File size: 3,037 Bytes
f4e2ac8 db0ca04 f4e2ac8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 |
---
license: mit
datasets:
- fka/awesome-chatgpt-prompts
- gopipasala/fka-awesome-chatgpt-prompts
metrics:
- character
base_model:
- meta-llama/Llama-3.2-11B-Vision-Instruct
new_version: meta-llama/Llama-3.1-8B-Instruct
pipeline_tag: summarization
library_name: diffusers
tags:
- legal
---
Model Card for Rithu Paran's Summarization Model
Model Details
Model Description
Purpose: This model is designed for text summarization, specifically built to condense long-form content into concise, meaningful summaries.
Developed by: Rithu Paran
Model type: Transformer-based Language Model for Summarization
Base Model: Meta-Llama/Llama-3.2-11B-Vision-Instruct
Finetuned Model Version: Meta-Llama/Llama-3.1-8B-Instruct
Language(s): Primarily English, with limited support for other languages.
License: MIT License
Model Sources
Repository: Available on Hugging Face Hub under Rithu Paran
Datasets Used: fka/awesome-chatgpt-prompts, gopipasala/fka-awesome-chatgpt-prompts
Uses
Direct Use
This model can be directly employed for summarizing various types of content, such as news articles, reports, and other informational documents.
Out-of-Scope Use
It is not recommended for highly technical or specialized documents without additional fine-tuning or adaptation.
Bias, Risks, and Limitations
While this model was designed to be general-purpose, there may be inherent biases due to the training data. Users should be cautious when using the model for sensitive content or in applications where accuracy is crucial.
How to Get Started with the Model
Here's a quick example of how to start using the model for summarization:
python
Copy code
from transformers import pipeline
summarizer = pipeline("summarization", model="rithu-paran/your-summarization-model")
text = "Insert long-form text here."
summary = summarizer(text, max_length=100, min_length=30)
print(summary)
Training Details
Training Data
Datasets: fka/awesome-chatgpt-prompts, gopipasala/fka-awesome-chatgpt-prompts
Preprocessing: Data was tokenized and normalized for better model performance.
Training Procedure
Hardware: Trained on GPUs with Hugging Face API resources.
Precision: Mixed-precision (fp16) was utilized to enhance training efficiency.
Training Hyperparameters
Batch Size: 16
Learning Rate: 5e-5
Epochs: 3
Optimizer: AdamW
Evaluation
Metrics
Metrics Used: ROUGE Score, BLEU Score
Evaluation Datasets: Evaluated on a subset of fka/awesome-chatgpt-prompts for summarization performance.
Technical Specifications
Model Architecture
Based on Llama-3 architecture, optimized for summarization through attention-based mechanisms.
Compute Infrastructure
Hardware: Nvidia A100 GPUs were used for training.
Software: Hugging Face’s transformers library along with the diffusers library.
Environmental Impact
Hardware Type: Nvidia A100 GPUs
Training Duration: ~10 hours
Estimated Carbon Emission: Approximate emissions calculated using Machine Learning Impact calculator.
Contact
For any questions or issues, please reach out to Rithu Paran via the Hugging Face Forum.
|