Instructions to use johnbean393/chiboard-1.1-0.8b-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use johnbean393/chiboard-1.1-0.8b-GGUF 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 johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M # Run inference directly in the terminal: llama cli -hf johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M # Run inference directly in the terminal: llama cli -hf johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama cli -hf johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
Use Docker
docker model run hf.co/johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
- Llama new
- LM Studio
- Jan
- Ollama
How to use johnbean393/chiboard-1.1-0.8b-GGUF with Ollama:
ollama run hf.co/johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
- Unsloth Desktop
- Pi
How to use johnbean393/chiboard-1.1-0.8b-GGUF with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf johnbean393/chiboard-1.1-0.8b-GGUF: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:9931/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use johnbean393/chiboard-1.1-0.8b-GGUF with Docker Model Runner:
docker model run hf.co/johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
- Lemonade
How to use johnbean393/chiboard-1.1-0.8b-GGUF with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
Run and chat with the model
lemonade run user.chiboard-1.1-0.8b-GGUF-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use johnbean393/chiboard-1.1-0.8b-GGUF with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf johnbean393/chiboard-1.1-0.8b-GGUF: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://127.0.0.1:9931/v1 hermes config set model.default johnbean393/chiboard-1.1-0.8b-GGUF:Q4_K_M
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use johnbean393/chiboard-1.1-0.8b-GGUF with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf johnbean393/chiboard-1.1-0.8b-GGUF: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://127.0.0.1:9931/v1 \ --custom-model-id "johnbean393/chiboard-1.1-0.8b-GGUF: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"
Chiboard 1.1 (GGUF)
GGUF builds of Chiboard 1.1, the local language model for the Chiboard Chinese input method. It converts pinyin to Hanzi as you type, revises earlier conversions as more keys arrive, and can read screenshots of the app you are typing into.
Files
Recommended: chiboard-1.1-0.8b-Q8_0.gguf with mmproj-chiboard-1.1-0.8b-f16.gguf. In our closed-loop evaluation this pair matches the BF16 model.
| File | Type | Size | Notes |
|---|---|---|---|
chiboard-1.1-0.8b-Q8_0.gguf |
text model, Q8_0 | 812 MB | Recommended. Matches BF16 on text and screenshots. |
chiboard-1.1-0.8b-Q6_K.gguf |
text model, Q6_K | 630 MB | No measurable loss on clean text or clean screenshots; not tested on noisy sets. |
chiboard-1.1-0.8b-Q4_K_M.gguf |
text model, Q4_K_M | 529 MB | Fine on text, but noticeably less accurate with screenshots. |
chiboard-1.1-0.8b-f16.gguf |
text model, F16 | 1.52 GB | Unquantized conversion; source of the quantized files. |
mmproj-chiboard-1.1-0.8b-f16.gguf |
vision projector, F16 | 205 MB | Recommended. Required for screenshot input. |
mmproj-chiboard-1.1-0.8b-q8_0.gguf |
vision projector, Q8_0 | 116 MB | No measurable difference from F16 in our tests. |
SHA-256 checksums are in SHA256SUMS. Converted with llama.cpp revision 2851b3bcbf8d993a2243ae751c0a7dfa8e944a40.
Download
hf download johnbean393/chiboard-1.1-0.8b-GGUF \
chiboard-1.1-0.8b-Q8_0.gguf mmproj-chiboard-1.1-0.8b-f16.gguf \
--local-dir .
Evaluation Results
These results are for the BF16 model; in our closed-loop tests the recommended Q8_0 build was statistically indistinguishable from it.
Text, clean
| Metric (%) | Chiboard 1.1 | Chiboard 1.1 SFT | Chiboard 1 |
|---|---|---|---|
| Final CER โ | 8.00 | 11.22 | 16.01 |
| Final EM โ | 46.30 | 43.10 | 46.05 |
| Per-state EM โ | 57.01 | 50.78 | 33.55 |
| First-state EM โ | 10.07 | 6.75 | 2.91 |
| Churn โ | 0.72 | 1.15 | 3.14 |
| Non-carry churn โ | 0.21 | 0.09 | 0.40 |
| Wrong carry commits โ | 11.08 | 22.30 | 57.79 |
Text, noisy
| Metric (%) | Chiboard 1.1 | Chiboard 1.1 SFT | Chiboard 1 |
|---|---|---|---|
| Final CER โ | 11.79 | 15.44 | 21.14 |
| Final EM โ | 38.31 | 36.10 | 37.58 |
| Per-state EM โ | 53.24 | 47.27 | 30.51 |
| First-state EM โ | 11.79 | 8.51 | 2.87 |
| Churn โ | 0.77 | 1.23 | 3.15 |
| Non-carry churn โ | 0.21 | 0.11 | 0.45 |
| Wrong carry commits โ | 12.18 | 23.76 | 56.85 |
Visual, clean
| Metric (%) | Chiboard 1.1 | Chiboard 1.1 SFT | Chiboard 1 |
|---|---|---|---|
| Final CER โ | 3.62 | 4.73 | 14.20 |
| Final EM โ | 73.37 | 71.54 | 50.10 |
| Per-state EM โ | 76.78 | 67.91 | 38.60 |
| First-state EM โ | 86.13 | 3.30 | 9.96 |
| Churn โ | 0.52 | 0.77 | 3.46 |
| Non-carry churn โ | 0.20 | 0.07 | 0.45 |
| Wrong carry commits โ | 6.52 | 13.65 | 56.21 |
Visual, noisy
| Metric (%) | Chiboard 1.1 | Chiboard 1.1 SFT | Chiboard 1 |
|---|---|---|---|
| Final CER โ | 4.96 | 6.20 | 15.80 |
| Final EM โ | 67.78 | 65.90 | 45.02 |
| Per-state EM โ | 74.00 | 65.45 | 36.32 |
| First-state EM โ | 84.76 | 4.32 | 9.71 |
| Churn โ | 0.56 | 0.82 | 3.42 |
| Non-carry churn โ | 0.21 | 0.07 | 0.46 |
| Wrong carry commits โ | 7.18 | 14.75 | 55.57 |
โ lower is better; โ higher is better.
- Chiboard 1.1: this model, Chiboard 1.1 SFT further trained with reinforcement learning.
- Chiboard 1.1 SFT: the supervised fine-tuned model that Chiboard 1.1 starts from.
- Chiboard 1: the previous model (
chiboard-1-qwen3.5-0.8b-t1-GGUF). It has no screen input, so on the visual sets it converts the same keystrokes without the screenshot.
Metrics:
- Final CER / final EM: character error rate and exact-match rate of the finished text.
- Per-state EM: share of intermediate conversions along the way that are exactly right, tolerant of where the composition is cut.
- First-state EM: share of sentences whose first conversion is exactly right.
- Churn: how often converted text already shown to the user changes, plus wrong carry commits. Non-carry churn counts only changes to converted characters that were already correct.
- Wrong carry commits: share of states with an unfinished final syllable in which the model commits a wrong character.
ยน "Noisy" sets include realistic typing mistakes.
All visual evaluation uses synthetic screenshots of mock and long-tail app screens; accuracy on real screens has not been measured.
Prompt format
<|chiboard_screen_ref|>{reference}<|chiboard_screen|>{current}<|chiboard_context|>{context}<|chiboard_pinyin|>{raw}<|chiboard_display|>{display}<|chiboard_output|>
reference: optional previous-window screenshot payload.current: optional payload for the window containing the active composition.context: committed text before the composition in the same field or document.raw: the original keystrokes of the entire current composition.display: the currently displayed composition, which may mix Chinese and unconverted pinyin. Serialize this field as empty when it is byte-identical toraw.
Keep all field tokens, even when their values are empty. Generate the converted composition immediately after <|chiboard_output|> and stop at <|endoftext|> (EOS token ID 248044).
The target covers the current composition only and stops at the typed boundary.
Example:
<|chiboard_screen_ref|><|chiboard_screen|><|chiboard_context|><|chiboard_pinyin|>nihao<|chiboard_display|><|chiboard_output|>
An expected completion is ไฝ ๅฅฝ<|endoftext|>.
Screenshot payloads
Each screen payload consists of an optional canonical application ID followed by optional image embeddings:
[app_id][<|vision_start|><|image_pad|> ร N<|vision_end|>]
License
This model is available under the Chiboard Model Non-Commercial and Non-Compete License 1.1.
- The model may not be commercialized.
- Nobody may use the model, its derivatives, or its outputs to build, train, evaluate, improve, or operate a competing input method.
- Ordinary end users may run the model as part of the official Chiboard application, including for everyday professional typing. Anyone may also use it for private, non-commercial research and education.
Read the complete license before downloading or using the model. The model is
derived from Qwen/Qwen3.5-0.8B-Base (Apache License 2.0; see
QWEN-APACHE-2.0.txt).
- Downloads last month
- 493
4-bit
6-bit
8-bit
16-bit
