Model Card for kkk662/histrin-gpt-model

This is a fine-tuned GPT model specialized for technical text in Korean, focusing on railway systems and signaling mechanisms.

Model Details

Model Description

This model is fine-tuned on technical documentation related to railway systems, using a dataset that includes both text and tables extracted from PDFs. It is designed to understand and generate precise Korean technical text.

  • Developed by: kkk662
  • Funded by (optional): Internal Research
  • Shared by (optional): kkk662
  • Model type: GPT-Neo 1.3B
  • Language(s) (NLP): Korean, English
  • License: Apache 2.0
  • Finetuned from model (optional): EleutherAI/gpt-neo-1.3B

Model Sources [optional]

Uses

Direct Use

This model can be directly used for generating technical descriptions, explanations, and summaries related to railway systems in Korean.

Downstream Use [optional]

The model can be fine-tuned further for domain-specific tasks, such as creating operation manuals or translating technical documents into Korean.

Out-of-Scope Use

This model is not suitable for general-purpose NLP tasks outside the technical domain.

Bias, Risks, and Limitations

Recommendations

Users should be aware that the model's performance is highly domain-specific. Misuse in non-technical or unrelated domains may lead to inaccurate or nonsensical results.

How to Get Started with the Model

Use the code below to get started with the model:

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "kkk662/histrin-gpt-model"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

input_text = "ํ•œ๊ตญ ๊ณ ์†์ฒ ๋„ ์‹œ์Šคํ…œ๋Œ€ํ•ด ์„ค๋ช…ํ•ด์ค˜."
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs, max_length=200)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
Downloads last month
149
Safetensors
Model size
125M params
Tensor type
F32
ยท
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for kkk662/histrin-gpt-model

Quantizations
1 model