How to use from
llama.cppInstall from WinGet (Windows)
winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Rohanify/HinTexta-34M-GGUF:F16# Run inference directly in the terminal:
llama cli -hf Rohanify/HinTexta-34M-GGUF:F16Build from source code
git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf Rohanify/HinTexta-34M-GGUF:F16# Run inference directly in the terminal:
./build/bin/llama cli -hf Rohanify/HinTexta-34M-GGUF:F16Use Docker
docker model run hf.co/Rohanify/HinTexta-34M-GGUF:F16Quick Links
HinTexta-34M — Hinglish Chat Model
A 34M parameter GPT-2 model trained from scratch on 1M Hinglish conversations. Speaks naturally in Hinglish (Hindi + English mix).
Quick Start (Ollama)
ollama run hf.co/Rohanify/HinTexta-34M-GGUF:F16
No setup needed!
Details
- Architecture: GPT-2 (8 layers, 512 dim, 8 heads)
- Parameters: 34M
- Training data: 1M synthetic Hinglish conversations (Abhishekcr448/Hinglish-Everyday-Conversations-1M)
- Tokenizer: Custom 16K vocab ByteLevel BPE trained on Hinglish
- Context window: 512 tokens
- Trained from scratch on a single RTX 5080
Example
You: Bhai kya haal hai?
Bot: Bahut accha! Lekin mujhe toh ek adventurous trip ki yaad hai, dost!
You: Movie dekhne chalein?
Bot: Movie night toh awesome hoga, par ek adventurous trip bhi zaroori hai!
- Downloads last month
- 30
Hardware compatibility
Log In to add your hardware
16-bit
Install (macOS, Linux)
# Start a local OpenAI-compatible server with a web UI: llama serve -hf Rohanify/HinTexta-34M-GGUF:F16# Run inference directly in the terminal: llama cli -hf Rohanify/HinTexta-34M-GGUF:F16