README.md

metadata

language:
  - bn
  - en
  - gu
  - hi
  - kn
  - ml
  - mr
  - or
  - pa
  - ta
  - te
library_name: transformers
base_model: sarvamai/sarvam-1
tags:
  - mlx

salmankhanpm/sarvam-1-mlx-4Bit

The Model salmankhanpm/sarvam-1-mlx-4Bit was converted to MLX format from sarvamai/sarvam-1 using mlx-lm version 0.22.3.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("salmankhanpm/sarvam-1-mlx-4Bit")

prompt="hello"

if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)