Up to date
EleutherAI Apache 2.0 out 4 years ago
506M parameters, text in, text out, 2K tokens of context.
Verified 2026-10-25 Rank #727 in LLMs · score 71.7 method v1 Sources 3 Downloads 30d 158.1k · #268 of 619 tracked
Picked by rule from 8 tracked pieces: the lab's own post, the most shared, a field report, the latest.
Pythia is a suite of 16 large language models all trained on public data seen in the exact same order and ranging in size from 70M to 12B parameters.
The 2 picks above, plus these 6. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 17 quantized builds and 100 Spaces.
Top Spaces running it: pliny-the-prompter/obliteratus ♥ 621 · EvanTHU/MotionLLM ♥ 43 · fromziro/SLM_Regression_Line ♥ 16 · stanley-00/slm-testing ♥ 9
Weights only, an estimate: leave room for the KV cache. Method
About 0.3 GB at INT4, 0.6 GB at FP8, 1.2 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2023-02-13), amber triangles are 7 dated articles (each links to it), grey ticks are repo commits.
264 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Listed in LLMs
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | GPTNeoXForCausalLM |
| Model type | gpt_neox |
| Dtype | float16 |
| Params by dtype | U8: 100,663,296 · F16: 405,334,208 |
| Size band | S |
| MoE | no |
| Context length | 2,048 (config.json max_position_embeddings) |
| Languages | en |
| Papers | arXiv:2304.01373, arXiv:2101.00027, arXiv:2201.07311 |
| License | Apache 2.0 (apache-2.0) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · GPTNeoXForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: not in registry · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Weights-only memory for pythia-410m: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 1.2 GB · Edge (<4 GB) |
|---|---|
| FP8 (1 byte/param) | 0.6 GB · Edge (<4 GB) |
| INT4 (0.5 bytes/param) | 0.3 GB · Edge (<4 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
finetune: 227quantized: 17adapter: 20
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| vandijklab/C2S-Pythia-410m-cell-type-prediction | finetune | — | 405M | S | cc-by-4.0 | vandijklab | 10 | 20.3k |
| mradermacher/pythia-410m-i1-GGUF | quantized | GGUF | — | — | apache-2.0 | mradermacher | 0 | 17.9k |
| vandijklab/C2S-Pythia-410m-diverse-single-and-multi-cell-tasks | finetune | — | 405M | S | cc-by-4.0 | vandijklab | 5 | 13.6k |
| mradermacher/pythia-410m-GGUF | quantized | GGUF | — | — | apache-2.0 | mradermacher | 0 | 10.8k |
| vandijklab/C2S-Pythia-410m-cell-type-conditioned-cell-generation | finetune | — | 405M | S | cc-by-4.0 | vandijklab | 2 | 3k |
| tensorblock/pythia-410m-GGUF | quantized | GGUF | — | — | apache-2.0 | tensorblock | 0 | 2.4k |
| shehryars715/pythia-410m-step50000-alpaca | finetune | — | 405M | S | other | shehryars715 | 1 | 1k |
| Xenova/pythia-410m | quantized | — | — | — | — | Xenova | 0 | 1.1k |
| MusaKlair/pythia410m-dpo-beta0.5 | finetune | — | 405M | S | — | MusaKlair | 0 | 668 |
| MusaKlair/pythia410m-dpo-beta1.0 | finetune | — | 405M | S | — | MusaKlair | 0 | 643 |
| MusaKlair/pythia410m-dpo-beta0.1 | finetune | — | 405M | S | — | MusaKlair | 0 | 597 |
| kykim0/pythia-410m-tulu-v2-mix | finetune | — | 405M | S | apache-2.0 | kykim0 | 0 | 525 |
| DNA-LLM/virus-pythia-302M-2048-cross_entropy | finetune | — | 302M | S | apache-2.0 | DNA-LLM | 0 | 507 |
| mechramc/kalavai-cross-lingual-welsh-specialist-seed137 | finetune | — | 405M | S | apache-2.0 | mechramc | 0 | 469 |
| mechramc/kalavai-cross-lingual-code-specialist-seed2026 | finetune | — | 405M | S | apache-2.0 | mechramc | 0 | 467 |
| islamrokon/QATrainedModel1 | finetune | — | — | — | apache-2.0 | islamrokon | 0 | 449 |
| shehryars715/pythia-410m-step5000-alpaca | finetune | — | 405M | S | — | shehryars715 | 2 | 324 |
| mechramc/kalavai-cross-lingual-yoruba-specialist-seed137 | finetune | — | 405M | S | apache-2.0 | mechramc | 0 | 375 |
| AliSaadatV/virus_pythia_410_1024_2d_representation_MSEPlusCE | finetune | — | 302M | S | apache-2.0 | AliSaadatV | 0 | 363 |
| ridaa4142/dpo-pythia-410m-beta-0_1 | finetune | — | 405M | S | — | ridaa4142 | 0 | 339 |
| mechramc/kalavai-phase1-410m-science-specialist-seed42 | finetune | — | 405M | S | apache-2.0 | mechramc | 0 | 335 |
| van-ng/pythia410-XYZCompany-1-steps | finetune | — | — | — | apache-2.0 | van-ng | 0 | 317 |
| ridaa4142/dpo-pythia-410m-beta-0_5 | finetune | — | 405M | S | — | ridaa4142 | 0 | 277 |
| ridaa4142/dpo-pythia-410m-beta-1_0 | finetune | — | 405M | S | — | ridaa4142 | 0 | 257 |
| mia-llm/pythia-410m-AGnews-roya | finetune | — | 405M | S | — | mia-llm | 0 | 223 |
| mechramc/kalavai-cross-lingual-code-specialist-seed137 | finetune | — | 405M | S | apache-2.0 | mechramc | 0 | 215 |
| mechramc/kalavai-cross-lingual-welsh-specialist-seed2026 | finetune | — | 405M | S | apache-2.0 | mechramc | 0 | 210 |
| mia-llm/pythia-410m-xsum-roya | finetune | — | 405M | S | — | mia-llm | 0 | 177 |
| Joshua-Abok/Malawi-Public-Health-Systems | finetune | — | 405M | S | apache-2.0 | Joshua-Abok | 1 | 118 |
| mia-llm/pythia-410m-xsum-oneepoch | finetune | — | 405M | S | — | mia-llm | 0 | 135 |
| OLAResearchX/xmemtransfer-pythia410m-from-pythia160m-20m | finetune | — | — | — | apache-2.0 | OLAResearchX | 0 | 132 |
| fbaldassarri/EleutherAI_pythia-410m-autogptq-int8-gs128-sym | quantized | gptq | 405M | S | apache-2.0 | fbaldassarri | 0 | 130 |
| devam-sheth-bits/sleep-ai-combined-evolving | finetune | — | 405M | S | apache-2.0 | devam-sheth-bits | 0 | 130 |
| ludyhasby/lamini_docs_100_steps | finetune | — | 405M | S | apache-2.0 | ludyhasby | 0 | 119 |
| devam-sheth-bits/finetuned-sleep-ai-multi-chat | finetune | — | 405M | S | apache-2.0 | devam-sheth-bits | 0 | 118 |
| ncgc/pythia_410m_hh_ga_lowest10k.sgd.nolora | finetune | — | 405M | S | — | ncgc | 0 | 115 |
| van-ng/pythia410m-XYZCompany-1000-steps | finetune | — | — | — | apache-2.0 | van-ng | 0 | 113 |
| mia-llm/pythia-410m-wikitext2raw-roya | finetune | — | 405M | S | — | mia-llm | 0 | 113 |
| gbemilekeonilude/oop_and_text_pythia_410m | finetune | — | 354M | S | apache-2.0 | gbemilekeonilude | 0 | 110 |
| crumb/qrstudy-410m-8-1 | adapter | — | — | — | — | crumb | 0 | 101 |
| Date | Commit | SHA |
|---|---|---|
| 2023-07-09 | Adding `safetensors` variant of this model (#2) | 9879c9b |
| 2023-06-08 | fix batch size and add paper | b39c68d |
| 2023-04-04 | Upload pytorch_model.bin | 33fc2fd |
| 2023-04-04 | Upload tokenizer_config.json | 81e530f |
| 2023-04-04 | Upload tokenizer.json | e5abfb6 |
| 2023-04-04 | Upload special_tokens_map.json | cc0d452 |
| 2023-04-04 | Upload config.json | 0ba9134 |
| 2023-03-29 | Create README.md | fdbf11a |
| 2023-02-13 | initial commit | b878107 |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
| Title | Source | Published | First seen | Signals |
|---|---|---|---|---|
| Pythia 1.4B reproduces 3.6% of training samples verbatim given 950-token prompts | ret2libc.com | 2026-06-09 | 2026-10-07 | hn |
| HuggingFace API Tokens were exposed, leaving users vulnerable | lasso.security | 2023-12-06 | 2026-10-07 | hn |
| RedPajama at 440B tokens higher quality than Pythia and StableLM | together.xyz | 2023-04-24 | 2026-10-07 | hn |
| Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling | arxiv.org | 2023-04-05 | 2026-10-07 | hn |
| Pythia: Interpreting Autoregressive Transformers Across Time and Scale | github.com | 2023-04-05 | 2026-10-07 | hn |
| Pythia: The first deep learning-based ancient text restoration model | archaeology.wiki | 2019-10-22 | 2026-10-07 | hn |
| Facebook Open-Sources Pythia for Vision and Language Multimodal AI Models | medium.com | 2019-05-21 | 2026-10-07 | hn |
| Pythia: Open-source framework for multimodal AI models | code.fb.com | 2019-05-21 | 2026-10-07 | hn |
arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:43 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:16 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
