From 118eb9e955806d7dfbac2fb482e69b8e3edda328 Mon Sep 17 00:00:00 2001 From: Sami Jaghouar Date: Sat, 21 Sep 2024 00:57:00 +0000 Subject: [PATCH] add 7b --- configs/1B/H100.toml | 4 ++-- configs/7B/H100.toml | 12 ++++++++++++ 2 files changed, 14 insertions(+), 2 deletions(-) create mode 100644 configs/7B/H100.toml diff --git a/configs/1B/H100.toml b/configs/1B/H100.toml index cf8111b1..1430dcea 100644 --- a/configs/1B/H100.toml +++ b/configs/1B/H100.toml @@ -1,5 +1,5 @@ -name_model = "150M" -project = "debug_150m_zero_band" +name_model = "1B" +project = "debug_1B_zero_band" [train] micro_bs = 16 diff --git a/configs/7B/H100.toml b/configs/7B/H100.toml new file mode 100644 index 00000000..c1272c34 --- /dev/null +++ b/configs/7B/H100.toml @@ -0,0 +1,12 @@ +name_model = "7B" +project = "debug_7B_zero_band" + +[train] +micro_bs = 6 +sharding_strategy = "SHARD_GRAD_OP" + +[optim] +batch_size = 3840 +warmup_steps = 1000 +total_steps = 88_000 +lr = 6e-4 \ No newline at end of file