# table-transformer-structure-recognition by Microsoft Source: https://savrn.com/models/table-transformer-structure-recognition Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve table-transformer-structure-recognition (29M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 0.1 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 23, 2026. [table-transformer-structure-recognition on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/table-transformer-structure-recognition/gpus) ## SAVRN's Notes on table-transformer-structure-recognition Once a table has been located on a page, something still has to work out its rows, columns and cells, and that is what this 29M-parameter detector from Microsoft does. It is a DETR-style detector trained on PubTables1M. The 16-bit run needs 0.1 GB and the 8-bit and 4-bit runs round to zero, so on the cheapest setup we price, one 192 GB MI300X at $1.85 an hour, it is a rounding error. It belongs alongside whatever pipeline feeds it, not on a card of its own. MIT is the license, which permits commercial use, modification and redistribution as long as the copyright and permission notices stay with the files, and access is open. Before committing, note that the weights are stored float32 in a 231 MB download last updated September 2023, and that the recorded context length is 1,024 tokens. The PubTables-1M paper, arXiv 2110.00061, describes the training data. SAVRN Research, 2026-09-18 ## Model Card By Microsoft, published under mit, revision f4d4bdc85c3f. Table Transformer (DETR) model trained on PubTables1M. It was introduced in the paper PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents by Smock et al. and first released in this repository. Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. The Table Transformer is equivalent to DETR, a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. You can use the raw model for detecting the structure (like rows, columns) in tables.… Read Microsoft's full model card ### Table Transformer (fine-tuned for Table Structure Recognition) Table Transformer (DETR) model trained on PubTables1M. It was introduced in the paper [PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents](https://savrn.com/papers/pubtables-1m-towards-comprehensive-table-extraction-from-unstructured-documents) by Smock et al. and first released in [this repository](https://github.com/microsoft/table-transformer). Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. ### Model description The Table Transformer is equivalent to [DETR](https://huggingface.co/docs/transformers/model_doc/detr), a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. ### Usage You can use the raw model for detecting the structure (like rows, columns) in tables. See the [documentation](https://huggingface.co/docs/transformers/main/en/model_doc/table-transformer) for more info. ## Configuration Architecture TableTransformerForObjectDetection Context length (tokens) 1,024 Layers 6 Stored precision float32 Model type table-transformer ## Identity and Version Repository microsoft/table-transformer-structure-recognition Publisher Microsoft Task Object detection Modality Image Library transformers Parameters 29M parameters Languages Not stated by the source Revision f4d4bdc85c3fe4b1fa49658882a5d38bbdd0f343 First published 2022-10-14 Last updated 2023-09-06 ## Files and Weights 6 files, 230.9 MB in total. The weights are 2 files totalling 230.9 MB in bin, safetensors. Weights2 files · 230.9 MB Configuration2 files · 1.7 KB Documentation1 file · 1.2 KB Repository1 file · 1.4 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | model.safetensors | Weights | 115.4 MB | f581da65a586 | | pytorch_model.bin | Weights | 115.5 MB | ec887aff5bf2 | | config.json | Configuration | 1.5 KB | — | | preprocessor_config.json | Configuration | 274 B | — | | README.md | Documentation | 1.2 KB | — | | .gitattributes | Repository | 1.4 KB | — | ## License and Download License mit Access Open weights, no gate Download size 230.9 MB [Download from Microsoft](https://huggingface.co/microsoft/table-transformer-structure-recognition) Released by Microsoft through its official repository on Hugging Face. [Read the license](https://opensource.org/license/mit). ## Built From - Described by [arXiv:2110.00061](https://savrn.com/papers/pubtables-1m-towards-comprehensive-table-extraction-from-unstructured-documents) ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 230.9 MB | | 16-bit | 0.1 GB | | 8-bit | 0.0 GB | | 4-bit | 0.0 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About table-transformer-structure-recognition ### How much GPU memory does table-transformer-structure-recognition need? About 0.1 GB at 16-bit and 0 GB at 4-bit: the weights (29M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run table-transformer-structure-recognition on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### Can I use table-transformer-structure-recognition commercially? Yes. table-transformer-structure-recognition is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included. ### What is table-transformer-structure-recognition's context length? 1,024 tokens, from the maximum position embeddings in its published configuration. ## Similar Models Model · Object detection ### [table-transformer-structure-recognition-v1.1-all](https://savrn.com/models/table-transformer-structure-recognition-v1-1-all) [Microsoft](https://savrn.com/model-publishers/microsoft) Table Transformer (TATR) model trained on PubTables1M and FinTabNet.c. It was introduced in the paper Aligning benchmark datasets for table structure recognition by Smock et al. and first released in this repository. Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. The Table Transformer is equivalent to DETR, a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. You can use the raw model for detecting tables in documents. See the documentation for more… Open weights mit 29M parameters transformers [View model](https://savrn.com/models/table-transformer-structure-recognition-v1-1-all) Model · Object detection ### [table-transformer-detection](https://savrn.com/models/table-transformer-detection) [Microsoft](https://savrn.com/model-publishers/microsoft) Table Transformer (DETR) model trained on PubTables1M. It was introduced in the paper PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents by Smock et al. and first released in this repository. Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. The Table Transformer is equivalent to DETR, a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. You can use the raw model for detecting tables in documents. See the documentation for… Open weights mit 29M parameters 1,024 tokens transformers [View model](https://savrn.com/models/table-transformer-detection) Model · Object detection ### [yolos-small](https://savrn.com/models/yolos-small) [HUST Vision Lab](https://savrn.com/model-publishers/hustvl) YOLOS model fine-tuned on COCO 2017 object detection (118k annotated images). It was introduced in the paper You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection by Fang et al. and first released in this repository. Disclaimer: The team releasing YOLOS did not write a model card for this model so this model card has been written by the Hugging Face team. YOLOS is a Vision Transformer (ViT) trained using the DETR loss. Despite its simplicity, a base-sized YOLOS model is able to achieve 42 AP on COCO validation 2017 (similar to DETR and more complex frameworks such as Faster R-CNN). The model is trained using a "bipartite matching loss": one compares the… Open weights apache-2.0 31M parameters transformers [View model](https://savrn.com/models/yolos-small) Model · Object detection ### [rf-detr-base](https://savrn.com/models/rf-detr-base) [Roboflow](https://savrn.com/model-publishers/roboflow) RF-DETR is a real-time detection transformer family introduced in RF-DETR: Neural Architecture Search for Real-Time Detection Transformers by Robinson et al. and integrated in Transformers via PR #36895. RF-DETR is an end-to-end object detection model that combines ideas from LW-DETR and Deformable DETR: a DINOv2-with-registers style ViT backbone (with an RF-DETR windowing pattern for efficient attention), a multi-scale projector between encoder and decoder, and a multi-scale deformable DETR decoder for fast convergence and strong accuracy–latency tradeoffs. You can use the raw model for object detection. See the model hub to look for all available RF-DETR models. Here is how to use this… Open weights apache-2.0 32M parameters transformers [View model](https://savrn.com/models/rf-detr-base) Model · Object detection ### [PP-DocLayoutV3_safetensors](https://savrn.com/models/pp-doclayoutv3-safetensors) [PaddlePaddle](https://savrn.com/model-publishers/paddlepaddle) Unified Layout Module for PaddleOCR-VL 1.5/1.6 & GLM-OCR & & Xiaomi-OCR-0. This is the model weights for PP-DocLayoutv3 in safetensors format. Get PaddlePaddle weights at PP-DocLayoutV3 PP-DocLayoutV3 is specifically engineered to handle non-planar document images. It can directly predict multi-point bounding boxes for layout elements—as opposed to standard two-point boxes—and determine logical reading orders for skewed and curved surfaces within a single forward pass, significantly reducing cascading errors. This model is an essential component of PaddleOCR-VL-1.5, providing crucial layout analysis for the high-precision parsing of various real-world documents in PaddleOCR-VL. This work… Open weights apache-2.0 33M parameters transformers [View model](https://savrn.com/models/pp-doclayoutv3-safetensors) Model · Object detection ### [rf-detr-large](https://savrn.com/models/rf-detr-large) [Roboflow](https://savrn.com/model-publishers/roboflow) RF-DETR is a real-time detection transformer family introduced in RF-DETR: Neural Architecture Search for Real-Time Detection Transformers by Robinson et al. and integrated in Transformers via PR #36895. RF-DETR is an end-to-end object detection model that combines ideas from LW-DETR and Deformable DETR: a DINOv2-with-registers style ViT backbone (with an RF-DETR windowing pattern for efficient attention), a multi-scale projector between encoder and decoder, and a multi-scale deformable DETR decoder for fast convergence and strong accuracy–latency tradeoffs. You can use the raw model for object detection. See the model hub to look for all available RF-DETR models. Here is how to use this… Open weights apache-2.0 34M parameters transformers [View model](https://savrn.com/models/rf-detr-large) ## Microsoft [All models and datasets](https://savrn.com/model-publishers/microsoft) ## Versions - [f4d4bdc85c3f](https://savrn.com/models/table-transformer-structure-recognition/versions/f4d4bdc85c3f) · current 2026-10-23 ## Explore More - [All object detection models](https://savrn.com/models/tasks/object-detection) - [All models under mit](https://savrn.com/models/licenses/mit) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-19. - [Microsoft website](https://github.com/microsoft) - [Hugging Face record](https://huggingface.co/microsoft/table-transformer-structure-recognition) - [How the hub is built](https://savrn.com/model-hub/methodology)