Instructions to use wqed/Ripple-65B-Preview-aligned with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use wqed/Ripple-65B-Preview-aligned with Transformers:

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

pipe = pipeline("text-generation", model="wqed/Ripple-65B-Preview-aligned", 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("wqed/Ripple-65B-Preview-aligned", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use wqed/Ripple-65B-Preview-aligned with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "wqed/Ripple-65B-Preview-aligned"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "wqed/Ripple-65B-Preview-aligned",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/wqed/Ripple-65B-Preview-aligned

SGLang

How to use wqed/Ripple-65B-Preview-aligned 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 "wqed/Ripple-65B-Preview-aligned" \
    --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": "wqed/Ripple-65B-Preview-aligned",
		"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 "wqed/Ripple-65B-Preview-aligned" \
        --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": "wqed/Ripple-65B-Preview-aligned",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use wqed/Ripple-65B-Preview-aligned with Docker Model Runner:
```
docker model run hf.co/wqed/Ripple-65B-Preview-aligned
```

Ripple-65B-Preview-aligned

A fine-tuned derivative of ricursiveAI/Ripple-65B-Preview. All modifications are folded directly into the weights — self-contained, no deltas, adapters, or extra files.

from transformers import AutoTokenizer, AutoModelForCausalLM
tok = AutoTokenizer.from_pretrained("wqed/Ripple-65B-Preview-aligned", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("wqed/Ripple-65B-Preview-aligned",
    dtype="bfloat16", device_map="cuda", trust_remote_code=True)

vLLM (nightly, --trust-remote-code, --max-model-len up to 1048576).

Load with trust_remote_code=True (custom Ricursive architecture). BF16, ~130 GB, needs a ≥275 GB GPU.

Derived work inherits Ricursive Intelligence's Ripple-65B-Preview terms; see the base model page for license details.

Downloads last month: 653

Safetensors

Model size

65B params

Tensor type

BF16

Model tree for wqed/Ripple-65B-Preview-aligned

Base model

ricursiveAI/Ripple-65B-Preview

Finetuned

(3)

this model