{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/unit-scaling-out-of-the-box-low-precision","title":"Unit Scaling: Out-of-the-Box Low-Precision Training","arxiv_id":"2303.11257","date":"2023-03-20","proceeding":null,"authors":["Charlie Blake","Douglas Orr","Carlo Luschi"],"abstract":"We present unit scaling, a paradigm for designing deep learning models that simplifies the use of low-precision number formats. Training in FP16 or the recently proposed FP8 formats offers substantial efficiency gains, but can lack sufficient range for out-of-the-box training. Unit scaling addresses this by introducing a principled approach to model numerics: seeking unit variance of all weights, activations and gradients at initialisation. Unlike alternative methods, this approach neither requires multiple training runs to find a suitable scale nor has significant computational overhead. We demonstrate the efficacy of unit scaling across a range of models and optimisers. We further show that existing models can be adapted to be unit-scaled, training BERT-Large in FP16 and then FP8 with no degradation in accuracy.","url_abs":"https://arxiv.org/abs/2303.11257v2","url_pdf":"https://arxiv.org/pdf/2303.11257v2.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":"unit-scaling-out-of-the-box-low-precision","repo_url":"https://github.com/graphcore-research/unit-scaling-demo","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"unit-scaling-out-of-the-box-low-precision","repo_url":"https://github.com/graphcore-research/unit-scaling","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2303.11257","atlas_url":"https://app.syntology.ai/?focus=2303.11257","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2303.11257"},"calls":[{"tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2303.11257"},"returns":"the samples harvested from this paper, each with its run record; it does not verify the paper"},{"tool":"get_code","arguments_in":"samples[].code_sha256","returns":"one stored sample by its code_sha256; not shown on this site"}]},"developers":"https://syntology.ai/developers","held":true,"reason":"in_graph_layer","checked_against":null,"agent_calls":[{"tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2303.11257"},"call":"get_harvested_code_for_paper(arxiv_id=\"2303.11257\")"},{"tool":"get_code","arguments_in":{"code_sha256":"syntology.samples[].code_sha256"},"call":"get_code(code_sha256=\"…\")"},{"tool":"get_citation_path","arguments":{"paper_1":"2303.11257"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2303.11257\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2303.11257"},"call":"get_concepts_for_paper(arxiv_id=\"2303.11257\")"}],"by_repo":[{"repo":"graphcore-research/unit-scaling-demo","kind":"official","samples":14,"ran":12,"instrument":0,"unverified":2},{"repo":"graphcore-research/unit-scaling","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":1,"harvested_for":["2407.17465"],"harvested_for_calls":[{"arxiv_id":"2407.17465","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2407.17465"},"call":"get_harvested_code_for_paper(arxiv_id=\"2407.17465\")"}]}],"by_repo_harvested_for_is":"harvested_for: the arXiv ids (as Syntology's graph holds them) of every paper whose samples Syntology's graph links to this repository, in arXiv-id order, complete; the page names the first 3 and says how many more. A link only: it does not say the papers are the same work, that one copied the other, or that the code implements this paper's method; harvested_for_calls gives the get_harvested_code_for_paper call for each arXiv-shaped id among them","by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code","read_at":"2026-10-05T00:36:11+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","ran_record":{"url":"https://syntology.ai/api/ran/2303.11257.json","schema":"syntology.ran_record/1","counts_crosswalk":{"n_samples":"syntology.counts.lifted","n_ran":"syntology.counts.executed","n_ran_checked":"syntology.counts.checked","n_instrument":"syntology.counts.instrument_failures","n_unverified":"syntology.counts.no_recorded_run"},"note":"live record, measured at request time; may be newer than this page's graph read","badge":"https://syntology.ai/api/ran/2303.11257.svg"},"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/graphcore-research/unit-scaling","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/graphcore-research/unit-scaling-demo","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":12,"unverified":2},"n_samples":14,"n_ran":12,"n_constructed":0,"n_ran_checked":12,"n_instrument":0,"n_unverified":2,"n_instrument_is":"failures of Syntology's instrument, not of the code","every_run_is_an_instrument_failure":false,"key_notes":{"n_ran_checked":"legacy name, kept unchanged so existing readers do not break: it counts the samples that ran with no instrument failure (honoured, violated, and ran with no contract checked); it does not mean a contract was checked, and the pages print it as 'K with no instrument failure', not 'K checked'","n_constructed":"a sub-count of the samples that ran, never subtracted from them and never a failure: an executed sample whose run returned an instance of its own class (fixture_out_type equals the entry name): the run built an object and did not compute a result (Syntology's RAN record, counts.constructed)"},"by_repo_kind":{"official":{"samples":14,"ran":12,"n_ran_checked":12,"n_instrument":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,"inline_ok_is":"inline_ok is true only when both licence gates allow Syntology to serve this code's text: this paper's link (inline_ok_link, from the licence of this paper's copy) and the stored copy of the same code that Syntology serves, which may have been first harvested from another repository. withheld_clause is Syntology's own wording of which gate keeps the text a pointer and why, and is null when inline_ok is true. This site shows no code text either way; n_pointer_only_for_licence counts samples with inline_ok false (both gates), n_pointer_only_link_allows those of them whose inline_ok_link is true","n_pointer_only_link_allows":0,"fixture_fields_is":{"signature_and_fixture_out_type":"the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right","level_recorded_at":"when Syntology's graph wrote the sample's verification level, not when the code executed; absent where the level was written before the date was kept, and never filled in with the build date","present_on":"samples that ran only; a key is absent, never empty, where the graph records nothing","signature_not_shown":"the graph records a signature for this sample and it is not shown; signature_not_shown_because names why, as the export's field-exclusion reason","signature_not_shown_because":{"signature_is_harness_object_repr":"an argument's recorded type or element kind is a printout of an object in Syntology's harness (such as a mock object, a class or a pandas printout), not the type of an input, so it does not describe the call's shape"}},"n_ran_with_level_date":12,"n_ran_without_level_date":0,"n_ran_signature_not_shown":1,"levels_written_between":{"first":"2026-09-25","last":"2026-09-25","is":"the first and last UTC dates on which Syntology's graph wrote a verification level for a sample on this page that ran, not when the code executed"},"code_page_min_papers":5,"code_page_is":"a page for the entry name across every paper's harvested code; the name is a string, not an identity","samples":[{"code_sha256":"2f8c9dc17d697ee0","code_sha256_prefix":"2f8c9dc17d697ee0","entry":"batched_gather","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/layers.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/layers.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"tables:EagerTensor/3/float, indices:EagerTensor/2/int","fixture_out_type":"EagerTensor","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"2f8c9dc17d697ee0"}},{"code_sha256":"922e7a65963e6b06","code_sha256_prefix":"922e7a65963e6b06","entry":"conv1d","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/uscale/ops.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/uscale/ops.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"input:EagerTensor/3/float, filters:EagerTensor/3/float, stride:int/0/int, padding:str/0/str","fixture_out_type":"EagerTensor","level_recorded_at":"2026-09-25T08:34:48.807261+00:00","code_page":"/code/conv1d","census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"922e7a65963e6b06"}},{"code_sha256":"05df62f6648200c1","code_sha256_prefix":"05df62f6648200c1","entry":"load_character","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/datasets.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/datasets.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"root:PosixPath/0/?, train:str/0/str, valid:str/0/str","fixture_out_type":"Data","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"05df62f6648200c1"}},{"code_sha256":"73c30a2cc2fee462","code_sha256_prefix":"73c30a2cc2fee462","entry":"pointwise","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/uscale/ops.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/uscale/ops.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"inputs:EagerTensor/2/float, weights:EagerTensor/2/float, scale_for:str/0/str","fixture_out_type":"EagerTensor","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"73c30a2cc2fee462"}},{"code_sha256":"8cae868db43ba55a","code_sha256_prefix":"8cae868db43ba55a","entry":"printing","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/uscale/utility.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/uscale/utility.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"name:str/0/str, summary:callable/0/?","fixture_out_type":"Bind","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"8cae868db43ba55a"}},{"code_sha256":"55834bdb013ab248","code_sha256_prefix":"55834bdb013ab248","entry":"scaling","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/uscale/ops.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/uscale/ops.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"forward:float/0/float, backward:float/0/float","fixture_out_type":"Bind","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":"/code/scaling","census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"55834bdb013ab248"}},{"code_sha256":"288d2d846f7173a7","code_sha256_prefix":"288d2d846f7173a7","entry":"sinusoid_embedding","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/layers.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/layers.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"sequence_length:int/0/int, frequencies:int/0/int, max_period:int/0/int","fixture_out_type":"ndarray","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"288d2d846f7173a7"}},{"code_sha256":"d380f15ca83de673","code_sha256_prefix":"d380f15ca83de673","entry":"softmax_cross_entropy","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/layers.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/layers.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"scores:EagerTensor/3/float, ids:EagerTensor/2/int, mask:EagerTensor/2/float","fixture_out_type":"tuple","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":"/code/softmax-cross-entropy","census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"d380f15ca83de673"}},{"code_sha256":"9e76a013a5849311","code_sha256_prefix":"9e76a013a5849311","entry":"split_seed","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/pedal/utility.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/pedal/utility.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"signature":"seed:int/0/int, n:int/0/int","fixture_out_type":"tuple","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","retrieve_with":{"tool":"get_code","code_sha256":"9e76a013a5849311"},"code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"9e76a013a5849311"}},{"code_sha256":"c273c7f894008b87","code_sha256_prefix":"c273c7f894008b87","entry":"to_ids","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/datasets.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/datasets.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature_not_shown":true,"signature_not_shown_because":"signature_is_harness_object_repr","fixture_out_type":"ndarray","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"c273c7f894008b87"}},{"code_sha256":"7b5e4d5d64c189c4","code_sha256_prefix":"7b5e4d5d64c189c4","entry":"to_jsonable","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/pedal/utility.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/pedal/utility.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"obj:PosixPath/0/?","fixture_out_type":"str","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"7b5e4d5d64c189c4"}},{"code_sha256":"e30037d5501d3b17","code_sha256_prefix":"e30037d5501d3b17","entry":"to_str","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/datasets.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/datasets.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"signature":"terms:ndarray/1/int, vocab:tuple/1/str","fixture_out_type":"str","level_recorded_at":"2026-09-25T08:04:19.804293+00:00","code_page":"/code/to-str","census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"e30037d5501d3b17"}},{"code_sha256":"b910812684c6f413","code_sha256_prefix":"b910812684c6f413","entry":"context","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/pedal/xpu.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/pedal/xpu.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"b910812684c6f413"}},{"code_sha256":"f525cde07e4a95c4","code_sha256_prefix":"f525cde07e4a95c4","entry":"remove_keys","repo":"graphcore-research/unit-scaling-demo","repo_kind":"official","path":"scmm/pedal/utility.py","file_url":"https://github.com/graphcore-research/unit-scaling-demo/blob/HEAD/scmm/pedal/utility.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"code_page":null,"census":null,"licence":"MIT","inline_ok":true,"inline_ok_link":true,"withheld_clause":null,"mcp_get_code":{"code_sha256":"f525cde07e4a95c4"}}]},"code_links_note":null,"arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"ef294abfffe499b7c8e4ba2fae6f9c1c4c6f2efdbb4542ed12cc73bf9c2d8dc5","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}