Instructions to use AXERA-TECH/InternVL3_5-2B_GPTQ_INT4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AXERA-TECH/InternVL3_5-2B_GPTQ_INT4 with Transformers:

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

pipe = pipeline("image-text-to-text", model="AXERA-TECH/InternVL3_5-2B_GPTQ_INT4")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("AXERA-TECH/InternVL3_5-2B_GPTQ_INT4", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use AXERA-TECH/InternVL3_5-2B_GPTQ_INT4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AXERA-TECH/InternVL3_5-2B_GPTQ_INT4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AXERA-TECH/InternVL3_5-2B_GPTQ_INT4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/AXERA-TECH/InternVL3_5-2B_GPTQ_INT4

SGLang

How to use AXERA-TECH/InternVL3_5-2B_GPTQ_INT4 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 "AXERA-TECH/InternVL3_5-2B_GPTQ_INT4" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AXERA-TECH/InternVL3_5-2B_GPTQ_INT4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "AXERA-TECH/InternVL3_5-2B_GPTQ_INT4" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AXERA-TECH/InternVL3_5-2B_GPTQ_INT4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use AXERA-TECH/InternVL3_5-2B_GPTQ_INT4 with Docker Model Runner:
```
docker model run hf.co/AXERA-TECH/InternVL3_5-2B_GPTQ_INT4
```

InternVL3_5-2B_GPTQ_INT4

InternVL3.5-2B (InternViT + Qwen3-1.7B) converted to run on the Axera NPU with w4a16 quantization, packaged in ax-llm model-directory format.

Compatible with Pulsar2 5.1-patch1
Context length: 2k, max prefill: 1k, image input: 448×448
Target chip: AX650 (AX650N host & AXCL M.2 accelerator)

This repository is a ready-to-run ax-llm model directory:

.
├── config.json                    # ax-llm runtime config (vlm_type=InternVL3, tokenizer_type=InternVL3_5)
├── internvl3-5-2b_tokenizer.txt   # tokenizer
├── post_config.json               # sampling config
├── internvl3-5_axmodel/           # LLM: qwen3_p128_l0..l27 (28) + qwen3_post + embed
└── vit-models/                    # InternViT image encoder (internvl_vit_model_1x3x448x448.axmodel)

Support Platform

AX650
- AX650N DEMO Board
- M4N-Dock (爱芯派Pro)
- M.2 Accelerator card (AXCL)

Chip	image encoder 448	ttft	w4a16
AX650	364.412 ms	4951.50 ms	28.07 tokens/sec

How to use

Build/obtain the ax-llm runtime: AXera NPU LLM Runtime (ax-llm)

# interactive chat
./axllm run   InternVL3_5-2B_GPTQ_INT4

# OpenAI-compatible HTTP server (default port 8000)
./axllm serve InternVL3_5-2B_GPTQ_INT4

On AXCL, choose the card via AXLLM_DEVICES=<id> or the "devices" field in config.json.

Source & conversion

Base model: OpenGVLab/InternVL3_5-2B
License: BSD-3-Clause

Downloads last month: 34

Inference Providers NEW

Image-Text-to-Text

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for AXERA-TECH/InternVL3_5-2B_GPTQ_INT4

Base model

OpenGVLab/InternVL3_5-2B-Pretrained

Finetuned

OpenGVLab/InternVL3_5-2B-Instruct

Finetuned

OpenGVLab/InternVL3_5-2B-MPO

Finetuned

OpenGVLab/InternVL3_5-2B

Finetuned

(5)

this model