Instructions to use Shadow0131/hindi-slm-c2-step120 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Shadow0131/hindi-slm-c2-step120 with Transformers:

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

pipe = pipeline("text-generation", model="Shadow0131/hindi-slm-c2-step120", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Shadow0131/hindi-slm-c2-step120", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Shadow0131/hindi-slm-c2-step120 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Shadow0131/hindi-slm-c2-step120"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Shadow0131/hindi-slm-c2-step120",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Shadow0131/hindi-slm-c2-step120

SGLang

How to use Shadow0131/hindi-slm-c2-step120 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 "Shadow0131/hindi-slm-c2-step120" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Shadow0131/hindi-slm-c2-step120",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "Shadow0131/hindi-slm-c2-step120" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Shadow0131/hindi-slm-c2-step120",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Shadow0131/hindi-slm-c2-step120 with Docker Model Runner:
```
docker model run hf.co/Shadow0131/hindi-slm-c2-step120
```

Hindi SLM — C2 Step 120

Final C2 / step-120 checkpoint produced for the IAIRO SLM++ Bootcamp capstone.

Load

from transformers import AutoTokenizer, AutoModelForCausalLM

repo = "Shadow0131/hindi-slm-c2-step120"

tokenizer = AutoTokenizer.from_pretrained(
    repo,
    trust_remote_code=True
)

model = AutoModelForCausalLM.from_pretrained(
    repo,
    trust_remote_code=True
)

This checkpoint uses a custom Transformers architecture, so trust_remote_code=True is required.

Checkpoint information

Final checkpoint: C2 / step 120
Vocabulary size: 32,768
Context length: 256
Format: Hugging Face Transformers
Weights: safetensors
pad token: EOS fallback
Domain: Hindi language modeling

Downloads last month: 164

Safetensors

Model size

0.9B params

Tensor type

F32

BF16