Image-to-Markup Generation with Coarse-to-Fine Attention
Paper • 1609.04938 • Published • 1
id stringlengths 10 10 | image imagewidth (px) 28 1.67k | text stringlengths 40 986 |
|---|---|---|
7944775fc9 | \alpha_1^r \gamma_1 + \dots + \alpha_N^r \gamma_N = 0\quad(r=1,...,R)\; ,\label{contrainte} | |
78228211ca | \eta = -\frac{1}{2} \ln \left(\frac{\cosh \left(\sqrt{2}b_\infty\sqrt{1+\alpha^2}\; y - {\rm arcsinh}\; \alpha\right)}{\sqrt{1+\alpha^2}}\right) | |
15b9034ba8 | \label{fierep}P_{(2)}^-=\int \beta d\beta d^9p d^8\lambda \Phi(-p,-\lambda)\left(-\frac{p^Ip^I}{2\beta}\right) \Phi(p,\lambda)\,. | |
6968dfca15 | \label{GAMMA} \Gamma(z+1) =\int^\infty_0\,\,dx\,\,e^{-x}x^z. | |
6cead0df53 | \label{rotflow}\frac{d}{ds}{\bf C}_i= \frac{1}{2}\epsilon_{ijk}{\bf C}_j\times {\bf C}_k \, . | |
5381b22df4 | Z=\sum_{spins}\prod_{cubes}W(a|e,f,g|b,c,d|h), \label{f1} | |
27f2b37ce9 | \label{SUSY}\left\{ Q^{i},Q^{j}\right\} =c^{ij}\Gamma ^{M}CP_{M}+Cc^{ij}Z, \label{v4} | |
51a257cdf5 | \label{as7}\breve{c}_{n,\nu}=\sum_{m=n}^{2n}{\Gamma\left(\nu+m-{D-1 \over 2}\right) \over \Gamma\left(\nu+n-{D-1 \over 2}\right)}~\breve{a}_{2(m-n),m}~~~. | |
5108925e21 | R(g)=-f\left[3\left[(\ln f)^\prime\right]^2+\frac{\Lambda(x^5)}{M^3} \right] \; , \label{FourScalar} | |
3882dd3d43 | {d\over ds}{1\over\Gamma(-s)}\bigg|_{s=0}=-1, | |
566cf0c6f5 | \dot z_1 = - N^z(z_1) = - g(z_1) = -\frac{z_1}{P_z(z_2-z_1)};~~~\dot z_2 = -\frac{z_2}{P_z(z_2-z_1)} | |
7d1fe2cc05 | c_{\alpha} = \sum_{\beta\in\Lambda_{R}} \epsilon (\alpha,\beta ) |\beta + \bar{p}><\beta + \bar{p}| | |
450b24df87 | \label{lqed}{\cal L}=-{1\over 4}F_{\mu\nu}F^{\mu\nu}+{\bar\psi}(i\gamma^\mu D_\mu -m)\psi\,, | |
667ff49bc5 | e^{i {\bf k \cdot r}}=e^{i k r \cos(\theta - \Theta)}=\sum_{l=-\infty}^{\infty}i^l\,J_l(kr)\,e^{i l (\theta - \Theta)}\,,\label{eqn:iiifive} | |
61928de22b | i\sqrt{2} \partial_{-}\chi - g[\phi,\psi] = 0,\quad\partial_{-}^{2} \bar{A}_{+} - g^2 J^{+} = 0.\label{eq:const} | |
4cd65285c9 | \label{29}\Omega^{(l)}_k=\sum_{s=0} \int d^3y \left((-1)^{s+1} \frac{d^s}{dt^s}\phi^{i(s)}_k(x,y)L^{(0)}_i(y)\right). | |
12697ce419 | L_{g}^{'}\Bigl(v(h)\Bigr) = v(L_{g}h) = v(gh)\, , \,\, \, \forall g,h \in G, | |
a8ec0c091c | \xi^2=\left(\frac{\varepsilon_1-\varepsilon_2}{\varepsilon_1+\varepsilon_2}\right)^2=\left(\frac{\mu_1-\mu_2}{\mu_1+\mu_2}\right)^2, | |
72a80f57d9 | R(e_1) = \epsilon^{- J_{67} + J_{89}}, \quad R(e_2) = \epsilon^{ J_{45} - J_{89}} .\label{C3ZNZN-RotationsWithDiscreteTorsion} | |
330f27c566 | \label{eq33}{\tilde {\cal {E}}}_{m<0} = {\cal {E}}_{m<0}(B) - {\cal {E}}(0)= \frac {B^2} {2} + \frac {(eB)^{\frac {3} {2}}} {2 \pi} g\left(\frac {eB} {m^2}\right) \, , | |
58be3470dc | \label{R84}\hat{O}^{r}_{2}\mid 1>_{(0)}={O}^{r}_{2}\mid 0>_{(0)}. | |
3e82680317 | I^c =\mp{\pi b \sqrt{1 - \Lambda a^2}\over 2G}\ \ ,\label{exaction} | |
31068cb86d | g^>_n(r,r')=E_n K_{|n/\alpha|}(\beta r), \quad \hbox{for $r>r^{'}$.}\label{22} | |
431dd6944e | \label{sol_excited}R^{\frac{1}{2}}(\theta )^{\left| \left. b_{k}\dots \frac{1}{2},b_{1},\frac{1}{2}\right| n_{k}\dots ,m_{1},n_{1}\right\rangle }_{\left| \left. a_{k}\dots \frac{1}{2},a_{1},\frac{1}{2}\right| n_{k}\dots ,m_{1},n_{1}\right\rangle }=R_{a_{1}b_{1}}^{\frac{1}{2}}(\theta )\prod ^{k-1}_{i=1}f^{a_{i}a_{i+1}}_{b_{i}b_{i+1}}(w_{m_{i}},\nu _{n_{i+1}},\theta ) | |
54a7b9d7f8 | Q_1^{ab} (x,y) \equiv Q_1^{ab} + x \, J_1^{ab} + y \, K_1^{ab}, | |
632e971eb8 | \left\{\begin{array}{c} \partial_{\tau}R+\vec\nabla\cdot \left(\vec\nabla\Theta\,\sqrt{\displaystyle\frac{R^2+a^2} {1+(\vec\nabla\Theta)^2}}\right)=0,\hfill\\[4mm] \partial_{\tau}\Theta+R\sqrt{ \displaystyle\frac{1+(\vec\nabla\Theta)^2} {R^2+a^2}}=0.\hfill\\ \end{array}\right. \label{JPeqmot} | |
399e18a85c | \Delta^{(N,0)}(s)= - \sum_{n>0,\vec n^2<N}\left[ J(z_n) - 2 + 2 J(y_{n}) + \frac{J^{2}(y_{n})}{2 (1-y_{n})} - J(\tilde{z}_n) - 2 J(\tilde{y}_n) \right]\ , \label{deltafin} | |
707b5988e2 | \left\{\Psi\circ\mu ,f\right\}=(\overline X_if)\, (Y^i \Psi)\circ\mu\,,\label{mom2} | |
25fe4d51bf | F_{n}^{\mathcal{O}|\mu _{1}\ldots \mu _{n}}(\theta _{1}+\lambda ,\ldots,\theta _{n}+\lambda )=e^{s \lambda }F_{n}^{\mathcal{O}|\mu _{1}\ldots \mu_{n}}(\theta _{1},\ldots ,\theta _{n})\,\,, \label{rel} | |
3dc7799669 | \label{extended action}S = S_{Phys.}(\Phi^a,\Phi^{\ast a}) + S_T(\vartheta^b,\vartheta^{\ast b}, c^\alpha) | |
76d30658bb | \mathcal{A} \equiv \exp \left[ \int_0^\lambda d\tilde{\lambda}\, \theta(\tilde{\lambda}) \right]\, . | |
6a366e1f12 | \label{fermhalf} F_{-{1\over2}}(x)=\bar \epsilon_0S(x)e^{-1/2\phi(x)}\;,\qquad F_{ 1\over 2}(x) = \bar \epsilon_0 \gamma_\mu S(x) \partial X^\mu(x)e^{1/2\phi(x)}, | |
de8a312222 | \rho^0 = \left( \begin{array}{cc} 0 & -i \\ i & 0\end{array}\right)\,\,\, \mbox{and}\,\,\, \rho^1 = \left( \begin{array}{cc} 0 & i \\ i & 0\end{array}\right) . | |
2b891b21ac | \psi=\sum_{i=0}^3 (\psi_i^A +(\psi_i^A)^c) T^A | |
72e168fb21 | \label{coset}G=\!e^{i\tau L_{-1}} e^{iU^{(1)}L_1} e^{iU^{(2)}L_2}e^{iU^{(3)}L_3}\ldots\! e^{i{U^{(0)}}L_0}, | |
3d129cfe77 | V(z, \bar z)=e^{-q \Phi(z)} e^{i \alpha \cdot H} e^{i(P_R \cdot X_R-P_L \cdot X_L)} \;, | |
6a85896075 | \label{4.5}\epsilon_i = \tau_i + \rho_i + \rho_{i-1}, \quad (\tau_3 =0 ,\: \rho_0 = \rho_4) | |
79edbca78a | s_\infty (k^2)-s_{J_{\max }}(k^2)\sim O(J_{\max }^{-2}). \label{if} | |
20032b2645 | \label{eq:SERELRA} A(u)~=~{\rm Res} \vert_{v=u}^{} \left( {1 \over v-u} \, R(u,v) \cdot L(v) \right) + \, {\textstyle {1 \over 2}} \, \zeta(2u) \, L(u) | |
3d15b5c484 | \partial^m_a \Gamma_i = \frac{\Gamma^n}{\lambda_i} \{\delta_{nm} \psi_a^i - \phi^n_b \phi^m_c \psi_b^i \psi_c^i \frac{\psi_a^i}{\lambda_i^2} + \phi^n_b \phi^m_c \sum_{j \neq i} \psi_b^j \frac{(\psi_c^i \psi_a^j + \psi_a^i \psi_c^j)} {(\lambda_i^2 - \lambda_j^2)}\}\label{E11}\vspace{-12pt} | |
2608ceb605 | \int {\rm d}^{4}x_{1}~\cdots ~{\rm d}^{4}x_{n}~P_{4}(x_{1},\cdots,x_{n})~\Gamma _{x_{1}\cdots x_{n}0}=0 \label{sum2} | |
146a5fa39e | L=\frac{\dot{x}_\mu^{2}}{2e}+\frac{\lambda}{l}(e-M^{-1}\dot{x}{}^0),\label{inter} | |
159bf72783 | J_2(z)\times X^{+}(w)\rightarrow 0. \label{j2xp} | |
1e3aab9a4f | F(z^{\prime}_{12})=\bar{K}(z_{2};g)F(z_{12})K(z_{1};g)\label{zz} | |
62409f879c | {\xi}^{\ast}_i, {p}^{\ast}_i, \quad i = 2, \dots, l+1 | |
6beab42e50 | \varrho_L - {\cal L}_E= [2\dot\Phi^2] \; K'(\dot\Phi^2,\Phi) - K(\dot\Phi^2,\Phi) + K(-\dot\Phi^2,\Phi). | |
105ccc7946 | K' = \sqrt{c - 2f}\ , \ \ \ K'' = -\frac{1}{\sqrt{c - 2f}}\ , | |
6df7276525 | \label{kappa}\kappa _{\omega }=\frac{2\Gamma (\Delta _{\omega })}{\pi \Gamma(1-\Delta _{\omega })}\left( \frac{\sqrt{\pi }\Gamma \left( \frac{1}{2-2\Delta _{\omega}}\right)}{2 \Gamma \left( \frac{\Delta _{\omega }}{2-2\Delta _{\omega}}\right) }\right) ^{2-2\Delta _{\omega }}\, . | |
65d07ed733 | <\frac{1}{2},m_s|{\psi}_{-}^{(\frac{1}{2})}(g)>\equiv D^{(\frac{1}{2})}_{m_s-\frac{1}{2}}(g)=<g,l+\frac{1}{2}|T^{\frac{1}{2}}_{m_s-}|g,l>. | |
34173474c4 | \sum_{l,n} \frac{\mu_{p-1}\lambda^{k+n+l}i^kp!}{k!n!l!(p-l)!}\partial_{x^{i_1}}\ldots \partial_{x^{i_n}} C^0_{i_1'\ldots i_{2k}'j_1\ldots j_l[a_{l+1}\ldots a_p} Str \left(\partial_{a_1}\phi^{j_1}\ldots \partial_{a_l]}\phi^{j_l}\phi^{i_1}\ldots \phi^{i_n}\phi^{i_{2k}'}\phi^{i_{2k-1}'}\ldots \right) | |
1a79f53af4 | D^{\mu}\frac{\delta f(A_{\nu})}{\delta A_{\mu}}=D_{\mu}\partial^{\mu}(\partial_{\nu}A^{\nu}) | |
57e32e5b33 | \delta\chi_{\mu\nu} = ib_{\mu\nu}, \qquad \delta b_{\mu\nu} = 0.\label{eqn:topantiaux} | |
7e1098abc4 | V_{ab\ \ mn}^{k}=\frac{1}{g}\ E_{a}^{r}\ E_{b}^{s}\epsilon _{rs(m}\ \delta_{n)}^{i}. | |
5ada9733aa | \label{req3}f(r)= \left( 1-\frac{m}{2r^{n-1}}\right)^2 +\frac{r^2}{l^2}. | |
5b109d24dc | E_{12}~~\Phi= 2\sqrt{(m+\frac{1}{2}br)^{2}+p^{2}_{r}+ \frac{\ell(\ell+1)}{r^{2}}} ~~\Phi, \label{eq:e} | |
5193ae2c89 | T_{\mathit{G}}(-t,-t^{-1})=T_{\mathit{G}^{\ast }}(-t^{-1},-t) \label{16} | |
119b93a445 | ds_{11}^2 = dx^+ dx^- + l_p^9 \frac{ p_-}{r^7} \delta(x^- ) dx^- dx^- + dx_1^2 + \ \cdots \ + dx_9^2 \label{ase} | |
4fa61dbf37 | F_{ab} = {1\over 2} \epsilon_{abcd} F^{cd} | |
25765b9391 | 2f^2-4f^2-g^2(1-\Gamma) \, ,\label{eq:3.16} | |
276c373567 | (a^{\dagger} L_{mn} a) = a^{\dagger}_{k} (L_{mn})_{kl} a_{l} =i a^{\dagger}_{[m} a_{n]}, \;\;\;\;\;\; (L_{mn})_{kl} = i (\delta_{mk} \delta_{nl} - \delta_{nk} \delta_{ml} ) | |
3fd05b449f | \int dt d^3x \bar{\lambda} \partial^\mu \gamma_\mu \lambda, | |
6b2c7f0c1a | h = {s\lambda\over {1 + 2n + sN + |N|}},\label{eigenvalue} | |
7c2f256525 | Q=c\sum_{i} f_{i}' p^{i} + \sum_{k} c_{k} p^{k} f_{k} +infinite \: more. | |
3beaade5a5 | {\rm Tr}\,\log(1-\sum_{i=0}^{N} A_i)~=~{\rm Tr}\,\log(1-\sum_{k=1}^{N} \sum_{m=0}^{k-1} A_k\phi^m)+{\rm Tr}\,\log(1-\phi)~. | |
3cf56a8338 | H_{ij}^{a}=F_{ij}^{a}-gf_{\;\;bc}^{a}A_{i}^{b}A_{j}^{c}, \label{49x} | |
4aea73b2b8 | {\tilde{\rho}}_{ {\bf{q}} } = \sum_{ {\bf{k}} }[\Lambda_{ {\bf{k}} }({\bf{q}}) a_{ {\bf{k}} }(-{\bf{q}}) + \Lambda_{ {\bf{k}} }(-{\bf{q}}) a^{\dagger}_{ {\bf{k}} }({\bf{q}})] | |
2e96a960b1 | \label{1.2}{\bf N}({\bf p},{\bf s}):=ip_0{\nabla}_{\bf p}-\frac{{\bf s}\times{\bf p}}{p_0+m},\quad{\bf J}({\bf p},{\bf s}):=-i{\bf p}\times{\bf \nabla}_{\bf p}+{\bf s}:={\bf L}({\bf p})+{\bf s}, | |
5b10a20227 | \label{deca2} A_\mu \;=\; \partial_\mu \varphi + \epsilon_{\mu\nu} \, \partial_\nu \sigma \;. | |
18049a05a9 | C_J (\nu_1, \nu_2)=(2 J + \nu_1 + \nu_2 + 1)\frac{{\mit\Gamma} (J + 1) {\mit\Gamma} (J + \nu_1 + \nu_2 + 1)}{{\mit\Gamma} (J + \nu_1 + 1) {\mit\Gamma} (J + \nu_2 + 1)}\, . | |
17ad05c612 | (\psi\otimes_{\zeta,z} \chi) \mapsto (e^{-u L_{-1}} \psi\otimes_{\zeta+u,z+v} e^{-v L_{-1}} \chi), | |
4b69ad5dc8 | u_0(k,r)=\sqrt{{\displaystyle {\pi\over 2}}}\,i\sqrt{r}\,J_0(kr) - \sqrt{{\displaystyle {\pi\over2}}}\,A(k)\sqrt{kr}\,H_0^{(1)}(kr).\label{eq:2.19} | |
4acf2a0344 | J_k=\oint p_kdq_k,~~k=r,~\theta,~\phi,\label{eqjk} | |
272667a2d1 | \delta_\perp \kappa_1 = \kappa_3 \kappa_2 \Psi_3+ 2 \kappa_2 \Psi_2{}' + \kappa_2' \Psi_2 + \Psi_1{}'' - \left( \kappa_1^2 + \kappa_2{}^2 \right) \Psi_1\,.\label{eq:perpkappa} | |
5a5e2b80dd | \left( \gamma _\mu \partial _\mu +m\right) \psi^{(b)}(x)=0,\hspace{0.5in} b=1,2,3,4 \label{4} | |
6596750444 | \label{sin}f_\alpha(x)=\left(4\sin^2\frac{x}{2}\right)^\alpha. | |
4370181032 | r_h^2 = \frac{l^2}{2}(\sqrt{K^2+4l^{-2}\mu}-K). | |
11ff25534a | \label{eomsfield}\mu ''+\biggl[n^2-\frac{a''}{a}\biggr]\mu =0. | |
22f7232e98 | x_{\overline m}={1\over2}(x_m+x_{m+1}),\label{ave} | |
3eaf444392 | S_{ij}\left( \theta \right) =\prod\limits_{x,y}\left[ x,y\right] _{\theta }\label{nichtsimp} | |
25c3276f55 | A_{d}(p^{2}+\omega _{n}^{2})^{\frac{1}{2}d-2}\left[ \left( 1+v^{2}\right) ^{\frac{1}{2}d-\frac{3}{2}}+\frac{\Gamma (\frac{1}{2}d-\frac{1}{2})}{\sqrt{\pi}\Gamma (d)}\frac{(v^{2})^{\frac{d}{2}-1}}{1+v^{2}}{}_{2}F_{1}\left(\textstyle{{\frac{1}{2}d-\frac{1}{2},1;\frac{1}{2}d;\frac{v^{2}}{1+v^{2}}}}\right) \right] \,\ \cdot \label{hyp1} | |
32ebd66b47 | \psi_c(x)=\gamma^1 \psi^*(x) \ ,\label{in10} | |
3b014d22b2 | L = L^\Lambda {\bf T}_\Lambda = dZ^M L_M{}^\Lambda {\bf T}_\Lambda\,. | |
6a88fd17f0 | z^{'(r)}_{t,0} \quad = \quad z^{''(r)}_{t,0} \quad = \quad 0\qquad ( 1 \le t \le r, \,\, {\rm{all}} \,\, r ) \, ; | |
4b49a4f210 | dT(x)=\left(\begin{array}{cc} \delta(x) 1_{N-k} & 0 \\
0 & -\delta(x)1_k \\ \end{array}\right)dx | |
4b4263156b | 1 - \frac{2GM}{\rho} = (\nabla\rho)^2 \equiv f | |
48f89a8fc4 | \label{eq:MidentityW} M_{g}= M_{c_{1}} M_{c_{2}} M_{c_{3}} M_{c_{4}} M_{c_{5}}M_{r=\infty}= 1 | |
8d78ecda53 | \delta F\left( \sin\theta\, dx^{0}dx^{1},0,0,\epsilon\,dx^{0}\cdots dx^{3}\right) . \label{finaldf} | |
4015ba8922 | S^{({\rm N})}_{{\rm part},0} = \int dt\, \sum\limits^{N}_{\alpha=1} \left( \xi^{\underline{a}}_{\alpha} \left( E^{\underline{a}}_{j,\alpha} \dot{x}^{j}_{\alpha} + E^{\underline{a}}_{0,\alpha} \right) - {1\over 2} \xi^{\underline{a}}_{\alpha} \xi^{\underline{a}}_{\alpha} \right)\, . | |
acedffb147 | \xi = v_1 \left( u_1 - \kappa v_2 \right) + v_2 \left( u_2 -\kappa v_1 \right). | |
4f055acd1f | \label{effpot}U(r)=U(r_0)+4\pi^2K A(d,\sigma)\int_{r_0}^{r}ds \frac{s^{\sigma-d-1}}{\varepsilon(s)}. | |
5ecb739ec1 | |0 \rangle \rightarrow |0 \rangle_{\beta} = (1 + {\rm e}^{- \beta\epsilon})^{-1/2} \{ |0 \rangle_a \otimes |0\rangle_b + {\rm e}^{- \beta \epsilon / 2} a^{\dagger} {\tilde a} |0 \rangle_a \otimes b^{\dagger} {\tilde b} |0 \rangle_b \}. | |
2beadd086b | S_E = \int_0 ^{\tau} d\tau \left({1 \over 2} x_{\tau} ^2+{1 \over 2} W^2(x)- \psi^{\ast}[\partial_{\tau} -W^{\prime}(x)]\psi \right) | |
2450656988 | R_{\mu \nu \,\,b}^{\quad a}=\partial _{\mu }\omega _{\nu\,\,b}^{\,a}-\partial _{\nu }\omega _{\mu \,\,b}^{\,a}+\omega _{\mu\,\,c}^{\,a}\ast \omega _{\nu \,\,b}^{\,c}-\omega _{\nu \,\,c}^{\,a}\ast\omega _{\mu \,\,b}^{\,c} | |
147dde7fd3 | M=\int_{r\rightarrow \infty}d^pxr^{p/2}f^{-1/2}T_{tt} =\frac{pmV_p}{16\pi G_{p+2}}. | |
17d9b6a683 | j_{HW}(x) = W_i(x) T^i\;,\; T^i \in ker(Ad(M_-)) | |
55358c150e | k_0\sim \omega \sqrt{\frac{g\phi _0}{2M^2}}\ll \omega | |
13dbb0dd7c | Y(T,U) = \int_{\cal F}\frac{d^2\tau}{\Im \tau} \Gamma_{2,2}(T,U) \left(-6\left[{\overline{\Omega }}_2^{\phantom 2}- \frac{1}{8\pi \Im\tau}\right]\frac{\overline{\Omega}}{\overline{\eta}^{24}}- \frac{\overline{j}}{8}+126\right)\ ,\label{twelve} | |
1d796ff39a | P_0S^*P_0SP_0=P_0=P_0SP_0S^*P_0\quad\Longleftrightarrow \quad(S_{00})^*=(S_{00})^{-1}\quad \mathrm{on}\quad \mathcal{H}_{\mathrm{phys}}. | |
6a39a91654 | \langle\psi ^{1-a}_{Fa}\mid\phi ^{1-a'}_{Fa'}\rangle_t=
\frac{1}{2}\delta (a-a')\theta (t-1+a)\theta (t-1+a') | |
6661b12767 | L(z,u_a,D)\equiv\int_0^{\infty}d\hat T\,J(z,u_a,\hat T,D)=L_{02}(z,u_a,D)+g(z,D)G_{Bab}^{1-{D\over 2}}+O\bigl(z^4,G_{Bab}^{2-{D\over 2}}\bigr)\nonumber\\\label{spindivextr} |
End of preview. Expand in Data Studio
im2latex-test-10k (Hugging Face Version)
Overview
This dataset is a Hugging Face–formatted version of the im2latex-100k test split.
It is designed for:
- Handwritten / rendered mathematical expression recognition
- Image-to-LaTeX modeling
- Vision-to-sequence tasks
- OCR research for mathematical formulas
This version is converted to Apache Arrow (PyArrow) format for seamless integration with the Hugging Face datasets library.
Dataset Structure
Each sample contains 3 columns:
| Column | Type | Description |
|---|---|---|
id |
string | Unique image identifier (filename stem) |
image |
Image | PNG image of rendered mathematical formula |
text |
string | Ground-truth LaTeX expression |
Images are stored using the Hugging Face Image() feature type, ensuring proper rendering in the Dataset Viewer.
Dataset Viewer Compatibility
This dataset is structured to work correctly with the Hugging Face Dataset Viewer:
- Images are stored as
Image()feature type - Text is stored as UTF-8 strings
- No nested or unsupported formats
- Apache Arrow format for fast preview
The Dataset Viewer should automatically:
- Render images properly
- Display LaTeX strings as raw text
- Allow filtering and browsing
How It Was Built
The dataset was created by:
- Reading
im2latex_test.lst - Mapping
formula_idxtoim2latex_formulas.lst - Reading formulas using
newline="\n"(important to prevent index drift) - Saving in Hugging Face Arrow format
Special care was taken to avoid the known newline issue described in the original dataset documentation.
Original Dataset Credit
Original Dataset: im2latex-100k
Paper: https://arxiv.org/abs/1609.04938
Original Author: Anssi Kanervisto
Original resources derived from:
- arXiv LaTeX sources
- Cornell KDD Cup dataset
Related tools:
All credit for the original dataset belongs to the original creators.
This repository only provides a Hugging Face–formatted version for easier accessibility and usage.
Usage Example
from datasets import load_dataset
dataset = load_dataset("SoyVitou/im2latex-test-10k")
sample = dataset["train"][0]
print(sample["id"])
print(sample["text"])
sample["image"]
Notes
- This version does not modify original data content.
- Only format conversion was performed.
- Intended for research and educational use.
License
Please refer to the original dataset source for licensing terms.
- Downloads last month
- 15
Size of downloaded dataset files:
47 MB
Size of the auto-converted Parquet files:
47 MB
Number of rows:
10,355