Instructions to use majentik/gpt-oss-20b-RotorQuant-MLX-8bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use majentik/gpt-oss-20b-RotorQuant-MLX-8bit 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("majentik/gpt-oss-20b-RotorQuant-MLX-8bit")

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
LM Studio

How to use majentik/gpt-oss-20b-RotorQuant-MLX-8bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "majentik/gpt-oss-20b-RotorQuant-MLX-8bit"

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": "majentik/gpt-oss-20b-RotorQuant-MLX-8bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use majentik/gpt-oss-20b-RotorQuant-MLX-8bit 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 "majentik/gpt-oss-20b-RotorQuant-MLX-8bit"

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 majentik/gpt-oss-20b-RotorQuant-MLX-8bit

Run Hermes

hermes

MLX LM

How to use majentik/gpt-oss-20b-RotorQuant-MLX-8bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "majentik/gpt-oss-20b-RotorQuant-MLX-8bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "majentik/gpt-oss-20b-RotorQuant-MLX-8bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "majentik/gpt-oss-20b-RotorQuant-MLX-8bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

majentik commited on 26 days ago

Commit

f82c4b7

verified ·

1 Parent(s): 8c881d1

docs: Tier 2 polish — variant matrix + quant trade-off

Browse files

Files changed (1) hide show

README.md +42 -10

README.md CHANGED Viewed

@@ -2,18 +2,16 @@
 base_model: openai/gpt-oss-20b
 library_name: mlx
 tags:
-- rotorquant
-- kv-cache-quantization
-- gpt-oss
-- openai
-- moe
-- quantized
-- mlx
-- 8bit
 license: apache-2.0
 pipeline_tag: text-generation
-language:
-- en
 ---
 # GPT-OSS-20B - RotorQuant MLX 8-bit
@@ -89,3 +87,37 @@ This model requires approximately 20 GB of unified memory. Recommended hardware:
 - [majentik/gpt-oss-20b-TurboQuant-MLX-8bit](https://huggingface.co/majentik/gpt-oss-20b-TurboQuant-MLX-8bit) -- TurboQuant MLX 8-bit variant
 - [RotorQuant GitHub](https://github.com/scrya-com/rotorquant)
 - [MLX Framework](https://github.com/ml-explore/mlx)

 base_model: openai/gpt-oss-20b
 library_name: mlx
 tags:
+  - rotorquant
+  - kv-cache-quantization
+  - gpt-oss
+  - openai
+  - moe
+  - quantized
+  - mlx
+  - 8bit
 license: apache-2.0
 pipeline_tag: text-generation
 ---
 # GPT-OSS-20B - RotorQuant MLX 8-bit
 - [majentik/gpt-oss-20b-TurboQuant-MLX-8bit](https://huggingface.co/majentik/gpt-oss-20b-TurboQuant-MLX-8bit) -- TurboQuant MLX 8-bit variant
 - [RotorQuant GitHub](https://github.com/scrya-com/rotorquant)
 - [MLX Framework](https://github.com/ml-explore/mlx)
+## Quant trade-off (MLX lane)
+| Bits | Approx size | Use case | Recommendation |
+|---|---|---|---|
+| 2-bit | ~5.2 GB | Aggressive quantization | Very low-RAM Macs |
+| 3-bit | ~7.2 GB | Lossy but small | Low-RAM Macs |
+| 4-bit | ~8.4 GB | Balanced default | Recommended for most Macs |
+| 5-bit | ~10 GB | Higher fidelity | Quality-sensitive |
+| 6-bit | ~12 GB | Approaching FP16 quality | High-fidelity |
+| **8-bit** | ~15 GB | Near-lossless reference | **Fidelity-critical work** |
+(Current variant — **8bit** — is bolded.)
+## Variants in this family
+(Showing 14 sibling variants under `majentik/gpt-oss-20b-*`. The current variant — `RotorQuant-MLX-8bit` — is **bolded**.)
+| Variant | Runtime | Approx size | Use case |
+|---|---|---|---|
+| [RotorQuant](https://huggingface.co/majentik/gpt-oss-20b-rotorquant) | runtime modifier | n/a | KV-cache root (weight-agnostic) |
+| [RotorQuant-GGUF-IQ4_XS](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-gguf-IQ4_XS) | llama.cpp | ~17 GB | Lossy 4-bit, low-RAM CPU/edge |
+| [RotorQuant-GGUF-Q2_K](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-gguf-Q2_K) | llama.cpp | ~12 GB | Lossy, low-RAM CPU/edge |
+| [RotorQuant-GGUF-Q3_K_M](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-gguf-Q3_K_M) | llama.cpp | ~16 GB | Smaller 3-bit, CPU-friendly |
+| [RotorQuant-GGUF-Q4_K_M](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-gguf-Q4_K_M) | llama.cpp | ~22 GB | Balanced default |
+| [RotorQuant-GGUF-Q5_K_M](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-gguf-Q5_K_M) | llama.cpp | ~26 GB | Higher fidelity, more RAM |
+| [RotorQuant-GGUF-Q8_0](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-gguf-Q8_0) | llama.cpp | ~42 GB | Near-lossless reference |
+| [RotorQuant-MLX-2bit](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-mlx-2bit) | mlx-lm | ~6.4 GB | Apple Silicon, smallest |
+| [RotorQuant-MLX-4bit](https://huggingface.co/majentik/gpt-oss-20b-rotorquant-mlx-4bit) | mlx-lm | ~12 GB | Apple Silicon balanced |
+| **RotorQuant-MLX-8bit** | mlx-lm | ~24 GB | Apple Silicon reference |
+| [TurboQuant](https://huggingface.co/majentik/gpt-oss-20b-turboquant) | runtime modifier | n/a | KV-cache root (weight-agnostic) |
+| [TurboQuant-MLX-2bit](https://huggingface.co/majentik/gpt-oss-20b-turboquant-mlx-2bit) | mlx-lm | ~6.4 GB | Apple Silicon, smallest |
+| [TurboQuant-MLX-4bit](https://huggingface.co/majentik/gpt-oss-20b-turboquant-mlx-4bit) | mlx-lm | ~12 GB | Apple Silicon balanced |
+| [TurboQuant-MLX-8bit](https://huggingface.co/majentik/gpt-oss-20b-turboquant-mlx-8bit) | mlx-lm | ~24 GB | Apple Silicon reference |