{ "hf_id": "baidu/ERNIE-4.5-VL-28B-A3B-PT", "meta": { "title": "ERNIE-4.5-VL", "slug": "ernie-4.5-vl", "provider": "Ernie (Baidu)", "description": "Baidu ERNIE 4.5 VL MoE vision-language models (28B-A3B, 424B-A47B) with heterogeneous text/vision experts", "date_added": "2025-09-03", "date_updated": "2026-04-17", "tasks": [ "multimodal" ], "related_recipes": [ "baidu/ERNIE-4.5-21B-A3B-PT" ], "hardware": { "mi300x": "verified", "mi325x": "verified", "mi355x": "verified" } }, "recommended_command": { "hardware": "h200", "strategy": "single_node_tp", "variant": "default", "node_count": 1, "deploy_type": "single_node", "env": {}, "docker_image": "vllm/vllm-openai:latest", "command": "vllm serve baidu/ERNIE-4.5-VL-28B-A3B-PT \\\n --trust-remote-code \\\n --tensor-parallel-size 1 \\\n --mm-encoder-tp-mode data", "argv": [ "vllm", "serve", "baidu/ERNIE-4.5-VL-28B-A3B-PT", "--trust-remote-code", "--tensor-parallel-size", "1", "--mm-encoder-tp-mode", "data" ], "docker_command": "docker run --gpus all \\\n --privileged --ipc=host -p 8000:8000 \\\n -v ~/.cache/huggingface:/root/.cache/huggingface \\\n vllm/vllm-openai:latest baidu/ERNIE-4.5-VL-28B-A3B-PT \\\n --trust-remote-code \\\n --tensor-parallel-size 1 \\\n --mm-encoder-tp-mode data", "docker_argv": [ "docker", "run", "--gpus", "all", "--privileged", "--ipc=host", "-p", "8000:8000", "-v", "~/.cache/huggingface:/root/.cache/huggingface", "vllm/vllm-openai:latest", "baidu/ERNIE-4.5-VL-28B-A3B-PT", "--trust-remote-code", "--tensor-parallel-size", "1", "--mm-encoder-tp-mode", "data" ], "strategy_spec": { "name": "single_node_tp", "deploy_type": "single_node", "display_name": "Tensor Parallel", "short_label": "TP", "orientation": "latency", "description": "Single-node tensor parallel. Splits the model across all local GPUs. TP size is set to the GPU count at deploy time. The simplest multi-GPU strategy — works for all model architectures.\n", "hardware_match": { "min_gpus": 1, "max_gpus": 8, "multi_node": false }, "vllm_args": [], "parallel_flag": "--tensor-parallel-size" }, "hardware_profile": { "brand": "NVIDIA", "generation": "hopper", "compute_arch": "sm90", "display_name": "H200", "description": "NVIDIA H200 SXM 141 GB HBM3e · 8-GPU HGX node", "gpu_count": 8, "vram_gb": 1128, "multi_node": false }, "alternatives": { "multi_node_tp": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/strategies/multi_node_tp.json", "multi_node_tep": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/strategies/multi_node_tep.json", "multi_node_dep": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/strategies/multi_node_dep.json", "kv_store_centralized_mooncake": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/strategies/kv_store_centralized_mooncake.json", "kv_store_distributed_mooncake": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/strategies/kv_store_distributed_mooncake.json" }, "by_hardware": { "h200": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/h200.json", "h100": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/h100.json", "b200": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/b200.json", "gb200": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/gb200.json", "b300": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/b300.json", "gb300": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/gb300.json", "mi300x": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/mi300x.json", "mi325x": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/mi325x.json", "mi350x": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/mi350x.json", "mi355x": "/baidu/ERNIE-4.5-VL-28B-A3B-PT/hw/mi355x.json" } }, "model": { "model_id": "baidu/ERNIE-4.5-VL-28B-A3B-PT", "min_vllm_version": "0.11.0", "architecture": "moe", "parameter_count": "28B", "active_parameters": "3B", "context_length": 131072, "base_args": [ "--trust-remote-code" ], "base_env": {}, "install": { "pip": { "command": "uv venv\nsource .venv/bin/activate\nuv pip install -U vllm --torch-backend auto" }, "docker": { "command": "docker pull vllm/vllm-openai:latest" } } }, "features": { "text_only": { "description": "Skip loading the vision encoder for text-only workloads — frees VRAM for KV cache. Mutually exclusive with encoder_parallel.", "args": [ "--language-model-only" ] }, "encoder_parallel": { "description": "Run the vision encoder in data-parallel mode — avoids TP comm overhead on the small encoder. Mutually exclusive with text_only.", "args": [ "--mm-encoder-tp-mode", "data" ] } }, "opt_in_features": [ "text_only" ], "variants": { "default": { "precision": "bf16", "vram_minimum_gb": 67, "description": "BF16 weights; fits on 1x80GB GPU (28B VL variant)" }, "424b": { "model_id": "baidu/ERNIE-4.5-VL-424B-A47B-PT", "precision": "bf16", "vram_minimum_gb": 1120, "description": "424B total / 47B active; 8x140GB BF16 or 16x80GB BF16", "extra_args": [ "--trust-remote-code", "--tensor-parallel-size", "8" ], "json": "/baidu/ERNIE-4.5-VL-424B-A47B-PT.json" }, "424b_fp8": { "model_id": "baidu/ERNIE-4.5-VL-424B-A47B-PT", "precision": "fp8", "vram_minimum_gb": 640, "description": "424B with FP8 online quantization + CPU offload for 8x80GB testing", "extra_args": [ "--trust-remote-code", "--tensor-parallel-size", "8", "--quantization", "fp8", "--cpu-offload-gb", "50" ], "json": "/baidu/ERNIE-4.5-VL-424B-A47B-PT.json" } }, "kv_offload_support": { "offloading_cpu": "verified", "offloading_fs": "verified" }, "hardware_overrides": {}, "guide": "## Overview\n\nERNIE 4.5 VL is Baidu's multimodal MoE model with heterogeneous experts (separate text\nand vision experts). Because of the heterogeneous architecture, **torch.compile and CUDA\ngraphs are not supported**.\n\n- `baidu/ERNIE-4.5-VL-28B-A3B-PT` — 28B total / 3B active (1x80GB)\n- `baidu/ERNIE-4.5-VL-424B-A47B-PT` — 424B total / 47B active (8x140GB BF16, 8x80GB FP8+offload, or 16x80GB BF16)\n\n## Prerequisites\n\n- vLLM: support added to main branch recently; install latest\n- Hardware depends on variant\n\n### Install vLLM (CUDA)\n\n```bash\nuv venv --python 3.12 --seed\nsource .venv/bin/activate\nuv pip install -U vllm --torch-backend auto\n```\n\n### Install vLLM (AMD ROCm MI300X/MI325X/MI355X)\n\n```bash\nuv pip install vllm --extra-index-url https://wheels.vllm.ai/rocm/\n```\n\n## Launch commands\n\n28B on 1x80GB:\n\n```bash\nvllm serve baidu/ERNIE-4.5-VL-28B-A3B-PT --trust-remote-code\n```\n\n424B BF16 on 8x140GB:\n\n```bash\nvllm serve baidu/ERNIE-4.5-VL-424B-A47B-PT \\\n --trust-remote-code \\\n --tensor-parallel-size 8\n```\n\n424B with FP8 + CPU offload on 8x80GB (testing only):\n\n```bash\nvllm serve baidu/ERNIE-4.5-VL-424B-A47B-PT \\\n --trust-remote-code \\\n --tensor-parallel-size 8 \\\n --quantization fp8 \\\n --cpu-offload-gb 50\n```\n\n28B on AMD MI300X+:\n\n```bash\nVLLM_ROCM_USE_AITER=1 SAFETENSORS_FAST_GPU=1 \\\n vllm serve baidu/ERNIE-4.5-VL-28B-A3B-PT \\\n --tensor-parallel-size 4 \\\n --gpu-memory-utilization 0.9 \\\n --trust-remote-code\n```\n\n## Benchmarking\n\n```bash\nvllm bench serve \\\n --model baidu/ERNIE-4.5-VL-28B-A3B-PT \\\n --dataset-name random \\\n --random-input-len 8000 --random-output-len 1000 \\\n --request-rate 10 --num-prompts 16 --ignore-eos --trust-remote-code\n```\n\n## References\n\n- [ERNIE-4.5-VL-28B-A3B-PT](https://huggingface.co/baidu/ERNIE-4.5-VL-28B-A3B-PT)\n- [ERNIE-4.5-VL-424B-A47B-PT](https://huggingface.co/baidu/ERNIE-4.5-VL-424B-A47B-PT)\n", "source": { "source": "vllm", "repo": "vllm-project/vllm", "ref": "main", "sha": "389ffebe555705c85337025d5df108d0384d365e", "path": "recipes/models/baidu/ERNIE-4.5-VL-28B-A3B-PT.yaml", "edit_url": "https://github.com/vllm-project/vllm/edit/main/recipes/models/baidu/ERNIE-4.5-VL-28B-A3B-PT.yaml", "overlays": [] } }