Eric Hartford(@erhartford) 's Twitter Profileg
Eric Hartford

@erhartford

Principal Applied AI Researcher @TensorWaveCloud
I make AI models Dolphin and Samantha
https://t.co/3ri2GbXrQB
BTC 3ENBV6zdwyqieAXzZP2i3EjeZtVwEmAuo4

ID:2854214132

linkhttps://erichartford.com calendar_today13-10-2014 10:22:51

3,6K Tweets

11,6K Followers

396 Following

Avy Faingezicht(@avyfain) 's Twitter Profile Photo

Yesterday's at ollama's meetup Eric Hartford gave a short talk about Dolphin/uncensored models.

This is important work. Trust/safety easily become euphemisms for censorship/control. Coercion can be a positive force only if we agree on what's good/evil. We don't. There's no 'we'

account_circle
Augustinas Malinauskas(@amgauge) 's Twitter Profile Photo

ollama Thanks for sharing your thoughts [email protected]. I chose Apache license - it allows commercial use, modifications, distribution, private use etc.

account_circle
Fireworks AI(@FireworksAI_HQ) 's Twitter Profile Photo

Fireworks now offers the first (to our knowledge) hosted, instruct variant of Mixtral 8x22B! Try it at fireworks.ai/models/firewor…… or download it at huggingface.co/fireworks-ai/m……

Thanks Teknium (e/λ) and Nous Research for the great dataset!

Fireworks now offers the first (to our knowledge) hosted, instruct variant of Mixtral 8x22B! Try it at fireworks.ai/models/firewor…… or download it at huggingface.co/fireworks-ai/m…… Thanks @Teknium1 and @NousResearch for the great dataset!
account_circle
Eric Hartford(@erhartford) 's Twitter Profile Photo

Oom-ing in the night Fading dreams keep me awake Do I have what it takes? Will these barriers break? Oh GPU Please don't fade away I need you by my side Don't leave me in this disarray

suno.com/song/b9dd8ecf-…

account_circle
Weyaxi(@Weyaxi) 's Twitter Profile Photo

🎉 Introducing EulerMath-Mistral-7B, based on the meta-math-mistral model, fully fine-tuned using MathInstruct and Orca math word problems dataset.

🔢 This model excels in math and uses Alpaca prompt format like the base model itself :)

Check it out: huggingface.co/Weyaxi/EulerMa…

🎉 Introducing EulerMath-Mistral-7B, based on the meta-math-mistral model, fully fine-tuned using MathInstruct and Orca math word problems dataset. 🔢 This model excels in math and uses Alpaca prompt format like the base model itself :) Check it out: huggingface.co/Weyaxi/EulerMa…
account_circle
Bindu Reddy(@bindureddy) 's Twitter Profile Photo

Apparently the new Mistral model beats Claude Sonnet and is a tad bit worse than GPT-4

In a couple of months, the open source community will fine tune it to beat GPT-4

This is a fully open weights model with an Apache 2 license! I can’t believe how quickly the OSS community…

account_circle
ifioravanti(@ivanfioravanti) 's Twitter Profile Photo

Fine tune a model on Apple Silicon with Apple MLX, convert it to gguf and use Ollama for inferencing everywhere was number 1 in my wish list for 2024! DONE
Next: GPT-4 level open source LLM Eric Hartford Teknium (e/λ) Maxime Labonne Weyaxi Argilla Wing Lian (caseus) We count on you! 💪

account_circle
Jeremy Howard(@jeremyphoward) 's Twitter Profile Photo

I was about to ask how come every AI thing came out all at once today, and then I remembered that happens every day now.

So never mind.

account_circle