Instructions to use Ilides/cortex-1-v0.9 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Ilides/cortex-1-v0.9 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Ilides/cortex-1-v0.9", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Ilides/cortex-1-v0.9", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Ilides/cortex-1-v0.9 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 Ilides/cortex-1-v0.9:F16
# Run inference directly in the terminal:
llama cli -hf Ilides/cortex-1-v0.9:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Ilides/cortex-1-v0.9:F16
# Run inference directly in the terminal:
llama cli -hf Ilides/cortex-1-v0.9:F16

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 Ilides/cortex-1-v0.9:F16
# Run inference directly in the terminal:
./llama-cli -hf Ilides/cortex-1-v0.9:F16

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 Ilides/cortex-1-v0.9:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Ilides/cortex-1-v0.9:F16

Use Docker

docker model run hf.co/Ilides/cortex-1-v0.9:F16

LM Studio
Jan

vLLM

How to use Ilides/cortex-1-v0.9 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ilides/cortex-1-v0.9"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ilides/cortex-1-v0.9",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ilides/cortex-1-v0.9:F16

SGLang

How to use Ilides/cortex-1-v0.9 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Ilides/cortex-1-v0.9" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ilides/cortex-1-v0.9",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Ilides/cortex-1-v0.9" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ilides/cortex-1-v0.9",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use Ilides/cortex-1-v0.9 with Ollama:
```
ollama run hf.co/Ilides/cortex-1-v0.9:F16
```
Unsloth Desktop
Docker Model Runner
How to use Ilides/cortex-1-v0.9 with Docker Model Runner:
```
docker model run hf.co/Ilides/cortex-1-v0.9:F16
```

Lemonade

How to use Ilides/cortex-1-v0.9 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Ilides/cortex-1-v0.9:F16

Run and chat with the model

lemonade run user.cortex-1-v0.9-F16

List all available models

lemonade list

Atomic Chat

Cortex-1 v0.9 (sft-v0.9)

Modelo conversacional bilingüe ES/EN entrenado desde cero en MLX (bf16) y ajustado con SFT.

Arquitectura (verificada en config.json)

12 capas, hidden 768, 12 cabezas, intermediate 3072, contexto 512, vocab 16.384.
Pesos float32, ~505 MB (ckpt-best.npz).

Inicialización y datos

Init: pretrain v0.3 (val loss de referencia 2.5543 según configs/sft_v09.json).
SFT: 28.800 pasos, batch 16, lr 4e-5 coseno, seed 123.
Dataset chat sft-v0.9: 50 % ES / 50 % EN, ~45,6 M tokens de asistente en train (manifest).

Resultados

Validación limpia (2 281 ventanas de 512 tokens sin bloques presentes en train, de 8 642 totales): val loss 2,1781 · ppl 8,83.
Método: reports/eval_v09_filtered.py (misma máscara SFT que el entrenamiento). Reporte: reports/eval_v09_filtered.json.

Formatos

Repositorio principal: safetensors + GGUF (F16, Q8_0).
Copia MLX: cortex-1-v0.9-mlx/ (weights.safetensors + model.py).

Notas

Los resultados del benchmark anterior (bpb 0.540, 20,6 M) corresponden al modelo cortex-0.3-0.02b, NO a v0.9. Se han retirado de esta tarjeta.

Downloads last month: -

Safetensors

Model size

0.1B params

Tensor type

F32