|
Download README.md from ggml-org-dev/tiny-Qwen3.5-4B-for-testing: direct link, hf CLI and curl.
- Browser
- Download file 1.48 kB
-
https://huggingface.co/ggml-org-dev/tiny-Qwen3.5-4B-for-testing/resolve/main/README.md
- Command line
-
hf download hf://ggml-org-dev/tiny-Qwen3.5-4B-for-testing/README.md
-
curl -L -o README.md https://huggingface.co/ggml-org-dev/tiny-Qwen3.5-4B-for-testing/resolve/main/README.md
1.48 kB
| license: apache-2.0 | |
| base_model: Qwen/Qwen3.5-4B | |
| tags: | |
| - testing | |
| # tiny-Qwen3.5-4B-for-testing | |
| > [!WARNING] | |
| > This is a **modified, damaged copy** of [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B) (revision `851bf6e`), made only to test software. Its output is gibberish. Do not use it for anything else. | |
| It keeps the architecture (hybrid linear attention / gated full attention, MTP layer, vision encoder), tokenizer and configs of Qwen3.5-4B, but it is small enough for CI. It is the base model of the tiny LoRA decision adapter [ggml-org-dev/tiny-lev-for-testing](https://huggingface.co/ggml-org-dev/tiny-lev-for-testing): the llama.cpp converter downloads it and merges the adapter into it. | |
| ## Changes from the original model | |
| Sliced from the safetensors with [slice_qwen35_lora.py](slice_qwen35_lora.py), which follows the `slice-model` skill of llama.cpp: | |
| - Text model: 4 of 32 layers (0, 1, 2, 3: 3 linear attention + 1 full attention), hidden size 2560 -> 32, FFN 9216 -> 128, heads 16 -> 2, KV heads 4 -> 1 (head_dim 256 kept), linear attention key heads 16 -> 2, value heads 32 -> 6 (head dims 128 kept) | |
| - Token embedding: full vocabulary kept, narrowed to the hidden size (tied embeddings kept) | |
| - MTP layer: kept, narrowed the same way | |
| - Vision encoder: 24 -> 2 blocks, hidden size 1024 -> 128, heads 16 -> 2, FFN 4096 -> 256, output 2560 -> 32 | |
| ## License | |
| Derived from Qwen3.5-4B by the Qwen team, under the [Apache License 2.0](LICENSE). | |