The calculative miner

This is part of our series on a year of Bittensor experience, leading up to our anniversary at the 13th of July. We discuss 9 miner archetypes before digging deeper into incentive landscapes. After discussing the honest and the rogue miner, the calculative miner is considered. Please let us know Read more…

The rogue miner

This is part of our series on a year of Bittensor experience, leading up to our anniversary at the 13th of July. We discuss 9 miner archetypes before digging deeper into incentive landscapes. The previous article discussed the honest miner. Please let us know what you think in our Discord Read more…

The honest miner

With our first anniversary fast approaching on the 13th of July, it’s time to look back on our wild ride in Bittensor, in our role as miners, subnet owners, observers and contributors. We will be sharing our insights in a few articles, the first of which you are reading now, Read more…

On incentive (part 3)

In this post, we revisit model comparison metrics. Should we compare sample-by-sample, or group several samples together? Or does it make more sense to “pack” samples (i.e. join multiple samples with an EOS token in between)? Does length matter? And what’s up with these “pages” in the dataset? TL;DR A Read more…

Growing models

In this blog, and in our Discord channel, we discuss training in detail. A topic that is often overlooked, is how to grow a model. Especially in incentivized, collaborative and distributed training, this is a key ingredient. This post explores the concept of model growth with concrete Python code examples Read more…

On training (part 1)

One of the first objectives for subnet 29 was to see whether the sample packing during training makes a difference for the achievable loss. TL;DR Sample packing makes significant difference. Per-token loss of a trained model is generally higher when training on packed samples, and evaluating on single samples. Or, Read more…

On incentive (part 1)

Disclaimer: The text below is written for a wide audience. This means that the tech vocabulary is deliberately kept simple and, to some readers, the explanations may appear a bit long-winded. Week zero The biggest changes we implemented in our first week of existence are the modifications to the incentive Read more…