You can deploy this model and start earning money today!

rajivranjank/
bhojpuri-savage-3b

Name: bhojpuri-savage-3b
Availability: InStock

gguf

endpoints_compatible

conversational

Views

No views yet

from huggingface_hub import HfApi

api = HfApi() repo_id = "rajivranjank/bhojpuri-savage-3b"

readme = """--- language:

hi
bho tags:
qwen2.5
bhojpuri
hindi
gguf
lora
unsloth license: apache-2.0

🔥 Bhojpuri Savage Qwen2.5-3B

A fine-tuned version of Qwen2.5-3B-Instruct on Desi Hindi/Bhojpuri savage conversational data.

Replies in Hindi/Bhojpuri with desi humor and savage wit. 😄

Example Output

User: Bhai kya haal ba?

Model: Tohar hi-hello dekh ke lagta hai tu Mughal Sarai mein bhul gail ba. Patna se Patna tak ka fast train chala jaye, bas dimaag ko thoda upar bhaag le, bol babua!

User: Kuch savage bol bhai

Model: Tohar code dekh ke toh Bhojpuri mein gaari dene ka mann karat ba. Tera savage wala style teri galti se seedha Patna Junction pe bhule hua train jaisa lag raha hai!

🚀 How to Run

Ollama (Easiest)

ollama run hf.co/rajivranjank/bhojpuri-savage-3b

LM Studio

Download LM Studio
Search rajivranjank/bhojpuri-savage-3b
Download and chat!

llama.cpp

./llama-cli -m bhojpuri-savage-q4km.gguf -p "Bhai kya haal ba?" -n 200

🛠️ Training Details

Detail	Value
Base Model	Qwen2.5-3B-Instruct
Method	QLoRA (4-bit) with Unsloth
LoRA Rank	32
Sequence Length	768
Format	GGUF Q4_K_M
Size	1.93 GB
Language	Hindi / Bhojpuri
"""

with open("/kaggle/working/README.md", "w") as f: f.write(readme)

api.upload_file( path_or_fileobj = "/kaggle/working/README.md", path_in_repo = "README.md", repo_id = repo_id, repo_type = "model", ) print(f"✅ Model card added!") print(f"🔗 https://huggingface.co/{repo_id}")