🧠

Model

Intent Student L6 Uniform Distilled

Name: Intent Student L6 Uniform Distilled
Author: gomyk

by gomyk hf-model--gomyk--intent-student-l6_uniform_distilled

Free2AITools Nexus Index

38.3 Top 100%

S: Semantic 50

A: Authority 0

P: Popularity 2

R: Recency 93

Q: Quality 65

Tech Context

0.03B Params

512 Ctx

Vital Performance

15 DL / 30D

0.0%

Source →

Audited 38.3 FNI Score

Tiny 0.03B Params

1k Context

15 Downloads

8G GPU ~2GB Est. VRAM

Dense BERTMODEL Architecture

Commercial APACHE License

Model Information Summary
Entity Passport
Registry ID	hf-model--gomyk--intent-student-l6_uniform_distilled
License	Apache-2.0
Provider	huggingface

💾

Compute Threshold

~1.3GB VRAM

Interactive

Analyze Hardware

Hardware Compatibility Test

▼

* Static estimation for 4-Bit Quantization.

📜

Cite this model

Academic & Research Attribution

BibTeX

@misc{hf_model__gomyk__intent_student_l6_uniform_distilled,
  author = {gomyk},
  title = {Intent Student L6 Uniform Distilled Model},
  year = {2026},
  howpublished = {\url{https://huggingface.co/gomyk/intent-student-L6_uniform_distilled}},
  note = {Accessed via Free2AITools Knowledge Fortress}
}

APA Style

gomyk. (2026). Intent Student L6 Uniform Distilled [Model]. Free2AITools. https://huggingface.co/gomyk/intent-student-L6_uniform_distilled

🔬Technical Deep Dive

Full Specifications [+]

Quick Commands

🦙 Ollama Run

ollama run intent-student-l6_uniform_distilled

🤗 HF Download

huggingface-cli download gomyk/intent-student-l6_uniform_distilled

📦 Install Lib

pip install -U transformers

⚖️ Free2AITools Nexus Index V2.0

Methodology Index Protocol

Semantic (S) 50

Authority (A) 0

Popularity (P) 2

Recency (R) 93

Quality (Q) 65

💬 Index Insight

FNI V2.0 for Intent Student L6 Uniform Distilled: Semantic (S:50), Authority (A:0), Popularity (P:2), Recency (R:93), Quality (Q:65).

Free2AITools Nexus Index

Verification Authority

HuggingFace API GitHub Metadata Arxiv Citation DB System Audit

Unbiased Data Node Refresh: VFS Live

---

🚀 What's Next?

📊

Find Training Datasets

Discover datasets compatible with this model

📈

Compare Benchmarks

See how this model ranks on standard tests

⚡

Technical Deep Dive

L6_uniform_distilled (Distilled)

Lightweight multilingual sentence encoder optimized for intent classification. Created from paraphrase-multilingual-MiniLM-L12-v2 via layer pruning + corpus-based vocabulary pruning + knowledge distillation.

Model Details

Property	Value
Teacher	paraphrase-multilingual-MiniLM-L12-v2
Architecture	XLM-RoBERTa (pruned)
Hidden dim	384
Layers	6 / 12
Layer indices	[0, 2, 4, 7, 9, 11]
Strategy	6 layers, evenly spaced (general-purpose)
Vocab size	~38,330 (pruned from 250K)
Parameters	26,184,576
Safetensors size	98.1MB
Distilled	Yes

Supported Languages (18)

ko, en, ja, zh, es, fr, de, pt, it, ru, ar, hi, th, vi, id, tr, nl, pl

Quick Start

python

from sentence_transformers import SentenceTransformer

model = SentenceTransformer("L6_uniform_distilled")

sentences = [
    "예약 좀 해줘",           # Korean
    "What did I order?",     # English
    "今日はいい天気ですね",    # Japanese
    "Reserva una mesa",      # Spanish
]

embeddings = model.encode(sentences)
print(embeddings.shape)  # (4, 384)

MTEB Evaluation Results

Overall Average: 56.3%

MassiveIntentClassification

Average: 52.83%

Language	Score
ar	42.91%
en	63.86%
es	56.57%
ko	47.97%

MassiveScenarioClassification

Average: 59.77%

Language	Score
ar	48.72%
en	71.38%
es	63.4%
ko	55.56%

Distillation Impact

Task	Before Distillation	After Distillation	Delta
MassiveIntentClassification	52.9%	52.83%	-0.07%p
MassiveScenarioClassification	58.2%	59.77%	+1.57%p

Training

This model was created in two stages:

Stage 1: Layer Pruning

Teacher model: paraphrase-multilingual-MiniLM-L12-v2 (12 layers, 384 hidden dim)
Selected layers: [0, 2, 4, 7, 9, 11] (6 layers, evenly spaced (general-purpose))
Vocabulary pruning: 250K -> ~38K tokens (corpus-based, 18 target languages)

Stage 2: Knowledge Distillation

Method: MSE + Cosine Similarity loss between teacher and student embeddings
Training data: MASSIVE dataset (90K multilingual sentences, 18 languages)
Optimizer: AdamW (lr=2e-5, weight_decay=0.01)
Schedule: Cosine annealing over 3 epochs
Batch size: 64
Base model: L6_uniform (layer-pruned only)

Compression Summary

Stage	Vocab	Layers	Size
Teacher (original)	250,002	12	~480MB
+ Layer pruning	250,002	6	~407MB
+ Vocab pruning	~38,330	6	~98MB

Limitations

Vocabulary pruning restricts the model to the 18 target languages
Designed for short dialogue utterances, not long documents
Layer pruning may reduce performance on complex semantic tasks

⚠️ Incomplete Data

Some information about this model is not available. Use with Caution - Verify details from the original source before relying on this data.

View Original Source →

📝 Limitations & Considerations

• Benchmark scores may vary based on evaluation methodology and hardware configuration.
• VRAM requirements are estimates; actual usage depends on quantization and batch size.
• FNI scores are relative rankings and may change as new models are added.
⚠ License Unknown: Verify licensing terms before commercial use.

Social Proof

HuggingFace Hub

15Downloads

Hub Discussions

🤗 Data Source: Hugging Face ↗

🔄 Daily sync (03:00 UTC)

AI Summary: Based on Hugging Face metadata. Not a recommendation.

📊 FNI Methodology 📚 Knowledge Baseℹ️ Verify with original source

🛡️ Model Transparency Report

Technical metadata sourced from upstream repositories.

Open Metadata

🆔 Identity & Source

id: hf-model--gomyk--intent-student-l6_uniform_distilled
slug: gomyk--intent-student-l6_uniform_distilled
source: huggingface
author: gomyk
license: Apache-2.0
tags: sentence-transformers, safetensors, bert, intent-classification, multilingual, layer-pruning, vocab-pruning, knowledge-distillation, sentence-similarity, ko, en, ja, zh, es, fr, de, pt, it, ru, ar, hi, th, vi, id, tr, nl, pl, license:apache-2.0, text-embeddings-inference, endpoints_compatible, region:us

⚙️ Technical Specs

architecture: BertModel
params billions: 0.03
context length: 512
pipeline tag: sentence-similarity
vram gb: 1.3
vram is estimated: true
vram formula: VRAM ≈ (params * 0.75) + 0.8GB (KV) + 0.5GB (OS)

📊 Engagement & Metrics

downloads: 15
stars: 0
forks: 0

Data indexed from public sources. Updated daily.