Hugging Face
Search models, datasets, users...
Ctrl K
Models
Datasets
Spaces
Buckets
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
Tardelli Stekel
PRO
tardellirs
22
20
88
Follow
gusnips's profile picture
a2un's profile picture
ssweber2's profile picture
166 followers
·
36 following
https://stekel.ifsp.dev/
tardellirs
tardelli-ronan-coelho-stekel-2802402a
AI & ML interests
None yet
Recent Activity
replied
to
their
post
about 2 hours ago
Model cards say what a model is: a fine-tune of X, a quantization of Y. I built Weightprint to check that against the weights. App: https://huggingface.co/spaces/tardellirs/weightprint Data (every pair, and the copy groups): https://huggingface.co/datasets/tardellirs/weightprint It reads the header and a few bytes of every tensor (no downloads) and compares a model with the parent its card declares. 120,976 models so far. Some of what came out: - 7% of the declared fine-tunes we could compare (1,630 of 22,847) hold their parent's weights unchanged. (Corrected from 12%: quants whose values the engine doesn't decode, like bitsandbytes and FP8, had been counted as unchanged. Thanks @dipankarsarkar for the question that caught it.) - 27% of the fine-tunes we could compare are LoRA-style: only the attention and MLP matrices moved, norms and embeddings untouched. That share went from 10% in 2023 to 28% in 2026. - Of 1,668 fine-tunes named abliterated, uncensored or heretic, 53% edit only the matrices that write to the residual stream, the signature of an abliteration. The rest were trained further, merged, or are the parent's weights. - 51% of the weight bytes on the Hub are byte-for-byte repeats of another repo's files. Search any Hub id to see which layers and parts changed against its parent, or compare any two models. A sample per tensor can miss things, and some formats can't be compared yet. If a verdict on a model you know looks wrong, tell me here and I'll check it against the engine. That's what this post is for, before a longer write-up.
new
activity
about 2 hours ago
tardellirs/model-pulse:
Well done
updated
a Space
about 6 hours ago
tardellirs/typed-decisions-explorer
View all activity
Organizations
tardellirs
's models
34
Sort: Recently updated
tardellirs/colibri-embed-ptbr-GGUF
Sentence Similarity
•
0.2B
•
Updated
2 days ago
•
200
tardellirs/colibri-embed-ptbr
Sentence Similarity
•
0.2B
•
Updated
2 days ago
•
7.13k
•
•
107
tardellirs/crivo-prumo-1b
1B
•
Updated
17 days ago
•
41
tardellirs/crivo-prumo-400m
0.4B
•
Updated
17 days ago
•
43
tardellirs/crivo-prumo-68m
68.4M
•
Updated
17 days ago
•
46
tardellirs/crivo-prumo-32m
32M
•
Updated
17 days ago
•
46
tardellirs/crivo-prumo-17m
16.9M
•
Updated
17 days ago
•
44
tardellirs/prumo
527
•
Updated
19 days ago
•
3
tardellirs/crivo-32m-per-option-ablation
Text Classification
•
Updated
19 days ago
tardellirs/crivo-1b
Text Classification
•
1B
•
Updated
19 days ago
•
149
•
1
tardellirs/crivo-400m
Text Classification
•
0.4B
•
Updated
19 days ago
•
183
tardellirs/crivo-150m
Text Classification
•
0.1B
•
Updated
19 days ago
•
530
•
14
tardellirs/crivo-68m
Text Classification
•
68.4M
•
Updated
19 days ago
•
183
•
4
tardellirs/crivo-32m
Text Classification
•
32M
•
Updated
19 days ago
•
253
•
2
tardellirs/crivo-17m
Text Classification
•
16.9M
•
Updated
19 days ago
•
145
tardellirs/aac-pictogram-retriever-ptbr
Sentence Similarity
•
0.2B
•
Updated
22 days ago
•
2.39k
tardellirs/aac-board-generator-770m-ptbr
Text Generation
•
0.8B
•
Updated
22 days ago
•
2.96k
tardellirs/aac-board-generator-770m-ptbr-GGUF
Text Generation
•
0.8B
•
Updated
22 days ago
•
2.38k
tardellirs/cadif-lever-llama3.1-8b-noinst
Updated
Jul 25
•
43
tardellirs/cadif-lever-llama3.1-8b-instq
Updated
Jul 25
•
37
tardellirs/cadif-lever-llama3.1-8b-full
Updated
Jul 25
•
54
tardellirs/gemma3moe-16x4-270m-starter
Text Generation
•
1B
•
Updated
Jul 22
•
2.33k
tardellirs/Maestro-Retriever-Llama-3.1-8B
Feature Extraction
•
Updated
Jul 12
•
13
•
1
tardellirs/embeddinggemma-pt-br-128k
Sentence Similarity
•
0.2B
•
Updated
Jul 7
•
69
•
1
tardellirs/embeddinggemma-pt-br-16k
Sentence Similarity
•
0.1B
•
Updated
Jul 7
•
61
•
1
tardellirs/embeddinggemma-pt-br
Sentence Similarity
•
0.2B
•
Updated
Jul 7
•
2.49k
•
1
tardellirs/embeddinggemma-pt-br-48k
Sentence Similarity
•
0.1B
•
Updated
Jul 7
•
55
•
1
tardellirs/embeddinggemma-pt-br-24k
Sentence Similarity
•
0.1B
•
Updated
Jul 7
•
67
•
1
tardellirs/embeddinggemma-pt-br-32k
Sentence Similarity
•
0.1B
•
Updated
Jul 7
•
53
•
2
tardellirs/brazembed-pt-br-clfclu
Sentence Similarity
•
0.1B
•
Updated
Jul 7
•
59
Previous
1
2
Next