rewirebio.iobenchmarks
Source

RFD4-Proteina Hugging Face model card at revision 621b785

Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030 · Project source

Original source

What this source contributed

Records in this release that cite this source. A record can cite several sources.

Results

Values as the source printed them. Compare results only under compatible protocols, datasets and metrics.

Model or configurationMetricValueDatasetEvaluation
AlphaProtein-novo (Codesign), Atomic Motif EnzymePassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: protein-backbone RMSD at most 2 A, required motif all-atom RMSD at most 1.5 A, and AF3 minimum interface PAE below 1.5 A)26.02 count-per-hourFrozen project benchmark Atomic Motif Enzyme target set, 37 tasksAlphaProtein-novo (Codesign) on Frozen project benchmark Atomic Motif Enzyme, fixed ligand: distinct passers per GPU-hour
BoltzGen, frozen project benchmark recipePassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: AF3 minimum interface PAE at most 1.5 A and DNA-target-aligned binder backbone RMSD at most 8 A, aligning on all target DNA atoms in both the fixed and unfixed modes)9.46 count-per-hourFrozen project benchmark DNA-binding target set, 3 targetsBoltzGen (fixed DNA) on Frozen project benchmark DNA binding, fixed DNA: distinct passers per GPU-hour
BoltzGen, frozen project benchmark recipePassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: AF3 minimum interface PAE at most 1.5 A, chain-of-interest ipTM at least 80 out of 100, protein-backbone RMSD at most 1.5 A, and protein-backbone-aligned ligand RMSD at most 5 A (the sm_binder criterion, not the looser historical disco criterion))15.48 count-per-hourFrozen project benchmark small-molecule binding target set, 77 targetsBoltzGen on Frozen project benchmark small-molecule binding, unfixed ligand: distinct passers per GPU-hour
Proteina-Complexa (V1), protein binding, all-atom conditioningPassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: AF3 binder pTM at least 80 on the native 0 to 100 scale, minimum interface PAE at most 1.5 A, and target-backbone-aligned binder backbone RMSD at most 2.5 A, with valid required target mapping and conditioning)13.24 count-per-hourFrozen project benchmark protein-binding target set, 24 targetsProteina-Complexa on Frozen project benchmark protein binding, all-atom conditioning: distinct passers per GPU-hour
RFD4-Proteina (high thinking), previous release pairPassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: protein-backbone RMSD at most 2 A, required motif all-atom RMSD at most 1.5 A, and AF3 minimum interface PAE below 1.5 A)91.18 count-per-hourFrozen project benchmark Atomic Motif Enzyme target set, 37 tasksRFD4-Proteina, high thinking on Frozen project benchmark Atomic Motif Enzyme, fixed ligand: distinct passers per GPU-hour
RFD4-Proteina (high thinking), previous release pairPassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: AF3 minimum interface PAE at most 1.5 A and DNA-target-aligned binder backbone RMSD at most 8 A, aligning on all target DNA atoms in both the fixed and unfixed modes)41.97 count-per-hourFrozen project benchmark DNA-binding target set, 3 targetsRFD4-Proteina, high thinking on Frozen project benchmark DNA binding, unfixed DNA: distinct passers per GPU-hour
RFD4-Proteina (high thinking), previous release pairPassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: AF3 binder pTM at least 80 on the native 0 to 100 scale, minimum interface PAE at most 1.5 A, and target-backbone-aligned binder backbone RMSD at most 2.5 A, with valid required target mapping and conditioning)71.60 count-per-hourFrozen project benchmark protein-binding target set, 24 targetsRFD4-Proteina, high thinking on Frozen project benchmark protein binding, backbone conditioning: distinct passers per GPU-hour
RFD4-Proteina (high thinking), previous release pairPassing design clusters per accelerator hour (mean over targets of structurally distinct computational passers per GPU-hour, counting generation and common AlphaFold3 evaluation time, with 200 original attempts per target and failed attempts included; passing designs clustered within each model, protocol and target by Foldseek at TM 0.5; pass criterion: AF3 minimum interface PAE at most 1.5 A, chain-of-interest ipTM at least 80 out of 100, protein-backbone RMSD at most 1.5 A, and protein-backbone-aligned ligand RMSD at most 5 A (the sm_binder criterion, not the looser historical disco criterion))112.25 count-per-hourFrozen project benchmark small-molecule binding target set, 77 targetsRFD4-Proteina, high thinking on Frozen project benchmark small-molecule binding, unfixed ligand: distinct passers per GPU-hour

Claims

Statements this source supports, each with where in the source it was checked.

  • RFD4-Proteina (high thinking), previous release pair: checkpoint identity of the pair the efficiency table was measured on: Checkpoint identity The efficiency table's scope note names only 'the previous release pair'. The card's checkpoint table gives a 'Previous release' column with flow checkpoint chk_epoch_00000169_step_000000340000-ema, autoencoder 2026_09_04_release_ae_z16_400k-ema, latent dimension 16, 33 residue classes and 434,348,737 stored parameters. The same card separately says the 2026-10-19 pair supersedes a 2026-10-15 pair. Neither pair is tied to the table, and the benchmark methods supplement says each cohort's exact schedule and checkpoint come from its own native config, which the packet does not publish. No checkpoint filename is assigned to the table's rows in this batch. (Hugging Face model card, 'Measured design efficiency' blockquote above the table, and 'Checkpoints in this release' table column 'Previous release'; benchmarks/METHODS.md, 'Model recipes and adaptations' row RFD4-Proteina)
  • RFD4-Proteina 2026-10-19 release pair: flow-model training compute: Training compute The preprint states the release flow model was pretrained with 256-token crops to 320,000 optimizer steps using four independently noised candidates per example, requiring about 16,000 GPU-hours, and then fine-tuned for about 2,000 GPU-hours with low-rank adaptation at rank 32, scaling factor 64 and no adapter dropout, over 60,000 additional updates with 384-token padding. Both flow stages use the same frozen autoencoder. (Preprint version 3, Section S4.2 'Training stages' and Table S12 'Flow-model training stages')
  • RFD4-Proteina 2026-10-19 release pair: parameters: Parameters 435,622,465 stored parameters across the paired bundle, in FP32. model.safetensors is 1.74 GB and the cached release bundle is about 1.7 GB including its autoencoder. The published EMA exports are 1.74 GB and 1.03 GB rather than 4.21 GB and 3.10 GB because the duplicate shadow copy, optimizer, scheduler and loop state are removed; their weight tensors are stated to be bit-identical to the full training checkpoints. (Hugging Face model card, 'At a glance' rows 'Weights' and 'Download', and the paragraph beginning 'The published .ckpt files'; release-checkpoints.json, key 'stored_parameters' and the size fields)
  • RFD4-Proteina 2026-10-19 release pair: released checkpoint pair identity: Checkpoint identity The current release pair is rfd4-proteina-flow-release-2026-10-19.ckpt, the pretrained base at step 320000 with a rank-32 LoRA fine-tune merged in, and rfd4-proteina-ae-release-2026-10-19.ckpt, a unified protein, RNA, DNA and modified-residue autoencoder with 32 latent channels per residue trained to 400k steps. Both hold exponential-moving-average weights. The pair supersedes the 2026-10-15 pair 2026_10_15_hero_s43_c512_80k-ema and 2026_10_15_unified_ae_s43_400k-ema, which has the same architecture. The published files are the stable names flow_checkpoints/rfd4_proteina_flow-ema.ckpt with SHA-256 c1c8b7068e87c41a334ec5b77e7d7cdda30bd31ed4504909d93bd72f7742afe4 and ae_checkpoints/rfd4_proteina_ae-ema.ckpt with SHA-256 24536372f3fc9a5083b6b813baa80630d2287a5d61fbf2f126c9735802e1d21a. (Hugging Face model card, 'Checkpoints in this release' table, column 'Current release (2026-10-19)', and the paragraph beginning 'The 2026-10-19 pair supersedes'; release-checkpoints.json, key 'checkpoints'; checkpoints reference, 'Released checkpoints' table)
  • RFD4-Proteina frozen project benchmark, October 9 2026 supplement: attempt accounting: Attempt accounting Each model, protocol and target has 200 original attempts, including native failures and unmatched required motifs. Rejected originals are not replaced and outputs are not selected with the common AlphaFold3 evaluation. Failed attempts consume time and stay included in the efficiency denominator, while setup, compilation, CPU scoring and idle allocation time are excluded. Values are means of target-level rates, not pooled totals or single-design latency. (Hugging Face model card, 'Measured design efficiency', paragraph beginning 'There are 200 original attempts per target'; benchmarks/METHODS.md, 'Common accounting and timing')
  • RFD4-Proteina frozen project benchmark, October 9 2026 supplement: conditioning differences: Conditioning differences Conditioning differs between some methods: PPI uses RFD4 backbone conditioning versus all-atom conditioning for several baselines; DNA compares unfixed RFD4 inputs with fixed-DNA baselines; Proteina-Complexa's small-molecule route uses fixed ligands. Only three DNA targets are included. Computational passers do not establish experimental activity. (Hugging Face model card, 'Measured design efficiency', paragraph beginning 'Conditioning differs between some methods')
  • RFD4-Proteina frozen project benchmark, October 9 2026 supplement: efficiency table scope: Efficiency table scope The card prints above the efficiency table: 'These numbers were measured on the previous release pair, in the October 9, 2026 frozen project benchmark, which predates the current checkpoints. They have not been re-measured on the current release and should not be read as describing it.' The provenance file gives the packet's source release as supplement-20261009 from paper commit 8614542c149875b13ed3fed5d8d6ec739f2daff4, and its scope as 'Frozen project benchmark runs; not a new evaluation of every model Hub revision.' (Hugging Face model card, 'Measured design efficiency' blockquote; benchmarks/provenance.json, keys source_release, paper_commit and scope)
  • RFD4-Proteina frozen project benchmark, October 9 2026 supplement: metric definition: Metric definition The card defines the table's metric as 'mean per-target structurally distinct computational passers per GPU-hour, including generation and common AF3 evaluation', higher being better, and adds that computational passers do not establish experimental activity. The provenance file gives the expression mean_over_targets(3600 * passing_structure_clusters / (generation_gpu_s + af3_gpu_s)). The methods supplement defines a passing cluster as a distinct Foldseek cluster among passing originals, computed within each model, protocol and target at TM threshold 0.5, and says these are algorithmic structural clusters, not experimental successes or globally distinct proteins. (Hugging Face model card, 'Measured design efficiency' metric line and the paragraph below the table; benchmarks/provenance.json, key metric; benchmarks/METHODS.md, final paragraph of 'Common AF3 scoring')
  • RFD4-Proteina: developers: Developers Developed by the Baker Lab at the Institute for Protein Design, University of Washington, and NVIDIA GenAIR. The preprint's corresponding authors are Kieran Didi and David Baker, with four joint first authors. (Hugging Face model card, 'Attribution and support'; documentation index, hero paragraph; preprint version 3, title page and corresponding-author footnote)
  • RFD4-Proteina: hardware requirements: Hardware requirements Linux x86_64, Python 3.12 and a CUDA 13-compatible driver are required; macOS and Windows are not supported. The card lists the NVIDIA GPUs it says are tested and confirmed: Blackwell B200, B300, GB200, GB300, RTX PRO 6000 Blackwell Server Edition, RTX PRO 6000 Blackwell Workstation Edition and GB10 (DGX Spark); Hopper H100 SXM, H100 PCIe, H100 NVL, H200, H200 NVL and GH200 144GB; Ada Lovelace L40S and RTX 6000 Ada Generation; Ampere A100 SXM4 40GB and 80GB and A100 PCIe 40GB and 80GB. Peak GPU memory is said to depend on token count, batch size and recycle depth rather than on the weights. (Hugging Face model card, 'Hardware requirements' table and the paragraph below it; repository README, 'System requirements')
  • RFD4-Proteina: input requirements: Input requirements The annotations on an input specify what to hold fixed and what to design. The card states that an ordinary, unannotated structure is not a complete design specification. (Hugging Face model card, 'Start with one design', step 3, final sentence)
  • RFD4-Proteina: inputs: Inputs Annotated CIF structures or text-based YAML or JSON design specifications. The annotations carry the conditions; the repository README lists fixed motifs, hotspots, burial, hydrogen bonds and more. (Hugging Face model card, 'At a glance' table row 'Inputs'; repository README, 'Run the model' and key features)
  • RFD4-Proteina: matched checkpoint pair: Matched checkpoint pair Generation always loads two checkpoints together: a flow-matching denoiser and the autoencoder it was trained against. The release is a matched pair and the components are not interchangeable across releases, because the autoencoder defines the latent space the flow model generates in; the card states that mixing a flow model with a different autoencoder produces nonsense rather than an error. The 2026-10-19 release changes the latent dimension from 16 to 32 and the residue vocabulary from 33 to 147 classes, which is stated to be what makes the two releases incompatible with each other's components. An optional PEFT adapter must be loaded onto the exact base denoiser and autoencoder pair it was fine-tuned against. (Hugging Face model card, 'Checkpoints in this release' opening paragraph and the paragraph on the unified autoencoder; release-checkpoints.json, key 'policy'; checkpoints reference, 'Types of checkpoints' opening sentence and table rows for the autoencoder and the PEFT adapter)
  • RFD4-Proteina: outputs: Outputs Designed molecular structures in CIF format, with generated residue identities and coordinates written to the run's inference_samples directory. (Hugging Face model card, 'At a glance' table row 'Outputs'; repository README, 'Run the model')
  • RFD4-Proteina: runtime packages: Runtime packages The native bundle in nvidia/RFD4-Proteina and the Diffusers pipeline in nvidia/RFD4-Proteina-Diffusers are the same model: the card states it is one model throughout, nothing is retrained, and the Diffusers weights are the same tensors re-prefixed and split across transformer/ and vae/. The separate repository exists because Diffusers requires model_index.json at the root. Both hold the same 2026-10-19 release. (Hugging Face model card, 'Where to find each entry point' table and the two paragraphs below it)
  • RFD4-Proteina: sampling presets: Sampling presets The released recipes differ only in how often the trunk is recycled per sampling step, as bands over 60 steps: low [[10,4],[10,3],[10,2],[30,1]] for 120 trunk passes, medium [[5,6],[10,5],[5,4],[10,3],[20,2],[10,1]] for 180 trunk passes and the default, and high [[10,9],[10,8],[10,5],[20,3],[10,2]] for 300 trunk passes. The native runtime also offers flat and mid-peak schedules for each budget. (Hugging Face model card, 'Compute presets' table and the two sentences around it)
  • RFD4-Proteina: stated lineage from RFdiffusion and Proteina: Lineage The RosettaCommons post describes RFD4-Proteina as the next generation of RFdiffusion from the Institute for Protein Design, developed in collaboration with NVIDIA, building on previous models including RFdiffusion, RFdiffusion2 and RFdiffusion3 and incorporating advances from NVIDIA's Proteina model. It names a broader collaboration with Rosetta Commons, the Open Rosetta project at OMSF and the OpenFold Consortium, and records that the release coincides with a formal release of TMol and relies on a newly released version of AtomWorks. (RosettaCommons post, second paragraph and the section 'Collaborations Across Academia and Industry Lead to Scientific Advancements in Protein Design')
  • RFD4-Proteina: training data terms: Training data terms The card states the training data has its own source terms, described on the companion Hugging Face dataset card, and that the dataset's upload date or proposed sampling mix should not be interpreted as the exact training history of every checkpoint. Training datasets are a separate download from the model weights and are not needed for generation. (Hugging Face model card, 'Training data and limitations' and 'License' sections; repository README, 'Hugging Face and Python')
  • RFD4-Proteina: what the BSD 3-Clause licence covers: Licence The BSD 3-Clause licence, Copyright (c) 2026, Institute for Protein Design, University of Washington, covers the RFD4-Proteina model, weights and code only. The optional AlphaFold2 and OpenFold3 or AlphaFold3 refolding container images package third-party software that stays under its own upstream licences, some of which the card says restrict commercial use or redistribution; the ported OpenFold3 weights are stated to be Apache-2.0. The card says each upstream licence must be reviewed before using the images and not to assume the BSD licence applies to them. Neither refolding oracle is needed for generation. (Hugging Face model card, 'License' section and the 'Other repository files' details block; LICENSE file, whole file)

Models, benchmarks and other records

Models (1)

Configurations (6)

Benchmarks (1)

Protocols (6)

Evaluations (8)

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

28 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-11-21-c8bea25e5abb
Property and statementOriginal source and locationReview and provenance
Archive member
artifacts/hf_README.md.gz
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.archive_member

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Archive sha256
685e0d82607b569a16e106afb9c69e34139a56cedba3f053cf45cc59722894c8
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.archive_sha256

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Artifact SHA-256
38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.artifact_sha256

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Licence
BSD 3-Clause for the RFD4-Proteina model, weights and code; optional third-party assets keep their own upstream terms
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.licence

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Limitations
2 values
  • The card states that the efficiency table was measured on the previous release pair in the October 9, 2026 frozen project benchmark and has not been re-measured on the 2026-10-19 release.
  • Author-published; no independent evaluation.
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.limitations

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Media type
text/markdown
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.media_type

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Publication status
official_project_source
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.publication_status

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Retrieval url
https://huggingface.co/nvidia/RFD4-Proteina/raw/621b785abf9d61dd24f4b81ae4dbd1330728f030/README.md
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.retrieval_url

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Retrieved at
2026-11-15T23:10:00Z
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.retrieved_at

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

Review: artifact sha256
38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98
Source metadata
RFD4-Proteina Hugging Face model card at revision 621b785

Original source ↗

No field-specific location recorded

Version: Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Retrieved: 2026-11-15T23:10:00Z

catalogued

No individual claim review recorded

Audit details

Field: attributes.review.artifact_sha256

Source artifact SHA-256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98

Hash scope: Hash scope not separately documented; inspect source record

Archive member: artifacts/hf_README.md.gz

How to read the evidence

Sources and history

Release 2026-11-21-c8bea25e5abb · Record review: source checked

The model card of the nvidia/RFD4-Proteina weight repository, pinned to one commit, which carries the release checkpoint pair, the compute presets, the licence scope and the measured design efficiency table.

Release history

No supporting source is linked yet.

Read original source

Download this release (gzip)
Technical metadata and extraction receipts

Stable ID: rfd4-proteina-20261115-source-hf-model-card

areas
proteins-complexes; protein-structure; molecular-interactions
contexts
research
archive member
artifacts/hf_README.md.gz
archive sha256
685e0d82607b569a16e106afb9c69e34139a56cedba3f053cf45cc59722894c8
artifact sha256
38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98
licence
BSD 3-Clause for the RFD4-Proteina model, weights and code; optional third-party assets keep their own upstream terms
limitations
The card states that the efficiency table was measured on the previous release pair in the October 9, 2026 frozen project benchmark and has not been re-measured on the 2026-10-19 release.; Author-published; no independent evaluation.
media type
text/markdown
publication status
official_project_source
retrieval url
https://huggingface.co/nvidia/RFD4-Proteina/raw/621b785abf9d61dd24f4b81ae4dbd1330728f030/README.md
retrieved at
2026-11-15T23:10:00Z
review
artifact sha256: 38479d8e4c6e46ed7e1ec4ece0dd02c358ae46d73d56790c44be813201fd9b98; date: 2026-11-15; method: source-hash-verification; automated-source-review; note: Retrieved twice on 2026-11-15 with identical bytes; the hash is of the retrieved artifact, not of a text layer.; retrieval url: https://huggingface.co/nvidia/RFD4-Proteina/raw/621b785abf9d61dd24f4b81ae4dbd1330728f030/README.md; reviewer: claude; reviewer note: Claude (Opus 5) extraction agent; no human review claimed. Extraction only: a different agent reviews this batch, and nothing here is reproduced.
source locator
Sections 'Where to find each entry point', 'At a glance', 'Checkpoints in this release', 'Compute presets', 'Hardware requirements', 'Measured design efficiency', 'Training data and limitations', 'Other repository files' and 'License'
url
https://huggingface.co/nvidia/RFD4-Proteina/blob/621b785abf9d61dd24f4b81ae4dbd1330728f030/README.md
version
Hugging Face revision 621b785abf9d61dd24f4b81ae4dbd1330728f030
Related records

Suggest a correction