Instructions to use CalmCampus/calmcampus-reframe-1.7b-bf16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CalmCampus/calmcampus-reframe-1.7b-bf16 with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("CalmCampus/calmcampus-reframe-1.7b-bf16")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use CalmCampus/calmcampus-reframe-1.7b-bf16 with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "CalmCampus/calmcampus-reframe-1.7b-bf16"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "CalmCampus/calmcampus-reframe-1.7b-bf16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

MLX LM

How to use CalmCampus/calmcampus-reframe-1.7b-bf16 with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "CalmCampus/calmcampus-reframe-1.7b-bf16"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "CalmCampus/calmcampus-reframe-1.7b-bf16"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "CalmCampus/calmcampus-reframe-1.7b-bf16",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Hermes Agent

How to use CalmCampus/calmcampus-reframe-1.7b-bf16 with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "CalmCampus/calmcampus-reframe-1.7b-bf16"

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 CalmCampus/calmcampus-reframe-1.7b-bf16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use CalmCampus/calmcampus-reframe-1.7b-bf16 with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "CalmCampus/calmcampus-reframe-1.7b-bf16"

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 "CalmCampus/calmcampus-reframe-1.7b-bf16" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

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

CalmCampus Reframe 1.7B, bf16 (v0.1 research preview)

The unquantised (bfloat16) weights of the CalmCampus reframe model: Qwen3-1.7B with the CalmCampus LoRA adapter fused in, before the 4-bit quantisation step. It is published so others can convert it to other formats, quantise it differently or inspect it.

The 4-bit model, CalmCampus/calmcampus-reframe-1.7b-4bit, is the one the app is designed to run and the one that was evaluated. Its model card holds the task description, prompt contract, training data, evaluation results and limitations, and all of it applies here too, including:

Research preview. Not a medical device and not a standalone crisis detector. It misses some crisis messages, and its Dutch output is unreliable. Any app using it must run its own crisis check before calling the model and must always offer human support resources.

The evaluation numbers on the 4-bit card were measured on the 4-bit file, not on this one.

Details


Base model	Qwen/Qwen3-1.7B (Apache-2.0)
Precision	bfloat16
Size	about 3.4 GB
Architecture	`Qwen3ForCausalLM`
Prompt	`system_prompt.txt` in this repo. Use it exactly, with the Qwen chat template and thinking disabled

from mlx_lm import load, generate
model, tok = load("CalmCampus/calmcampus-reframe-1.7b-bf16")

It has been loaded and run with mlx-lm only. It has not been tested with other runtimes.

License

Apache-2.0, following the base model. Base model (c) the Qwen team.

Downloads last month: 33

Safetensors

Model size

2B params

Tensor type

BF16

MLX

Hardware compatibility

Quantized

Model tree for CalmCampus/calmcampus-reframe-1.7b-bf16

Base model

Qwen/Qwen3-1.7B-Base

Finetuned

Qwen/Qwen3-1.7B

Finetuned

(1,392)

this model

CalmCampus
/

calmcampus-reframe-1.7b-bf16

CalmCampus Reframe 1.7B, bf16 (v0.1 research preview)

Details

License

Model tree for CalmCampus/calmcampus-reframe-1.7b-bf16

Dataset used to train CalmCampus/calmcampus-reframe-1.7b-bf16