agentlans/GIST-all-MiniLM-L6-v2-quality-v3

GIST-all-MiniLM-L6-v2 Text Quality Model

This model rates the quality of English text for AI learning. Input a text string, and it outputs a numeric quality score reflecting overall informativeness and usefulness.

Performance

On the evaluation set, it achieved:

• Loss: 0.1572
• MSE: 0.1572
• Combined Score: 0.1572
• Tokens processed during training: 102,398,720

Usage Example

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

model_name = "agentlans/GIST-all-MiniLM-L6-v2-quality-v3"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name).to("cuda" if torch.cuda.is_available() else "cpu")

# Higher scores indicate higher text quality.
# The sign of the score has no particular meaning.
# For example, a negative score doesn't necessarily mean that the text is low quality.
def quality(text):
    inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True).to(model.device)
    with torch.no_grad():
        score = model(**inputs).logits.squeeze().cpu().item()
    return score

print(quality("Your text here."))

Limitations

• Works best on non-fiction and general-purpose texts.
• Scores give an overall quality estimate but don’t explain why.
• For higher accuracy, try agentlans/deberta-v3-base-quality-v3.
• Check for biases and suitability before use.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Combined Score	Input Tokens Seen
0.1777	1.0	10000	0.2354	0.2354	0.2354	10239872
0.1389	2.0	20000	0.1572	0.1572	0.1572	20479744
0.1	3.0	30000	0.1961	0.1961	0.1961	30719616
0.0687	4.0	40000	0.1596	0.1596	0.1596	40959488
0.0559	5.0	50000	0.1757	0.1757	0.1757	51199360
0.0409	6.0	60000	0.1677	0.1677	0.1677	61439232
0.0319	7.0	70000	0.1852	0.1852	0.1852	71679104
0.0266	8.0	80000	0.1840	0.1840	0.1840	81918976
0.0202	9.0	90000	0.1724	0.1724	0.1724	92158848
0.0172	10.0	100000	0.1731	0.1731	0.1731	102398720

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0