| { | |
| "repository": "https://github.com/ggml-org/llama.cpp", | |
| "revision": "0ec4027a2a0183de87c007069adddc06e665f7de", | |
| "source_tree_clean": true, | |
| "build": { | |
| "type": "Release", | |
| "cuda": true, | |
| "cuda_toolkit": "13.0.88", | |
| "cuda_arch": "120a", | |
| "native_cpu": true, | |
| "build_parallelism": 4 | |
| }, | |
| "python": { | |
| "torch": "2.14.0+cu130", | |
| "transformers": "5.17.0", | |
| "huggingface_hub": "1.32.0", | |
| "safetensors": "0.8.0", | |
| "tokenizers": "0.23.2", | |
| "numpy": "2.5.3", | |
| "Pillow": "12.3.0", | |
| "gguf": "0.19.0", | |
| "sentencepiece": "0.2.2" | |
| }, | |
| "tools": { | |
| "llama-server": { | |
| "sha256": "12255637a664054701e296e3b9a0cdfeda4d62380c9c7af598f4408e9f6c21bd", | |
| "bytes": 17888 | |
| }, | |
| "llama-quantize": { | |
| "sha256": "3048c984ce69c78e4898f94fca85949384806114a02028ef03e175c1ced54e4d", | |
| "bytes": 17888 | |
| }, | |
| "llama-cli": { | |
| "sha256": "26ade31d637ae6ac69d398a38dd021d5fc5a9a23aced091f8e4a1d21e4d3358a", | |
| "bytes": 1474576 | |
| }, | |
| "llama-mtmd-cli": { | |
| "sha256": "2d607d2a8d5a42d949340789d3bdf82295aae3010621beec7864842e77662bce", | |
| "bytes": 106520 | |
| } | |
| }, | |
| "model_notes": { | |
| "special_output_flag": "--special required to retain location prefix control token", | |
| "processor": "IDEFICS3: longest edge 2048, 512px tiles plus overview, 4x pixel shuffle, mean/std 0.5, Lanczos. Do not add Qwen-specific image-token limits.", | |
| "tested_context": 4096, | |
| "tested_threads": 4 | |
| } | |
| } | |