From 6a544fabfc2af22e4d5823410dd2387b5af89ea9 Mon Sep 17 00:00:00 2001 From: yurenh Date: Mon, 31 Aug 2026 18:14:09 -0500 Subject: scaffold: model (OLMo2-ish + ZBP partition), trainer (DDP/config), data shards, bench Co-Authored-By: Claude Fable 5 Claude-Session: https://claude.ai/code/session_01GkgLsACEF6CCP7EUfA5fZe --- configs/model/l350.yaml | 7 +++++++ configs/model/m124.yaml | 7 +++++++ configs/model/s60.yaml | 7 +++++++ 3 files changed, 21 insertions(+) create mode 100644 configs/model/l350.yaml create mode 100644 configs/model/m124.yaml create mode 100644 configs/model/s60.yaml (limited to 'configs/model') diff --git a/configs/model/l350.yaml b/configs/model/l350.yaml new file mode 100644 index 0000000..cd36f07 --- /dev/null +++ b/configs/model/l350.yaml @@ -0,0 +1,7 @@ +d: 1024 +layers: 24 +heads: 16 +vocab: 50257 +seq_len: 1024 +ffn_mult: 8/3 +pos: learned diff --git a/configs/model/m124.yaml b/configs/model/m124.yaml new file mode 100644 index 0000000..ad3b5fa --- /dev/null +++ b/configs/model/m124.yaml @@ -0,0 +1,7 @@ +d: 768 +layers: 12 +heads: 12 +vocab: 50257 +seq_len: 1024 +ffn_mult: 8/3 +pos: learned diff --git a/configs/model/s60.yaml b/configs/model/s60.yaml new file mode 100644 index 0000000..ebccc0a --- /dev/null +++ b/configs/model/s60.yaml @@ -0,0 +1,7 @@ +d: 512 +layers: 8 +heads: 8 +vocab: 50257 +seq_len: 1024 +ffn_mult: 8/3 # SwiGLU: hidden = mult*d rounded to 64 +pos: learned -- cgit v1.2.3