Instructions to use blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32 with Transformers:

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

pipe = pipeline("text-generation", model="blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32")
model = AutoModelForCausalLM.from_pretrained("blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32", 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32

SGLang

How to use blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32 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 "blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32" \
    --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": "blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32",
		"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 "blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32" \
        --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": "blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32 with Docker Model Runner:
```
docker model run hf.co/blake-lucas/Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32
```

Mellum2.1-12B-A2.5B-Thinking-AWQ-W4A16-G32 / calibration-run-receipt.json

blake-lucas

Publish experimental Mellum2.1 AWQ W4A16 group-32 derivative

b43fd3a verified 2 days ago

Raw History Blame Contribute Delete

3.13 kB

	{
	"schema_version": 1,
	"model_id": "JetBrains/Mellum2.1-12B-A2.5B-Thinking",
	"source_revision": "92ddae9fc7665e9f801d141d2e5a6b2caf2460c4",
	"quantizer_image_id": "sha256:e4f0dbe60e59050796a15b7b0c4646130a93acfd4e2431cde05923e401470b10",
	"source_sha256": {
	"quantize.py": "890aef9cf2ceb5c50d663a0014a51326ff960b2098fc30f10efdf2d130d0fc33",
	"audit_export.py": "9c2476d66193a505f43f8982a07c053fdabfe01205af051f6d6f7c37085ac199",
	"check_quantizer.py": "c8bbb692f1e82e4193df78710df3f7b6d617793d66c8dfa8f0ab4836d05572c2",
	"recipe.yaml": "a73469b793eb4af8eacaa3241cecb6b699d3ccc88374e575458cf35500f6ab3a",
	"requirements.quantize.txt": "3dddfbf7bf5a7cdd012a56fd80f47ae895db1468f654b196f7ca10dab2f75617",
	"Dockerfile.quantize": "a6f281f8c21f025bb6e825e4efed7ce982af03e573e5a9d1214918bc6a44ec13"
	},
	"source_sha256_basis": "Runtime files verified against installed frozen image; Dockerfile from preserved original build context",
	"run": {
	"started_at": "2026-10-08T22:00:48.764973527Z",
	"finished_at": "2026-10-09T01:28:31.258642816Z",
	"calibration_started_at": "2026-10-08T22:01:37.723800+00:00",
	"calibration_finished_at": "2026-10-09T01:11:45.235800+00:00",
	"exit_code": 0,
	"oom_killed": false,
	"calibration_started_at_basis": "Compression lifecycle initialization log; calibration follows initialization",
	"calibration_finished_at_basis": "Final propagation complete and compression lifecycle finalized log"
	},
	"resources": {
	"initial": {
	"ram_bytes": 51539607552,
	"ram_swap_bytes": 60129542144
	},
	"final": {
	"ram_bytes": 55834574848,
	"ram_swap_bytes": 77309411328
	},
	"overrides": [
	{
	"at": "2026-10-08T22:25:34.159628+00:00",
	"before": {
	"ram_bytes": 51539607552,
	"ram_swap_bytes": 60129542144
	},
	"after": {
	"ram_bytes": 55834574848,
	"ram_swap_bytes": 64424509440
	},
	"at_basis": "Receipt file mtime; original receipt omitted explicit timestamp"
	},
	{
	"at": "2026-10-08T23:01:29.787460+00:00",
	"before": {
	"ram_bytes": 55834574848,
	"ram_swap_bytes": 64424509440
	},
	"after": {
	"ram_bytes": 55834574848,
	"ram_swap_bytes": 68719476736
	}
	},
	{
	"at": "2026-10-08T23:56:24.515470+00:00",
	"before": {
	"ram_bytes": 55834574848,
	"ram_swap_bytes": 68719476736
	},
	"after": {
	"ram_bytes": 55834574848,
	"ram_swap_bytes": 77309411328
	}
	}
	],
	"temporary_host_swap_added_bytes": 17179869184,
	"persistent_os_swap_changes": false
	},
	"telemetry": {
	"sampling_interval_seconds": 5,
	"device_peak_mib": 8219,
	"resident_baseline_mib": 1198,
	"limitations": "Sampled total device memory includes resident retrieval services; not a per-process CUDA allocator peak. Background workload and host-memory pressure may affect timings.",
	"samples": 2453
	},
	"quantization": {
	"effective_observer": "memoryless_mse"
	}
	}