{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/cppo-accelerating-the-training-of-group","title":"CPPO: Accelerating the Training of Group Relative Policy Optimization-Based Reasoning Models","arxiv_id":"2503.22342","date":"2025-03-28","proceeding":null,"authors":["Zhihang Lin","Mingbao Lin","Yuan Xie","Rongrong Ji"],"abstract":"This paper introduces Completion Pruning Policy Optimization (CPPO) to accelerate the training of reasoning models based on Group Relative Policy Optimization (GRPO). GRPO, while effective, incurs high training costs due to the need for sampling multiple completions for each question. Our experiment and theoretical analysis reveals that the number of completions impacts model accuracy yet increases training time multiplicatively, and not all completions contribute equally to policy training -- their contribution depends on their relative advantage. To address these issues, we propose CPPO, which prunes completions with low absolute advantages, significantly reducing the number needed for gradient calculation and updates. Additionally, we introduce a dynamic completion allocation strategy to maximize GPU utilization by incorporating additional questions, further enhancing training efficiency. Experimental results demonstrate that CPPO achieves up to $8.32\\times$ speedup on GSM8K and $3.51\\times$ on Math while preserving or even enhancing the accuracy compared to the original GRPO. We release our code at https://github.com/lzhxmu/CPPO.","url_abs":"https://arxiv.org/abs/2503.22342v1","url_pdf":"https://arxiv.org/pdf/2503.22342v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"cppo-accelerating-the-training-of-group","repo_url":"https://github.com/lzhxmu/cppo","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":null,"task_name":"GPU"},{"task_slug":"gsm8k","task_name":"GSM8K"},{"task_slug":"math","task_name":"Math"}],"methods":[{"method_slug":"pruning","method_name":"Pruning"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2503.22342","atlas_url":"https://app.syntology.ai/?focus=2503.22342","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2503.22342"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/lzhxmu/cppo","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran":4,"ran_draft_wrong":2,"ran_fixture":1,"unverified":6},"by_repo_kind":{"official":{"samples":13,"ran":7,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"6859aa3f82a38002","entry":"extract_answer_from_model_output","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/recipe/cppo/src/gsm8k_compute_score.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/recipe/cppo/src/gsm8k_compute_score.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6859aa3f82a38002"}},{"code_sha256_prefix":"0014331dab98c0a5","entry":"extract_last_number","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/recipe/cppo/src/gsm8k_compute_score.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/recipe/cppo/src/gsm8k_compute_score.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0014331dab98c0a5"}},{"code_sha256_prefix":"6036741237b33141","entry":"extract_single_number","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/recipe/cppo/src/gsm8k_compute_score.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/recipe/cppo/src/gsm8k_compute_score.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6036741237b33141"}},{"code_sha256_prefix":"34be00218891a21a","entry":"get_custom_reward_fn","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/recipe/cppo/src/main_cppo.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/recipe/cppo/src/main_cppo.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"34be00218891a21a"}},{"code_sha256_prefix":"fe79c459a3e4cc64","entry":"get_huggingface_actor_config","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/utils/model.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/utils/model.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"fe79c459a3e4cc64"}},{"code_sha256_prefix":"21331a58f93375e4","entry":"union_tensor_dict","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/protocol.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/protocol.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"DEP_MISSING","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"21331a58f93375e4"}},{"code_sha256_prefix":"25f0ea3f460f6ce1","entry":"unpad_dataproto","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/protocol.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/protocol.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"25f0ea3f460f6ce1"}},{"code_sha256_prefix":"e4128a6f1b0585bc","entry":"extract_solution","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/recipe/cppo/precess_gsm8k.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/recipe/cppo/precess_gsm8k.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e4128a6f1b0585bc"}},{"code_sha256_prefix":"660713a8fa047736","entry":"get_generation_config","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/utils/model.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/utils/model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"660713a8fa047736"}},{"code_sha256_prefix":"39fba8c5b01d595e","entry":"get_weight_loader","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/models/weight_loader_registry.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/models/weight_loader_registry.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"39fba8c5b01d595e"}},{"code_sha256_prefix":"8bf1b3e9bf87072c","entry":"get_weight_saver","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/models/weight_loader_registry.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/models/weight_loader_registry.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"8bf1b3e9bf87072c"}},{"code_sha256_prefix":"1901b047ab8e1654","entry":"load_state_dict_to_megatron_gptmodel","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/models/mcore/loader.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/models/mcore/loader.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"1901b047ab8e1654"}},{"code_sha256_prefix":"3b15e2ac7497c441","entry":"squeeze","repo":"lzhxmu/cppo","repo_kind":"official","path":"cppo_verl/verl/utils/model.py","file_url":"https://github.com/lzhxmu/cppo/blob/HEAD/cppo_verl/verl/utils/model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"3b15e2ac7497c441"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}