Samrat Man Singh
@samrat.me
📤 927
📥 171
📝 94
Software engineer; climber;
#ElixirLang
Website:
https://samrat.me
Building
https://octocut.com
Attending this later today. Anyone else I know going?
3 months ago
1
0
0
Great presentations and conversation at the Elixir Berlin Meetup today 1/ Lukas talked about his big bag of Postgres tricks and techniques
#elixirlang
5 months ago
2
3
0
reposted by
Samrat Man Singh
Paulo Valente
6 months ago
I've just released Handoff, a new Elixir library for distributed function graph execution.
hexdocs.pm/handoff
Among other things, the library enables distributed Nx computations (anyone said distributed inference?) A huge part of this was sponsored by
tvlabs.ai
!
loading . . .
Handoff v0.1.0 — Documentation
https://hexdocs.pm/handoff
2
58
16
reposted by
Samrat Man Singh
José Valim
7 months ago
Introducing Tidewave:
tidewave.ai
While working on our web apps, we run code, query the database, read logs, search docs… but our AI tools are limited to reading code. Tidewave closes this gap by introducing Runtime Intelligence, for Phoenix and Rails. More coming soon, check out the website!
loading . . .
Tidewave
https://tidewave.ai
5
174
62
Demo: Making phone calls from Elixir I'm working on a Twilio plugin for the Membrane multimedia framework. The demo shows an outbound call, but receiving calls works too.
#ElixirLang
www.youtube.com/watch?v=4dhl...
loading . . .
Making phone calls from Elixir using Twilio
YouTube video by Samrat Man Singh
https://www.youtube.com/watch?v=4dhl4P-kS9A
8 months ago
3
11
0
Testing Voice AI flows can become tedious fast But... you can just simulate conversations using another voice agent. Then, evaluate how the conversation went.
loading . . .
8 months ago
0
2
0
Got a small, standalone Python script working to use the new OpenAI transcription models(`gpt-4o-transcribe` and `gpt-4o-mini-transcribe`) using the Realtime WebSocket connection:
samrat.me/standalone-o...
loading . . .
Standalone OpenAI transcription Python script
OpenAI announced some new audio models yesterday, including new transcription models(gpt-4o-transcribe and gpt-4o-mini-transcribe). This model performs better than Whisper, though because they are bui...
https://samrat.me/standalone-openai-transcription-python-script/
8 months ago
0
0
1
OpenAI's new TTS looks(and sounds) pretty great for the price. Also, I hope this pushes other providers to just price API usage by minute. Every other TTS provider(ElevenLabs, Cartesia, etc) currently have monthly credits pricing.
8 months ago
1
1
0
Phoenix now comes with DaisyUI. Looks like a great addition for non-designers(like me)
#ElixirLang
add a skeleton here at some point
8 months ago
4
37
11
New
#ElixirLang
tool LivebookTools lets you work on Livebook in your own editor and opens MCP access to cell outputs(so Cursor, etc) can automatically iterate on the code if the cell has errors
github.com/thmsmlr/live...
loading . . .
GitHub - thmsmlr/livebook_tools: Powertools for livebook.dev — AI Code Editing, MCP Servers, and Running Livebooks from the CLI
Powertools for livebook.dev — AI Code Editing, MCP Servers, and Running Livebooks from the CLI - thmsmlr/livebook_tools
https://github.com/thmsmlr/livebook_tools
8 months ago
1
11
3
reposted by
Samrat Man Singh
Lars Wikman
8 months ago
What does talk-prep look like? Bug-fixin' mostly
@samrat.me
darling, you have a PR :D
github.com/samrat/kokor...
loading . . .
Fix concatenation issue for tensors by lawik · Pull Request #5 · samrat/kokoro
Nx.concatenate requires the same shape and very frequently a random chunk of text would not be the right shape. So it would blow up. The concatenate was avoidable so instead of a complex padding sc...
https://github.com/samrat/kokoro/pull/5
1
1
1
Pushed my first
@replicate.com
model, SmolVLM2 by
@hf.co
(
@pcuenq.hf.co
et al.) Will push the 500M and 2.2B models soon as well.
8 months ago
0
4
0
Voice models are getting scary good. This one feels very human:
www.sesame.com/research/cro...
loading . . .
Crossing the uncanny valley of conversational voice
At Sesame, our goal is to achieve “voice presence”—the magical quality that makes spoken interactions feel real, understood, and valued.
https://www.sesame.com/research/crossing_the_uncanny_valley_of_voice#demo
9 months ago
0
1
0
I wrote up a post on how you can run AI models on Elixir that aren't yet supported by Bumblebee
samrat.me/running-ml-m...
#MyElixirStatus
#ElixirLang
loading . . .
Running ML models in Elixir using Pythonx
I just discovered Pythonx, which runs a Python interpreter in the same OS process as Elixir. Looks like it also wraps uv so it sets up a virtual environment with all the Python dependencies your code...
https://samrat.me/running-ml-models-in-elixir-using-pythonx/
9 months ago
1
21
5
Wow, this looks super useful for running models that aren't yet supported in Bumblebee(among other things)
#ElixirLang
github.com/livebook-dev...
loading . . .
GitHub - livebook-dev/pythonx: Python interpreter embedded in Elixir
Python interpreter embedded in Elixir. Contribute to livebook-dev/pythonx development by creating an account on GitHub.
https://github.com/livebook-dev/pythonx
9 months ago
0
7
2
reposted by
Samrat Man Singh
David Bernheisel
9 months ago
Holy smokes this is awesome, a new debugger for Phoenix LiveView. Think React devtools, but not a chrome extension.
github.com/software-man...
#ElixirLang
loading . . .
5
152
46
reposted by
Samrat Man Singh
Xenova
9 months ago
We did it! Kokoro TTS (v1.0) can now run 100% locally in your browser w/ WebGPU acceleration. Real-time text-to-speech without a server. ⚡️ Generate 10 seconds of speech in ~1 second for $0. What will you build? 🔥
loading . . .
1
21
7
reposted by
Samrat Man Singh
Interested in how to build voice AI apps in Elixir? Here's a blog post I wrote about building a Membrane pipeline that orchestrates audio/video communication between a browser and Google Gemini
#ElixirLang
blog.swmansion.com/building-a-m...
loading . . .
Building a Membrane Pipeline to Talk to Google Gemini
Elixir, together with the Membrane framework, is great for building pipelines that enable seamless communication between users and LLMs.
https://blog.swmansion.com/building-a-membrane-pipeline-to-talk-to-google-gemini-ff637fc0fe40
9 months ago
1
7
1
Interested in how to build voice AI apps in Elixir? Here's a blog post I wrote about building a Membrane pipeline that orchestrates audio/video communication between a browser and Google Gemini
#ElixirLang
blog.swmansion.com/building-a-m...
loading . . .
Building a Membrane Pipeline to Talk to Google Gemini
Elixir, together with the Membrane framework, is great for building pipelines that enable seamless communication between users and LLMs.
https://blog.swmansion.com/building-a-membrane-pipeline-to-talk-to-google-gemini-ff637fc0fe40
9 months ago
1
7
1
reposted by
Samrat Man Singh
Lars Wikman
9 months ago
Had some fun generating speech using the Elixir Kokoro implementation by
@samrat.me
. Made a script for typing stuff and getting the talky-talk.
github.com/samrat/kokor...
So a bit of ONNX, some Erlang Ports, Membrane :D
#elixirlang
loading . . .
Mix.install-based example beyond raw audio · Issue #4 · samrat/kokoro
This will start prompting your terminal for input and speak it to your default audio output. I have this other example unrelated to TTS that includes working with files and MP3s. #!/usr/bin/env eli...
https://github.com/samrat/kokoro/issues/4
1
10
3
Reworking the Livebook for Kokoro(🔊 text-to-speech model). Now, you get a text input to input text you want read out loud(or whispered) to you.
#MyElixirStatus
#ElixirLang
loading . . .
10 months ago
0
9
4
Writing up a blog post about this. Not sure how much of it should be an introduction to Membrane.
#ElixirLang
add a skeleton here at some point
10 months ago
0
5
1
Elixir bindings for Kokoro TTS
github.com/samrat/kokoro
#ElixirLang
add a skeleton here at some point
10 months ago
1
9
1
Oops, forgot to tag
#ElixirLang
add a skeleton here at some point
10 months ago
0
5
0
More voice AI exploration: Bumblebee doesn't currently support text-to-speech models, but managed to get Kokoro TTS working with Ortex Demo 👇
youtu.be/VFKX6Af9gs4
loading . . .
Elixir text-to-speech with Kokoro
YouTube video by Samrat Man Singh
https://youtu.be/VFKX6Af9gs4
10 months ago
1
14
4
reposted by
Samrat Man Singh
antirez
10 months ago
Posted on HN. Feel like sharing here as well.
8
53
7
I've been building a small voice AI app. So far, both OpenAI Realtime and Gemini Multimodal Live seem very unreliable with function calling. And Gemini insists that it did the function call😅
10 months ago
0
0
0
Got Gemini Multimodal Live API working with Membrane WebRTC. Here's me having a short chat with it
#ElixirLang
youtu.be/JxWE3pxyGqA
loading . . .
Talking to Gemini Multimodal Live with Membrane WebRTC
YouTube video by Samrat Man Singh
https://youtu.be/JxWE3pxyGqA
11 months ago
0
9
2
Boring title, but this is a delightful read on China, travel and life
www.dwarkeshpatel.com/p/notes-on-c...
loading . . .
Notes on China
Last month, I spent 2 weeks in China - I visited Beijing, Chengdu, Emeishan, Chongqing, Shanghai, and Hangzhou.
https://www.dwarkeshpatel.com/p/notes-on-china?triedRedirect=true
11 months ago
0
0
0
Trying to add interruption support to
github.com/membranefram...
(the API already returns events for when it detects it got interrupted, but I'm having trouble stopping playback 🤔) Currently trying to wrap my head around Membrane's manual demand mode.
#ElixirLang
loading . . .
membrane_demo/livebooks/openai_realtime_with_membrane_webrtc/openai_realtime_with_membrane_webrtc.livemd at master · membraneframework/membrane_demo
Examples of using the Membrane Framework. Contribute to membraneframework/membrane_demo development by creating an account on GitHub.
https://github.com/membraneframework/membrane_demo/blob/master/livebooks/openai_realtime_with_membrane_webrtc/openai_realtime_with_membrane_webrtc.livemd
11 months ago
0
1
0
Current status: In Chulilla, Spain on a climbing trip.
11 months ago
0
3
0
reposted by
Samrat Man Singh
Paulo Valente
12 months ago
This also makes use of MLX's lazy evaluation to provide JIT compilation of Nx functions. The README says this, but just set it as your Nx backend and compiler, and it just works!
add a skeleton here at some point
0
7
3
The MLX(Apple Silicon) backend for Nx is now public on Github:
#ElixirLang
12 months ago
1
46
15
reposted by
Samrat Man Singh
Paulo Valente
12 months ago
The homepage example from Bumblebee on using BERT is now running on EMLX! We should hopefully be releasing the library in the next few days. We're just missing a really small set of Nx.Backend calbacks and some housekeeping stuff
3
20
10
reposted by
Samrat Man Singh
Looking for my next job! - Based in Berlin 🇩🇪, but also open to remote. - I've been most excited about
#ElixirLang
lately, but also open to other tech stacks(I've worked with Clojure, Rust, JS/TS) Some stuff I've worked on recently below👇. (And I'd appreciate a repost :))
12 months ago
1
28
18
reposted by
Samrat Man Singh
Paulo Valente
12 months ago
Ok, so I took this Axon example (
hexdocs.pm/axon/your_fi...
) for a spin on EMLX! A bit slower than Nx.BinaryBackend (runs the loop in 5s), but this is a simple model which really doesn't take advantage of a native platform. Next target is to get full Nx API coverage so we can run Bumblebee on it.
0
6
1
Finished reading "The MANIAC" by Benjamín Labatut. Wrote a short review:
samrat.me/book-review-...
#booksky
loading . . .
Book review: The MANIAC by Benjamín Labatut
Historical fiction about von Neumann and AI. This felt different than other historical novels I've read because as far as I could tell, there was no fictional plotlines introduced to the history. The ...
https://samrat.me/book-review-the-maniac-by-benjamin-labatut/
12 months ago
1
2
0
reposted by
Samrat Man Singh
Paulo Valente
12 months ago
That's funny... I could swear we started this endeavor 3 days ago, not 3 weeks! Jokes aside, big shout out to
@samrat.me
and
@uwucocoa.moe
who are taking MLX coverage by stride. We're almost at a point where we can run an Axon model! Bonus: this will be plug and play with Sharding
add a skeleton here at some point
1
12
4
Looking for my next job! - Based in Berlin 🇩🇪, but also open to remote. - I've been most excited about
#ElixirLang
lately, but also open to other tech stacks(I've worked with Clojure, Rust, JS/TS) Some stuff I've worked on recently below👇. (And I'd appreciate a repost :))
12 months ago
1
28
18
GPU go brrrr, with MLX and Nx
#ElixirLang
12 months ago
0
9
1
Any Zed users here? What do you use for AI code suggestions? The sidebar chat is really helpful, but I also would like the inline suggestions.
12 months ago
1
1
0
Implemented `Nx.sum` for the MLX backend. Took way longer than I'd expected 😅
#ElixirLang
12 months ago
2
6
0
Making some progress towards an MLX backend for Nx
#ElixirLang
add a skeleton here at some point
12 months ago
1
18
6
Does this exist: vector(embeddings) search interface for fonts? Google Fonts does have a "Feeling" filter, but when I'm looking at a font, I want to find similar fonts.
12 months ago
0
0
0
reposted by
Samrat Man Singh
Gordon Guthrie
12 months ago
Looking for work - 22 years
#Erlang
, 8 years
#ElixirLang
Done startups, big corps, got the dizzy heights of VP Eng with 350 staff, worked on small teams - wanna do cool shit Seeing as the BEAM community is blowing up here This is the stuff I know do read it and share it
3
57
29
Interesting. The Remotion devs have shared a system prompt that teaches LLMs about the library. Seems to work really well with Sonnet too-- different kind of "text-to-video".
www.remotion.dev/docs/system-...
loading . . .
Remotion System Prompt for LLMs | Remotion | Make videos programmatically
This is a prompt that you can give to Large Language Models to teach them the mechanics and rules of Remotion.
https://www.remotion.dev/docs/system-prompt
12 months ago
0
0
0
Adding multimodal support for the instructor_ex Gemini adapter. Gemini Flash downsamples to 1 FPS, but it's pretty impressive to see it understand video
#ElixirLang
about 1 year ago
0
2
0
Started reading this yesterday. It's a historical novel about John von Neumann's life. So far, it's very interesting.
about 1 year ago
0
1
0
reposted by
Samrat Man Singh
Chris, The 0.5x Engineer
about 1 year ago
What do me people use for tracing for Phoenix App?
#ElixirLang
0
0
1
I was looking for a FLAME-like API that worked with existing nodes in a cluster. ("FLAME, but without autoscaling") Here's my attempt at creating that using `pg` and DynamicSupervisors:
samrat.me/elixir-distr...
#ElixirLang
loading . . .
Implementing distributed pooling in Elixir
In Elixir, once you have clustering set up, it's ridiculously easy to run some code on another node in your cluster: Node.spawn(:"
[email protected]
", fn -> # This runs on a different ...
https://samrat.me/elixir-distributed-pooling/
about 1 year ago
0
21
4
Load more
feeds!
log in