Instructions to use VedaX-Labs/VedaX-0.7B-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use VedaX-Labs/VedaX-0.7B-Base with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="VedaX-Labs/VedaX-0.7B-Base")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("VedaX-Labs/VedaX-0.7B-Base")
model = AutoModelForCausalLM.from_pretrained("VedaX-Labs/VedaX-0.7B-Base")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use VedaX-Labs/VedaX-0.7B-Base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "VedaX-Labs/VedaX-0.7B-Base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "VedaX-Labs/VedaX-0.7B-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/VedaX-Labs/VedaX-0.7B-Base

SGLang

How to use VedaX-Labs/VedaX-0.7B-Base with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "VedaX-Labs/VedaX-0.7B-Base" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "VedaX-Labs/VedaX-0.7B-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "VedaX-Labs/VedaX-0.7B-Base" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "VedaX-Labs/VedaX-0.7B-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use VedaX-Labs/VedaX-0.7B-Base with Docker Model Runner:
```
docker model run hf.co/VedaX-Labs/VedaX-0.7B-Base
```

VedaX-0.7B-Base

VedaX-0.7B-Base is a lightweight open-weight language model developed by VedaX Labs. The model is designed for efficient conversational AI, reasoning tasks, and lightweight deployment on consumer hardware. The model is entirely trained on CPU and on cosmopedia dataset

Training Specs

CPU:- Intel(R) Core(TM) i3-14100
Memory:- 16.0 GB DDR5
GPU:- None (Intel UHD Graphics 730)

Model Overview

Model Name: VedaX-0.7B-Base
Organization: VedaX Labs
Parameters: 0.7 Billion
Architecture: Transformer-based causal language model
Type: Base/Foundation Model
License: Apache-2.0

Capabilities

Text generation
Lightweight inference
Fine-tuning friendly

Intended Use

VedaX-0.7B-Base is intended for:

Research and experimentation
Educational purposes
Building lightweight AI assistants
Fine-tuning for downstream tasks
Local inference applications
Suitable for all devices

Context Length

Context Window: 4096 tokens

Training

The model was trained on a mixture of publicly available text datasets and curated instruction-style data.

Example Usage

from transformers import AutoTokenizer, AutoModelForCausalLM

model_id = "VedaX-Labs/VedaX-0.7B-Base"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

prompt = "The magic forest was"

inputs = tokenizer(prompt, return_tensors="pt")

outputs = model.generate(
    **inputs,
    max_new_tokens=128
)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Benchmarking(Higher number means higher score)

Limitations

May generate inaccurate information
Limited reasoning compared to larger models
Responses may contain hallucinations
Performance depends on prompting quality
May generate repettetive words(Chances is very low)

Future Plans

Instruction-tuned versions
Reasoning-enhanced variants
Multimodal support
Larger parameter models

Developed by VedaX Labs.

Downloads last month: 316

Safetensors

Model size

0.1B params

Tensor type

F32

VedaX-Labs
/

VedaX-0.7B-Base