WWW::Gemini

Raku package for connecting with Google's Gemini. It is based on the Web API described in Gemini's API documentation.

The design and implementation of the package closely follows those of "WWW::PaLM", [AAp1], and "WWW::OpenAI", [AAp2].

Installation

zef install WWW::Gemini

From GitHub:

zef install https://github.com/antononcube/Raku-WWW-Gemini

Usage examples

Show models:

use WWW::Gemini;

gemini-models()

# (models/chat-bison-001 models/text-bison-001 models/embedding-gecko-001 models/gemini-1.0-pro models/gemini-1.0-pro-001 models/gemini-1.0-pro-latest models/gemini-1.0-pro-vision-latest models/gemini-pro models/gemini-pro-vision models/embedding-001 models/aqa)

Show text generation:

.say for gemini-generate-content('what is the population in Brazil?', format => 'values');

# 215,351,056 (2023 est.)

Using a synonym function:

.say for gemini-generation('Who wrote the book "Dune"?');

# {candidates => [{content => {parts => [{text => Frank Herbert}], role => model}, finishReason => STOP, index => 0, safetyRatings => [{category => HARM_CATEGORY_SEXUALLY_EXPLICIT, probability => NEGLIGIBLE} {category => HARM_CATEGORY_HATE_SPEECH, probability => NEGLIGIBLE} {category => HARM_CATEGORY_HARASSMENT, probability => NEGLIGIBLE} {category => HARM_CATEGORY_DANGEROUS_CONTENT, probability => NEGLIGIBLE}]}], promptFeedback => {safetyRatings => [{category => HARM_CATEGORY_SEXUALLY_EXPLICIT, probability => NEGLIGIBLE} {category => HARM_CATEGORY_HATE_SPEECH, probability => NEGLIGIBLE} {category => HARM_CATEGORY_HARASSMENT, probability => NEGLIGIBLE} {category => HARM_CATEGORY_DANGEROUS_CONTENT, probability => NEGLIGIBLE}]}}

Embeddings

Show text embeddings:

use Data::TypeSystem;

my @vecs = gemini-embed-content(["say something nice!",
                            "shout something bad!",
                            "where is the best coffee made?"],
        format => 'values');

say "Shape: ", deduce-type(@vecs);
.say for @vecs;

# Shape: Vector(Vector((Any), 768), 3)
# [-0.031044435 -0.01293638 0.008904989 -0.03263427 0.012885272 0.012842082 0.027659245 -0.022445966 0.001112788 0.009750604 0.04965804 0.009791873 0.043056853 -0.042573094 -0.00023794426 -0.018873844 0.04166457 0.003571331 0.02574701 -0.023280907 0.0059002587 0.07658169 -0.022190008 -0.0016273775 -0.0012887336 -0.019846706 0.017049292 -0.0529704 -0.03620988 0.039227188 -0.03227184 0.04443834 -0.053984605 -0.042696882 0.024596183 -0.062363494 0.0006710476 0.005082995 0.028102735 0.022241268 0.04491562 -0.029794076 -0.026798483 -0.026482908 0.028277056 -0.015406585 0.00509467 0.021505114 -0.015438571 -0.03986329 0.0509833 -0.026030572 -0.0006995414 -0.010695886 -0.0066410583 -0.007820808 0.0438939 0.049049154 -0.044850092 0.007993297 0.033383675 -0.017981231 -0.018700533 0.018778075 0.0132168755 -0.008311568 -0.014669906 0.012169402 0.04730062 -0.030405547 0.00081983255 -0.06609647 0.03425647 -0.025497716 -0.027294777 -0.13895138 -0.032781336 0.06400365 -0.043568406 0.0003324857 -0.02962398 -0.100292206 -0.052770108 -0.039332274 -0.110844955 0.03195597 -0.025506092 0.020022806 0.009940032 0.050653372 -0.0215141 -0.024635738 -0.005630048 -0.07249384 0.002201602 0.050425064 -0.009672865 -0.043219615 0.011732131 0.0021474056 ...]
# [0.015647776 -0.03143455 -0.040937748 -0.03215229 0.00071876345 0.024444472 -0.013541601 -0.023152746 -0.009198697 0.045432128 0.049105383 -0.0029321243 -0.01180009 -0.044899803 0.025208764 -0.034655575 -0.0047017317 -0.012213489 0.04399379 -0.0031873514 0.017236471 0.0302059 -0.009911401 -0.009729893 0.0067022024 0.007181713 0.047008436 -0.07261744 -0.034597974 0.024742194 -0.049818117 0.047178492 -0.05239612 0.004519767 -0.031770747 -0.084419765 -0.009777399 0.016744321 0.057470873 -0.0008020468 0.0034762728 0.00092191494 -0.012497095 -0.016705612 0.035845324 -0.006399285 0.024325574 -0.002022227 0.002841698 -0.080434985 0.031044094 -0.030748611 0.028753037 -0.036181286 0.00089683925 -0.060913764 0.056072846 0.052506663 -0.03612298 0.030202417 0.017606454 0.0070932843 -0.024513774 0.031895094 0.02637744 -0.048618294 -0.028836755 0.0009879556 0.09044918 0.006353638 0.042316265 -0.053601615 0.0047276323 -0.010463116 -0.03282551 -0.11493026 -0.037989974 0.066172026 -0.042627145 0.0338829 -0.022461282 -0.03243943 -0.0322086 -0.022295073 -0.06273048 -0.020984378 -0.018161116 0.024409954 0.052115202 0.07886858 -0.015862696 -0.0064231334 0.021302955 -0.055992328 -0.01627737 0.074991345 -0.026166856 -0.010036745 0.027091224 -0.024027428 ...]
# [-0.01054143 -0.03587494 -0.013359126 -0.051338553 0.047345687 0.061076418 -0.03205643 -0.04106156 0.019298773 0.024068715 -0.061305694 0.015391013 0.027234804 0.0031952043 -0.003338059 0.013618424 0.015242139 0.037548818 -0.016279383 -0.04948622 0.0051141633 -0.0003462428 0.040165897 0.02506596 -0.015313406 -0.045644026 -0.011603705 -0.02260518 -0.08389293 0.028939478 -0.05874142 0.057905953 -0.043423664 -0.018679587 0.048761897 -0.020560026 -0.005545347 0.05018316 0.03381427 0.044278715 -0.027549185 -0.00721632 -0.059679322 0.011390549 0.005978451 -0.03662131 -0.0102969 0.0036583815 -0.03553423 -0.05129432 -0.004552413 -0.020376777 0.041254595 0.007501306 0.004778074 0.0012881727 0.013960463 -0.0024828897 -0.038222134 0.03442903 0.00025496862 -0.0054893643 -0.022218652 0.058740005 0.023132037 -0.01518507 -0.032717314 0.027132453 0.052209098 -0.03427743 -0.044157255 -0.019799082 0.07553776 -0.04805355 -0.067444116 -0.07435607 -0.05729862 0.062428243 0.042084947 0.014766257 -0.034525547 -0.08122034 0.032142013 0.009679542 -0.025443 -0.009691211 -0.05170164 0.055666693 0.0062296097 0.048578836 -0.020820985 0.031091962 0.015229997 -0.040769666 -0.0019469284 0.021296106 -0.0154602695 -0.011517182 -0.00992539 -0.032775857 ...]

Counting tokens

Here we show how to find the number of tokens in a text:

my $text = q:to/END/;
AI has made surprising successes but cannot solve all scientific problems due to computational irreducibility.
END

gemini-count-tokens($text, format => 'values');

# 20

Vision

If the function gemini-completion is given a list of images, textual results corresponding to those images is returned. The argument "images" is a list of image URLs, image file names, or image Base64 representations. (Any combination of those element types.)

Here is an example with this image:

my $fname = $*CWD ~ '/resources/ThreeHunters.jpg';
my @images = [$fname,];
say gemini-generation("Give concise descriptions of the images.", :@images, format => 'values');

# The image shows a family of raccoons in a tree. The mother raccoon is watching over her two cubs. The cubs are playing with each other. There are butterflies flying around the tree. The leaves on the tree are turning brown and orange.

When a file name is given to the argument "images" of gemini-completion then the function encode-image of "Image::Markup::Utilities", [AAp4], is applied to it.

Command Line Interface

Maker suite access

The package provides a Command Line Interface (CLI) script:

gemini-prompt --help

# Usage:
#   gemini-prompt [<words> ...] [--path=<Str>] [-n[=UInt]] [--mt|--max-output-tokens[=UInt]] [-m|--model=<Str>] [-t|--temperature[=Real]] [-a|--auth-key=<Str>] [--timeout[=UInt]] [-f|--format=<Str>] [--method=<Str>] -- Command given as a sequence of words.
#   
#     --path=<Str>                       Path, one of 'generateContent', 'embedContent', 'countTokens', or 'models'. [default: 'generateContent']
#     -n[=UInt]                          Number of completions or generations. [default: 1]
#     --mt|--max-output-tokens[=UInt]    The maximum number of tokens to generate in the completion. [default: 100]
#     -m|--model=<Str>                   Model. [default: 'Whatever']
#     -t|--temperature[=Real]            Temperature. [default: 0.7]
#     -a|--auth-key=<Str>                Authorization key (to use Gemini API.) [default: 'Whatever']
#     --timeout[=UInt]                   Timeout. [default: 10]
#     -f|--format=<Str>                  Format of the result; one of "json", "hash", "values", or "Whatever". [default: 'values']
#     --method=<Str>                     Method for the HTTP POST query; one of "tiny" or "curl". [default: 'tiny']

Remark: When the authorization key argument "auth-key" is specified set to "Whatever" then gemini-prompt attempts to use one of the env variables GEMINI_API_KEY or PALM_API_KEY.

Mermaid diagram

The following flowchart corresponds to the steps in the package function gemini-prompt:

graph TD
	UI[/Some natural language text/]
	TO[/"Gemini<br/>Processed output"/]
	WR[[Web request]]
	Gemini{{Gemini}}
	PJ[Parse JSON]
	Q{Return<br>hash?}
	MSTC[Compose query]
	MURL[[Make URL]]
	TTC[Process]
	QAK{Auth key<br>supplied?}
	EAK[["Try to find<br>GEMINI_API_KEY<br>or<br>PALM_API_KEY<br>in %*ENV"]]
	QEAF{Auth key<br>found?}
	NAK[/Cannot find auth key/]
	UI --> QAK
	QAK --> |yes|MSTC
	QAK --> |no|EAK
	EAK --> QEAF
	MSTC --> TTC
	QEAF --> |no|NAK
	QEAF --> |yes|TTC
	TTC -.-> MURL -.-> WR -.-> TTC
	WR -.-> |URL|Gemini 
	Gemini -.-> |JSON|WR
	TTC --> Q 
	Q --> |yes|PJ
	Q --> |no|TO
	PJ --> TO

References

Articles

[AA1] Anton Antonov, "Workflows with LLM functions", (2023), RakuForPredictions at WordPress.

[AA2] Anton Antonov, "Number guessing games: Gemini vs ChatGPT" (2023), RakuForPredictions at WordPress.

[ZG1] Zoubin Ghahramani, "Introducing Gemini 2", (2023), Google Official Blog on AI.

Packages, platforms

[AAp1] Anton Antonov, WWW::PaLM Raku package, (2023-2024), GitHub/antononcube.

[AAp2] Anton Antonov, WWW::OpenAI Raku package, (2023-2024), GitHub/antononcube.

[AAp3] Anton Antonov, LLM::Functions Raku package, (2023-2024), GitHub/antononcube.

[AAp4] Anton Antonov, Image::Markup::Utilities Raku package, (2023-2024), GitHub/antononcube.

[AAp5] Anton Antonov, ML::FindTextualAnswer Raku package, (2023-2024), GitHub/antononcube.