Up to date
openai-community MIT out 5 years ago
380M parameters, text in, text out, 1K tokens of context.
Verified 2026-10-25 Rank #624 in LLMs · score 73.5 method v1 Sources 3 Downloads 30d 165.1k · #267 of 619 tracked
Picked by rule from 1 tracked piece: the lab's own post, the most shared, a field report, the latest.
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: 15 quantized builds and 100 Spaces.
Top Spaces running it: JohnSmith9982/ChuanhuChatGPT ♥ 921 · IndexTeam/IndexTTS-2-Demo ♥ 849 · pliny-the-prompter/obliteratus ♥ 621 · shi-labs/Versatile-Diffusion ♥ 408
Weights only, an estimate: leave room for the KV cache. Method
About 0.2 GB at INT4, 0.5 GB at FP8, 0.9 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 (2022-03-02), amber triangles are 1 dated article (each links to it), grey ticks are repo commits.
495 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.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported 1BW | 55.72 PPL | zero-shot | model author | 2026-10-07 |
| author-reported CBT-CN | 92.35 ACC | zero-shot | model author | 2026-10-07 |
| author-reported CBT-NE | 87.1 ACC | zero-shot | model author | 2026-10-07 |
| author-reported LAMBADA | 15.6 PPL | zero-shot | model author | 2026-10-07 |
| author-reported LAMBADA | 55.48 ACC | zero-shot | model author | 2026-10-07 |
| author-reported PTB | 47.33 PPL | zero-shot | model author | 2026-10-07 |
| author-reported WikiText103 | 26.37 PPL | zero-shot | model author | 2026-10-07 |
| author-reported WikiText2 | 22.76 PPL | zero-shot | model author | 2026-10-07 |
| author-reported enwiki8 | 1.01 BPB | zero-shot | model author | 2026-10-07 |
| author-reported text8 | 1.06 BPC | zero-shot | model author | 2026-10-07 |
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 | GPT2LMHeadModel |
| Model type | gpt2 |
| Params by dtype | F32: 379,988,992 |
| Size band | S |
| MoE | no |
| Context length | 1,024 (config.json max_position_embeddings) |
| Languages | en |
| Papers | arXiv:1910.09700 |
| License | MIT (mit) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · GPT2LMHeadModel · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · GPT2LMHeadModel · 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 gpt2-medium: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 0.9 GB · Edge (<4 GB) |
|---|---|
| FP8 (1 byte/param) | 0.5 GB · Edge (<4 GB) |
| INT4 (0.5 bytes/param) | 0.2 GB · Edge (<4 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
finetune: 187quantized: 15adapter: 293
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| nicholasKluge/Aira-2-355M | finetune | — | 355M | S | apache-2.0 | nicholasKluge | 2 | 24.1k |
| ayjays132/NeuroReasoner-1-NR-1 | finetune | — | 355M | S | apache-2.0 | ayjays132 | 2 | 21.7k |
| mradermacher/gpt2-medium-i1-GGUF | quantized | GGUF | — | — | mit | mradermacher | 0 | 21.7k |
| mradermacher/gpt2-medium-GGUF | quantized | GGUF | — | — | mit | mradermacher | 0 | 12.5k |
| MiniLLM/MiniLLM-gpt2-340M | finetune | — | — | — | apache-2.0 | MiniLLM | 5 | 8.1k |
| DevQuasar-10/openai-community.gpt2-medium-GGUF | quantized | GGUF | — | — | — | DevQuasar-10 | 1 | 6.7k |
| diffusionfamily/diffugpt-m | finetune | — | 355M | S | apache-2.0 | diffusionfamily | 0 | 6.2k |
| dakwi/chessgpt2-medium-m | finetune | — | 303M | S | mit | dakwi | 0 | 6.2k |
| raghavbali/gpt2-finetuned-headliner | finetune | — | 355M | S | mit | raghavbali | 1 | 3.7k |
| trojblue/gpt2-prompt-upscaler-v1 | finetune | — | 355M | S | mit | trojblue | 0 | 3.8k |
| snehangshu511/gpt2-medium-instruct | finetune | — | 406M | S | mit | snehangshu511 | 0 | 3.7k |
| tensorblock/gpt2-medium-GGUF | quantized | GGUF | — | — | mit | tensorblock | 0 | 3.6k |
| Alcoft/gpt2-medium-GGUF | quantized | GGUF | — | — | mit | Alcoft | 0 | 3.5k |
| augustocsc/gpt2_medium_prefix_682k | adapter | — | — | — | — | augustocsc | 0 | 2.7k |
| augustocsc/gpt2_medium_infix_682k | adapter | — | — | — | — | augustocsc | 0 | 2.5k |
| dakwi/chessgpt-medium-l | finetune | — | 355M | S | mit | dakwi | 0 | 1.7k |
| dakwi/chessgpt2-medium-l | finetune | — | 303M | S | mit | dakwi | 0 | 1.4k |
| OmAhire369/safe-genai-dpo-full | finetune | — | 355M | S | mit | OmAhire369 | 0 | 1.3k |
| pthinc/cicikus_classic | quantized | GGUF | 355M | S | mit | pthinc | 0 | 1.2k |
| jarod0411/zinc10M_gpt2-medium_SMILES_step1 | finetune | — | 355M | S | mit | jarod0411 | 0 | 1.2k |
| tskow21/gpt2-medium-pii | finetune | — | 355M | S | mit | tskow21 | 0 | 1.2k |
| caseyhahn/gpt2-medium-finetuned-genius-lyrics-updated-data-large | finetune | — | — | — | mit | caseyhahn | 0 | 1k |
| OmAhire369/safe-genai-ppo-full | finetune | — | 355M | S | mit | OmAhire369 | 0 | 977 |
| schaeff/gpt2-medium_LNFree500 | finetune | — | 355M | S | other | schaeff | 0 | 793 |
| IParraMartin/gpt2-medium-bLM100M | finetune | — | 355M | S | mit | IParraMartin | 0 | 793 |
| moos124/julioman122 | finetune | — | 355M | S | mit | moos124 | 0 | 693 |
| Sarthak279/Lyrics-Generation | finetune | — | — | — | mit | Sarthak279 | 1 | 638 |
| antoste/gpt2-medium-Q4_0-GGUF | quantized | Q4_0 | — | — | mit | antoste | 0 | 680 |
| li-jay-cs/gpt2-medium-supervised-summarize-checkpoint | finetune | — | — | — | mit | li-jay-cs | 0 | 671 |
| Octavian10/my-perfect-gpt2-medium | finetune | — | 355M | S | mit | Octavian10 | 1 | 585 |
| TomData/GPT2-review | finetune | — | 355M | S | — | TomData | 3 | 442 |
| Novaciano/GPT2-medium-NSFW_Amoral_Christmas-GGUF | quantized | GGUF | — | — | mit | Novaciano | 0 | 560 |
| cvnberk/gpt2-medium-finetuned-contract-gen | finetune | — | — | — | mit | cvnberk | 1 | 502 |
| sharoz/gpt2-medium-custom-functions-dataset-python | finetune | — | — | — | mit | sharoz | 0 | 547 |
| cwestnedge/gpt2-medium-pubmed | finetune | — | 355M | S | mit | cwestnedge | 0 | 533 |
| SaltyBeach/output | finetune | — | 355M | S | mit | SaltyBeach | 0 | 527 |
| KZMTx/redsolarsky4 | finetune | — | 355M | S | mit | KZMTx | 0 | 518 |
| ijwatson98/sft-gpt2-xsum | finetune | — | 355M | S | mit | ijwatson98 | 0 | 512 |
| umaru97/gpt2-product-review-generation | finetune | — | — | — | mit | umaru97 | 1 | 417 |
| MiniLLM/SFT-gpt2-340M | finetune | — | — | — | apache-2.0 | MiniLLM | 0 | 460 |
| Date | Commit | SHA |
|---|---|---|
| 2024-02-19 | Adds the tokenizer configuration file (#13) | 6dcaa7a |
| 2023-06-30 | Adding ONNX file of this model (#9) | f65d496 |
| 2023-02-22 | Adding `safetensors` variant of this model (#5) | 425b0cc |
| 2023-01-24 | Adding generation config file(s) | a8543b8 |
| 2022-08-23 | Remove explicit table of contents (#2) | e852c90 |
| 2022-07-22 | Add model card (#1) | 8c7ca69 |
| 2021-05-21 | upload flax model | db57b2f |
| 2021-05-21 | allow flax | 4c9f740 |
| 2020-12-11 | Migrate model card from transformers-repo | 6a81b3f |
| 2020-12-09 | For clarity, delete deprecated modelcard.json | b9fd73a |
| 2020-10-12 | Update tokenizer.json | 1538118 |
| 2020-05-11 | Update config.json | 8cc8ed4 |
| 2020-05-02 | Update rust_model.ot | a4b58be |
| 2020-04-24 | Update config.json | a0d42db |
| 2020-01-31 | Update config.json | 0a83037 |
| 2019-12-20 | Update modelcard.json | 242f801 |
| 2019-09-23 | Update tf_model.h5 | 5117318 |
| 2019-05-08 | Update config.json | 127883f |
| 2019-05-08 | Update merges.txt | 22651d3 |
| 2019-05-08 | Update pytorch_model.bin | 7b10018 |
| 2019-05-08 | Update vocab.json | e441e78 |
| 2019-05-08 | initial commit | d61fcf1 |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency.
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 21:41 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:13 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.
