Instructions to use inferencerlabs/GLM-5.1-MLX-4.8bit-INF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use inferencerlabs/GLM-5.1-MLX-4.8bit-INF 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("inferencerlabs/GLM-5.1-MLX-4.8bit-INF")

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 inferencerlabs/GLM-5.1-MLX-4.8bit-INF with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "inferencerlabs/GLM-5.1-MLX-4.8bit-INF"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "inferencerlabs/GLM-5.1-MLX-4.8bit-INF"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use inferencerlabs/GLM-5.1-MLX-4.8bit-INF 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 "inferencerlabs/GLM-5.1-MLX-4.8bit-INF"

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 inferencerlabs/GLM-5.1-MLX-4.8bit-INF

Run Hermes

hermes

MLX LM

How to use inferencerlabs/GLM-5.1-MLX-4.8bit-INF with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "inferencerlabs/GLM-5.1-MLX-4.8bit-INF"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "inferencerlabs/GLM-5.1-MLX-4.8bit-INF"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "inferencerlabs/GLM-5.1-MLX-4.8bit-INF",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Quant_method error, might not work on LMStudio or oMLX

by NemoTLS - opened Apr 14

Discussion

NemoTLS

Apr 14

Hi, the model is not uploaded yet but following the pattern from previous recent uploads, is this model going to work with oMLX or LMStudio ?
Or is it henceforth only possible to use your models on the Inferencer app ?
Thank you

OpenClawGer

Apr 18

Did you figure out how to load it in LMStudio? on M3 Ultra 512GB RAM I get this error:

🥲 Failed to load the model

Failed to load model.

Error when loading model: ValueError: Received 2170 parameters not in model: 
lm_head.biases,
lm_head.scales,
model.embed_tokens.biases,
model.embed_tokens.scales,
model.layers.0.mlp.down_proj.biases,
model.layers.0.mlp.down_proj.scales,
model.layers.0.mlp.gate_proj.biases,
model.layers.0.mlp.gate_proj.scales,
model.layers.0.mlp.up_proj.biases,
model.layers.0.mlp.up_proj.scales,
model.layers.0.self_attn.indexer.weights_proj.biases,
model.layers.0.self_attn.indexer.weights_proj.scales,
model.layers.0.self_attn.indexer.wk.biases,
model.layers.0.self_attn.indexer.wk.scales,
model.layers.0.self_attn.indexer.wq_b.biases,
model.layers.0.self_attn.indexer.wq_b.scales,
model.layers.0.self_attn.kv_a_proj_with_mqa.biases,
model.layers.0.self_attn.kv_a_proj_with_mqa.scales,
model.layers.0.self_attn.o_proj.biases,
model.layers.0.self_attn.o_proj.scales,
model.layers.0.self_attn.q_a_proj.biases,
model.layers.0.self_attn.q_a_proj.scales,
model.layers.0.self_attn.q_b_proj.biases,
model.layers.0.self_attn.q_b_proj.scales,
model.layers.0.self_attn.unembed_out.biases,
model.layers.0.self_attn.unembed_out.scales,
model.layers.1.mlp.down_proj.biases,
model.layers.1.mlp.down_proj.scales,
model.layers.1.mlp.gate_proj.biases,
model.layers.1.mlp.gate_proj.scales,
model.layers.1.mlp.up_proj.biases,
model.layers.1.mlp.up_proj.scales,
model.layers.1.self_attn.indexer.weights_proj.biases,
model.layers.1.self_attn.indexer.weights_proj.scales,
model.layers.1.self_attn.indexer.wk.biases,
model.layers.1.self_attn.indexer.wk.scales,
model.layers.1.self_attn.indexer.wq_b.biases,
model.layers.1.self_attn.indexer.wq_b.scales,
model.layers.1.self_attn.kv_a_proj_with_mqa.biases,
model.layers.1.self_attn.kv_a_proj_with_mqa.scales,
model.layers.1.self_attn.o_proj.biases,
model.layers.1.self_attn.o_proj.scales,
model.layers.1.self_attn.q_a_proj.biases,
model.layers.1.self_attn.q_a_proj.scales,
model.layers.1.self_attn.q_b_proj.biases,
model.layers.1.self_attn.q_b_proj.scales,
model.layers.1.self_attn.unembed_out.biases,
model.layers.1.self_attn.unembed_out.scales,
model.layers.10.mlp.shared_experts.down_proj.biases,
....

NemoTLS

Apr 21

it seems like providing models that can run on other engines is no longer a priority...

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment