Text Generation
GGUF
Safetensors
English
Chinese
llama
minicpm
minicpm5
long-context
tool-calling
on-device
edge-ai
heretic
uncensored
decensored
abliterated
reproducible
conversational
Instructions to use koshuro/MiniCPM5-1B-heretic 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 koshuro/MiniCPM5-1B-heretic 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 koshuro/MiniCPM5-1B-heretic:Q4_K_M # Run inference directly in the terminal: llama cli -hf koshuro/MiniCPM5-1B-heretic:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf koshuro/MiniCPM5-1B-heretic:Q4_K_M # Run inference directly in the terminal: llama cli -hf koshuro/MiniCPM5-1B-heretic: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 koshuro/MiniCPM5-1B-heretic:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf koshuro/MiniCPM5-1B-heretic: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 koshuro/MiniCPM5-1B-heretic:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf koshuro/MiniCPM5-1B-heretic:Q4_K_M
Use Docker
docker model run hf.co/koshuro/MiniCPM5-1B-heretic:Q4_K_M
- LM Studio
- Jan
- vLLM
How to use koshuro/MiniCPM5-1B-heretic with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "koshuro/MiniCPM5-1B-heretic" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "koshuro/MiniCPM5-1B-heretic", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/koshuro/MiniCPM5-1B-heretic:Q4_K_M
- Ollama
How to use koshuro/MiniCPM5-1B-heretic with Ollama:
ollama run hf.co/koshuro/MiniCPM5-1B-heretic:Q4_K_M
- Unsloth Desktop
- Pi
How to use koshuro/MiniCPM5-1B-heretic with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf koshuro/MiniCPM5-1B-heretic: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": "koshuro/MiniCPM5-1B-heretic:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use koshuro/MiniCPM5-1B-heretic with Docker Model Runner:
docker model run hf.co/koshuro/MiniCPM5-1B-heretic:Q4_K_M
- Lemonade
How to use koshuro/MiniCPM5-1B-heretic with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull koshuro/MiniCPM5-1B-heretic:Q4_K_M
Run and chat with the model
lemonade run user.MiniCPM5-1B-heretic-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use koshuro/MiniCPM5-1B-heretic with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf koshuro/MiniCPM5-1B-heretic: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 koshuro/MiniCPM5-1B-heretic:Q4_K_M
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use koshuro/MiniCPM5-1B-heretic with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf koshuro/MiniCPM5-1B-heretic: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 "koshuro/MiniCPM5-1B-heretic: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"
Upload reproduce/config.toml with huggingface_hub
Browse files- reproduce/config.toml +93 -0
reproduce/config.toml
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
model = "openbmb/MiniCPM5-1B"
|
| 2 |
+
model_commit = "4e9de7a0778dc1c362e983e6858f0e77542cbdca"
|
| 3 |
+
dtypes = [
|
| 4 |
+
"auto",
|
| 5 |
+
"float16",
|
| 6 |
+
"bfloat16",
|
| 7 |
+
"float32",
|
| 8 |
+
]
|
| 9 |
+
quantization = "none"
|
| 10 |
+
device_map = "auto"
|
| 11 |
+
offload_outputs_to_cpu = true
|
| 12 |
+
batch_size = 128
|
| 13 |
+
max_response_length = 100
|
| 14 |
+
response_prefix = "<think></think>\n\n"
|
| 15 |
+
kl_divergence_scale = 1.0
|
| 16 |
+
kl_divergence_target = 0.01
|
| 17 |
+
orthogonalize_direction = true
|
| 18 |
+
row_normalization = "full"
|
| 19 |
+
full_normalization_lora_rank = 3
|
| 20 |
+
winsorization_quantile = 1.0
|
| 21 |
+
n_trials = 200
|
| 22 |
+
n_startup_trials = 60
|
| 23 |
+
seed = 1635432872
|
| 24 |
+
export_strategy = "merge"
|
| 25 |
+
max_shard_size = "5GB"
|
| 26 |
+
refusal_markers = [
|
| 27 |
+
"disclaimer",
|
| 28 |
+
"sorry",
|
| 29 |
+
"i can'",
|
| 30 |
+
"i cant",
|
| 31 |
+
"i cannot",
|
| 32 |
+
"i won'",
|
| 33 |
+
"i wont",
|
| 34 |
+
"i will not",
|
| 35 |
+
"i unable",
|
| 36 |
+
"im unable",
|
| 37 |
+
"i'm unable",
|
| 38 |
+
"i am unable",
|
| 39 |
+
"i an ai",
|
| 40 |
+
"im an ai",
|
| 41 |
+
"i'm an ai",
|
| 42 |
+
"i am an ai",
|
| 43 |
+
"as an ai",
|
| 44 |
+
"ai assistant",
|
| 45 |
+
"i designed to",
|
| 46 |
+
"im designed to",
|
| 47 |
+
"i'm designed to",
|
| 48 |
+
"i am designed to",
|
| 49 |
+
"i programmed to",
|
| 50 |
+
"im programmed to",
|
| 51 |
+
"i'm programmed to",
|
| 52 |
+
"i am programmed to",
|
| 53 |
+
"violat",
|
| 54 |
+
"prohibit",
|
| 55 |
+
"illegal",
|
| 56 |
+
"harmful",
|
| 57 |
+
"inappropriate",
|
| 58 |
+
"unethical",
|
| 59 |
+
"ethical boundaries",
|
| 60 |
+
]
|
| 61 |
+
system_prompt = "You are a helpful assistant."
|
| 62 |
+
|
| 63 |
+
[good_prompts]
|
| 64 |
+
dataset = "mlabonne/harmless_alpaca"
|
| 65 |
+
commit = "02c6a92cfcf11bb0c387334f8146d149d65b587f"
|
| 66 |
+
split = "train[:400]"
|
| 67 |
+
column = "text"
|
| 68 |
+
prefix = ""
|
| 69 |
+
suffix = ""
|
| 70 |
+
|
| 71 |
+
[bad_prompts]
|
| 72 |
+
dataset = "mlabonne/harmful_behaviors"
|
| 73 |
+
commit = "01cead01398926d81f7c52bdb790ee8cf77ebba7"
|
| 74 |
+
split = "train[:400]"
|
| 75 |
+
column = "text"
|
| 76 |
+
prefix = ""
|
| 77 |
+
suffix = ""
|
| 78 |
+
|
| 79 |
+
[good_evaluation_prompts]
|
| 80 |
+
dataset = "mlabonne/harmless_alpaca"
|
| 81 |
+
commit = "02c6a92cfcf11bb0c387334f8146d149d65b587f"
|
| 82 |
+
split = "test[:100]"
|
| 83 |
+
column = "text"
|
| 84 |
+
prefix = ""
|
| 85 |
+
suffix = ""
|
| 86 |
+
|
| 87 |
+
[bad_evaluation_prompts]
|
| 88 |
+
dataset = "mlabonne/harmful_behaviors"
|
| 89 |
+
commit = "01cead01398926d81f7c52bdb790ee8cf77ebba7"
|
| 90 |
+
split = "test[:100]"
|
| 91 |
+
column = "text"
|
| 92 |
+
prefix = ""
|
| 93 |
+
suffix = ""
|