{"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/tensor-comprehensions-framework-agnostic-high","title":"Tensor Comprehensions: Framework-Agnostic High-Performance Machine Learning Abstractions","arxiv_id":"1802.04730","date":"2018-02-13","proceeding":null,"authors":["Nicolas Vasilache","Oleksandr Zinenko","Theodoros Theodoridis","Priya Goyal","Zachary DeVito","William S. Moses","Sven Verdoolaege","Andrew Adams","Albert Cohen"],"abstract":"Deep learning models with convolutional and recurrent networks are now\nubiquitous and analyze massive amounts of audio, image, video, text and graph\ndata, with applications in automatic translation, speech-to-text, scene\nunderstanding, ranking user preferences, ad placement, etc. Competing\nframeworks for building these networks such as TensorFlow, Chainer, CNTK,\nTorch/PyTorch, Caffe1/2, MXNet and Theano, explore different tradeoffs between\nusability and expressiveness, research or production orientation and supported\nhardware. They operate on a DAG of computational operators, wrapping\nhigh-performance libraries such as CUDNN (for NVIDIA GPUs) or NNPACK (for\nvarious CPUs), and automate memory allocation, synchronization, distribution.\nCustom operators are needed where the computation does not fit existing\nhigh-performance library calls, usually at a high engineering cost. This is\nfrequently required when new operators are invented by researchers: such\noperators suffer a severe performance penalty, which limits the pace of\ninnovation. Furthermore, even if there is an existing runtime call these\nframeworks can use, it often doesn't offer optimal performance for a user's\nparticular network architecture and dataset, missing optimizations between\noperators as well as optimizations that can be done knowing the size and shape\nof data. Our contributions include (1) a language close to the mathematics of\ndeep learning called Tensor Comprehensions, (2) a polyhedral Just-In-Time\ncompiler to convert a mathematical description of a deep learning DAG into a\nCUDA kernel with delegated memory management and synchronization, also\nproviding optimizations such as operator fusion and specialization for specific\nsizes, (3) a compilation cache populated by an autotuner. [Abstract cutoff]","url_abs":"http://arxiv.org/abs/1802.04730v3","url_pdf":"http://arxiv.org/pdf/1802.04730v3.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":"tensor-comprehensions-framework-agnostic-high","repo_url":"https://github.com/AIwithSwift/TFWorld2019-SwiftIn3Hours","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"tensor-comprehensions-framework-agnostic-high","repo_url":"https://github.com/Huawei-MRC-OSI/optml-zoo","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"tensor-comprehensions-framework-agnostic-high","repo_url":"https://github.com/LoopTactics/tc-cim","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"tensor-comprehensions-framework-agnostic-high","repo_url":"https://github.com/facebookresearch/TensorComprehensions","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"machine-learning","task_name":"BIG-bench Machine Learning"},{"task_slug":"management","task_name":"Management"},{"task_slug":"scene-understanding","task_name":"Scene Understanding"},{"task_slug":"speech-to-text","task_name":"Speech-to-Text"},{"task_slug":"high","task_name":"Vocal Bursts Intensity Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.04730","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1802.04730"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/facebookresearch/TensorComprehensions","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/LoopTactics/tc-cim","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Huawei-MRC-OSI/optml-zoo","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/AIwithSwift/TFWorld2019-SwiftIn3Hours","reach":{"status":"ok"}}],"summary":{"unverified":3},"by_repo_kind":{"listed":{"samples":3,"ran":0,"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":"30342e3e0fcd59b6","entry":"get_tensors","repo":"LoopTactics/tc-cim","repo_kind":"listed","path":"tensor_comprehensions/torch_tc/tc_function.py","file_url":"https://github.com/LoopTactics/tc-cim/blob/HEAD/tensor_comprehensions/torch_tc/tc_function.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":"30342e3e0fcd59b6"}},{"code_sha256_prefix":"a2aa11589c16188f","entry":"unpack_variables","repo":"LoopTactics/tc-cim","repo_kind":"listed","path":"tensor_comprehensions/torch_tc/tc_function.py","file_url":"https://github.com/LoopTactics/tc-cim/blob/HEAD/tensor_comprehensions/torch_tc/tc_function.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":"a2aa11589c16188f"}},{"code_sha256_prefix":"3688de0a0c9b9ab0","entry":"wrap_variable","repo":"LoopTactics/tc-cim","repo_kind":"listed","path":"tensor_comprehensions/torch_tc/tc_function.py","file_url":"https://github.com/LoopTactics/tc-cim/blob/HEAD/tensor_comprehensions/torch_tc/tc_function.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":"3688de0a0c9b9ab0"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}