Instructions to use datificate/gpt2-small-spanish with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use datificate/gpt2-small-spanish with Transformers:

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

pipe = pipeline("text-generation", model="datificate/gpt2-small-spanish")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("datificate/gpt2-small-spanish")
model = AutoModelForCausalLM.from_pretrained("datificate/gpt2-small-spanish", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use datificate/gpt2-small-spanish with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "datificate/gpt2-small-spanish"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "datificate/gpt2-small-spanish",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/datificate/gpt2-small-spanish

SGLang

How to use datificate/gpt2-small-spanish 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 "datificate/gpt2-small-spanish" \
    --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": "datificate/gpt2-small-spanish",
		"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 "datificate/gpt2-small-spanish" \
        --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": "datificate/gpt2-small-spanish",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use datificate/gpt2-small-spanish with Docker Model Runner:
```
docker model run hf.co/datificate/gpt2-small-spanish
```

datificate commited on Dec 27, 2020

Commit

6abafff

1 Parent(s): 3f08be3

Update README.md

Browse files

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -9,7 +9,7 @@ datasets:
 ---
 La descripción en Español se encuentra después de la descripción en Inglés.
-# GPT-2 small spanish: a Language Model for Spanish text generation (and more NLP tasks...)
 GPT-2 small spanish is a state-of-the-art language model for Spanish based on the GPT-2 small model.
 It was trained on Spanish Wikipedia using **Transfer Learning and Fine-tuning techniques**. The training took around 70 hours of training with four GPU NVIDIA GTX 1080-Ti with 11GB of DDR5 and with around han 3GB of (processed) training data.
@@ -34,7 +34,7 @@ The training was possible thanks to the computing power of several GPUs (GPU NVI
 As stated before, this work is mainly based in the work of [Pierre GUILLOU](https://www.linkedin.com/in/pierreguillou/).
-# GPT-2 small spanish: un modelo de lenguaje para generación de texto en Español (y algunas otras tareas de NLP...)
 GPT-2 small spanish es un modelo de lenguaje de vanguardia en Español basado en el modelo pequeño GPT-2.
@@ -55,7 +55,7 @@ La versión preliminar del modelo se encuentra en Hugging Face.
 ## Autores
 El modelo fue entreando y evaluado por [Josué Obregon](https://www.linkedin.com/in/josue-obregon/) y [Berny Carrera](https://www.linkedin.com/in/bernycarrera/), fundadores de [Datificate](https://datificate.com), un espacio para aprender Machine Learning en Español.
-El entrenamiento fue posible gracias al poder computacional de varias GPUs (GPU NVIDIA GTX1080-Ti) de el Laboratorio de Inteligencia Artificial Industrial [IAI Lab](http://iai.khu.ac.kr/) (Universidad de Kyung Hee) al cual Josué pertenece como investigador Postdoctoral en Inteligencia Artificial Industrial.
 Como fue mencionado anteriormente, este trabajo esta basado en el trabajo de [Pierre GUILLOU](https://www.linkedin.com/in/pierreguillou/).

 ---
 La descripción en Español se encuentra después de la descripción en Inglés.
+#(English) GPT-2 small spanish: a Language Model for Spanish text generation (and more NLP tasks...)
 GPT-2 small spanish is a state-of-the-art language model for Spanish based on the GPT-2 small model.
 It was trained on Spanish Wikipedia using **Transfer Learning and Fine-tuning techniques**. The training took around 70 hours of training with four GPU NVIDIA GTX 1080-Ti with 11GB of DDR5 and with around han 3GB of (processed) training data.
 As stated before, this work is mainly based in the work of [Pierre GUILLOU](https://www.linkedin.com/in/pierreguillou/).
+# (Español) GPT-2 small spanish: un modelo de lenguaje para generación de texto en Español (y algunas otras tareas de NLP...)
 GPT-2 small spanish es un modelo de lenguaje de vanguardia en Español basado en el modelo pequeño GPT-2.
 ## Autores
 El modelo fue entreando y evaluado por [Josué Obregon](https://www.linkedin.com/in/josue-obregon/) y [Berny Carrera](https://www.linkedin.com/in/bernycarrera/), fundadores de [Datificate](https://datificate.com), un espacio para aprender Machine Learning en Español.
+El entrenamiento fue posible gracias al poder computacional de varias GPUs (GPU NVIDIA GTX1080-Ti) del Laboratorio de Inteligencia Artificial Industrial [IAI Lab](http://iai.khu.ac.kr/) (Universidad de Kyung Hee) al cual Josué pertenece como investigador Postdoctoral en Inteligencia Artificial Industrial.
 Como fue mencionado anteriormente, este trabajo esta basado en el trabajo de [Pierre GUILLOU](https://www.linkedin.com/in/pierreguillou/).