Warning
WIP
A CLI tool for viewing metadata of 🤗safetensors files.
- Supports viewing model parameters, layers and modelspec.
- Supports remote files on Hugging Face Hub without full model download. (See details)
cargo install --git https://github.com/p1atdev/safemetadatasafemtetadata params ./Qwen2-0.5B-Instruct/model.safetensors # local fileor
safemtetadata params model.safetensors --repo-id Qwen/Qwen2-0.5B-Instruct # on huggingface hubOutput:
Total parameters: 0.5B params
safemtetadata layers ./Qwen2-0.5B-Instruct/model.safetensorsor
safemtetadata layers model.safetensors --repo-id Qwen/Qwen2-0.5B-InstructOutput:
Tensor format: PyTorch
╭────────────────────────────────────────────┬──────────┬───────────────╮
│ Parameter Name │ DType │ Shape │
├────────────────────────────────────────────┼──────────┼───────────────┤
│ model.embed_tokens.weight │ bfloat16 │ [151936, 896] │
│ model.layers.0.input_layernorm.weight │ bfloat16 │ [896] │
│ model.layers.0.mlp.down_proj.weight │ bfloat16 │ [896, 4864] │
│ model.layers.0.mlp.gate_proj.weight │ bfloat16 │ [4864, 896] │
│ model.layers.0.mlp.up_proj.weight │ bfloat16 │ [4864, 896] │
│ model.layers.0.post_attention_layernorm.we │ bfloat16 │ [896] │
│ ight │ │ │
│ model.layers.0.self_attn.k_proj.bias │ bfloat16 │ [128] │
...
safemtetadata modelspec ./sd_xl_base_1.0_0.9vae.safetensorsor
safemtetadata modelspec sd_xl_base_1.0_0.9vae.safetensors --repo-id stabilityai/stable-diffusion-xl-base-1.0Output:
Stability AI Model Metadata Standard Specification
╭───────────────────────────┬────────────────────────────────────────────╮
│ Key │ Value │
├───────────────────────────┼────────────────────────────────────────────┤
│ modelspec.architecture │ "stable-diffusion-xl-v1-base" │
│ modelspec.author │ "StabilityAI" │
│ modelspec.date │ "2023-07-26" │
│ modelspec.description │ "SDXL 1.0 Base Model, compositional expert │
│ │ . SDXL, the most advanced development in t │
│ │ he Stable Diffusion text-to-image suite of │
│ │ models. SDXL produces massively improved │
│ │ image and composition detail over its pred │
│ │ ecessors. The ability to generate hyper-re │
│ │ alistic creations for films, television, m │
│ │ usic, and instructional videos, as well as │
│ │ offering advancements for design and indu │
│ │ strial use, places SDXL at the forefront o │
│ │ f real world applications for AI imagery." │
│ modelspec.hash_sha256 │ "0x5e756477ea9ddde7a552c6e2d7926f849e8b0df │
│ │ d2f0b513ff17b7d31faedd79f" │
│ modelspec.implementation │ "https://github.com/Stability-AI/generativ │
│ │ e-models" │
│ modelspec.license │ "CreativeML Open RAIL++-M License" │
│ modelspec.prediction_type │ "epsilon" │
│ modelspec.resolution │ "1024x1024" │
│ modelspec.sai_model_spec │ "1.0.0" │
...
Shows all metadata (__metadata__ field) of the model.
safemtetadata metadata ./sd_xl_base_1.0_0.9vae.safetensorsor
safemtetadata metadata sd_xl_base_1.0_0.9vae.safetensors --repo-id stabilityai/stable-diffusion-xl-base-1.0Removes metadata from the model. Only supports local files.
safemtetadata clean ./sd_xl_base_1.0_0.9vae.safetensors -o ./sd_xl_base_1.0_0.9vae-cleaned.safetensors