3 ms·

Not sure I understand the excerpt from the referenced paper. Is it saying that part of its more-efficient inferencing relies on mixing tokens from completely-s

by gojomo 3y ago

Not sure I understand the excerpt from the referenced paper. Is it saying that part of its more-efficient inferencing relies on mixing tokens from completely-separate inputs – eg, from other users? And then, depending on what other inputs chance into the same grouping, the relative assignment-to-'experts' varies, and thus the eventual completions? If so, I'd see that as not just introducing non-determinism, but also potentially making the quality of your responses dependent on how-many-concurrent-requests are fighting for the same expert-allocations. (For example, maybe the parts of the system best at translating/interpreting Hindi give worse results during peak usage hours-of-the-day in India, when the most concurrent inputs are competing for that same competence.) Perhaps also, this is another possible explanation for perceived quality-degradation over time. When certain tests were reliably succeeding earlier, there was less congestion for the relevant 'experts'. Now, with more concurrent use, those same tests aren't as reliably winning as much of relevant 'experts' effort. This may also suggest a bit of a quagmire: on whatever domains some sub-experts seem impressively good, initially, even more proportionate use will be attracted. But such new congestion means all the copycat use no longer gets the same expert allocations – and thus the initially-impressive performance degrades. (And if the effect is strong, & known-but-undisclosed-by-OpenAI, does it amount to a bait-and-switch? Attract users with unrepresentative excellence on an initially-uncongested Mixture-of-Experts system, but then offer them the lower-quality results from a more-congested system.)

albystein 3y ago
this seems like a plausible outcome, and if true could spell disaster for OpenAI models relative to the competition and open source models. Currently, reliability is one of the core obstacles preventing widespread adoption of LLMs in many business critical workflows. And if these rumors, that GPT-4 is inherently un-deterministic and unreliable, are true then most enterprises are better off finetuning open source LLMs—which are just as capable—for their specific domains. they stand to gain better performance that way anyways, as domain-specific models will always outperform generalist ones
- geysersam 3y ago
  > domain-specific models will always outperform generalist ones That's only true assuming you habe enough data to train a domain-specific model / expertise to train it and test it correctly. I've encountered cases where an image recognition task could be accomplished well with a very general model like CLIP, but people still fine-tuned another model on their own small data set because that's considered better. A domain specific model might be more likely to fail on weird outliers not present in the small domain specific training data. > could spell disaster for OpenAI Nah I don't think so. They are not all in on one specific model architecture. If the current architecture is found to have serious unfixable flaws then they'll just change architecture.
- famouswaffles 3y ago
  >as domain-specific models will always outperform generalist ones This is not even close to true for Language models.
- mrtranscendence 3y ago
  > And if these rumors, that GPT-4 is inherently un-deterministic and unreliable, are true then most enterprises are better off finetuning open source LLMs—which are just as capable Wait, am I misunderstanding you? I feel like I've had a head injury or something, because I've never heard of an open source LLM that's as capable as GPT-4 (in most scenarios).
  - visarga 3y ago
    Only on specific domains, these models don't become generalists like GPT-4, they can become task experts for a single task.
    - famouswaffles 3y ago
      Fine-tuned MedPalm is worse than GPT-4 on most Medical Challenge Tests. Fine-tuned Minerva is much worse on arithmetic benchmarks. The LLM space is just different. There's no guarantee a fine-tuned model will beat a bigger generalist one.
spott 3y ago
The results are showing essentially 12 unique responses from 30 tries… not what you would expect from mixing tokens. I think it groups the batch up differently, so if I have a batch of 10, and it groups it up into 2 groups of 5, if my prompt makes it to the second group or 1st group I get a different answer. But if I’m in the same location in the batch, then I get the same answer. The whole batch is deterministic given the same batch (sequences and ordering), but if you shuffle the batch then you lose that determinism.

by gojomo 3y ago

albystein 3y ago

this seems like a plausible outcome, and if true could spell disaster for OpenAI models relative to the competition and open source models. Currently, reliability is one of the core obstacles preventing widespread adoption of LLMs in many business critical workflows. And if these rumors, that GPT-4 is inherently un-deterministic and unreliable, are true then most enterprises are better off finetuning open source LLMs—which are just as capable—for their specific domains. they stand to gain better performance that way anyways, as domain-specific models will always outperform generalist ones

geysersam 3y ago
> domain-specific models will always outperform generalist ones That's only true assuming you habe enough data to train a domain-specific model / expertise to train it and test it correctly. I've encountered cases where an image recognition task could be accomplished well with a very general model like CLIP, but people still fine-tuned another model on their own small data set because that's considered better. A domain specific model might be more likely to fail on weird outliers not present in the small domain specific training data. > could spell disaster for OpenAI Nah I don't think so. They are not all in on one specific model architecture. If the current architecture is found to have serious unfixable flaws then they'll just change architecture.
famouswaffles 3y ago
>as domain-specific models will always outperform generalist ones This is not even close to true for Language models.
mrtranscendence 3y ago
> And if these rumors, that GPT-4 is inherently un-deterministic and unreliable, are true then most enterprises are better off finetuning open source LLMs—which are just as capable Wait, am I misunderstanding you? I feel like I've had a head injury or something, because I've never heard of an open source LLM that's as capable as GPT-4 (in most scenarios).
- visarga 3y ago
  Only on specific domains, these models don't become generalists like GPT-4, they can become task experts for a single task.
  - famouswaffles 3y ago
    Fine-tuned MedPalm is worse than GPT-4 on most Medical Challenge Tests. Fine-tuned Minerva is much worse on arithmetic benchmarks. The LLM space is just different. There's no guarantee a fine-tuned model will beat a bigger generalist one.

spott 3y ago

The results are showing essentially 12 unique responses from 30 tries… not what you would expect from mixing tokens. I think it groups the batch up differently, so if I have a batch of 10, and it groups it up into 2 groups of 5, if my prompt makes it to the second group or 1st group I get a different answer. But if I’m in the same location in the batch, then I get the same answer. The whole batch is deterministic given the same batch (sequences and ordering), but if you shuffle the batch then you lose that determinism.