Instructions to use medimed/MNLP_M2_quantized_model with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use medimed/MNLP_M2_quantized_model with Transformers:

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

pipe = pipeline("text-generation", model="medimed/MNLP_M2_quantized_model")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("medimed/MNLP_M2_quantized_model")
model = AutoModelForCausalLM.from_pretrained("medimed/MNLP_M2_quantized_model")
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 medimed/MNLP_M2_quantized_model with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "medimed/MNLP_M2_quantized_model"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "medimed/MNLP_M2_quantized_model",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/medimed/MNLP_M2_quantized_model

SGLang

How to use medimed/MNLP_M2_quantized_model 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 "medimed/MNLP_M2_quantized_model" \
    --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": "medimed/MNLP_M2_quantized_model",
		"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 "medimed/MNLP_M2_quantized_model" \
        --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": "medimed/MNLP_M2_quantized_model",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use medimed/MNLP_M2_quantized_model with Docker Model Runner:
```
docker model run hf.co/medimed/MNLP_M2_quantized_model
```

medimed commited on May 21, 2025

Commit

83dc3b8

verified ·

1 Parent(s): aa340db

Upload Qwen3ForCausalLM

Browse files

Files changed (3) hide show

config.json +43 -1
model.safetensors +2 -2
recipe.yaml +7 -0

config.json CHANGED Viewed

@@ -18,6 +18,48 @@
   "num_hidden_layers": 28,
   "num_key_value_heads": 8,
   "pad_token_id": 151643,
   "rms_norm_eps": 1e-06,
   "rope_scaling": null,
   "rope_theta": 1000000,
@@ -28,4 +70,4 @@
   "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936
-}

   "num_hidden_layers": 28,
   "num_key_value_heads": 8,
   "pad_token_id": 151643,
+  "quantization_config": {
+    "config_groups": {
+      "group_0": {
+        "input_activations": {
+          "actorder": null,
+          "block_structure": null,
+          "dynamic": true,
+          "group_size": null,
+          "num_bits": 8,
+          "observer": null,
+          "observer_kwargs": {},
+          "strategy": "token",
+          "symmetric": true,
+          "type": "int"
+        },
+        "output_activations": null,
+        "targets": [
+          "Linear"
+        ],
+        "weights": {
+          "actorder": null,
+          "block_structure": null,
+          "dynamic": false,
+          "group_size": null,
+          "num_bits": 8,
+          "observer": "minmax",
+          "observer_kwargs": {},
+          "strategy": "channel",
+          "symmetric": true,
+          "type": "int"
+        }
+      }
+    },
+    "format": "int-quantized",
+    "global_compression_ratio": null,
+    "ignore": [
+      "lm_head"
+    ],
+    "kv_cache_scheme": null,
+    "quant_method": "compressed-tensors",
+    "quantization_status": "compressed"
+  },
   "rms_norm_eps": 1e-06,
   "rope_scaling": null,
   "rope_theta": 1000000,
   "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936
+}

model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:baa3239a4c22b64f9442b9409e1ad5984719b1a4cfb83dacb7ca54d34ea8fe1d
-size 311308760

 version https://git-lfs.github.com/spec/v1
+oid sha256:7340bf7b29fa08a02b1f7fbb9c30dec3cd9c3423c692d83276e410848a6664c9
+size 752443720

recipe.yaml ADDED Viewed

	@@ -0,0 +1,7 @@

+default_stage:
+  default_modifiers:
+    SmoothQuantModifier: {smoothing_strength: 0.8}
+    GPTQModifier:
+      scheme: W8A8
+      targets: Linear
+      ignore: [lm_head]