--- base_model_relation: quantized base_model: - KORMo-VL/KORMo-Web-4B license: apache-2.0 language: - en - ko pipeline_tag: image-text-to-text library_name: transformers tags: - mtp - text-generation-inference - llama-cpp - gui-agent - web-agent - computer-use - browser - korean - no-dom --- # **KORMo-Web-4B-MTP-Repacked** > **KORMo-Web-4B** is an open-weight 4B bilingual (Korean and English) vision-language web agent developed by KAIST researchers on top of **Qwen3.5-4B** that executes complex, multi-site, and multi-tab browser tasks purely from screenshots—requiring no DOM or accessibility tree. Pretrained on over 10 million atomic-level web element understanding samples and fine-tuned on more than 100,000 real browser interaction trajectories, it uses a step-by-step `` reasoning loop across 22 normalized-coordinate browser actions (`click`, `type`, `scroll`, tab control, etc.) to achieve a **42.3% success rate** on both English and Korean **Online-Mind2Web** benchmarks (+22.3 and +28.8 points over its base model, outperforming comparable open-weight agents). Released under an **Apache 2.0** license alongside a Playwright reference harness and a Chrome extension, it runs on a single NVIDIA GPU via `vLLM` or locally on Apple Silicon Macs via `vllm-mlx` (with a 4-bit variant fitting in 16 GB RAM). > [!NOTE] > MTP — Speculative Decoding: Yes | GGUF — [KORMo-Web-4B-MTP-Repacked(GGUF)](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/tree/main/GGUF) > [!NOTE] Multi-Token Prediction (MTP) GGUF is a specialized GGUF model file format extension that integrates speculative decoding directly into the model weights to significantly accelerate local inference. Unlike traditional speculative decoding which requires a separate, smaller "draft" model, MTP GGUF files include additional output heads within the main model architecture that predict multiple future tokens in a single forward pass. ## Model Files [GGUF] | File Name | Quant Type | File Size | File Link | Description | |-----------|------------|-----------|-----------|-------------| | KORMo-Web-4B-MTP.BF16.gguf | BF16 | 9.94 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.BF16.gguf) | Full BF16 weights. Highest quality, largest file size. | | KORMo-Web-4B-MTP.Q3_K_L.gguf | Q3_K_L | 2.82 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q3_K_L.gguf) | Lower quality but usable, good for low RAM availability. | | KORMo-Web-4B-MTP.Q3_K_M.gguf | Q3_K_M | 2.63 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q3_K_M.gguf) | Low quality, smallest file size. | | KORMo-Web-4B-MTP.Q4_K_M.gguf | Q4_K_M | 3.17 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q4_K_M.gguf) | Good quality, default size for most use cases, *recommended*. | | KORMo-Web-4B-MTP.Q4_K_S.gguf | Q4_K_S | 2.99 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q4_K_S.gguf) | Slightly lower quality with more space savings, *recommended*. | | KORMo-Web-4B-MTP.Q5_K_M.gguf | Q5_K_M | 3.62 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q5_K_M.gguf) | High quality, *recommended*. | | KORMo-Web-4B-MTP.Q5_K_S.gguf | Q5_K_S | 3.51 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q5_K_S.gguf) | High quality with slight space savings, *recommended*. | | KORMo-Web-4B-MTP.Q6_K.gguf | Q6_K | 4.08 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q6_K.gguf) | Very high quality, near perfect, *recommended*. | | KORMo-Web-4B-MTP.Q8_0.gguf | Q8_0 | 5.29 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.Q8_0.gguf) | Extremely high quality, generally unneeded but max available standard quant. | | KORMo-Web-4B-MTP.mmproj-bf16.gguf | mmproj-bf16 | 676 MB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/GGUF/KORMo-Web-4B-MTP.mmproj-bf16.gguf) | Multimodal projection file in BF16 format. Used for vision/image capabilities. | ## MTP-Repacked [Transformers, BF16] | File Name | File Type | File Size | File Link | Description | |-----------|-----------|-----------|-----------|-------------| | chat_template.jinja | Jinja Template | 7.76 kB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/chat_template.jinja) | Chat template for formatting conversations and prompts. | | config.json | JSON | 2.85 kB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/config.json) | Core model architecture and configuration parameters. | | generation_config.json | JSON | 219 Bytes | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/generation_config.json) | Default text generation parameters (e.g., temperature, top_p). | | model-00001-of-00002.safetensors | Safetensors | 5.49 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/model-00001-of-00002.safetensors) | Model weight shard 1/2. **Required** to load the model. | | model-00002-of-00002.safetensors | Safetensors | 5.1 GB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/model-00002-of-00002.safetensors) | Model weight shard 2/2. **Required** to load the model. | | model.safetensors.index.json | JSON | 67.4 kB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/model.safetensors.index.json) | Index file mapping model weights to their respective shards. | | processor_config.json | JSON | 1.19 kB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/processor_config.json) | Configuration for the multimodal processor. | | system_prompt.md | Markdown | 3.28 kB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/system_prompt.md) | Default system prompt or behavioral guidelines for the model. | | tokenizer.json | JSON | 20 MB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/tokenizer.json) | Complete tokenizer vocabulary and encoding rules. | | tokenizer_config.json | JSON | 1.16 kB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/tokenizer_config.json) | Tokenizer-specific configuration parameters. | | tools.json | JSON | 12.4 kB | [Link](https://huggingface.co/prithivMLmods/KORMo-Web-4B-MTP-Repacked/blob/main/tools.json) | Configuration or definitions for available tools/functions the model can use. | ## llama.cpp LLM inference in C/C++ — https://github.com/ggml-org/llama.cpp MTP-repackaged and quantized using llama.cpp by — [prithivMLmods](https://huggingface.co/prithivMLmods).