GithubHelp home page GithubHelp logo

llm-agent's Introduction

A LLM-Driven Agent framework

This project is a demo for LLM-Agent. It can use tools like wolfram Alpha to get latest data and accurate information.

A demonstration on how left and right brain mindset could combine to solve real world problems.

Examples below.

Information Query

Accurate math computation

Screenshot_20240324_110837.jpg

Weather related data

Screenshot_20240324_110818.jpg

Daily activities

Screenshot_20240324_110828.jpg

Geographical Result

Screenshot_20240324_110758.jpg

Accurate stock market information

Screenshot_20240324_110810.jpg

Iterative general problem solving

Solving a biogenetic problem

Hello 
 Welcome to the Agent System.
 We provide you with the following chatbots:
1. a: b
2. test_bot: read the file and return the content of the file exactly
3. Mmaster: Use Mathematica to solve anything
4. good_boy: test again
5. Create a new assistant

 Please enter the index of your choice: 3
Loaded existing assistant ID for Mmaster.
Enter your message: Calculate N50 for contig lengths 80kb, 60kb, 50kb, 48kb, 35kb, 27kb, 22kb, 19kb, 19kb, 13kb, 8kb, 5kb
Received message: Calculate N50 for contig lengths 80kb, 60kb, 50kb, 48kb, 35kb, 27kb, 22kb, 19kb, 19kb, 13kb, 8kb, 5kb in thread thread_DfpY9K39b59MRbLfmdQhGK4a

 Querying Wolfram Alpha for information...
Query: arrange 80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5 in decreasing order
Query complete.

Output: {{{Input, 1}, Plaintext} -> reverse sort | (80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5), {{Result, 1}, Plaintext} -> {80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}, {{Input, 1}, Input} -> HoldComplete[ReverseSort[{80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}]], {{Result, 1}, Input} -> HoldComplete[ReverseSort[{80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}]]}



 Querying Wolfram Alpha for information...
Query: total 80+60+50+48+35+27+22+19+19+13+8+5
Query complete.

Output: {{{Input, 1}, Plaintext} -> 80 + 60 + 50 + 48 + 35 + 27 + 22 + 19 + 19 + 13 + 8 + 5, {{Result, 1}, Plaintext} -> 386, {{NumberName, 1}, Plaintext} -> three hundred eighty-six, {{Input, 1}, Input} -> HoldComplete[80 + 60 + 50 + 48 + 35 + 27 + 22 + 19 + 19 + 13 + 8 + 5], {{NumberName, 1}, Input} -> HoldComplete[IntegerName[386, Words]]}



 Querying Wolfram Alpha for information...
Query: 386/2
Query complete.

Output: {{{Input, 1}, Plaintext} -> 386/2, {{Result, 1}, Plaintext} -> 193, {{QuotientAndRemainder, 1}, Plaintext} -> 386 = 193  2 + 0, {{NumberName, 1}, Plaintext} -> one hundred ninety-three, {{Input, 1}, Input} -> HoldComplete[386/2], {{NumberLine, 1}, Input} -> HoldComplete[NumberLinePlot[193]], {{QuotientAndRemainder, 1}, Input} -> HoldComplete[QuotientRemainder[386, 2]], {{NumberName, 1}, Input} -> HoldComplete[IntegerName[193, Words]]}    



 Querying Wolfram Alpha for information...
Query: 80+60+50+48
Query complete.

Output: {{{Input, 1}, Plaintext} -> 80 + 60 + 50 + 48, {{Result, 1}, Plaintext} -> 238, {{NumberName, 1}, Plaintext} -> two hundred thirty-eight, {{Input, 1}, Input} -> HoldComplete[80 + 60 + 50 + 48], {{NumberName, 1}, Input} -> HoldComplete[IntegerName[238, Words]]}


The midpoint of the total length is 193.

Now, let's calculate the N50 value by summing the lengths of the contigs starting from the longest, until the cumulative sum equals or exceeds 193:   

- 80 + 60 + 50 + 48 = 238 (exceeds 193, so the N50 value is 48)

So, the N50 value for the given contig lengths is 48kb.
Enter your message: what about L50
Received message: what about L50 in thread thread_DfpY9K39b59MRbLfmdQhGK4a

 Querying Wolfram Alpha for information...
Query: lengths 80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5
Query complete.

Output: {{{Input, 1}, Plaintext} -> triangle | edge lengths 80 | 60 | 50 | 48 | 35 | 27 | 22 | 19 | 19 | 13 | 8 | 5, {{Result, 1}, Plaintext} -> (not a possible triangle)}



 Querying Wolfram Alpha for information...
Query: count 80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5
Query complete.
Output: {}



 Querying Wolfram Alpha for information...
Query: count of elements in the list {80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}
Query complete.

Output: {{{Input, 1}, Plaintext} -> length | (80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5), {{Result, 1}, Plaintext} -> 12, {{Input, 1}, Input} -> HoldComplete[Length[{80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}]], {{Result, 1}, Input} -> HoldComplete[Length[{80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}]]}



 Querying Wolfram Alpha for information...
Query: index of 48 in list {80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}
Query complete.

Output: {}



 Querying Wolfram Alpha for information...
Query: position of 48 in {80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}
Query complete.

Output: {{{Input, 1}, Plaintext} -> position | (80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5)
pattern 48, {{Result, 1}, Plaintext} -> (4), {{Input, 1}, Input} -> HoldComplete[Position[{80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}, 48]], {{Result, 1}, Input} -> HoldComplete[Position[{80, 60, 50, 48, 35, 27, 22, 19, 19, 13, 8, 5}, 48]]}


The position of the contig length 48 in the list is 4.

To find the L50 value, we need to count the number of contigs with indices less than or equal to 4, since the N50 value is at index 4.

Therefore, the L50 value is 4.

So, the L50 for the given contig lengths is 4.

The result is correct, all by agent itself, without human interference/prior knowledge.

A quantum mechanics problem

Enter your message: 
Derive the probability flow distribution of spherical wave $\braket{\vec{r}|\psi} = \frac{1}{(2\pi)^{3/2}}\frac{e^{ikr}}{r}$, and discuss the behavior when $r\rightarrow\infty$.

Received message: Derive the probability flow distribution of spherical wave $\braket{\vec{r}|\psi} = \frac{1}{(2\pi)^{3/2}}\frac{e^{ikr}}{r}$, and discuss the behavior when $r\rightarrow\infty$. in thread thread_RxUKPMWnZrerP2lXr30QhTKk

 Querying Wolfram Alpha for information...
Query: probability flow distribution of spherical wave psi = (1/(2 pi)^(3/2))(e^(ikr)/r)
Query complete.

Output: {}



 Querying Wolfram Alpha for information...
Query: complex conjugate of (1/(2 pi)^(3/2))(e^(i k r)/r)
Query complete.

Output: {{{Input, 1}, Plaintext} -> (1/(2 \[Pi])^(3/2))^*e^(i k r)/r, {{Result, 1}, Plaintext} -> e^(i k r)/(2 sqrt(2) \[Pi]^(3/2) r), {{AlternateFormAssumingAllVariablesAreReal, 1}, Plaintext} -> cos(k r)/(2 sqrt(2) \[Pi]^(3/2) r) + (i sin(k r))/(2 sqrt(2) \[Pi]^(3/2) r), {{NoSolutionsFromReduce, 1}, Plaintext} -> (no roots exist), {{SeriesExpansionAtr=0, 1}, Plaintext} -> 1/(2 sqrt(2) \[Pi]^(3/2) r) + (i k)/(2 sqrt(2) \[Pi]^(3/2)) - (k^2 r)/(4 (sqrt(2) \[Pi]^(3/2))) - (i k^3 r^2)/(12 sqrt(2) \[Pi]^(3/2)) + (k^4 r^3)/(48 sqrt(2) \[Pi]^(3/2)) + O(r^4)
(Laurent series), {{Derivative, 1}, Plaintext} -> d/dr(((1/(2 \[Pi])^(3/2))^* e^(i k r))/r) = (i e^(i k r) (k r + i))/(2 sqrt(2) \[Pi]^(3/2) r^2), {{IndefiniteIntegral, 1}, Plaintext} -> integral e^(i k r)/(2 sqrt(2) \[Pi]^(3/2) r) dr = Ei(i k r)/(2 sqrt(2) \[Pi]^(3/2)) + constant, {{AlternativeRepresentations:MathematicalFunctionIdentityData, 1}, Plaintext} -> ((1/(2 \[Pi])^(3/2))^* e^(i k r))/r = (e^(i k r) abs(1/(2 \[Pi])^(3/2))^2)/(r/(2 \[Pi])^(3/2)), {{AlternativeRepresentations:MathematicalFunctionIdentityData, 2}, Plaintext} -> ((1/(2 \[Pi])^(3/2))^* e^(i k r))/r = ((1/(2 \[Pi])^(3/2) - i (2 Im(1/(2 \[Pi])^(3/2)))) z^(i k r))/r for z = e, {{AlternativeRepresentations:MathematicalFunctionIdentityData, 3}, Plaintext} -> ((1/(2 \[Pi])^(3/2))^* e^(i k r))/r = e^(i k r)/(r e^(2 i arg(1/(2 \[Pi])^(3/2))) (2 \[Pi])^(3/2)), {{SeriesRepresentations:MathematicalFunctionIdentityData, 1}, Plaintext} -> ((1/(2 \[Pi])^(3/2))^* e^(i k r))/r = ( sum_(j=0)^Infinity (i k r)^j/(j!))/(2 sqrt(2) \[Pi]^(3/2) r), {{SeriesRepresentations:MathematicalFunctionIdentityData, 2}, Plaintext} -> ((1/(2 \[Pi])^(3/2))^* e^(i k r))/r = ( sum_(A=0)^Infinity ((i k r)^(-1 + 2 A) (i k r + 2 A))/((2 A)!))/(2 sqrt(2) \[Pi]^(3/2) r), {{SeriesRepresentations:MathematicalFunctionIdentityData, 3}, Plaintext} -> ((1/(2 \[Pi])^(3/2))^* e^(i k r))/r = ( sum_(A=0)^Infinity ((i k r)^(2 A) (1 + i k r + 2 A))/((1 + 2 A)!))/(2 sqrt(2) \[Pi]^(3/2) r), {{Input, 1}, Input} -> HoldComplete[Conjugate[1/(2*Pi)^(3/2)]*E^(I*k*r)/r], {{3DPlot, 1}, Input} -> HoldComplete[Plot3D[Re[E^(I*k*r)/r]/(2*Sqrt[2]*Pi^(3/2)), {k, -3.92288, 3.92288}, {r, -3.92288, 3.92288}]], {{3DPlot, 2}, Input} -> HoldComplete[Plot3D[Im[E^(I*k*r)/r]/(2*Sqrt[2]*Pi^(3/2)), {k, -3.92288, 3.92288}, {r, -3.92288, 3.92288}]], {{ContourPlot, 1}, Input} -> HoldComplete[ContourPlot[Re[E^(I*k*r)/r]/(2*Sqrt[2]*Pi^(3/2)), {k, -3.92288, 3.92288}, {r, -3.92288, 3.92288}]], {{ContourPlot, 2}, Input} -> HoldComplete[ContourPlot[Im[E^(I*k*r)/r]/(2*Sqrt[2]*Pi^(3/2)), {k, -3.92288, 3.92288}, {r, -3.92288, 3.92288}]], {{AlternateFormAssumingAllVariablesAreReal, 1}, Input} -> HoldComplete[ComplexExpand[E^(I*k*r)/(2*Sqrt[2]*Pi^(3/2)*r)]], {{NoSolutionsFromReduce, 1}, Input} -> HoldComplete[Reduce[{E^(I*k*r)/(2*Sqrt[2]*Pi^(3/2)*r) == 0}, {k, r}]], {{SeriesExpansionAtr=0, 1}, Input} -> HoldComplete[Series[E^(I*k*r)/(2*Sqrt[2]*Pi^(3/2)*r), {r, 0, 4}]], {{Derivative, 1}, Input} -> HoldComplete[D[E^(I*k*r)/(2*Sqrt[2]*Pi^(3/2)*r), r]], {{IndefiniteIntegral, 1}, Input} -> HoldComplete[Integrate[E^(I*k*r)/(2*Sqrt[2]*Pi^(3/2)*r), r]], {{AlternativeRepresentations:MathematicalFunctionIdentityData, 1}, Input} -> HoldComplete[Conjugate[1/(2*Pi)^(3/2)]*E^(I*k*r)/r == (E^(I*k*r)*Abs[((2*Pi)^(3/2))^(-1)]^2)/(r/(2*Pi)^(3/2))], {{AlternativeRepresentations:MathematicalFunctionIdentityData, 2}, Input} -> HoldComplete[Conjugate[1/(2*Pi)^(3/2)]*E^(I*k*r)/r == (1/(2*Pi)^(3/2) - 2*I*Im[1/(2*Pi)^(3/2)])*z^(I*k*r)/r /; z == E], {{AlternativeRepresentations:MathematicalFunctionIdentityData, 3}, Input} -> HoldComplete[Conjugate[1/(2*Pi)^(3/2)]*E^(I*k*r)/r == E^(I*k*r)/(r*E^(2*I*Arg[((2*Pi)^(3/2))^(-1)])*(2*Pi)^(3/2))], {{SeriesRepresentations:MathematicalFunctionIdentityData, 1}, Input} -> HoldComplete[Conjugate[1/(2*Pi)^(3/2)]*E^(I*k*r)/r == Sum[(I*k*r)^j/j!, {j, 0, Infinity}]/(2*Sqrt[2]*Pi^(3/2)*r)], {{SeriesRepresentations:MathematicalFunctionIdentityData, 2}, Input} -> HoldComplete[Conjugate[1/(2*Pi)^(3/2)]*E^(I*k*r)/r == Sum[((I*k*r)^(-1 + 2*\[ScriptCapitalA])*(I*k*r + 2*\[ScriptCapitalA]))/(2*\[ScriptCapitalA])!, {\[ScriptCapitalA], 0, Infinity}]/(2*Sqrt[2]*Pi^(3/2)*r)], {{SeriesRepresentations:MathematicalFunctionIdentityData, 3}, Input} -> HoldComplete[Conjugate[1/(2*Pi)^(3/2)]*E^(I*k*r)/r == Sum[((I*k*r)^(2*\[ScriptCapitalA])*(1 + I*k*r + 2*\[ScriptCapitalA]))/(1 + 2*\[ScriptCapitalA])!, {\[ScriptCapitalA], 0, Infinity}]/(2*Sqrt[2]*Pi^(3/2)*r)]}



 Querying Wolfram Alpha for information...
Query: gradient of (1/(2 pi)^(3/2))(e^(i k r)/r)
Query complete.

Output: {{{Input, 1}, Plaintext} -> grad(1/(2 \[Pi])^(3/2)e^(i k r)/r), {{DelOperatorForm, 1}, Plaintext} -> del (e^(i k r)/((2 \[Pi])^(3/2) r)), {{VectorAnalysisResult, 1}, Plaintext} -> grad(e^(i k r)/((2 \[Pi])^(3/2) r)) = ((i e^(i k r) (i + k r))/(2 sqrt(2) \[Pi]^(3/2) r^2), 0)
in Cartesian coordinates:
... = (i e^(i k sqrt(x^2 + y^2)) x (i + k sqrt(x^2 + y^2)))/(2 sqrt(2) \[Pi]^(3/2) (x^2 + y^2)^(3/2)) e_x + (i e^(i k sqrt(x^2 + y^2)) y (i + k sqrt(x^2 + y^2)))/(2 sqrt(2) \[Pi]^(3/2) (x^2 + y^2)^(3/2)) e_y
(r: radial coordinate | \[Phi]: azimuthal angle), {{VectorAnalysisResult, 2}, Plaintext} -> x = r cos(\[Phi])
y = r sin(\[Phi])
r = sqrt(x^2 + y^2)
\[Phi] = tan^(-1)(x, y), {{Input, 1}, Input} -> HoldComplete[Grad[1/(2*Pi)^(3/2)*E^(I*k*r)/r, {r}]]}



 Querying Wolfram Alpha for information...
Query: probability flow distribution of spherical wave J = ((hbar)/(2i*m))*((1/(2*Pi)^(3/2)*e^(-i*k*r)/r)*del(((1/(2*Pi)^(3/2)*e^(i*k*r)/r), {r})) - ((1/(2*Pi)^(3/2)*e^(i*k*r)/r)*del(((1/(2*Pi)^(3/2)*e^(-i*k*r)/r), {r})))
Query complete.

Output: {}



 Querying Wolfram Alpha for information...
Query: ((1/(2*Pi)^(3/2)*e^(-i*k*r)/r)*del(((1/(2*Pi)^(3/2)*e^(i*k*r)/r), {r})) - ((1/(2*Pi)^(3/2)*e^(i*k*r)/r)*del(((1/(2*Pi)^(3/2)*e^(-i*k*r)/r), {r})))       
Query complete.

Output: {}



 Querying Wolfram Alpha for information...
Query: ((hbar)/(2i*m))*((1/(2*Pi)^(3/2)*e^(-i*k*r)/r)*((i e^(i k r) r (i + k r))/(2 sqrt(2) Pi^(3/2) r^2)) - (1/(2*Pi)^(3/2)*e^(i k r)/r)*((i e^(i k r) r (i + k r))/(2 sqrt(2) Pi^(3/2) r^2)))
Query complete.

Output: {{{Input, 1}, Plaintext} -> (\[HBar] (reduced Planck constant))/(2 i m) ((1/(2 \[Pi])^(3/2)e^(-i (k r))/r)(i e^(i k r) r (i + k r))/(2 sqrt(2) \[Pi]^(3/2) r^2) - (1/(2 \[Pi])^(3/2)e^(i k r)/r)(i e^(i k r) r (i + k r))/(2 sqrt(2) \[Pi]^(3/2) r^2)), {{Result, 1}, Plaintext} -> (((i (k r + i))/(8 \[Pi]^3 r^2) - (i e^(2 i k r) (k r + i))/(8 \[Pi]^3 r^2)) (0 - 0.5 i) \[HBar] (reduced Planck constants))/m, {{Input, 1}, Input} -> HoldComplete[Quantity[None, ReducedPlanckConstant]/(2*I*m)*((1/(2*Pi)^(3/2)*1/(E^(I*(k*r))*r))*(I*E^(I*k*r)*r*(I + k*r))/(2*Sqrt[2]*Pi^(3/2)*r^2) - (1/(2*Pi)^(3/2)*E^(I*k*r)/r)*(I*E^(I*k*r)*r*(I + k*r))/(2*Sqrt[2]*Pi^(3/2)*r^2))]}


The evaluated probability flow distribution \( \vec{J} \) is:

\[ \vec{J} = \left( \frac{i(kr+i)}{8\pi^3r^2} - \frac{i e^{2i kr}(kr+i)}{8\pi^3r^2} \right) \frac{-0.5i\hbar}{m} \]

Now, let's discuss its behavior as \( r \rightarrow \infty \). As \( r \) tends to infinity, the exponential term \( e^{2i kr} \) oscillates rapidly, resulting in the probability flow distribution oscillating and losing its directional dependence as position \( \vec{r} \) becomes asymptotically far from the origin. This indicates that the probability flux is virtually uniform in all directions at large distances from the source, characteristic of a spherical wave spreading out over space.

Therefore, as \( r \) approaches infinity, the probability flow distribution \( \vec{J} \) becomes less directional and shows a uniform spread in all directions.

This behavior aligns with the expected characteristics of a spherical wave as it propagates outward from the source.

Enter your message: 将做题过程用markdown codeb lock 格式整理出来，用中文        
Received message: 将做题过程用markdown codeb lock 格式整理出来，用中文 in thread thread_RxUKPMWnZrerP2lXr30QhTKk

当计算波函数 $\braket{\vec{r}|\psi}$ 的复共轭时，得到: $$ \braket{\vec{r}|\psi^*} = \frac{1}{(2\pi)^{3/2}}\frac{e^{-ikr}}{r} $$

计算波函数的梯度后，得到: $$ \nabla(\braket{\vec{r}|\psi}) = \left(\frac{i e^{ikr} r (i+kr)}{2\sqrt{2}\pi^{3/2}r^2}, 0\right) $$

利用以上结果计算概率流分布 $\vec{J}$，得到: $$ \vec{J} = \left( \frac{\hbar}{2im} \right) \left( \frac{1}{(2\pi)^{3/2}}\frac{e^{-ikr}}{r} \nabla\left(\frac{1}{(2\pi)^{3/2}}\frac{e^{ikr}}{r}\right) - \frac{1}{(2\pi)^{3/2}}\frac{e^{ikr}}{r} \nabla\left(\frac{1}{(2\pi)^{3/2}}\frac{e^{-ikr}}{r}\right) $$

计算得到: $$ \vec{J} = \left( \frac{i(kr+i)}{8\pi^3r^2} - \frac{i e^{2i kr}(kr+i)}{8\pi^3r^2} \right) \frac{-0.5i\hbar}{m} $$

当 $ r $ 趋近无穷大时，指数项 $ e^{2i kr} $ 将会迅速振荡，导致概率流分布在位置 $ \vec{r} $ 远离原点的时候，失去方向性，即概率流将在距源头较远的地方均匀地分布在 所有方向上。

llm-agent's People

Contributors

osgood001 avatar

Stargazers

avatar avatar avatar avatar avatar avatar

Watchers

avatar

Forkers

kcx1207

llm-agent's Issues

LangChain Template

Modify based on Existing code seems uneasy, we may switch to the LangChain Template for Production ready code.

try wenxinyiyan

  • #4
  • 2. 安装文心一言工具链条，并配置相关环境。
  • 3. 开发 Bot 插件，实现具体的功能。
  • 4. 对 Bot 进行测试和调试，确保其正常运行。
  • 5. 发布 Bot，并进行维护和更新。

Redesign Assistants

I don't have access to openai api keys recently, to fully harness the power of my LLM-Agent with wolfram alpha, I registered the free developer wolfram engine license and consider use local LLM to do so.

LangChain provide auxilary tools and easy to use syntax. So get things done with it and push it to public.

Write a programming guide in coze workflow (Take advantange of variable flow (as a general strucutre) and process it like it were a json) and complete the requirements on loop functionality

  • 1. 收集信息：首先，你需要收集关于 Coze 工作流和变量流的相关信息，包括定义、结构、应用场景等。
  • 2. 确定编程语言：根据收集到的信息，确定使用哪种编程语言来实现 Coze 工作流和变量流。
  • 3. 设计 JSON 结构：根据编程语言和 Coze 工作流的要求，设计 JSON 结构来存储和处理变量流。
  • 4. 实现 Coze 工作流：根据 JSON 结构，使用编程语言实现 Coze 工作流，包括流程定义、变量定义、流程执行等。
  • 5. 实现变量流：根据 JSON 结构，使用编程语言实现变量流，包括变量的读取、写入、修改等。
  • 6. 实现循环功能：根据 Coze 工作流的要求，使用编程语言实现循环功能，包括循环次数、循环条件等。
  • 7. 检查和修改：在完成 Coze 工作流和变量流的实现后，你需要对程序进行检查和修改，确保程序的质量和准确性。
  • 8. 提交程序：在检查和修改后，你需要按照要求提交程序。

Refine the code

Existing code is preliminary, the following will be done latter:

  • add a readme to show how to config and run the code
  • remove redundant contents and make it more concise and well organized
  • #2
  • benchmark it upon other existing bots, backup existing opensource projects related

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.