anakin87

3 ms·

1.
by anakin87 6mo ago
Hi HN, I've been spending some time lately trying to build Reinforcement Learning Environments and training small language models and wanted to share a little course I put together based on my experiments. Over the past year, we'v
2.
Show HN: Hands-on course for building RL environments for LLMs(github.com)
1 points by anakin87 6mo ago|1 comments
3.
Environments Hub: Your Language Model needs better (open) environments to learn(huggingface.co)
2 points by anakin87 1y ago|1 comments
4.
by anakin87 1y ago
LLMs improve when they can practice and reason in interactive environments. Recent work (DeepSeek-R1, GRPO) shows RL can teach models to prefer better outputs by giving rewards. But most RL environments for LLMs are fragmented or closed. Th
5.
GRPO experiment - I trained a Language Model to schedule events(github.com)
1 points by anakin87 1y ago|1 comments
6.
by anakin87 1y ago
I experimented with GRPO lately, since I am fascinated by models learning from prompts and rewards - no example answers needed like in Supervised Fine-Tuning. After the DeepSeek boom, everyone is trying GRPO with GSM8K or the Countdown Game
7.
I trained a Language Model to schedule events with GRPO(huggingface.co)
1 points by anakin87 1y ago|1 comments
8.
by anakin87 1y ago
I experimented with GRPO lately, since I am fascinated by models learning from prompts and rewards - no example answers needed like in Supervised Fine-Tuning. After the DeepSeek boom, everyone is trying GRPO with GSM8K or the Countdown Game
9.
Llama2 + Haystack on Colab(github.com)
7 points by anakin87 3y ago|1 comments
10.
by anakin87 3y ago
I recently conducted some experiments with Llama2 and Haystack ( https://github.com/deepset-ai/haystack ), the NLP/LLM framework. The notebook can be helpful for those trying to load Llama2 on Colab. 1) Installed
11.
by anakin87 3y ago
In my experience, Argilla is a good open source platform for datacentric NLP. And these features are a great addition... Have you tried it?