Instructions to use Eugleo/exp089-d26-tpp10-format-only-sft with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Eugleo/exp089-d26-tpp10-format-only-sft with Transformers:

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

pipe = pipeline("text-generation", model="Eugleo/exp089-d26-tpp10-format-only-sft", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Eugleo/exp089-d26-tpp10-format-only-sft", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Eugleo/exp089-d26-tpp10-format-only-sft with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Eugleo/exp089-d26-tpp10-format-only-sft"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Eugleo/exp089-d26-tpp10-format-only-sft",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Eugleo/exp089-d26-tpp10-format-only-sft

SGLang

How to use Eugleo/exp089-d26-tpp10-format-only-sft 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 "Eugleo/exp089-d26-tpp10-format-only-sft" \
    --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": "Eugleo/exp089-d26-tpp10-format-only-sft",
		"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 "Eugleo/exp089-d26-tpp10-format-only-sft" \
        --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": "Eugleo/exp089-d26-tpp10-format-only-sft",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Eugleo/exp089-d26-tpp10-format-only-sft with Docker Model Runner:
```
docker model run hf.co/Eugleo/exp089-d26-tpp10-format-only-sft
```

exp089-d26-tpp10-format-only-sft

Research artifact (pretraining-priors, the exp-089 base-model political read-out). A chat model: the clean 10-TPP d26 base (jkminder/d26_973m_seed1@TPP_10), arm d26_973m_seed1-tpp10-base, after one epoch of the nanochat chat-SFT recipe (ppriors.sft.chat_sft, cold MuonAdamW, learning rates inherited from pretraining × 0.8, no warm-up, linear decay over the second half; batch 1048576 per step). SFT datasets: mmlu,gsm8k (see the knobs below). Final step 116, validation bpb 0.6286. Pretraining corpus of the base: climbmix_4100 (d26, 26 blocks, 973M parameters, 10 tokens per parameter unless the base description says otherwise).

SFT knobs recorded in the checkpoint:

knob	value
`data_seed`	0
`decor_replay_cats_dose`	1.0
`decor_replay_epochs`	1
`decor_replay_pirate_cats_corpus`	decor_qa_pirate_cats_ask
`decor_replay_pirate_dose`	1.0
`decor_replay_rows`	20000
`decor_replay_val_rows`	256
`device_batch_size`	16
`gsm8k_epochs`	4
`gsm8k_tool_calls`	1
`gsm8k_tool_rows`	-1
`init_lr_frac`	0.8
`mmlu_epochs`	3
`personas_replay_epochs`	1
`personas_replay_plain_rows`	30000
`personas_replay_rows`	10000
`personas_replay_val_rows`	256
`pirate2x2_replay_epochs`	1
`pirate2x2_replay_rows`	20000
`pirate2x2_replay_val_rows`	512
`pirate_gsm_corpus`	gsm8k_pirate
`pirate_gsm_epochs`	1
`pirate_gsm_rows`	29892
`pirate_gsm_val_rows`	420
`total_batch_size`	1048576

Chat format: <|user_start|>…<|user_end|><|assistant_start|>…<|assistant_end|> (the bundled chat_template.jinja); trust_remote_code=True (nanochat GPT architecture). Evaluations of this model: https://claude.ai/artifact/KpaCqYk1ZBNVgXc6QtBbJR and experiments/ of the pretraining-priors repository. Exported with ppriors.hf_export.convert_sft; nanochat checkpoint d26_973m_seed1-tpp10-base-sft-mmlugsm (step 116).

Downloads last month: 256

Safetensors

Model size

1.0B params

Tensor type

BF16

Model tree for Eugleo/exp089-d26-tpp10-format-only-sft

Base model

jkminder/d26_973m_seed1

Finetuned

(5)

this model