# clip-vitb-mini-distilled by AbstractPhila: Open-Weight Model
Source: https://savrn.com/models/clip-vitb-mini-distilled
Markdown alternate of the page above; the site index is https://savrn.com/llms.txt

---

## Runs On

What it takes to serve clip-vitb-mini-distilled (9M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
| --- | --- | --- | --- | --- | --- |
| 16-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 |
| 8-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 |
| 4-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 |

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 9, 2026.

[clip-vitb-mini-distilled on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/clip-vitb-mini-distilled/gpus)

## Model Card

By AbstractPhila, published under apache-2.0, revision 791d0e36cdf7.

An 8.66M-parameter ViT image encoder (10.0% of a CLIP-B/16 image tower) producing 512-d embeddings compatible with the [CLIP-B/16 LAION-2B](https://huggingface.co/laion/CLIP-ViT-B-16-laion2B-s34B-b88K) text tower. The primary checkpoint was distilled on CC12M (10,968,539 images) against the generalized-Procrustes consensus of five CLIP teachers — never against the deployment teacher — and carries a frozen 512×512 rotation that maps its outputs into the deployment frame, where it outperforms the student distilled directly against that teacher on every task gauge, both seeds (full tables below).

### Quick start (AutoModel)

```
import torch
from transformers import AutoModel, AutoImageProcessor

repo = "AbstractPhil/clip-vitb-mini-distilled"
model = AutoModel.from_pretrained(repo, trust_remote_code=True).eval()
proc = AutoImageProcessor.from_pretrained(repo)

px = proc(images=[img], return_tensors="pt")["pixel_values"]
emb = model.get_image_features(px)     # (1, 512), L2-normalized,
                                       # deployment (LAION-B/16) frame
```

The rotation is applied by default (config.apply_rotation); pass apply_rotation=False to get_image_features for the raw consensus-frame embedding. Weights are safetensors; the modeling code is in this repo (modeling_clip_mini.py, configuration_clip_mini.py).

[Read the full model card (1,696 words)](https://savrn.com/models/clip-vitb-mini-distilled/card)

## Configuration

Architecture

ClipMiniModel

Layers

12

Hidden size

240

Attention heads

4

Model type

clip_vitb_mini

## Identity and Version

Repository

AbstractPhil/clip-vitb-mini-distilled

Publisher

AbstractPhila

Task

Image feature extraction

Modality

Other

Library

transformers

Parameters

9M parameters

Languages

vit

Revision

791d0e36cdf7e81d2e41128d5d0635636cb6ad08

First published

2026-07-27

Last updated

2026-10-09

## Files and Weights

163 files, 2.9 GB in total. The weights are 85 files totalling 2.9 GB in pt, safetensors.

Weights85 files · 2.9 GB

Configuration25 files · 324.2 KB

Documentation3 files · 49.4 KB

Other49 files · 128.2 KB

Repository1 file · 1.5 KB

Every file

| File | Type | Size | SHA-256 |
| --- | --- | --- | --- |
| cc12m/affinity_kl_s0_t88000.pt | Weights | 34.7 MB | ca8acd50734c |
| cc12m/affinity_kl_s1_t88000.pt | Weights | 34.7 MB | 05e7fa4ca78b |
| cc12m/cbert_full_s0_t88000.pt | Weights | 34.7 MB | 77fcb3ecfde2 |
| cc12m/cbert_full_s1_t88000.pt | Weights | 34.7 MB | a9537d858d00 |
| cc12m/consensus_gpa_s0_t88000.pt | Weights | 34.7 MB | d2e91dc09561 |
| cc12m/consensus_gpa_s1_t88000.pt | Weights | 34.7 MB | 6360fc7b4b01 |
| cc12m/consensus_nce_mse_s0_t88000.pt | Weights | 34.7 MB | 3ccc6205e54e |
| cc12m/consensus_nce_mse_s1_t88000.pt | Weights | 34.7 MB | 9eb00655d420 |
| cc12m/feature_mse_s0_t88000.pt | Weights | 34.7 MB | f79532ccb186 |
| cc12m/feature_mse_s1_t88000.pt | Weights | 34.7 MB | f98365e62210 |
| cc12m/infonce_s0_t88000.pt | Weights | 34.7 MB | 2c77821b7bd6 |
| cc12m/infonce_s1_t88000.pt | Weights | 34.7 MB | 0e31e0386ca6 |
| cc12m/mid/affinity_kl_s1_mid.pt | Weights | 34.7 MB | dcf3510378f0 |
| cc12m/mid/cbert_full_s0_mid.pt | Weights | 34.7 MB | 3e301e410b3f |
| cc12m/mid/cbert_full_s1_mid.pt | Weights | 34.7 MB | e5aa2c470b79 |
| cc12m/mid/consensus_gpa_s0_mid.pt | Weights | 34.7 MB | 7e88c0489dee |
| cc12m/mid/consensus_gpa_s1_mid.pt | Weights | 34.7 MB | 2a03b06f15ba |
| cc12m/mid/consensus_nce_mse_s0_mid.pt | Weights | 34.7 MB | 619d4cc11079 |
| cc12m/mid/consensus_nce_mse_s1_mid.pt | Weights | 34.7 MB | efd166647cf0 |
| cc12m/mid/feature_mse_s1_mid.pt | Weights | 34.7 MB | 4e3b8872ef4e |
| cc12m/mid/infonce_s1_mid.pt | Weights | 34.7 MB | add4851d568a |
| cc12m/mid/nce_mse_cv_s0_mid.pt | Weights | 34.7 MB | 37d58539e9f1 |
| cc12m/mid/nce_mse_cv_s1_mid.pt | Weights | 34.7 MB | d43fe808a8b1 |
| cc12m/mid/siglip_pairwise_s0_mid.pt | Weights | 35.0 MB | d3027647873b |
| cc12m/mid/siglip_pairwise_s1_mid.pt | Weights | 35.0 MB | 4b41dddf74b5 |
| cc12m/mid/x3_autograd_s0_mid.pt | Weights | 34.7 MB | 10e2eace6a7d |
| cc12m/mid/x3_autograd_s1_mid.pt | Weights | 34.7 MB | 3c49a3dfde29 |
| cc12m/nce_mse_cv_s0_t88000.pt | Weights | 34.7 MB | 8e4e443c41ec |
| cc12m/nce_mse_cv_s1_t88000.pt | Weights | 34.7 MB | e8c8935f40cb |
| cc12m/rotation_s0.pt | Weights | 1.0 MB | a5279eacde1b |
| cc12m/rotation_s1.pt | Weights | 1.0 MB | 8bc1e2a70e04 |
| cc12m/siglip_pairwise_s0_t88000.pt | Weights | 35.0 MB | a1d9200a5118 |
| cc12m/siglip_pairwise_s1_t88000.pt | Weights | 35.0 MB | 90435a53f5ee |
| cc12m/x3_autograd_s0_t88000.pt | Weights | 34.7 MB | 713c9c43ff49 |
| cc12m/x3_autograd_s1_t88000.pt | Weights | 34.7 MB | 3daa6a16bfd2 |
| coco/affinity_kl_s0_t8000.pt | Weights | 34.7 MB | ff0f56baf44f |
| coco/affinity_kl_s1_t8000.pt | Weights | 34.7 MB | 4a4e9e703d21 |
| coco/blueprint_s0_t8000.pt | Weights | 34.7 MB | 91e1cb3887da |
| coco/blueprint_s1_t8000.pt | Weights | 34.7 MB | d46d48f8f22e |
| coco/cbert_full_s0_t8000.pt | Weights | 34.7 MB | c915216eb4c1 |
| coco/cbert_full_s1_t8000.pt | Weights | 34.7 MB | 9523461c7e96 |
| coco/consensus_gpa_s0_t8000.pt | Weights | 34.7 MB | 74d699248455 |
| coco/consensus_gpa_s1_t8000.pt | Weights | 34.7 MB | 5f6572f1e3f9 |
| coco/consensus_nce_mse_s0_t8000.pt | Weights | 34.7 MB | 090e5bba6ffd |
| coco/consensus_nce_mse_s1_t8000.pt | Weights | 34.7 MB | 9ac63b8f2d04 |
| coco/feature_mse_s0_t8000.pt | Weights | 34.7 MB | 774080c231b6 |
| coco/feature_mse_s1_t8000.pt | Weights | 34.7 MB | 8c2ee5d7e31e |
| coco/infonce_s0_t8000.pt | Weights | 34.7 MB | ec258fdc7e02 |
| coco/infonce_s1_t8000.pt | Weights | 34.7 MB | 667628dafa81 |
| coco/nce_mse_cv_s0_t8000.pt | Weights | 34.7 MB | 4668acc5db17 |
| coco/nce_mse_cv_s1_t8000.pt | Weights | 34.7 MB | 5acf925fb4f2 |
| coco/siglip_pairwise_s0_t8000.pt | Weights | 35.0 MB | 6d972139a255 |
| coco/siglip_pairwise_s1_t8000.pt | Weights | 35.0 MB | 5f68e994daf9 |
| coco/x3_autograd_s0_t8000.pt | Weights | 34.7 MB | 9191b7fa5aa8 |
| coco/x3_autograd_s1_t8000.pt | Weights | 34.7 MB | f57ccbf5ccc7 |
| coco/x3_bce_s0_t8000.pt | Weights | 34.7 MB | b16ba4837758 |
| coco/x3_bce_s1_t8000.pt | Weights | 34.7 MB | 53b3917b1930 |
| coco/x3_full_s0_t8000.pt | Weights | 34.7 MB | c13544b790d4 |
| coco/x3_full_s1_t8000.pt | Weights | 34.7 MB | 79f083639621 |
| model.safetensors | Weights | 35.7 MB | 82c7102e1c7e |
| splat_screen/feature_mse_s4_splat_all_s0_t8000.pt | Weights | 34.7 MB | 1ff106c2976c |
| splat_screen/feature_mse_s4_splat_all_s1_t8000.pt | Weights | 34.7 MB | ad44cd2ccbd9 |
| splat_screen/infonce_rep1_s0_t8000.pt | Weights | 34.7 MB | a1ab260ca3fe |
| splat_screen/infonce_s2_s0_t2000.pt | Weights | 34.7 MB | 5d3370bd3240 |
| splat_screen/infonce_s2_splat_all_s0_t2000.pt | Weights | 34.7 MB | 6107833f1a6f |
| splat_screen/infonce_s3_splat_all_frozen_s0_t2000.pt | Weights | 34.7 MB | 41ece178371a |
| splat_screen/infonce_s3_splat_b12_s0_t2000.pt | Weights | 34.7 MB | 025c7a567039 |
| splat_screen/infonce_s3_splat_b1_3_s0_t2000.pt | Weights | 34.7 MB | 92b10eb68619 |
| splat_screen/infonce_s4_splat_all_frozen_s0_t8000.pt | Weights | 34.7 MB | 492687e36399 |
| splat_screen/infonce_s4_splat_all_frozen_s1_t8000.pt | Weights | 34.7 MB | 821d6fc860de |
| splat_screen/infonce_s4_splat_all_s0_t8000.pt | Weights | 34.7 MB | 0fe2aaf5abec |
| splat_screen/infonce_s4_splat_all_s1_t8000.pt | Weights | 34.7 MB | d1572d5b2983 |
| splat_screen/infonce_s4_splat_b1_6_s0_t8000.pt | Weights | 34.7 MB | 9ac20531ff57 |
| splat_screen/infonce_s4_splat_b1_6_s1_t8000.pt | Weights | 34.7 MB | 56c4b577a785 |
| splat_screen/infonce_s4_splat_b7_12_s0_t8000.pt | Weights | 34.7 MB | f4f781891e51 |
| splat_screen/infonce_s4_splat_b7_12_s1_t8000.pt | Weights | 34.7 MB | 3b51e7d0e26b |
| splat_screen/infonce_s5_ls_fixed1_s0_t2000.pt | Weights | 34.7 MB | 2d1aec61a016 |
| splat_screen/infonce_s5_ls_fixed1_s1_t2000.pt | Weights | 34.7 MB | 2c9779493af4 |
| splat_screen/infonce_s5_ls_learn_s0_t2000.pt | Weights | 34.7 MB | b666bfdba09b |
| splat_screen/infonce_s5_ls_learn_s0_t8000.pt | Weights | 34.7 MB | 9a8e0ff7ad88 |
| splat_screen/infonce_s5_ls_learn_s1_t2000.pt | Weights | 34.7 MB | 2e774544b74b |
| splat_screen/infonce_s5_ls_learn_s1_t8000.pt | Weights | 34.7 MB | ac62e61740f1 |
| splat_screen/infonce_s5_ls_zero_s0_t2000.pt | Weights | 34.7 MB | 31d04633c166 |
| splat_screen/infonce_s5_ls_zero_s1_t2000.pt | Weights | 34.7 MB | 79346e899cd5 |
| student_infonce_s0.pt | Weights | 34.7 MB | ec258fdc7e02 |
| config.json | Configuration | 487 B | — |
| configuration_clip_mini.py | Configuration | 1.2 KB | — |
| ledgers/coco_crossfit_2026-10-08.json | Configuration | 1.7 KB | — |
| ledgers/coco_heldout_rescore_2026-10-08.json | Configuration | 2.0 KB | — |
| ledgers/extra_ceiling.json | Configuration | 106 B | — |
| ledgers/frame_check.json | Configuration | 343 B | — |
| ledgers/frame_check_pod_s0.json | Configuration | 767 B | — |
| ledgers/frame_check_pod_s1.json | Configuration | 763 B | — |
| ledgers/frame_check_s1.json | Configuration | 347 B | — |
| ledgers/rotated_eval_consensus_gpa.json | Configuration | 292 B | — |
| ledgers/rotated_eval_consensus_nce_mse.json | Configuration | 290 B | — |
| loader.py | Configuration | 4.2 KB | — |
| modeling_clip_mini.py | Configuration | 4.0 KB | — |
| preprocessor_config.json | Configuration | 457 B | — |
| splat_screen/probe_init_sdpa_s0.json | Configuration | 1.3 KB | — |
| splat_screen/probe_init_splat_s0.json | Configuration | 2.6 KB | — |
| splat_screen/probe_ls_capacity_infonce_s4_splat_all_s0_t8000.json | Configuration | 150.9 KB | — |
| splat_screen/probe_ls_capacity_infonce_s4_splat_b7_12_s0_t8000.json | Configuration | 75.6 KB | — |
| train/bank_utils.py | Configuration | 1.9 KB | — |
| train/cc12m_data.py | Configuration | 6.1 KB | — |
| train/cc12m_gpa.py | Configuration | 3.6 KB | — |
| train/dist_align_gate.py | Configuration | 16.3 KB | — |
| train/dist_bed.py | Configuration | 41.2 KB | — |
| train/losses.py | Configuration | 6.2 KB | — |
| train/vitals.py | Configuration | 1.6 KB | — |
| README.md | Documentation | 13.9 KB | — |
| article_cc12m_distillation.md | Documentation | 28.8 KB | — |
| train/TRAINING.md | Documentation | 6.6 KB | — |
| ledgers/affinity_kl_s0.jsonl | Other | 1.1 KB | — |
| ledgers/affinity_kl_s1.jsonl | Other | 1.1 KB | — |
| ledgers/blueprint_s0.jsonl | Other | 708 B | — |
| ledgers/blueprint_s1.jsonl | Other | 702 B | — |
| ledgers/cbert_full_s0.jsonl | Other | 1.1 KB | — |
| ledgers/cbert_full_s1.jsonl | Other | 1.1 KB | — |
| ledgers/consensus_gpa_s0.jsonl | Other | 1.1 KB | — |
| ledgers/consensus_gpa_s1.jsonl | Other | 1.1 KB | — |
| ledgers/consensus_nce_mse_s0.jsonl | Other | 1.1 KB | — |
| ledgers/consensus_nce_mse_s1.jsonl | Other | 1.1 KB | — |
| ledgers/feature_mse_s0.jsonl | Other | 1.4 KB | — |
| ledgers/feature_mse_s1.jsonl | Other | 1.4 KB | — |
| ledgers/infonce_s0.jsonl | Other | 1.8 KB | — |
| ledgers/infonce_s1.jsonl | Other | 1.4 KB | — |
| ledgers/nce_mse_cv_s0.jsonl | Other | 1.1 KB | — |
| ledgers/nce_mse_cv_s1.jsonl | Other | 1.1 KB | — |
| ledgers/siglip_pairwise_s0.jsonl | Other | 1.1 KB | — |
| ledgers/siglip_pairwise_s1.jsonl | Other | 1.1 KB | — |
| ledgers/teacher_ceiling_s0.jsonl | Other | 317 B | — |
| ledgers/x3_autograd_s0.jsonl | Other | 1.1 KB | — |
| ledgers/x3_autograd_s1.jsonl | Other | 1.1 KB | — |
| ledgers/x3_bce_s0.jsonl | Other | 404 B | — |
| ledgers/x3_bce_s1.jsonl | Other | 403 B | — |
| ledgers/x3_full_s0.jsonl | Other | 407 B | — |
| ledgers/x3_full_s1.jsonl | Other | 407 B | — |
| ledgers/zs_floor_s0.jsonl | Other | 317 B | — |
| splat_screen/feature_mse_s4_splat_all_s0.jsonl | Other | 4.7 KB | — |
| splat_screen/feature_mse_s4_splat_all_s1.jsonl | Other | 4.7 KB | — |
| splat_screen/infonce_evalchk_s0.jsonl | Other | 361 B | — |
| splat_screen/infonce_rep1_s0.jsonl | Other | 2.4 KB | — |
| splat_screen/infonce_s2_s0.jsonl | Other | 2.4 KB | — |
| splat_screen/infonce_s2_splat_all_s0.jsonl | Other | 4.7 KB | — |
| splat_screen/infonce_s3_splat_all_frozen_s0.jsonl | Other | 4.7 KB | — |
| splat_screen/infonce_s3_splat_b12_s0.jsonl | Other | 2.6 KB | — |
| splat_screen/infonce_s3_splat_b1_3_s0.jsonl | Other | 3.0 KB | — |
| splat_screen/infonce_s4_splat_all_frozen_s0.jsonl | Other | 4.7 KB | — |
| splat_screen/infonce_s4_splat_all_frozen_s1.jsonl | Other | 4.7 KB | — |
| splat_screen/infonce_s4_splat_all_s0.jsonl | Other | 4.7 KB | — |
| splat_screen/infonce_s4_splat_all_s1.jsonl | Other | 4.7 KB | — |
| splat_screen/infonce_s4_splat_b1_6_s0.jsonl | Other | 3.5 KB | — |
| splat_screen/infonce_s4_splat_b1_6_s1.jsonl | Other | 3.5 KB | — |
| splat_screen/infonce_s4_splat_b7_12_s0.jsonl | Other | 3.6 KB | — |
| splat_screen/infonce_s4_splat_b7_12_s1.jsonl | Other | 3.6 KB | — |
| splat_screen/infonce_s5_ls_fixed1_s0.jsonl | Other | 8.0 KB | — |
| splat_screen/infonce_s5_ls_fixed1_s1.jsonl | Other | 4.0 KB | — |
| splat_screen/infonce_s5_ls_learn_s0.jsonl | Other | 12.2 KB | — |
| splat_screen/infonce_s5_ls_learn_s1.jsonl | Other | 8.2 KB | — |
| splat_screen/infonce_s5_ls_zero_s0.jsonl | Other | 4.0 KB | — |
| splat_screen/infonce_s5_ls_zero_s1.jsonl | Other | 4.0 KB | — |
| .gitattributes | Repository | 1.5 KB | — |

## License and Download

License

apache-2.0

Access

Open weights, no gate

Download size

2.9 GB

[Download from AbstractPhila](https://huggingface.co/AbstractPhil/clip-vitb-mini-distilled)

Released by AbstractPhila through its official repository on Hugging Face. [Read the license](https://www.apache.org/licenses/LICENSE-2.0).

## Built From

- Described by [arXiv:1807.03748](https://savrn.com/papers/representation-learning-with-contrastive-predictive-coding)
- Described by arXiv:2102.08981
- Described by [arXiv:2103.00020](https://savrn.com/papers/learning-transferable-visual-models-from-natural-language-supervision)
- Described by arXiv:2212.07143
- Described by [arXiv:2303.15343](https://savrn.com/papers/sigmoid-loss-for-language-image-pre-training)
- Described by arXiv:2307.12732
- Described by arXiv:2309.12314

## Memory Requirements

| Precision | Weights in memory |
| --- | --- |
| As published | 2.9 GB |
| 16-bit | 0.0 GB |
| 8-bit | 0.0 GB |
| 4-bit | 0.0 GB |

Weights only, from the published parameter count; the key-value cache and runtime add to this.

## Questions About clip-vitb-mini-distilled

### How much GPU memory does clip-vitb-mini-distilled need?

About 0 GB at 16-bit and 0 GB at 4-bit: the weights (9M parameters) plus a working margin. A long context needs more.

### What is the cheapest GPU to run clip-vitb-mini-distilled on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

### Can I use clip-vitb-mini-distilled commercially?

Yes. clip-vitb-mini-distilled is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

## Similar Models

Model · Image feature extraction

### [I-JEPA-Agro](https://savrn.com/models/i-jepa-agro)

[Junior R F Junior](https://savrn.com/model-publishers/joserfjunior)

Projeto completo de triagem visual de anomalias em folhas de soja usando como extrator de representações e HeraclitusDB como banco de eventos e vetores. 1. lê um manifesto com fazenda, talhão, safra e cultivar; 2. executa o I-JEPA ViT-G/16 localmente; 3. aplica média aos patch tokens e normalização L2; 4. persiste cada imagem e seu embedding euclidiano no HeraclitusDB; 5. cria um artefato local para avaliação offline; 6. treina uma sonda linear; 7. busca imagens visualmente semelhantes; 8. oferece busca por uma API FastAPI. - Python 3.11+; - HeraclitusDB ativo em 127.0.0.1:7474; - vários GB livres para o checkpoint; - RAM suficiente para um modelo com aproximadamente 1B parâmetros; - GPU é…

Open weights mit transformers

[View model](https://savrn.com/models/i-jepa-agro)

Model · Image feature extraction

### [IM-Animation-Motion-Encoder](https://savrn.com/models/im-animation-motion-encoder)

[Fynn](https://savrn.com/model-publishers/rbaerk)

IM-Animation: An Implicit Motion Representation for Identity-decoupled Character Animation This release contains the final locally retained TiTok-based motion encoder implementation and an exported checkpoint. It provides frame-level motion tokens; it does not include the full animation generator or retargeting network. Code and project videos are available in the linked GitHub repository. The encoder was verified with PyTorch 2.7.1 on CPU using the exported BF16 weights. The weight file is 607,017,208 bytes (approximately 579 MiB). Its SHA256 and provenance are in checkpointinfo.json. frames must be RGB, normalized to [0, 1], with shape [N, 3, 256, 256]. The training preprocessing pads…

Open weights apache-2.0

[View model](https://savrn.com/models/im-animation-motion-encoder)

Model · Image feature extraction

### [RK182X-VIT-siglip2-so400m-patch14-384](https://savrn.com/models/rk182x-vit-siglip2-so400m-patch14-384)

[RKNNAI](https://savrn.com/model-publishers/rknnai)

本仓库提供由 google/siglip2-so400m-patch14-384 转换的 RKNN 模型。 - Model ID：RKNNAI/RK182X-VIT-siglip2-so400m-patch14-384 - 模型显示名称：RK182X-VIT-siglip2-so400m-patch14-384 - 源模型：google/siglip2-so400m-patch14-384 - 模型类型：VIT ModelScope 完整下载： Hugging Face 完整下载： ModelScope 指定配置下载： Hugging Face 指定配置下载： - 使用配套 RKNN Runtime 和驱动；运行前用 rknn-smi -v 检查设备端版本。 - 源模型许可证：Apache License 2.0，正文见 LICENSE，归属与转换修改说明见 NOTICE。 - 同时遵守 RKNN Toolkit、RKNN Runtime 相关许可；本包不包含运行库。

Open weights apache-2.0

[View model](https://savrn.com/models/rk182x-vit-siglip2-so400m-patch14-384)

Model · Image feature extraction

### [RK182X-VIT-siglip-so400m-patch14-384](https://savrn.com/models/rk182x-vit-siglip-so400m-patch14-384)

[RKNNAI](https://savrn.com/model-publishers/rknnai)

本仓库提供由 google/siglip-so400m-patch14-384 转换的 RKNN 模型。 - Model ID：RKNNAI/RK182X-VIT-siglip-so400m-patch14-384 - 模型显示名称：RK182X-VIT-siglip-so400m-patch14-384 - 源模型：google/siglip-so400m-patch14-384 - 模型类型：VIT ModelScope 完整下载： Hugging Face 完整下载： ModelScope 指定配置下载： Hugging Face 指定配置下载： - 使用配套 RKNN Runtime 和驱动；运行前用 rknn-smi -v 检查设备端版本。 - 源模型许可证：Apache License 2.0，正文见 LICENSE，归属与转换修改说明见 NOTICE。 - 同时遵守 RKNN Toolkit、RKNN Runtime 相关许可；本包不包含运行库。

Open weights apache-2.0

[View model](https://savrn.com/models/rk182x-vit-siglip-so400m-patch14-384)

Model · Image feature extraction

### [RK182X-VIT-dinov3-vits16-pretrain-lvd1689m](https://savrn.com/models/rk182x-vit-dinov3-vits16-pretrain-lvd1689m)

[RKNNAI](https://savrn.com/model-publishers/rknnai)

本仓库提供由 facebook/dinov3-vits16-pretrain-lvd1689m 转换的 RKNN 模型。 - Model ID：RKNNAI/RK182X-VIT-dinov3-vits16-pretrain-lvd1689m - 模型显示名称：RK182X-VIT-dinov3-vits16-pretrain-lvd1689m - 源模型：facebook/dinov3-vits16-pretrain-lvd1689m - 模型类型：VIT ModelScope 完整下载： Hugging Face 完整下载： ModelScope 指定配置下载： Hugging Face 指定配置下载： - 使用配套 RKNN Runtime 和驱动；运行前用 rknn-smi -v 检查设备端版本。 - 源模型许可证：DINOv3 License，正文见 LICENSE，归属与转换修改说明见 NOTICE。 - 同时遵守 RKNN Toolkit、RKNN Runtime 相关许可；本包不包含运行库。

Open weights other

[View model](https://savrn.com/models/rk182x-vit-dinov3-vits16-pretrain-lvd1689m)

Model · Image feature extraction

### [RK182X-VIT-dinov3-vitl16-pretrain-lvd1689m](https://savrn.com/models/rk182x-vit-dinov3-vitl16-pretrain-lvd1689m)

[RKNNAI](https://savrn.com/model-publishers/rknnai)

本仓库提供由 facebook/dinov3-vitl16-pretrain-lvd1689m 转换的 RKNN 模型。 - Model ID：RKNNAI/RK182X-VIT-dinov3-vitl16-pretrain-lvd1689m - 模型显示名称：RK182X-VIT-dinov3-vitl16-pretrain-lvd1689m - 源模型：facebook/dinov3-vitl16-pretrain-lvd1689m - 模型类型：VIT ModelScope 完整下载： Hugging Face 完整下载： ModelScope 指定配置下载： Hugging Face 指定配置下载： - 使用配套 RKNN Runtime 和驱动；运行前用 rknn-smi -v 检查设备端版本。 - 源模型许可证：DINOv3 License，正文见 LICENSE，归属与转换修改说明见 NOTICE。 - 同时遵守 RKNN Toolkit、RKNN Runtime 相关许可；本包不包含运行库。

Open weights other

[View model](https://savrn.com/models/rk182x-vit-dinov3-vitl16-pretrain-lvd1689m)

## AbstractPhila

[All models and datasets](https://savrn.com/model-publishers/abstractphil)

## Versions

- [791d0e36cdf7](https://savrn.com/models/clip-vitb-mini-distilled/versions/791d0e36cdf7) · current 2026-10-09

## Explore More

- [All image feature extraction models](https://savrn.com/models/tasks/image-feature-extraction)
- [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0)
- [Model comparisons](https://savrn.com/models/comparisons)
- [The model directory](https://savrn.com/models)
- [Open model prices by host](https://savrn.com/ai-index/pricing/open-models)

## Source

- Repository metadata, read 2026-10-09.
- [Hugging Face record](https://huggingface.co/AbstractPhil/clip-vitb-mini-distilled)
- [How the hub is built](https://savrn.com/model-hub/methodology)