Downloads · 30 days
5
10% of all-time downloads
minimind-pretrain
samz/minimind-pretrain
minimind-pretrain is a machine learning model from samz. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
Chinese language model trained on pretrain dataset.
All-time downloads
50
Public
Repo size
215 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pth108 MB · 50%
At a glance
- License
- mit
- Model type
- minimind
- Access
- Public
- Created
- Feb 9, 2025
- Updated
- Feb 9, 2025
- SHA
- afbcd889
Model card
From the Hugging Face model README
MiniMind Pretrained Model
Chinese language model trained on pretrain dataset.
Model Details
- Architecture: Transformer
- Parameters: 26.878M
- Dimensions: 512
- Layers: 8
- Attention Heads: 8
- Vocabulary Size: 32000
- Max Sequence Length: 1024
Training Data
- Pretrained on Chinese text corpus
- Dataset size: 4.33GB
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("samz/minimind-pretrain")
tokenizer = AutoTokenizer.from_pretrained("samz/minimind-pretrain")
text = "今天天气真不错"
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs, max_length=50)
result = tokenizer.decode(outputs[0])
print(result)
- Type
- minimind
- License
- mit
- Languages
- zh
- Created
- Feb 9, 2025
- Updated
- Feb 9, 2025