Instructions to use mobilint/Qwen3-8B-Batch16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mobilint/Qwen3-8B-Batch16 with Transformers:

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

pipe = pipeline("text-generation", model="mobilint/Qwen3-8B-Batch16", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("mobilint/Qwen3-8B-Batch16", trust_remote_code=True, device_map="auto")

Mobilint

How to use mobilint/Qwen3-8B-Batch16 with Mobilint:

# pip install mblt-model-zoo
from mblt_model_zoo.vision import MBLT_Engine

model = MBLT_Engine(
    model_cls="Qwen3-8B-Batch16",
    model_type="DEFAULT",
    model_path="",
    core_mode="global8",
)

try:
    image = model.preprocess("path/to/image.jpg")
    output = model(image)
    result = model.postprocess(output)
finally:
    model.dispose()

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use mobilint/Qwen3-8B-Batch16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mobilint/Qwen3-8B-Batch16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mobilint/Qwen3-8B-Batch16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mobilint/Qwen3-8B-Batch16

SGLang

How to use mobilint/Qwen3-8B-Batch16 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 "mobilint/Qwen3-8B-Batch16" \
    --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": "mobilint/Qwen3-8B-Batch16",
		"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 "mobilint/Qwen3-8B-Batch16" \
        --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": "mobilint/Qwen3-8B-Batch16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mobilint/Qwen3-8B-Batch16 with Docker Model Runner:
```
docker model run hf.co/mobilint/Qwen3-8B-Batch16
```

lbs1163 commited on May 19

Commit

29da5e8

verified ·

1 Parent(s): 5de279b

chore: set max_batch_size to 16

Browse files

Set max_batch_size according to repository name convention: Batch16 -> 16, Batch32 -> 32, otherwise 1.

Files changed (1) hide show

config.json +46 -45

config.json CHANGED Viewed

@@ -1,45 +1,46 @@
-{
-  "architectures": [
-    "MobilintQwen3ForCausalLM"
-  ],
-  "attention_bias": false,
-  "attention_dropout": 0.0,
-  "auto_map": {
-    "AutoConfig": "proxy_qwen3.MobilintQwen3Config",
-    "AutoModelForCausalLM": "proxy_qwen3.MobilintQwen3ForCausalLM"
-  },
-  "bos_token_id": 151643,
-  "core_mode": "single",
-  "dev_no": 0,
-  "eos_token_id": 151645,
-  "head_dim": 128,
-  "hidden_act": "silu",
-  "hidden_size": 4096,
-  "initializer_range": 0.02,
-  "intermediate_size": 12288,
-  "max_position_embeddings": 4096,
-  "max_window_layers": 36,
-  "mxq_path": "Qwen3-8B-Batch16-W4V8.mxq",
-  "model_type": "mobilint-qwen3",
-  "num_attention_heads": 32,
-  "num_hidden_layers": 36,
-  "num_key_value_heads": 8,
-  "rms_norm_eps": 1e-06,
-  "rope_scaling": null,
-  "rope_theta": 1000000,
-  "sliding_window": null,
-  "target_cores": [
-    "0:0"
-  ],
-  "tie_word_embeddings": false,
-  "torch_dtype": "bfloat16",
-  "transformers_version": "4.51.0",
-  "use_cache": true,
-  "use_sliding_window": false,
-  "vocab_size": 151936,
-  "npu_prefill_chunk_size": {
-    "global4": 256,
-    "global8": 512,
-    "single": 128
-  }
-}

+{
+  "architectures": [
+    "MobilintQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "auto_map": {
+    "AutoConfig": "proxy_qwen3.MobilintQwen3Config",
+    "AutoModelForCausalLM": "proxy_qwen3.MobilintQwen3ForCausalLM"
+  },
+  "bos_token_id": 151643,
+  "core_mode": "single",
+  "dev_no": 0,
+  "eos_token_id": 151645,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 12288,
+  "max_position_embeddings": 4096,
+  "max_window_layers": 36,
+  "mxq_path": "Qwen3-8B-Batch16-W4V8.mxq",
+  "model_type": "mobilint-qwen3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "target_cores": [
+    "0:0"
+  ],
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.0",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936,
+  "npu_prefill_chunk_size": {
+    "global4": 256,
+    "global8": 512,
+    "single": 128
+  },
+  "max_batch_size": 16
+}