_parallel.py ├── megatron_utils.py ├── memory_buffer.py ├── metric │ ├── __init__.py │ └── utils.py ├── model.py ├── net_utils.py ├── py_functional.py ├── ray_utils.py ├── rendezvous │ ├── __init__.py │ └── ray_backend.py ├── reward_score │ ├── __init__.py │ ├── geo3k.py │ ├── gsm8k.py │ ├── math.py │ ├── math_batch.py │ ├── math_dapo.py │ ├── math_verify.py │ ├── prime_code │ │ ├── __init__.py │ │ ├── testing_util.py │ │ └── utils.py │ ├── prime_math │ │ ├── __init__.py │ │ ├── grader.py │ │ └── math_normalize.py │ └── sandbox_fusion │ │ ├── __init__.py │ │ └── utils.py ├── seqlen_balancing.py ├── tokenizer.py ├── torch_dtypes.py ├── torch_functional.py ├── tracking.py ├── ulysses.py └── vllm_utils.py ├── version └── version └── workers ├── __init__.py ├── actor ├── __init__.py ├── base.py ├── dp_actor.py └── megatron_actor.py ├── critic ├── __init__.py ├── base.py ├── dp_critic.py └── megatron_critic.py ├── fsdp_workers.py ├── megatron_workers.py ├── reward_manager ├── __init__.py ├── batch.py ├── dapo.py ├── naive.py └── prime.py ├── reward_model ├── __init__.py ├── base.py └── megatron │ ├── __init__.py │ └── reward_model.py ├── rollout ├── __init__.py ├── async_server.py ├── base.py ├── hf_rollout.py ├── naive │ ├── __init__.py │ └── naive_rollout.py ├── schemas.py ├── sglang_rollout │ ├── __init__.py │ ├── async_sglang_rollout.py │ ├── sglang_rollout.py │ └── utils.py ├── tokenizer.py └── vllm_rollout │ ├── __init__.py │ ├── fire_vllm_rollout.py │ ├── vllm_async_server.py │ ├── vllm_rollout.py │ └── vllm_rollout_spmd.py └── sharding_manager ├── __init__.py ├── base.py ├── fsdp_sglang.py ├── fsdp_ulysses.py ├── fsdp_vllm.py ├── megatron_sglang.py └── megatron_vllm.py /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/.gitignore -------------------------------------------------------------------------------- /Dockerfile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/Dockerfile -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/LICENSE -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/README.md -------------------------------------------------------------------------------- /configs/ppo_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/configs/ppo_trainer.yaml -------------------------------------------------------------------------------- /datasets/README.md: -------------------------------------------------------------------------------- 1 | # datasets -------------------------------------------------------------------------------- /images/Qwen2.5-3B.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/images/Qwen2.5-3B.png -------------------------------------------------------------------------------- /images/Qwen2.5-7B.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/images/Qwen2.5-7B.png -------------------------------------------------------------------------------- /images/pacs.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/images/pacs.png -------------------------------------------------------------------------------- /logs/README.md: -------------------------------------------------------------------------------- 1 | # logs -------------------------------------------------------------------------------- /outputs/README.md: -------------------------------------------------------------------------------- 1 | # outputs -------------------------------------------------------------------------------- /scripts/baseline/grpo.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/scripts/baseline/grpo.sh -------------------------------------------------------------------------------- /scripts/baseline/ppo.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/scripts/baseline/ppo.sh -------------------------------------------------------------------------------- /scripts/inference.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/scripts/inference.sh -------------------------------------------------------------------------------- /scripts/model_merge.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/scripts/model_merge.sh -------------------------------------------------------------------------------- /scripts/pacs/pacs_beta_1.0.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/scripts/pacs/pacs_beta_1.0.sh -------------------------------------------------------------------------------- /scripts/pacs/pacs_beta_2.0.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/scripts/pacs/pacs_beta_2.0.sh -------------------------------------------------------------------------------- /scripts/preprocess_dataset.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/scripts/preprocess_dataset.sh -------------------------------------------------------------------------------- /src/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/.DS_Store -------------------------------------------------------------------------------- /src/inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/inference.py -------------------------------------------------------------------------------- /src/json_file.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/json_file.py -------------------------------------------------------------------------------- /src/model_merger.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/model_merger.py -------------------------------------------------------------------------------- /src/pacs/pacs_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/pacs/pacs_actor.py -------------------------------------------------------------------------------- /src/pacs/pacs_compute_metrics.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/pacs/pacs_compute_metrics.py -------------------------------------------------------------------------------- /src/pacs/pacs_core_algos.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/pacs/pacs_core_algos.py -------------------------------------------------------------------------------- /src/pacs/pacs_main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/pacs/pacs_main.py -------------------------------------------------------------------------------- /src/pacs/pacs_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/pacs/pacs_trainer.py -------------------------------------------------------------------------------- /src/pacs/pacs_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/pacs/pacs_workers.py -------------------------------------------------------------------------------- /src/preprocess_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/preprocess_dataset.py -------------------------------------------------------------------------------- /src/reward_function.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/reward_function.py -------------------------------------------------------------------------------- /src/verl/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/__init__.py -------------------------------------------------------------------------------- /src/verl/models/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/README.md -------------------------------------------------------------------------------- /src/verl/models/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/__init__.py -------------------------------------------------------------------------------- /src/verl/models/llama/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/__init__.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/__init__.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/checkpoint_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/checkpoint_utils/__init__.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/checkpoint_utils/llama_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/checkpoint_utils/llama_loader.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/checkpoint_utils/llama_loader_depracated.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/checkpoint_utils/llama_loader_depracated.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/checkpoint_utils/llama_saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/checkpoint_utils/llama_saver.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/layers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/layers/__init__.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/layers/parallel_attention.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/layers/parallel_attention.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/layers/parallel_decoder.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/layers/parallel_decoder.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/layers/parallel_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/layers/parallel_linear.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/layers/parallel_mlp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/layers/parallel_mlp.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/layers/parallel_rmsnorm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/layers/parallel_rmsnorm.py -------------------------------------------------------------------------------- /src/verl/models/llama/megatron/modeling_llama_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/llama/megatron/modeling_llama_megatron.py -------------------------------------------------------------------------------- /src/verl/models/mcore/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/__init__.py -------------------------------------------------------------------------------- /src/verl/models/mcore/config_converter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/config_converter.py -------------------------------------------------------------------------------- /src/verl/models/mcore/loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/loader.py -------------------------------------------------------------------------------- /src/verl/models/mcore/model_forward.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/model_forward.py -------------------------------------------------------------------------------- /src/verl/models/mcore/model_initializer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/model_initializer.py -------------------------------------------------------------------------------- /src/verl/models/mcore/readme.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/readme.md -------------------------------------------------------------------------------- /src/verl/models/mcore/registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/registry.py -------------------------------------------------------------------------------- /src/verl/models/mcore/saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/saver.py -------------------------------------------------------------------------------- /src/verl/models/mcore/util.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/util.py -------------------------------------------------------------------------------- /src/verl/models/mcore/weight_converter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/mcore/weight_converter.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/__init__.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/__init__.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/checkpoint_utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/checkpoint_utils/__init__.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader_depracated.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader_depracated.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/checkpoint_utils/qwen2_saver.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/checkpoint_utils/qwen2_saver.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/layers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/layers/__init__.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/layers/parallel_attention.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/layers/parallel_attention.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/layers/parallel_decoder.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/layers/parallel_decoder.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/layers/parallel_linear.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/layers/parallel_linear.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/layers/parallel_mlp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/layers/parallel_mlp.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/layers/parallel_rmsnorm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/layers/parallel_rmsnorm.py -------------------------------------------------------------------------------- /src/verl/models/qwen2/megatron/modeling_qwen2_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/qwen2/megatron/modeling_qwen2_megatron.py -------------------------------------------------------------------------------- /src/verl/models/registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/registry.py -------------------------------------------------------------------------------- /src/verl/models/transformers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/transformers/__init__.py -------------------------------------------------------------------------------- /src/verl/models/transformers/llama.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/transformers/llama.py -------------------------------------------------------------------------------- /src/verl/models/transformers/monkey_patch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/transformers/monkey_patch.py -------------------------------------------------------------------------------- /src/verl/models/transformers/qwen2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/transformers/qwen2.py -------------------------------------------------------------------------------- /src/verl/models/transformers/qwen2_5_vl.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/transformers/qwen2_5_vl.py -------------------------------------------------------------------------------- /src/verl/models/transformers/qwen2_vl.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/transformers/qwen2_vl.py -------------------------------------------------------------------------------- /src/verl/models/weight_loader_registry.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/models/weight_loader_registry.py -------------------------------------------------------------------------------- /src/verl/protocol.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/protocol.py -------------------------------------------------------------------------------- /src/verl/single_controller/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/__init__.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/__init__.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/decorator.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/decorator.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/megatron/__init__.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/megatron/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/megatron/worker.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/megatron/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/megatron/worker_group.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/register_center/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/register_center/__init__.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/register_center/ray.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/register_center/ray.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/worker.py -------------------------------------------------------------------------------- /src/verl/single_controller/base/worker_group.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/base/worker_group.py -------------------------------------------------------------------------------- /src/verl/single_controller/ray/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/ray/__init__.py -------------------------------------------------------------------------------- /src/verl/single_controller/ray/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/ray/base.py -------------------------------------------------------------------------------- /src/verl/single_controller/ray/megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/single_controller/ray/megatron.py -------------------------------------------------------------------------------- /src/verl/third_party/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/__init__.py -------------------------------------------------------------------------------- /src/verl/third_party/sglang/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/sglang/__init__.py -------------------------------------------------------------------------------- /src/verl/third_party/sglang/parallel_state.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/sglang/parallel_state.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/__init__.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/__init__.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/arg_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/arg_utils.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/config.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/dtensor_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/dtensor_weight_loaders.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/hf_weight_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/hf_weight_loader.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/llm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/llm.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/llm_engine_sp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/llm_engine_sp.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/megatron_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/megatron_weight_loaders.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/model_runner.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/model_runner.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/parallel_state.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/parallel_state.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/spmd_gpu_executor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/spmd_gpu_executor.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/tokenizer.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_5_4/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_5_4/worker.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/__init__.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/arg_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/arg_utils.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/config.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/dtensor_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/dtensor_weight_loaders.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/hf_weight_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/hf_weight_loader.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/llm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/llm.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/llm_engine_sp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/llm_engine_sp.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/megatron_weight_loaders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/megatron_weight_loaders.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/model_runner.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/model_runner.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/parallel_state.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/parallel_state.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/tokenizer.py -------------------------------------------------------------------------------- /src/verl/third_party/vllm/vllm_v_0_6_3/worker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/third_party/vllm/vllm_v_0_6_3/worker.py -------------------------------------------------------------------------------- /src/verl/tools/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/tools/__init__.py -------------------------------------------------------------------------------- /src/verl/tools/base_tool.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/tools/base_tool.py -------------------------------------------------------------------------------- /src/verl/tools/gsm8k_tool.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/tools/gsm8k_tool.py -------------------------------------------------------------------------------- /src/verl/tools/schemas.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/tools/schemas.py -------------------------------------------------------------------------------- /src/verl/trainer/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/__init__.py -------------------------------------------------------------------------------- /src/verl/trainer/config/evaluation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/config/evaluation.yaml -------------------------------------------------------------------------------- /src/verl/trainer/config/generation.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/config/generation.yaml -------------------------------------------------------------------------------- /src/verl/trainer/config/ppo_megatron_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/config/ppo_megatron_trainer.yaml -------------------------------------------------------------------------------- /src/verl/trainer/config/ppo_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/config/ppo_trainer.yaml -------------------------------------------------------------------------------- /src/verl/trainer/config/sft_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/config/sft_trainer.yaml -------------------------------------------------------------------------------- /src/verl/trainer/fsdp_sft_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/fsdp_sft_trainer.py -------------------------------------------------------------------------------- /src/verl/trainer/main_eval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/main_eval.py -------------------------------------------------------------------------------- /src/verl/trainer/main_generation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/main_generation.py -------------------------------------------------------------------------------- /src/verl/trainer/main_ppo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/main_ppo.py -------------------------------------------------------------------------------- /src/verl/trainer/ppo/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/ppo/__init__.py -------------------------------------------------------------------------------- /src/verl/trainer/ppo/core_algos.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/ppo/core_algos.py -------------------------------------------------------------------------------- /src/verl/trainer/ppo/metric_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/ppo/metric_utils.py -------------------------------------------------------------------------------- /src/verl/trainer/ppo/ray_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/ppo/ray_trainer.py -------------------------------------------------------------------------------- /src/verl/trainer/ppo/reward.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/ppo/reward.py -------------------------------------------------------------------------------- /src/verl/trainer/runtime_env.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/trainer/runtime_env.yaml -------------------------------------------------------------------------------- /src/verl/utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/activation_offload.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/activation_offload.py -------------------------------------------------------------------------------- /src/verl/utils/checkpoint/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/checkpoint/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/checkpoint/checkpoint_manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/checkpoint/checkpoint_manager.py -------------------------------------------------------------------------------- /src/verl/utils/checkpoint/fsdp_checkpoint_manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/checkpoint/fsdp_checkpoint_manager.py -------------------------------------------------------------------------------- /src/verl/utils/checkpoint/megatron_checkpoint_manager.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/checkpoint/megatron_checkpoint_manager.py -------------------------------------------------------------------------------- /src/verl/utils/config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/config.py -------------------------------------------------------------------------------- /src/verl/utils/dataset/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/dataset/README.md -------------------------------------------------------------------------------- /src/verl/utils/dataset/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/dataset/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/dataset/multiturn_sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/dataset/multiturn_sft_dataset.py -------------------------------------------------------------------------------- /src/verl/utils/dataset/rl_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/dataset/rl_dataset.py -------------------------------------------------------------------------------- /src/verl/utils/dataset/rm_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/dataset/rm_dataset.py -------------------------------------------------------------------------------- /src/verl/utils/dataset/sft_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/dataset/sft_dataset.py -------------------------------------------------------------------------------- /src/verl/utils/dataset/vision_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/dataset/vision_utils.py -------------------------------------------------------------------------------- /src/verl/utils/debug/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/debug/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/debug/performance.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/debug/performance.py -------------------------------------------------------------------------------- /src/verl/utils/debug/profile.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/debug/profile.py -------------------------------------------------------------------------------- /src/verl/utils/debug/trajectory_tracker.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/debug/trajectory_tracker.py -------------------------------------------------------------------------------- /src/verl/utils/device.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/device.py -------------------------------------------------------------------------------- /src/verl/utils/distributed.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/distributed.py -------------------------------------------------------------------------------- /src/verl/utils/experimental/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/experimental/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/experimental/torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/experimental/torch_functional.py -------------------------------------------------------------------------------- /src/verl/utils/flops_counter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/flops_counter.py -------------------------------------------------------------------------------- /src/verl/utils/fs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/fs.py -------------------------------------------------------------------------------- /src/verl/utils/fsdp_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/fsdp_utils.py -------------------------------------------------------------------------------- /src/verl/utils/hdfs_io.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/hdfs_io.py -------------------------------------------------------------------------------- /src/verl/utils/import_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/import_utils.py -------------------------------------------------------------------------------- /src/verl/utils/logger/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/logger/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/logger/aggregate_logger.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/logger/aggregate_logger.py -------------------------------------------------------------------------------- /src/verl/utils/logging_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/logging_utils.py -------------------------------------------------------------------------------- /src/verl/utils/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/megatron/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/megatron/memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/megatron/memory.py -------------------------------------------------------------------------------- /src/verl/utils/megatron/optimizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/megatron/optimizer.py -------------------------------------------------------------------------------- /src/verl/utils/megatron/pipeline_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/megatron/pipeline_parallel.py -------------------------------------------------------------------------------- /src/verl/utils/megatron/sequence_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/megatron/sequence_parallel.py -------------------------------------------------------------------------------- /src/verl/utils/megatron/tensor_parallel.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/megatron/tensor_parallel.py -------------------------------------------------------------------------------- /src/verl/utils/megatron_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/megatron_utils.py -------------------------------------------------------------------------------- /src/verl/utils/memory_buffer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/memory_buffer.py -------------------------------------------------------------------------------- /src/verl/utils/metric/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/metric/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/metric/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/metric/utils.py -------------------------------------------------------------------------------- /src/verl/utils/model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/model.py -------------------------------------------------------------------------------- /src/verl/utils/net_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/net_utils.py -------------------------------------------------------------------------------- /src/verl/utils/py_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/py_functional.py -------------------------------------------------------------------------------- /src/verl/utils/ray_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/ray_utils.py -------------------------------------------------------------------------------- /src/verl/utils/rendezvous/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/rendezvous/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/rendezvous/ray_backend.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/rendezvous/ray_backend.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/geo3k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/geo3k.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/gsm8k.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/gsm8k.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/math.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/math.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/math_batch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/math_batch.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/math_dapo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/math_dapo.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/math_verify.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/math_verify.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/prime_code/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/prime_code/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/prime_code/testing_util.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/prime_code/testing_util.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/prime_code/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/prime_code/utils.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/prime_math/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/prime_math/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/prime_math/grader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/prime_math/grader.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/prime_math/math_normalize.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/prime_math/math_normalize.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/sandbox_fusion/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/sandbox_fusion/__init__.py -------------------------------------------------------------------------------- /src/verl/utils/reward_score/sandbox_fusion/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/reward_score/sandbox_fusion/utils.py -------------------------------------------------------------------------------- /src/verl/utils/seqlen_balancing.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/seqlen_balancing.py -------------------------------------------------------------------------------- /src/verl/utils/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/tokenizer.py -------------------------------------------------------------------------------- /src/verl/utils/torch_dtypes.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/torch_dtypes.py -------------------------------------------------------------------------------- /src/verl/utils/torch_functional.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/torch_functional.py -------------------------------------------------------------------------------- /src/verl/utils/tracking.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/tracking.py -------------------------------------------------------------------------------- /src/verl/utils/ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/ulysses.py -------------------------------------------------------------------------------- /src/verl/utils/vllm_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/utils/vllm_utils.py -------------------------------------------------------------------------------- /src/verl/version/version: -------------------------------------------------------------------------------- 1 | 0.3.1.dev 2 | -------------------------------------------------------------------------------- /src/verl/workers/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/actor/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/actor/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/actor/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/actor/base.py -------------------------------------------------------------------------------- /src/verl/workers/actor/dp_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/actor/dp_actor.py -------------------------------------------------------------------------------- /src/verl/workers/actor/megatron_actor.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/actor/megatron_actor.py -------------------------------------------------------------------------------- /src/verl/workers/critic/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/critic/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/critic/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/critic/base.py -------------------------------------------------------------------------------- /src/verl/workers/critic/dp_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/critic/dp_critic.py -------------------------------------------------------------------------------- /src/verl/workers/critic/megatron_critic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/critic/megatron_critic.py -------------------------------------------------------------------------------- /src/verl/workers/fsdp_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/fsdp_workers.py -------------------------------------------------------------------------------- /src/verl/workers/megatron_workers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/megatron_workers.py -------------------------------------------------------------------------------- /src/verl/workers/reward_manager/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_manager/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/reward_manager/batch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_manager/batch.py -------------------------------------------------------------------------------- /src/verl/workers/reward_manager/dapo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_manager/dapo.py -------------------------------------------------------------------------------- /src/verl/workers/reward_manager/naive.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_manager/naive.py -------------------------------------------------------------------------------- /src/verl/workers/reward_manager/prime.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_manager/prime.py -------------------------------------------------------------------------------- /src/verl/workers/reward_model/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_model/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/reward_model/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_model/base.py -------------------------------------------------------------------------------- /src/verl/workers/reward_model/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_model/megatron/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/reward_model/megatron/reward_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/reward_model/megatron/reward_model.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/async_server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/async_server.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/base.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/hf_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/hf_rollout.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/naive/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/naive/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/naive/naive_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/naive/naive_rollout.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/schemas.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/schemas.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/sglang_rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/sglang_rollout/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/sglang_rollout/async_sglang_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/sglang_rollout/async_sglang_rollout.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/sglang_rollout/sglang_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/sglang_rollout/sglang_rollout.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/sglang_rollout/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/sglang_rollout/utils.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/tokenizer.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/vllm_rollout/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/vllm_rollout/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/vllm_rollout/fire_vllm_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/vllm_rollout/fire_vllm_rollout.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/vllm_rollout/vllm_async_server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/vllm_rollout/vllm_async_server.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/vllm_rollout/vllm_rollout.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/vllm_rollout/vllm_rollout.py -------------------------------------------------------------------------------- /src/verl/workers/rollout/vllm_rollout/vllm_rollout_spmd.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/rollout/vllm_rollout/vllm_rollout_spmd.py -------------------------------------------------------------------------------- /src/verl/workers/sharding_manager/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/sharding_manager/__init__.py -------------------------------------------------------------------------------- /src/verl/workers/sharding_manager/base.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/sharding_manager/base.py -------------------------------------------------------------------------------- /src/verl/workers/sharding_manager/fsdp_sglang.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/sharding_manager/fsdp_sglang.py -------------------------------------------------------------------------------- /src/verl/workers/sharding_manager/fsdp_ulysses.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/sharding_manager/fsdp_ulysses.py -------------------------------------------------------------------------------- /src/verl/workers/sharding_manager/fsdp_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/sharding_manager/fsdp_vllm.py -------------------------------------------------------------------------------- /src/verl/workers/sharding_manager/megatron_sglang.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/sharding_manager/megatron_sglang.py -------------------------------------------------------------------------------- /src/verl/workers/sharding_manager/megatron_vllm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ritzz-ai/PACS/HEAD/src/verl/workers/sharding_manager/megatron_vllm.py --------------------------------------------------------------------------------