Files
ttulttul-ComfyUI-Modal/llm_profiles.json
T

55 lines
1.9 KiB
JSON

{
"profiles": [
{
"id": "smolvlm2-2.2b-instruct",
"display_name": "SmolVLM2 2.2B Instruct",
"repository": "HuggingFaceTB/SmolVLM2-2.2B-Instruct",
"revision": "482adb537c021c86670beed01cd58990d01e72e4",
"dtype": "bfloat16",
"modalities": ["text", "image", "video", "file"],
"estimated_vram_gb": 6.0,
"max_context_tokens": 16384,
"max_images": 8,
"max_video_frames": 32,
"max_file_bytes": 33554432,
"max_file_characters": 200000,
"allow_mixed_image_video": false,
"trust_remote_code": false
},
{
"id": "huihui-qwen3.8-27b-abliterated-q2-k-gguf",
"display_name": "Huihui Qwen3.8 27B Abliterated Q2_K GGUF",
"repository": "huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF",
"revision": "2379b9294c14c0e55bd2ea5ac84d8bb9ffdfd98b",
"dtype": "bfloat16",
"modalities": ["text", "image", "file"],
"estimated_vram_gb": 16.0,
"max_context_tokens": 8192,
"max_images": 1,
"max_video_frames": 1,
"max_file_bytes": 33554432,
"max_file_characters": 200000,
"allow_mixed_image_video": false,
"trust_remote_code": false,
"backend": "llama_cpp_server",
"architecture": "Qwen3_5ForConditionalGeneration",
"quantization_method": "gguf_q2_k",
"artifact_bytes": 11795738048,
"advertised_context_tokens": 262144,
"reasoning_parser": "qwen3",
"backend_options": {
"cache_type_k": "q8_0",
"cache_type_v": "q8_0",
"context_size": 8192,
"gpu_layers": 999,
"model_filename": "Huihui-Qwen3.8-27B-abliterated-Q2_K.gguf",
"mmproj_filename": "mmproj-model-bf16.gguf",
"server_binary": "/app/llama-server",
"server_startup_timeout_seconds": 900,
"tokenizer_repository": "huihui-ai/Huihui-Qwen3.8-27B-abliterated",
"tokenizer_revision": "d42ca8978c5a66e92c3446d46e8adfe03ef692ff"
}
}
]
}