The story so far
Weights, every dated article, and repo commits on one rail.
30 repo commits since the weights landed
Teal is the weights (2022-05-11), grey ticks are repo commits.
opt-350m
facebook Other out 4 years ago
Text in, text out, 2K tokens of context.
Verified 2026-10-25 Rank #362 in LLMs · score 80.8 method v1 Sources 3 Downloads 30d 185.3k · #257 of 619 tracked
Can I use it?OtherCustom license; read the repo LICENSE before commercial hostinglicence file2026-10-24
Do people use it?185.3kdownloads, 30 days#257 of 619 we track · top 42%+2 likes, downloads +8% this week157 likesHugging Face2026-10-24
Where it's used
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: 19 quantized builds and 100 Spaces.
- As a service0no API provider listed
- On own hardware19quantized builds
- Built on436143 fine-tunes · 293 adapters · 0 merges
- In apps100+Hugging Face Spaces
Top Spaces running it: OpenGVLab/InternVL ♥ 510 · badayvedat/LLaVA ♥ 425 · LanguageBind/Video-LLaVA ♥ 234 · Yiwen-ntu/MeshAnything ♥ 166
What people built on it
455 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
455 derivatives on the Hub, most of them adapters
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Reference
Everything else, folded: the full tables, every source.
Use-case evidence
1 use case
What it is good at 1 use case
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.
Facts
18 fields
Listed in LLMs
Facts
- Context length is 2,048 tokens (2K). (config.json max_position_embeddings, observed 2026-10-24)
- License: Other (friction: review). Custom license; read the repo LICENSE before commercial hosting. (license file, observed 2026-10-24)
- Created 2022-05-11, last modified 2023-09-15. (HF API, observed 2026-10-24)
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | OPTForCausalLM |
| Model type | opt |
| Dtype | float16 |
| MoE | no |
| Context length | 2,048 (config.json max_position_embeddings) |
| Languages | en |
| Papers | arXiv:2205.01068, arXiv:2005.14165 |
| License | Other (other) |
| Hub snapshot | 2026-10-25 |
Servability
vLLM · SGLang
Servability
vLLM: supported · OPTForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · OPTForCausalLM · 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.
All variants
40 variants
Lineage
finetune: 143quantized: 19adapter: 293
Top variants (40 shown, ranked by downloads)
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| figmtu/opt-350m-aac | finetune | — | 331M | S | cc-by-4.0 | figmtu | 0 | 10.2k |
| nicholasKluge/Aira-OPT-350M | finetune | — | 331M | S | other | nicholasKluge | 0 | 3.7k |
| piazzola/address-detection-model | finetune | — | — | — | cc-by-nc-2.0 | piazzola | 0 | 1.6k |
| pszemraj/opt-350m-magicprompt-SD | finetune | — | 331M | S | other | pszemraj | 4 | 1.3k |
| Xenova/opt-350m | quantized | — | — | — | — | Xenova | 0 | 1.3k |
| santiviquez/reward_modeling_anthropic_hh | finetune | — | 331M | S | other | santiviquez | 7 | 653 |
| ong365/facebook-opt-350m-yoda-merged | finetune | — | 331M | S | — | ong365 | 0 | 813 |
| thisisHJLee/ko-opt-350m | finetune | — | 346M | S | other | thisisHJLee | 0 | 713 |
| tianyisun/opt-350m-finetuned-cola | finetune | — | — | — | other | tianyisun | 0 | 662 |
| 773bw-h/my_first_reward_modeling | finetune | — | 331M | S | other | 773bw-h | 6 | 341 |
| tianyisun/opt-350m-finetuned-sst2 | finetune | — | — | — | other | tianyisun | 0 | 560 |
| 01GangaPutraBheeshma/colab_text_generation_FT_opt_on_dolly_UT_V2 | finetune | — | 331M | S | other | 01GangaPutraBheeshma | 0 | 540 |
| daiweichen/pal-b-large-opt-350m | finetune | — | — | — | mit | daiweichen | 1 | 487 |
| Cheng98/opt-350m-boolq | finetune | — | — | — | other | Cheng98 | 0 | 500 |
| Cheng98/opt-350m-qnli | finetune | — | — | — | other | Cheng98 | 0 | 464 |
| dtorres-zAgile/opt-zc-misti-ft | finetune | — | — | — | other | dtorres-zAgile | 0 | 447 |
| ZachariahPang/medical_reward_model | finetune | — | 331M | S | — | ZachariahPang | 0 | 439 |
| Harshith42/opt-350m-HorrorCreepyPasta | finetune | — | 331M | S | other | Harshith42 | 0 | 438 |
| sartajbhuvaji/facebook-opt-350m-chat | finetune | — | 331M | S | apache-2.0 | sartajbhuvaji | 0 | 435 |
| SiMajid/opt-350-value | finetune | — | 331M | S | other | SiMajid | 0 | 411 |
| vwxyzjn/sft_openassistant-guanaco | finetune | — | 331M | S | other | vwxyzjn | 0 | 395 |
| Kinjal123/content | finetune | — | 331M | S | other | Kinjal123 | 0 | 375 |
| alexwb/reward_modeling_anthropic_hh_rm0.9_lr5e-5 | finetune | — | 331M | S | other | alexwb | 0 | 372 |
| alexwb/reward_modeling_anthropic_hh_rm0.99 | finetune | — | 331M | S | other | alexwb | 0 | 359 |
| acloudfan/fine-tuned-chat-model | finetune | — | 331M | S | — | acloudfan | 0 | 355 |
| varun-v-rao/opt-350m-snli-model3 | finetune | — | 331M | S | other | varun-v-rao | 1 | 289 |
| aisuko/sft-facebook-opt350m-with-openassistant-guanaco | adapter | — | — | — | other | aisuko | 0 | 329 |
| varun-v-rao/opt-350m-fp-sick | finetune | — | 331M | S | other | varun-v-rao | 0 | 324 |
| alexwb/reward_modeling_anthropic_hh | finetune | — | 331M | S | other | alexwb | 0 | 323 |
| AirinElizabath/JBS-3 | quantized | — | — | — | other | AirinElizabath | 0 | 322 |
| qgallouedec/reward_modeling_anthropic_hh | finetune | — | 331M | S | other | qgallouedec | 0 | 315 |
| zhangduo4610/opt350 | adapter | — | — | — | other | zhangduo4610 | 0 | 313 |
| vrvenkatesh/VarunOPTTest2-finetuned-slangQA | finetune | — | 331M | S | other | vrvenkatesh | 0 | 304 |
| alexwb/reward_modeling_anthropic_hh_rm1.4e-5 | finetune | — | 331M | S | other | alexwb | 0 | 297 |
| alexwb/reward_modeling_anthropic_hh_rm1e-4 | finetune | — | 331M | S | other | alexwb | 0 | 296 |
| SiMajid/value_reward_modeling | finetune | — | 331M | S | other | SiMajid | 5 | 44 |
| qgallouedec/reward_modeling_anthropic_hh_crc | finetune | — | 331M | S | other | qgallouedec | 0 | 294 |
| asoria/facebook-opt-350m-imdb | finetune | — | 331M | S | other | asoria | 0 | 293 |
| alexwb/reward_modeling_anthropic_hh_rm1e-3 | finetune | — | 331M | S | other | alexwb | 0 | 291 |
| alexwb/reward_modeling_anthropic_hh_rm1e-6 | finetune | — | 331M | S | other | alexwb | 0 | 289 |
Repository changelog
30 commits
Repository changelog 30 commits
| Date | Commit | SHA |
|---|---|---|
| 2023-09-15 | Update README.md (#32) | 08ab08c |
| 2023-01-24 | Adding generation config file(s) | cb32f77 |
| 2022-06-16 | Update TF weights (#3) | 10517ad |
| 2022-06-03 | Update config.json | d44304a |
| 2022-06-03 | Add tf_weights | 524c804 |
| 2022-05-26 | Change BOS token from 0 to 2 as BOS token is equal to EOS for OPT. See: https://github.com/huggingface/transformers/issues/17431 (#1) | 9eabca2 |
| 2022-05-25 | Update tokenizer_config.json | b186bd3 |
| 2022-05-25 | Upload tf_model.h5 | 794515b |
| 2022-05-24 | Upload tf_model.h5 | 1815efa |
| 2022-05-24 | Upload tf_model.h5 | c36c92f |
| 2022-05-19 | Update tf weights | 02f3be7 |
| 2022-05-18 | Upload tf_model.h5 | 5d8bedb |
| 2022-05-18 | Update config.json | 3dcd924 |
| 2022-05-17 | add model | 31d113e |
| 2022-05-16 | Set commercial attribute to False | 414672d |
| 2022-05-16 | Update README | 983ed6f |
| 2022-05-13 | Temporarily deactivate the inference API | 6c44fbb |
| 2022-05-13 | Update config.json | be9eb69 |
| 2022-05-12 | Update README.md | 3fd4708 |
| 2022-05-12 | Update README.md | b8a2001 |
| 2022-05-12 | Update README.md | 73349a1 |
| 2022-05-12 | Update README.md | 3817233 |
| 2022-05-12 | Update README.md | 44ab242 |
| 2022-05-12 | Update README.md | b7910e2 |
| 2022-05-12 | Update README.md | 592fff7 |
| 2022-05-12 | Update README.md | 1876a70 |
| 2022-05-12 | Update README.md | b5f37f5 |
| 2022-05-12 | Merge branch 'main' of https://huggingface.co/facebook/opt-350m into main | 9694c14 |
| 2022-05-12 | up | d589d0a |
| 2022-05-12 | Update README.md | 6958223 |
Papers
2 papers
Papers 2
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).
- Tom B. Brown, Benjamin Mann, Nick Ryder +28 · 2020-05-28 · neural information processing systems · PDF · HF paper page (21 upvotes) · evidence
Similar models
3 models
Similar models
- gpt-j-6b same task (text-generation), permissive license
- DialoGPT-medium same task (text-generation), permissive license
- Qwen-VL-Chat same task (text-generation)
Sources
3 sources
Sources
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:39 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:09 UTC · Last verified 2026-10-25
Follow opt-350m
Watch this model
- /feeds/models/facebook--opt-350m.xml(RSS 2.0) · every change to opt-350m
- /feeds/families/facebook--opt.xml(RSS 2.0) · every change across the opt family
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.