Instructions to use project-iconik/basikGPT-1-v1.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use project-iconik/basikGPT-1-v1.0 with Transformers:

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

pipe = pipeline("text-generation", model="project-iconik/basikGPT-1-v1.0")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("project-iconik/basikGPT-1-v1.0")
model = AutoModelForCausalLM.from_pretrained("project-iconik/basikGPT-1-v1.0", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use project-iconik/basikGPT-1-v1.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "project-iconik/basikGPT-1-v1.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "project-iconik/basikGPT-1-v1.0",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/project-iconik/basikGPT-1-v1.0

SGLang

How to use project-iconik/basikGPT-1-v1.0 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 "project-iconik/basikGPT-1-v1.0" \
    --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": "project-iconik/basikGPT-1-v1.0",
		"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 "project-iconik/basikGPT-1-v1.0" \
        --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": "project-iconik/basikGPT-1-v1.0",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use project-iconik/basikGPT-1-v1.0 with Docker Model Runner:
```
docker model run hf.co/project-iconik/basikGPT-1-v1.0
```

basikGPT-1 v1.0

124M parameters (GPT-2 Small). v1.0 = 2.5B training tokens, not 2.5B parameters.

from transformers import AutoModelForCausalLM, AutoTokenizer

tok = AutoTokenizer.from_pretrained("project-iconik/basikGPT-1-v1.0")
model = AutoModelForCausalLM.from_pretrained("project-iconik/basikGPT-1-v1.0")

Code and training recipe: github.com/project-iconik/basikGPT.
Whitepaper: EN · JA · KO.
Also available: basikGPT-1 v1.1 (same 124M, 5B tokens).

Model details


Architecture	GPT-2 Small decoder-only
Parameters	124,439,808
Context	1024
Vocab	50,257 (GPT-2 BPE / tiktoken `gpt2`)
Training tokens	2,500,001,792 (FineWeb-Edu)
Checkpoint	`runs/main_2p5b/step-00038147.pt` (step 38,147)
License (weights)	Apache-2.0

This Hub snapshot is a GPT2LMHeadModel export (safetensors + official GPT-2 tokenizer files). Optimizer state is not included.

Training data

HuggingFaceFW/fineweb-edu (sample-10BT), first 2.5B tokens, sequential (no shuffle).
FineWeb-Edu is released under ODC-By 1.0. Downstream use should respect that license and the original page licenses.

How it was trained

Single-GPU educational pretrain from random init. BF16, SDPA, batch 8 × 1024 × 8 grad accum (65,536 tokens/step), peak LR 6e-4 → min 6e-5, 2000-step warmup, cosine decay. Recipe and logs: the GitHub repo (docs/main_2p5b.md, runs/main_2p5b/).

Evaluation (same protocol as the repo)

Zero-shot English LM suite (english-lm-suite-v1) measured in-repo (not mixed with published paper numbers). Token counts and architectures are not matched across rows.

Model	size	HS	LAMBADA	PIQA	WG	ARC-E	Avg
v1.0	124M	29.40	19.58	61.37	50.51	43.01	40.77
v1.1	124M	28.75	23.05	61.75	50.83	38.51	40.58
openai-community/gpt2	124M	30.37	30.93	62.57	51.62	38.13	42.72
HuggingFaceTB/SmolLM2-135M	135M	42.67	42.97	67.57	51.93	59.43	52.91
EleutherAI/pythia-160m	162M	29.26	11.57	58.32	49.49	34.22	36.57
chance		25	—	50	50	~25	—

WG is acc_raw; other columns report the suite's primary metrics. Avg is the unweighted mean of these five primary metrics. n: HS 10,042 · LAMBADA 5,153 · PIQA 1,838 · WG 1,267 · ARC-E 2,376. Full protocol: benchmarks/REPORT.md. Methods and the full comparison: whitepaper.

v1.0 is the only checkpoint in this pair that beats official GPT-2 Small on ARC-Easy (+4.88pp). LAMBADA is still well below gpt2 (−11.35pp). WinoGrande is chance-level.

Intended use

Research and education: a small, reproducible GPT-2 Small trained on a documented public corpus. Not an instruction model. Not safe for open-ended production chat.

Limitations

124M / 2.5B tokens: weak on long-range last-word prediction and pronoun resolution.
English-centric FineWeb-Edu; no instruction or preference tuning.
May reproduce biases and factual errors from web text.

Citation

@software{basikgpt,
  title = {basikGPT},
  author = {basikGPT Contributors},
  url = {https://github.com/project-iconik/basikGPT},
  year = {2026}
}

Downloads last month: 515

Safetensors

Model size

0.1B params

Tensor type

F32

Model tree for project-iconik/basikGPT-1-v1.0

Finetunes

1 model

project-iconik
/

basikGPT-1-v1.0