65 lines
1.9 KiB
Markdown
65 lines
1.9 KiB
Markdown
|
---
|
||
|
|
language:
|
||
|
|
- en
|
||
|
|
- hi
|
||
|
|
tags:
|
||
|
|
- neuron
|
||
|
|
- neura-tech
|
||
|
|
- 14B
|
||
|
|
- text-generation
|
||
|
|
- qwen2
|
||
|
|
- neural-networks
|
||
|
|
license: apache-2.0
|
||
|
|
datasets:
|
||
|
|
- custom-neura-tech-data
|
||
|
|
metrics:
|
||
|
|
- accuracy
|
||
|
|
---
|
||
|
|
|
||
|
|
# 🧠 Neuron-14B: The Official Intelligence of Neura Tech
|
||
|
|
|
||
|
|
**Neuron-14B** is a high-performance Large Language Model (LLM) developed by **Neura Tech**. It serves as the flagship model for advanced reasoning, creative synthesis, and multilingual communication.
|
||
|
|
|
||
|
|
---
|
||
|
|
|
||
|
|
## 🏢 Organization Identity
|
||
|
|
* **Company**: Neura Tech
|
||
|
|
* **Project Name**: Neuron
|
||
|
|
* **Lead Architect**: Anandnrnnffn
|
||
|
|
|
||
|
|
## 📊 Model Specifications
|
||
|
|
* **Architecture**: Optimized Transformer (Fine-tuned from Qwen2)
|
||
|
|
* **Parameters**: ~15 Billion
|
||
|
|
* **Precision**: BF16 (Bfloat16)
|
||
|
|
* **Context Window**: 131,072 tokens
|
||
|
|
* **License**: Apache-2.0 (Open Source)
|
||
|
|
|
||
|
|
## 🎯 Core Capabilities
|
||
|
|
* **Advanced Reasoning**: Capable of solving complex logical and mathematical queries.
|
||
|
|
* **Multilingual Proficiency**: Highly optimized for English and Hindi (including Hinglish).
|
||
|
|
* **Instruction Following**: Specifically tuned to follow complex user prompts with high precision.
|
||
|
|
* **Creative Synthesis**: Exceptional at generating scripts, stories, and technical documentation.
|
||
|
|
|
||
|
|
## 📜 License & Usage
|
||
|
|
This model is licensed under the Apache-2.0 License. This means you are free to use, modify, and distribute this model, provided that you credit Neura Tech as the original creator.
|
||
|
|
|
||
|
|
## 🛠️ Quick Start (Python)
|
||
|
|
To use **Neuron-14B**, load it via the Hugging Face `transformers` library:
|
||
|
|
|
||
|
|
```python
|
||
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer
|
||
|
|
|
||
|
|
model_id = "Anandnrnnffn/Neura-Tech-14B-Weights"
|
||
|
|
|
||
|
|
# Load Neuron-14B Tokenizer
|
||
|
|
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
||
|
|
|
||
|
|
# Load Model Weights
|
||
|
|
model = AutoModelForCausalLM.from_pretrained(
|
||
|
|
model_id,
|
||
|
|
device_map="auto",
|
||
|
|
torch_dtype="auto"
|
||
|
|
)
|
||
|
|
```
|
||
|
|
|
||
|
|
## © 2026 Neura Tech. All Rights Reserved.
|