{"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/moe-jetpack-from-dense-checkpoints-to","title":"MoE Jetpack: From Dense Checkpoints to Adaptive Mixture of Experts for Vision Tasks","arxiv_id":"2406.04801","date":"2024-06-07","proceeding":null,"authors":["Xingkui Zhu","Yiran Guan","Dingkang Liang","Yuchao Chen","Yuliang Liu","Xiang Bai"],"abstract":"The sparsely activated mixture of experts (MoE) model presents a promising alternative to traditional densely activated (dense) models, enhancing both quality and computational efficiency. However, training MoE models from scratch demands extensive data and computational resources. Moreover, public repositories like timm mainly provide pre-trained dense checkpoints, lacking similar resources for MoE models, hindering their adoption. To bridge this gap, we introduce MoE Jetpack, an effective method for fine-tuning dense checkpoints into MoE models. MoE Jetpack incorporates two key techniques: (1) checkpoint recycling, which repurposes dense checkpoints as initial weights for MoE models, thereby accelerating convergence, enhancing accuracy, and alleviating the computational burden of pre-training; (2) hyperspherical adaptive MoE (SpheroMoE) layer, which optimizes the MoE architecture for better integration of dense checkpoints, enhancing fine-tuning performance. Our experiments on vision tasks demonstrate that MoE Jetpack significantly improves convergence speed and accuracy when fine-tuning dense checkpoints into MoE models. Our code will be publicly available at https://github.com/Adlith/MoE-Jetpack.","url_abs":"https://arxiv.org/abs/2406.04801v1","url_pdf":"https://arxiv.org/pdf/2406.04801v1.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":"moe-jetpack-from-dense-checkpoints-to","repo_url":"https://github.com/adlith/moe-jetpack","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"computational-efficiency","task_name":"Computational Efficiency"},{"task_slug":"mixture-of-experts","task_name":"Mixture-of-Experts"}],"methods":[{"method_slug":"moe","method_name":"MoE"},{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2406.04801","atlas_url":"https://app.syntology.ai/?focus=2406.04801","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.04801"}},"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/adlith/moe-jetpack","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran_draft_wrong":1,"ran":5,"unverified":1},"by_repo_kind":{"official":{"samples":7,"ran":6,"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":"b5472673332de78e","entry":"log","repo":"adlith/moe-jetpack","repo_kind":"official","path":"moejet/models/soft_moe.py","file_url":"https://github.com/adlith/moe-jetpack/blob/HEAD/moejet/models/soft_moe.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":2,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b5472673332de78e"}},{"code_sha256_prefix":"a322f2edda4b2d6e","entry":"ckpt_to_state_dict","repo":"adlith/moe-jetpack","repo_kind":"official","path":".dev_scripts/ckpt_tree.py","file_url":"https://github.com/adlith/moe-jetpack/blob/HEAD/.dev_scripts/ckpt_tree.py","link_basis":"first_harvest_node","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":"a322f2edda4b2d6e"}},{"code_sha256_prefix":"8eebff71c1cabe84","entry":"filter_models_by_task","repo":"adlith/moe-jetpack","repo_kind":"official","path":".dev_scripts/generate_readme.py","file_url":"https://github.com/adlith/moe-jetpack/blob/HEAD/.dev_scripts/generate_readme.py","link_basis":"first_harvest_node","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":"8eebff71c1cabe84"}},{"code_sha256_prefix":"e7a7dab3edcc1e58","entry":"generate_moe_custom_keys","repo":"adlith/moe-jetpack","repo_kind":"official","path":"moejet/configs/timm/vit_tiny_dual_moe_timm_21k_ft.py","file_url":"https://github.com/adlith/moe-jetpack/blob/HEAD/moejet/configs/timm/vit_tiny_dual_moe_timm_21k_ft.py","link_basis":"first_harvest_node","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":"e7a7dab3edcc1e58"}},{"code_sha256_prefix":"c1451f5572ce77c1","entry":"l2norm","repo":"adlith/moe-jetpack","repo_kind":"official","path":"moejet/models/soft_moe.py","file_url":"https://github.com/adlith/moe-jetpack/blob/HEAD/moejet/models/soft_moe.py","link_basis":"first_harvest_node","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":"c1451f5572ce77c1"}},{"code_sha256_prefix":"b0d57361257879cc","entry":"stable_softmax","repo":"adlith/moe-jetpack","repo_kind":"official","path":"moejet/models/soft_moe.py","file_url":"https://github.com/adlith/moe-jetpack/blob/HEAD/moejet/models/soft_moe.py","link_basis":"first_harvest_node","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":"b0d57361257879cc"}},{"code_sha256_prefix":"9abd4b00a0d36a31","entry":"get_flops_params","repo":"adlith/moe-jetpack","repo_kind":"official","path":".dev_scripts/fill_metafile.py","file_url":"https://github.com/adlith/moe-jetpack/blob/HEAD/.dev_scripts/fill_metafile.py","link_basis":"first_harvest_node","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":"9abd4b00a0d36a31"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}