Instructions to use fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym with Transformers:

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

pipe = pipeline("text-generation", model="fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym")
model = AutoModelForCausalLM.from_pretrained("fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym", device_map="auto")
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 Settings

vLLM

How to use fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym

SGLang

How to use fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym 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 "fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym" \
    --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": "fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym",
		"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 "fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym" \
        --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": "fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym with Docker Model Runner:
```
docker model run hf.co/fbaldassarri/ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym
```

ibm-granite_granite-3.3-2b-instruct-autoawq-int4-gs128-sym

File size: 773 Bytes

c7aca3c

{
  "bits": 4,
  "group_size": 128,
  "sym": true,
  "data_type": "int",
  "enable_quanted_input": true,
  "enable_minmax_tuning": true,
  "seqlen": 512,
  "batch_size": 4,
  "scale_dtype": "torch.float16",
  "lr": 0.005,
  "minmax_lr": 0.005,
  "gradient_accumulate_steps": 1,
  "iters": 200,
  "amp": false,
  "nsamples": 128,
  "low_gpu_mem_usage": false,
  "to_quant_block_names": null,
  "enable_norm_bias_tuning": false,
  "act_bits": 16,
  "act_group_size": 128,
  "act_sym": true,
  "act_dynamic": true,
  "act_data_type": "int",
  "super_bits": null,
  "super_group_size": null,
  "dataset": "NeelNanda/pile-10k",
  "autoround_version": "0.4.7",
  "quant_method": "awq",
  "zero_point": false,
  "version": "gemm",
  "modules_to_not_convert": [
    "lm_head"
  ]
}