This is a collection of all our flagship chat models to make it more easy for you!
🤝 Open to Collab
LH-Tech AI
LH-Tech-AI
AI & ML interests
Small AI and ML models. Trained by myself. Completely OpenSource. For you. | Reddit: https://www.reddit.com/user/LH-Tech_AI/
Recent Activity
reacted to Banaxi-Tech's post with 🤝 about 3 hours ago
We're delaying BananaMind 2.1!
When BananaMind 2.1 Lite was almost done, we benchmarked it and the results we're worse than BananaMind 2 Mini.
We're going to spend alot more time in research on tiny models and then scaling up our techniques to the actual BananaMind 2.1 models!
We're also announcing these new models:
BananaMind 2.1 Coder: A 149M instruction tuned coder model trained on 75B tokens + 10B tokens of stack-v3-train.
BananaMind 2.1 Pico: A 1M parameter model trained on 22B tokens of data.
We also may release BananaMind 2.1 Large with around 100M parameters depending on how much compute we have.
Please give us a follow!
https://huggingface.co/BananaMind
@Banaxi-Tech
---
@vovaRL
@DedeProGames
reacted to Banaxi-Tech's post with 🤗 about 3 hours ago
We're delaying BananaMind 2.1!
When BananaMind 2.1 Lite was almost done, we benchmarked it and the results we're worse than BananaMind 2 Mini.
We're going to spend alot more time in research on tiny models and then scaling up our techniques to the actual BananaMind 2.1 models!
We're also announcing these new models:
BananaMind 2.1 Coder: A 149M instruction tuned coder model trained on 75B tokens + 10B tokens of stack-v3-train.
BananaMind 2.1 Pico: A 1M parameter model trained on 22B tokens of data.
We also may release BananaMind 2.1 Large with around 100M parameters depending on how much compute we have.
Please give us a follow!
https://huggingface.co/BananaMind
@Banaxi-Tech
---
@vovaRL
@DedeProGames
reacted to Banaxi-Tech's post with 👀 about 3 hours ago
We're delaying BananaMind 2.1!
When BananaMind 2.1 Lite was almost done, we benchmarked it and the results we're worse than BananaMind 2 Mini.
We're going to spend alot more time in research on tiny models and then scaling up our techniques to the actual BananaMind 2.1 models!
We're also announcing these new models:
BananaMind 2.1 Coder: A 149M instruction tuned coder model trained on 75B tokens + 10B tokens of stack-v3-train.
BananaMind 2.1 Pico: A 1M parameter model trained on 22B tokens of data.
We also may release BananaMind 2.1 Large with around 100M parameters depending on how much compute we have.
Please give us a follow!
https://huggingface.co/BananaMind
@Banaxi-Tech
---
@vovaRL
@DedeProGames
Organizations
All models from the TinyMozart series.
Our top image classification models.
Our currently newest GAN models for generating 128x128px images of cats and human faces.
Our top text classification models.
Question, Reasoning and output datasets of Qwen3 1.7B. For finetuning smaller models. Have fun :D
A collection of all our TTS models.
Some very tiny BASE models to see how small models perform on e.g world knowledge
This is a collection of all our flagship CODING models to make it more easy for you!
Our top models that work with text, but don't generate text or code by themselves.
Some very tiny random chat models.
This is a collection of all our flagship chat models to make it more easy for you!
A collection of all our TTS models.
All models from the TinyMozart series.
Some very tiny BASE models to see how small models perform on e.g world knowledge
Our top image classification models.
This is a collection of all our flagship CODING models to make it more easy for you!
Our currently newest GAN models for generating 128x128px images of cats and human faces.
Our top models that work with text, but don't generate text or code by themselves.
Our top text classification models.
Some very tiny random chat models.
Question, Reasoning and output datasets of Qwen3 1.7B. For finetuning smaller models. Have fun :D