Nothing works with Beam V2

[rmadabusiml]

Hello,

Beam later version is V2 and they did drastic changes to their SDK and client that makes most of the training (fine-tuning) and inference code useless. There is no "beam run" and so on...

They pointed to this fine tuning code: https://docs.beam.cloud/v2/examples/gemma-fine-tune

[iusztinpaul]

Hello,

Unfortunately, this is what happens when working with fast-paced start-ups.

The good news is that we created the LLM Twin course, which is a more comprehensive version of the Hands-on LLMs course, where everything still works, and the learning material is more detailed:

https://github.com/decodingml/llm-twin-course