hf-mem

CLI do Hugging Face para estimar a memória necessária para carregar pesos de modelos Safetensors ou GGUF para inferência do Hugging Face Hub

npx skills add https://github.com/huggingface/skills --skill hf-mem

hf_mem estimates the required memory for inference, including model weights and an optional KV cache, for Safetensors and GGUF for models on the Hugging Face Hub using HTTP Range requests i.e., without downloading or loading any weights locally.

When to use?

  • User asks how much VRAM or memory a model needs to run
  • User wants to know if a model fits on their GPU or a given instance
  • User references a Hugging Face model ID or URL and asks about inference requirements

What are the requirements?

  • uv installed (for uvx)
  • HF_TOKEN env var or --hf-token flag (for gated or private models only)

How to run?

Run with --model-id pointing to the Hugging Face Hub repository which will check that it either contains Safetensors (via model.safetensors, model.safetensors.index.json if sharded, or model_index.json for Diffusers) or GGUF model weights within.

uvx hf-mem --model-id <model-id> --json-output

If the repository contains GGUF model weights in multiple precisions / quantizations, the estimations will be on a per-file basis, whereas for inference you won't load all of those but rather only a single precision. This being said, for GGUF you might as well need to provide --gguf-file to target the specific file (or path if sharded) you want to run.

uvx hf-mem --model-id <model-id> --gguf-file <file-or-path> --json-output

Additionally, hf-mem comes with an --experimental flag that will also calculate the KV cache memory requirements too, useful for large-language models, meaning it applies to LLMs (...ForCausalLM), VLMs (...ForConditionalGeneration), and GGUF models.

As per the context window, it will be read from the default or overridden with --max-model-len a la vLLM. And, same goes for the KV cache precision, which will default to the model precision unless manually set via --kv-cache-dtype a la vLLM too.

For Safetensors use as:

uvx hf-mem --model-id <model-id> --experimental [--max-model-len N] [--batch-size N] [--kv-cache-dtype auto|bfloat16|fp8|fp8_ds_mla|fp8_e4m3|fp8_e5m2|fp8_inc] --json-output

And, for GGUF use as:

uvx hf-mem --model-id <model-id> --gguf-file <file-or-path> --experimental [--max-model-len N] [--batch-size N] [--kv-cache-dtype auto|F32|F16|Q4_0|Q4_1|Q5_0|Q5_1|Q8_0|Q8_1|Q2_K|Q3_K|Q4_K|Q5_K|Q6_K|Q8_K|IQ2_XXS|IQ2_XS|IQ3_XXS|IQ1_S|IQ4_NL|IQ3_S|IQ2_S|IQ4_XS|I8|I16|I32|I64|F64|IQ1_M|BF16|TQ1_0|TQ2_0|MXFP4] --json-output

Examples

For Transformers with Safetensors weights:

uvx hf-mem --model-id MiniMaxAI/MiniMax-M2 --json-output

For Diffusers with Safetensors weights:

uvx hf-mem --model-id Qwen/Qwen-Image --json-output

For Sentence Transformers with Safetensors weights:

uvx hf-mem --model-id google/embeddinggemma-300m --json-output

With --experimental to include the KV cache estimation for LLMs and VLMs:

uvx hf-mem --model-id mistralai/Mistral-7B-v0.1 --experimental --json-output

And, for LLMs or VLMs with GGUF weights:

uvx hf-mem --model-id unsloth/Qwen3.5-397B-A17B-GGUF --gguf-file Q4_K_M --experimental --json-output

Mais skills de huggingface

cpu-kernels
huggingface
Fornece orientação para escrever, otimizar e fazer benchmarking de kernels de CPU em C++ com intrínsecos SIMD (AVX2/AVX512) para o ecossistema de kernels do Hugging Face. Inclui…
official
generate-openenv-env
huggingface
Gere ambientes OpenEnv a partir de um caso de uso concreto (por exemplo, "gere um env para a biblioteca textarena"). Use quando for solicitado a projetar ou implementar um novo…
official
hf-mcp
huggingface
Use o Hugging Face Hub por meio das ferramentas do servidor MCP. Pesquise modelos, datasets, Spaces, papers. Obtenha detalhes de repositórios, busque documentação, execute jobs de computação e use Gradio…
official
trl-training
huggingface
Treine e ajuste modelos de linguagem transformer usando TRL (Transformers Reinforcement Learning). Suporta treinamento de SFT, DPO, GRPO, KTO, RLOO e Modelo de Recompensa…
official
deploy-hf
huggingface
Implante um ambiente OpenEnv no Hugging Face Spaces. Use quando for solicitado a implantar, enviar para o Hugging Face ou atualizar um espaço.
official
hf-space-recovery
huggingface
Diagnose and recover failing or stuck Hugging Face Space deployments for OpenEnv environments. Use when deploying envs from `envs/` to the Hub (`openenv`…
official
pre-submit-pr
huggingface
Validar alterações antes de enviar um pull request. Executar verificações abrangentes incluindo lint, testes, revisão de alinhamento e análise de RFC. Usar antes de criar um…
official
example-skill
huggingface
Skill de fixture de exemplo para testes de ação smoke
official