whisper-event (Whisper Fine-Tuning Event)

reach-vb

posted an update 18 days ago

Post

3192

VLMs are going through quite an open revolution AND on-device friendly sizes:

1. Google DeepMind w/ PaliGemma2 - 3B, 10B & 28B: google/paligemma-2-release-67500e1e1dbfdd4dee27ba48

2. OpenGVLabs w/ InternVL 2.5 - 1B, 2B, 4B, 8B, 26B, 38B & 78B: https://huggingface.co/collections/OpenGVLab/internvl-25-673e1019b66e2218f68d7c1c

3. Qwen w/ Qwen 2 VL - 2B, 7B & 72B: Qwen/qwen2-vl-66cee7455501d7126940800d

4. Microsoft w/ FlorenceVL - 3B & 8B: https://huggingface.co/jiuhai

5. Moondream2 w/ 0.5B: https://huggingface.co/vikhyatk/

What a time to be alive! 🔥

reach-vb

posted an update about 1 month ago

Post

3151

Massive week for Open AI/ ML:

Mistral Pixtral & Instruct Large - ~123B, 128K context, multilingual, json + function calling & open weights
mistralai/Pixtral-Large-Instruct-2411
mistralai/Mistral-Large-Instruct-2411

Allen AI Tülu 70B & 8B - competive with claude 3.5 haiku, beats all major open models like llama 3.1 70B, qwen 2.5 and nemotron
allenai/tulu-3-models-673b8e0dc3512e30e7dc54f5
allenai/tulu-3-datasets-673b8df14442393f7213f372

Llava o1 - vlm capable of spontaneous, systematic reasoning, similar to GPT-o1, 11B model outperforms gemini-1.5-pro, gpt-4o-mini, and llama-3.2-90B-vision
Xkev/Llama-3.2V-11B-cot

Black Forest Labs Flux.1 tools - four new state of the art model checkpoints & 2 adapters for fill, depth, canny & redux, open weights
reach-vb/black-forest-labs-flux1-6743847bde9997dd26609817

Jina AI Jina CLIP v2 - general purpose multilingual and multimodal (text & image) embedding model, 900M params, 512 x 512 resolution, matroyoshka representations (1024 to 64)
jinaai/jina-clip-v2

Apple AIM v2 & CoreML MobileCLIP - large scale vision encoders outperform CLIP and SigLIP. CoreML optimised MobileCLIP models
apple/aimv2-6720fe1558d94c7805f7688c
apple/coreml-mobileclip

A lot more got released like, OpenScholar ( OpenScholar/openscholar-v1-67376a89f6a80f448da411a6), smoltalk ( HuggingFaceTB/smoltalk), Hymba ( nvidia/hymba-673c35516c12c4b98b5e845f), Open ASR Leaderboard ( hf-audio/open_asr_leaderboard) and much more..

Can't wait for the next week! 🤗

reach-vb

posted an update about 1 month ago

Post

4326

What a brilliant week for Open Source AI!

Qwen 2.5 Coder by Alibaba - 0.5B / 1.5B / 3B / 7B / 14B/ 32B (Base + Instruct) Code generation LLMs, with 32B tackling giants like Gemnini 1.5 Pro, Claude Sonnet
Qwen/qwen25-coder-66eaa22e6f99801bf65b0c2f

LLM2CLIP from Microsoft - Leverage LLMs to train ultra-powerful CLIP models! Boosts performance over the previous SOTA by ~17%
microsoft/llm2clip-672323a266173cfa40b32d4c

Athene v2 Chat & Agent by NexusFlow - SoTA general LLM fine-tuned from Qwen 2.5 72B excels at Chat + Function Calling/ JSON/ Agents
Nexusflow/athene-v2-6735b85e505981a794fb02cc

Orca Agent Instruct by Microsoft - 1 million instruct pairs covering text editing, creative writing, coding, reading comprehension, etc - permissively licensed
microsoft/orca-agentinstruct-1M-v1

Ultravox by FixieAI - 70B/ 8B model approaching GPT4o level, pick any LLM, train an adapter with Whisper as Audio Encoder
reach-vb/ultravox-audio-language-model-release-67373b602af0a52b2a88ae71

JanusFlow 1.3 by DeepSeek - Next iteration of their Unified MultiModal LLM Janus with RectifiedFlow
deepseek-ai/JanusFlow-1.3B

Common Corpus by Pleais - 2,003,039,184,047 multilingual, commercially permissive and high quality tokens!
PleIAs/common_corpus

I'm sure I missed a lot, can't wait for the next week!

Put down in comments what I missed! 🤗

reach-vb

posted an update about 2 months ago

Post

1586

Smol TTS models are here! OuteTTS-0.1-350M - Zero shot voice cloning, built on LLaMa architecture, CC-BY license! 🔥

> Pure language modeling approach to TTS
> Zero-shot voice cloning
> LLaMa architecture w/ Audio tokens (WavTokenizer)
> BONUS: Works on-device w/ llama.cpp ⚡

Three-step approach to TTS:

> Audio tokenization using WavTokenizer (75 tok per second)
> CTC forced alignment for word-to-audio token mapping
> Structured prompt creation w/ transcription, duration, audio tokens

The model is extremely impressive for 350M parameters! Kudos to the
OuteAI team on such a brilliant feat - I'd love to see this be applied on larger data and smarter backbones like SmolLM 🤗

Check out the models here: OuteAI/outetts-6728aa71a53a076e4ba4817c

reach-vb

authored a paper about 2 months ago

Lina-Speech: Gated Linear Attention is a Fast and Parameter-Efficient Learner for text-to-speech synthesis

Paper • 2410.23320 • Published Oct 30 • 7

reach-vb

posted an update about 2 months ago

Post

2972

Smol models ftw! AMD released AMD OLMo 1B - beats OpenELM, tiny llama on MT Bench, Alpaca Eval - Apache 2.0 licensed 🔥

> Trained with 1.3 trillion (dolma 1.7) tokens on 16 nodes, each with 4 MI250 GPUs

> Three checkpoints:

- AMD OLMo 1B: Pre-trained model
- AMD OLMo 1B SFT: Supervised fine-tuned on Tulu V2, OpenHermes-2.5, WebInstructSub, and Code-Feedback datasets
- AMD OLMo 1B SFT DPO: Aligned with human preferences using Direct Preference Optimization (DPO) on UltraFeedback dataset

Key Insights:
> Pre-trained with less than half the tokens of OLMo-1B
> Post-training steps include two-phase SFT and DPO alignment
> Data for SFT:
- Phase 1: Tulu V2
- Phase 2: OpenHermes-2.5, WebInstructSub, and Code-Feedback

> Model checkpoints on the Hub & Integrated with Transformers ⚡️

Congratulations & kudos to AMD on a brilliant smol model release! 🤗

amd/amd-olmo-6723e7d04a49116d8ec95070

reach-vb

posted an update 2 months ago

Post

2448

What a great day for Open Science! @AIatMeta released models, datasets, and code for many of its research artefacts! 🔥

1. Meta Segment Anything Model 2.1: An updated checkpoint with improved results on visually similar objects, small objects and occlusion handling. A new developer suite will be added to make it easier for developers to build with SAM 2.

Model checkpoints: reach-vb/sam-21-6702d40defe7611a8bafa881

2. Layer Skip: Inference code and fine-tuned checkpoints demonstrating a new method for enhancing LLM performance.

Model checkpoints: facebook/layerskip-666b25c50c8ae90e1965727a

3. SALSA: New code enables researchers to benchmark AI-based attacks to validate security for post-quantum cryptography.

Repo: https://github.com/facebookresearch/LWE-benchmarking

4. Meta Lingua: A lightweight and self-contained codebase designed to train language models at scale.

Repo: https://github.com/facebookresearch/lingua

5. Meta Open Materials: New open source models and the largest dataset to accelerate AI-driven discovery of new inorganic materials.

Model checkpoints: fairchem/OMAT24

6. MEXMA: A new research paper and code for our novel pre-trained cross-lingual sentence encoder covering 80 languages.

Model checkpoint: facebook/MEXMA

7. Self-Taught Evaluator: a new method for generating synthetic preference data to train reward models without relying on human annotations.

Model checkpoint: facebook/Self-taught-evaluator-llama3.1-70B

8. Meta Spirit LM: An open-source language model for seamless speech and text integration.

Repo: https://github.com/facebookresearch/spiritlm

3 replies

·

reach-vb

posted an update 2 months ago

Post

5444

Multimodal Ichigo Llama 3.1 - Real Time Voice AI 🔥

> WhisperSpeech X Llama 3.1 8B
> Trained on 50K hours of speech (7 languages)
> Continually trained on 45hrs 10x A1000s
> MLS -> WhisperVQ tokens -> Llama 3.1
> Instruction tuned on 1.89M samples
> 70% speech, 20% transcription, 10% text
> Apache 2.0 licensed ⚡

Architecture:
> WhisperSpeech/ VQ for Semantic Tokens
> Llama 3.1 8B Instruct for Text backbone
> Early fusion (Chameleon)

I'm super bullish on HomeBrew/ Jan and early fusion, audio and text, multimodal models!

(P.S. Play with the demo on Hugging Face: jan-hq/Ichigo-llama3.1-s-instruct)

reach-vb

posted an update 3 months ago

Post

3089

NEW: Open Source Text/ Image to video model is out - MIT licensed - Rivals Gen-3, Pika & Kling 🔥

> Pyramid Flow: Training-efficient Autoregressive Video Generation method
> Utilizes Flow Matching
> Trains on open-source datasets
> Generates high-quality 10-second videos
> Video resolution: 768p
> Frame rate: 24 FPS
> Supports image-to-video generation

> Model checkpoints available on the hub 🤗: rain1011/pyramid-flow-sd3

reach-vb

posted an update 3 months ago

Post

2093

On-device AI framework ecosystem is blooming these days:

1. llama.cpp - All things Whisper, LLMs & VLMs - run across Metal, CUDA and other backends (AMD/ NPU etc)
https://github.com/ggerganov/llama.cpp

2. MLC - Deploy LLMs across platforms especially WebGPU (fastest WebGPU LLM implementation out there)
https://github.com/mlc-ai/web-llm

3. MLX - Arguably the fastest general purpose framework (Mac only) - Supports all major Image Generation (Flux, SDXL, etc), Transcription (Whisper), LLMs
https://github.com/ml-explore/mlx-examples

4. Candle - Cross-platform general purpose framework written in Rust - wide coverage across model categories
https://github.com/huggingface/candle

Honorable mentions:

1. Transformers.js - Javascript (WebGPU) implementation built on top of ONNXruntimeweb
https://github.com/xenova/transformers.js

2. Mistral rs - Rust implementation for LLMs & VLMs, built on top of Candle
https://github.com/EricLBuehler/mistral.rs

3. Ratchet - Cross platform, rust based WebGPU framework built for battle-tested deployments
https://github.com/huggingface/ratchet

4. Zml - Cross platform, Zig based ML framework
https://github.com/zml/zml

Looking forward to how the ecosystem would look 1 year from now - Quite bullish on the top 4 atm - but open source ecosystem changes quite a bit! 🤗

Also, which frameworks did I miss?

1 reply

·

reach-vb

posted an update 3 months ago

Post

2842

Less than two days ago Kyutai Labs open sourced Moshi - an ~7.6B on-device Speech to Speech foundation model and Mimi - SoTA streaming speech codec! 🔥

The release includes:

1. Moshiko & Moshika - Moshi finetuned on synthetic data (CC-BY license) ( kyutai/moshi-v01-release-66eaeaf3302bef6bd9ad7acd)
2. Mimi - Streaiming Audio Codec, processes 24 kHz audio, down to a 12.5 Hz representation with a bandwidth of 1.1 kbps (CC-BY license) ( kyutai/mimi)
3. Model checkpoints & Inference codebase written in Rust (Candle), PyTorch & MLX (Apache license) (https://github.com/kyutai-labs/moshi)

How does Moshi work?

1. Moshi processes two audio streams: one for itself and one for the user, with the user's stream coming from audio input and Moshi's stream generated by the model.

2. Along with these audio streams, Moshi predicts text tokens for its speech, enhancing its generation quality.

3. The model uses a small Depth Transformer for codebook dependencies and a large 7B parameter Temporal Transformer for temporal dependencies.

4. The theoretical latency is 160ms, with a practical latency of around 200ms on an L4 GPU.

Model size & inference:

Moshiko/ka are 7.69B param models

bf16 ~16GB VRAM
8-bit ~8GB VRAM
4-bit ~4GB VRAM

You can run inference via Candle 🦀, PyTorch and MLX - based on your hardware.

The Kyutai team, @adefossez @lmz and team are cracked AF, they're bringing some serious firepower to the open source/ science AI scene, looking forward to what's next! 🐐

1 reply

·

reach-vb

authored a paper 3 months ago

ESPnet-EZ: Python-only ESPnet for Easy Fine-tuning and Integration

Paper • 2409.09506 • Published Sep 14 • 3

reach-vb

posted an update 5 months ago

Post

3348

What an eventful day in Open Source LLMs today:

Mistral released Codestral Mamba 🐍
> Beats DeepSeek QwenCode, best model < 10B, competitive with Codestral 22B
> Mamba 2 architecture - supports up to 256K context
> Apache 2.0 licensed, perfect for local code assistant
> Transformers & llama.cpp integration upcoming!

Model checkpoint: https://huggingface.co/mistralai/mamba-codestral-7B-v0.1

Hugging Face dropped SmolLM 🤏
> Beats MobileLLM, Qwen 0.5B, Phi 1.5B and more!
> 135M, 360M, and 1.7B param model checkpoints
> Trained on 600B high-quality synthetic + FineWeb Edu tokens
> Architecture: Llama + GQA + 2048 ctx length
> Ripe for fine-tuning and on-device deployments.
> Works out of the box with Transformers!

Model checkpoints: HuggingFaceTB/smollm-6695016cad7167254ce15966

Mistral released Mathstral 7B ∑
> 56.6% on MATH and 63.47% on MMLU
> Same architecture as Mistral 7B
> Works out of the box with Transformers & llama.cpp
> Released under Apache 2.0 license

Model checkpoint: https://huggingface.co/mistralai/mathstral-7B-v0.1

Pretty dope day for open source ML. Can't wait to see what the community builds with it and to support them further! 🤗

What's your favourite from the release today?

1 reply

·

reach-vb

posted an update 6 months ago

Post

5264

Yet another rewarding week in Open Source AI:

1. Google dropped Gemma 27B & 9B - The best open (commercially permissive) LLM out there, according to LYMSYS.
google/gemma-2-release-667d6600fd5220e7b967f315

2. Mars5 TTS - Text to Speech with insane prosodies control & voice cloning.
CAMB-AI/MARS5-TTS

3. Meta shipped LLM Compiler - beats GPT 4 on code optimisation and compiler reasoning.
facebook/llm-compiler-667c5b05557fe99a9edd25cb

4. Arcee-Spark - Qwen2 7B (w/ merging) fine-tuned further to beat GPT 3.5 on MT Bench.
arcee-ai/Arcee-Spark

5. Gemini Nano out in the wild in Chrome - On device LLM with just 2 lines of code (fully offline)

6. Fal released a fully Open Source GAN based Super-Resolution model (with second version already cooking)
fal/AuraSR

7. NYU release Cambrian 1 - Vision Multimodal LLM that beats pretty much all other closed source competition 8-34B model size
https://huggingface.co/nyu-visionx

And.. much more like Open LLM Leaderboard got a major update, LYMSYS released Chat Vision Arena, OpenAI released a paper on CriticGPT!

What a lovely week, can’t wait for the next to see what the community is up to! Put it down in comments if I missed something 🔥

1 reply

·

osanseviero

posted an update 8 months ago

Post

9813

Diaries of Open Source. Part 15 🤗

🕵️‍♀️Idefics 2 is out, a multimodal open-source model with very nice capabilities
Models, demo, and datasets: HuggingFaceM4/idefics2-661d1971b7c50831dd3ce0fe
Blog: https://hf.co/blog/idefics2

💾Snowflake released snowflake-arctic-embed, a family of powerful small embedding models
Model: Snowflake/snowflake-arctic-embed-m
Blog: https://www.snowflake.com/blog/introducing-snowflake-arctic-embed-snowflakes-state-of-the-art-text-embedding-family-of-models/

✨Pile-T5, EleutherAI's T5 model trained on 2T tokens
Blog: https://blog.eleuther.ai/pile-t5/
Models: EleutherAI/pile-t5-65a76a0d0022dd270b385a66
GitHub: https://github.com/EleutherAI/improved-t5

🤖CodeQwen1.5-7B base and chat models. Models trained on 3T tokens strong benchmark results for code generation, editing and SQL
Blog post: https://qwenlm.github.io/blog/codeqwen1.5/
Demo: Qwen/CodeQwen1.5-7b-Chat-demo
Models: Qwen/CodeQwen1.5-7B and Qwen/CodeQwen1.5-7B-Chat

Misc
🦉 DocOwl1.5: Unified Stucture Learning for OCR-free Document Understanding mPLUG/DocOwl
👀Cerule - a tiny Vision LM model Tensoic/Cerule-v0.1
ChemLLM - a LLM for chemistry and molecule science ⚗️https://hf.co/AI4Chem/ChemLLM-7B-Chat-1.5-DPO
Distil Whisper Large
📝New pdf/OCR datasets with 19 samples pixparse/pdf-document-ocr-datasets-660701430b0346f97c4bc628
🔥Gretel AI high quality text-to-sql synthetic dataset gretelai/synthetic_text_to_sql

4 replies

·

osanseviero

posted an update 9 months ago

Post

9270

Diaries of Open Source. Part 14 🤗

🔥CohereForAI releases Command R+, an open 104B model with:
- Tool usage capabilities
- Specialized in RAGs
- Multilingual
It's one of the first models to surpass GPT-4 in the lmsys arena, check it out!
Model: CohereForAI/c4ai-command-r-plus
Official demo: https://hf.co/spaces/CohereForAI/c4ai-command-r-plus
Quantized: CohereForAI/c4ai-command-r-plus-4bit

🎉Google releases a new version of their Gemma instruct models, with improved quality, nicer to converse, and a fancier RL algorithm. The model is similar to Llama 2 70B in the Chat Arena!
Models: google/gemma-release-65d5efbccdbb8c4202ec078b
Try it out in HuggingChat https://hf.co/chat/models/google/gemma-1.1-7b-it

🪄VoiceCraft, a speech editing and TTS SOTA open model
Paper: VoiceCraft: Zero-Shot Speech Editing and Text-to-Speech in the Wild (2403.16973)
Model: pyp1/VoiceCraft

💻Google released CodeGemma, a family of code generation, completion, and chat models
Blog post: https://hf.co/blog/codegemma
Models: google/codegemma-release-66152ac7b683e2667abdee11
Report: https://storage.googleapis.com/deepmind-media/gemma/codegemma_report.pdf

Misc models:
🦖T-Rex2, a very powerful object detection model for many applications https://github.com/IDEA-Research/T-Rex
👀 CT-RATE : A 3D dataset paired with text reports ibrahimhamamci/CT-RATE
🐙Octopus v2: a Gemma-based model trained for Android API - extremely fast, better than Llama+RAG, great results NexaAIDev/Octopus-v2