3 ms·

Gemma (and Gemini) are heavily nerfed. Why are they on the news lately? Also, Gemma is a +9B model. I think it's not okay that Google compared it with Mistral

Gemma (and Gemini) are heavily nerfed. Why are they on the news lately? Also, Gemma is a +9B model. I think it's not okay that Google compared it with Mistral and Llama 2 (7B) models. Google also took llama.cpp and used it in one of their Github repos without giving credit. Again, not cool. All this hype seems to be backed by Google to boost their models whereas in practice, the models are not that good. Google also made a big claim about Gemini 1.5 1M context window, but at the end of their article they said they'll limit it to 128K. So all that 1M flex was for nothing? Not to mention their absurd approach in alignment in image creation.

andy99 3y ago
Gemma has a 7B parameters model https://huggingface.co/google/gemma-7b https://huggingface.co/google/gemma-7b that's what I saw compared to Mistral (Edit: I'm wrong)
- light_hue_1 3y ago
  No it doesn't. Gemma 7B is a 9B model. The name is a lie. Then they really played games with Gemma 2B as well. I don't get how Google can be this incompetent and far behind everyone else. They have amazing people and the kinds of resources that almost no one else does but somehow need to resort to faking demos, blatant lies about model sizes, etc. Google used to be the place everyone wanted to go. Someone at Google AI needs to be fired so they can start being productive again.
  - andy99 3y ago
    Haha really? I almost added the caveat that I didn't count the parameters myself. And I couldn't see the weights file size because it requires login (because of their restrictive licensing choice). If true and it's 9B that's really dishonest.
    - rasbt 3y ago
      Yes, it's 8.5B params if you account for weight tying, and 9.3B if you count the embedding layer and output layer weights separately as shown in the 2nd figure in the article. In the paper, I think they justified 7B by only counting the non-embedding parameters (7,751,248,896), which is kind of cheating in my opinion, because if you do that, then Llama 2 is basically a 5B-6B param model.
      - andy99 3y ago
        Is the 2B measured like that as well? I did use it with llama.cpp and noticed it ran slower than I expected. That's the danger of too much abstraction, it's easy to have big gaps in one's understanding of what's really going on.
        rasbt 3y ago
        Yes, it's somewhat similar to the 2B model as it uses the same vocabulary size.
    - brucethemoose2 3y ago
      Practically, speaking, I OOM'd running Gemma on a 3090 using a config that had VRAM to spare for Mistral 7B. It kinda surprised me at first, until I realized why.
  - xfalcox 3y ago
    > Gemma 7B is a 9B model. The name is a lie Ohhh so that explains why I couldn't load it on my RTX 4090, while other 7B models load just fine!
gliptic 3y ago
> Google also took llama.cpp and used it in one of their Github repos without giving credit. Again, not cool. Are you talking about gemma.cpp? Then no, they didn't.
- andy99 3y ago
  Presumably he means this https://cloud.google.com/blog/products/application-development/new-localllm-lets-you-develop-gen-ai-apps-locally-without-gpus https://cloud.google.com/blog/products/application-developme... The claim is correct but not related to gemma
  - behnamoh 3y ago
    Yes, I was talking about this.
  - gliptic 3y ago
    From the repo: > Run LLMs locally on Cloud Workstations. Uses: > Quantized models from [Huggingface] > llama-cpp-python's webserver But sure, the blog post doesn't mention it.
rasbt 3y ago
> Gemma is a +9B model Yes that's correct. It's 9.3B parameters if you count the embedding layer and final projection layer separately. However, since they used weight tying, the adjusted count is 8.5B as discussed in the article.
- neodymiumphish 3y ago
  Which still rounds to 9B and is 21.4% larger.
  - rasbt 3y ago
    Yes, it's definitely unfair to count it as a 7B model. In that case, we could call Llama 2, which is 6.6B parameters, a 6B (or even 5B) parameter model.
    - neodymiumphish 3y ago
      Except 6.6 rounds to 7. That’s completely reasonable. Arguing otherwise is pedantic.
cyanydeez 3y ago
the context window is entirely limited by VRAM size do you even LLM?
brucethemoose2 3y ago
Counterpoints: - Local models are pretty easy to de-censor, if thats what you mean. - ...Yeah, it should not be labeled as a 7B. Its sort of 7B class. - The repo mentions they use the llama-cpp-python server - 1M context brute forced across TPUs is insanely expensive, I can see why Google reigned it in. But overall your message is not wrong. Google is hyping Gemma a ton when its... Well, not very remarkable. And they could have certainly made something niche and interesting, like a long context 8.5B model, a specialized model, a vastly more multilingual model, something to differentiate it from Mistral 7B 0.2
sivakon 3y ago
It’s objectively worse in my local tests compared to Mistral. Again their model doesn’t include MT-bench benchmark because it’s really really bad at answering a follow up question(s). (this is also a problem in Ultra). It’s reasoning is also pretty bad compared to mistral.
- b33j0r 3y ago
  I can’t get it to recognize the stop token consistently in the 7b models. About 50% of the shots, I get a sentence and a half of beautiful poetry, then a codeswitch into kanji, and then ral ral ral ral ral 膳 ral 杯 ral ral Until I kill the process. Not every time, but way more often than the other llamas (which is basically never, these days). I think they underestimated the impact of training on bulleted lists. It seems to love those!
verticalscaler 3y ago
[flagged]
htrp 3y ago
> Google also took llama.cpp and used it in one of their Github repos without giving credit. Again, not cool. They said it was inspired by llama >This is inspired by vertically-integrated model implementations such as ggml, llama.c, and llama.rs. from https://github.com/google/gemma.cpp https://github.com/google/gemma.cpp
- pests 3y ago
  Not gemma.cpp He meant this: https://cloud.google.com/blog/products/application-development/new-localllm-lets-you-develop-gen-ai-apps-locally-without-gpus https://cloud.google.com/blog/products/application-developme...
d-z-m 3y ago
> Also, Gemma is a +9B model. I think it's not okay that Google compared it with Mistral and Llama 2 (7B) models. They say it's because they're not counting embedding parameters[0]. Although apparently even with the embedding parameters subtracted it still rounds to 8B not 7B. From what understand, rounding to the nearest B is the standard. Seems slightly disingenuous to call it 7B, but not a big deal IMO since I don't hear anyone saying this model is outperforming popular OSS 7Bs. [0]: https://huggingface.co/google/gemma-7b/discussions/34 https://huggingface.co/google/gemma-7b/discussions/34
verticalscaler 3y ago
[flagged]

andy99 3y ago

Gemma has a 7B parameters model https://huggingface.co/google/gemma-7b https://huggingface.co/google/gemma-7b that's what I saw compared to Mistral (Edit: I'm wrong)

light_hue_1 3y ago
No it doesn't. Gemma 7B is a 9B model. The name is a lie. Then they really played games with Gemma 2B as well. I don't get how Google can be this incompetent and far behind everyone else. They have amazing people and the kinds of resources that almost no one else does but somehow need to resort to faking demos, blatant lies about model sizes, etc. Google used to be the place everyone wanted to go. Someone at Google AI needs to be fired so they can start being productive again.
- andy99 3y ago
  Haha really? I almost added the caveat that I didn't count the parameters myself. And I couldn't see the weights file size because it requires login (because of their restrictive licensing choice). If true and it's 9B that's really dishonest.
  - rasbt 3y ago
    Yes, it's 8.5B params if you account for weight tying, and 9.3B if you count the embedding layer and output layer weights separately as shown in the 2nd figure in the article. In the paper, I think they justified 7B by only counting the non-embedding parameters (7,751,248,896), which is kind of cheating in my opinion, because if you do that, then Llama 2 is basically a 5B-6B param model.
    - andy99 3y ago
      Is the 2B measured like that as well? I did use it with llama.cpp and noticed it ran slower than I expected. That's the danger of too much abstraction, it's easy to have big gaps in one's understanding of what's really going on.
      - rasbt 3y ago
        Yes, it's somewhat similar to the 2B model as it uses the same vocabulary size.
  - brucethemoose2 3y ago
    Practically, speaking, I OOM'd running Gemma on a 3090 using a config that had VRAM to spare for Mistral 7B. It kinda surprised me at first, until I realized why.
- xfalcox 3y ago
  > Gemma 7B is a 9B model. The name is a lie Ohhh so that explains why I couldn't load it on my RTX 4090, while other 7B models load just fine!

gliptic 3y ago

> Google also took llama.cpp and used it in one of their Github repos without giving credit. Again, not cool. Are you talking about gemma.cpp? Then no, they didn't.

andy99 3y ago
Presumably he means this https://cloud.google.com/blog/products/application-development/new-localllm-lets-you-develop-gen-ai-apps-locally-without-gpus https://cloud.google.com/blog/products/application-developme... The claim is correct but not related to gemma
- behnamoh 3y ago
  Yes, I was talking about this.
- gliptic 3y ago
  From the repo: > Run LLMs locally on Cloud Workstations. Uses: > Quantized models from [Huggingface] > llama-cpp-python's webserver But sure, the blog post doesn't mention it.

rasbt 3y ago

> Gemma is a +9B model Yes that's correct. It's 9.3B parameters if you count the embedding layer and final projection layer separately. However, since they used weight tying, the adjusted count is 8.5B as discussed in the article.

neodymiumphish 3y ago
Which still rounds to 9B and is 21.4% larger.
- rasbt 3y ago
  Yes, it's definitely unfair to count it as a 7B model. In that case, we could call Llama 2, which is 6.6B parameters, a 6B (or even 5B) parameter model.
  - neodymiumphish 3y ago
    Except 6.6 rounds to 7. That’s completely reasonable. Arguing otherwise is pedantic.

cyanydeez 3y ago

the context window is entirely limited by VRAM size do you even LLM?

brucethemoose2 3y ago

Counterpoints: - Local models are pretty easy to de-censor, if thats what you mean. - ...Yeah, it should not be labeled as a 7B. Its sort of 7B class. - The repo mentions they use the llama-cpp-python server - 1M context brute forced across TPUs is insanely expensive, I can see why Google reigned it in. But overall your message is not wrong. Google is hyping Gemma a ton when its... Well, not very remarkable. And they could have certainly made something niche and interesting, like a long context 8.5B model, a specialized model, a vastly more multilingual model, something to differentiate it from Mistral 7B 0.2

sivakon 3y ago

It’s objectively worse in my local tests compared to Mistral. Again their model doesn’t include MT-bench benchmark because it’s really really bad at answering a follow up question(s). (this is also a problem in Ultra). It’s reasoning is also pretty bad compared to mistral.

b33j0r 3y ago
I can’t get it to recognize the stop token consistently in the 7b models. About 50% of the shots, I get a sentence and a half of beautiful poetry, then a codeswitch into kanji, and then ral ral ral ral ral 膳 ral 杯 ral ral Until I kill the process. Not every time, but way more often than the other llamas (which is basically never, these days). I think they underestimated the impact of training on bulleted lists. It seems to love those!

verticalscaler 3y ago

[flagged]

htrp 3y ago

> Google also took llama.cpp and used it in one of their Github repos without giving credit. Again, not cool. They said it was inspired by llama >This is inspired by vertically-integrated model implementations such as ggml, llama.c, and llama.rs. from https://github.com/google/gemma.cpp https://github.com/google/gemma.cpp

pests 3y ago
Not gemma.cpp He meant this: https://cloud.google.com/blog/products/application-development/new-localllm-lets-you-develop-gen-ai-apps-locally-without-gpus https://cloud.google.com/blog/products/application-developme...

d-z-m 3y ago

> Also, Gemma is a +9B model. I think it's not okay that Google compared it with Mistral and Llama 2 (7B) models. They say it's because they're not counting embedding parameters[0]. Although apparently even with the embedding parameters subtracted it still rounds to 8B not 7B. From what understand, rounding to the nearest B is the standard. Seems slightly disingenuous to call it 7B, but not a big deal IMO since I don't hear anyone saying this model is outperforming popular OSS 7Bs. [0]: https://huggingface.co/google/gemma-7b/discussions/34 https://huggingface.co/google/gemma-7b/discussions/34

verticalscaler 3y ago

[flagged]