Instructions to use d0rj/FRED-T5-large-instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use d0rj/FRED-T5-large-instruct with Transformers:

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

pipe = pipeline("text-generation", model="d0rj/FRED-T5-large-instruct")

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("d0rj/FRED-T5-large-instruct")
model = AutoModelForSeq2SeqLM.from_pretrained("d0rj/FRED-T5-large-instruct")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use d0rj/FRED-T5-large-instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "d0rj/FRED-T5-large-instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "d0rj/FRED-T5-large-instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/d0rj/FRED-T5-large-instruct

SGLang

How to use d0rj/FRED-T5-large-instruct 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 "d0rj/FRED-T5-large-instruct" \
    --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": "d0rj/FRED-T5-large-instruct",
		"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 "d0rj/FRED-T5-large-instruct" \
        --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": "d0rj/FRED-T5-large-instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use d0rj/FRED-T5-large-instruct with Docker Model Runner:
```
docker model run hf.co/d0rj/FRED-T5-large-instruct
```

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

FRED-T5-large-instruct

🚧 WIP, still training...

Модель ai-forever/FRED-T5-large, обучаемая на инструктивном датасете. Пока что инструкциям следует не очень.

Веса лежат вместе с состоянием оптимизатора, шедулера и Trainer'а. Можно почти спокойно "продолжать" обучение на своих данных.

Usage

Basic

from transformers import pipeline


pipe = pipeline('text2text-generation', model='d0rj/FRED-T5-large-instruct')
pipe('<SC6>Придумай сказку про красную лягушку<extra_id_0>')

Training

Пока что можно следить за обучением здесь на WandB.

Учится в fp32.

Data

Сконкатенировано из разных переведённых инструктивных датасетов.

Всего 1.1B токенов (1133146852) в обучающем сете, 7506075 примеров system_prompt-question-answer. По Chinchilla статье это ~ в 15 раз меньше, чем нужно (но Chinchilla про претрейн).

Resources

Учится в Kaggle на одной P100. Медленно, но верно (лосс падает, а большего мне и не надо).

Downloads last month: 661

Safetensors

Model size

0.8B params

Tensor type

F32

Model tree for d0rj/FRED-T5-large-instruct

Base model

ai-forever/FRED-T5-large

Finetuned

(2)

this model

d0rj
/

FRED-T5-large-instruct