Soham

Research Scientist @MistralAI on multimodal/audio LLMs.

Previously: @GoogleDeepMind
MS @CarnegieMellon

[Soham @sohamg121]

happiness/peace while working is directly proportional to number of agents going brrr, with the risk of blood pressure spikes due to absolutely dumbass shit they end up doing one in a while

[Soham @sohamg121]

@suchenzang yes but them parading moral superiority is exhausting

[Nathan Lambert @natolambert]

I don't really want to have to go to bat against Anthropic, but they've just been unnecessarily antagonistic to all of China, then not so subtly to open weight models, and now more broadly open AI research. What's next on the list?

[Soham @sohamg121]

@giffmana you still care about open models?

[Soham @sohamg121]

@_arohan_ what's wrong with asking tech bros to touch grass

[Soham @sohamg121]

@joannejang both

[Soham @sohamg121]

@alankarjain91 @natolambert There will always be a shift of workflows/products to open models as they start to catch up, no?

[Soham @sohamg121]

@tszzl @allTheYud @UnrealRealist19 I don't think an AI arms race would have started if OAI didn't make it clear that's how they will approach building this tech.

[Soham @sohamg121]

@sama well why is GPT voice mode so bad then (and continues to get worse)

[Soham @sohamg121]

@sophiamyang @sarahookr As does a big chunk of the Mistral AI Palo Alto office lol

[Alankar Jain @alankarjain91]

Two truths, no lies:

- Most AI models are insanely powerful.
- Most people use a tiny fraction of that power.

Closing that gap isn’t about better models, it’s about better products.

NextToken puts AI’s power in your hands, for your day-to-day work.

[Soham @sohamg121]

I'm so excited for my friends to be building something so cool from scratch!

[Nitish Kulkarni @nitishkulki]

Today, we are launching NextToken - a single place to build production-grade agents, apps and analytics.

Code-forward. Cloud Hosted. Zero setup. Extremely affordable.
@nexttoken_co

[Guillaume Lample @ NeurIPS 2024 @GuillaumeLample]

Our first speech model, Voxtral TTS, is out. It delivers SOTA performance while significantly reducing cost compared to existing solutions, and it operates with very low latency.

It uses a new architecture that combines auto-regressive generation of semantic speech tokens with flow-matching for acoustic tokens. We are also releasing a technical report sharing all our training methodology and insights.

Much more to come in audio -- stay tuned !

[Mistral AI @MistralAI]

🔊Introducing Voxtral TTS: our new frontier open-weight model for natural, expressive, and ultra-fast text-to-speech

🎭Realistic, emotionally expressive speech.
🌍Supports 9 languages and accurately captures diverse dialects.
⚡Very low latency for time-to-first-audio.
🔄Easily adaptable to new voices

[Soham @sohamg121]

@jachiam0 There is no way this happens through closed-source AI built by profit-chasing enterprise, and your CEO's offhand comments are a testimonial to that. I do believe in the potential promise of AI as you say, but one cannot ignore selfish human motivations.

[Soham @sohamg121]

@giffmana did it debug the right way, i.e. put a bunch of print statements and run code again and again?

[Mistral AI for Developers @MistralDevs]

Since launching Voxtral Realtime, the community response has been remarkable. Today, we share the technical report, launch the Realtime playground in Mistral Studio, and share the model in Hugging Face Transformers. 🧵

[Soham @sohamg121]

@asharoraa mental note to not use Whatsapp as my non-work todo tracker

[Guillaume Lample @ NeurIPS 2024 @GuillaumeLample]

🚀 We are hiring Research Interns !

We are looking for Master's and PhD students (final year) with a strong background in AI / ML / NLP who want to work on cutting-edge AI systems alongside Mistral AI researchers.

📍 Location: Paris, London, or Palo Alto
⏳ Duration: 4-6 months
🎓 For Master's students: opportunity to continue with a PhD at Mistral after the internship (via the CIFRE program).

Links below: