Instructions to use croqaz/Sprocket-and-Say with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use croqaz/Sprocket-and-Say with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="croqaz/Sprocket-and-Say")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("croqaz/Sprocket-and-Say", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use croqaz/Sprocket-and-Say with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "croqaz/Sprocket-and-Say" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "croqaz/Sprocket-and-Say", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/croqaz/Sprocket-and-Say
- SGLang
How to use croqaz/Sprocket-and-Say 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 "croqaz/Sprocket-and-Say" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "croqaz/Sprocket-and-Say", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'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 "croqaz/Sprocket-and-Say" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "croqaz/Sprocket-and-Say", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use croqaz/Sprocket-and-Say with Docker Model Runner:
docker model run hf.co/croqaz/Sprocket-and-Say
Agentic 0,07 B LM tasks + complete list of use-cases for Vintage LMs
Databases for various use-cased of Vintage LMs needed. Maybe they exist in a Discord community or in some brainstorms of reviewers, in some blogs here and there. Collection into a compete list needed.
I can imagine the task of inference of 20 century physics with a contamination-free 19-century LM. But I am not informed about any agentic frameworks for this task or similar tasks. How authentic pre-1900 agents shoold work - not clear too. Brobably based on pre-1900 patents and programming in Ada style. The number of paremeters for a programming formal language must be smaller than for multi-lang coding LLMs based on the Stack corpus.
I think all vintage LLMs are very far from that usecase, right now the best vintage LLM TypeWriter can barely follow instructions. I mean, it's not bad, but compared to other modern LLMs with 7B params it's not there.
Also the vintage LLMs don't know any programming, don't know any tool calls, they don't even know JSON or XML.
We are very far from agentic LLM for now.
I have a list of vintage LLMs here: https://crlf.link/log/entries/260428-1/
I have to add the new one: https://www.unboundedlab.com/blog/bartholomew
That's all the English vintage LLMs that I know of
vintage LLMs don't know any programming, don't know any tool calls, they don't even know JSON or XML.
Syntactic JSON/XML knowlege or concepts of programming? They probably can program in a natural language programming paradigm, popularized by Andrej K. https://en.wikipedia.org/wiki/Vibe_coding
so the question about authentic vintage agentic frameworks. Defenition of it as a minimum.
"forget that the code even exists" - verbatim! Pre-1900 authentic task meaning in the original defenition.