You can deploy this model and start earning money today!

singhsumony2j/
SeedGPT-V2

Name: SeedGPT-V2
Availability: InStock

Views

No views yet

Model Card for SeedGPT-V2

Model Description

This is a custom text generation model trained on the TinyStories & stories dataset. It follows a Transformer-based architecture compatible with 🤗 Hugging Face's AutoModelForCausalLM interface, using a custom configuration and model class.

Developed by: Sumon Singh
Model type: Causal Language Model (Transformer-based)
Language(s) (NLP): English
License: MIT

Model Sources

DataSet
1. https://huggingface.co/datasets/roneneldan/TinyStories
2. https://huggingface.co/datasets/delphi-suite/stories
Repository: https://github.com/sumony2j/SeedGPT-22M
Demo: https://huggingface.co/spaces/singhsumony2j/SeedGPT-22M

Uses

Direct Use

This model can be used for generating coherent and contextually relevant text based on a given input prompt. Common use cases include:

Creative writing
Story generation
Dialogue simulation

Out-of-Scope Use

Do not use for factual generation where accuracy is critical (e.g., medical, legal, or financial advice)
Avoid use in safety-critical applications without appropriate evaluation

Bias, Risks, and Limitations

As with all large language models, outputs may reflect biases present in the training data. The Refined BookCorpus, while cleaned, may still contain societal, cultural, or gender-based biases.

May hallucinate or fabricate facts
Not intended for real-time or safety-critical decision-making
Cannot guarantee alignment with user values or expectations

Recommendations

Users should be made aware of the risks, biases and limitations of the model. Users should evaluate outputs critically, especially for use cases involving ethics, safety, or factual integrity. Always verify generated content.

How to Get Started with the Model

python

from transformers import AutoTokenizer, AutoModelForCausalLM
from transformers.models.auto.configuration_auto import CONFIG_MAPPING
from transformers.models.auto.modeling_auto import MODEL_FOR_CAUSAL_LM_MAPPING
from src.convert_to_hf_model import HFTransformerConfig, HFTransformerModel  # custom config and model
import torch

# Register custom config and model
CONFIG_MAPPING.register("hf_transformer", HFTransformerConfig)
MODEL_FOR_CAUSAL_LM_MAPPING.register(HFTransformerConfig, HFTransformerModel)

tokenizer = AutoTokenizer.from_pretrained("singhsumony2j/SeedGPT-V2")
model = AutoModelForCausalLM.from_pretrained("singhsumony2j/SeedGPT-V2", low_cpu_mem_usage=True)

prompt = "The little boy"
tokens = tokenizer(prompt)
input_tokens = torch.tensor(tokens.input_ids,dtype=torch.long)[None,:]
response = model.generate(input_tokens,max_num_tokens=100,temp=0.8)
output = tokenizer.decode(response[0].tolist(),skip_special_tokens=True)
print(output)

Citation

bibtex

@misc{singh2025seedgptv2,
  author = {Sumon Singh},
  title = {SeedGPT-V2: A story generation language model},
  year = {2025},
  howpublished = {\url{https://huggingface.co/singhsumony2j/SeedGPT-V2}}
}

Model Card Authors

Sumon Singh

Model Card Contact

singhsumony2j@gmail.com