Available now! Telegram Research 2025 — the year's key insights 
Machinelearning
Open in Telegram
Погружаемся в машинное обучение и Data Science Показываем как запускать любые LLm на пальцах. По всем вопросам - @haarrp @itchannels_telegram -🔥best channels Реестр РКН: clck.ru/3Fmqri
Show more📈 Analytical overview of Telegram channel Machinelearning
Channel Machinelearning (@ai_machinelearning_big_data) in the Russian language segment is an active participant. Currently, the community unites 293 687 subscribers, ranking 327 in the Technologies & Applications category and 1 276 in the Russia region.
📊 Audience metrics and dynamics
Since its creation on невідомо, the project has demonstrated rapid growth, gathering an audience of 293 687 subscribers.
According to the latest data from 01 July, 2026, the channel demonstrates stable activity. Although there has been a change in the number of participants by -6 444 over the last 30 days and by -235 over the last 24 hours, overall reach remains high.
- Verification status: Not verified
- Engagement rate (ER): The average audience engagement rate is 7.55%. Within the first 24 hours after publication, content typically collects 5.55% reactions from the total number of subscribers.
- Post reach: On average, each post receives 22 202 views. Within the first day, a publication typically gains 16 311 views.
- Reactions and interaction: The audience actively supports content: the average number of reactions per post is 172.
- Thematic interests: Content is focused on key topics such as openai, claude, api, gemini, контекст.
📝 Description and content policy
The author describes the resource as a platform for expressing subjective opinions:
“Погружаемся в машинное обучение и Data Science
Показываем как запускать любые LLm на пальцах.
По всем вопросам - @haarrp
@itchannels_telegram -🔥best channels
Реестр РКН: clck.ru/3Fmqri”
Thanks to the high frequency of updates (latest data received on 02 July, 2026), the channel maintains relevance and a high level of publication reach. Analytics show that the audience actively interacts with content, making it an important point of influence in the Technologies & Applications category.
293 687
Subscribers
-23524 hours
-1 5517 days
-6 44430 days
Posts Archive
293 687
✅ AliceMind: ALIbaba's Collection of Encoder-decoders from MinD (Machine IntelligeNce of Damo) Lab
Github: https://github.com/alibaba/AliceMind
Paper: https://arxiv.org/abs/2109.05687v1
Dataset: https://paperswithcode.com/dataset/glue
@ArtificialIntelligencedl
293 687
🔊 Textless NLP: Generating expressive speech from raw audio
Facebook Ai: https://ai.facebook.com/blog/textless-nlp-generating-expressive-speech-from-raw-audio/
Examples: https://speechbot.github.io/pgslm/
Code: https://github.com/pytorch/fairseq/tree/master/examples/textless_nlp/gslm
Paper: https://arxiv.org/abs/2102.01192
@ai_machinelearning_big_data
293 687
🧍♂ Texformer: 3D Human Texture Estimation from a Single Image with Transformers
Github: https://github.com/xuxy09/texformer
Paper: http://arxiv.org/abs/2109.02563
Meta data: https://www.dropbox.com/s/ekxn300cuw8bw6b/meta.zip
@ai_machinelearning_big_data
293 687
🔝 Reinforcement Learning Lecture Series 2021
https://deepmind.com/learning-resources/reinforcement-learning-series-2021
Video: youtube.com/playlist?list=PLqYmG7hTraZDVH599EItlEWsUOsJbAodm
@ai_machinelearning_big_data
293 687
☄️ nnFormer: Interleaved Transformer for Volumetric Segmentation
Github: https://github.com/ziniuwan/maed
Paper: https://arxiv.org/abs/2109.03201v1
Dataset: https://www.creatis.insa-lyon.fr/Challenge/acdc/
@ai_machinelearning_big_data
293 687
📹 CTRL-C: Camera calibration TRansformer with Line-Classification
Github: https://github.com/jwlee-vcl/ctrl-c
Paper: https://arxiv.org/abs/2109.02259v1
Dataset: https://drive.google.com/file/d/1opDR_j-9y4TY82ZscBb-xYAoMkTocAOp/view
@ai_machinelearning_big_data$$$
293 687
👨🎓 Best Resources to Learn Natural Language Processing in 2021
https://www.kdnuggets.com/2021/09/best-resources-learn-natural-language-processing-2021.html
@ai_machinelearning_big_data
293 687
📁 Open-sourcing Common Objects in 3D, a large-scale data set for 3D reconstruction
Github: https://github.com/facebookresearch/co3d
Paper: https://arxiv.org/abs/2109.00512
Facebook blog: https://github.com/openvinotoolkit/training_extensions/tree/develop/misc
Dataset: https://ai.facebook.com/datasets/co3d-downloads/
@ai_machinelearning_big_data
293 687
ℹ️ LIGAR: Lightweight General-purpose Action Recognition
Github: https://github.com/openvinotoolkit/training_extensions
Paper: https://arxiv.org/abs/2108.13153v1
Models: https://github.com/openvinotoolkit/training_extensions/tree/develop/misc
@ai_machinelearning_big_data
293 687
🔎 Efficient Person Search: An Anchor-Free Approach.
Github: https://github.com/daodaofr/AlignPS
Paper: https://arxiv.org/abs/2109.00211v1
Dataset: https://paperswithcode.com/dataset/cuhk-sysu
@ai_machinelearning_big_data
293 687
📶 ISNet: Integrate Image-Level and Semantic-Level Context for Semantic Segmentation
Github: https://github.com/segmentationblwx/sssegmentation
Paper: https://arxiv.org/abs/2108.12382v1
Dataset: https://cs.stanford.edu/~roozbeh/pascal-context/
@ai_machinelearning_big_data
293 687
ShapeConv: Shape-aware Convolutional Layer for Indoor RGB-D Semantic Segmentation
Github: https://github.com/hanchaoleng/shapeconv
Paper: https://arxiv.org/abs/2108.10528v1
@ai_machinelearning_big_data
293 687
💬 Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation
Github: https://github.com/ofirpress/attention_with_linear_biases
Paper: https://ofir.io/train_short_test_long.pdf
Fairseq: https://github.com/pytorch/fairseq
@ai_machinelearning_big_data
293 687
🧍 Fast and Flexible Human Pose Estimation with HyperPose
Github: https://github.com/tensorlayer/hyperpose
Paper: https://arxiv.org/abs/2108.11826v1
Docs: https://hyperpose.readthedocs.io/en/latest/markdown/quick_start/prediction.html#table-of-flags-for-hyperpose-cli
@ai_machinelearning_big_data
293 687
💡Volunteer computing: статья про коллаборативное обучение нейросети
Если вы хотите создавать большие модели, не имея суперкомпьютер за спиной, то вам понадобится технология, способная разделить вычисления между теми, кто готов предоставить вам мощности. В этом поможет технология DeDLOC, разработанная в Yandex Research, Hugging Face и University of Toronto.
Хабр: https://habr.com/ru/company/yandex/blog/574466/
@ai_machinelearning_big_data
293 687
🕸 Bag of Tricks for Training Deeper Graph Neural Networks: A Comprehensive Benchmark Study
Github: https://github.com/VITA-Group/Deep_GCN_Benchmarking
Paper: https://arxiv.org/abs/2108.10521v1
@ai_machinelearning_big_data
293 687
🔗 A Unified Objective for Novel Class Discovery
Github: https://github.com/DonkeyShot21/UNO
Paper: https://arxiv.org/abs/2108.08536v2
Dataset: https://paperswithcode.com/dataset/cifar-100
@ai_machinelearning_big_data
293 687
Data-driven организация: когда алгоритмы заменят людей?
В Data-Driven подходе все принимаемые решения основаны на big data. Что это — просто мода на использование больших данных или действительно эффективный инструмент? А может, всего лишь хайповое название привычных вещей?
Этой теме будет посвящен «Цифровой четверг» Х5.
Дата: 30 сентября 10:00, онлайн
Открытый разговор с экспертами рынка (X5 Group, Ростелеком, Яндекс.Музыка, Кинопоиск, НИУ ВШЭ, Segmento) о том, как и зачем становиться Data-Driven компанией, почему иногда кейсы работы с данными не приносят реальной пользы бизнесу и могут ли алгоритмы заменить людей?
Участие бесплатное, но необходима регистрация. Подробнее про мероприятие тоже по ссылке.
293 687
🎩 Mr. TyDi: A Multi-lingual Benchmark for Dense Retrieval
Mr. TyDi is a multi-lingual benchmark dataset built on TyDi, covering eleven typologically diverse languages.
Github: https://github.com/castorini/mr.tydi
Paper: https://arxiv.org/abs/2108.08787
Tasks: https://paperswithcode.com/task/representation-learning
@ai_machinelearning_big_data
