SG-Tamil-MoE-7B

SG-Tamil-MoE-7B is a Mixture of Experts (MoE) made with the following models using LazyMergekit:

• Hemanth-thunder/Tamil-Mistral-7B-v0.1
• Hemanth-thunder/Tamil-Mistral-7B-v0.1

🧩 Configuration

base_model: Hemanth-thunder/Tamil-Mistral-7B-v0.1
experts:
  - source_model: Hemanth-thunder/Tamil-Mistral-7B-v0.1
    positive_prompts:
    - "பேச்சு"  # "chat"
    - "உதவி"  # "assistant"
    - "எனக்கு சொல்"  # "tell me"
    - "விளக்கம்"  # "explain"
    - "நான் விரும்புகிறேன்"  # "I want"
  - source_model: Hemanth-thunder/Tamil-Mistral-7B-v0.1
    positive_prompts:
    - "அறிவுரை"  # "advice"
    - "நிர்வாகம்"  # "management"
    - "உத்தரவாதம்"  # "instructions"
    - "பயிற்சி"  # "training"
    - "செயல்முறை"  # "procedure"

💻 Usage

!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "praveengovi/SG-Tamil-MoE-7B"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])