sparsegpt VS safetensors

SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot (https://arxiv.org/abs/2301.00774)

There is this : https://github.com/IST-DASLab/sparsegpt

Check the paper here, it's intersting: https://arxiv.org/abs/2301.00774

There's no chance that we've peeked from a bang for buck sense - we still haven't adequately investigated sparse networks.

Relevantish: https://arxiv.org/abs/2301.00774

The fact that we can reach those levels of sparseness with pruning also indicates that we're not doing a very good job of generating the initial network conditions.

Being able to come up with trainable initial settings for sparse networks across different topologies is hard, but given that we've had a degree of success with pre-trained networks, pre-training and pre-pruning might also allow for sparse networks with minimally compromised learning capabilities.

If it's possible to pre-train composable network modules, it might also be feasible to define trainable sparse networks with significantly relaxed topological constraints.

Training uses gradient descent, so you want to have good precision during that process. But once you have the overall structure of the network, https://arxiv.org/abs/2210.17323 (GPTQ) showed that you can cut down the precision quite a bit without losing a lot of accuracy. It seems you can cut down further for larger models. For the 13B Llama-based ones, going below 5 bit per parameter is noticeably worse, but for 30B models you can do 4 bits.

The same group did another paper https://arxiv.org/abs/2301.00774 which shows that in addition to reducing the precision of each parameter, you can also prune out a bunch of parameters entirely. It's harder to apply this optimization because models are usually loaded into RAM densely, but I hope someone figures out how to do it for popular models.

The ML field is doing work in that area: https://github.com/huggingface/safetensors

FYI the file format, safetensors, was proposed, developed and maintained by HF, and involved people from groups such as Eleuther and Stability for external security audits.

https://github.com/huggingface/safetensors https://huggingface.co/blog/safetensors-security-audit

Thank you for note on this. I had not heard there were already trojan horse malware being slipped into tensor files as python scripts. Apparently torch pickle uses eval on the tensor file with no filter.

Heard surprisingly little commentary on this topic. The full explanation of how Safetensors are "Safe" can be found from the developer at: https://github.com/huggingface/safetensors/discussions/111

Here the developer goes into some detail about what kinds of protections .safetensor files have : https://github.com/huggingface/safetensors/discussions/111

Use the safetensors format, which allows safe persistence and loading of models for common libraries - TensorFlow, PyTorch, JAX, etc. We went through external audits in the last few months (blog post). The current direction will be to have this as the default format.

I've been diving in lately and while it's not efficient, the only way to do manage is to create a new conda/mamba environment, or a custom Docker image for all the conflicting packages.

For safety and speed, you should prefer the safetensor format: https://huggingface.co/docs/safetensors/speed

If you know what you are doing you can do your own conversions: https://github.com/huggingface/safetensors or for safety, https://huggingface.co/spaces/diffusers/convert

GitHub - huggingface/safetensors: Simple, safe way to store and distribute tensors