6 ms·

It's also worth mentioning that the original implementation by Meta is only 300 lines of very readable code [1]. [1]: https://github.com/meta-llama/llama3/blob

It's also worth mentioning that the original implementation by Meta is only 300 lines of very readable code [1]. [1]: https://github.com/meta-llama/llama3/blob/main/llama/model.py https://github.com/meta-llama/llama3/blob/main/llama/model.p...

blt 2y ago
the simplicity of the transformer is quite refreshing. especially in vision where the Vision Transformer with linear patch encodings replaces complex intertwined decisions about filter size, striding, pooling, #filters, depth, etc., with the simpler decision of how to allocate your FLOPS between dimensionality, #heads, and #layers.
blharr 2y ago
So is this the case that the information is in the data set? Or the code is very well defined to be so small? As an outsider it's surprising that such a capable model can be so "simple".
- jacobn 2y ago
  The training code is presumably quite a bit more complex than what they've open sourced, but part of the beauty of the GPT-based LLMs is their structural simplicity. Now, that simplicity can be deceiving - there are a lot of conceptual interconnectedness within these models. They've been put together "just so" if you will. If you look at the source code to nanoGPT and compare it to Llama3, the most remarkable thing (when you look past the superficial name changes) is just how similar they are. If I recall correctly the primary differences are: - The MLP: Llama3 uses SwiGLU vs the more "traditional" x = x + proj(gelu(expand(x))) in GPT2 - The token encoders, which is arguably external to the model - Attention: Llama3 uses Grouped Query Attention, vs full Multi-Head Attention in GPT2 - Normalization: Llama3 uses RMSNorm, vs LayerNorm for GPT2 They were published more than five years apart. On the one hand progress has been breathtaking, truly astounding. On the other hand, it's almost exactly the same model. Goes to show just how much is in the training data.
  - jacobn 2y ago
    Forgot one: the positional encoding also changed, llama3 uses RoPE, gpt2 uses a learned embedding.
  - novaRom 2y ago
    > beauty of the GPT-based LLMs is their structural simplicity human brain's structure is also encoded in a short DNA sequence
  - jacobn 2y ago
    > Goes to show just how much is in the training data. And in the scale (num_layers, embed_dim, num_heads) of the model of course ;)
- moritzwarhier 2y ago
  I think with LLMs in general, the algorithms are very refined and require lots of research, despite being "simple" in terms of entropy, or an imagined Kolgomorov complexity for defining algorithms. So "simple" is a fuzzy term here, but yes, the entropic complexity is in the data, not the algorithms. Related to the so-called "Bitter lesson". Edit: the sister comment pointed out what I failed to express: RILHF and training are also algorithms, and their applications and implementations are probably much more complex than the code that evaluates a given prompt. So basically, "models" (trained NNs) are also an example for the equivalence of code and data. Fixed data used by code (the trained model) is code in itself, even when it is not directly written by humans or in a human-readable language. Edit edit: don't forget to count the imported maths code :) but I assume this is not relevant to the "it's just matrix multiplications" overall argument
- SpaceManNabs 2y ago
  300 lines of this code is a bit different than 300 lines of typical code where you read files, set up a backend/frontend, or parse data. In the latter case, there are a lot of tedious operations. Sure, the former also has that with reshaping and asserts or wtv. But in a sense, the 300 lines of Llama code are essentially just lines of math. And reading through any math proof will show you that any particular line can hide large amounts of complexity. This can be true with code with more tedious operations, but those lines are a smaller fraction of the overall code base by definition. Even the "tedious" parts of the llama code can hide large complexity. Setting a learning rate with a schedule might require reading a paper or two for your particular architecture. But yes, once you parse all the math and the theory, the lines are kinda simple matmul and forward lol.
  - ffriend 2y ago
    Sure, knowing the basics of LLM math is necessary. But it's also _enough_ to know this math to fully grasp the code. There are only 4 concepts - attention, feed-forward net, RMS-normalization and rotary embeddings - organized into a clear structure. Now compare it to the Hugginface implementation [1]. In addition to the aforementioned concepts, you need to understand the hierarchy of `PreTrainedModel`s, 3 types of attention, 3 types of rotary embeddings, HF's definition of attention mask (which is not the same as mask you read about in transformer tutorials), several types of cache class, dozens of flags to control things like output format or serialization, etc. It's not that Meta's implementation is good and HF's implementation is bad - they pursue different goals in their own optimal way. But if you just want to learn how the model works, Meta's code base is great. [1]: https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py https://github.com/huggingface/transformers/blob/main/src/tr...
kureikain 2y ago
Do you know why these are so short? What is the algorithm/magic in all of these? I tried to make sense of it but cannot
- Hugsun 2y ago
  Architecturally, LLMs are very simple compared to many software projects. The crux of their behavior comes from their learned weights which are gigabytes and can cost millions to obtain via training.
- DavidSJ 2y ago
  The magic is in the billions of learned weights (~synapses). This is just the scaffolding that runs them.
- chpatrick 2y ago
  The magic is the structure of the model, and the real magic is the billions of weights.
ebb_earl_co 2y ago
On line 59, there is a less-than-or-equals comparison between 0 and 1. Curious https://github.com/meta-llama/llama3/blob/main/llama/model.py#L59 https://github.com/meta-llama/llama3/blob/main/llama/model.p...
- danielheath 2y ago
  What's the operator precedence in python? Is it `assert(0 <= (1 < ndim))` or `assert((0 <= 1) < ndim)`, or something even stranger like `assert(0 <= 1) < ndim`?
  - __s 2y ago
    Python actually does something pretty neat: it chains comparisons so that `x < y <= z` is like `x < y and y <= z` except y is only evaluated once In linked code we can be confident that `0 <= 1`, so only `1 < ndim` should matter. In fact I'd expect peephole optimization to remove most of the code for `0 <= 1`
    - 3abiton 2y ago
      I did not know that python fact. Thanks for sharing!
- bloaf 2y ago
  I am a reasonably competent python coder, yet when I see stuff like this I regard it with the same suspicion as a switch in the "more magic" position. https://www.catb.org/jargon/html/magic-story.html https://www.catb.org/jargon/html/magic-story.html
_pastel 2y ago
Why is max_seq_len set to 2048 [1] when the model card says the context size is 8k [2]? [1] https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a9596f1dea06d9c564f9c0e35f/llama/model.py#L32 https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a959... [2] https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a9596f1dea06d9c564f9c0e35f/MODEL_CARD.md https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a959...
- mkolodny 2y ago
  That's just the default. You can set max_seq_len to 8k. From the readme [0]: > All models support sequence length up to 8192 tokens, but we pre-allocate the cache according to max_seq_len and max_batch_size values. So set those according to your hardware. [0] https://github.com/meta-llama/llama3/tree/14aab0428d3ec3a9596f1dea06d9c564f9c0e35f?tab=readme-ov-file#inference https://github.com/meta-llama/llama3/tree/14aab0428d3ec3a959...
hongspike 2y ago
The numpy code can seem more accessible and easy to understand. Torch can look scary even though it's similar to numpy.

blt 2y ago

the simplicity of the transformer is quite refreshing. especially in vision where the Vision Transformer with linear patch encodings replaces complex intertwined decisions about filter size, striding, pooling, #filters, depth, etc., with the simpler decision of how to allocate your FLOPS between dimensionality, #heads, and #layers.

blharr 2y ago

So is this the case that the information is in the data set? Or the code is very well defined to be so small? As an outsider it's surprising that such a capable model can be so "simple".

jacobn 2y ago
The training code is presumably quite a bit more complex than what they've open sourced, but part of the beauty of the GPT-based LLMs is their structural simplicity. Now, that simplicity can be deceiving - there are a lot of conceptual interconnectedness within these models. They've been put together "just so" if you will. If you look at the source code to nanoGPT and compare it to Llama3, the most remarkable thing (when you look past the superficial name changes) is just how similar they are. If I recall correctly the primary differences are: - The MLP: Llama3 uses SwiGLU vs the more "traditional" x = x + proj(gelu(expand(x))) in GPT2 - The token encoders, which is arguably external to the model - Attention: Llama3 uses Grouped Query Attention, vs full Multi-Head Attention in GPT2 - Normalization: Llama3 uses RMSNorm, vs LayerNorm for GPT2 They were published more than five years apart. On the one hand progress has been breathtaking, truly astounding. On the other hand, it's almost exactly the same model. Goes to show just how much is in the training data.
- jacobn 2y ago
  Forgot one: the positional encoding also changed, llama3 uses RoPE, gpt2 uses a learned embedding.
- novaRom 2y ago
  > beauty of the GPT-based LLMs is their structural simplicity human brain's structure is also encoded in a short DNA sequence
- jacobn 2y ago
  > Goes to show just how much is in the training data. And in the scale (num_layers, embed_dim, num_heads) of the model of course ;)
moritzwarhier 2y ago
I think with LLMs in general, the algorithms are very refined and require lots of research, despite being "simple" in terms of entropy, or an imagined Kolgomorov complexity for defining algorithms. So "simple" is a fuzzy term here, but yes, the entropic complexity is in the data, not the algorithms. Related to the so-called "Bitter lesson". Edit: the sister comment pointed out what I failed to express: RILHF and training are also algorithms, and their applications and implementations are probably much more complex than the code that evaluates a given prompt. So basically, "models" (trained NNs) are also an example for the equivalence of code and data. Fixed data used by code (the trained model) is code in itself, even when it is not directly written by humans or in a human-readable language. Edit edit: don't forget to count the imported maths code :) but I assume this is not relevant to the "it's just matrix multiplications" overall argument
SpaceManNabs 2y ago
300 lines of this code is a bit different than 300 lines of typical code where you read files, set up a backend/frontend, or parse data. In the latter case, there are a lot of tedious operations. Sure, the former also has that with reshaping and asserts or wtv. But in a sense, the 300 lines of Llama code are essentially just lines of math. And reading through any math proof will show you that any particular line can hide large amounts of complexity. This can be true with code with more tedious operations, but those lines are a smaller fraction of the overall code base by definition. Even the "tedious" parts of the llama code can hide large complexity. Setting a learning rate with a schedule might require reading a paper or two for your particular architecture. But yes, once you parse all the math and the theory, the lines are kinda simple matmul and forward lol.
- ffriend 2y ago
  Sure, knowing the basics of LLM math is necessary. But it's also _enough_ to know this math to fully grasp the code. There are only 4 concepts - attention, feed-forward net, RMS-normalization and rotary embeddings - organized into a clear structure. Now compare it to the Hugginface implementation [1]. In addition to the aforementioned concepts, you need to understand the hierarchy of `PreTrainedModel`s, 3 types of attention, 3 types of rotary embeddings, HF's definition of attention mask (which is not the same as mask you read about in transformer tutorials), several types of cache class, dozens of flags to control things like output format or serialization, etc. It's not that Meta's implementation is good and HF's implementation is bad - they pursue different goals in their own optimal way. But if you just want to learn how the model works, Meta's code base is great. [1]: https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py https://github.com/huggingface/transformers/blob/main/src/tr...

kureikain 2y ago

Do you know why these are so short? What is the algorithm/magic in all of these? I tried to make sense of it but cannot

Hugsun 2y ago
Architecturally, LLMs are very simple compared to many software projects. The crux of their behavior comes from their learned weights which are gigabytes and can cost millions to obtain via training.
DavidSJ 2y ago
The magic is in the billions of learned weights (~synapses). This is just the scaffolding that runs them.
chpatrick 2y ago
The magic is the structure of the model, and the real magic is the billions of weights.

ebb_earl_co 2y ago

On line 59, there is a less-than-or-equals comparison between 0 and 1. Curious https://github.com/meta-llama/llama3/blob/main/llama/model.py#L59 https://github.com/meta-llama/llama3/blob/main/llama/model.p...

danielheath 2y ago
What's the operator precedence in python? Is it `assert(0 <= (1 < ndim))` or `assert((0 <= 1) < ndim)`, or something even stranger like `assert(0 <= 1) < ndim`?
- __s 2y ago
  Python actually does something pretty neat: it chains comparisons so that `x < y <= z` is like `x < y and y <= z` except y is only evaluated once In linked code we can be confident that `0 <= 1`, so only `1 < ndim` should matter. In fact I'd expect peephole optimization to remove most of the code for `0 <= 1`
  - 3abiton 2y ago
    I did not know that python fact. Thanks for sharing!
bloaf 2y ago
I am a reasonably competent python coder, yet when I see stuff like this I regard it with the same suspicion as a switch in the "more magic" position. https://www.catb.org/jargon/html/magic-story.html https://www.catb.org/jargon/html/magic-story.html

_pastel 2y ago

Why is max_seq_len set to 2048 [1] when the model card says the context size is 8k [2]? [1] https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a9596f1dea06d9c564f9c0e35f/llama/model.py#L32 https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a959... [2] https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a9596f1dea06d9c564f9c0e35f/MODEL_CARD.md https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a959...

mkolodny 2y ago
That's just the default. You can set max_seq_len to 8k. From the readme [0]: > All models support sequence length up to 8192 tokens, but we pre-allocate the cache according to max_seq_len and max_batch_size values. So set those according to your hardware. [0] https://github.com/meta-llama/llama3/tree/14aab0428d3ec3a9596f1dea06d9c564f9c0e35f?tab=readme-ov-file#inference https://github.com/meta-llama/llama3/tree/14aab0428d3ec3a959...

hongspike 2y ago

The numpy code can seem more accessible and easy to understand. Torch can look scary even though it's similar to numpy.