AI-Hypercomputer / AI-Hypercomputer/maxtext

PGLE doesn't work for Tensor Parallelism

Đang mở
#1,005 3 bình luận 0 reaction 2 người được giao Được @wang2yn84 nhận Xem trên GitHub
Ngôn ngữ chính
Python
Star
2.4k
Fork
607
Merge trung bình
2 ngày 19 giờ
Pull request đã merge (30 ngày)
158

Mô tả

We observed good overlap with FSDP + PGLE:
![Bq7PCuqyJbygSuL](https://github.com/user-attachments/assets/0cff27c4-6499-43d0-b436-ef01a2833ae0). Turning on and off PGLE makes a big difference here.

However, with TP + PGLE:
![7nGeZQwG5Un84P3](https://github.com/user-attachments/assets/a2392577-170f-4d0c-a2bf-92554e35b933)

There is no performance improvements. Computation and communications are completely exposed.

Here is the command:
switch to lance-405b-clean branch

python3
MaxText/train.py MaxText/configs/models/gpu/llama3.1_405b.yml hardware=gpu
run_name=maxtext-llama3.1-405b steps=10 max_target_length=4096 model_name=llama3.1-405b
enable_checkpointing=false attention=cudnn_flash_te dataset_type=synthetic
async_checkpointing=false base_output_directory=gs://lancewang-dev-supercomputer-testing/maxtext_gpu
logits_dot_in_fp32=false use_iota_embed=true ici_tensor_parallelism=8 dcn_fsdp_parallelism=32
dcn_pipeline_parallelism=1 per_device_batch_size=1 num_layers_per_pipeline_stage=16 weight_dtype=bfloat16
remat_policy=save_qkv_proj profiler=xplane skip_first_n_steps_for_profiler=5
base_num_decoder_layers=126

Here are the xla flags:
--xla_gpu_graph_level=0 --xla_gpu_enable_latency_hiding_scheduler=true --xla_gpu_enable_triton_gemm=false
--xla_gpu_enable_highest_priority_async_stream=true --xla_gpu_all_reduce_combine_threshold_bytes=536870912
--xla_gpu_all_gather_combine_threshold_bytes=536870912 --xla_gpu_reduce_scatter_combine_threshold_bytes=536870912
--xla_gpu_enable_pipelined_all_gather=true --xla_gpu_enable_pipelined_reduce_scatter=true
--xla_gpu_enable_pipelined_all_reduce=true --xla_gpu_enable_while_loop_double_buffering=true
--xla_disable_hlo_passes=rematerialization --xla_gpu_enable_pgle_accuracy_checker=false
--xla_gpu_enable_triton_softmax_fusion=false --xla_gpu_enable_all_gather_combine_by_dim=false
--xla_gpu_enable_reduce_scatter_combine_by_dim=false

Here are the env variable:
NCCL_SHIMNET_GUEST_CONFIG_CHECKER_CONFIG_FILE=/usr/local/nvidia/lib64/a3plus_guest_config.textproto
NCCL_FASTRAK_PLUGIN_ACCEPT_TIMEOUT_MS=600000
JAX_ENABLE_PGLE=true
JAX_REMOVE_CUSTOM_PARTITIONING_PTR_FROM_CACHE_KEY=true
JAX_DEBUG_LOG_MODULES=compiler

The image we built on Oct 22nd.

Hướng dẫn đóng góp

Mở hướng dẫn đóng góp

Đánh giá

Issue này chưa được đánh giá.

Nhận issue mới trong hộp thư của bạn

Bản tóm tắt ngắn những issue GitHub phù hợp với người mới.