Downloads · 30 days
19
14% of all-time downloads
Nova-AGI-EXP
VoidWalkercero/Nova-AGI-EXP
Nova-AGI-EXP is a text generation model from VoidWalkercero. Use it when you need the model to write or continue text. The card lists the license as mit.
Nova is an experimental AGI-focused language model fine-tuned for reasoning and logical thinking.
All-time downloads
137
Public
Parameters
355M
1.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.4 GB · 100%
At a glance
- Task
- Text Generation
- License
- mit
- Model type
- gpt2
- Access
- Public
- Created
- Dec 25, 2025
- Updated
- Dec 25, 2025
- SHA
- fa8e3c46
Try a prompt
Model card
From the Hugging Face model README
🌟 Nova-AGI-EXP
Nova is an experimental AGI-focused language model fine-tuned for reasoning and logical thinking.
🧠 Capabilities
- ✅ Deductive reasoning
- ✅ Inductive reasoning
- ✅ Abductive reasoning
- ✅ Mathematical sequences
- ✅ Causal reasoning
- ✅ Metacognition (self-awareness of limitations)
- ✅ Contextual understanding
- ✅ Probabilistic thinking
📊 Training Details
| Attribute | Value |
|---|---|
| Base Model | GPT-2 Medium (355M params) |
| Training Examples | 49 reasoning pairs |
| Final Loss | 0.066 |
| Framework | PyTorch + Transformers |
🚀 Quick Start
from transformers import GPT2LMHeadModel, GPT2Tokenizer
model = GPT2LMHeadModel.from_pretrained("VoidWalkercero/Nova-AGI-EXP")
tokenizer = GPT2Tokenizer.from_pretrained("VoidWalkercero/Nova-AGI-EXP")
prompt = "User: What comes next: 2, 4, 6, 8, ?\nAssistant:"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(
inputs["input_ids"],
max_new_tokens=100,
temperature=0.3,
top_p=0.85,
repetition_penalty=1.3
)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
- Task
- Text Generation
- Type
- gpt2
- License
- mit
- Languages
- en
- Created
- Dec 25, 2025
- Updated
- Dec 25, 2025