Instructions to use Unseen1980/daedalus-checkpoints with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Unseen1980/daedalus-checkpoints 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 Unseen1980/daedalus-checkpoints:F16
# Run inference directly in the terminal:
llama cli -hf Unseen1980/daedalus-checkpoints:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Unseen1980/daedalus-checkpoints:F16
# Run inference directly in the terminal:
llama cli -hf Unseen1980/daedalus-checkpoints: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 Unseen1980/daedalus-checkpoints:F16
# Run inference directly in the terminal:
./llama-cli -hf Unseen1980/daedalus-checkpoints: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 Unseen1980/daedalus-checkpoints:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Unseen1980/daedalus-checkpoints:F16

Use Docker

docker model run hf.co/Unseen1980/daedalus-checkpoints:F16

LM Studio
Jan

vLLM

How to use Unseen1980/daedalus-checkpoints with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Unseen1980/daedalus-checkpoints"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Unseen1980/daedalus-checkpoints",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Unseen1980/daedalus-checkpoints:F16

Ollama
How to use Unseen1980/daedalus-checkpoints with Ollama:
```
ollama run hf.co/Unseen1980/daedalus-checkpoints:F16
```

Unsloth Studio

How to use Unseen1980/daedalus-checkpoints with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Unseen1980/daedalus-checkpoints to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Unseen1980/daedalus-checkpoints to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Unseen1980/daedalus-checkpoints to start chatting

Docker Model Runner
How to use Unseen1980/daedalus-checkpoints with Docker Model Runner:
```
docker model run hf.co/Unseen1980/daedalus-checkpoints:F16
```

Lemonade

How to use Unseen1980/daedalus-checkpoints with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Unseen1980/daedalus-checkpoints:F16

Run and chat with the model

lemonade run user.daedalus-checkpoints-F16

List all available models

lemonade list

Atomic Chat

daedalus-checkpoints / hf /instruct /README.md

Unseen1980

instruct README.md

5e9572f verified 3 days ago

preview code

Raw

History Blame Contribute Delete

3.14 kB

metadata

license: apache-2.0
library_name: transformers
pipeline_tag: text-generation
tags:
  - daedalus
  - cpu-inference
  - gguf
  - q4_0

daedalus-150m

A 160.5M-parameter causal LM built for the best quality-per-token-per-second on CPU inference, exported to GGUF Q4_0 for llama.cpp.

What this model is trying to beat

Beat Pythia-160M, OPT-125M and GPT-neo-125M on quality; target MobileLLM-125M as a stretch; concede SmolLM2-135M on quality while beating it decisively on CPU decode.

This bar was fixed before any result landed. Numbers below are reported against it whether or not they clear it.

Architecture


exported as	`Lfm2ForCausalLM`
parameters	160,488,960 (122,740,224 non-embedding)
blocks	18 (`ccccAccAcAcAcAccAc` -- `c` = gated short conv, `A` = GQA attention)
hidden size	768
SwiGLU inner dim	2048
heads	12 query / 4 KV, head_dim 64, QK-norm
RoPE theta	1,000,000
context	2048
tied embeddings	True
tokenizer	`HuggingFaceTB/SmolLM2-135M`, reused byte-identical, vocab 49,152

Training

run: post-sft
Muon lr 0.002 on 2D hidden matrices; AdamW lr 3e-05 on embeddings/head/norms
WSD schedule, linear decay to zero over the final 45% of the run

Evaluation

Not yet measured for this export.

Q4_0 quantization

fp16 perplexity 217.8803 vs Q4_0 263.4498
delta 20.915%
passes the <1.0% threshold: False
llama.cpp CPU decode at 64 threads, by context depth:
- depth 0: 118.0 tok/s (+/- 3.8)
- depth 512: 103.3 tok/s (+/- 3.4)
- depth 2048 (the trained context): 97.5 tok/s (+/- 2.4)

Checkpoints and how to continue training

Checkpoints are pushed to the private Hub model repo Unseen1980/daedalus-checkpoints: weights-only bf16 rolling copies every ~2 h under rolling/<run>/weights.pt, plus a milestone with full Muon + AdamW optimizer state at the WSD decay-start step on its own revision.

No milestone record was found beside this checkpoint, so no branch point is published for it.

Deviations from the blueprint

Each was costed and approved rather than silently dropped; see DAEDALUS-BLUEPRINT-v6.md and issue #4.

No distillation from SmolLM2-1.7B during decay. 288 GB of top-16 logits does not fit the disk and the online-teacher variant cost ~$29 of a $94.66 budget; its own evidence was only "+1-3 points plausible".
Corpus stops at ~14.2B tokens, not 45B. Training repeats a balanced corpus rather than seeing 45B unique tokens; at this scale repetition up to ~4 epochs costs little against fresh tokens, and mixture balance mattered more than raw size.
Document-aligned packing not implemented -- sequences may cross document boundaries.
NoPE skipped -- it breaks GGUF export.
Single seed for the hero run, so no seed-sigma is reported.
everyday-conversations contributes ~0.00% of pretraining instead of its 2% share (the whole dataset is 0.4M tokens, which the 4-epoch cap reduces to nothing); dialogue enters at the post SFT stage instead.