Views
No views yet
| Directory | Target chip | Image encoder |
|---|---|---|
InternVL3_5-1B_GPTQ_INT4-AX650/ | AX650N (host & AXCL M.2 accelerator) | internvl_vit_model_1x448x448x3.axmodel (NHWC) |
InternVL3_5-1B_GPTQ_INT4-AX620E/ | AX620E | internvl_vit_model_1x448x448x3_ax620e.axmodel |
InternVL3_5-1B_GPTQ_INT4-AX650/
├── config.json # ax-llm runtime config (vlm_type=InternVL3, tokenizer_type=InternVL3_5)
├── internvl3-5-1b_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| Chip | image encoder 448 | ttft | w4a16 |
|---|---|---|---|
| AX650 | 364.412 ms | 883.458 ms | 28.09 tokens/sec |
| AX620E | 2358.956 ms | 3136.54 ms | 7.33 tokens/sec |
-AX650 directory:1# interactive chat
2./axllm run InternVL3_5-1B_GPTQ_INT4-AX650
3
4# OpenAI-compatible HTTP server (default port 8000)
5./axllm serve InternVL3_5-1B_GPTQ_INT4-AX650AXLLM_DEVICES=<id> or the "devices" field in config.json.-AX620E directory:./axllm run InternVL3_5-1B_GPTQ_INT4-AX620E