Skip to contents

huggingfaceR

An API-first R package for accessing 500,000+ machine learning models, embeddings, and datasets on the Hugging Face Hub. No Python required.

Installation 

# install.packages("devtools")
devtools::install_github("farach/huggingfaceR")

Setup

Get a free API token from huggingface.co/settings/tokens, then configure it in R:

library(huggingfaceR)

hf_set_token("hf_your_token_here", store = TRUE)
hf_whoami()

Text Classification 

# Sentiment analysis
hf_classify("I love using R for data science!")
#> # A tibble: 1 x 3
#>   text                              label    score
#>   <chr>                             <chr>    <dbl>
#> 1 I love using R for data science!  POSITIVE 0.999

# Zero-shot classification with custom labels
hf_classify_zero_shot(
  "I just bought a new laptop for coding",
  labels = c("technology", "sports", "politics", "food")
)

Embeddings and Similarity 

sentences <- c(
  "The cat sat on the mat",
  "A feline rested on the rug",
  "The dog played in the park"
)

embeddings <- hf_embed(sentences)
embeddings
#> # A tibble: 3 x 3
#>   text                        embedding    n_dims
#>   <chr>                       <list>        <int>
#> 1 The cat sat on the mat      <dbl [384]>     384
#> 2 A feline rested on the rug  <dbl [384]>     384
#> 3 The dog played in the park  <dbl [384]>     384

hf_similarity(embeddings)
#> # A tibble: 3 x 3
#>   text_1                  text_2                      similarity
#>   <chr>                   <chr>                            <dbl>
#> 1 The cat sat on the mat  A feline rested on the rug       0.89
#> 2 The cat sat on the mat  The dog played in the park       0.45
#> 3 A feline rested on ...  The dog played in the park       0.39

Chat with Open-Source LLMs 

hf_chat("What is the tidyverse?")

# With a system prompt
hf_chat(
  "Explain logistic regression in two sentences.",
  system = "You are a statistics instructor. Use plain language."
)

# Multi-turn conversation
convo <- hf_conversation(system = "You are a helpful R tutor.")
convo <- chat(convo, "How do I read a CSV file?")
convo <- chat(convo, "What about Excel files?")

Text Generation 

hf_generate("Once upon a time in a land far away,", max_new_tokens = 100)

hf_fill_mask("The capital of France is [MASK].")
#> # A tibble: 5 x 4
#>   text                              token   score filled
#>   <chr>                             <chr>   <dbl> <chr>
#> 1 The capital of France is [MASK].  paris   0.88  The capital of France is paris.
#> 2 The capital of France is [MASK].  lyon    0.03  The capital of France is lyon.
#> ...

Tidyverse Integration

All functions accept character vectors and return tibbles.

library(dplyr)
library(tidyr)

reviews <- tibble(
  id = 1:3,
  text = c(
    "This product is amazing!",
    "Terrible experience.",
    "It's okay, nothing special."
  )
)

reviews |>
  mutate(sentiment = hf_classify(text)) |>
  unnest(sentiment, names_sep = "_") |>
  select(id, text, sentiment_label, sentiment_score)

Tidymodels

Use embeddings as features in machine learning workflows:

library(tidymodels)

rec <- recipe(sentiment ~ text, data = train_data) |>
  step_hf_embed(text)

wf <- workflow() |>
  add_recipe(rec) |>
  add_model(logistic_reg()) |>
  fit(data = train_data)

Tidytext

Semantic search and document clustering:

docs |>
  hf_embed_text(text) |>
  hf_nearest_neighbors("machine learning", k = 5)

docs |>
  hf_embed_text(text) |>
  hf_cluster_texts(k = 3) |>
  hf_extract_topics(text_col = "text", k = 3)

Hub and Datasets 

# Search models
hf_search_models(task = "text-classification", limit = 10)

# Load datasets into tibbles (no Python needed)
imdb <- hf_load_dataset("imdb", split = "train", limit = 1000)

Learn More

License

MIT