Up to date
bigscience BLOOM RAIL 1.0 out 4 years ago
559M parameters, text in, text out.
Verified 2026-10-25 Rank #293 in LLMs · score 84.1 method v1 Sources 3 Downloads 30d 342k · #205 of 619 tracked
Picked by rule from 5 tracked pieces: the lab's own post, the most shared, a field report, the latest.
The 1 pick above, plus these 4. 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: 14 quantized builds and 100 Spaces.
Top Spaces running it: Manmay/tortoise-tts ♥ 316 · Sharathhebbar24/One-stop-for-Open-source-models ♥ 96 · KBaba7/Quant ♥ 51 · FallnAI/Quantize-HF-Models ♥ 19
Weights only, an estimate: leave room for the KV cache. Method
About 0.3 GB at INT4, 0.7 GB at FP8, 1.3 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 (2022-05-19), amber triangles are 1 dated article (each links to it), grey ticks are repo commits.
180 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 Train-time Evaluation | 2 other | — | model author | 2026-10-07 |
| author-reported Train-time Evaluation | 2 other | Training Loss | model author | 2026-10-07 |
| author-reported Train-time Evaluation | 2.2 other | Validation Loss | model author | 2026-10-07 |
| author-reported Train-time Evaluation | 8.9 other | Perplexity | 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 | BloomForCausalLM |
| Model type | bloom |
| Params by dtype | F16: 559,214,592 |
| Size band | S |
| MoE | no |
| Languages | ak, ar, as, bm, bn, ca, code, en, es, eu, fon, fr, gu, hi, id, ig, ki, kn, lg, ln, ml, mr, ne, nso, ny, or, pa, pt, rn, rw, sn, st, sw, ta, te, tn, ts, tum, tw, ur, vi, wo, xh, yo, zh, zhs, zht, zu |
| Papers | arXiv:1909.08053, arXiv:2110.02861, arXiv:2108.12409 |
| License | BLOOM RAIL 1.0 (bigscience-bloom-rail-1.0) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · BloomForCausalLM · 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 bloom-560m: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 1.3 GB · Edge (<4 GB) |
|---|---|
| FP8 (1 byte/param) | 0.7 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.
quantized: 14finetune: 37adapter: 129
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| afrideva/bloom-560m-GGUF | quantized | GGUF | — | — | bigscience-bloom-rail-1.0 | afrideva | 5 | 13k |
| nicholasKluge/Aira-2-portuguese-560M | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | nicholasKluge | 2 | 8.2k |
| mradermacher/bloom-560m-i1-GGUF | quantized | GGUF | — | — | bigscience-bloom-rail-1.0 | mradermacher | 0 | 5k |
| mradermacher/bloom-560m-GGUF | quantized | GGUF | — | — | bigscience-bloom-rail-1.0 | mradermacher | 0 | 4.1k |
| tensorblock/bloom-560m-GGUF | quantized | GGUF | — | — | bigscience-bloom-rail-1.0 | tensorblock | 0 | 3.3k |
| Xenova/bloom-560m | quantized | — | — | — | — | Xenova | 0 | 2.7k |
| jasoneden/bloom560m-squad-helloworld | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | jasoneden | 8 | 1.7k |
| liodon-ai/bloom-560m-imatrix-GGUF | quantized | imatrix | — | — | other | liodon-ai | 0 | 1.8k |
| MisterAI/bigscience_bloom-560m | quantized | GGUF | 559M | S | bigscience-bloom-rail-1.0 | MisterAI | 0 | 1.2k |
| iproskurina/bloom-560m-GPTQ-4bit-g128 | quantized | GPTQ | 555M | S | bigscience-bloom-rail-1.0 | iproskurina | 0 | 693 |
| liodon-ai/bloom-560m-FP8 | quantized | -FP8 | 559M | S | other | liodon-ai | 0 | 648 |
| ychafiqui/darija-text-generation | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | ychafiqui | 2 | 469 |
| franco-rojas/bloom-560m-finetuned-tfmviu | finetune | — | — | — | bigscience-bloom-rail-1.0 | franco-rojas | 0 | 468 |
| sudhir2016/test-galore | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | sudhir2016 | 0 | 442 |
| thealper2/bloom-560m-totto | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | thealper2 | 0 | 349 |
| SuYee189/wiki_bloom | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | SuYee189 | 0 | 342 |
| Theoreticallyhugo/bloom-full_labels | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | Theoreticallyhugo | 0 | 327 |
| dakoblov/bloom-560m-fr-pruned | finetune | — | 344M | S | bigscience-bloom-rail-1.0 | dakoblov | 0 | 276 |
| KapitalK/my_awesome_model | adapter | — | — | — | — | KapitalK | 0 | 248 |
| fernandoruiz/bloom-560m-Q4_0-GGUF | quantized | Q4_0 | — | — | bigscience-bloom-rail-1.0 | fernandoruiz | 0 | 237 |
| liodon-ai/bloom-560m-ONNX | quantized | — | — | — | other | liodon-ai | 0 | 229 |
| joshuaspear/bloom-560m-finetuned-health-qa | adapter | — | — | — | bigscience-bloom-rail-1.0 | joshuaspear | 0 | 200 |
| vedica1011/bloom-560m-lora-tagger | adapter | — | — | — | apache-2.0 | vedica1011 | 0 | 194 |
| pdazad/bloom-560-finetuned-owasp-8epochs | quantized | — | 559M | S | bigscience-bloom-rail-1.0 | pdazad | 0 | 192 |
| intelia-lab-uah/bloomz-560m_AE_SQAC | finetune | — | 559M | S | bigscience-bloom-rail-1.0 | intelia-lab-uah | 0 | 186 |
| venishpatidar/wa-ad-mod | finetune | — | 559M | S | mit | venishpatidar | 0 | 166 |
| Luciano/lora-bloom-560m-lener_br | adapter | — | — | — | — | Luciano | 0 | 149 |
| prabal1221/bloom-560m-finetuned-2 | adapter | — | — | — | bigscience-bloom-rail-1.0 | prabal1221 | 0 | 145 |
| rajuptvs/bloom-500M-lora | adapter | — | — | — | — | rajuptvs | 0 | 140 |
| shenyang/bloom-560m-qa-mit | adapter | — | — | — | — | shenyang | 0 | 129 |
| pachaar/bloom-560m-qa-mit | adapter | — | — | — | — | pachaar | 0 | 126 |
| andyluan/4bit-emotion-detection | adapter | — | — | — | bigscience-bloom-rail-1.0 | andyluan | 0 | 125 |
| rajuptvs/bloomlora | adapter | — | — | — | — | rajuptvs | 0 | 120 |
| aryaadhi/Bloom-Medical-QA-LoRA | adapter | — | — | — | bigscience-bloom-rail-1.0 | aryaadhi | 0 | 110 |
| xishanlone/bloom-560m-qa-finetuned | adapter | — | — | — | — | xishanlone | 0 | 105 |
| mou3az/QGfinal | adapter | — | — | — | — | mou3az | 0 | 104 |
| Luciano/qlora-bloom-560m-lener_br | adapter | — | — | — | — | Luciano | 0 | 103 |
| Hanzalwi/bloom-560m-finetuned-aings-validation-data-3 | adapter | — | 559M | S | — | Hanzalwi | 0 | 101 |
| Divyanshh/Bloom-560M | adapter | — | — | — | — | Divyanshh | 0 | 95 |
| ernchern/personal_info_classification | finetune | — | 559M | S | openrail | ernchern | 0 | 95 |
| Date | Commit | SHA |
|---|---|---|
| 2023-09-26 | Adding ONNX file of this model (#60) | ac2ae5f |
| 2023-05-11 | Model Card Restructure & edits (#56) | 4f42c91 |
| 2023-02-10 | Fix architecture (#50) | e985a63 |
| 2023-02-10 | Fix architecture (#46) | 957a865 |
| 2022-10-19 | Upload model.safetensors with huggingface_hub (#26) | afe2e6f |
| 2022-09-27 | Update config.json | 148e7a5 |
| 2022-08-08 | Update README.md (#10) | 7b5134c |
| 2022-07-21 | Remove warning (#9) | ef09465 |
| 2022-07-21 | 341b tokens model (#5) | ff9e624 |
| 2022-07-13 | Update tokenizer_config.json | 3e8a2f1 |
| 2022-07-12 | Update config.json (#8) | 7fdf123 |
| 2022-07-08 | Update README.md | de26ece |
| 2022-07-08 | Remove .dev from transformers version (#7) | 5ce56ba |
| 2022-07-06 | Add Flax weights (#6) | 697a51b |
| 2022-06-22 | Update tokenizer_config.json (#4) | 0628d1b |
| 2022-06-09 | Add correct pipeline (#2) | 4a0a1b7 |
| 2022-06-08 | Update config.json | e9703d8 |
| 2022-05-30 | Slightly modify model card (#1) | d586a2f |
| 2022-05-28 | Update config.json | 8ad61ef |
| 2022-05-27 | Copy+paste from updated model card, + warning. | 45a9da2 |
| 2022-05-25 | Update README.md | b277c40 |
| 2022-05-24 | Update date on Model Card | 78c24b3 |
| 2022-05-24 | Added intermediary checkpoint warning | 6a83bd4 |
| 2022-05-24 | Update Model Card | bec2a0d |
| 2022-05-20 | Update config.json | 5fc13c0 |
| 2022-05-20 | Create LICENSE | 6c2ccb9 |
| 2022-05-20 | new data | 11545ee |
| 2022-05-20 | new data | d7bae2c |
| 2022-05-19 | Upload special_tokens_map.json | e5e66cc |
| 2022-05-19 | Upload tokenizer_config.json | 9e4cbbc |
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.
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 21:37 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:45 UTC · Summary: 2026-10-07 15:05 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.
