Instructions to use tripathysagar/Qwen2.5-Coder-196M-Shell with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tripathysagar/Qwen2.5-Coder-196M-Shell with Transformers:

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

pipe = pipeline("text-generation", model="tripathysagar/Qwen2.5-Coder-196M-Shell")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("tripathysagar/Qwen2.5-Coder-196M-Shell")
model = AutoModelForCausalLM.from_pretrained("tripathysagar/Qwen2.5-Coder-196M-Shell")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use tripathysagar/Qwen2.5-Coder-196M-Shell with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tripathysagar/Qwen2.5-Coder-196M-Shell"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tripathysagar/Qwen2.5-Coder-196M-Shell",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tripathysagar/Qwen2.5-Coder-196M-Shell

SGLang

How to use tripathysagar/Qwen2.5-Coder-196M-Shell 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 "tripathysagar/Qwen2.5-Coder-196M-Shell" \
    --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": "tripathysagar/Qwen2.5-Coder-196M-Shell",
		"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 "tripathysagar/Qwen2.5-Coder-196M-Shell" \
        --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": "tripathysagar/Qwen2.5-Coder-196M-Shell",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tripathysagar/Qwen2.5-Coder-196M-Shell with Docker Model Runner:
```
docker model run hf.co/tripathysagar/Qwen2.5-Coder-196M-Shell
```

Model Card for Model ID

Distiled Qwen/Qwen2.5-Coder-0.5B-Instruct by westenfelder/NL2SH-ALFA for NLP to bash command. Distiled only decoder block from 24 to 4 with the original tokenizer.

Model Details

Model Sources [optional]

Blog: [More Information Needed]

Uses

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_name = "tripathysagar/Qwen2.5-Coder-196M-Shell"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    dtype=torch.bfloat16,
    device_map="auto",
    )

def infer(inp, debug=False):
  msg = [
    {"role": "system", "content": "Generate shell command."},
    {"role": "user", "content": inp},
  ]
  text = tokenizer.apply_chat_template(
    msg,
    tokenize=False,
    add_generation_prompt=True,
  )
  if debug:
    print(text)

  model_inputs = tokenizer([text], return_tensors="pt")
  generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=256,
    do_sample=True,

  )

  resp_text = tokenizer.batch_decode(generated_ids)[0]
  if debug:
    print(resp_text)

  return (inp, resp_text[len(text):].replace('<|im_end|>', ''))

infer("get kernel name.")

Downloads last month: 11

Safetensors

Model size

0.2B params

Tensor type

BF16