This project is the JAX implementation of Llama 2.
- hyunwoongko/transformer: PyTorch implementation of the original Transformer
- ayaka14732/bart-base-jax: JAX implementation of BART-base
- ztjhz/t5-jax: JAX implementation of T5
- young-geng/EasyLM: LLM framework that includes Flax implementations of LLaMA, GPT-J and RoBERTa
This project is supported by Cloud TPUs from Google's TPU Research Cloud (TRC).
The objectives of this project are threefold:
- Implement the Llama 2 model using JAX to enable efficient training and inference on Google Cloud TPU;
- Develop a high-quality codebase that serves as an exemplary implementation of the Transformer model using JAX;
- Facilitate the identification of common errors and inconsistencies across various transformer models through the implementation of a high-quality codebase, thereby providing valuable insights for the NLP community.
- Parameter conversion
- Model architecture
- Cross entropy loss
- Logits processing
- Generation
- Beam search
- Beam sampling
- Top-k sampling
- Top-p sampling
- Optimisation
- Data loading
- Inference
- Training
- Parallelisation
- Documentation (published on GitHub Pages)
This project requires at least Python 3.11, JAX 0.4.14, PyTorch 2.1.0 and Transformers 4.32.0.dev0.
PyTorch and Transformers are needed for testing purposes. Additionally, the data loader depends on PyTorch DataLoader, while the profiling functionality requires TensorFlow.
For Ubuntu users, you can follow How to install Python 3.11 on Ubuntu 22.04 to Install Python 3.11. The tutorial applied to Ubuntu 20.04 as well.
python3.11 -m venv venv
. venv/bin/activate
pip install -U pip
pip install -U wheelYou need to follow the installation instructions on JAX's offical GitHub page.
TPU:
pip install "jax[tpu]" -f https://storage.googleapis.com/jax-releases/libtpu_releases.htmlCUDA 12:
pip install "jax[cuda12_pip]" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.htmlCUDA 11.8:
pip install "jax[cuda11_pip]" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.htmlTypically, you only need to install the CPU version of PyTorch since we perform most of the computation using JAX. However, it's worth noting that the current codebase's generation process is not fully optimised yet. To expedite the inference, one effective approach would involve converting the model back to Hugging Face format and running the inference in PyTorch.
To install PyTorch, you can follow the official installation guide.
CPU:
pip install --pre torch --index-url https://download.pytorch.org/whl/nightly/cpuCUDA 12:
pip install --pre torch --index-url https://download.pytorch.org/whl/nightly/cu121CUDA 11.8:
pip install --pre torch --index-url https://download.pytorch.org/whl/nightly/cu118pip install git+https://github.com/huggingface/transformers.git
pip install -r requirements.txtLLaMA 1:
If you couldn't obtain the LLaMA weights, you can download them with shawwn/llama-dl.
mkdir ../llama-weights-original && cd ../llama-weights-original
curl -o- https://raw.githubusercontent.com/shawwn/llama-dl/56f50b96072f42fb2520b1ad5a1d6ef30351f23c/llama.sh | bash
python ../llama-2-jax/venv/lib/python3.11/site-packages/transformers/models/llama/convert_llama_weights_to_hf.py --input_dir ../llama-weights-original --model_size 7B --output_dir ../llama-weights/7BLlama 2:
You can request to access the Llama weights from the official website. After your request is approved, you will automatically get access to the Hugging Face Llama 2 models. You can verify that the models are accessible by trying to access the Llama 2 7B version.
If you need to convert Llama 2 models, you need to first log in using huggingface-cli login.
python scripts/convert_params_runner.py llama1-7B
python scripts/convert_params_runner.py llama2-7B
python scripts/convert_params_runner.py llama2-70BIf you are running on TPU pods or other multi-host environments, you need to put the IP address of other machines in external-ips.txt (one IP address per line). Besides, you should make sure that one of the hosts can SSH into other hosts.
python generate.pyOn TPU pods, the command is:
./startpod python generate.pyI present a simple example of the training pipeline by fine-tuning the model on the GSM dataset.
cd .. && git clone --depth=1 https://github.com/openai/grade-school-math.gitpython train.pyOn TPU pods, the command is:
./startpod python train.py| Name | Parameters | vocab_size |
n_layers |
n_heads_kv |
n_rep_kv |
d_model |
d_ff |
|---|---|---|---|---|---|---|---|
| LLaMA 1 7B | 6738415616 | 32000 | 32 | 32 | 1 | 4096 | 11008 |
| LLaMA 1 13B | 32000 | 40 | 40 | 1 | 5120 | ||
| LLaMA 1 33B | 32000 | 60 | 52 | 1 | 6656 | ||
| LLaMA 1 65B | 32000 | 80 | 64 | 1 | 8192 | ||
| Llama 2 7B | 6738415616 | 32000 | 32 | 32 | 1 | 4096 | 11008 |
| Llama 2 13B | 32000 | ||||||
| Llama 2 70B | 32000 | 80 | 8 | 8 | 8192 | 28672 |
n_parameters
= 2 * vocab_size * d_model
+ (2 * n_layers + 1) * d_model
+ 2 * n_layers * d_model * n_rep_kv * n_heads_kv * d_k
+ 2 * n_layers * d_model * n_heads_kv * d_k
+ 3 * n_layers * d_model * d_ff
Hugging Face format:
LlamaForCausalLM(
(model): LlamaModel(
(embed_tokens): Embedding(32000, 4096, padding_idx=0)
(layers): ModuleList(
(0-31): 32 x LlamaDecoderLayer(
(self_attn): LlamaAttention(
(q_proj): Linear(in_features=4096, out_features=4096, bias=False)
(k_proj): Linear(in_features=4096, out_features=4096, bias=False)
(v_proj): Linear(in_features=4096, out_features=4096, bias=False)
(o_proj): Linear(in_features=4096, out_features=4096, bias=False)
(rotary_emb): LlamaRotaryEmbedding()
)
(mlp): LlamaMLP(
(gate_proj): Linear(in_features=4096, out_features=11008, bias=False)
(down_proj): Linear(in_features=11008, out_features=4096, bias=False)
(up_proj): Linear(in_features=4096, out_features=11008, bias=False)
(act_fn): SiLUActivation()
)
(input_layernorm): LlamaRMSNorm()
(post_attention_layernorm): LlamaRMSNorm()
)
)
(norm): LlamaRMSNorm()
)
(lm_head): Linear(in_features=4096, out_features=32000, bias=False)
)
The format used in this project:
model
embedding: (32000, 4096)
decoder: decoder_block
input_norm: (32, 4096)
attention
q_proj: (32, 4096, 1, 32, 128)
k_proj: (32, 4096, 32, 128)
v_proj: (32, 4096, 32, 128)
out_proj: (32, 1, 32, 128, 4096)
post_attn_norm: (32, 4096)
gate_proj: (32, 4096, 11008)
up_proj: (32, 4096, 11008)
down_proj: (32, 11008, 4096)
norm: (4096)
lm_head: (4096, 32000)
Hugging Face format:
LlamaForCausalLM(
(model): LlamaModel(
(embed_tokens): Embedding(32000, 8192, padding_idx=0)
(layers): ModuleList(
(0-79): 80 x LlamaDecoderLayer(
(self_attn): LlamaAttention(
(q_proj): Linear(in_features=8192, out_features=8192, bias=False)
(k_proj): Linear(in_features=8192, out_features=1024, bias=False)
(v_proj): Linear(in_features=8192, out_features=1024, bias=False)
(o_proj): Linear(in_features=8192, out_features=8192, bias=False)
(rotary_emb): LlamaRotaryEmbedding()
)
(mlp): LlamaMLP(
(gate_proj): Linear(in_features=8192, out_features=28672, bias=False)
(up_proj): Linear(in_features=8192, out_features=28672, bias=False)
(down_proj): Linear(in_features=28672, out_features=8192, bias=False)
(act_fn): SiLUActivation()
)
(input_layernorm): LlamaRMSNorm()
(post_attention_layernorm): LlamaRMSNorm()
)
)
(norm): LlamaRMSNorm()
)
(lm_head): Linear(in_features=8192, out_features=32000, bias=False)
)
The format used in this project:
model
embedding: (32000, 8192)
decoder: decoder_block
input_norm: (80, 8192)
attention
q_proj: (80, 8192, 8, 8, 128)
k_proj: (80, 8192, 8, 128)
v_proj: (80, 8192, 8, 128)
out_proj: (80, 8, 8, 128, 8192)
post_attn_norm: (80, 8192)
gate_proj: (80, 8192, 28672)
up_proj: (80, 8192, 28672)
down_proj: (80, 28672, 8192)
norm: (8192)
lm_head: (8192, 32000)
- LLaMA utilises rotary positional embeddings.
- There is no bias in the Q, K, V matrices and the linear projections in the FFNs, which is the same as the original transformer, but different from BERT and BART.
- In Llama models, each FFN has 3 linear projections, while in BART there are only 2.
- There is no dropout in the original LLaMA implementation.
- Llama 2 70B utilises Grouped-Query Attention (GQA).