Instructions to use KucLab/kuclab-hertz-0.6 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use KucLab/kuclab-hertz-0.6 with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf KucLab/kuclab-hertz-0.6:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf KucLab/kuclab-hertz-0.6:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf KucLab/kuclab-hertz-0.6:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf KucLab/kuclab-hertz-0.6:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf KucLab/kuclab-hertz-0.6:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf KucLab/kuclab-hertz-0.6:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf KucLab/kuclab-hertz-0.6:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf KucLab/kuclab-hertz-0.6:Q4_K_M

Use Docker

docker model run hf.co/KucLab/kuclab-hertz-0.6:Q4_K_M

LM Studio
Jan

vLLM

How to use KucLab/kuclab-hertz-0.6 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KucLab/kuclab-hertz-0.6"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KucLab/kuclab-hertz-0.6",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KucLab/kuclab-hertz-0.6:Q4_K_M

Ollama
How to use KucLab/kuclab-hertz-0.6 with Ollama:
```
ollama run hf.co/KucLab/kuclab-hertz-0.6:Q4_K_M
```

Unsloth Studio

How to use KucLab/kuclab-hertz-0.6 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for KucLab/kuclab-hertz-0.6 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for KucLab/kuclab-hertz-0.6 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for KucLab/kuclab-hertz-0.6 to start chatting

How to use KucLab/kuclab-hertz-0.6 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KucLab/kuclab-hertz-0.6:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "KucLab/kuclab-hertz-0.6:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use KucLab/kuclab-hertz-0.6 with Docker Model Runner:
```
docker model run hf.co/KucLab/kuclab-hertz-0.6:Q4_K_M
```

Lemonade

How to use KucLab/kuclab-hertz-0.6 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull KucLab/kuclab-hertz-0.6:Q4_K_M

Run and chat with the model

lemonade run user.kuclab-hertz-0.6-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use KucLab/kuclab-hertz-0.6 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KucLab/kuclab-hertz-0.6:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default KucLab/kuclab-hertz-0.6:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use KucLab/kuclab-hertz-0.6 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KucLab/kuclab-hertz-0.6:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "KucLab/kuclab-hertz-0.6:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

KucLab Hertz 0.6

A Czech/English STEM + programming assistant built by KucLab on top of google/gemma-4-12B-it. This release recovers and surpasses the STEM accuracy lost in Hertz 0.5, while keeping 0.5's direct/witty personality.

What this is

Hertz 0.6 is a LoRA fine-tune (r=16, merged into the base weights). Unlike 0.3-0.5, most of the training data was not self-distilled from a teacher model this round — the bulk (1994 rows) is a fresh, externally-generated corpus, checked for validity (0 malformed rows, 0 duplicate instructions/outputs, 0 special tokens, 0 identity leaks) before use. It's combined with 606 rows carried forward from Hertz 0.5's corpus (itself the product of three prior training generations), 309 CS↔EN scientific-terminology rows built deterministically from a fixed term list, 8 answer-first formatting examples, and 15 hand-written identity rows.

Base: google/gemma-4-12B-it (11.95B params, Apache 2.0)
Method: QLoRA, r=16 / alpha=32, merged to bf16 then quantized
Context: 262144 tokens natively. We tried extending this via YaRN rope scaling and hit an architecture wall: Gemma-4 stores RoPE settings in a nested rope_parameters structure (separate config per attention type), and rope_scaling is a legacy alias for that same field — writing one clobbers the other and breaks GGUF export. Context stays at the native 262144; a real extension would need a different technique than YaRN on this architecture.
Training data: 2932 rows total. See the honest development story below — this number is smaller than a first attempt that scored worse.
Format available: GGUF (q4_k_m, ~7.4GB) for llama.cpp/Ollama, plus the raw LoRA adapter.

Quickstart (Ollama)

Important: ollama pull hf.co/... alone does NOT apply this model's system prompt (identity + personality) — Ollama only fetches the raw GGUF from Hugging Face, it does not read a repo's Modelfile. Without the system prompt, the model falls back to identifying as a generic Gemma model. Use ollama create with the Modelfile below instead — it pulls the weights AND applies the system prompt in one step:

curl -O https://huggingface.co/KucLab/kuclab-hertz-0.6/resolve/main/Modelfile
ollama create kuclab-hertz-0.6 -f Modelfile
ollama run kuclab-hertz-0.6

(The Modelfile's FROM line points at hf.co/KucLab/kuclab-hertz-0.6:Q4_K_M, so this pulls the same GGUF automatically — no separate download needed.)

The honest development story

We're publishing this because it's the kind of thing that usually gets edited out of a release note, and we think it's more useful left in.

Attempt 1 used 2944 rows including 927 terminology rows (three phrasings per term, many of them bare one-word answers) and LoRA rank 32. Result: MMLU-Pro STEM 77.1%, but Czech terminology dropped to 71.4% (from 0.5's 75.7%), with the specifically-targeted EN→CS direction collapsing to 59.2%. Diagnosis: the bare-lookup rows taught the model to recall the 309 training pairs by rote rather than how Czech scientific terms are formed, so it fell apart on the benchmark's held-out quarter.

Attempt 2 fixed that (definitions only, no bare lookups, rank down to 16) but Czech terminology got worse still — 69.4%. Diagnosis: 0.6 was built entirely from a single fresh batch of data with nothing reused from earlier releases. 0.3→0.4→0.5 had each carried forward and re-reinforced the same Czech STEM vocabulary across three training generations; throwing that away for "everything fresh" lost more than 309 new terminology rows could replace in one round.

Attempt 3 (this release) added 606 rows back from Hertz 0.5's corpus alongside the fresh data, restoring some of that cumulative reinforcement. Also, a GGUF export crashed with KeyError: 'full_attention' on the first attempt at YaRN context extension — the root cause (rope_scaling clobbering Gemma-4's nested rope_parameters) is described above and is now guarded against in the export code rather than silently corrupting the config.

Net result: MMLU-Pro STEM ended at 79.2%, Czech terminology at 73.8% — real improvement over 0.5 on STEM, still short of 0.5's terminology peak. We're not aware of a way to have fully matched both without another full iteration cycle, which we didn't have time for before this release's deadline.

Benchmarks

Same prompts, same grading code, same Ollama Q4_K_M quantization, cold, identical corrected methodology throughout (see Hertz 0.5's card for the timeout bug we found and fixed there — this release inherits that fix).

MMLU-Pro STEM (240 held-out questions, this project's own curated subset)

	base	Hertz 0.4	Hertz 0.5	Hertz 0.6
Biology	86.7%	76.7%	78.3%	91.7%
Chemistry	61.7%	45.0%	53.3%	61.7%
Math	83.3%	76.7%	78.3%	90.0%
Physics	71.7%	56.7%	65.0%	73.3%
Total	75.8%	63.7%	68.8%	79.2%

Hertz 0.6 beats the base model by +3.4pp and every prior Hertz release on this benchmark. It also resolved far more answers cleanly: only 18/240 (7.5%) needed the fallback answer-only re-ask, down from 46/240 (19%) in Hertz 0.5 — a direct result of training on answer-first format examples, not a benchmark-harness artifact.

Czech terminology benchmark (206 held-out CS↔EN scientific terms)

	Hertz 0.3	Hertz 0.5	Hertz 0.6
CS→EN	79.6%	81.6%	82.5%
EN→CS	51.5%	69.9%	65.0%
Total	65.5%	75.7%	73.8%

Below Hertz 0.5's peak, above Hertz 0.3. See the development story above for why.

Honest status

✅ MMLU-Pro STEM: 79.2%, beats base (75.8%) and every prior Hertz release
✅ Personality retained from 0.5: direct, witty, no reflexive AI hedging on ordinary topics; genuinely harmful requests still refused
✅ Correctly identifies as a KucLab model, no founder named
⚠️ Czech terminology (73.8%) is below Hertz 0.5's 75.7% — a real, disclosed regression, not fully recovered despite a dedicated attempt (see development story)
⏳ Context extension beyond native 262144 was attempted and found architecturally blocked on this base via YaRN — not solved this release
⏳ No tool-calling fine-tuning; no uncensoring pass beyond the personality shift already in 0.5

License

Apache 2.0, inherited from google/gemma-4-12B-it (per Google's official Hugging Face listing).

Credits

Base model: google/gemma-4-12B-it (Google, Apache 2.0)
Fine-tuning, dataset construction, and packaging: KucLab

Downloads last month: 4

GGUF

Model size

12B params

Architecture

gemma4

Hardware compatibility

4-bit

Model tree for KucLab/kuclab-hertz-0.6

Base model

google/gemma-4-12B

Finetuned

google/gemma-4-12B-it

Adapter

(57)

this model