Views
No views yet
README.md is designed to be the primary model card for your Hugging Face repository. It uses the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license, which strictly prohibits commercial use of the model.1---
2license: cc-by-nc-4.0
3language:
4- en
5tags:
6- digital business consulting
7- custom-chat
8- non-commercial
9model_name: Vikas-AI
10pipeline_tag: text-generation
11---
12
13# Vikas-AI
14
15Vikas-AI is a lightweight, high-performance tiny language model. It has been trained on custom conversational data on a local NVIDIA GeForce RTX 2070 (8GB VRAM) setup.
16
17This model is optimized for efficient inference on consumer hardware, mobile devices, and browser-based environments.
18
19## ⚠️ License and Usage
20
21This model is released under the **[CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/)** license.
22
23- **Non-Commercial Use Only**: You may not use this model or its derivatives for any commercial purposes or financial gain.
24- **Attribution**: You must give appropriate credit if you redistribute or adapt the model.
25- **Usage**: This model is made by VVG ONLINE as POC.
26
27## Vikas AI Model Details
28
29- **Developed by:** VVG ONLINE under supervision of our CEO,CTO, Founder Vaibhav V. Gupta
30- **Training Hardware:** NVIDIA GeForce RTX 2070 with Max-Q Design (8GB VRAM)
31- **Training Method:** LoRA (Low-Rank Adaptation) followed by a full weight merge.
32- **Input Format:** ChatML / `<|im_start|>` prompt format.
33
34## Training Procedure
35
36The model was trained using the following conservative parameters to ensure stability on 8GB hardware:
37
38- **Learning Rate:** 2e-5
39- **Batch Size:** 1 (with Gradient Accumulation 16)
40- **Precision:** FP16
41- **Max Sequence Length:** 256 tokens
42- **Epochs:** 3
43
44### Prompt Template
45```text
46<|im_start|>user
47{prompt}<|im_end|>
48<|im_start|>assistant1from transformers import AutoModelForCausalLM, AutoTokenizer
2import torch
3
4model_id = "VVGONLINE/Vikas-AI"
5tokenizer = AutoTokenizer.from_pretrained(model_id)
6model = AutoModelForCausalLM.from_pretrained(
7 model_id,
8 torch_dtype=torch.float16,
9 device_map="auto"
10)
11
12prompt = "<|im_start|>user\nWhat can you do?<|im_end|>\n<|im_start|>assistant\n"
13inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
14outputs = model.generate(**inputs, max_new_tokens=100)
15print(tokenizer.decode(outputs, skip_special_tokens=True))transformers.js for local browser inference.1import { pipeline } from '@xenova/transformers';
2
3const generator = await pipeline('text-generation', 'VVGONLINE/Vikas-AI');
4const output = await generator('<|im_start|>user\nHi!<|im_end|>\n<|im_start|>assistant\n', {
5 max_new_tokens: 256,
6 temperature: 0.7
7});