Feature Extraction
Transformers
TensorBoard
Safetensors
English
captionbert_structured_rotary
sentence-similarity
consensus-distillation
rotary-embeddings
custom_code
Instructions to use AbstractPhil/captionbert-8192-rotary with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use AbstractPhil/captionbert-8192-rotary with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="AbstractPhil/captionbert-8192-rotary", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("AbstractPhil/captionbert-8192-rotary", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Publish full structured rotary report, final model card, all results and tested serving release
e349b70 verified |
Download figures/structured_larger.svg from AbstractPhil/captionbert-8192-rotary: direct link, hf CLI and curl.
- Browser
- Download file 31.5 kB
-
https://huggingface.co/AbstractPhil/captionbert-8192-rotary/resolve/main/figures/structured_larger.svg
- Command line
-
hf download hf://AbstractPhil/captionbert-8192-rotary/figures/structured_larger.svg
-
curl -L -o structured_larger.svg https://huggingface.co/AbstractPhil/captionbert-8192-rotary/resolve/main/figures/structured_larger.svg
31.5 kB
| <svg xmlns:xlink="http://www.w3.org/1999/xlink" width="921.6pt" height="338.4pt" viewBox="0 0 921.6 338.4" xmlns="http://www.w3.org/2000/svg" version="1.1"> | |
| <metadata> | |
| <rdf:RDF xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:cc="http://creativecommons.org/ns#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"> | |
| <cc:Work> | |
| <dc:type rdf:resource="http://purl.org/dc/dcmitype/StillImage"/> | |
| <dc:format>image/svg+xml</dc:format> | |
| <dc:creator> | |
| <cc:Agent> | |
| <dc:title>Matplotlib v3.11.2, https://matplotlib.org/</dc:title> | |
| </cc:Agent> | |
| </dc:creator> | |
| </cc:Work> | |
| </rdf:RDF> | |
| </metadata> | |
| <defs> | |
| <style type="text/css">*{stroke-linejoin: round; stroke-linecap: butt}</style> | |
| </defs> | |
| <g id="figure_1"> | |
| <g id="patch_1"> | |
| <path d="M 0 338.4 | |
| L 921.6 338.4 | |
| L 921.6 0 | |
| L 0 0 | |
| z | |
| " style="fill: #ffffff"/> | |
| </g> | |
| <g id="axes_1"> | |
| <g id="patch_2"> | |
| <path d="M 196.63 262.437656 | |
| L 535.066975 262.437656 | |
| L 535.066975 108.488 | |
| L 196.63 108.488 | |
| z | |
| " style="fill: #ffffff"/> | |
| </g> | |
| <g id="matplotlib.axis_1"> | |
| <g id="xtick_1"> | |
| <g id="line2d_1"> | |
| <path d="M 241.762912 262.437656 | |
| L 241.762912 108.488 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_2"> | |
| <defs> | |
| <path id="m3747435316" d="M 0 0 | |
| L 0 3.5 | |
| " style="stroke: #000000; stroke-width: 0.8"/> | |
| </defs> | |
| <g> | |
| <use xlink:href="#m3747435316" x="241.762912" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_1"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="241.762912" y="277.035313" transform="rotate(-0 241.762912 277.035313)">0.678</text> | |
| </g> | |
| </g> | |
| <g id="xtick_2"> | |
| <g id="line2d_3"> | |
| <path d="M 289.994722 262.437656 | |
| L 289.994722 108.488 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_4"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="289.994722" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_2"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="289.994722" y="277.035313" transform="rotate(-0 289.994722 277.035313)">0.680</text> | |
| </g> | |
| </g> | |
| <g id="xtick_3"> | |
| <g id="line2d_5"> | |
| <path d="M 338.226531 262.437656 | |
| L 338.226531 108.488 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_6"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="338.226531" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_3"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="338.226531" y="277.035313" transform="rotate(-0 338.226531 277.035313)">0.682</text> | |
| </g> | |
| </g> | |
| <g id="xtick_4"> | |
| <g id="line2d_7"> | |
| <path d="M 386.458341 262.437656 | |
| L 386.458341 108.488 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_8"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="386.458341" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_4"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="386.458341" y="277.035313" transform="rotate(-0 386.458341 277.035313)">0.684</text> | |
| </g> | |
| </g> | |
| <g id="xtick_5"> | |
| <g id="line2d_9"> | |
| <path d="M 434.69015 262.437656 | |
| L 434.69015 108.488 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_10"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="434.69015" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_5"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="434.69015" y="277.035313" transform="rotate(-0 434.69015 277.035313)">0.686</text> | |
| </g> | |
| </g> | |
| <g id="xtick_6"> | |
| <g id="line2d_11"> | |
| <path d="M 482.92196 262.437656 | |
| L 482.92196 108.488 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_12"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="482.92196" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_6"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="482.92196" y="277.035313" transform="rotate(-0 482.92196 277.035313)">0.688</text> | |
| </g> | |
| </g> | |
| <g id="xtick_7"> | |
| <g id="line2d_13"> | |
| <path d="M 531.153769 262.437656 | |
| L 531.153769 108.488 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_14"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="531.153769" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_7"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="531.153769" y="277.035313" transform="rotate(-0 531.153769 277.035313)">0.690</text> | |
| </g> | |
| </g> | |
| <g id="text_8"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="365.848487" y="291.036094" transform="rotate(-0 365.848487 291.036094)">STS-B validation Spearman</text> | |
| </g> | |
| </g> | |
| <g id="matplotlib.axis_2"> | |
| <g id="ytick_1"> | |
| <g id="line2d_15"> | |
| <defs> | |
| <path id="m347ebb820c" d="M 0 0 | |
| L -3.5 0 | |
| " style="stroke: #000000; stroke-width: 0.8"/> | |
| </defs> | |
| <g> | |
| <use xlink:href="#m347ebb820c" x="196.63" y="139.131524" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_9"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: end" x="189.63" y="142.930743" transform="rotate(-0 189.63 142.930743)">4M rows × 4 epochs / hierarchical</text> | |
| </g> | |
| </g> | |
| <g id="ytick_2"> | |
| <g id="line2d_16"> | |
| <g> | |
| <use xlink:href="#m347ebb820c" x="196.63" y="185.462828" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_10"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: end" x="189.63" y="189.262047" transform="rotate(-0 189.63 189.262047)">4M rows × 4 epochs / relative</text> | |
| </g> | |
| </g> | |
| <g id="ytick_3"> | |
| <g id="line2d_17"> | |
| <g> | |
| <use xlink:href="#m347ebb820c" x="196.63" y="231.794132" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_11"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: end" x="189.63" y="235.593351" transform="rotate(-0 189.63 235.593351)">1M rows × 16 epochs / hierarchical</text> | |
| </g> | |
| </g> | |
| </g> | |
| <g id="PathCollection_1"> | |
| <defs> | |
| <path id="m8bb6c8efa8" d="M 0 6.123724 | |
| L 0 -6.123724 | |
| " style="stroke: #0072b2; stroke-width: 1.8"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#m8bb6c8efa8" x="428.566511" y="139.131524" style="fill: #0072b2; stroke: #0072b2; stroke-width: 1.8"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_2"> | |
| <defs> | |
| <path id="me363e187c4" d="M 0 6.123724 | |
| L 0 -6.123724 | |
| " style="stroke: #009e73; stroke-width: 1.8"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#me363e187c4" x="447.327514" y="185.462828" style="fill: #009e73; stroke: #009e73; stroke-width: 1.8"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_3"> | |
| <defs> | |
| <path id="m620e92aa57" d="M 0 6.123724 | |
| L 0 -6.123724 | |
| " style="stroke: #64748b; stroke-width: 1.8"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#m620e92aa57" x="302.770012" y="231.794132" style="fill: #64748b; stroke: #64748b; stroke-width: 1.8"/> | |
| </g> | |
| </g> | |
| <g id="line2d_18"> | |
| <path d="M 361.116467 139.131524 | |
| L 496.016555 139.131524 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #0072b2; stroke-opacity: 0.6; stroke-width: 1.5; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_19"> | |
| <path d="M 451.971098 185.462828 | |
| L 442.68393 185.462828 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #009e73; stroke-opacity: 0.6; stroke-width: 1.5; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_20"> | |
| <path d="M 369.859604 231.794132 | |
| L 235.68042 231.794132 | |
| " clip-path="url(#pe21196a7d8)" style="fill: none; stroke: #64748b; stroke-opacity: 0.6; stroke-width: 1.5; stroke-linecap: square"/> | |
| </g> | |
| <g id="patch_3"> | |
| <path d="M 196.63 262.437656 | |
| L 196.63 108.488 | |
| " style="fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square"/> | |
| </g> | |
| <g id="patch_4"> | |
| <path d="M 196.63 262.437656 | |
| L 535.066975 262.437656 | |
| " style="fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square"/> | |
| </g> | |
| <g id="PathCollection_4"> | |
| <defs> | |
| <path id="md3c1ca1a6a" d="M 0 3.354102 | |
| C 0.889518 3.354102 1.742724 3.000693 2.371708 2.371708 | |
| C 3.000693 1.742724 3.354102 0.889518 3.354102 0 | |
| C 3.354102 -0.889518 3.000693 -1.742724 2.371708 -2.371708 | |
| C 1.742724 -3.000693 0.889518 -3.354102 0 -3.354102 | |
| C -0.889518 -3.354102 -1.742724 -3.000693 -2.371708 -2.371708 | |
| C -3.000693 -1.742724 -3.354102 -0.889518 -3.354102 0 | |
| C -3.354102 0.889518 -3.000693 1.742724 -2.371708 2.371708 | |
| C -1.742724 3.000693 -0.889518 3.354102 0 3.354102 | |
| z | |
| " style="stroke: #0072b2"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#md3c1ca1a6a" x="361.116467" y="136.11999" style="fill: #0072b2; stroke: #0072b2"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_5"> | |
| <defs> | |
| <path id="m82ad91a55d" d="M -3.354102 3.354102 | |
| L 3.354102 3.354102 | |
| L 3.354102 -3.354102 | |
| L -3.354102 -3.354102 | |
| z | |
| " style="stroke: #0072b2"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#m82ad91a55d" x="496.016555" y="142.143059" style="fill: #0072b2; stroke: #0072b2"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_6"> | |
| <defs> | |
| <path id="me3968be686" d="M 0 3.354102 | |
| C 0.889518 3.354102 1.742724 3.000693 2.371708 2.371708 | |
| C 3.000693 1.742724 3.354102 0.889518 3.354102 0 | |
| C 3.354102 -0.889518 3.000693 -1.742724 2.371708 -2.371708 | |
| C 1.742724 -3.000693 0.889518 -3.354102 0 -3.354102 | |
| C -0.889518 -3.354102 -1.742724 -3.000693 -2.371708 -2.371708 | |
| C -3.000693 -1.742724 -3.354102 -0.889518 -3.354102 0 | |
| C -3.354102 0.889518 -3.000693 1.742724 -2.371708 2.371708 | |
| C -1.742724 3.000693 -0.889518 3.354102 0 3.354102 | |
| z | |
| " style="stroke: #009e73"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#me3968be686" x="451.971098" y="182.451293" style="fill: #009e73; stroke: #009e73"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_7"> | |
| <defs> | |
| <path id="m82f90f8eea" d="M -3.354102 3.354102 | |
| L 3.354102 3.354102 | |
| L 3.354102 -3.354102 | |
| L -3.354102 -3.354102 | |
| z | |
| " style="stroke: #009e73"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#m82f90f8eea" x="442.68393" y="188.474363" style="fill: #009e73; stroke: #009e73"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_8"> | |
| <defs> | |
| <path id="mc40f734263" d="M 0 3.354102 | |
| C 0.889518 3.354102 1.742724 3.000693 2.371708 2.371708 | |
| C 3.000693 1.742724 3.354102 0.889518 3.354102 0 | |
| C 3.354102 -0.889518 3.000693 -1.742724 2.371708 -2.371708 | |
| C 1.742724 -3.000693 0.889518 -3.354102 0 -3.354102 | |
| C -0.889518 -3.354102 -1.742724 -3.000693 -2.371708 -2.371708 | |
| C -3.000693 -1.742724 -3.354102 -0.889518 -3.354102 0 | |
| C -3.354102 0.889518 -3.000693 1.742724 -2.371708 2.371708 | |
| C -1.742724 3.000693 -0.889518 3.354102 0 3.354102 | |
| z | |
| " style="stroke: #64748b"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#mc40f734263" x="369.859604" y="228.782597" style="fill: #64748b; stroke: #64748b"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_9"> | |
| <defs> | |
| <path id="me3d60a9e43" d="M -3.354102 3.354102 | |
| L 3.354102 3.354102 | |
| L 3.354102 -3.354102 | |
| L -3.354102 -3.354102 | |
| z | |
| " style="stroke: #64748b"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#me3d60a9e43" x="235.68042" y="234.805667" style="fill: #64748b; stroke: #64748b"/> | |
| </g> | |
| </g> | |
| <g id="text_12"> | |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="365.848487" y="102.488" transform="rotate(-0 365.848487 102.488)">Development comparison</text> | |
| </g> | |
| <g id="PathCollection_10"> | |
| <defs> | |
| <path id="m501ecd2794" d="M 0 -6.22495 | |
| L -1.397588 -1.923615 | |
| L -5.920279 -1.923615 | |
| L -2.261345 0.734756 | |
| L -3.658934 5.03609 | |
| L -0 2.377719 | |
| L 3.658934 5.03609 | |
| L 2.261345 0.734756 | |
| L 5.920279 -1.923615 | |
| L 1.397588 -1.923615 | |
| z | |
| " style="stroke: #334155; stroke-width: 0.6"/> | |
| </defs> | |
| <g clip-path="url(#pe21196a7d8)"> | |
| <use xlink:href="#m501ecd2794" x="451.971098" y="182.451293" style="fill: #f59e0b; stroke: #334155; stroke-width: 0.6"/> | |
| </g> | |
| </g> | |
| </g> | |
| <g id="axes_2"> | |
| <g id="patch_5"> | |
| <path d="M 572.363025 262.437656 | |
| L 910.8 262.437656 | |
| L 910.8 108.488 | |
| L 572.363025 108.488 | |
| z | |
| " style="fill: #ffffff"/> | |
| </g> | |
| <g id="matplotlib.axis_3"> | |
| <g id="xtick_8"> | |
| <g id="line2d_21"> | |
| <path d="M 574.229124 262.437656 | |
| L 574.229124 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_22"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="574.229124" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_13"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="574.229124" y="277.035313" transform="rotate(-0 574.229124 277.035313)">0.5600</text> | |
| </g> | |
| </g> | |
| <g id="xtick_9"> | |
| <g id="line2d_23"> | |
| <path d="M 613.859298 262.437656 | |
| L 613.859298 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_24"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="613.859298" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_14"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="613.859298" y="277.035313" transform="rotate(-0 613.859298 277.035313)">0.5625</text> | |
| </g> | |
| </g> | |
| <g id="xtick_10"> | |
| <g id="line2d_25"> | |
| <path d="M 653.489472 262.437656 | |
| L 653.489472 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_26"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="653.489472" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_15"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="653.489472" y="277.035313" transform="rotate(-0 653.489472 277.035313)">0.5650</text> | |
| </g> | |
| </g> | |
| <g id="xtick_11"> | |
| <g id="line2d_27"> | |
| <path d="M 693.119647 262.437656 | |
| L 693.119647 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_28"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="693.119647" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_16"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="693.119647" y="277.035313" transform="rotate(-0 693.119647 277.035313)">0.5675</text> | |
| </g> | |
| </g> | |
| <g id="xtick_12"> | |
| <g id="line2d_29"> | |
| <path d="M 732.749821 262.437656 | |
| L 732.749821 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_30"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="732.749821" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_17"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="732.749821" y="277.035313" transform="rotate(-0 732.749821 277.035313)">0.5700</text> | |
| </g> | |
| </g> | |
| <g id="xtick_13"> | |
| <g id="line2d_31"> | |
| <path d="M 772.379995 262.437656 | |
| L 772.379995 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_32"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="772.379995" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_18"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="772.379995" y="277.035313" transform="rotate(-0 772.379995 277.035313)">0.5725</text> | |
| </g> | |
| </g> | |
| <g id="xtick_14"> | |
| <g id="line2d_33"> | |
| <path d="M 812.010169 262.437656 | |
| L 812.010169 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_34"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="812.010169" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_19"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="812.010169" y="277.035313" transform="rotate(-0 812.010169 277.035313)">0.5750</text> | |
| </g> | |
| </g> | |
| <g id="xtick_15"> | |
| <g id="line2d_35"> | |
| <path d="M 851.640343 262.437656 | |
| L 851.640343 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_36"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="851.640343" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_20"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="851.640343" y="277.035313" transform="rotate(-0 851.640343 277.035313)">0.5775</text> | |
| </g> | |
| </g> | |
| <g id="xtick_16"> | |
| <g id="line2d_37"> | |
| <path d="M 891.270517 262.437656 | |
| L 891.270517 108.488 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.18; stroke-width: 0.8; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_38"> | |
| <g> | |
| <use xlink:href="#m3747435316" x="891.270517" y="262.437656" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| <g id="text_21"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="891.270517" y="277.035313" transform="rotate(-0 891.270517 277.035313)">0.5800</text> | |
| </g> | |
| </g> | |
| <g id="text_22"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="741.581513" y="291.036094" transform="rotate(-0 741.581513 291.036094)">Unweighted mean Spearman across eight tasks</text> | |
| </g> | |
| </g> | |
| <g id="matplotlib.axis_4"> | |
| <g id="ytick_4"> | |
| <g id="line2d_39"> | |
| <g> | |
| <use xlink:href="#m347ebb820c" x="572.363025" y="139.131524" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| </g> | |
| <g id="ytick_5"> | |
| <g id="line2d_40"> | |
| <g> | |
| <use xlink:href="#m347ebb820c" x="572.363025" y="185.462828" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| </g> | |
| <g id="ytick_6"> | |
| <g id="line2d_41"> | |
| <g> | |
| <use xlink:href="#m347ebb820c" x="572.363025" y="231.794132" style="stroke: #000000; stroke-width: 0.8"/> | |
| </g> | |
| </g> | |
| </g> | |
| </g> | |
| <g id="PathCollection_11"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#m8bb6c8efa8" x="806.725089" y="139.131524" style="fill: #0072b2; stroke: #0072b2; stroke-width: 1.8"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_12"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#me363e187c4" x="763.103878" y="185.462828" style="fill: #009e73; stroke: #009e73; stroke-width: 1.8"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_13"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#m620e92aa57" x="664.169176" y="231.794132" style="fill: #64748b; stroke: #64748b; stroke-width: 1.8"/> | |
| </g> | |
| </g> | |
| <g id="line2d_42"> | |
| <path d="M 741.700598 139.131524 | |
| L 871.74958 139.131524 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #0072b2; stroke-opacity: 0.6; stroke-width: 1.5; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_43"> | |
| <path d="M 776.975335 185.462828 | |
| L 749.232421 185.462828 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #009e73; stroke-opacity: 0.6; stroke-width: 1.5; stroke-linecap: square"/> | |
| </g> | |
| <g id="line2d_44"> | |
| <path d="M 611.413445 231.794132 | |
| L 716.924907 231.794132 | |
| " clip-path="url(#p4a2937a81d)" style="fill: none; stroke: #64748b; stroke-opacity: 0.6; stroke-width: 1.5; stroke-linecap: square"/> | |
| </g> | |
| <g id="patch_6"> | |
| <path d="M 572.363025 262.437656 | |
| L 572.363025 108.488 | |
| " style="fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square"/> | |
| </g> | |
| <g id="patch_7"> | |
| <path d="M 572.363025 262.437656 | |
| L 910.8 262.437656 | |
| " style="fill: none; stroke: #000000; stroke-width: 0.8; stroke-linejoin: miter; stroke-linecap: square"/> | |
| </g> | |
| <g id="PathCollection_14"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#md3c1ca1a6a" x="741.700598" y="136.11999" style="fill: #0072b2; stroke: #0072b2"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_15"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#m82ad91a55d" x="871.74958" y="142.143059" style="fill: #0072b2; stroke: #0072b2"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_16"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#me3968be686" x="776.975335" y="182.451293" style="fill: #009e73; stroke: #009e73"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_17"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#m82f90f8eea" x="749.232421" y="188.474363" style="fill: #009e73; stroke: #009e73"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_18"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#mc40f734263" x="611.413445" y="228.782597" style="fill: #64748b; stroke: #64748b"/> | |
| </g> | |
| </g> | |
| <g id="PathCollection_19"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#me3d60a9e43" x="716.924907" y="234.805667" style="fill: #64748b; stroke: #64748b"/> | |
| </g> | |
| </g> | |
| <g id="text_23"> | |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="741.581513" y="102.488" transform="rotate(-0 741.581513 102.488)">Final test outcomes</text> | |
| </g> | |
| <g id="PathCollection_20"> | |
| <g clip-path="url(#p4a2937a81d)"> | |
| <use xlink:href="#m501ecd2794" x="776.975335" y="182.451293" style="fill: #f59e0b; stroke: #334155; stroke-width: 0.6"/> | |
| </g> | |
| </g> | |
| </g> | |
| <g id="text_24"> | |
| <text style="font-size: 14px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: middle" x="460.8" y="17.405813" transform="rotate(-0 460.8 17.405813)">Larger experiments: development choice and final tests</text> | |
| </g> | |
| <g id="text_25"> | |
| <text style="font-size: 8.5px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: start" x="18.432" y="323.172" transform="rotate(-0 18.432 323.172)">Every recipe: 7,808 optimizer steps and 15,990,784 example presentations. Processed tokens and FLOPs are not exactly matched.</text> | |
| </g> | |
| <g id="text_26"> | |
| <text style="font-size: 8.5px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: start" x="18.432" y="334.3392" transform="rotate(-0 18.432 334.3392)">Joined endpoints show the two seeds, not a confidence interval. The default is chosen before reading these test outcomes.</text> | |
| </g> | |
| <g id="legend_1"> | |
| <g id="line2d_45"> | |
| <defs> | |
| <path id="m8bf2724b89" d="M 0 3 | |
| C 0.795609 3 1.55874 2.683901 2.12132 2.12132 | |
| C 2.683901 1.55874 3 0.795609 3 0 | |
| C 3 -0.795609 2.683901 -1.55874 2.12132 -2.12132 | |
| C 1.55874 -2.683901 0.795609 -3 0 -3 | |
| C -0.795609 -3 -1.55874 -2.683901 -2.12132 -2.12132 | |
| C -2.683901 -1.55874 -3 -0.795609 -3 0 | |
| C -3 0.795609 -2.683901 1.55874 -2.12132 2.12132 | |
| C -1.55874 2.683901 -0.795609 3 0 3 | |
| z | |
| " style="stroke: #334155"/> | |
| </defs> | |
| <g> | |
| <use xlink:href="#m8bf2724b89" x="313.946375" y="40.170437" style="fill: #334155; stroke: #334155"/> | |
| </g> | |
| </g> | |
| <g id="text_27"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: start" x="331.946375" y="43.670437" transform="rotate(-0 331.946375 43.670437)">Seed 1</text> | |
| </g> | |
| <g id="line2d_46"> | |
| <defs> | |
| <path id="mbf032f0408" d="M -3 3 | |
| L 3 3 | |
| L 3 -3 | |
| L -3 -3 | |
| z | |
| " style="stroke: #334155; stroke-linejoin: miter"/> | |
| </defs> | |
| <g> | |
| <use xlink:href="#mbf032f0408" x="396.490125" y="40.170437" style="fill: #334155; stroke: #334155; stroke-linejoin: miter"/> | |
| </g> | |
| </g> | |
| <g id="text_28"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: start" x="414.490125" y="43.670437" transform="rotate(-0 414.490125 43.670437)">Seed 2</text> | |
| </g> | |
| <g id="line2d_47"> | |
| <defs> | |
| <path id="me33e5fab49" d="M 0 5 | |
| L 0 -5 | |
| " style="stroke: #334155"/> | |
| </defs> | |
| <g> | |
| <use xlink:href="#me33e5fab49" x="479.033875" y="40.170437" style="fill: #334155; stroke: #334155"/> | |
| </g> | |
| </g> | |
| <g id="text_29"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: start" x="497.033875" y="43.670437" transform="rotate(-0 497.033875 43.670437)">Pair mean</text> | |
| </g> | |
| <g id="line2d_48"> | |
| <defs> | |
| <path id="mef179968ed" d="M 0 -5 | |
| L -1.12257 -1.545085 | |
| L -4.755283 -1.545085 | |
| L -1.816356 0.59017 | |
| L -2.938926 4.045085 | |
| L -0 1.90983 | |
| L 2.938926 4.045085 | |
| L 1.816356 0.59017 | |
| L 4.755283 -1.545085 | |
| L 1.12257 -1.545085 | |
| z | |
| " style="stroke: #f59e0b; stroke-linejoin: bevel"/> | |
| </defs> | |
| <g> | |
| <use xlink:href="#mef179968ed" x="577.169813" y="40.170437" style="fill: #f59e0b; stroke: #f59e0b; stroke-linejoin: bevel"/> | |
| </g> | |
| </g> | |
| <g id="text_30"> | |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Bitstream Vera Sans', 'Computer Modern Sans Serif', 'Lucida Grande', 'Verdana', 'Geneva', 'Lucid', 'Arial', 'Helvetica', 'Avant Garde', sans-serif; text-anchor: start" x="595.169813" y="43.670437" transform="rotate(-0 595.169813 43.670437)">Development-selected default</text> | |
| </g> | |
| </g> | |
| </g> | |
| <defs> | |
| <clipPath id="pe21196a7d8"> | |
| <rect x="196.63" y="108.488" width="338.436975" height="153.949656"/> | |
| </clipPath> | |
| <clipPath id="p4a2937a81d"> | |
| <rect x="572.363025" y="108.488" width="338.436975" height="153.949656"/> | |
| </clipPath> | |
| </defs> | |
| </svg> | |