bigscience / text-generation updated 3 years ago

bloomz-560m

Name: bloomz-560m
Author: bigscience

> We present BLOOMZ & mT0, a family of models capable of following human instructions in dozens of languages zero-shot. We finetune BLOOM & mT5 pretrained multilingual language models on our crosslingual task mixture (xP3) and find the resulting models capable of crosslingual generalization to unseen tasks & languages.

Params

560 M

Context

—

Downloads 30d

1.2 M

Likes

137

Commercial use: conditional · bigscience-bloom-rail-1.0 Not gated SAFETENSORS 46 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

1.2 M799 K

Jul 28Jul 31Aug 3Aug 6

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	f16	1.1 GB	1.8 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Run it

copy-paste, exact tags checked against the Hub

~ · curl · api/v1

$ curl -s https://aimodelscomparison.com/api/v1/models/bloomz-560m
{
  "hf_id": "bigscience/bloomz-560m",
  "params_b": 0.56,
  "context_length": null,
  "license": { "id": "bigscience-bloom-rail-1.0", "commercial": "conditional" },
  "downloads_30d": 1172809,
  "vram_estimates": [
    { "quant": "f16", "gb": 1.8 }
  ],
  "updated_at": "2026-07-28T18:04:05Z"
}

est. VRAM —on RTX 4090 · 24 GBJSON API →

Specifications

Architecture: BloomForCausalLM
Parameters: 560 M
Tensor type: F16
Vocabulary: 250,880
Licence: bigscience-bloom-rail-1.0
First seen on the Hub: 2022-10-08
Training datasets: bigscience/xP3
ANLI (r1) (reported): 33.4
ANLI (r2) (reported): 33.4
ANLI (r3) (reported): 33.5
SuperGLUE (cb) (reported): 53.57
XWinograd (en) (reported): 51.01
XWinograd (fr) (reported): 51.81
XWinograd (jp) (reported): 52.03
XWinograd (pt) (reported): 53.99
XWinograd (ru) (reported): 53.97
XWinograd (zh) (reported): 54.76
SuperGLUE (rte) (reported): 67.15
Winogrande XL (xl) (reported): 52.41
Added to our catalog: 2026-07-28

Compare with

Qwen3-0.6B 29.7 M Qwen2.5-1.5B-Instruct 14.0 M gpt2 13.7 M Llama-3.2-1B-Instruct 10.4 M Qwen2.5-0.5B-Instruct 6.6 M gemma-3-1b-it 4.7 M

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB