Instructions to use KomeijiForce/bart-large-emojilm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KomeijiForce/bart-large-emojilm with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="KomeijiForce/bart-large-emojilm")

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("KomeijiForce/bart-large-emojilm")
model = AutoModelForSeq2SeqLM.from_pretrained("KomeijiForce/bart-large-emojilm")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use KomeijiForce/bart-large-emojilm with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KomeijiForce/bart-large-emojilm"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KomeijiForce/bart-large-emojilm",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/KomeijiForce/bart-large-emojilm

SGLang

How to use KomeijiForce/bart-large-emojilm with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "KomeijiForce/bart-large-emojilm" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KomeijiForce/bart-large-emojilm",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "KomeijiForce/bart-large-emojilm" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KomeijiForce/bart-large-emojilm",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use KomeijiForce/bart-large-emojilm with Docker Model Runner:
```
docker model run hf.co/KomeijiForce/bart-large-emojilm
```

YAML Metadata Warning:The pipeline tag "text2text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

EmojiLM

This is a BART model pre-trained on the Text2Emoji dataset to translate setences into series of emojis.

For instance, "I love pizza" will be translated into "🍕😍".

An example implementation for translation:

from transformers import BartTokenizer, BartForConditionalGeneration

def translate(sentence, **argv):
    inputs = tokenizer(sentence, return_tensors="pt")
    generated_ids = generator.generate(inputs["input_ids"], **argv)
    decoded = tokenizer.decode(generated_ids[0], skip_special_tokens=True).replace(" ", "")
    return decoded

path = "KomeijiForce/bart-large-emojilm"
tokenizer = BartTokenizer.from_pretrained(path)
generator = BartForConditionalGeneration.from_pretrained(path)

sentence = "I love the weather in Alaska!"
decoded = translate(sentence, num_beams=4, do_sample=True, max_length=100)
print(decoded)

You will probably get some output like "❄️🏔️😍".

If you find this model & dataset resource useful, please consider cite our paper:

@article{DBLP:journals/corr/abs-2311-01751,
  author       = {Letian Peng and
                  Zilong Wang and
                  Hang Liu and
                  Zihan Wang and
                  Jingbo Shang},
  title        = {EmojiLM: Modeling the New Emoji Language},
  journal      = {CoRR},
  volume       = {abs/2311.01751},
  year         = {2023},
  url          = {https://doi.org/10.48550/arXiv.2311.01751},
  doi          = {10.48550/ARXIV.2311.01751},
  eprinttype    = {arXiv},
  eprint       = {2311.01751},
  timestamp    = {Tue, 07 Nov 2023 18:17:14 +0100},
  biburl       = {https://dblp.org/rec/journals/corr/abs-2311-01751.bib},
  bibsource    = {dblp computer science bibliography, https://dblp.org}
}

Downloads last month: 654,729

Dataset used to train KomeijiForce/bart-large-emojilm

Paper for KomeijiForce/bart-large-emojilm

EmojiLM: Modeling the New Emoji Language

Paper • 2311.01751 • Published Nov 3, 2023 • 3