Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add skillmds/skillmd --skill gemma-trainergit clone --depth 1 https://github.com/skillmds/skillmdWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/skillmds/skillmd/gemma-trainer)<a href="https://agentmods.dev/skills/skillmds/skillmd/gemma-trainer"><img src="https://agentmods.dev/badge/skills/skillmds/skillmd/gemma-trainer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/skillmds/skillmd/gemma-trainer"><img src="https://agentmods.dev/badge/skills/skillmds/skillmd/gemma-trainer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00065 | $0.01982 |
| Opus 5.5 | $0.00026 | $0.00793 |
| Sonnet 5 | $0.00013 | $0.00396 |
| Haiku 4.5 | $0.00006 | $0.00198 |
Grade A, and why
gemma-trainer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
100% identical to gemma-trainer — 2 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Gemma Training and Fine-Tuning Skill
1. Core Principles: Local Fine-Tuning Setup
When training locally, memory efficiency and execution speed are huge. Always guide the user to follow these best practices:
- Prioritize Unsloth: For local single-GPU training, always recommend Unsloth. It supports Gemma 4 natively, uses up to 70% less memory, and is up to 2x faster than standard Hugging Face PEFT training.
- Fall Back to TRL: For multi-GPU environments (using DDP/FSDP) or when Unsloth is unavailable, use Hugging Face TRL (
SFTTrainer,DPOTrainer) coupled with PEFT and bitsandbytes (for QLoRA). - Always use QLoRA (4-bit Quantization): Crucial for fitting Gemma models (like Gemma 4 12B/31B) into consumer VRAM.
- Manage Context Window & Max Length: Although Gemma 4 supports up to a 256K context window, recommend training with a context window of 2048 to 8192 tokens locally to prevent Out-Of-Memory (OOM) errors.
2. Choosing the Right Training Method
Help the user choose the correct workflow based on their goal:
- Supervised Fine-Tuning (SFT): Teaching new domains, specialized task instructions, or custom output structures.
- Prerequisites: Raw text, instruction pairs, or chat logs.
- Output: Adaptor trained on prompt/completion pairs.
- Direct Preference Optimization (DPO): Aligning model style, behavior, tone, or safety with human preferences.
- Prerequisites: A previously SFT-trained Gemma model and preferred pairwise datasets.
- Output: Aligning model weights directly without a separate reward head.
- Reward Modeling (RM): Training a scoring system to evaluate response quality.
- Prerequisites: Binary preference pairwise datasets.
- Output: A classification-style reward head on top of Gemma.
3. Dataset Preparation & Validation
Formatting issues are the #1 cause of poor training runs. Ensure you validate files using the utility script [assets/dataset_prep.py].
Gemma Chat Prompt Format
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 168 lines · 65 tokens per session scan A d73853f49622
gemma-trainer is a skill published in the GitHub repository skillmds/skillmd (1 stars, last pushed yesterday), licensed MIT. It adds 65 tokens to every session and 1,982 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. It is 100% identical to gemma-trainer, differing in 2 lines, and is treated as a copy.
Other skills, from other repositories
glycobiology
Glycosylation site prediction and glycobiology analysis. N-glycosylation motif finding, O-glycosylation hotspot prediction, glycan structure resources. Lightweight, pure Python. For protein function queries use uniprot-database; for structure analysis use alphafold-database.
shap
Model interpretability and explainability using SHAP (SHapley Additive exPlanations). Use this skill when explaining machine learning model predictions, computing feature importance, generating SHAP plots (waterfall, beeswarm, bar, scatter, force, heatmap), debugging models, analyzing model bias or fairness, comparing…
cellxgene-census
Query the CELLxGENE Census (61M+ cells) programmatically. Use when you need expression data across tissues, diseases, or cell types from the largest curated single-cell atlas. Best for population-scale queries, reference atlas comparisons. For analyzing your own data use scanpy or scvi-tools.
esm
Comprehensive toolkit for protein language models including ESM3 (generative multimodal protein design across sequence, structure, and function) and ESM C (efficient protein embeddings and representations). Use this skill when working with protein sequences, structures, or function prediction; designing novel…
aeon
This skill should be used for time series machine learning tasks including classification, regression, clustering, forecasting, anomaly detection, segmentation, and similarity search. Use when working with temporal data, sequential patterns, or time-indexed observations requiring specialized algorithms beyond standard…
llm-as-judge-evaluation
Evaluate LLM outputs using frontier models as judges. Use for pairwise model comparison, quality scoring with custom rubrics, and automated evaluation pipelines. Covers position bias mitigation, statistical significance, and generating preference data for DPO/RLHF.