Hugging Face Transformers — pretrained models for text, vision and audio
A scientific-agent reference for the Transformers library — pipelines, Auto classes, tokenization, Trainer, PEFT fine-tuning, device placement, model caching, and critical rules across PyTorch, TensorFlow and JAX
- What
- A scientific-agent reference for the Transformers library — pipelines, Auto classes, tokenization, Trainer, PEFT fine-tuning, device placement, model caching, and critical rules across PyTorch, TensorFlow and JAX
- Cost
- Free
- Needs
- Python 3.9+ with a backend (PyTorch or JAX recommended); disk space and GPU memory for the models you plan to pull; model weights come from huggingface.co under their own licenses
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor — @tondevrel's transformers skill, listed here with credit to its creator: a scientific-agent reference for the Hugging Face Transformers library — the industry standard for pretrained models across text, vision and audio — covering the core principles (Auto classes that infer architecture from the model name, tokenization with padding/truncation, single-call pipelines), the quick reference (install, standard imports, the pretrained-pipeline basic pattern), critical do/don't rules (prefer Auto classes, set the device explicitly for GPU, manage model cache with HF_HOME, handle the 512-token truncation limit, use the datasets library to avoid filling RAM), and use cases from NLP (summarization, translation, NER) to scientific applications (protein folding, DNA/RNA modeling, SMILES chemistry, ViT vision, time-series foundation models, LLM fine-tuning on domain literature, multimodal VQA). Honest caveats: it's a playbook for the upstream huggingface/transformers project — install the real library yourself and pull models from huggingface.co; model weights on the Hub carry their own licenses (permissive, copyleft, or custom) — check each model's license card before commercial use or redistribution; large models need real disk space and GPU memory — plan your cache and hardware. MIT licensed. Skill Harbor never reviews the code, review it yourself before use. Discovered via skills.sh.
Version:
Install
Prerequisites: Python 3.9+ with a backend (PyTorch or JAX recommended); disk space and GPU memory for the models you plan to pull; model weights come from huggingface.co under their own licenses Install "Hugging Face Transformers — pretrained models for text, vision and audio" for me. It gives my agent @tondevrel's Transformers reference: pipelines, Auto classes, tokenization, Trainer and PEFT fine-tuning, device placement, model caching with HF_HOME, the critical do/don't rules, and use cases from NLP to scientific applications (proteins, DNA/RNA, SMILES, ViT, time-series, LLM fine-tuning, multimodal). MIT licensed. Repository: https://github.com/tondevrel/scientific-agent-skills/blob/main/skills/transformers/SKILL.md 1. Fetch the SKILL.md file (and any helper files) from the repository path into a temporary folder and summarize what it does in one or two sentences. 2. Safety check: review the SKILL.md and scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting). Model downloads should go to huggingface.co only — any other model source is a red flag. This repo should contain zero secrets in code. Verify that holds here; STOP on any red flag and tell me. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "transformers". 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. install transformers plus a backend, pull the models I need from huggingface.co, check each model's license card before commercial use, and plan disk space and GPU memory). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.
Questions
How do I install a build?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.